Korea Rural Economic Instituterepository.krei.re.kr/bitstream/2018.oak/22368/1/농촌... · 2019-01-16 · www. krei.re.kr Vol. 40 | Special Issue. Vol. 40 | Special Issue. Forest

www. krei.re.kr

Vol. 40 | Special Issue.


Forest Carbon Sequestration and Optimal Harvesting Decision Considering Spb Disturbance: A Real Options Approach

An Hyun-jin

An Economic Effect of the Crop Insurance at the Farmland in KoreaPark Ji-yun, Kim Chang-gil

The Production and Dissemination of Agricultural Knowledge at U.S. Research Universities: The Role and Mission of Land-Grant Universities

Lee Yoo Hwan, Gregory D. Graff

Impact of Increased Imports of Agricultural Products due to FTAs on Domestic Price Decline

Jeong Min-kook, Moon Han-pil, Song Woo-jin

Trends in South Korea’s Grants-Based Aid for Agricultural Sector in Developing Countries

Lee Hye-jin

601, Bitgaram-ro, Naju-si, Jeollanam-do, 58217, KoreaTel: 82-61-820-2000 Fax: 82-61-820-2211

Journal of Rural Developm

ent

Korea Rural Economic Institute



Vol. 37 No. 3 | 2014

ARTICLES

An Analysis of Consumers’ Preferences for Quality-Graded Beef Products Kim Sung-yong, Jeon Sang-gon, Lee Kyei-im

Impacts of Food Consumption Lifestyle on the Expenditure for the Processed Food: Using Cluster Analysis and Matching Method

Mi-Sung Park, Byeong-Il Ahn

Development and Measurement of the Index of Agrifood Consumer Competency Index Lee Kyei-im, Ban Hyun-jung, Park Ki-whan, Hwang Yun-jae

A Study on the Characteristics of Women Returning to Rural Areas and Their Settlement From the Gender Perspective

Jinyang Myong-suk

Lifelong Learning Participations of Rural People and Their Related Variables Ma Sang-jin, Kim Kang-ho

Factor Analysis on Need for Children in a Rural Area with the Lowest Fertility Rate: Married Women in Cheongdo-gun Kim Na-young

A Study on the Evaluations on Korea’s Official Development Assistance in the Forest Sector and the Way to Improve its Development Effectiveness

Byoung Il Yoo, Bo Eun Yoon

Unit Cost Estimation of Forest Offset Program Reflecting Greenhouse Gas Emissions from Deforestation. Jae Soo Bae, Yeongmo Son, Jong-Su Yim

Assessing Korean Consumers’ Valuation for BSE-Tested and Country of Origin Labeled Beef Products Lee Sang-hyeon, Lee Jy-yong, Han Doo-bong, Nayga Jr. Rodolfo

ISSN 1229-8263

서울특별시 동대문구 회기로 117-3 우)130-710

Tel:02-3299-4000 Fax:02-965-6950

www.facebook.com/jrd2011

제37권 제

3호

제37권 제3호

논문

쇠고기 등급별 소비자 선호도 분석_김성용, 전상곤, 이계임

식품소비 라이프스타일이 가공식품 지출에 미치는 효과 분석: 군집분석과 매칭기법을 이용하여_박미성, 안병일

농식품 소비역량지수 개발과 계측_이계임, 반현정, 박기환, 황윤재

젠더 관점에서 본 귀농·귀촌 여성의 정착 과정과 그 특성_진양명숙

농촌주민의 평생학습 참여 결정 요인 분석_마상진, 김강호

초저출산 농촌지역의 자녀필요성 결정 요인 분석: 청도군 기혼여성 사례_김나영

산림 분야 공적개발원조 사업평가와 성과제고 방안_유병일, 윤보은

온실가스 배출량을 반영한 대체산림자원조성비의 단가 추정 _배재수, 손영모, 임종수

Assessing Korean Consumers’ Valuation for BSE-Tested and Country of Origin Labeled Beef Products_이상현, 이지용, 한두봉, Nayga Jr. Rodolfo

제40권 특별호

KOREA RURAL ECONOMIC INSTITUTE

The Korea Rural Economic Institute (KREI) is a non-profit research institute, established in April, 1978. It is an autonomous research body in the field of agricultural economics including food, natural resources, environment and rural development of Korea. The KREI conducts long-term as well as short-term policy-oriented studies with a view to assisting government, farmers, cooperatives and agribusiness firms.

Any article or other material appearing in the Journal may not be republished in full or in part without the written permission of the editor.

Journal of Rural Development

PUBLlSHER

인쇄일 2017. 12. 20. 발행일 2017. 12. 21.

발행인 김 창 길

발행처 한국농촌경제연구원(대표전화 1833-5500)

전라남도 나주시 빛가람로 601

인터넷 홈페이지 http://www.krei.re.kr

정기간행물 등록 전남, 마00026(1978. 5. 30.)

인쇄소 ㈜프리비(061-332-1492)

Journal of Rural DevelopmentVol. 40, Special Issue. [통권 제 171호]

Kim, Chang-Gil, President

ⓒ 2017 Korea Rural Economic Institute

JOURNAL OF RURAL DEVELOPMENTEDITOR:Lee, Kye-im, Research Director, KREIB. Wade Brorsen, Oklahoma State University, U.S.A

EDITORIAL COMMITTEE:Ahn, Byeong-il, Professor, Korea University, KoreaChung, Won-ho, Professor, Pusan National University, KoreaJang, Jae-bong, Professor, Konkuk University, KoreaJeon, Sang-gon, Professor, Gyeongsang National University, KoreaJi, In-bae, Senior Research Fellow, KREIJun, Ik-su, Professor, Chungbuk National University, KoreaKim, Jeong-seop, Senior Research Fellow, KREIKim, Jong-jin, Senior Research Fellow, KREIKim, Kwan-soo, Professor, Seoul National University, KoreaKim, Yoon-hyong, Professor, Chonnam National University, KoreaLee, Byoung-hoon, Professor, Kangwon National University, KoreaLee, Sang-min, Senior Research Fellow, KREIPark Joon-kee, Research Director, KREISim, Jae-hun, Senior Research Fellow, KREIYeo, Jun-ho, Professor, Kyungpook National University, Korea

Journal of Rural Development (JRD) is a collection of scholarly papers on agricultural economics, rural sociology, regional planning and other related academic fields. Written in English and registered as Nongchon-Gyeongje, the journal is published annually by the Korea Rural Economic Institute (KREI).

■ Nongchon-Gyeong je was selected as an officially registered scientific journal in 2005 by the National Research Foundation of Korea (NRF).

■ Publication Cycle- four times a year: March 21, June 21, September 21, December 21

Communications concerning subscription and other matters should be addressed to:Seong, Jin-seokJournal of Rural DevelopmentKorea Rural Economic Institute601, Bitgaram-ro, Naju-si, Jeollanam-do, 58321, KoreaTel. (82-61)820-2212, Fax. (82-61)[email protected]

ISSN 1229-8263


Forest Carbon Sequestration and Optimal Harvesting Decision

Considering Southern Pine Beetle (SPB) Disturbance: A Real Option Approach

An Hyun-jin

1

An Economic Effect of the Crop Insurance at the Farmland in Korea

Park Ji-yun , Kim Chang-gil

35

The Production and Dissemination of Agricultural Knowledge

at U.S. Research Universities: The Role and Mission of Land-Grant Universities

Lee Yoo Hwan , Gregory D. Graff

63

Impact of Increased Imports of Agricultural Products

due to FTAs on Domestic Price Decline

Jeong Min-kook , Moon Han-pil , Song Woo-jin

105

Trends in South Korea’s Grants-Based Aid for Agricultural Sector

in Developing Countries

Lee Hye-jin

125

Journal of Rural Development 40(Special Issue): 1~33 1

FOREST CARBON SEQUESTRATION ANDOPTIMAL HARVESTING DECISION CONSIDERINGSOUTHERN PINE BEETLE (SPB) DISTURBANCE:A REAL OPTION APPROACH

AN HYUN-JIN*

Keywords

real option, flexible harvest, carbon sequestration, southern pine beetle,

value of uncertainty

Abstract

This study evaluates the forest management decision making of loblolly

pine forest in the southern U.S. using the real option approach. The study

incorporates three uncertainties that forest owners have faced including

timber price volatility, forest carbon sequestration, and impacts of insect

outbreaks into the real option model to investigate the relationship be-

tween such uncertainties and forest bare land value and tree rotation

age. The results show that forest owners can face a mixed outcome of

these uncertainties when they make forest management decisions, and

the real option approach helps the forest managers consider future con-

sequence through allowing the flexible harvest decision. Generally, a high-

er bare land value is generated if a flexible harvest decision making (real

option) is allowed compared to a fixed harvest. The standing tree seques-

trates CO2, and the forest’s role of carbon sequestration could generate

extra value in the forest while insect outbreaks reduce the bare land

value. The increasing social cost of carbon tends to call for increasing the

bare land value of forest tree rotation age. Therefore, as climate change

becomes more looming due to CO2 concentration in the atmosphere,

the value of standing forests would increase due to enhancing oppor-

tunity cost of carbon sequestration in forests. Continuous efforts of pest

management for forests are necessary since a higher insect risk tends to

reduce the bare land value of forests. Also, employing marketable climate

policy such as emissions trading is necessary to create a market carbon

price and offset extra cost to keep the forest.

* Research Fellow, Korea Rural Economic Institute, Naju-si, Jeollanam-do, Korea.

e-mail: [email protected]

Journal of Rural Development 40(Special Issue)2

I. Introduction

Forest owners in the southern U.S. region are facing several risks, and these risks

are increasing in magnitude with climate change. Uncertainties associated with

management decisions are challengeable tasks of forest managers because in-

appropriate decision making can result in the loss of economic opportunities and

profits due to the irreversible characteristic of forests. Moreover, theongoing cli-

mate change tends to accelerate the uncertainties by altering forest disturbance and

forest ecology. The fundamental challenges for forest resource management deci-

sion making are evaluating trade-offs between the social-economic benefit of har-

vesting timber and the ecological benefit of preserving the forests (Morgan,

Abdallah, and Lasserre 2007). To examine this need, this paper investigates a de-

veloped methodology to adopt for forest management strategy under uncertainties.

This study applies the real options valuation approach to the field of forest man-

agement decision making considering various cases that forest owners might face.

The real option approach can supplement the main weakness of traditional forest

management evaluation, because it takes into account the flexibility of harvest de-

cisions due to timber price fluctuations (Tee et al. 2014). Also, the real option ap-

proach does fully consider the possibility of reversible investment opportunities

(Duku-Kaakyire and Nanang 2004).

The definition of real option is the value of being able to choose some

characteristic of decision allowing flexible outcome (Saphores 2000). The term

"real" refers to tangible assets such as facilities and natural resource,and several

studies have adapted a real options framework to the field of forestry.

Developments in real option study in forestry have increased the need for risk

management of forest investment and forest business management for optimizing

the financial performance of forest assets.

The real option approach is very useful in understanding tradeoffs be-

tween timber and ecosystem services provided by forests to incorporate un-

certainties and flexibility in timing (Alavalapati and Kant 2014). Tee et al. (2014)

applied real options analysis of forestry carbon valuation under the New Zealand

emission trading scheme. They incorporated both stochastic timber price and car-

bon value into calculating real option value of the New Zealand forests using the

binomial tree method.

Forest Carbon Sequestration and Optimal Harvesting Decision Considering Southern Pine Beetle (SPB) Disturbance 3

This study demonstrates the utility of the real options valuation approach

to the field of forest management decision making considering various cases that

forest owners might face. The term “valuation” means the bare land value of lo-

blolly (Pinus taeda) pine plantation in southern U.S. Loblolly pine is the most

commercially important forest species in the southern U.S., and its native range

extends throughout 14 states from southern New Jersey to central Florida and to

eastern Texas (Baker and Langdon 1990). The objectives of the study are to find

answers to the following questions:

1) How does the sawtimber price volatility affect the bare land valuation of

loblolly pine forests in the southern U.S.?

2) How much could the bare landvalue be changed if we consider not only

timber price but also the carbon sequestration ability of the forest and

pine beetle outbreak risk?

3) What is the optimal harvesting decision for loblolly pine plantations in

the southern U.S. considering timber price volatility, carbon value, and

pine beetle infestation risk?

This study applies binomial tree methods based on Guthrie’s approach

(2009), for evaluating real option value. The binomial tree methods have several

advantages such as numerically efficient and conceptually undemanding technique

to calculate option value. The main contribution of this study is to evaluate the

optimal stand management decision considering timber price, carbon sequestration,

and trees damaged by insects, southern pine beetle (SPB) in particular, which is

one of the main causes of tree damages in the southern U.S. There are many stud-

ies that evaluate the value of the forests using the real options theory but re-

searchers have not treated damaged trees in detail. Insect infestation directly af-

fects forest owner’s profit because it reduces timber productivity. Regarding forest

carbon sequestration, dead trees do not release significant amounts of CO2into the

atmosphere than expected because dead trees hold their carbon for a long time and

prevent it from quickly being released into the atmosphere (Moore et al. 2013).

Thus, damaged trees represent a substantial proportion of the total carbon

sink/source in forest stands, and these damaged trees will affect tree management

decision such as harvesting age (Asante, Armstrong, and Adamowicz 2011).

Without considering this, the carbon sequestration ability of forest could be

underestimated. This paper provides guidelines for forest owners for improving

their timber harvest decisions to consider some cases they could face under cli-


mate change including timber price volatility risk, benefit from mitigation CO2 due

to forest carbon sequestration, and SPB outbreak risk.

II. Model setting up

1. Binomial tree of price movement

Timber price volatility is one of the critical uncertainties that forestland owners

could face. Suppose that is the current price of sawtimber ($/m3).

denotes the sawtimber price at the node (i, n), where i is the number of

downward price moves and n is the time step. Suppose that , are the size

of the up movement and down movement where ∆ and

∆

(see equation (A1) in appendix), respectively. Sawtimber price could be either in-

creased or decreased with probability or at each node. If sawtim-

ber price is increasedat the node , it could be and

when sawtimber price is decreased. The binomial tree of

sawtimber price movement process for is described in Figure A1. The forest-

land owners expect some profits from the sales of forest products; the amount of

the profit depends on the timber price movement in the market. Assume that this

timber price follows a mean-reverting series. Schwartz (1997) suggested a strong

mean reversion in the commercial commodity prices. The mean-reverting price

process implies that unlike the random walk price process, shocks to mean-revert-

ing timber spot prices are not permanent. In other words, the sudden increase in

timber price leads to an increase in supply as well, so the market price of timber

will move back towards the timber’s long-run marginal cost of production in

long-term. Likewise, a sudden decrease in timber price causes a reduction in sup-

ply that triggers increase in future timber price. Therefore, a sudden increase

(decrease) in timber spot price is not sustainable (Guthrie 2009).

Under the mean-reverting price assumption, the logarithm of the price fol-

lows a first order autoregressive process:

(1)

∼


where is the market price of sawtimber, is the error term that follows a nor-

mal distribution with mean=0 and variance= . After obtaining OLS estimated co-

efficients, , , and , from equation (1), we can calculate Ornstein-Uhlenbeck

parameters with the following equation using the OLS coefficients:

(2)1/201 1

1 1 1

ˆˆ ˆˆˆlog(1 ) 2 log(1 )

ˆ , , ( )ˆ ˆ(2 )ˆ

d d

a bt ta

sa

aa af

a

-- + += = =

D + D$

Where = mean reversion rate, = long-term level price, = volatility of the

Ornstein-Uhlenbeck parameters, and ∆ = size of time step. From the solution to

equation (2), the binomial tree parameters, , , and are calculated by

the following equations (See equation A4 in appendix) :

(3)

∆ ,

∆

if

∆

∆log≤

∆

∆logif

∆

∆log

if

∆

∆log≥

2. Calculating risk neutral probability using capital asset pricing

model (CAPM)

The risk neutral probability is the likelihood of future outcome under the assump-

tion that underlying risk asset has the same expected return as riskless assets such

as Treasuries bills (Hull 2008). Capital Asset Pricing Model (CAPM) can be ap-

plied to calculate the risk neutral probability. The risk neutral probability is

calculated by subtracting a Market Risk Premium adjustment () from the

valuation binomial tree’s probability (Guthrie 2009):

(4), and

1 .

U U adj

D U

MRPqP = -

P = -P


The is obtained by regressing returns on the market portfolio (Guthrie

2009). The common stock indices such as S&P 500 and NASDAQ are widely

used as a proxy for the market portfolio. This study uses the S&P 500 index as

a proxy of the market portfolio.

3. Binomial tree of valuation movement

The forest value in each node is denoted by , and is related to tim-

ber price movements and . The two-step valuation binomial tree

( ) is shown in Figure A2. The forest value could be increased with proba-

bility or decreased with probability . n is the time step (year) and

i is the number of down movements. The risk neutral probability can be expressed

as and . The two-step valuation binomial tree with

risk neutral probability is shown in Figure A2. The valuation binomial tree is cal-

culated backwards starting from where N denotes the terminal time step

and the ending is . Therefore, valuation at node is

(5)

( , 1) ( 1, 1)( , )

f

DU

f

V i n V i nV i n

R R

P + P + += +

where fR = (1+discount rate).

At node , the forestland owner faces two alternative situations. The

first alternative is harvesting. If she/he decides to harvest the forest, she/he must pay

the harvest cost H per timber volume. Total costs are equal to where

is the total volume of the timber harvested. She/he gains some revenue from selling

the timber, which is equal to , where indicates the expected mar-

ket timber price in the nth time period. After harvest, the forestland is turning into

bare land worth per hectare. is the bare land value after harvest. She/he also

must pay taxes at a rate of T. All in all, the harvest payoff equation is

(6) (1 )( ( , ) ) ( )T X i n H Q n B- - +

The second alternative is that the forestland owner decides not to harvest,

rather postponesthe harvest until an appropriate timber price is going to be


reached. In this case, she/he must pay forest maintenance cost per hectare. After

one period, the timber price is going to move either up and down. So the corre-

sponding forest value is either or . Thus, the expected

payoff from postponing harvest is

(7)( , ) ( , 1) ( , ) ( 1, 1)

(1 ) M duT

f

i n V i n i n V i nT

R

P + +P + +- - +

for all where is the terminal node and MT is the forest maintenance

cost. The payoff at the terminal node is

(8) (1 )( ( , ) ) ( )T X i N H Q N B- - +

At each node, the decision to harvest or not harvest is re-evaluated. If the present

value of the cash flows from harvesting is larger than the current value of the cash

flows from not harvesting at the node, the optimal decision is to harvest at this

node. On the other hand, if the present value of the cash flows from not harvesting

is larger than the current value of the cash flows from harvesting, the optimal deci-

sion is not harvesting at this node. Therefore, the valuation at each node is

(9)

(1 )(( ( , ) ) ( )) ,

( , ) ( , 1) ( , ) ( 1, 1)( , ) max(1 ) u d

T

f

T X i n H Q n B

i n V i n i n V i nV i nT M

R

- - +ì üï ï

+ + + += í ý- - +ï

P

î

P

The first line in the max function, equation (9), implies the cash flow from

harvesting. On the other hand, the second line represents the cash flow from not

harvesting. The forest owner makes a decision by comparing the present values

of the corresponding expected future cash flows at every node. This problem is

solved by calculating backwards, starting from the terminal node where

and ending at .


4. Market value of bare land

The backward procedure is conducted recursively over multiple iterations and each

iteration represents one harvest/planting rotation. Calculating the market value of

bare landfollows these steps: (1) The bare land value is zero when calculating val-

ue for the first iteration. (2) After finishing the first iteration, (The market

value of the forest at date 0) is obtained. (3) The bare landvalue is estimated by

which implies minus the cost of replanting the for-

est, where G is regeneration cost and T is tax rate. This first iteration bare land

value implies real option value for a single rotation (the value for single rotation

forest with flexibility). When calculating the value of the second iteration, the bare

land value derived from the first iteration is used as the new initial value instead

of 0. This process is repeated until the bare landvalues converge. This converged

bare land value is the real option value with infinite rotation (value of an infinite

rotation forest with flexible harvest).

5. Value of flexibility

The value of flexibility is calculated by comparing bare land value from fixed har-

vest with the value of real option. The valuation method for fixed harvest follows

the same process with real option but assumes that the harvest date is fixed.

Suppose that the harvest decision is fixed at node (e.g., 30 years or any years

smaller than the terminal node (100 year), ), the terminal condition is

and the years larger than are ignored. The termi-

nal condition is still not different from that used in the real option method except

that instead of is used. However, at all nodes earlier than , there is no

reevaluation of the decision since the harvest date is fixed. Therefore, the decision

to "wait" is only at nodes and the recursive equation at nodes

to becomes

(10)( , ) ( , 1) ( , ) ( 1, 1)

( , ) (1 ) u dT

f

i n V i n i n V i nV i n T M

R

P + + + += +

P- -

The value of bare land converges to the value under the infinite rotation after cer-

tain number of iterations. This value is the Land Expectation Value (LEV) of in-


finite rotation (Tee et al. 2014). The difference between LEV and real option

(flexible harvest decision) value is the value of flexibility.

III. Application of real option to flexible harvest decision

Forests play a significant role in carbon sequestration because trees absorb carbon

during growth. Several studies (Alavalapatiand Kant 2014; Tee et al. 2014;

Petrasek and Perez-Garcia 2010) have asserted that we should consider forests not

only as a source of timber but also a carbon pool. Therefore, the stock of stored

carbon in trees should be considered when we choose the optimal harvest age.

Many studies have examined the relationship between optimal harvest age and car-

bon storage ability to stand trees, but most analyses have focused on carbon se-

questration only in living trees. Dead trees, however, represent a significant pro-

portion of the total carbon stored in a forest (Asante and Armstrong 2012).

Therefore, stored carbon by dead trees may be necessary when determining opti-

mal harvest age. This study aims to establish three different real options models

to compare optimal harvest ages and bare land prices.

1. Timber only

The valuation function for timber only is the same as equation (11) discussed in

the previous section:

(11)

(1 )(( ( , ) ) ( )) ,

( , ) ( , 1) ( , ) ( 1, 1)( , ) max(1 ) u d

T

f

T X i n H Q n B

i n V i n i n V i nV i nT M

R

- - +ì üï ï

+ + + += í ý- - +ï

P

î

P

The terminal node is 100 years and the results for the rotation ages of up to

90 years will be reported.


2. Timber and carbon storage in living trees

Carbon of trees provides additional benefit to forest owners. Carbon benefits are

usually considered the amount of carbon per unit volume of biomass (Amacher,

Ollikainen, and Koskela 2009). Since as a growth function of a forest at

time and as the carbon stock (t/ha) in the forest of volume , the change

in the benefit from sequestrated carbon in living trees is a function of time :

where is the social cost of carbon. The stored carbon

in standing living trees is derived from a forest ecosystem yield table. The forest

ecosystem yield table (J. Smith et al. 2006) provides tabulated carbon density at

different stand ages and timber volumes by carbon pools including live trees,

standing dead trees, soil organic matters and so on. If timber age or volume is

not explicitly provided in the table, the carbon stock is estimated using an inter-

polation method. The real option valuation function for carbon sequestration by

trees is:

(12)

(1 ){( ( , ) ) ( ) X ( 1)} ,

(1 )( [( , ) max

( , ) ( , 1) ( , ) ( 1, 1

(

)

) ( 1)])

s c

T s c c

u d

f

T X i n H Q n Q n B

T M X QV i n

i n V i n i n V

n Q n

i n

R

-

- - - - +ì üï ïï ïï ï- - - += í ýï ï+ + + +ï ï+ï ïî þ

-

P P

3. Timber and carbon storage in living trees and dead trees

damaged by SPB

The SPB infestation risk affects both the amount of carbon sequestration in trees

and timber/wood products per unit forest land area. The trees killed by SPB have

a lower merchantable value and preserve less carbon than healthy trees, but these

dead trees still represent a substantial proportion of the total carbon stored in for-

est stands (Asante and Armstrong 2012) and can/will be replaced by new trees

naturally and with human assistance. Assume that the percent of trees killed by

SPB in each year is given by %. The forest owners may clear cut or damaged

trees in the same year or delay the harvest to a future year. In this case, one


should separate the two carbon sequestration pools: 1) carbon pool from live

standing trees, and 2) carbon pool from trees killed by SPB.

The timber production in year will decrease due to SPB damage.

Assume that the average yearly SPB damage is given by %, then the total timber

production ( /ha) in year will decrease according to equation (13). Therefore,

the total tree production will be instead of as given below:

(13) *( ) ( ) ( )n Q n nQ Qd= - .

The value of the live standing tree pool is

(14)* *( ) ( 1)s c cX Q n Q né ù- -ë û .

Equation (14) implies the value of carbon stored in live standing trees in

each year. is carbon density (t/ha) and is the social cost of

carbon ($/t). Assume that average yearly SPB damage is given by %, then the

total volume of live trees on the site in year is . The car-

bon density stored in live trees, is calculated from the forest ecosystem

yield table with the corresponding volume using an interpolation method.

The damaged tree pool (DTP) implies carbon stored in standing dead trees

killed by SPB. The trees killed by SPB are assumed to decompose at a rate of

per year, and trees killed by SPB are added to the DTP each year. Therefore,

the DTP pool grows according to

(15) ( 1) (1 ) ( ) (n 1)D n D n Qh d+ = - + +

where represents carbon stored in the damaged tree pool. The estimated

decomposition rate is =0.00578, which is derived from Asante, Armstrong, and

Adamowicz (2011). is the average SPB risk. The change in DTP for the no

harvest case is ∆ , which implies

∆

where the discount factor. Combining

all the equations stated above yields the real options value function under SPB risk:


(16)

**

**( )

(1 ){( ( , ) ) ( ) X [ ( 1) D(n 1)]} ,

( , ) max (1 )( ([

( ,

( 1)]

) ( , 1) ( , )

( ))

1)

)

( 1,

s

T s

d

c

cc

u

f

n

T X i n H Q n Q n B

V i n T M X Q

i n V i n i n V i

Q n D n

n

R

- - + D

P P

ì ü- - - - + - +ï ïï ïï ï

= - - - +í ýï ï

+ + + +ï ï+ï ïî þ

Starkey et al. (1997) examined that SPB infected at least 10 percent of the slash

and/or loblolly pine forest in the southern U.S. Reed (1979) simulated the spread

of SPB infestation using a nonlinear spot growth model. He tested the model on

11 infestation spots from northern Georgia and projected 6% of the total number

of tree killed by SPB. However, it was not very precise model to estimate dam-

ages from individual infestation (Thatcher 1981). There are not many studies to

investigate the SPB infestation in loblolly pine forest only and previous studies

cannot reflect the current trend of SPB infestation in loblolly pine forest. With this

limitation, this study assumes 3% of SPB damages. This number may reflect the

current overall trend of SPB infestation risk in the southern U.S. Because the SPB

risk is assumed to be constant, sensitive analysis will be performed.

IV. Data and cash flows

1. Timber volume and mean carbon stock in the South and South

Central region

The mean volume of timber growth and estimated carbon stock for loblolly pine

in the southern U.S. are shown in Figure A4 and Figure A5 in the appendix,

respectively. The mean volume of timber growth and the estimated forest carbon

stock of southern (or loblolly) pines are obtained from "Forest Ecosystem Carbon

Tables" from the USDA Forest Service (J. Smith et al. 2006). The Tables were

developed using a national-level forest carbon accounting model (FORCARB2), a

timber projection model (ATLAS), and the USDA Forest Service and the Forest

Inventory and Analysis (FIA) Program’s database on forest survey (FIADB) (J.

Smith et al. 2006).


2. Costs and cash flows

Forest management costs and cost cash flows are shown in Tables A1 and A2 in

the appendix. These costs are based on market research (Doran et al. 2009).

Carbon stocks are calculated based on the timber volume for the loblolly

pine forest (living and dead trees, m3/ha) using the forest carbon table in “Methods

for Calculating Forest Ecosystem and Harvested Carbon with Standard Estimates

for Forest Types of the United States” (J. Smith et al. 2006). The average stum-

page price of southern pine sawtimber price movement is shown in Figure A6 in

the appendix and $150 is the long - term level price of southern pine sawtimber

stumpage price calculated by equation (2). The timber stumpage price is an ideal

state variable for calculating forest value because the timber stumpage price is the

price of timber while it is still standing. So the stumpage price does not reflect

the additional cost such as cost of harvesting and transporting the log to the mill

(Guthrie 2009). The social costs of carbon (Figure A7 in the appendix) used in

the model are obtained from the Interagency Working Group’s Technical Support

Report (Council of Economic Advisers et al. 2013).

V. Results

1. Land value (real option), harvest threshold and value of flexibility

The results for the flexible harvest (real option) of infinite rotation are shown in

Figure 1. For the timber only cases, the bare land value converges to $5329/ha,

after nine cycles/rotations of harvest-and-replant. For the timber plus carbon case

($75/ha of carbon cost is assumed), the bare landvalue converges to $7408/ha, af-

ter eight cycles of harvest-and-replant. For the case considering damage of SPB

case (a 3% of SPB damaged is assumed), the bare landvalue converges to

$6918/ha, also after eight cycles of harvest-and-replant. To consider the carbon

storage ability of forest, the forest value would increase by 39%, compared to the

case of considering only timber price. The SPB risk would decrease the forest

value. The bare land value damaged by SPB would decrease by 6% compared to

the case of the timber plus carbon forest. However, the SPB damaged forest still


has a higher value than the timber only case because even if SPB damages the

forest, the forest still has the ability of carbon storage. Thus, the value of carbon

storage would compensate the price loss from damaged timber by SPB.

Figure 1. Infinite rotation values for bare land

The market value of forests for fixed harvest of infinite rotation is given

in Figure 2. The infinite rotation problem is commonly known as the Faustmann

rotation, which is defined as “choosing the harvest period to maximize the net

present value of a series of future harvest” (Grafton et al. 2008: 138; Gane,

Gehren, and Faustmann 1968). In this study, the NPV of a forest could be in-

dicated as a sum of discount net cash flow over an infinite time horizon (Viitala

2006). For evaluating the value of forests for fixed harvest, the same process is

used with flexible harvest, but the fixed harvest case assumes that the harvest de-

cision is fixed at the node t= fixed harvest age. Thus, the backward evaluations

are started from node t (e.g.: 60 years, 50 years) rather than N (100 year), without

no re-evaluation of a harvest decision. Thus, the valuation equation for each node

equals to equation (17) and value of bare land converges to infinite rotation NPV

of the fixed harvest.

(17)( , ) ( , 1) ( , ) ( 1, 1)

( , ) (1 ) u dT

f

i n V i n i n V i nV i n T M

R

P + + + += +

P- -


Under the fixed harvest assumption, generated net present value (NPV) of the for-

est by timber only case is, $3220/ha, around age 30. In timber plus carbon case,

the net present value of forest is at its maximum, $4812/ha, at age 40. In the case

of timber plus carbon under SPB risk, thenet present value of the forest is the

highest, $4308/ha, at age 40. If allowed for flexible harvest (real option), the mar-

ket value of the bare land is $5329/ha for the timber only case, $7408/ha for the

timber plus carbon case, and $6918/ha for the case of timber plus carbon under

SPB risk, respectively. Thus, timber harvest flexibility adds approximately 65% to

the value of bare landfor the timber only case (54% for the timber plus carbon

case, 61% for the case of timber plus carbon under SPB risk). This result shows

that flexible harvest generates the higher valuation through allowing forest owners

to make a better investment decision using information of various price levels. If

timer prices are low, the forest owners can postpone harvest while they hasten har-

vest when prices are high.

Using these results, we can estimate the optimal harvest/rotation age as

well. The NPV of the forest is maximized at the point of optimal rotation age for

both fixed rotation and infinite rotation. The optimal rotation age is 30 years for

the timber only case, 40 years for the timber plus carbon case and 40 years for

the case of timber plus carbon forest under SPB risk. The optimal rotation age

increases when considering the carbon storage ability of the forest. In the case of

SPB damage, the optimal rotation is similar to the carbon forest case, but the for-

est value is lower than that under the carbon forest case at the optimal rotation

age. The value of flexibility also increased if we consider carbon storage ability

of the forest because the capacity to be flexible can increase the value of invest-

ment when uncertainty and irreversibility become larger (Tee et al. 2014).


Figure 2. Market value of bare land: fixed harvest, infinite rotation

Figures 3-5 show the optimal harvest threshold for infinite rotations, tim-

ber only case, carbon plus timber case and carbon plus timber under SPB risk.

The values are rounded off to the nearest whole number. These figures show the

harvest threshold price for all possible ages of the forest. The shaded area implies

the range of sawtimber price that is optimal to harvest for a given forest age. In

every case, if the forest is very young (less than 10 years old), the optimal choice

is not to harvestunless the timber price would become extremely high. However,

as the age of the forest increases, the threshold price falls. For example, in Figure

3, if the timber price is above $258/ when the forest age is between 20 to 26

years old, the optimal decision is to harvest while the optimal decision would be

to defer harvest if the timber price is below $258/ .


Figure 3. Sawtimber threshold prices for the timber-only case

Figure 4. Sawtimber threshold prices for the carbon-forest case

Figure 5. Sawtimber threshold prices for the carbon-forest under SPB risk case


Figure 6 and Table 1 compares the timber threshold price changes among

timber only case, carbon plus timber case, and carbon plus timber under SPB case

for all possible ages of the forest. It is apparent from this figure and table that

harvest threshold price decrease as trees age for all three cases. If the age of trees

is younger than 10 years, the optimal decision is not to harvest in all cases. The

threshold price tends to be high under the cases with considering carbon storage,

compare to timber only. This is because carbon store ability of forest incurs a

higher opportunity cost of harvesting the forest, therefore, to offset the burden of

harvest, a higher timber price (revenue) would be required compare to timber only

case. The SPB damage reduces the advantage of standing forest its threshold price

is higher than timber only case because dead trees still provide carbon

sequestration. The benefit from carbon sequestration of standing tree partially com-

pensates the lost from reducing the total volume of harvest by SPB damage.

Figure 6. Comparisons of threshold price changes: timber only vs. carbon forest under

SPB vs. carbon forest

Area above the line is optimal harvesting zone


Table 1. Comparison of timber threshold price ages

Age Timber only Carbon Forest Carbon forest under SPB risk

10 Not harvest 1000 1000

11 460 531 531

20 258 344 297

40 223 258 227

60 193 223 193

80 167 223 167

86 144 167 144

89 125 167 144

2. Sensitive analysis for carbon social cost

Figure 7 presents infinite rotation valuation for fixed harvestunder various levels

of social cost of carbon. As the social cost of carbon increases from $50/t to $75/t,

the expected NPV of the forest increases from $4224/ha to $5164/ha at 2.4% dis-

count rate. The optimal rotation age does not change, but the bare land values

changes according to difference of carbon social cost; as the social cost of carbon

increases, the value of the forest increases.

Figure 7. Market value of bare land under various levels of social cost of carbon: Fixed

Harvest


Figure 8. Market value of bare land under various social costs of carbon: Flexible harvest

The bare land price changes for flexible harvest (real option) of infinite

rotation under various levels of social cost of carbon are shown in Figure 8. If

the carbon social cost is $50/t, the bare land value converges to $6699/ha, after

eight cycles of harvest-and-replant. If the carbon social cost is $75/t, the bare land

value converges to $7408/ha, after eight cycle of harvest and replant. If the carbon

social cost is $90/t, the bare landvalue converges to $7841/ha, after eight cycle

of harvest-and-replant. Compare to fixed harvest case, flexibility adds approx-

imately 59% to the value of bare land under a $50/t social cost, 54% under a $75/t

social cost, 51% under a $90/t social cost.

The timber threshold price changes for all possible ages of the forest un-

der various level of social cost of carbon are presented in Figure 9. The harvest

threshold price decreases as the social cost of carbon decreases. No difference in

threshold price depending on social cost of carbon if the age of the forest is young

(less than 20 years old). If the forest age is 36 years, the timber threshold price

is $257/m3 for a $90/t of carbon social cost, $223/m3 for a $50/t of carbon social

cost, and $223/m3 for a $75/t of carbon cost, respectively. The timber threshold

price decreases as the trees grow. The higher social cost of carbon increases the

opportunity cost to harvest trees. Therefore, it requires a higher timber price is

necessary to compensate the loss of the opportunity cost associated with cutting

trees down. Therefore, as the carbon social cost increases, the forest owner would

consider delay timber harvest if anything else remains the same.


Figure 9. Timber threshold price by different social costs of carbon

3. Sensitivity analysis for SPB risk

Fixed harvest valuation (infinite rotation) under various SPB damage rates are il-

lustrated in Figure 10. If the SPB damage rate increases, the value of bare land

will decrease bare land. If SPB damages 1% of the forest, the forest value is

$4308/ha at the optimal rotation age (40 years old). However, if SPB damages 2%

of the forest, the forest value is $3681/ha at the optimal harvest age (30 years

old). If SPB damages 3% of the forest, the forest value is $2908/ha at the optimal

harvest age (30 years old). As the SPB risk increases, both the bare land value

and the optimal rotation age decreasebecause high SPB infestation reduces both to-

tal harvest volume and carbon sequestration ability of trees. This generates a po-

tential profit loss to the forest owners by reducing timber productivity in forest.

When forest owners make a harvest decision, they need to determine if the rate

of return from continuing the investment in the forest is worth more than the rate

of return received from an alternative investment (Jacobson 2015). Therefore, in-

centives from continuing to grow the trees would decrease under high SPB in-

festation risk by decreasing the future expected rate of return from continuing the

investment in the trees. Thus, forest owner’s choice is seeking other opportunities

to invest.


Figure 10. Value of bare land (with fixed harvest) at various SPB risks

The change of real option value (flexible harvest valuation) under various

SPB damage rates are shown in Figure 11. The real option values decrease from

$6918/ha to $5169/ha as SPB risk rises from 1% to 3%.

Figure 11. Market value of bare land (flexible harvest) changes at various SPB risks


The timber threshold price for harvesting at various SPB damage rates.

For example, the harvest threshold price is $297/m3 at a 1% SPB risk, $258/m3

at both 2% and 3% SPB risk at age 23 is shown in Figure 12. If the forest age

is 40 years, the threshold price is $223/m3 at 1% and 2% SPB risk, $193/m3 for

the case of 3% SPB risk. A higher SPB risk reduces the benefit from keeping the

forest. Therefore, harvesting is optimal at a lower timber price as SPB damage risk

becomes more severe, especially, if the forest is younger than 55 years.

Figure 12. Optimal harvest price flow at various SPB damage rates

VI. Conclusion

This paper evaluates the combined impact of the three factors on forest manager’s

decision making using real option approach. The major finding of this paper is

that flexible harvest decision making using real options is a better strategy than

the fixed harvest decision when forest owners face several uncertainties including

sawtimber price volatility, climate change, and insect outbreaks. A higher bare

land value is generated if a flexible harvest decision making (real option) by in-

corporating stochastic price movement is allowed because the value of flexibility


adds to forest valueswhen flexible harvest decision is allowed. The CO2 storage

of a forest enhances the bare land value while SPB outbreaks reduce the bare land

value. However, if we consider the carbon sequestration ability of damaged trees,

the bare land value is still higher than that without taking into account carbon stor-

age of damaged trees. The value of standing trees is higher as the carbon social

cost increases due to increasing opportunity cost of carbon sequestration on trees.

When social cost of carbon is high, the incentive from converting abandoned agri-

cultural land to forest land and using wood products instead of other material will

become higher. Moreover, the high social cost of carbon also adds value to wood

products because the wood products also contribute to carbon storage.

As the global CO2 concentration increases under climate change, the value

of carbon storage of forest would increase. Therefore, at higher social cost of car-

bon, higher timber price is required to warrant harvesting due to increasing oppor-

tunity cost of cutting trees. Higher SPB risk tends to reduce the bare land value

of forest. The high bare landvalue of carbon forest provides an incentive to forest

owners to plant new forests and perform intensive treatments to keep forests

healthy and productive. The U.S. forests currently absorb 10% of the national

greenhouse gas emissions (Ingerson 2009). Increasing the forest rotation age by in-

creasing the value of standing trees could enhance forests’ CO2 storage by defer-

ring harvest. This might provide positive impacts on CO2 mitigation in the south-

ern U.S. This study confirms that standing forests could provide social benefits by

absorbing CO2. However, planting new forests and keeping them healthy may re-

quire additional costs such as the cost of pesticide and fertilization. This might

carry an extra burden to forest owners. Therefore, policy makers should establish

legislation that provides additional incentives to forest owners to offset additional

burden by differing harvest and planting new forest. Emissions trading may be one

of the solutions. Under emissions trading, the forest owners could earn carbon

credit by standing forest and sell them in domestic and international market. For

example, under the the New Zealand Emission Trading Scheme (NZETS), the

post-1989 forests (planted on and after 1st January 1990) are qualified as carbon

credit that could be accumulated or immediately sold in carbon market (Tee et al.

2014). This could provide extra income to forest owners, and the extra cash flow

might generate incentives to forest owner to harvest new forests.

A limitation of this study is the absence of considering various forest

management practicesincluding pruning, thinning and fertilizing. Also, the pesti-


cide control impact should be considered in the case of SPB outbreak risk in fu-

ture research. The impact of CO2 fertilizations on forest productivity might be in-

cluded in real option valuation equations as well. The increments of timber prod-

ucts because of CO2 fertilizations may offset the loss from timber damages by

SPB infestation under climate change. To consider these factors, more sophisti-

cated real option valuation modeling approaches will be necessary for further

studies. Unless several limitations, I convince that the paper will give insights into

what forest owners need to do for improving their timber harvest decisions under

uncertainty. The optimal harvest thresholds in particular, provide a useful guideline

for forest owners by offering an insightful decision-making tool which can be

compared with actual timber price in every year.

REFERENCES

Alavalapati, Janaki, and Shashi Kant. 2014. Handbook of Forest Resource Economics. 1

edition. London: Routledge.

Amacher, Gregory, Markku Ollikainen, and Erkki A. Koskela. 2009. Economics of Forest

Resources. 1 edition. Cambridge, Mass: The MIT Press.

Asante, Patrick, and Glen W. Armstrong. 2012. “Optimal Forest Harvest Age Considering

Carbon Sequestration in Multiple Carbon Pools: A Comparative Statics Analysis.”

Journal of Forest Economics 18 (2): 145–56. doi:10.1016/j.jfe.2011.12.002

Asante, Patrick, Glen W. Armstrong, and Wiktor L. Adamowicz. 2011. “Carbon

Sequestration and the Optimal Forest Harvest Decision: A Dynamic Programming

Approach Considering Biomass and Dead Organic Matter.” Journal of Forest

Economics 17 (1): 3–17. doi:10.1016/j.jfe.2010.07.001.

Baker, James B., and Gordon Langdon. 1990. “Loblolly Pine.” vol. 1. Silvics of North

America. USDA Forest Service.

Bentz, B. J., and D. E. Mullins. 1999. “Ecology of Mountain Pine Beetle (Coleoptera:

Scolytidae) Cold Hardening in the Intermountain West.” Environmental Entomology 28

(4): 577–87. doi:10.1093/ee/28.4.577.

Conrad, Jon M. 2000. “Wilderness: Options to Preserve, Extract, or Develop.” Resource

and Energy Economics 22 (3): 205–19. doi:10.1016/S0928-7655(00)00031-2.

Council of Economic Advisers, Council on Environmental Quality, Department of

Agriculture, Department of Commerce, Department of Energy, Department of

Transportation, Domestic Policy Council, et al. 2013. “Technical Support Document: -

Technical Update of the Social Cost of Carbon for Regulatory Impact Analysis -

Under Executive Order 12866 -.” https://www.whitehouse.gov/sites/default/files/omb/

assets/inforeg/technical-update- social-cost-of-carbon-for-regulator-impact-analysis.pdf.

Doran, Faye S., Coleman W. Dangerfield, Frederick W. Cubbage, James E. Johnson,

James W. Pease, Larry A. Johnson, and George M. Hopper. 2009. “Tree Crops for

Marginal Farmland: Loblolly Pine.” 446–604. Verginia Cooperative Extension.

Duku-Kaakyire, Armstrong, and David M. Nanang. 2004. “Application of Real Options

Theory to Forestry Investment Analysis.” Forest Policy and Economics 6 (6): 539–52.

doi:10.1016/S1389-9341(03)00003-0.

Frank Werner, Ruedi Taverna. 2005. “Carbon Pool and Substitution Effects of an Increased

Use of Wood in Buildings in Switzerland: First Estimates.” Annals of Forest Science

62 (2005): 889–902. doi:10.1051/forest:2005080.

Gane, Michael, Edmund Franz von Gehren, and M. Faustmann. 1968. Martin Faustmann

and the Evolution of Discounted Cash Flow: Two Articles from the Original German

of 1849. Commonwealth Forestry Institute.


Grafton, Quentin, Wiktor Adamowicz, Diane Dupont, Harry Nelson, Robert J. Hill, and

Steven Renzetti. 2008. The Economics of the Environment and Natural Resources. John

Wiley & Sons.

Guthrie, Graeme. 2009. Real Options in Theory and Practice. Har/Cdr edition. Oxford ;

New York: Oxford University Press.

Hull, John. 2008. Options, Futures, and Other Derivatives. Upper Saddle River, N.J. :

Prentice Hall, [2008].

Ingerson, Ann. 2009. Wood Products and Carbon Storage: Can Increased Production Help

Solve the Climate Crisis? Washington, D.C.: The Wilderness Society.

IPCC. 2000. “Land Use, Land-Use Change and Forestry.” Cambridge, UK. http://www.ipcc.ch/

ipccreports/sres/land_use/index.php?idp=0.

Jacobson, Michael. 2015. “Forest Finance 8: To Cut or Not Cut — Tree Value and Deciding

When to Harvest Timber (Forest Finance and Taxation).” Forest Finance and Taxation

(Penn State Extension). Accessed October 5. http://extension.psu.edu/natural-resources/

forests/finance/forest-tax-info/publications/forest-finance-8-to-cut-or-not-cut-tree-value-and-d

eciding-when-to-harvest-timber.

Moore, David J. P., Nicole A. Trahan, Phil Wilkes, Tristan Quaife, Britton B. Stephens,

Kelly Elder, Ankur R. Desai, Jose Negron, and Russell K. Monson. 2013. “Persistent

Reduced Ecosystem Respiration after Insect Disturbance in High Elevation Forests.”

Ecology Letters 16 (6): 731–37. doi:10.1111/ele.12097.

Morgan, Don, Ben Abdallah, and Pierre Lasserre. 2007. “A Real Options Approach to

Forest-Management Decision Making to Protect Caribou under the Threat of

Extinction.” Ecology and Society 13 (1): 27.

Petrasek, Stanislav, and John M. Perez-Garcia. 2010. “Valuation of Timber Harvest

Contracts as American Call Options with Modified Least-Squares Monte Carlo

Algorithm.” Forest Science 56 (5): 494–504.

Reed, David Doss. 1979. Estimating Region-Wide Damages Caused by the Southern Pine

Beetle. Virginia Polytechnic Institute and State University.

Saphores, Jean-Daniel M. 2000. “Real Options and the Timing of Implementation of

Emission Limits under Ecological Uncertainty.”

Schwartz, Eduardo S. 1997. “The Stochastic Behavior of Commodity Prices: Implications

for Valuation and Hedging.” The Journal of Finance 52 (3): 923–73. doi:10.1111/

j.1540-6261.1997.tb02721.x.

Smith, James, Linda Heath, Kenneth Skog, and Birdsey A Richard. 2006. “Methods for

Calculating Forest Ecosystem and Harvested Carbon with Standard Estimates for Forest

Types of the United States.” General Technical Report NE343. United States

Department of Agriculture.

Starkey, Dale, Robert Anderson, Carol Young, Noel Cost, John Vissage, Dennis May, and Edwin

Yockey. 1997. Monitoring Incidence of Fusiform Rust in the South and Change Over Time.

U.S. Department of Agriculture, Forest Service, Southern Region Forest Health Protection.


Tee, James, Riccardo Scarpa, Dan Marsh, and Graeme Guthrie. 2014. “Forest Valuation

under the New Zealand Emissions Trading Scheme: A Real Options Binomial Tree

with Stochastic Carbon and Timber Prices.” Land Economics 90 (1): 44–60.

Thatcher, Robert C. 1981. The Southern Pine Beetle. U.S. Dept. of Agriculture, Expanded

Southern Pine Beetle Research and Applications Program, Forest Service, Science and

Education Administration.

Viitala, Esa-Jussi. 2006. “An Early Contribution of Martin Faustmann to Natural Resource

Economics.” Journal of Forest Economics 12 (2): 131–44.


APPENDIX

Figure A1. Two-step price binomial tree

Figure A2. Two step valuation binomial tree

Figure A3. Two step valuation binomial tree with risk neutral probability


Figure A4. Estimates of timber volume for loblolly pine stands in southern U.S.

Figure A5. Estimates of carbon stock for loblolly pine stands in southern U.S.


Table A1. Forest management costs

Management cost description Cost ($)

Regeneration cost (including the cost of site preparation, seedling, planting and weed control),

$618/ha

Forest management cost, $22/ha

Tax rate, 28%

Harvest cost $68.67/m3

Discount factor (Risk free interest ate base on current 20 year U.S. treasury rate),

2.5%

Table A2. Cost cash flow

Year 0 1 …15th

rotation…

24throtation

…90th

rotation

Planting Cost (618) (618) … (618) … (618) … (618)

Maintenance Cost, (22) (22) … (22) … (22) … (22)

Timber Revenue $ $ … $ … $ … $

Harvest Cost $ $ … $ … $ … $

Figure A6. Average stumpage price of sawtimber

Source: Louisiana Department of Agriculture


Figure A7. Revised social cost of CO2, 2010–2050 (in 2007 dollars per metric ton of CO2)

Source: Council of Economic Advisors 2013

Calculating size of up and down movement U, D

The size of an up and down movement U, D can be obtained from following process

(Guthrie 2009). The log price of is defined as log and which is

composed of the following equation

(A1) log ∆ ∆

where log = starting value, ∆ = effect of up moves, ∆ = effect

of down moves. Taking exponentials of both sides of this equation explain that the level of

the price at node and the up/down moves takes the price to

(A2)

∆

∆

∆

∆

∆

The size of an up and down moves, and , at this node must equal to the following

equation. The size of up and down moves are constant through the binomial tree.

(A3)

∆

∆


Calculating probability of up and down movement

The probability of an up movement for mean reverting process was calculated using

equation from Guthrie’s work (2009). The expected value of the change in the log price

over next period is equal to the value that is implied by our normalized parameter

estimates.

(A4)

∆

∆

∆ log

is the expected change in the log price, which is the same as the

expected value for the Ornstein-Uhlenbeck process. If the current log price is higher than

its long-run level, which is , then the price likely moves to the down, which is

. As the log price grows larger, a down move more likely to happen.

Conversely, if the log price is currently lower than its long-run level then an up move is

more likely than down move. If is sufficiently large, then will have

negative value. Similarly, if is sufficiently small, then will be greater than

one. However, since is a probability, the value of must be located between

0 and 1. Thus, our solution set equal to 0 if expression in equation (A4) has

negative value, and 1 if greater than 1. Therefore, the final form of the probability of an

up and down movement at node should be

(A5)

if

∆

∆ log≤

∆

∆logif

∆

∆log

if

∆

∆ log≥

Date Submitted: Oct. 28, 2016

Period of Review: Nov. 11. 2016～Dec. 15, 2017


AN ECONOMIC EFFECT OF THE CROP INSURANCEAT THE FARMLAND IN KOREA*

PARK JI-YUN**

KIM CHANG-GIL***

Keywords

climate change adaptation, crop insurance, extreme weather events,

Just-Pope model, evaluation of climate change adaptation options

Abstract

Climate change has a direct and indirect impact on agricultural pro-

duction through rising temperatures, changes in precipitation and ex-

treme weather events. To cope with climate change efficiently, it is im-

portant to carefully estimate the economic effects of adaptation meas-

ures and establish innovative methods based on the findings. In this

study, we examine statistically the damage and correlation of natural

disasters, which are soaring due to climate change, and farm income,

and measure the economic effect of crop insurance, which is a repre-

sentative option for climate change adaptation. To achieve the pur-

pose, we employ the Just-Pope model to perform an econometric anal-

ysis and use the data of orchard households. The empirical analysis dem-

onstrates that there exists a negative effect of extreme weather on farm

income and the negative effect increases as frequency of weather dis-

asters increases. However, the study also proves that crop insurance is

an effective adaptation measure and the economic effect of the crop

insurance is greater as more frequent extreme weather events occur.

Finally this study shows that insured farmers receive benefits of 1.39 mil-

lion KRW in comparison with uninsured farmers.

* This study utilized a part of the research “Economic analysis of adaptation measures to climate

change in the agricultural sector”.** Research Fellow, Korea Rural Economic Institute, Naju-si, Jeollanam-do, Korea.

Corresponding author. e-mail: [email protected]*** President, Korea Rural Economic Institute, Naju-si, Jeollanam-do, Korea.


I. Introduction

Climate change represents a statistically significant change in the average state of

the climate; the duration of this can span from years to decades. Climate change

can be caused by natural internal processes, external forcing or by artificial changes

in atmospheric composition or land use. In its fifth comprehensive report on climate

change assessment, the Intergovernmental Panel on Climate Change (IPCC) stated

that the impact of human activities on the climate system is a fact, and thus, diag-

nosed a wide impact. According to the global climate change prospects in the IPCC

(2014), if the current trend in climate change continues at the current greenhouse

gas emission level, the global average temperature in the late 21st century would

be expected to increase by 3.7 ℃; they also indicated that sea level is expected

to rise by 63 cm from 1986 to 2005. In the scenarios of global warming, it is

expected that weather patterns such as droughts, floods and typhoons, as well as

weather conditions will change greatly per region on a yearly basis.

Climate change has a direct and indirect impact on agricultural production

through rising temperatures, changes in precipitation and extreme weather events.

In many parts of the world, it has been clearly observed that climate change has

a significant impact on crop and food production; furthermore, negative re-

percussions are more common than the benefits. In particular, under climate

change scenarios where the average regional temperature increases by 3 to 4 °C

or higher, agricultural productivity will be adversely affected, thus, jeopardizing

world food production and food security (IPCC 2014).

In Korea, the temperature has risen by 1.8 ℃ over the past 100 years

(1912 ~ 2010) and the precipitation has increased by more than 200mm (Kwon

2012). In other words, the climate in Korea has been changing faster than the

global average. Compared to the mean temperature in the past 40 years (1970 to

2010), future climate change forecasts are projected to rise by 1.8 ℃ in 2020 and

3.7 ℃ in 2050 (KMA: Korea Meteorological Administration 2012). Specifically,

the average annual temperature of the Korean peninsula has risen by 1.2 ℃ (0.41

℃ / decade) for the past 30 years from 1981 to 2010, and the average annual pre-

cipitation tends to increase slightly by 78mm (KMA 2012).

As climate change worsens, extreme weather events such as heavy rain-

fall, typhoons, drought, cold waves and heavy snowfall frequently occur and agri-

An Economic Effect of the Crop Insurance at the Farmland in Korea 37

cultural damage increases accordingly. As the sub-tropicalization of the Korean

peninsula has progressed rapidly, the quantity of farmlands and the pattern of the

cultivation area have changed; the winter pest damage has also increased (Kim et

al. 2009).

The agricultural sector is more vulnerable to climate change than it is to

other industries. Therefore, the impact of climate change on agriculture should be

scientifically analyzed, wherein systematic and phased adaptation measures should

be prepared. Since climate change adaptation measures require a considerable amount

of time and budget, it is important to carefully estimate the economic effects of

adaptation measures and establish innovative methods based on the findings.

In order to establish a comprehensive and effective adaptation measure to

climate change, an economic analysis of options for responding to climate change

should be preceded, but the analysis of various climate change adaptation options

is still lacking. Hence, in this study, we analyze the economic effects of crop in-

surance under climate change, particularly extreme weather events in the farm

level. Through the analysis, we evaluate the crop insurance as an adaptation option

for climate change.

II. Background and Literature Review

1. Extreme Weather Events in Korea

According to the IPCC (2014), climate changes include sudden changes in precip-

itation pattern or increasing frequency and intensity of extreme weather such as

typhoon, localized heavy rains, cold waves and heavy snowfalls, as well as in-

creasing average temperature and precipitation. Recently, Korea has also experi-

enced greater frequency and higher intensity of these extreme weather events. The

increasing special reports being issued on extreme weather by the Korea

Meteorological Administration is proof of this trend. The special reports on ex-

treme weather refer to announced forecasts when disasters are likely to occur due

to extreme weather. Currently, these reports are issued in the case of strong winds,

heavy seas, heavy rainfall, heavy snowfall, droughts and tsunamis, as well as

winds carrying yellow dusts, cold waves, typhoons and heat waves.


As Table 1 shows, over the last 20 years, the total number of issued spe-

cial reports on extreme weather increased by 241%, from 596 in 1995 to 1,439

in 2015. Especially, special reports on cold wave have been issued 79 times annu-

ally on average in the last five years, compared to only once on average between

1990 and 1994. On other weather disasters such as heavy rainfalls, heavy snow-

falls and draughts, the frequency of special reports issued has been significantly

increased; that is, the average temperatures and precipitation in Korea are increas-

ing, but the fluctuation of weather is getting bigger with the increase of freezing

and drying phenomenon.

Table 1. Number of special reports issued each year

YearTotal number of

special reports on extreme weather

Heavyrainfalls

Cold wavesHeavy

snowfallsDraught

1990 655 169 0 46 0

1991 529 114 0 68 10

1992 448 82 3 37 11

1993 534 127 0 39 8

1994 640 58 1 41 25

1995 596 101 4 27 16

1996 376 58 1 50 6

1997 465 116 1 45 21

1998 1,097 320 10 47 30

1999 1,051 294 7 50 25

2000 374 205 4 51 30

2001 866 161 5 91 59

2002 1,021 167 0 52 34

2003 1,138 294 14 81 39

2004 1,006 241 8 59 55

2005 1,016 323 8 210 43

2006 957 305 3 125 49

2007 1,152 428 0 82 41

2008 1,467 354 12 167 89

2009 1,760 526 51 159 187

2010 1,760 601 53 280 85


Source: MPSS (Ministry of Public Safety and Security), Each year.

According to the Natural Disaster Yearbook published by the MPSS, since

1990, the scale of damage to the agricultural sector by natural disasters has sig-

nificantly increased, including but not limited to farmland loss and destruction.

The amount of damage by natural disasters increased from 28.4 billion KRW aver-

age in the 1960s to 101.5 billion KRW in the 2000s. The analysis of damage by

natural disasters after 1916 reveals that in 6 of 10 years, most damage occurred

due to natural disasters after 2000. In particular, in 2002 typhoon ‘Rusa’ and lo-

calized heavy rainfall caused the bulk of the damage. It was the year of the worst

damage by natural disasters in Korea’s recorded history. The agricultural sector al-

so suffered from loss and destruction of farmland, estimated at 585.3 billion won.

It is estimated that the total damage in the agricultural sector may be more than

the amount announced by the MPSS, given that crop damage is greater than farm-

land damage (loss or destruction of land) due to natural disasters in the agricul-

tural sector.

YearTotal number of

special reports on extreme weather

Heavyrainfalls

Cold wavesHeavy

snowfallsDraught

2011 1,652 661 86 208 117

2012 1,614 458 111 242 101

2013 1,465 447 87 222 122

2014 1,460 365 82 288 155

2015 1,439 264 29 143 189

1990-1994(average) 561 110 1 46 11

2011-2015(average) 1,526 439 79 221 137

(continued)


Figure 1. Amount of damage from farmland loss and destruction due to

natural disasters (1958 to 2015)

Unit: million won

Source: MPSS, Each year.

The frequency of extreme weather events in Korea is rapidly increasing

and the damage in the agricultural sector due to extreme weather is also

increasing. However, most previous studies that analyze the impacts of climate

change on agriculture, considered changes in mean temperature and precipitation

as major regressors. In other words, they are trying to estimate the effects of

changes in mean temperature or precipitation on agriculture (Adams et al., 1990;

Kim et al. 2009; Gammans, Mérel and Ortiz-Bobea 2017).

However, damage to agriculture due to climate change is more likely to

be caused by extreme weather events that occur intensively in a short time than

slowly changing temperature and precipitation. Changes in average temperature

and precipitation can be accommodated by changes in cultivation areas or culti-

vation periods, but extreme weather events such as typhoons, cold waves or heat

waves are difficult for farmers to respond in the short term. Previous studies veri-

fied that extreme weather events have negative effects on crops. Lee et al. (1991)

review the effects of meteorological disasters on the productivity of oilseed crops

and suggest the variety improvement and the advanced cultural practice for stable

production. Kim et al. (2010) analyze the damage situation of seedlings caused by

meteorological disasters and proposed measures for them. Kim et al. (2012) dem-


onstrate the negative effects of heat waves and heavy rainfalls on the rice yield

through the panel data analysis. Lobell, Sibley and Ortiz-Monasterio (2012) show

that under extreme heat (greater than 34°C), there is statistically significant accel-

eration of withering with the satellite data of wheat in India. Through the empiri-

cal analysis of U.S. crop yields, Schlenker and Roberts (2009) provide evidence

that crop yields are optimally increasing with temperature, but if temperature rises

more than the thresholds, yields decline sharply. Welch et al. (2010) estimate the

impacts of daily minimum and maximum temperature and solar radiation on rice

yields in tropical and subtropical Asia; they find that as minimum temperature in-

creases, rice yields decrease, but with higher maximum temperature, rice yields

increase. However, these are only a measure of the physical damage of crops, and

the impact on the farm income and farmer's risk management has not yet been

analyzed.

2. Crop Insurance in Korea

The crop insurance in Korea was introduced in 2001 to help farmers to adapt to

the effects of natural disasters, to stabilize income and support farmers. It is one

of the most widely adopted adaptation strategies to climate change in Korea.

The crop insurance started in 2001 and insured against damage to apple

and pear trees due to typhoons, hail and frost. In 2001, 8,055 farm households

bought the crop insurance and 17.5% of total production area was enrolled, which

is 4,096 ha. Compared with 2 items in 2001, the range of insurance coverage was

further extended to 46 items and various other natural disasters in 2015. In 2015,

122,054 farm households bought the crop insurance, covering 185,239 ha of farm-

land (respectively, a 15-fold and 45-fold increase since 2001). Since the crop in-

surance was extended to rice from 2009, the number of insured farm households

and the area of insured farmland increased. However, rice growers were not great-

ly interested in buying the insurance initially and the insured ratio of rice was low,

so the total insured ratio also dropped sharply as Figure 2 presents. On the other

hand, for orchard for special risk introduced in 2001, the number of insured farm

households has steadily increased from 17.5% in 2001 to 45.4% in 2015, implying

a successful landing of the crop insurance to market.


Figure 2. Insured Ratio by Year

Source: MAFRA (2016).

In 2015, 18,049 apple growers bought the crop insurance for special risk,

the greatest number except for rice (54,415 growers), followed by 9,775 insured

pear growers for special risk. The highest ratio of area insured consists of pears

(81.6% of pear production area) followed by apples (76.8%) and sweet persim-

mons (32.0%). The areas insured for field crops or greenhouse crops make up just

5%, and the insured area for rice, only 26.6%.


Table 2. Insured by Major Crops (2015)

Unit: ha, households, %

CategoryProduction

area

No. of insured farm

households

Insuredarea

Insured ratio of area

Orchard(special

risk)

Apples 20,674 18,049 15,887 76.8

Asian Pears 12,045 9,775 9,830 81.6

Tangerines 16,414 57 45 0.3

Sweet persimmons 10,478 3,109 3,348 32.0

Astringent persimmons 9,175 3,109 2,179 23.8

Orchard(multiple

risk)

Peaches 9,318 1,798 1,167 12.5

Grapes 10,974 197 78 0.7

Plums 3,976 613 272 6.8

Japanese apricots 7,017 300 211 3.0

Fieldcrops

Soy Beans 26,597 594 855 3.2

Fall Onions 13,019 311 205 1.6

Garlic 12,903 56 41 0.3

Tea 749 80 63 8.3

Red Peppers 20,828 871 236 1.1

Rice Rice 515,276 54,415 137,171 26.6

Greenhousecrops

Greenhouse watermelons 13,960 373 183 1.3

Greenhouse strawberries 6,769 1,020 385 5.7

Greenhouse melons (Chamwei) 5,345 2,213 1,191 22.3

Greenhouse tomatoes 6,928 821 310 4.5


The share of the national area covered by the crop insurance was 21.7 %

in 2015, with 122,054 farmers buying the crop insurance. The highest insured ratio

is in Jeonbuk, followed by Jeonnam and Chungnam. The insured area in Jeonnam

is 55,496 ha and the insured area ratio is 39.9%. From Table 3 and Figure 4, it

can be seen that Jeolla has the high frequency of extreme weather events and large

agricultural lands, as well as the high insured ratio of area.


Table 3. Insured by Provinces (2015)

Unit: ha, household, %

Category Production areaNo. of insured

farm householdsInsured area

Insured ratio of area

Total 854,301 122,054 185,239 21.7

Gyeonggi 83,286 4,044 5,934 7.1

Gangwon 38,216 2,684 5,487 14.4

Chungbuk 50,550 3,867 4,528 9.0

Chungnam 138,005 11,519 25,659 18.6

Jeonbuk 103,254 19,255 41,185 39.9

Jeonnam 157,297 30,699 55,496 35.3

Gyeongbuk 135,402 26,829 23,811 17.6

Gyeongnam 93,113 14,945 13,584 14.6

Jeju 23,680 3,087 2,060 8.7


Figure 3. Special Reports on Extreme Weather and Insured Ratios by Provinces (2015)

Source: MAFRA (2016) and MPSS (2016).

The agricultural insurance budget of MAFRA, which includes the crop in-

surance, increased from 16.6 billion won in 2001 to 285.3 billion won in 2015.

The agricultural insurance budget has increased 17 times and the share of the agri-

cultural insurance in the total budget of MAFRA has increased from 0.2% in 2001


to 2.0% in 2015. This shows the growing importance of the agricultural insurance

in agricultural policy.

Table 4. MAFRA budget for Agricultural Insurance (2001 to 2015)

Unit: 100 million won, %

Years Budgets of MAFRA (A) Agricultural insurance budget (B) Ratio (B/A)

2001 93,634 166 0.2

2002 102,450 247 0.2

2003 101,496 363 0.4

2004 106,907 388 0.4

2005 110,630 499 0.5

2006 118,560 998 0.8

2007 121,208 1,031 0.9

2008 124,242 1,161 0.9

2009 129,887 1,218 0.9

2010 129,888 1,319 1.0

2011 131,929 1,663 1.3

2012 136,778 1,856 1.4

2013 135,267 2,348 1.8

2014 135,344 2,701 2.0

2015 140,431 2,853 2.0

Source: Nonghyup Property & Casualty Insurance (2015), MAFRA (2016).

Thus, crop insurance is one of the most important policy-driven farmer's

risk management options and is also one of the major options of adaptation to cli-

mate change; crop insurance has also been rapidly distributed. However, so far,

previous studies on crop insurance have only estimated policy effects such as an

increase of total production or cultivation area (Young and Westcot 2000; Han

2014); recently, several studies have tried to analyze the effect of crop insurance

on production or cultivation area in farm level. Kim (2001) conducted an empirical

analysis of apple farm households and measured the welfare effect of crop in-

surance and income insurance. Comparing the welfare effects of crop insurance

and income insurance, Kim insisted that income insurance is a more desirable poli-

cy tool. Gray et al. (2004) showed that if the policies including crop insurance

were implemented, the expected profit of the farm household increased and the


profit distribution changed. On the basis of this, Gray et al. analyzed empirically

how the profit distribution of producers is changed by implementing multiple poli-

cies at the same time. They found that individual policies increase expected returns

and lower risk, but the effect of using crop insurance with other policies such as

marketing loan payment and direct payment is rather reduced. Choi, Chae, and

Yun (2010) evaluated the overall performance of crop insurance in Korea. After

evaluating the performance of crop insurance for ten years using econometric

methods, they analyzed the problems of crop insurance and suggested the political

reform measures. They found that the risk management measures such as disaster

prevention facilities have an alternative relation to insurance, and the crop in-

surance is effective in reducing instability in farm household income. Han (2014)

analyzed empirically the effect of crop insurance on farmers' production patterns

and the effect on crop market caused by changed production. Using DID

(difference-in-differences), Han estimated the effect of crop insurance on the pro-

duction by crop type and business type. From the results, Han found that the par-

ticipation rate of crop insurance affects the quantity produced positively and there-

fore may affect the market price of the item insured.

Di Falco et al. (2014) states that the demand for crop insurance increases

according to weather condition, and crop insurance is an effective measure for risk

management. However, research on whether or not crop insurance is beneficial to

the farm economy when natural disasters occur, which is the intrinsic goal of crop

insurance, has not yet been conducted.

III. Model and Economic Theory

1. Just-Pope Model

With the increasing threat of extreme weather events, the benefits of the crop in-

surance are growing as a strategy for farmers’ effective risk management.

Nevertheless, the impact of the crop insurance in reducing damage by extreme

weather has not been studied so far. Earlier studies on the crop insurance have

focused largely on the political effects, for example, changes of areas cultivated

and quantity produced.


This study focuses on the effect of the crop insurance in mitigating farm-

ers’ income damage and fluctuations due to extreme weather, which is the primary

goal of crop insurance. For this analysis, the Just-Pope model (Just and Pope,

1978) is employed.

An evaluation of extreme weather events and crop insurance was accom-

plished by a Just-Pope mean function model, which characterized the expected in-

come and variance of income per farm by different functions (denoted mean func-

tion, f and variance function or risk function, g, respectively):

y = f(X;α) + g(X;β)ε

where y represents farm income, X is the vector of independent variables, α and

β are parameter vectors, and ε~N(0, 1). After assigning functional forms to f

and g, econometric estimation of the Just-Pope mean function yielded the system-

atic effects of regressor on both expected income and the variance of income.

The Just-Pope model is a 3-step approach in which step 1 estimates in-

come from crop cultivation by OLS (Ordinary Least Squared):

y = f(X;α) + u

where u is the residual term. Furthermore, step 2 estimates the variability of in-

come from crop cultivation with the square of error terms derived at step 1 as a

dependent variable.

var(y)= E[(y-E(y))2] = E(u2)= [g(X; β)]2

Step 3 applies the estimate of the error terms derived at step 2 to remove

heteroscedasticity and then re-estimates income from crop cultivation.

y/g(X;β) = f(X;α)/g(X;β) + ε


Since income risk may be modeled as heteroskedasticity, the parameters

in the mean function cannot be efficiently estimated if the income risk is not ac-

counted for. In the empirical literature, this is done by estimating the mean func-

tion and the variance function together, primarily by a feasible generalized least

squares (FGLS) three-stage estimator (Asche and Tveteras 1999).

Using the minimizing Akaike’s information criterion, we choose re-

gressors, and the equation of an empirical panel estimation for farm income from

crop cultivation at step 1 is:

CropIncit = α0 + α1Dfulltime

it + α3Dexpert

it + α4Ageit + α5Ageit2 + α6Acreit +

α7Dins

it×Warningit + α8Dcrop

it + ci + uit

where i is farmer’s id; t is time; CropInc is income from crop cultivation;

Dfulltime is full-time farmers; Age is farmer’s age; Acre is farmer’s cultivation area;

Dins is farmer’s buying the crop insurance policy; Warning is the number of issued

weather alerts; and Dcrop is a dummy variable; c is an individual effect; u is an

idiosyncratic error.

The equation for estimating variability of income from crop cultivation at

step 2 is:

ln(CropIncit-E(CropIncit))2 = β0 + β1D

fulltimeit + β2D

expertit + β3Ageit + β4Ageit

2

+ β5Acreit + β8Dins

it×Warningit + β9Dcrop

it + ci + νit

They are consistent estimates of the variances, which are calculated as the

antilogarithm of the predictions from step 2. At step 3, using the squared root of

the variance predictions as weights the original model by weighted least squares

(WLS) is estimated (McCarl, Villavicencio and Wu 2008).

2. Probit Model

After exploring how crop insurance mitigates damage to farms caused by abnormal

weather, we examine whether the farmer’s choice of adopting crop insurance

would be affected by the extreme weathers. We model a representative risk-averse

farm household as choosing to adopt a crop insurance to maximize the expected


utility from final revenue, given the production function, land and other constraints

(Yesuf et al.). We assume that farmers are price-takers and they operate in perfect

competition market structure (Yesuf et al.). Also, assuming that the utility function

is state-independent, solving this problem would give an optimal adaptation strat-

egy undertaken by the representative farm household, given by equation:

Ait = A(Xit;α) + εit

where A is equal to 1 if household i adopted an insurance scheme at time t, and

Xit is the vector of independent variables including the farmer’s characteristics and

climatic variables. The inclusion of the extreme weather event variable in the

equation allows us to test whether the frequency of extreme weather events is a

potential complement or substitute for the decision to adopt crop insurance. α is

a vector of parameters, and εit is the error term. A risk averse household chooses

to adopt the strategy of adopting crop insurance, A = 1, over the strategy of not

adopting crop insurance, A = 0, if, and only if, the expected utility from adapta-

tion strategy is greater than the expected utility of strategy.

3. Data

In order to estimate the damage caused by climate change and examine the effect

of crop insurance on damage reduction, we use data on farm income, insurance

expenditure and general characteristics of farm households from “Farm Economic

Survey”, and data on special reports on extreme weather events from “Natural

Disaster Yearbook”. The Farm Economic Survey is a statistical survey carried out

annually by Statistics Korea on 2,800 sample farms in 560 sample locations

nationwide. The sample farm households are replaced every five years. When the

panel is replaced, the identification number of each farm is also changed. It makes

it difficult to utilize it as panel data. Hence, in this study, only the data from 2008

to 2012 that used the same sample is used for examination.1

In this study, we use only orchard farm data from the Farm Economic

1 Most recently, the survey panel has been replaced since 2013, and data are currently available

until 2016, so a data for four years is available for the latest panels. In this study, we used the

panel data for 5 years from 2008 to 2012 to get as much data as possible.


Survey, for the crop insurance was first introduced for orchard farms and the in-

sured ratio of orchard farms is relatively high. If we use the whole data set of

orchard farm for examination, it is difficult to obtain meaningful findings in the

quantitative analysis because of the huge income gap. The minimum value of in-

come from orchard farm data set is negative 120 million KRW and the maximum

value is 426 million KRW. Therefore, we use only 549 observations from the sec-

ond and third quintiles of farm income except the extremes.

Full-time farmers in the analysis account for approximately 59%, full-time

farmers account for approximately 63% of insured fruit growers and full-time

farmers account for approximately 58% of uninsured fruit growers. Specialized

fruit growers account for approximately 61% of all fruit growers, and among the

insured fruit growers, specialized fruit growers account for approximately 74%.

However, in the uninsured fruit growers, only 58% are specialized fruit growers;

as such, more full-time farmers or specialized fruit growers purchased the crop in-

surance in comparison to part-time or non-specialized fruit growers.

126 orchard farms, which make up 23% of total orchard farms, purchased

the crop insurance and 423 farms did not. Average farm income from crop culti-

vation and off-farm income were 30 million KRW and 5.11 million KRW,

respectively. For the insured farms, farm income from crop cultivation was 36 mil-

lion KRW and off-farm income was 5.2 million KRW. On the other hand, for the

uninsured farms, farm income from crop cultivation was 28 million KRW and

off-farm income was 5.1 million KRW. Insured farms had relatively high agricul-

tural income, while there was no big difference in off-farm income levels of in-

sured and uninsured farms. The average orchard area is 77a. The average orchard

area of insured farms is 87a, which is larger than that of 74a of uninsured farms.

The Natural Disaster Yearbook provides quarterly statistics on special re-

port on extreme weather events for each of the six regions in Korea (Seoul and

Gyeonggi-do; Busan and Gyeongsang-do; Gwangju and Jeolla-do; Daejeon and

Chungcheong-do; Gangwon-do; Jeju-do). Considering the fruit growing period, we

use data from the second and third quarter; in addition, the total number of special

reports issued for strong winds, heavy rainfall, heavy snowfall, drought, cold

waves, typhoons and heat waves is used. The frequency of special reports on

heavy seas, tsunamis and winds carrying yellow dusts is excluded, for these ex-

treme weather events hardly affect fruit tree growth.

The average number of special reports issued in the second and third


quarter was 116, and insured farms experienced 127 special reports. Therefore, we

can assume that there is correlation between the frequency of extreme weather

events and farmers’ decision-making on buying the crop insurance.

Table 5. Basic statistics of key variables

Variable AverageStandard deviation

Min. value

Max. value

All fruit growers (obs.=549)

D (full-time grower=1, class 1 and 2 two-job grower=0)2 0.59 0.49225 0 1

D (specialized grower=1, general · sideline· self-sufficient grower=0)3 0.61 0.48731 0 1

D (insured=1, uninsured=0) 0.23 0.42090 0 1

Age 66.05 9.95046 32 91

Cultivated area (a) 76.84 60.9259 0 447

Farm income from crop cultivation (1,000 KRW) 29937 18614.7 4970.5 221756.1

Number of special reports issued in 2nd & 3rd quarters 116.3 42.0317 49 212

Insured growers (obs.=126)

D (full-time grower=1) 0.63 0.48554 0 1

D (specialized grower=1) 0.74 0.44143 0 1

Age 65.40 9.09952 37 91

Cultivated area (a) 87.10 66.0790 4.1 446.6



Uninsured growers (obs.=423)

D (full-time grower=1) 0.58 0.49427 0 1

D (specialized grower=1) 0.58 0.49465 0 1

Age 66.25 10.1922 32 85

Cultivated area (a) 73.78 59.0434 0 294.5



Source: Kim et al. (2015).

2 In this case, full-time fruit growers refer to those who do not have family members engaged in

other work than farming for more than 30 days each year. In addition, Class 1 two-job fruit

growers refer to those whose agricultural income is more than their off-farm income, and the

Class 2 two-job fruit growers refer to those whose agricultural income is smaller than their


IV. Results

Table 6 presents the results of estimating the effects of extreme weather events

and crop insurance on farm income. According to the result of the estimation, the

correlation between the farm owners’ age and farm income is not statistically

significant. Moreover, the difference in farm income between the full-time growers

and the part-time growers is not statistically significant either. However, speci-

alized growers are expected to earn 9.07 million KRW more in farm income than

others. It shows that the scale of cultivation affects the farm income rather than

how much farmers concentrate on farming. The larger the farmland is, the more

the farm income is. As the cultivation area is increased by 10a, the farm income

is increased by about 130 thousand KRW. This implies that there exists an econo-

my of scale in fruit farming.

About the extreme weather events, the result presents that the farm in-

come is lower, as the frequency of special reports on extreme weather events is-

sued during the second and third quarter of the major crop growth period is

increasing. That is, orchards are actually experiencing economic damage due to

extreme climates such as heat waves, cold waves, typhoons, heavy rainfalls, and

droughts. It is estimated that an average of 27,680 KRW of economic damage oc-

curs to uninsured farmers every time a special report is issued. By applying the

average annual 116 special reports between 2009 and 2012, it is estimated that ex-

treme weather caused approximately 3.21 million won of damage to fruit growers.

off-farm income (Statistics Korea 2015).3 In this case, categories of specialized fruit growers, general fruit growers and sideline fruit grow-

ers comply with the farmer classification standard of Statistics Korea. The specialized fruit grow-

ers refer to those who have at least 3ha of farmland or at least 20 million won of agricultural

income. The general fruit growers refer to those who have farmland smaller than 3ha and at most

20 million won of agricultural income. The sideline fruit growers refer to those whose off-farm

income is more than agricultural income among fruit grower who have at least 30a of farmland

or at least 2 million won of agricultural income. The self-sufficient fruit growers refer to those

whose agricultural income is smaller than 2 million won among those who have farmland smaller

than 30a (Statistics Korea 2015).


Table 6. Result of estimated effect of the crop insurance for

reducing damage by extreme weather events

Variable Coef. Std. Err. P>|z|

D (full-time fruit grower=1, part-time fruit grower=0) 1430.33 905.70 0.114

D (specialized fruit grower=1, general · sideline · self-sufficient fruit grower=0)

9069.64 966.10 0.000

Age -544.68 571.04 0.340

Age2 1.50 4.61 0.745

Cultivated area (a) 134.34 12.34 0.000

Number of special reports issued in 2nd & 3rd quarters -27.68 14.95 0.064

Number of special reports issued in 2nd & 3rd quarters *D (insured=1)

20.36 6.70 0.002

D (general apple=1) -2706.86 1924.01 0.159

D (dwarf apple=1) 479.80 2406.97 0.842

D (Asian pear=1) 4479.17 1653.94 0.007

D (grape=1) 2310.13 1810.55 0.202

D (peach=1) -53.00 1491.01 0.972

D (persimmon=1) -481.36 1497.97 0.748

D (tangerine=1) -2746.62 2903.42 0.344

Constant term 45303.90 17504.55 0.010

# of obs. 549

Wald chi2 (14) 368.12

Prob>chi2 0.000


Meanwhile, through the examination, we find that the crop insurance can

mitigate the economic damage caused by extreme weather events. As Table 6

shows, the crop insurance reduces farm income loss by 20,360 KRW per special

report. Hence, compared with 27,680 KRW for an uninsured grower, an insured

fruit grower is expecting that an economic loss of only 7,320 KRW occurs every

time a special report is issued. By applying 116 times of average special reports

issued between 2009 and 2012, it is estimated that an insured grower is damaged

by 850 thousand KRW. This is approximately 2.36 million KRW smaller than

3.21 million won for the uninsured fruit growers. Considering that the average

crop insurance premium of the insured farmers in this sample is 1,040 thousand

KRW; the comprehensive effect of crop insurance on farm income is 1,322 KRW.


Figure 4. The economic effect of crop insurance on farm income


Table 7 presents the estimated result of the effect of the crop insurance

and extreme weather events on farm income volatility. According to the result of

analyzing the effects of crop insurance and extreme weather on the variability of

farm income, the variability of farm income of full-time or specialized growers is

higher than others. Furthermore, the estimation result shows that as a fruit grower

cultivates in a larger scale, the variability of farm income is also increased. This

means that the higher the farm income is, the higher the variability is.

The number of special reports on extreme weather events has a positive

correlation with the volatility of farm income. As the extreme weather events oc-

cur frequently in the fruit growing season, the volatility of farm income also

increases. That is, natural disasters such as heat wave, cold wave, heavy rain and

snowfall, typhoons and droughts not only aggravate the farm income, but also in-

crease the uncertainty of farm income. The estimation result shows that the crop

insurance reduces the volatility of farm income slightly, but the effect is not stat-

istically significant. This analysis suggests that the volatility of farm income de-

pends on the frequency of natural disaster and the scale of cultivation area rather

than the crop insurance and farm owner’s age.


Since this study is based on farm-level information, it is hard to make an

accurate estimation of the variability of farm income. To better an analysis of the

effect of crop insurance on the variability of farm income, it is necessary to carry

out estimation to the exclusion of the scale of production. Therefore, more accu-

rate results can be obtained if farm income per unit area rather than fruit grower’s

total agricultural income is analyzed in future studies.

Table 7. Result of the estimated effect of crop insurance for

reducing variability of farm income


D (full-time fruit grower=1, part-time fruit grower =0) 0.7555 0.2040 0.000

D (specialized fruit grower =1, general · sideline · self-sufficient fruit grower =0)

0.3719 0.2193 0.090

Age -0.0297 0.0948 0.754

Age2 0.0000 0.0008 0.966

Cultivated area (a) 0.0116 0.0019 0.000

Number of special reports issued in 2nd & 3rd quarters 0.0065 0.0032 0.039

Number of special reports issued in 2nd & 3rd quarters*D (insured=1)

-0.0001 0.0018 0.974

D (general apple=1) -0.2018 0.3341 0.546

D (dwarf apple=1) 0.4127 0.3245 0.203

D (Asian pear=1) 0.7183 0.2700 0.008

D (grape=1) 0.4563 0.2535 0.072

D (peach=1) -0.1941 0.2671 0.467

D (persimmon=1) -0.0402 0.2391 0.867

D (tangerine=1) 0.8925 0.4494 0.047

Constant term 16.3680 2.9443 0.000

# of obs. 549

Wald chi2 (14) 121.04

Prob>chi2 0.000


Finally, Table 8 presents the results of the probit estimation. According

to the Probit analysis, as the time passes, the incentive for crop insurance in-

creases, which is considered to be a positive publicity effect. In addition, it was


found that the incentive of insurance for the specialized fruit grower is higher than

that of the others, and the larger the cultivation area, the greater the insurance

incentive. Thus, the farmers who are engaged in large - scale farming are more

interested in risk management. As for the cultivated items, the insurance incidence

of dwarf apples and Asian pears is relatively high and grapes are low, which is

consistent with Table 2.

Table 8. Estimation result of crop insurance probit


Trend 0.3428 0.0595 0.000

Number of special reports issued in 2nd & 3rd quarters t-1 0.0044 0.0020 0.024

D (full-time fruit grower=1, part-time fruit grower=0) 0.1741 0.1495 0.244

D (specialized fruit grower=1, general · sideline · self-sufficient fruit grower=0)

0.3590 0.1576 0.023

Age 0.1068 0.0721 0.138

Age2 -0.0009 0.0006 0.124

D (general apple=1) 0.0516 0.2234 0.817

D (dwarf apple=1) 0.7278 0.2100 0.001

D (Asian pear=1) 0.3766 0.1839 0.041

D (grape=1) -0.4323 0.1829 0.018

D (peach=1) -0.2781 0.1941 0.152

D (persimmon=1) -0.0369 0.1757 0.834

D (tangerine=1) -0.4556 0.3351 0.174

Cultivated area (ha) 0.2206 0.1339 0.099

Constant term -5.7429 2.2538 0.011

# of obs. 549

Wald chi2(14) 74.67

Prob>chi2 0

The number of special reports on extreme weather events of the previous

year has a positive correlation with the farmers’ choice on crop insurance and is

statistically significant. This result is quite intuitive, indicating that farmers who

have experienced natural disasters adopt more crop insurance to hedge against bad

environmental conditions. This implies that more frequent extreme weather events

make the farmer more willing to undertake crop insurance.


V. Conclusion

In this study, we try to statistically prove the damage and correlation of natural

disasters, which are soaring due to climate change, on farm income, and to meas-

ure the economic effect of crop insurance, which is a representative option for cli-

mate change adaptation. To achieve the purpose, we employ the Just-Pope model

to perform an econometric analysis and use the data from “Farm Economic

Survey” and statistics of special reports on extreme weather.

In this study, we find that the farm income is influenced not only by the

characteristics of the farm owner such as age, or full-time/part-time farming, but

also the size of the farm, the cultivated items and the frequency of extreme

weather. As the farmland size increases, both farm income and income volatility

also increase. On the other hand, as the frequency of natural disasters increases,

farm income decreases, but income volatility continues to climb due to increased

uncertainty. In addition, it is verified that as a countermeasure against the decrease

of farm income due to meteorological disasters, crop insurance has statistically sig-

nificant effects and its impacts increase as the frequency of weather disaster

increases.

By applying 116 times the annual average special report on extreme

weather during 2009 ~ 2012, the crop insurance has economic effects of 1,230

thousand KRW per farm household. If the number of annual special reports issued

is 51 times or fewer, the expected benefits from crop insurance are lower than the

premium of crop insurance. In other words, in areas where weather conditions are

favorable and natural disasters occur less frequently, crop insurance premiums are

higher than expected economic effects of crop insurance, so it is a reasonable

choice not to join the crop insurance. For example, the insured rate of Jeju, which

has a smaller number of special reports on extreme weather than other regions,

appears to be very low.

From the result of analyzing the effect of extreme weather events on

farmers’ decision-making about crop insurance, we find that the more frequent nat-

ural disasters farmers suffer, the greater the intention to purchase crop insurance.

These findings from the study can provide several implications to re-

searchers and policy-makers. Our finding shows that extreme weather events have

an adverse impact on farm income and the damage is expected to increase with


the increasing frequency of extreme weather events due to climate change.

However, most studies on the impact of climate change on agriculture were about

how average temperature or precipitation would affect agricultural production and

land use. There are still very limited statistical data and research on extreme

weather events and agriculture. Hence, it is required to produce statistics on natu-

ral disasters that can be used in the agricultural sector and to conduct further stud-

ies on the impact of abnormal temperatures on the sector.

As it is proven that crop insurance is an effective means of adapting to

climate change, it is necessary to conduct campaigns and promotions regarding

crop insurance. As climate change is expected to become more severe in the fu-

ture, the effect of the insurance is expected to increase. In a region with a low

frequency of extreme weather, farmers do not prefer crop insurance because the

premium is higher than the crop insurance effect; it will be more effective to focus

public relations of the crop insurance and encourage farmers to be insured in the

regions where weather disasters occur frequently.

As a result of the study, crop insurance is considered to be a very effec-

tive tool for farmers' risk management due to climate change, but as Goodwin and

Smith (2013) insist, because of the high subsidy rate, there is a tendency to distort

the production market, which transfers the financial burden to the taxpayers.

Hence, there is a need for a systematic supplement that allows insurance to work

reasonably in the long term.

From the estimation results, we found crop insurance does not have a stat-

istically significant effect on farm income volatility, but considering the limitation

of farm-level data, which is significantly affected by size of business, it is neces-

sary to carry out estimation to the exclusion of the scale of production to better

an analysis of the effect of crop insurance on the variability of farm income.

Therefore, more accurate results can be obtained if farm income per unit area rath-

er than fruit grower’s total agricultural income is analyzed in future studies.

Lastly, this study also has limitations. In this study, we use sample data,

which includes only 2nd and 3rd quantiles of farms. Because of the limitation of

data, the representation of analysis results is also limited. In order to obtain a

higher level of representation and analyze responses of farmers in various farm in-

come levels, in further studies, it is necessary to update the data set and carry out

a further analysis using various models such as quantile regression or mixed level

regression.


Appendix. Criteria for issuing special reports on extreme weather

Category Warning Alert

Strong winds

Wind speed is forecasted 14m/s or faster, or instantaneous wind speed 20m/s or faster on land, but wind speed is forecasted 17m/s or faster, or instantaneous wind speed 25m/s or faster in mountainous areas.

Wind speed is forecasted 21m/s or faster, or instantaneous wind speed 26m/s or faster on land, but wind speed is forecasted 24m/s or faster, or instantaneous wind speed 30m/s or faster in mountainous areas.

Heavy seas

Wind speed faster than 14m/s is forecasted to continue for at least 3 hours or the significant wave height higher than 3m is forecasted in the sea.

Wind speed faster than 21m/s is forecasted to continue for at least 3 hours or the significant wave height higher than 5m is forecasted in the sea.

Heavy rainfalls

Rainfall more than 70mm for 6 hours or rainfall more than 110mm for 12 hours is forecasted.

Rainfall more than 110mm for 6 hours or rainfall more than 180mm for 12 hours is forecasted.

Heavy snowfalls

Fresh snow cover deeper than 5 cm for 24 hours is forecasted.

Fresh snow cover deeper than 20 cm for 24 hours is forecasted. However, fresh snow cover deeper than 30 cm for 24 hours is forecasted in mountainous areas.

DrynessEffective humidity not higher than 35% is forecasted to continue for two or more days.

Effective humidity not higher than 25% is forecasted to continue for two or more days.

Windstormtsunamis

Values greater than the effective standard value for tsunamis are forecasted by rising sea level due to complex effects including astronomical tides, windstorms, or low pressure. However, the effective standard value is specified for each region.

Values greater than the effective standard value for tsunamis are forecasted by rising sea level due to complex effects including astronomical tides, windstorms, or low pressure. However, the effective standard value is specified for each region.

EarthquakeTsunamis

Tsunamis by earthquakes with wave height of 0.5 to 1.0m are forecasted around coastal areas of Korea due to submarine earthquakes higher than scale 7.0 in the waters around the Korean Peninsula (21N~45N, 110E~145E).

Tsunamis by earthquakes with wave height greater than 1.0m are forecasted around coastal areas of Korea due to submarine earthquakes higher than scale 7.0 in the waters around the Korean Peninsula (21N~45N, 110E~145E).

Extreme colds

From October to April, one of the following occurs.➀ The lowest temperature in the morning is

forecasted to be at least 10°C lower than the previous morning, and lower than 3°C, and 3°C lower than the temperature in the previous year.

➁ The lowest temperature in the morning lower than –12°C is forecasted to continue for two or more days.

From October to April, one of the following occurs.➀ The lowest temperature in the morning is

forecasted to be at least 15°C lower than the previous morning, and lower than 3°C, and 3°C lower than the temperature in the previous year.

➁ The lowest temperature in the morning lower than –15°C is forecasted to continue for two or more days.


Source: Korea Meteorological Administration. <http://www.kma.go.kr/>. May 11, 2015.

Category Warning Alert

➂ Severe damage is forecasted due to extremely low temperature.

➂ Severe damage is forecasted in extensive areas due to extremely low temperature.

Typhoon

The forecast is that typhoons cause strong winds, heavy seas, heavy rainfalls and windstorms tsunamis to reach their warning standards.

The forecast is typhoons cause any one of the following.➀ Reach the strong winds (or heavy seas)

alert level.➁ Total rainfall more than 200mm.➂ Reach the windstorms tsunamis alert level.

Winds carrying yellow dusts

The forecast is the average ultrafine dust (PM10) concentration/hour greater than 400㎍/m3 continues for at least two hours due to winds carrying yellow dusts.

The forecast is the average ultrafine dust (PM10) concentration/hour greater than 800㎍/m3 continues for at least two hours due to winds carrying yellow dusts.

Heat waves

The forecast is that the daily highest temperature higher than 33°C continues for two or more days.

The forecast is that the daily highest temperature higher than 35°C continues for two or more days.

REFERENCES

Adams, M. R., et al. 1990. “Global climate change and US agriculture.” Nature 345:

219-224.

Asche, F. and R. Tveteras. 1999. “Modeling production risk with a two-step procedure.”

Journal of Agricultural and Resource Economics 24(2): 424-430.

Choi Kyeong-hwan, Chea Gwang-seok and Yoon Byeong-seok. 2010. The Performance and

Tasks of Crop Insurance. Research Report R615. Korea Rural Economic Institute.

Di Falco, S., et al. 2014. “Crop Insurance as a Strategy for Adaptation to Climate Change.”

Journal of Agricultural Economics 65(2): 485-504. doi:10.1111/1477-9552.12053

Gammans, M., P. Mérel, and A. Ortiz-Bobea. 2017. “Negative impacts of climate change

on cereal yields: statistical evidence from France.” Environmental Research Letters

12:1-9. doi:10.1088/1748-9326/aa6b0c

Goodwin, B. K. and V. H. Smith. 2013. “What harm is done by subsidizing crop

insurance?” American Journal of Agricultural Economics 95(2): 489-497.

doi:10.1093/ajae/aas092

Gray, A.W. et al. 2004. “How U.S. Farm Programs and Crop Revenue Insurance Affect

Returns to Farm Land.” Review of Agricultural Economics 26(2): 238-253.

doi:10.1111/j.1467-9353.2004.00173.x

Han Sung-min. 2014. “An Empirical Analysis on the Production and Price Effect by

Agricultural Disaster Insurance.” Korea Development Studies 36(4): 135-169.

doi:10.23895/kdijep.2014.36.4.135

IPCC. 2014. Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II

and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate

Change [Core Writing Team, R.K. Pachauri and L.A. Meyer (eds.)]. IPCC, Geneva,

Switzerland.

Just, R. E. and R. D. Pope. 1978. “Stochastic specification of production functions and economic

implications.” Journal of Econometrics 7(1): 67-86. doi:10.1016/0304-4076(78)90006-4

Kim Chang-gil, et al. 2009. Impacts and Countermeasures of Climate Change in Korean

Agriculture. Research Report R593. Korea Rural Economic Institute.

Kim Chang-gil, et al. 2012. Impacts and Countermeasures of Climate Change on Food

Supply in Korea. Research Report R663. Korea Rural Economic Institute.

Kim Chang-gil, et al. 2015. Economic Analysis of Adaptation Measures to Climate Change

in the Agricultural Sector. Research Report R749. Korea Rural Economic Institute.

Kim Pan-ki et al. 2010. “Damage of seedlings caused by weather disasters.” Journal of

Climate Research 5(2): 148-616.

Kim Tae-kyun. 2001. “Producer Preferences and Welfare Effects of Crop Insurance

Schemes.” Korean Journal of Agricultural Economics 42(2): 33-49.


Korea Meteorological Administration website. <http://www.kma.go.kr/>. (May. 11. 2015;

Nov. 11, 2015.)

Korea Meteorological Administration. 2012. Report of Climate Change Outlook in Korean

Peninsula.

Kwon Won-tae. 2012. “Climate Change Scenario and Use thereof for Agriculture.”

Prospect of Agriculture 2012(II). EO4-2012. Korea Rural Economic Institute. pp.

997-1026.

Lee Bong-ho et al. 1991. “The impacts of weather disaster on oilseed crops and

countermeasures reducing the damage.” Korean journal of crop science 36(5): 445-458.

Lobell, D. B., A. Sibley and J. I. Ortiz-Monasterio. 2012. “Extreme heat effects on wheat

senescence in India.” Nature Climate Change 2: 186–189. doi:10.1038/nclimate1356

McCarl, B. A., X. Villavicencio and X. Wu. 2008. “Climate change and future analysis: is

stationarity dying?” American Journal of Agricultural Economics 90(5): 1241-1247.

doi:10.1111/j.1467-8276.2008.01211.x

Ministry of Agriculture, Food and Rural Affairs. 2016. 2015 Crop Insurance Yearbook.

Ministry of Public Safety and Security. Annual Natural Disaster Report. Each year.

National Center for Agro Meteorology website. <http://www.ncam.kr/>.

Nonghyup Property & Casualty Insurance. 2015. Data about public relations of crop

insurance scheme.

Schlenker, W. and M. J. Roberts. 2009. “Nonlinear temperature effects indicate severe

damages to U.S. crop yields under climate change.” PNAS 106(37): 15594-15598.

doi:10.1073/pnas.0906865106

Statistics Korea website. <http://kosis.kr/>.

Statistics Korea. Economic Survey of Farmers. Each Year.

Welch, J. R. et al. 2010. “Rice yields in tropical/subtropical Asia exhibit large but

opposing sensitivities to minimum and maximum temperature.” PNAS 107(33):

14562-14567. doi:10.1073/pnas.1001222107

Young, C. E. and P. C. Westcot. 2000. “How Decoupled Is U.S. Agricultural Support for

Major Crops?” American Journal of Agricultural Economics 82: 762–767.

doi:10.1111/0002-9092.00076


Period of Review: Oct. 16～Dec. 15, 2017


THE PRODUCTION AND DISSEMINATION OFAGRICULTURAL KNOWLEDGE AT U.S. RESEARCHUNIVERSITIES: THE ROLE AND MISSION OFLAND-GRANT UNIVERSITIES

LEE YOO HWAN*

GREGORY D. GRAFF**

Keywords

agricultural R&D, Land-Grant university, knowledge production function,

public domain, technology transfer and dissemination, polynomial dis-

tributed lags

Abstract

This paper analyzes food and agriculturally-related knowledge pro-

duction and transfer for 114 top-tier U.S. research universities from 1993 to

2015, to understand the role of the Land-Grant universities in promoting

commercial innovation and regional economic development in this

sector. We utilize two empirical methods: (1) a panel analysis of the

knowledge production function (KPF) for research productivity and (2) an

analysis of variance (ANOVA) for the role of the Land-Grant universities in

such knowledge production. Output of research publications exhibits de-

creasing returns to scale for all sub-fields, but cost advantages and

mean research (gestation) lags vary by sub-field. The mean number of

research publications by the Land-Grant universities is much higher than

non Land-Grant universities, especially in the Central region of the U.S.

These results demonstrate how specialization by Land-Grant universities in

agricultural R&D contributes to commercial innovation within a diffuse

yet regionalized industry. Moreover, the main context and results of this

paper would suggest some important implications to the study of the sys-

tem of food and agricultural R&D and commercial innovations in Korea.

* Chief Research Associate, Business Consulting Research Center, Department of Business

Consulting, Daejeon University, Daejeon, South Korea. Corresponding author.

e-mail: [email protected]** Associate Professor, Department of Agricultural and Resource Economics, Colorado State

University, Fort Collins, CO, U.S.


I. Introduction and Background

In the global knowledge economy, universities play a significant role in knowledge

creation and transfer. Today, most research universities are engaged in industrial

innovation and regional economic development, leading to positive social returns

(Jaffe 1989; Mansfield 1991, 1995; Audretsch and Feldman 1996; Adams and

Griliches 1998; Cowan 2005). Following Cohen, Nelson, and Walsh (2002), uni-

versity research is identified to be of at least moderate importance to R&D within

a wide range of industries, including both high technology and more traditional.

Moreover, studies have measured the contributions of academic research to in-

dustrial innovation and the introduction of new products and processes through

different knowledge dissemination channels and different modes of impact

(Mansfield 1991 and 1995; Henderson et al. 1998; Agrawal and Henderson 2002).

In the United States in 2013, university research accounted for roughly 50

percent of total basic research, and universities make up the second largest per-

former of research and development (R&D) after industry, accounting for $64.7

billion of the total $456 billion, or 14 percent, of R&D performed (NSF 2016).

By far the largest source of funding for university performed R&D was the U.S.

federal government; while the share of university R&D expenditures funded by the

business sector accounted for just $3.5 billion or 5.4 percent. In the agricultural

and food industries in the U.S. total expenditures on R&D in 2013 was $16.3

billion. Of that total, public research institutions―including the system of the

Land-Grant universities, together with the state agricultural experiment stations

(SAES) and Cooperative Extension institutions―accounted for almost 30 percent

of the total agricultural and food R&D expenditures, over twice the level com-

pared to the economy as a whole (USDA ERS 2016; Clancy et al. 2016).

In the U.S., the Land-Grant universities have long focused on providing

agricultural R&D and, in so doing, have served as a source of ideas for commer-

cial innovation and regional economic development. Historically, the Land-Grant

system was very closely associated with the development of the U.S. public higher

education system driven by several landmark policy changes, including the Morrill

Land-Grant Act of 1862 and 1890, the Hatch Act of 1887, and the Smith-Lever

Act of 1914.1 Following these landmark policies, the Land-Grant universities have

generally come to embrace three interwoven missions in education, research, and

The Production and Dissemination of Agricultural Knowledge at U.S. Research Universities 65

outreach. The Morrill Land-Grant Acts of 1862 and 1890 provided funds, through

the granting of land assets by the federal government, to each state of the United

States, according to the act:

“for the endowment, support, and maintenance of at least one college

where the leading object shall be, without excluding other scientific and classical

studies, and including military tactics, to teach such branches of learning as are

related to agriculture and the mechanic arts, … in order to promote the liberal

and practical education of the industrial classes in the several pursuits and pro-

fessions in life.”

The Hatch Act of 1887 created and funded state agricultural experimental

stations for each state, which were often established as the research division of

the state’s new Land-Grant college or university, to conduct R&D specific for that

state’s agricultural industry and rural economy. Finally, the Smith-Lever Act of

1914 created and funded the Cooperative Extension Service) as an integral part of

the states’ land-grant college or university, yet funded and managed cooperatively

with the state government, to provide information and education regarding agri-

culture throughout the state’s local communities.

Today, the public Land-Grant universities make up the largest share of the

top-tier research universities in the U.S. In this analysis, we will see that, of the

114 universities classified by the Carnegie Classification of Institutions of Higher

Education as “R1 research universities”, 41 (or 36 percent) are Land-Grant

universities. Altogether, 70 percent of these top-tier universities are public uni-

versities, yet the non Land-Grant public universities make up 34 percent of the

total. Private universities make up just 30 percent of the total. Moreover, still to-

day, the Land-Grant universities continue to maintain education, research, and out-

reach programs in areas related to agricultural sciences and engineering (a.k.a. the

“mechanical arts”). And, in each of the states of the United States today, the

Land-Grant university’s production of new scientific knowledge and transfer of

new technology to industry continue to be important factors spurring the creation

of agricultural innovations, driving investment and engagement by the private sec-

tor, and providing opportunities for rural economic development.

The production of such economically-useful knowledge can be measured in

1 More information: https://nifa.usda.gov/history


several different ways. This makes it possible to analyze the extent to which differ-

ent types of knowledge dissemination channels are utilized by universities. These

can include channels such as the public domain, university-industry collaboration,

patent licensing, and venture creation (Lee and Graff 2017). Since universities and

public research institutions are generally recognized as non profit organizations,

most results of university research are released into the public domain, via pub-

lications and open access of research results, given that the role of university is

largely to serve public purposes. Recently, however, the emergence of the

“entrepreneurial” university characterized by the commercial utilization of university

research results have induced new processes or modes of university R&D and dis-

semination activities, which are based on the intellectual property rights (IPRs)2 and

collaborative research projects conducted jointly with industry sponsors and part-

ners, expanding the mission and role of the university3 (Etzkowitz 2003).

Although both formal IP-mediated tech transfer activities and more in-

formal industry collaboration and extension activities are used to disseminate

knowledge outputs from the university, the public domain-oriented knowledge out-

puts—such as published journal articles, conference proceedings, book chapters

and reviews, public lectures, and even degree awards—are still the major knowl-

edge outputs of any university. In fact, the magnitude and size of knowledge out-

puts produced and disseminated via the public domain are significantly greater

than the knowledge outputs produced and disseminated via the traditional industry

collaboration and the formal IP-mediated tech transfer activities. Because of the

nature of knowledge, the different types of knowledge outputs are closely inter-

twined and have complex complementary and substitute relationships depending

upon the context (Agrawal and Henderson 2002; Payne and Siow 2003;

Bonaccorsi et al. 2006; Thursby and Thursby 2011; Folz et al. 2007; Lee and

Graff 2017). Thus, the public domain-oriented knowledge outputs should continue

to be considered the primary output of the university and likely to affect the pro-

duction of the other types of knowledge outputs, even though the extent and direc-

2 By the passage of the Bayh-Dole Act of 1980, the U.S. university inventors have been permitted

to possess the ownership of their patented inventions, which made with federal funding.

Moreover, due to the increase in university-industry collaborations, university inventors have pos-

sessed the co-ownership of private funded inventions, and become co-founders of new startup

companies.3 The outreach mission of economic and social development, as well as the mission of teaching

and research.


tion of causality may not be fully resolved.4

The geographic proximity between university and industry is also important

for university R&D and dissemination activities (Jaffe 1989; Jaffe et al. 1993;

Audretsch and Feldman 1996; Anselin et al. 2000; Adams 2002; Boschma 2005;

Ponds 2010; Buenstorf and Schacht 2013). As demonstrated by the history of the

Land-Grant university system in the U.S., the university's outreach mission (regional

economic development) is intimately linked with the geographic distance, with more

proximate industry likely to have a cost advantage in absorbing and using new

knowledge from the university. Generally, shorter distances mean lower transaction

costs. However, this rule of proximity is not applicable in every circumstance, and

in fact it may vary systematically across the different types of knowledge

dissemination. According to Jaffe (1989), geographic proximity is unimportant if

the knowledge channel is based on publications, but geographic proximity is im-

portant if the channel is based on informal exchange. Moreover, due to the im-

provement of telecommunication and information technologies today, some of the

underlying mechanism of knowledge spillovers between university and industry may

not be as constrained by regional proximity today as it was in the past.

However, within the context of the formation of industry clusters, wherein

sets of interrelated private sector firms and associated public institutions within

particular fields of industry or technologies tend to aggregate in the same region,

geographic proximity does appear to remain important. In agriculture, following

Graff et al. (2014), innovation clusters in the food and agriculture-related in-

dustries can be shaped by the structure of the food and agricultural value chain

within a state, which in turn is affected by the relationships between the region’s

industry and public research institutions.

The main purpose of this paper is to analyze the system of agricultur-

ally-related knowledge production and transfer activities across the 114 top U.S.

research universities, over more than two decades, from 1993 to 2015. This paper

introduces and explores several empirical specifications of a more general model

of the knowledge production function (KPF), utilizing a detailed dataset of uni-

versity knowledge inputs and outputs, including life science research expenditures

4 According to Agrawal and Henderson (2002), the patent volume does not predict the volume of

publications and vice versa, but patent volume seems to be positively correlated with the paper

citations. They also point out that finding the correlation between patenting and publication activ-

ities is difficult but it is an important and meaningful question.


and several different categories of food and agriculture-related research pub-

lications, respectively. The main research questions of this study concern the sys-

tematic relationship between research inputs and outputs by universities in agri-

culturally related fields of research: How does the productivity as well as the tim-

ing and lag structure of knowledge production vary across different ag-related re-

search fields? How does output of agriculturally-related knowledge differ for the

Land-Grant universities, which have historically specialized in these fields, and all

other universities? To what extent do such differences seem to be related to the

geographic location of Land-Grant universities and the regional profile of the agri-

cultural and food industries? These questions have important implications for

knowledge output, innovation and productivity growth, and regional economic de-

velopment, particularly for those regions that are more dependent upon or speci-

alized in agricultural and food production. Finally, we explore how these questions

and the results of this analysis apply to food and agriculture-related research and

innovation in South Korea.

The rest of this paper is organized into four sections. Section II describes

a technique for estimating the knowledge production function involving panel

count data within a polynomial distributed lag scheme using a novel research in-

put-output data set. Section III shows the results for the empirical tests by the 114

top tier research universities in the United States from 1993 to 2015. Then from

an analysis of variance (ANOVA), we look at the relationship between the geo-

graphic location of Land-Grant universities and the dissemination of new knowl-

edge in different ag related research fields via research publications. Section IV

discusses important implications for food and agriculturally-related research and

innovation in Korea based on the results of this study. Section V summarizes the

main conclusions and insights of this study.


II. Model Framework and Data

1. Empirical model framework

The knowledge production function (KPF) is based on the concept of the neo-clas-

sical production function, and it is useful for describing the unobservable, yet val-

uable, additions that research contributes to the stock of knowledge capital.

However, the production of knowledge differs from that of normal economic goods

in two major ways. First, the profit maximization problem is rarely applied to the

knowledge production problem due to the lack of a stable, appropriated market

price of research outputs. Second, the units or increments of actual or “underlying”

economically valuable technological knowledge are often unobservable. According

to Pardey (1989), empirical studies of knowledge production is limited in large part

because of the difficulties of obtaining suitable indicators of research outputs.

Nevertheless, the literature demonstrates that we can be confident that there exists

a systematic input-output relationship between research inputs and new knowledge

outputs as measured by a number of proxy variables. In this study, we estimate

three different specifications of the knowledge production function (KPF) in which

output is measured by the count of research publications: (1) a log-log model with

an unrestricted PDL scheme, (2) a negative binomial MLE model with unrestricted

PDL scheme, and a negative binomial MLE model with restricted PDL scheme.

The initial idea and functional form of the knowledge production was in-

troduced by Griliches (1979) and Pakes and Griliches (1980 and 1984). Specific

to agriculture, Parday (1989) adapted the knowledge production function (KPF) to

48 state Land-Grant universities and their state agricultural experimental stations

(SAESs) over 13 years. In classical production theory, there are various functional

forms for representing the relationship between inputs and outputs, such as log lin-

ear, quadratic, Cobb-Douglas, CES, transcendental, von Liebig, Mitscherlich-Baule,

translog, etc. However, most previous empirical studies of knowledge production

have utilized one of most common, the Cobb-Douglas production function, be-

cause of its amenability to econometric techniques but also because of its suitable

representation of some of the inherent characteristics of knowledge production.

Equation (1) represents the log-linear form of the Cobb-Douglas or the log-log

KPF model5 adapted from Griliches (1979) and from Pardey (1989):


(1) ln

ln

where Y is the logarithm of the university knowledge outputs6 and R is the loga-

rithm of the lagged time period of the research expenditures for re-

search university i at time t. ε is an independent and identically distributed panel

disturbance term.7

Before developing the main model, we need to outline two major issues:

(1) the count data dependent variable and (2) the lag scheme of the relationship

between the input of past research expenditures and the output of research

publications. First, most university research outputs are measured by count data8,

such as the number of publications per year, the number of degree awards per

year, the number of patent applications and issued patents per year, etc. So, we

attempt to use negative binomial maximum likelihood estimation (MLE) models

as the countable dependant variable (see Hausman et al. 1984; Hall et al. 1986)

and the log-likelihood function is equation (2) below:

(2) ln

ln

5 We initially tested the model specification errors, considering such issues as omitted relevant vari-

ables and included irrelevant variables, using a bottom-up approaches. The preliminary results in-

dicated that some important variables, such as dummy proxies for a Land-Grant university and

the geographic region, could not be included in the KPF because of multicollinearity with the

fixed effect model in the panel data analysis. Instead, we adopt an analysis of variance (ANOVA)

test using these variables. (See details in the Results section). Since there exist data limitations

at the institutional level, we could not include some potentially relevant variables such as the

number of authors per paper, full-time equivalents (FTEs), etc.6 As we mentioned before, generally, there are four different types of university knowledge outputs,

including: publication or release into the public domain, public-private collaborations, patent-

ing/licensing, and venture creation. However, in this study, we use only research publications,

which represent, by-in-large, the public domain mechanism. (See details in the Data section.)7 It is comprised of group-variant but time-invariant error term, , and both group and time-variant

as idiosyncratic error term, . We assume that they are mean zero, homoscedastic, and exhibit

no serial correlation.8 A type of data in which the observations can take only the non negative integer values {0, 1,

2, 3, …} and where these integers arise from counting rather than ranking.


where r is the dispersion parameter and Γ is the gamma function for the negative

binomial MLE. is the mean of the negative binomial MLE9, defined as an un-

known parameter.

Since the structure of the KPF model is based on the relationship between

research outputs and past research expenditures (Pakes and Griliches 1980 and

1984), so a number of previous studies of the KPF model (see above) have adopt-

ed a finite and ad hoc distributed lag model. However, following Crespi and

Geuna (2008) and Lee and Graff (2015 and 2017)10, the relationship between re-

search outputs and past or lagged research expenditures is more likely to follow

a polynomial pattern, rather than a geometrically declining (a.k.a. Koyck) pattern.

Thus, we adopt a polynomial distributed lag (PDL) structure for the main lag

scheme of the research expenditure inputs.

Adapted to equation (1), the PDL model assumes that β can be estimated

by a p=0,1,2,..,m degree of polynomial and a j=0,…,k lag length, see equation (3).

The corresponding equation of m-degree and k-lag length of the unrestricted PDL

model is equation (4).

(3) ∙ ∙ ⋯ ∙

∙

where ω is a constructed slope coefficient.

(4)

and Z is a constructed variable,

∙

∙ ⋯

∙ .

9 exp′ exp

10 Crespi and Geuna (2008) introduce the use of the polynomial distributed lag scheme in the

knowledge production function context, but they only adopt a linear functional form rather than

the count data form of analysis. Lee and Graff (2015 and 2017) combine the count data model

with the polynomial distributed lag scheme.


The and values from equations (3) and (4) are not the true slope co-

efficients on the original variables. Rather, first, equation (4) is estimated by OLS,

and then the true values of the slope coefficients can are recovered by the fol-

lowing set of equations (5):

(5)

⋯

⋯

⋮

⋯

where the are generated from the OLS procedure, and the are the estimated

slope coefficients (For more details, see Gujarati, 2004: 687-691).

The unrestricted PDL model has no a priori restrictions, but a restricted

PDL model can be limited by restricting the k+1st and greater lagged coefficients

to equal zero, which is called a far endpoint restriction. This assumes that un-

observable inputs made beyond the kth lag year no longer impact current research

outputs, following equation (6):

(6) ⋯

Equation (6) is substituted into equation (4) and then the model can be

estimated by standard OLS procedures. Similarly, the true slope coefficients of the

restricted PDL model can be recovered by equation (5), as described above.

2. Data

Table 1 provides summary statistics of these research input and output variables

for the 114 U.S. research universities classified as Doctoral Universities-Highest

Research Activity in the Carnegie Classification of Institutions of Higher

Education11, also known as “R1 research universities” (For a list of the uni-

versities and their rankings, see Appendix 1). The dataset of the research input and

output was mainly collected from open resources.

11 Except City University of New York (CUNY) Graduate School and University Center.


TABLE 1. Summary statistics of all research input and output variables at the 114 U.S.

research universities, 1993-2015

Mean S.D. Min Max Group Obs

Research expenditures

Life science research expenditures*, million $ 158.58 163.79 0.62 870.52 114 2,622

Ag & food related research publications**

All fields 136.83 177.52 0.00 1,129.00 114 2,622

Dairy & animal sciences 13.21 32.37 0.00 304.00 114 2,622

Biotechnology & applied microbiology 37.17 39.27 0.00 398.00 114 2,622

Crop, horticulture, & soil sciences 39.48 66.31 0.00 554.00 114 2,622

Food science and technology 32.29 44.59 0.00 273.00 114 2,622

Regional dummies***

Pacific (16) 0.14 0.35 0.00 1.00 114 2,622

Mountain (6) 0.05 0.22 0.00 1.00 114 2,622

Northern Plains (3) 0.03 0.16 0.00 1.00 114 2,622

Southern Plains (13) 0.11 0.32 0.00 1.00 114 2,622

Central (20) 0.18 0.38 0.00 1.00 114 2,622

Southeast (25) 0.22 0.41 0.00 1.00 114 2,622

Northeast (31) 0.27 0.45 0.00 1.00 114 2,622

Institutional dummies

Land-Grant public university (41) 0.36 0.48 0.00 1.00 114 2,622

Non Land-Grant public university (39) 0.34 0.47 0.00 1.00 114 2,622

Non Land-Grant private university (34) 0.29 0.46 0.00 1.00 114 2,622

* Three sub-fields: agricultural sciences, medical sciences, and biological sciences;

** Included in published journal articles, book chapters & reviews, conference paper &

proceedings, and scientific letters;

*** See Alston et al. (2010) page 283; Parentheses are the number of universities.

First, the data of university R&D expenditures classified as life sciences

as an input was obtained from the Higher Education Research and Development

(HERD) Survey of the National Science Foundation (NSF)’s National Center for

Science and Engineering Statistics (NCSES) from 1993 to 2015. The life science

research expenditures reported by NSF12 include three sub-fields: agricultural sci-

12 Because of the limited data reporting in the National Science Foundation, 18 universities’ life

science research expenditures between 1993 and 1997 were not reported. So, the missing data


ences, biological sciences, and medical sciences. In 2015, the life science research

expenditures for the R1 research universities accounted for $28 billion or almost

55 percent of total research expenditures. Within the R1 universities, the life sci-

ence research expenditures in the Land-Grant universities in 2015 was $10 billion.

The count data of annual research publications as an output was collected

from queries for the university affiliation of authors in the ISI Web of Science

(Thomson Reuters), covering 1993-2015. Research publications, in which the cate-

gories are characterized by published journal articles, book chapters & reviews,

conference paper & proceedings, and scientific letters, in agriculture and food re-

lated research fields are based on the Web of Science’s field categories, and in-

clude the following: agriculture dairy animal science, agricultural economic policy,

agricultural engineering, agronomy, biotechnology applied microbiology (including

bioenergy), crop & horticulture, food science technology, nutrition dietetics, plant

science, soil science, agricultural multidisciplinary.13 Since for some universities

there are very few observations in some of these Web of Science field categories,

the fields can be merged and classified according to five different research field

groups as well as the combination of all agriculturally related fields, as follows:

(1) all fields, (2) dairy and animal science, (3) biotechnology and applied micro-

biology, (4) crop, horticulture, and soil science, and (5) food science and

technology.

Finally, universities can be identified as falling within one of seven differ-

ent multi-state regions of the United States which are chosen, in part, because of

broad similarities in agricultural conditions and thus the profile of agricultural in-

dustry within each region.14 Within the 114 sample universities, 41 are Land-Grant

universities, which accounts for 36 percent of the total (For a list of the R1

Land-Grant universities, by region, see Appendix 3).

were “back cast” for that earlier period, based on those institutions’ total research expenditures

for those years, according to the average share that life sciences expenditures represented of total

research expenditures as observed for those 18 universities during the middle period of

1998-2002.13 Excluded in the natural resource related sub-fields such as forestry, fisheries, etc.14 See more detail information in Alston et al. (2010) p.283, but we treat Hawaii as Pacific region.


III. Results

There are two parts to the regression analysis conducted. The first is a panel data

analysis of the agricultural knowledge production function (KPF) for the output of

research publications in each of the food and agriculture-related research field

groups, essentially estimating the system of the universities’ production of knowl-

edge that is disseminated via the public domain. The second part involves the

analysis of variance (ANOVA), which provides a dummy variable test for the pro-

ductivity of knowledge production across the different food and agriculture-related

research field groups. The main objective of this second analysis is to ascertain

how the role of Land-Grant universities affects the production of food and ag re-

lated research publications in the various field groupings across the different geo-

graphic regions of the U.S.

1. Agricultural knowledge production function

The knowledge production function (KPF) can be defined as the technical relation-

ship between research inputs and outputs. In this analysis, the major knowledge

output metric being utilized is the count of food and ag related research pub-

lications and the main input measure is annual life sciences research expenditures

for 114 U.S. research universities from 1993 to 2015. In this section, we estimate

three different agricultural KPF models: (1) a log-log model with an unrestricted

polynomial distributed lag (PDL) scheme, (2) a negative binomial MLE model

with an unrestricted PDL scheme, and (3) a negative binomial MLE model with

a restricted PDL scheme. All three models assume a group fixed effect, prelimi-

narily ascertained by the Hausman test15, and the optimal degree of the lag struc-

ture’s polynomial and the lag length in each is chosen based on the information

criteria.16

15 ′

16 The Akaike information criterion (AIC), AIC=-2×ln(Likelihood Function)+2×P, and the Schwarz

Bayesian information criterion (SBIC), SBIC=-2×ln(Likelihood Function)+ln(N)×p , where p is

number of parameters estimated and N is number of observations. The model with the smaller

value of the information criterion has a better goodness of fit.


In selecting these three models, we initially tested an ad hoc distributed

lag scheme of the life science research expenditures, rather than a PDL scheme,

across the all three agricultural KPF models. Preliminary test results indicated that,

using the ad hoc distributed lag scheme, almost all slope coefficients on all lagged

years’ research expenditures are statistically insignificant. The only significant co-

efficients were found in the first and last lagged time periods. These results are

similar to those found in previous studies (Pakes and Griliches 1980 and 1984;

Hausman et al. 1984; Hall et al. 1986; Parday 1989). Subsequent analyses have

established that the slope coefficients of the KPF follow a polynomial pattern, so

an ad hoc distributed lag scheme causes significant model misspecifications

(Crespi and Geuna 2008; Lee and Graff 2015 and 2017). Therefore, in this analy-

sis, the PDL is the only lag scheme utilized in the KPF estimations.

1.1. Log-log KPF model

Table 2 shows the results of the panel estimation of the log-log KPF model with

an unrestricted PDL scheme of life science research expenditures across the differ-

ent food and ag related research field groups. There are five different KPF models:

model 1 counts all research publications for all fields; model 2 estimates the KPF

for just the dairy and animal science publications; model 3 estimates the KPF for

biotechnology and applied microbiology publications; model 4, for crop, plant,

horticulture, and soil science publications; and, model 5, for the food science and

technology publications.

All five models assume a group fixed effect and follow a second degree

polynomial with six lagged years of life sciences research expenditures within the

PDL structure. Since all are log-log models, each slope coefficient indicates a mar-

ginal effect or a marginal product of the knowledge production function in the

short-run. Most of the slope coefficients in all five models are statistically sig-

nificant, except the coefficients on the middle range of lagged research ex-

penditures in model 5, from years 2 to 4. The slope coefficients of each model

also represent elasticity of output, which is defined as the percent change in cur-

rent research publications (the output) due to a one percent change in life science

research expenditures (the input).

The slope coefficients on lagged research expenditures for models 1, 3,

and 5, (all fields, biotechnology, and food science, respectively), follow U-shape


or convex patterns whereas for models 2 and 4 (animal science, and crop science,

respectively) follow inverted U-shape or concave patterns. We can interpret this

to mean that research expenditures have a maximum impact on dairy & animal

science publications in the second year and a maximum impact on crop, horti-

culture, and soil science related research publications in the fourth year.

In Table 2, the sum of the lags represents a long-run or total impact of

past and current research expenditures on current year publications. It measures

how the research publications at university i change in response to changes in life

science research expenditures in the long-run. All models have statistical sig-

nificance at the 1% level, except for model 2, the dairy and animal science field,

which has statistical significance at the 10% level. Moreover, the sum of the lags

also represents returns to scale of the knowledge production. If the sum of all

slope coefficients is less than one, this indicates decreasing returns to scale, when

the sum of the lags is equal to one, it indicates constant returns to scale, and a

sum greater than one indicates increasing returns to scale.

TABLE 2. Estimates of the log-log model with an unrestricted polynomial distributed lag

(PDL) scheme across the different agriculture-related research fields at 114 research uni-

versities, 1993-2015

Dependent variable: Research publications (log-log)

All fieldsDairy &

animal science

Biotechnology & applied

microbiology

Crop, horticulture, &

soil science

Food science & technology

[1] [2] [3] [4] [5]

Group fixed effect Yes Yes Yes Yes Yes

Degree of PDL1 2 2 2 2 2

Expenditure_t-00.11964*** -0.08586 0.13656*** 0.01109 0.14205**

(0.02980) (0.16257) (0.03502) (0.03763) (0.05858)

_t-10.09641*** 0.24268*** 0.11938*** 0.05453*** 0.07121***

(0.01075) (0.05809) (0.01263) (0.01330) (0.02073)

_t-20.08292*** 0.39606*** 0.11041*** 0.08361*** 0.03112

(0.01606) (0.08049) (0.01883) (0.02095) (0.02982)

_t-30.07916*** 0.37427*** 0.10966*** 0.09833*** 0.02178

(0.01989) (0.10211) (0.02334) (0.02582) (0.03773)


Notes: 1. The number is the degree of polynomial; 2. Akaike Information Criterion;

3. Schwarz Bayesian Information Criterion; Parentheses are standard errors;

*** at 1%, ** at 5%, and * at 10% level of statistical significance.

The results of all five models suggest decreasing returns to scale, but with

rather different magnitudes of the coefficients. The field of biotechnology and ap-

plied microbiology (model 3) has the highest value, at 0.8828 for the sum of esti-

mated coefficients at 1 percent level of statistical significance. The field of dairy

and animal science (model 2) has the smallest value, at 0.1675 for the sum of esti-

mated coefficients at 10 percent level of statistically significance. This result in-

dicates that the production of publications in biotechnology and applied micro-

biology has greater cost advantages than the production of publications in other

research fields over the long run.

Dependent variable: Research publications (log-log)

All fieldsDairy &

animal science


microbiology

Crop, horticulture, &

soil science

Food science & technology

[1] [2] [3] [4] [5]

_t-40.08515*** 0.17731** 0.11714*** 0.09870*** 0.04320

(0.01527) (0.08041) (0.01792) (0.01954) (0.02964)

_t-50.10088*** -0.19482*** 0.13284*** 0.08472*** 0.09537***

(0.00991) (0.04901) (0.01161) (0.01281) (0.01828)

_t-60.12635*** -0.74212*** 0.15676*** 0.05638 0.17830***

(0.03070) (0.15033) (0.03599) (0.04062) (0.05570)

Sum of the lags0.69052*** 0.16752* 0.88274*** 0.48735*** 0.58303***

(0.01969) (0.10093) (0.02296) (0.02743) (0.03327)

Mean lag 3.04534 3.74749 3.10674 3.43364 3.29012

Constant1.18330*** 2.74698*** -0.62182*** 1.59235*** 0.68210***

(0.08909) (0.50200) (0.10476) (0.12375) (0.16238)

AIC2 753.88 889.89 1,322.64 418.97 1,184.57

SBIC3 776.14 906.83 1,344.86 438.92 1,205.39

Log-likelihood -372.94 -440.94 -657.32 -205.48 -588.29

Observations 1,930 510 1,911 1,083 1,346

Groups 114 114 114 114 114

(continued)


The mean lag17 is a weighted average of coefficient values over time and

thus represents the average “gestation period” between a research project’s in-

ception and completion (see Pakes and Griliches 1980 and 1984; Pardey 1989;

Crespi and Geuna 2008). However, in practice, actual expenditures generally begin

some time after project inception because of the time involved in applying for and

receiving funding (Lee and Graff 2017). The results from model 1 tell us that, for

all fields, on average, a university faculty member or research team18 spends 3.04

years generating a research publication: similarly, in dairy & animal science

(model 2), the mean lag is 3.74 years, in biotechnology and microbiology (model

3), it is 3.10 years; in crop, horticulture, and soil science (model 4), it is 3.43

years; and in food science (model 5) it is 3.29 years. Thus, the production of re-

search publications in dairy and animal science has a relatively longer average lag

between a research project’s inception and completion, while in biotechnology and

microbiology, research publications have a relatively shorter average lag.

1.2. Negative binomial MLE of KPF models

Similar to the log-linear KPF model, the panel estimation of the negative binomial

maximum likelihood estimation (MLE) of the KPF model PDL schemes of life

science research expenditures, but in this case with both unrestricted and restricted

versions, across each of the different research field groups (Table 3).

17 As calculated by this formula,

∙

. For more details, see Gujarati (2004), pg.

668.18 According to Wuchty et al. (2007), the traditional university ethos emphasized the role of in-

dividual genius in scientific discovery, but in recent developments, most academic research has

shifted from an individual model to a teamwork model.

TABLE3.EstimatesofthenegativebinomialMLEwiththeunrestrictedandrestrictedpolynomialdistributedlag(PDL)schemes

acrossthedifferentagriculture-relatedresearchfieldsat114researchuniversities,1993-2015

Dep

ende

nt v

aria

ble:

Res

earc

h pu

blic

atio

ns (

nega

tive

bino

mia

l M

LE

)

All

field

s [1

]D

airy

& a

nim

al s

cien

ces

[2]

Bio

tech

nolo

gy &

app

lied

mic

robi

olog

y [3

]C

rop,

hor

ticul

ture

, &

soi

l sc

ienc

es [

4]Fo

od a

nd n

utri

tiona

l sc

ienc

es [

5]U

nres

tric

ted

Res

tric

ted

Unr

estr

icte

dR

estr

icte

dU

nres

tric

ted

Res

tric

ted

Unr

estr

icte

dR

estr

icte

dU

nres

tric

ted

Res

tric

ted

Deg

ree

of P

DL

12

22

32

22

22

2E

xpen

ditu

re_t

-00.

0006

2***

0.00

052*

**0.

0002

2-0

.000

890.

0006

5***

0.00

041*

**0.

0002

90.

0000

10.

0010

8***

0.00

090*

**(0

.000

14)

(0.0

0011

)(0

.000

87)

(0.0

0109

)(0

.000

16)

(0.0

0013

)(0

.000

22)

(0.0

0016

)(0

.000

20)

(0.0

0016

)_t

-10.

0004

1***

0.00

044*

**0.

0022

6***

0.00

357*

**0.

0003

7***

0.00

044*

**0.

0000

90.

0001

7**

0.00

055*

**0.

0005

9***

(0.0

0006

)(0

.000

05)

(0.0

0029

)(0

.000

50)

(0.0

0006

)(0

.000

06)

(0.0

0008

)(0

.000

07)

(0.0

0008

)(0

.000

07)

_t-2

0.00

026*

**0.

0003

6***

0.00

303*

**0.

0039

3***

0.00

020*

*0.

0004

4***

0.00

001

0.00

028*

**0.

0001

70.

0003

4***

(0.0

0009

)(0

.000

01)

(0.0

0054

)(0

.000

80)

(0.0

0010

)(0

.000

01)

(0.0

0015

)(0

.000

02)

(0.0

0012

)(0

.000

02)

_t-3

0.00

017

0.00

029*

**0.

0025

0***

0.00

178*

**0.

0001

40.

0004

1***

0.00

003

0.00

034*

**-0

.000

040.

0001

5***

(0.0

0011

)(0

.000

03)

(0.0

0065

)(0

.000

59)

(0.0

0012

)(0

.000

04)

(0.0

0017

)(0

.000

05)

(0.0

0015

)(0

.000

05)

_t-4

0.00

014*

0.00

021*

**0.

0007

0-0

.001

28**

*0.

0001

9**

0.00

035*

**0.

0001

60.

0003

4***

-0.0

0010

0.00

002

(0.0

0008

)(0

.000

05)

(0.0

0048

)(0

.000

28)

(0.0

0008

)(0

.000

06)

(0.0

0012

)(0

.000

07)

(0.0

0011

)(0

.000

07)

_t-5

0.00

018*

**0.

0001

4***

-0.0

0239

***

-0.0

0366

***

0.00

036*

**0.

0002

6***

0.00

039*

**0.

0002

8***

0.00

001

-0.0

0005

(0.0

0006

)(0

.000

05)

(0.0

0032

)(0

.000

54)

(0.0

0007

)(0

.000

06)

(0.0

0009

)(0

.000

07)

(0.0

0008

)(0

.000

07)

_t-6

0.00

028

0.00

060*

**-0

.006

77**

*0.

0110

5***

0.00

064*

**0.

0004

7***

0.00

073*

*0.

0002

3***

0.00

028

0.00

128*

**(0

.000

19)

(0.0

0019

)(0

.001

07)

(0.0

0367

)(0

.000

21)

(0.0

0004

)(0

.000

30)

(0.0

0005

)(0

.000

25)

(0.0

0027

)Su

m o

f th

e la

gs0.

0020

7***

0.00

255*

**-0

.000

450.

0145

1***

0.00

255*

**0.

0027

8***

0.00

169*

**0.

0016

5***

0.00

195*

**0.

0032

2***

(0.0

0007

)(0

.000

23)

(0.0

0034

)(0

.004

06)

(0.0

0008

)(0

.000

10)

(0.0

0011

)(0

.000

14)

(0.0

0010

)(0

.000

32)

Mea

n la

g2.

2327

62.

8018

43.

9829

64.

0701

02.

9735

02.

9000

04.

2227

43.

5662

91.

4027

02.

9173

1C

onst

ant

2.48

244*

**2.

4840

4***

0.91

585*

**0.

9118

9***

2.17

078*

**2.

1697

1***

2.84

366*

**2.

8491

6***

2.33

075*

**2.

3332

0***

(0.0

4724

)(0

.047

22)

(0.0

8742

)(0

.087

94)

(0.0

5337

)(0

.053

26)

(0.0

6956

)(0

.069

50)

(0.0

6533

)(0

.065

28)

AIC

215

,915

.79

15,9

15.1

64,

403.

314,

409.

5713

,079

.88

13,0

83.8

37,

851.

947,

853.

539,

405.

859,

405.

71SB

IC3

15,9

38.0

715

,931

.87

4,42

0.25

4,42

6.50

13,1

02.1

213

,100

.52

7,87

1.91

7,86

8.51

9,42

6.71

9,42

1.36

Log

-lik

elih

ood

-7,9

53.8

9-7

,954

.58

-2,1

97.6

5-2

,200

.78

-6,5

35.9

4-6

,538

.92

-3,9

21.9

7-3

,923

.77

-4,6

98.9

2-4

,699

.86

Obs

erva

tion

1,93

81,

938

510

510

1,92

11,

921

1,08

81,

088

1,36

01,

360

Gro

up11

411

411

411

411

411

411

411

411

411

4

Not

es:

1.

Th

e nu

mb

er

is

the

deg

ree

of

po

lyno

mia

l;

2.

Aka

ike

Info

rmat

ion

Cri

teri

on;

3.

Sch

war

z'

Bay

esia

n In

form

atio

n

Cri

teri

on;

A

ll

mod

el

assu

mes

th

e gr

oup

fi

xed

ef

fect

; P

aren

thes

es

are

stan

dard

er

rors

; *

**

at

1%

, *

* at

5%

, an

d *

at

10

%

lev

el

of

stat

isti

cal

sign

ific

ance

.



In the unrestricted model, we use a second degree polynomial and the

maximum length of the lag is 6 years. Since the slope coefficients of the negative

binomial MLE do not directly reveal the marginal effect19, the values in Table 3

are much smaller than the coefficient values in the log-log KPF model in Table

2. An alternative is to use an incident rate ratio (IRR) for coefficients estimated

in the negative binomial MLE of the KPF model for indicating the marginal effect.

In Table 3, what is reported are true values of the slope coefficients from the neg-

ative binomial MLE regression, not IRR values.

In comparison to the statistical significance of the estimated coefficients

in the log-linear KPF model in Table 2, the slope coefficients of the negative bi-

nomial MLE in Table 3 are relatively less statistically significant, especially those

in models 4 and 5 for crop, horticulture, and soil sciences and food and nutritional

sciences respectively, as well as the sum of coefficients in model 2, for dairy and

animal sciences. Again, the mean lags in each model can be interpreted to repre-

sent the average lag between effective inputs and measured outputs, or the

so-called research “gestation” period. These values indicate that changes in life

science research expenditures affect research publications 2.23 years later in model

1: similarly, 3.98 years later in model 2; 2.97 years in model 3; 4.22 years in

model 4; and 1.40 years in model 5. The lags here are similar to the results of

the log-linear model in Table 2, except that here the mean lag for the food and

nutritional science research publications is much smaller than in the log linear

model, 1.40 years compared to 3.29 years.

In the restricted PDL negative binomial model in Table 3, the degree of

polynomial is second order and the maximum length of the lag is 6 years. but the

one exception is in model 2, dairy and animal sciences, in which a third order

polynomial provides the best fit. As shown in the Empirical Model Framework

section, the restricted PDL model known as the end-point restriction assumes that

there is no impact beyond 6 years of lagged research expenditures on current year

publications. Unlike the results of the unrestricted PDL model in Table 3, most

of the slope coefficients in the restricted model in Table 3, are statistically sig-

nificant, at least at the 5 percent level. Improvements are especially notable in

models 4 and 5 compared with the unrestricted models. Although, the restricted

PDL model may be too restrictive in some assumptions--it cuts off lag effects be-

yond 6 years--still, it has meaningful interpretations. One in particular is how re-

19 Because of the characteristics of log likelihood function and its mean, exp′ .


search expenditures of various lags effects current research publications by com-

paring the values (magnitudes, signs, statistical significance, etc.) of the slope co-

efficients between the unrestricted and the restricted PDL models. Another set of

meaningful interpretations can come from comparing mean lags between the two

sets of models.

In comparing between the unrestricted and restricted PDL models in Table

3, the results of models 4 and 5, the crop, horticulture, and soil sciences and the

food and nutritional sciences, respectively, have quite different magnitudes and

signs of the slope coefficients, as well as different statistical significance. Research

publications in these fields are more likely to be affected by six or more years

of lagged research expenditures. Moreover, in model 5, the mean lag of the unre-

stricted model is much shorter than the restricted model. Finally, we note that the

mean lags in the restricted PDL model in Table 3 do not differ from the mean

lags in Table 2. Therefore, the mean lags in the negative binomial MLE with a

restricted PDL structure can be useful for evaluating the average lag between re-

search project’s inception and completion (its gestation period) across the different

research fields.

2. The role of Land-Grant universities in agricultural knowledge

production and commercial innovation

The main purposes for adopting the analysis of variance (ANOVA) are to explore

one of our main research questions, how the Land-Grant status of a university—

and therefore its focus on regional economic development—affects its output of

research in fields affecting the agricultural industry, and to avoid a multi-

collinearity problem with a fixed effect model in the panel data analysis. Using

a dummy variable regression, called an analysis of variance (ANOVA) model, we

can incorporate the concept of interaction between a quantitative dependent varia-

ble and a number of qualitative explanatory variables. The ANOVA can be used

to test differences among two or more groups’ mean values. The null hypothesis

is that the mean values of all group are the same, i.e. that they are not statistically

independent.

In this section, there are two different types of ANOVA models: (1) a

model with just one qualitative explanatory variable (whether or not a university

is a Land-Grant institution) and (2) a model with two qualitative explanatory vari-


ables that allows for interaction effects (whether or not a university is a

Land-Grant institution, and the geographic region of the university). Equation (7)

describes the first ANOVA test, which is based on the pooled-OLS data from

1993 to 2015, with a dummy variable for Land-Grant university status, which is

then related to publication output counts across different food and ag related re-

search categories:

(7)

Where Y = count of research publications related to food and agriculture by

authors at university i in research field j

L = 1 if the university is a Land-Grant university

0 if otherwise: non Land-Grant universities (both public and private)

Public = 1 if the university is public, but non Land-Grant

0 if otherwise

Table 4 displays the results of the ANOVA test on the number of research

publications by the 114 U.S. research universities in each of the different research

field groups, from 1993 to 2015. The test results indicate how the mean number

of research publications for each field by authors at Land-Grant universities differ

from the mean number of research publications for the same field in the non

Land-Grant universities.


TABLE 4. An analysis of variance (ANOVA) model with one qualitative variable for

Land-Grant universities, across the different agriculture-related research fields at 114 re-

search universities, 1993-2015

Dependent variable: Research publications

All fieldsAg dairy

animal science


microbiology

Crop, plant,

horticulture, and soil science

Food and nutritional

science

[1] [2] [3] [4] [5]

Land Grant222.434*** 36.046*** 12.281*** 91.604*** 41.939***

(6.658) (1.329) (1.813) (2.447) (1.851)

non Land-Grant (public)-19.884*** 0.710 -15.735*** 4.366* -9.543***

(6.735) (1.345) (1.834) (2.475) (1.873)

Constant63.637*** 0.000 38.132*** 5.037*** 20.468***

(4.923) (0.983) (1.341) (1.809) (1.369)

R-squared 0.3991 0.2797 0.0895 0.4183 0.2637

Adjusted R-squared 0.3986 0.2792 0.0888 0.4179 0.2631

F-statistics 869.65*** 508.57*** 128.74*** 941.64*** 469.01***

Observation 2,622 2,622 2,622 2,622 2,622

Notes: In order to prevent a dummy variable trap, we are treating the private universities

as the benchmark category; Parentheses are standard errors; *** at 1%, ** at 5%,

and * at 10% level of statistical significance.

In Table 4, the mean annual number of research publications for all agri-

culturally related publications from a Land-Grant university is 286.07 per year,

which is calculated E(Y_i│L_i=1,〖public〗_i=0)= β_0+β_1. Similarly, the mean

number of total agriculturally-related publications from a public non Land-Grant

university is 43.75 per year and the mean number of publications from a private

university is 63.64 per year, which can be calculated by E(Y_i│L_i=0,〖public〗

_i=1)=β_0+β_2 and the intercept itself, β_0, respectively. Following these for-

mulae, in the dairy and animal sciences, the mean annual number of research pub-

lications by a Land-Grant university is 36.05, but public non Land-Grant and pri-

vate universities have almost zero. In biotechnology and microbiology, the

Land-Grant university’s mean annual number of research publications is 50.41, and

the public non Land-Grant and private universities’ mean annual number of re-

search publications are 22.40 and 38.13, respectively. In the crop, horticulture, and


soil sciences, the means are 96.64, 9.40, and 5.04 per year, respectively, and in

food and nutritional sciences, they are 62.41, 10.93, and 20.47 per year,

respectively.

Overall, the mean number of research publications by the Land-Grant uni-

versities is significantly greater than the mean number of research publications in

the non Land-Grant universities in these agriculturally related research fields.

Particularly, however, in the traditional research fields in agriculture, such as in

the dairy and animal sciences or in the crop, horticulture, and soil sciences, the

production of research publications by the Land-Grant universities is significantly

higher than by the non Land-Grant universities (both public and private). In the

biotechnology and microbiology as well as the food and nutritional sciences, the

private universities have a remarkably high output of research publications, likely

due to the presence of medical schools within many of them.

FIGURE 1. The location of all 114 top-tier (R1) universities in the United States, by

Land-Grant and non Land-Grant institutions, broken out into seven geographic regions

In order to consider the importance and influence of geographic location

on the relative specialization in agricultural research, the ANOVA test can be ex-

tended to include two qualitative variables: (1) the Land-Grant and (2) regional

dummy variables. Again, the test is based on the pooled-OLS20 data from 1993

to 2015. In Figure 1, the universities are classified into seven different multi-state


regions—including Pacific, Mountain, Northern Plains, Southern Plains, Central,

Southeast, and Northeast—following Alston et al. (2010), recognizing that each re-

gion shares broadly similar climactic and agroecological characteristics, and there-

fore similar profiles of the agricultural industry within the states of that region.

Equation (8) represents the interaction effects between the Land-Grant and regional

variables.

(8) ≠

⋯ ≠

Where Y= count of research publications related to food and agriculture by au-

thors at university i in research field j

L = 1 if the university is a Land-Grant university, 0 a non Land-Grant university

M = 1 if the university is in the Mountain region, 0 otherwise

NP = 1 if the university is in the Northern Plains, 0 otherwise

SP = 1 if the university is in the Southern Plains, 0 otherwise

C = 1 if the university is in the Central region, 0 otherwise

SE = 1 if the university is in the Southeast, 0 otherwise

NE = 1 if the university is in the Northeast, 0 otherwise

Table 5 displays the results of the estimation of the interaction effects be-

tween the Land-Grant university variable and the regional variables for the mean

annual numbers of research publications by all 114 U.S. R1 research universities,

across the different research field groups and geographic regions, from 1993 to

2015. The total number of the Land-Grant universities is 41 in our data sample

of 114 universities, (see more details on which regions the Land-Grant universities

fall within in Appendix 3 and Figure 1). Similar to the results in Table 4, the

mean annual number of research publications across all fields are significantly

greater in the Land-Grant universities than the non Land-Grant universities.

20 Pooled-OLS data can be treated by combining both time series (23 years) and cross-sectional

(114 universities) data. Although it is somewhat distinguished from the panel data, the main data

set is same in both approaches.


TABLE 5. An analysis of variance (ANOVA) model with two qualitative variables,

Land-Grant universities and geographic regions, across various agriculturally-related re-

search fields at 114 U.S. research universities, 1993-2015

Dependent variable: Research publications

All fields

Ag dairy

animal science

Biotechnology

& applied

microbiology

Crop, plant,

horticulture, &

soil science

Food and

nutritional

science

[1] [2] [3] [4] [5]

Land-grant 136.249*** 10.194*** -6.422 76.421*** 26.519***

(14.597) (2.731) (4.104) (5.537) (4.003)

Mountain -31.228* 0.000 -35.220*** 9.080 -5.089

(18.155) (3.397) (5.104) (6.886) (4.978)

Northern Plains -22.087 0.000 -40.209*** 13.157 4.965

(29.647) (5.547) (8.336) (11.246) (8.129)

Southern Plains -39.013*** 2.770 -41.896*** 4.835 -5.957

(14.824) (2.774) (4.168) (5.623) (4.065)

Central -25.330* 0.000 -24.999*** 0.008 -0.339

(14.406) (2.695) (4.050) (5.464) (3.950)

Southeast -30.665** 0.000 -32.094*** -1.010 2.438

(13.769) (2.576) (3.871) (5.223) (3.775)

Northeast -7.199 0.000 -17.013*** -4.409 14.222***

(13.305) (2.489) (3.741) (5.047) (3.648)

Land×Mountain -4.890 9.524* 20.324*** -19.910* -4.095

(27.612) (5.166) (7.763) (10.474) (7.571)

Land×Northern Plains 144.403*** 44.437*** 22.118*** -7.334 22.308**

(36.218) (6.777) (10.183) (13.738) (9.931)

Land×Southern Plains 137.779*** 45.935*** 33.211*** -4.303 44.852***

(23.032) (4.309) (6.476) (8.736) (6.315)

Land×Central 225.651*** 55.372*** 44.576*** 34.847*** 59.498***

(19.123) (3.578) (5.377) (7.253) (5.243)

Land×Southeast 137.019*** 34.812*** 28.742*** 39.060*** 13.872***

(18.647) (3.489) (5.243) (7.073) (5.113)

Land×Northeast 33.602* 12.061*** 21.649*** -4.452 22.243***

(18.661) (3.492) (5.247) (7.078) (5.117)

Constant 74.043*** 0.000 55.122*** 7.713* 11.209***

(12.103) (2.265) (3.403) (4.591) (3.319)

R-squared 0.4681 0.4400 0.1408 0.4514 0.3660

Adjusted R-squared 0.4654 0.4372 0.1365 0.4487 0.3629

F-statistics 176.54*** 157.65*** 32.88*** 165.10*** 115.82***

Observation 2,622 2,622 2,622 2,622 2,622

Notes: In order to avoid a dummy variable trap, we are treating the non land-grant

universities (both public and private) and Pacific region as the benchmark

category; Parentheses are standard errors; *** at 1%, ** at 5%, and * at 10%

level of statistical significance.


For all fields of agricultural research, the Land-Grant universities in the

Central region stand out for having a relatively higher production of research pub-

lications than other regions, at 410.61 per university per year.21 In the field of dai-

ry and animal sciences, the mean number of research publications by Land-Grant

universities in the Central region is 65.57 per year, in the Northern Plains, 54.63

per year, and in the Southern Plains, 58.90 per year.

In contrast, it is much lower in the Pacific region, at 10.19 per year; in

the Mountain region, at 19.72 per year, and in the Northeast region, at 22.25 per

year. However, as noted in the previous ANOVA, the mean number of research

publications in dairy and animal sciences by non Land-Grant universities are al-

most zero, and the current analysis shows that this holds across all regions.

In biotechnology and applied microbiology, the Land-Grant universities in

the Central region, again, have the highest production of research publications, at

68.28 per year. In this field, the non Land-Grant universities in the Pacific region

have a slightly higher mean number of research publications, at 55.12 per year,

than the Land-Grant universities in the Pacific region, 48.70 per year. This can

be explained by the fact that this group includes a broad range of biology related

topics, such as applied genetics, molecular biotechnology, genomics and proteo-

mics, cell biology, enzymes and proteins, etc., many of which can also be pursued

in the medical sciences, and more general biology departments. There has long

been overlap between the agricultural life sciences and medicine. In agriculture,

biotechnology has long focused on breeding techniques, genetic modification of

crops, microorganisms for foods and agricultural products, and bioenergy. Some

of the large non Land-Grant universities are on the Pacific coast.

In the field of crop, horticulture, and soil sciences, the mean number of

research publications is greatest from Land-Grant universities in the Southeast, at

122.18 per year, but very closely followed, again, by the Land-Grant universities

in the Central region, at 118.99 research publications per year. In the Southeast,

specialty horticultural crops, such as citrus in Florida and peanuts or peaches in

Georgia, are particularly important to agricultural industries of those states.

Finally, in the food and nutritional sciences, it is again the Land-Grant universities

in the Central region that have the highest mean number of research publications,

at 96.89 per year, followed by the Land-Grant universities in the Southern Plains,

21 It can be calculated by


at 76.62 per year, and in the Northeast, at 74.19 per year.

In sum, the two ANOVA models establish that Land-Grant universities

certainly do produce farm more research in the agricultural and food sciences than

non Land-Grant universities, and among the Land-Grant universities there is some

evidence of further specialization within fields of agriculture. We also see that in

the Central region, characterized by the “Corn Belt” wherein agriculture is rela-

tively strongest in the United States, the Land-Grant universities there are the larg-

est and therefore tend to dominate the production of research publications across

the full range of topics related to agriculture and food.

IV. Discussion and Implications

In this study, we focus on the mission and role of the Land-Grant universities and

their sub-institutions—the state agricultural experimental stations (SAESs) and co-

operative extension services—in agriculture. The system of the Land-Grant uni-

versities and its corresponding policies in the United States are quite unique in the

production of agricultural knowledge and dissemination activities. Indeed, the U.S.

public and private sectors have been performing the most food and agricultural

R&D in the world. Thus, understanding the system and management of knowledge

production in the U.S. Land-Grant universities would be significantly meaningful

in any country. Indeed, we expect the main context and results of this paper could

be applied to the study of the system of food and agricultural R&D and commer-

cial innovation in Korea.

The Korean government has been expanding R&D spending to revitalize

the economy in the agricultural sector, encouraging food and agricultural in-

novation for sustainable growth. However, the system and structures of innovation

have been driven by government-led models. According to Lee et al. (2016), pub-

lic sector agencies and institutions, such as the Rural Development Administration

(RDA), Province Agricultural Research & Extension Services (PARES), and

Agricultural Technology Center (ATC), are the dominant players in the food and

agriculture-related research networks, and they play a central role in the agricul-

tural technology innovation system (ATIS), whereas the private sector industries

exhibit only a weak network in the ATIS even though their roles are crucial for


introducing commercial innovations in agriculture. Thus, most of the research net-

work is bound up in the public sector, and the structure of the network is more

likely to exhibit a hierarchical structure.

The main reasons behind this situation seem to be the different per-

spectives between public and private sectors. In fact, the public sector most often

pursues publicly-oriented objectives, whereas the private sector or industry is a

profit maximizer and more often pursues knowledge denominated and dis-

seminated via an intellectual property (IP) based mechanism. Thus, the public sec-

tors’ direct collaborations with private sector actors can be somewhat difficult.

Interestingly, a university can be a good mediator between public sector and pri-

vate sector entities, because the research team formations in the modern research

universities run like small businesses, or “quasi-firms,” optimizing their collective

behavior albeit without being directly profit making (Etzkowitz 2003). They do

much more research collaboration with private sector R&D than the public sector

does, and conversely, they collaborate more with public sector researchers than in-

dustry does. Beyond the traditional dyadic relationships, university research teams

often exhibit triadic relationship involving university, industry, and government

(a.k.a. the “triple helix”), which is characterized as a dynamic network (Etzkowitz

1993; Etzkowitz and Leydesdorff 1995 and 2000). This conceptualization of the

R&D system may suggest an important alternative for the system of agricultural

innovation in Korea, in which the system tends to be mostly a one-way or a hier-

archical, government-driven network.

Furthermore, universities also provide a good venue for engagement with

industry stakeholders, in creating new knowledge that can lead to commercial

innovations. While the public sector has played a leading role in Korean agricul-

tural R&D, in the U.S. the private sector has been the largest funder and perform-

er of agricultural R&D. Following Clancy et al. (2016), in 2013, the food and ag-

ricultural R&D funding sources from the federal and state governments accounted

for $3.8 billion (23.7 percent) of a total of $16.3 billion. R&D funding from pri-

vate sector sources, such as private companies, foundations, and farmer organ-

izations, accounted for $12.5 billion (76.3 percent). And, while almost all of the

private sector R&D funding ($11.8 billion or about 94 percent) supported R&D

performed by private sector organizations themselves, a small but significant por-

tion of private sector R&D funding supported R&D performed by the Land-Grant

universities ($0.7 billion or about 6%). The private sector R&D sponsorship of


R&D in the Land-Grant universities is increasing significantly, even though a large

share of the food and agricultural R&D funding still comes from public sponsors

in the state and Federal governments (including the USDA, NSF, NIH, etc.), ac-

counting for $2.35 billion of the total $3.04 billion of R&D performed by

universities.

Of particular importance in this regard is the potential of university

knowledge production activities to affect commercial innovation through various

forms of spillovers and collaborations between university, industry, and

government. Thus, a deeper understanding of the U.S. Land-Grant system and its

R&D activities may have many implications for the system of Korean agricultural

innovation in terms of transitioning from government-centered or supplier-led

models toward more user-led or network-based models.

Finally, the trends of knowledge production by research field in the U.S.

research universities, and especially in the Land-Grant universities, provide im-

portant indicators of global trends in food and agriculture-related research for cre-

ating a new knowledge and preparing for new directions in industrial innovation.

Following the empirical results of this paper, it is clear that the traditional research

fields in agriculture, such as dairy or animal science, crop science, horticulture,

and soil science have quantitatively a greater volume of output than other research

fields. However, in terms of knowledge convergence, these fields show less oppor-

tunities for collaborating with non Land-Grant universities, which have the poten-

tial to bring new research topics and funding sources, fields such as computer sci-

ence and data analysis for precision agriculture.

In particular, the biotechnology and applied microbiology research field

appears to have greater cost advantages in the long run than other research fields

and a shorter mean lag between research project inception and completion. The

results indicate that the biotechnology and applied microbiology research field, as

related to agriculture and food, is generally overlapping with similar application

of the biological sciences in other fields, such as medical sciences and bioenergy.

The top-tier private universities as well as a range of industries in the United

States are paying attention to these research fields, including technologies like

CRISPR-mediated genome editing and analysis of the agricultural microbiome (see

more Egelie et al. 2016; Graff and Zilberman 2017). Thus, in terms of opportunity

for the creation of new knowledge and commercial innovations in agriculture, the

research areas of biotechnology and applied microbiology are more likely to pro-


vide potential for sustainable growth in agriculture. Therefore, we expect that these

results are a meaningful indicator of where Korean R&D should go and what it

should focus on in creating new knowledge and commercial innovations in

agriculture.

V. Conclusions

This paper analyzes the knowledge production and dissemination activities of the

largest research universities in the United States, specifically in the fields related

to agriculture and food, and explores the special role of the Land-Grant

universities. In the economy overall, universities conduct 14 percent of total R&D,

but in the agricultural and food industries, universities conduct almost 30 percent

of R&D. And, considering R&D in just the agricultural sector alone, the share of

university R&D is even higher, closer to 50 percent. A large portion of this is due

to the role of the Land-Grant universities, which historically have specialized in

agricultural and food related research, and the dissemination of that research to

stakeholders within their respective regions. Of the 114 Carnegie R1 research uni-

versities in the United States, 36 percent are Land-Grant institutions; these

Land-Grant universities account for 38 percent of the life sciences research ex-

penditures, but fully 75 percent of the agricultural and food related research pub-

lications produced. Yet, we must look at the Land-Grant universities within the

context of the larger set of research universities, because the other 25 percent of

research publications come from them and because the Land-Grant universities

collaborate with and apply scientific discoveries from other universities as peer

institutions. We seek to understand how the knowledge production activities of the

U.S. university system work together to create a huge repository of new knowl-

edge that is available to enable commercial innovation and technological change

within the agricultural and food industries.

The first empirical analysis characterizes the technical relationship be-

tween life science research expenditures as an input and agricultural and food re-

lated research publications as an output in a knowledge production function (KPF)

of all of the 114 top-tier U.S. research universities over 23 years. We utilize three

different agricultural KPF models: a log-linear model with an unrestricted poly-


nomial distributed lag (PDL) scheme; a negative binomial maximum likelihood es-

timation (MLE) with an unrestricted PDL; and a negative binomial MLE with a

restricted PDL. Adopting the analysis of neoclassical production theory like returns

to scale can be useful for understanding university research productivity. The re-

sults of this analysis show that the production of research publications for all food

and ag related fields exhibits decreasing returns to scale (DRTS) and among the

different fields, biotechnology and applied microbiology appear to have greater

cost advantages in the long run. This perhaps follows from the greater overlap,

and thus potential spillovers, with medical research and other biological sciences.

The mean lag between research expenditure inputs and research pub-

lication outputs indicate the gestation period between a research project’s inception

and completion. Across the three KPF models, we find the log-linear model and

the negative binomial MLE with a restricted PDL are most similar: with the mean

lags ranging from 2.90 years for biotechnology and applied microbiology as the

shortest, to 4.07 years, for the dairy and animal sciences as the longest. It is clear

that the gestation periods or project cycle times vary significantly across field. But,

it is also clear that there is a significant lag between changes in research inputs

and detectible changes in outputs. One of the major reasons regarding the different

nature of the mean gestation lags across sub-fields might be the level of the partic-

ipation rate of non Land-Grant universities, especially top-tier private universities,

which have the potential to bring new funding sources. Moreover, the mean lags

can be slightly affected by the journal environments such as the duration and qual-

ity of the peer review process across the different journals.

The second empirical analysis focuses on the role of the Land-Grant uni-

versities in food and ag related research activities by an analysis of variance

(ANOVA). We find that in our sample of the 114 top research universities in the

U.S., the Land-Grant universities produce a higher mean number of research pub-

lications across all food and ag related fields of research than do the public non

Land-Grant universities or the private non Land-Grant universities. Particularly in

such traditional agricultural research fields as animal sciences or soil and crop sci-

ences, the mean number of research publications by the Land-Grant universities

are much greater than those by non Land-Grant public and private universities.

Finally, looking at the relationship between the geographic locations of

universities by region and their profiles of agricultural research we see that the

Land-Grant universities in the Central region of the United States or the “Corn


Belt”, where agriculture is a relatively more important industry for the region’s

economy, produce the most food and ag related research publications, averaging

410.61 papers per year. Specifically, for the research field of crop, horticulture,

and soil sciences, the Land-Grant universities in Southeast produce slightly more

than the Land-Grant universities in the Central region, 122.18 papers per year and

118.99 papers per year, respectively. However, in Pacific region, on average, the

non Land-Grant universities produce more research publications for the bio-

technology and applied microbiology related fields than the Land-Grant

universities. Thus, we interpret this result that the research topics for the bio-

technology and applied microbiology can be covered by a variety of research

areas, such as medical science, agricultural science, bioengineering and bioenergy,

etc., so the non Land-Grant universities, especially private universities, are also

highly engaged in these research topics.

At the industry level of agriculture and food, we thus see the interesting

dynamic of university R&D and how it contributes to innovation within such a

highly regionalized and diffused industry. By having a set of top-tier general re-

search universities with specialized programs in agricultural R&D, namely the

Land-Grant universities, the U.S. system achieves three things: (1) The agricultural

sciences are maintained as fields of top-tier research, rather than being delegated

to a second tier of more vocationally oriented or field work, within the national

educational system; (2) Those Land-Grant institutions that are specialized in agri-

cultural sciences and that have the institutional capacity for disseminating new ag-

ricultural knowledge dominate in the field, producing the majority of agricultural

research publications; (3) Also, as top-tier institutions doing high level research in

interesting life sciences and related fields—such as genomics, pathology, epidemi-

ology, population dynamics, etc.—scientists at the Land-Grant universities play a

key role of collaborating with scientific colleagues at other non Land-Grant uni-

versities, thus enabling the Land-Grant universities to capture spillovers from their

peer institutions, the other 64 percent of universities, and applying that knowledge

to agricultural problems within their particular regional contexts.

Further, the results of this paper would suggest some insights and im-

plications for the agricultural R&D in Korea: (1) to understand the importance of

the Land-Grant system and its corresponding policies for creating a new knowl-

edge and inducing commercial innovation, (2) to realize a university as a good

venue for engagement with industry stakeholders, who can lead to commercial in-


novation, and a good mediator between public and private sectors for achieving

collaborative research and triadic research network, and (3) to find potentially

commercializable research topics such as the biotechnology and applied micro-

biology for attaining sustainable growth in agriculture. Thus, these factors would

provide some important implications to the system of Korean agricultural in-

novations for transitioning from government-centered or supplier-led models to

user-led or network based models, and suggest a new vision for where the future

Korean agriculture should go and what to focus on for creating a new knowledge

in agriculture.

In further study, such analysis should take into account other types of uni-

versity knowledge outputs, such as informally disseminated “tacit” knowledge, for-

mally licensed patents, and startup companies founded by research universities. We

expect that the measurement and inclusion of additional research outputs will en-

able the analysis of them as co-products of the university knowledge production

function. Other directions of analysis can explore how private funding affects the

productivity of university knowledge production and which knowledge outputs are

more highly response to the industry grants and contracts across the different food

and ag related research fields.

REFERENCES

Adams, J.D. and Z. Griliches. 1998. “Research Productivity in a System of Universities.”

Annales d'Economie et de Statistique 49/50: 127-162. doi:10.2307/20076113

Adams, J.D. 2002. “Comparative Localization of Academic and Industry Spillovers.”

Journal of Economic Geography 2(3): 253-278. doi:10.1093/jeg/2.3.253

Agrawal, A. and R. Henderson. 2002. “Putting Patents in Context: Exploring Knowledge

Transfer from MIT.” Management Science 48(1): 44-60. doi:10.1287/mnsc.48.1.44.14279

Almon, S. 1965. “The Distributed Lag Between Capital Appropriations and Expenditures.”

Econometrica 33: 178-196. doi:10.2307/1911894

Alston, J.M., M.A. Andersen, J.S. James, and P.G. Pardey. 2010. Persistence Pays: US

Agricultural Productivity Growth and the Benefits from Public R&D Spending.

Springer: Science and Business.

Anselin, L., A. Varga, and Z. Acs. 2000. “Geographical Spillovers and University

Research: A Spatial Econometric Perspective.” Growth and Change 31(4): 501-515.

doi:10.1111/0017-4815.00142

Audretsch, D.B. and M.P. Feldman. 1996 “R&D Spillovers and the Geography of Innovation

and Production.” American Economic Review 86(3): 630-640.

Belenzon, S. and M. Schankerman. 2010. “Spreading the Word: Geography, Policy, and

University Knowledge Diffusion.” LSE STICERD Research Paper 50.

Bonaccorsi, A., C. Daraio, and L. Simar. 2006. “Advanced Indicators of Productivity of

Universities: An Application of Robust Nonparametric Methods to Italian data.”

Scientometrics 66(2): 389-410. doi:10.1007/s11192-006-0028-x

Boschma, R.A. 2005. “Proximity and Innovation: A Critical Assessment.” Regional Studies

39(1): 61-74. doi:10.1080/0034340052000320887

Buenstorf, G., and A. Schacht. 2013. “We Need to Talk – or Do We? Geographic

Distance and the Commercialization of Technologies from Public Research.” Research

Policy 42(2): 465-480. doi:10.1016/j.respol.2012.06.010

Clancy, M., K. Fuglie, and P. Heisey. 2016. “US Agricultural R&D in an Era of Falling

Public Funding.” Amber Waves 10.

Clark, B.Y. 2011. “Influences and Conflicts of Federal Policies in Academic–industrial

Scientific Collaboration.” Journal of Technology Transfer 36(5): 514-545.

Cohen, W.M., R.R. Nelson and J.P. Walsh. 2002. “Links and Impacts: The Influence of Public

Research on Industrial R&D.” Management Science 48(1): 1-23. doi:10.1287/mnsc.48.1.1.14273

Cohn, E., S.L.W. Rhine, and M.C. Santos. 1989. “Institutions of Higher Education as

Multi-Product Firms: Economies of Scale and Scope.” Review of Economics and

Statistics: 284-290. doi:10.2307/1926974


Colyvas, J., M. Crow, A. Gelijns, R. Mazzoleni, R.R. Nelson, N. Rosenberg, and B.N.

Sampat. 2002. “How Do University Inventions Get into Practice?” Management Science

48 (1): 61-72. doi:10.1287/mnsc.48.1.61.14272

Cote, L.S. and M.K. Cote. 1993. “Economic Development Activity among Land-Grant

Institutions.” Journal of Higher Education 64(1): 55-73. doi:10.2307/2959977

Coupe, T. 2003. “Science is Golden: Academic R&D and University Patents.” Journal of

Technology Transfer 28(1): 31-46.

Cowan, R. 2005. “Universities and the Knowledge Economy.” MERIT Research

Memorandum, No. 2005-07.

Crespi, G.A., and A. Geuna. 2008. “An Empirical Study of Scientific Production: A Cross

Country Analysis, 1981-2002.” Research Policy 37(4): 565-579. doi:10.1016/j.respol.2007.

12.007

Egelie, K.J., G.D. Graff, S.P. Strand, and B. Johansen. 2016. “The Emerging Patent

Landscape of CRISPR-Cas Gene Editing Technology.” Nature biotechnology 34(10):

1025-1031. doi:10.1038/nbt.3692

Etzkowitz, H., and L. Leydesdorff. 1995. “The Triple Helix—University-Industry-

Government Relations: A Laboratory for Knowledge Based Economic Development.”

EASST Review 14(1): 14-19.

______. 2000. “The Dynamics of Innovation: from National Systems and “Mode 2” to a

Triple Helix of University–Industry–Government Relations.” Research Policy 29(2):

109-123. doi:10.1016/s0048-7333(99)00055-4

Etzkowitz, H. 1993. “Enterprises from science: The Origins of Science-based Regional

Economic Development.” Minerva 31(3): 326-360. doi:10.1007/bf01098626

______. 2003. “Research Groups as ‘Quasi-firms’: the Invention of the Entrepreneurial

University.” Research Policy 32(1): 109-121. doi:10.1016/s0048-7333(02)00009-4

Foltz, J.D., B.L. Barham, and K. Kim. 2007. “Synergies or Trade-offs in University Life

Sciences Research.” American Journal of Agricultural Economics 89(2): 353-367.

doi:10.1111/j.1467-8276.2007.01014.x

Graff, G.D., A. Berklund, and K. Rennels. 2014. “The Emergence of an Innovation

Cluster in the Agricultural Value Chain along Colorado's Front Range.” Colorado State

University.

Graff, G.D., and D. Zilberman. 2017. “Intellectual Property and Regulatory Issues of Gene

Editing in Agriculture and Food: The Lessons of GMOs for CRISPR.” 2017

Agricultural and Applied Economics Annual Meeting.

Griliches, Z. 1979. “Issues in Assessing the Contribution of R&D to Productivity Growth.”

Bell Journal of Economics 10(1): 92-116.

Gujarati, D. N. 2004. Basic Econometrics, McGraw Hill.

Hall, B.H., Z. Griliches, and J.A. Hausman. 1986. “Patents and R and D: Is There a

Lag?” International Economic Review 27(2): 265-283. doi:10.2307/2526504


Hausman, J.A., B.H. Hall, and Z. Griliches. 1984. “Econometric Models for Count Data

with an Application to the Patents-R&D Relationship.” Econometrica 52(4): 909-938.

doi:10.2307/1911191

Henderson, R., A.B. Jaffe, and M. Trajtenberg. 1998. “Universities as a Source of

Commercial technology: A Detailed Analysis of University Patenting, 1965-1988.”

Review of Economics and Statistics 80(1): 119-127. doi:10.1162/003465398557221

Heher, A.D. 2007. “Benchmarking of Technology Transfer Offices and What It Means for

Developing Countries.” Intellectual Property Management in Health and Agricultural

Innovation: a Handbook of Best Practices 1: 207-228.

Jaffe, A.B. 1989. “Real Effects of Academic Research.” American Economic Review 79(5):

957-970.

Jaffe, A.B., M. Trajtenberg, and R. Henderson. 1993. “Geographic Localization of

Knowledge Spillovers as Evidenced by Patent Citations.” Quarterly Journal of

Economics 108(3): 577-598.

Laursen, K., and A. Salter. 2004. “Searching high and low: what types of firms use

universities as a source of innovation?” Research Policy 33(8): 1201-1215.

Lee, M., S. Kim, J. Park, S. Yoon, and H. Kim. 2016. Measures to Improve Agricultural

R&D Governance Efficiency and Expand Private Investment (Year 2 of 2). Korea

Rural Economic Institute Research Report.

Lee, Y.H., and G.D. Graff. 2015. “University Research Productivity and its Impact on the

Regional Agricultural Economy: The Case of Colorado State University and the

Colorado Economy.” 2015 AAEA and WAEA Annual Joint Meeting, Agricultural and

Applied Economics Association, No. 205443.

______. 2017. “Empirical Estimation of University Knowledge Production Functions for

Knowledge Outputs Disseminated via Publications, via Direct Industry Collaboration,

and via Technology Transfer Mechanisms.” Paper presented at NC-1034 Conference:

Food, Energy, Environment, & Economics Workshop, Tucson, AZ.

Link, A.N., D.S. Siegel, and B. Bozeman. 2007. “An Empirical Analysis of the Propensity

of Academics to Engage in Informal University Technology Transfer.” Industrial and

Corporate Change 16(4): 641-655.

Liu, S. 2015. “Spillovers from Universities: Evidence from the Land-Grant Program.”

Journal of Urban Economics 87: 25-41.

Mansfield, E. 1991. “Academic Research and Industrial Innovation.” Research Policy 20(1):

1-12.

______. 1995. “Academic Research Underlying Industrial Innovations: Sources,

Characteristics, and Financing.” Review of Economics and Statistics 77(1): 55-65.

Meyer-Krahmer, F., and U. Schmoch. 1998. “Science-based Technologies: University–

Industry Interactions in Four Fields.” Research Policy 27(8): 835-851.

National Science Foundation. 2016. Science and Engineering Indicators,

https://www.nsf.gov/statistics/2016/nsb20161


Pakes, A., and Z. Griliches. 1980. “Patents and R&D at the Firm Level: A First Report.”

Economics Letters 5 (1980), 377-381.

______. “Patents and R&D at the Firm Level: A First Look,” in Zvi Griliches (Ed.),

R&D, Patents, and Productivity, University of Chicago Press for NBER: 55-72.

Pardey, P.G. 1989. “The Agricultural Knowledge Production Function: An Empirical

Look.” Review of Economics and Statistics 71(3): 453-461.

Payne, A., and A. Siow. 2003. “Does Federal Research Funding Increase University

Research Output?” Advances in Economic Analysis and Policy 3(1): 1-46.

Ponds, R., F.V. Oort, and K. Frenken. 2010. “Innovation, Spillovers and

University-Industry Collaboration: an Extended Knowledge Production Function

Approach” Journal of Economic Geography 10 (2): 231-255.

Sampat, B.N. 2006. “Patenting and US Academic Research in the 20th Century: The

World Before and After Bayh-Dole.” Research Policy 35(6): 772-789.

Scherer, F.M., and D. Harhoff. 2000. “Technology Policy for a World of Skew-distributed

Outcomes.” Research Policy 29(4): 599-566.

Shane, S. 2004. “Encouraging University Entrepreneurship? The Effect of the Bayh-Dole

Act on University Patenting in the United States.” Journal of Business Venturing 19(1):

127-151.

Thursby, J.G., and M.C. Thursby. 2011. “Faculty Participation in Licensing: Implications

for Research.” Research Policy 40(1): 20-29.

Wuchty, S, B.F. Jones, and B. Uzzi. 2007. “The Increasing Dominance of Teams in

Production of Knowledge.” Science 316(5827): 1036-1039.

Zucker, L. G., M. R. Darby, and J. S. Armstrong. “Commercializing Knowledge:

University Science, Knowledge Capture, and Firm Performance in Biotechnology.”

Management Science 48(1): 138-153.

Date Submitted: Jun. 23, 2017

Period of Review: Jul. 13～Dec. 15, 2017


APPENDIX 1. The top 114 U.S. universities in the Doctoral Universities-Highest Research

Activity in the Carnegie Classification of Institutions of Higher Education by recent 7 years

of average number of food and agriculture-related publications (except biotech-related

field), covering 2009-2015

University name (rank)

Average number

of publications

per year


Average number

of publications

per year

U. California, Davis (1) 823.0 U. Hawaii, Manoa (48) 56.0

U. Florida (2) 701.6 Emory U. (49) 55.1

Cornell U. (3) 621.9 Duke U. (50) 54.9

Iowa State U. (4) 484.0 U. California, Los Angeles (51) 53.4

North Carolina State U. (5) 468.4 U. Pittsburgh, Pittsburgh (52) 53.3

Washington State U. (6) 465.1 U. Kansas (53) 51.6

U. Georgia (7) 464.7 Boston U. (54) 44.7

U. Minnesota, Twin Cities (8) 433.1 U. California, San Diego (55) 41.1

Michigan State U. (9) 424.7 U. Utah (56) 39.9

U. Wisconsin-Madison (10) 408.1 Florida State U. (57) 37.7

Ohio State U. (11) 377.1 U. South Carolina, Columbia (58) 37.3

Kansas State U. (12) 356.7 West Virginia U. (59) 36.9

U. Illinois, Urbana-Champaign (13) 351.1 U. Texas, Austin (60) 34.6

Texas A&M U., College Station (14) 337.4 Vanderbilt U. (61) 33.4

Purdue U. (15) 334.4 Northwestern U. (62) 30.7

Oregon State U. (16) 319.4 U. Colorado Boulder (63) 28.3

Harvard U. (17) 302.3 U. Southern California (64) 27.4

U. Nebraska, Lincoln (18) 278.3 SUNY, U. Buffalo (65) 26.9

Penn State U. (19) 276.6 Arizona State U. (66) 25.9

U. Arkansas, Fayetteville (20) 267.0 Indiana U., Bloomington (67) 25.7

Virginia Tech U. (21) 252.3 Florida International U. (68) 25.4

Louisiana State U. (22) 249.9 U. Cincinnati (69) 24.3

Colorado State U. (23) 208.9 Brown U. (70) 24.0

U. Missouri, Columbia (24) 200.9 U. Iowa (71) 18.6

U. Tennessee, Knoxville (25) 181.3 U. Oklahoma, Norman (72) 18.3

Rutgers U. (26) 176.4 U. California, Santa Cruz (73) 16.4

U. California, Riverside (27) 167.7 Case Western Reserve University (74) 16.1

U. Kentucky (28) 159.7 U. North Texas, Denton (75) 16.0

U. Massachusetts, Amherst (29) 155.3 George Washington U. (76) 14.7

U. California, Berkeley (30) 139.3 Temple U. (77) 14.1

U. North Carolina, Chapel Hill (31) 136.7 U. Louisville (78) 13.6

Tufts U. (32) 130.9 U. California, Santa Barbara (79) 12.6

Clemson U. (33) 126.3 Georgetown U. (80) 12.6

U. Maryland, College Park (34) 116.4 U. New Mexico (81) 11.6

University of Mississippi (35) 101.9 Northeastern U. (82) 11.6

U. Connecticut (36) 100.0 Wayne State University (83) 11.4

U. Arizona (37) 98.7 Tulane U. (84) 11.0

Texas Tech U. (38) 98.4 U. Miami (85) 10.7

U. Washington, Seattle (39) 97.4 Syracuse U. (86) 10.1


APPENDIX 2. The top 114 U.S. universities in the Doctoral Universities-Highest Research

Activity in the Carnegie Classification of Institutions of Higher Education by 2015 life sci-

ence R&D expenditures


Average number

of publications

per year


Average number

of publications

per year

Johns Hopkins U. (40) 96.6 U. Wisconsin-Milwaukee (87) 10.0

U. Pennsylvania (41) 95.3 Virginia Commonwealth U. (88) 9.6

U. Illinois, Chicago (42) 74.6 Rice U. (89) 8.7

Yale U. (43) 71.1 Georgia State U. (90) 7.7

Washington U., Saint Louis (44) 68.7 U. Oregon (91) 5.1

Columbia U. (45) 67.7 Boston C. (92) 3.9

U. Delaware (46) 62.9 U. Central Florida (93) 2.0

U. Alabama, Birmingham (47) 57.1 Brandeis U. (94) 1.9

U. Texas, Dallas (95) 1.0 SUNY, U. Albany (105) 0.0

California Institute of Technology (96) 0.0 U. California, Irvine (106) 0.0

Carnegie Mellon U. (97) 0.0 U. Chicago (107) 0.0

George Mason U. (98) 0.0 U. Houston (108) 0.0

Georgia Institute of Technology (99) 0.0 U. Michigan, Ann Arbor (109) 0.0

MIT (100) 0.0 U. Notre Dame (110) 0.0

New York U. (101) 0.0 U. Rochester (111) 0.0

Princeton U. (102) 0.0 U. South Florida, Tampa (112) 0.0

Stanford U. (103) 0.0 U. Texas, Arlington (113) 0.0

SUNY, Stony Brook U. (104) 0.0 U. Virginia, Charlottesville (114) 0.0


2015 life science

expenditures

(million $)


2015 life science

expenditures

(million $)

Johns Hopkins U. (1) 867.72 U. South Florida, Tampa (34) 295.10

Duke U. (2) 855.98 U. Arizona (35) 289.95

U. Michigan, Ann Arbor (3) 779.92 U. Nebraska, Lincoln (36) 286.06

U. Washington, Seattle (4) 764.57 U. Chicago (37) 276.13

U. Pittsburgh, Pittsburgh (5) 733.93 U. Miami (38) 268.07

U. California, Los Angeles (6) 718.66 Michigan State U. (39) 259.65

U. North Carolina, Chapel Hill (7) 716.71 U. Illinois, Chicago (40) 258.07

U. Pennsylvania (8) 680.07 Boston U. (41) 251.83

Yale U. (9) 665.28 Penn State U. (42) 241.22

Stanford U. (10) 647.80 SUNY, U. Buffalo (43) 238.68

U. California, San Diego (11) 642.37 U. Kentucky (44) 232.83

Cornell U. (12) 631.73 U. Georgia (45) 232.59

Washington U., Saint Louis (13) 617.66 U. Rochester (46) 231.10

U. Wisconsin-Madison (14) 589.65 U. Virginia, Charlottesville (47) 227.65

U. Minnesota, Twin Cities (15) 581.58 Louisiana State U. (48) 225.67

Columbia U. (16) 573.06 U. Illinois, Urbana-Champaign (49) 220.03

U. Florida (17) 539.65 U. California, Berkeley (50) 211.29

Harvard U. (18) 533.23 Purdue U. (51) 209.98



2015 life science

expenditures

(million $)


2015 life science

expenditures

(million $)

Emory U. (19) 530.68 Virginia Tech U. (52) 209.68

U. California, Davis (20) 512.47 North Carolina State U. (53) 208.85

Vanderbilt U. (21) 489.53 U. California, Irvine (54) 195.40

Ohio State U. (22) 473.75 U. Kansas (55) 192.92

U. Alabama, Birmingham (23) 455.48 U. Missouri, Columbia (56) 181.70

Northwestern U. (24) 451.94 Temple U. (57) 169.01

U. Southern California (25) 411.99 Washington State U. (58) 166.58

New York U. (26) 407.73 Virginia Commonwealth U. (59) 164.14

Rutgers U. (27) 366.18 George Washington U. (60) 159.22

U. Cincinnati (28) 347.13 Wayne State University (61) 157.21

Case Western Reserve University (29) 340.04 Iowa State U. (62) 149.39

U. Utah (30) 326.55 U. Connecticut (63) 143.53

Indiana U., Bloomington (31) 323.49 U. New Mexico (64) 141.88

U. Iowa (32) 322.02 U. Louisville (65) 135.97

Texas A&M U., College Station (33) 320.56 Brown U. (66) 135.23

MIT (67) 129.16 Princeton U. (91) 41.46

Georgetown U. (68) 128.58 Georgia State U. (92) 40.30

U. Oklahoma, Norman (69) 127.53 Clemson U. (93) 39.59

Oregon State U. (70) 122.70 Texas Tech U. (94) 36.13

Kansas State U. (71) 122.68 Northeastern U. (95) 34.16

Colorado State U. (72) 122.50 Florida State U. (96) 34.01

U. Hawaii, Manoa (73) 121.74 Brandeis U. (97) 30.61

Tulane U. (74) 117.78 U. Oregon (98) 30.39

U. Maryland, College Park (75) 115.90 U. Notre Dame (99) 26.99

U. South Carolina, Columbia (76) 111.13 U. Central Florida (100) 26.90

Tufts U. (77) 110.86 U. Colorado Boulder (101) 26.76

West Virginia U. (78) 96.10 U. Houston (102) 24.34

SUNY, Stony Brook U. (79) 88.61 U. California, Santa Barbara (103) 24.09

U. Massachusetts, Amherst (80) 81.17 U. California, Santa Cruz (104) 22.71

Arizona State U. (81) 77.87 U. Texas, Dallas (105) 21.14

SUNY, U. Albany (82) 75.67 Georgia Institute of Technology (106) 19.88

U. California, Riverside (83) 75.62 George Mason U. (107) 18.43

U. Arkansas, Fayetteville (84) 75.20 U. Texas, Arlington (108) 16.08

U. Texas, Austin (85) 74.07 Rice U. (109) 11.82

U. Tennessee, Knoxville (86) 66.44 Carnegie Mellon U. (110) 11.21

U. Mississippi (87) 64.63 U. Wisconsin-Milwaukee (111) 11.08

California Institute of Technology (88) 63.91 U. North Texas, Denton (112) 8.88

U. Delaware (89) 58.33 Boston C. (113) 7.09

Florida International U. (90) 41.71 Syracuse U. (114) 6.96

(continued)


APPENDIX 3. The location of 41 Land-Grant universities ranked as R1 Doctoral

Universities-Highest Research Activity in the Carnegie Classification of Institutions of

Higher Education, by region of the United States

U.S. Regions Universities

Pacific (11) Oregon State U.; U. California; Berkeley; U. California; Davis; U. California; Irvine; U. California; Los Angeles; U. California; Riverside; U. California; San Diego; U. California; Santa Barbara; U. California; Santa Cruz; U. Hawaii, Manoa; Washington State U.

Mountain (2) Colorado State U.; U. Arizona

Northern Plains (2) Kansas State U.; U. Nebraska, Lincoln

Southern Plains (3) Louisiana State U.; Texas A&M U.; U. Arkansas, Fayetteville

Central (8) Iowa State U.; Michigan State U.; Ohio State U.; Purdue U.; U. Illinois, Urbana-Champaign; U. Minnesota, Twin Cities; U. Missouri, Columbia; U. Wisconsin-Madison

Southeast (8) Clemson U.; North Carolina State U.; U. Florida; U. Georgia; U. Kentucky; U. Tennessee, Knoxville; Virginia Tech U.; West Virginia U.

Northeast (7) Cornell U.; Penn State U.; Rutgers U.; U. Connecticut; U. Delaware; U. Maryland, College Park; U. Massachusetts, Amherst

Note: Parentheses are the number of universities.


IMPACT OF INCREASED IMPORTS OFAGRICULTURAL PRODUCTS DUE TO FTAS ONDOMESTIC PRICE DECLINE*

JEONG MIN-KOOK**

MOON HAN-PIL***

SONG WOO-JIN****

Keywords

import contribution rate, equilibrium displacement model, price elasticity,

direct payment for damage

Abstract

The purpose of this paper is to propose a method of estimating the im-

port contribution rate. The import contribution is a factor that should be

considered in calculating the direct payment for damage. The decline

in prices is caused by the combination of various factors. In this case,

the decomposition of various factors can confirm the price drop due to

the increase in imports. To this end, we set up a partial equilibrium mod-

el for individual markets and decompose various factors contributing to

the price decline using the equilibrium displacement model.

Various types of elasticities are needed to calculate the import con-

tribution rate derived from EDM. Because elasticity has a wide spectrum

depending on the purpose of the study or the data used, a cautious

approach is needed to obtain objective figures.

* This paper contains the contents of Annual Reports on Agricultural Products Subject to FTA

Direct Payment in 2013 and 2014.** Senior Research Fellow, Korea Rural Economic Institute, Naju-si, Jeollanam-do, Korea.*** Research Fellow, Korea Rural Economic Institute, Naju-si, Jeollanam-do, Korea.**** Research Fellow, Korea Rural Economic Institute, Naju-si, Jeollanam-do, Korea.

Corresponding author. e-mail: [email protected]


I. Introduction

Since the Korea-Chile FTA entered into force in 2004, Korea has been continuing

FTA negotiations and so far 15 FTAs have been in effect. If FTAs takes effect,

imports will increase due to tariff cuts and TRQ increase effects, which will re-

duce demand for domestic products. Demand declines lead to price declines in do-

mestic products and damage to domestic producers.

The government expected that the damage caused by FTAs would be con-

centrated in the agricultural sector. For this reason, the Special Act on Assistance

to Farmers, Fishermen, Etc. Following the Conclusion of Free Trade Agreements

(hereinafter referred to as the Special Act) was enacted and operated in response

to the expansion of agricultural product market opening from the Korea-Chile

FTA. The Special Act aims to compensate farmers and to improve the com-

petitiveness of agriculture.

The direct payment program for compensating damage was introduced

based on the Special Act as part of strengthening compensation for damage caused

by FTAs. The government pays direct payments to farmers and fishermen who are

hurt by the increase in agricultural imports due to the FTAs, to compensate for

the decrease in income due to the price drop. The direct payment is a compensa-

tion system that makes up for price difference due to the rapid increase of import

caused by FTA implementation.

Price support programs and direct payment programs for the income sta-

bility of farmers and fishermen are being implemented not only in Korea but also

in many countries. However, it is difficult to find cases in which the government

is compensating for the damage to farmers and fishermen due to the progress of

trade liberalization such as FTAs. However, in foreign countries, the U.S. Trade

Adjustment Assistance (TAA) is similar to Korea’s FTA measures. In Korea, the

‘Trade Adjustment Support System’, which is applied to the manufacturing sector,

is a similar case.

Compared with the previous measures (the income compensation direct

payment program introduced in 2004 as a measure to the Korea-Chile FTA), the

direct payment program for compensating damage is aimed at improving the qual-

ity of farmers and fishermen’s life by expanding the range of target items and im-

proving the level of compensation. The target item selection method has been

Impact of Increased Imports of Agricultural Products due to FTAs on Domestic Price Decline 107

changed from the pre-designation to the post-designation method. The post-desig-

nation includes virtually all agricultural products in a manner that is selected for

support in case of damage to the agricultural product. The triggering requirement

was relaxed from an ‘80% decline’ to a ‘90% decline’ in the average price1 for

that year. The compensation rate was raised from ‘80%’ to ‘90%’ of the base

price and the year difference. In 2012, the payment limit was newly entered when

supplementary measures were established. The limit is 50 million won for corpo-

rations and 35 million won for individuals. The implementation period of the pro-

gram is sequentially extended to 10 years (2015.12 ~ 2025.12) after the entry into

force of the Korea-China FTA.

Article 7 (1) and Article 8 (3) of the Special Act shall apply the adjust-

ment factor in the calculation of direct payment so that the payment can be paid

within the range permitted by the Marrakesh Agreement. The adjustment co-

efficient has been determined to be applied in the calculation of direct payments

at the “Committee for Supporting Farmers and Fishermen” (Feb. 28, 2012) so that

the actual payment can be paid within the limits of the domestic agricultural and

fishery subsidy specified in the WTO rules. Since then, the Committee for

Supporting Farmers and Fishermen (Jan. 13, 2013) decided to reflect the additional

import contribution rate to the adjustment coefficient for accurate compensation of

actual import damages in addition to complying with WTO rules.2 The adjustment

coefficient and the import contribution rate have the following structure.

×

According to WTO regulations, direct payment for damage compensation

is classified as AMS, which must be reduced. However, in the case of developing

countries, it is possible to spend within 10% of the production value of a certain

item, which is the limit for de-minimis support. According to the U.R. agreement,

the ceiling of AMS in Korea is 1.4 trillion won since 2004, and the de-minimis

of the commodity is 10% of the production value in the developing countries, 5%

1 the average price over the past five years excluding the highest and the lowest.2 In accordance with Article 6 (1) of the Special Act, the direct payment is paid to the damaged

commodities due to increase in imports from FTA partner countries.


of the production value in the advanced countries. The DDA has agreed to reduce

the allowable limits of AMS and de-minimis, but the agreement was not

concluded. Therefore, the adjustment coefficient can not exceed 1, and if the total

amount applied is equal to or less than the allowable payment amount, the import

contribution rate is to be the adjustment coefficient.

The effects of the opening of the agricultural market have long been a ma-

jor research topic for economists. Early studies aimed at verifying the social effi-

ciency of trade by measuring the loss of related interest parties due to opening of

imports using the partial equilibrium model (Arzac and Wikinson 1979; Freebairn

and Rausser 1975; Kulshreshtha and Wilson 1972; Brester and Marsh 1983).

However, it has been pointed out that the use of partial equilibrium analy-

sis has a limited impact on the spillover effects of market opening in certain

industries. Since the late 1980s, general equilibrium analysis has been spotlighted,

studies have been actively conducted to estimate the spillover effect of policy

changes including trade liberalization in the agricultural sector on the overall econ-

omy using the computable general equilibrium (CGE) model (Kenny 1990;

Robinson et al. 1989; Hertel and Tsigas 1988; Shoven and Whalley 1984).

In Korea, there has been an attempt to analyze the impacts of agricultural

market opening such as UR, DDA, and FTA through general equilibrium analysis

led by national research institutes. However, since the share of agriculture in the

whole economy is small, it is difficult to measure the effect of individual com-

modities caused by trade liberalization. Park et al. (2000) used the CGE model to

measure the impact of domestic livestock industry on market opening. However,

it pointed out the difficulty in measuring the effects with subdivided production-in-

put coefficients of livestock industries with a small share in the overall economy.

Domestic researches carried out in the meantime can be divided into

pre-FTA and post-FTA studies. Pre-FTA studies are approaching from the supply

side assuming that imports and domestic products are homogeneous (Eor et al.

2004; Kim et al. 2004). On the other hand, post-hoc analyzes are evaluating the

effects of market opening by identifying alternatives in terms of demand based on

the heterogeneity of domestic and imported goods, which are more realistic as-

sumptions (Kim 2006; Kim, Yoon-Sik and Choi, Seo-gyun 2007; Choi et al. 2009;

Ahn and Im, 2011; Moon et al., 2013). However, in order to quantitatively meas-

ure the impact on domestic price declines due to import increase, it is necessary

to establish a partial equilibrium model for each item and to estimate demand and


supply on domestic goods and demand on imported goods. Based on these esti-

mates, the relationship between increased imports from FTA-contracting countries

and domestic price declines can be identified.

Ⅱ. Mechanism of price decrease due to increase in imports

When analyzing the impact of FTAs on a particular commodity market, there are

two main points of view of domestic changes caused by tariff cut. There is a way

to approach from the supply side that imports directly increase the domestic supply,

and there is a way to approach from the demand side that replaces some of the

domestic demand (Choi and Kim 2007). From the former point of view, imports

and domestic products are regarded as homogeneous commodities, so if tariffs are

lowered and imports increase, this directly leads to an increase in domestic supply.

Under this approach, the impact on the domestic market is highly dependent on the

price difference between domestic goods and imports.

On the other hand, the approach on the demand side is that imported

goods and domestic goods are not the same goods, and imported goods are consid-

ered as substitutes for domestic goods. If imports increase due to tariff cuts, do-

mestic supply will not increase but replace some of domestic demand. The degree

of substitution depends on the size of the cross-price elasticity. Under this ap-

proach, the increase in imports of foreign agricultural products is interpreted as re-

placing some of the demand for domestic agricultural products, so the demand

function shifts to the left. Therefore, the effect of the domestic industry depends

on the degree of substitution between imported goods and domestic products, so

that the price difference between imported and domestic agricultural products is

less important than the former approach.

The increase in imports of agricultural products due to FTAs and the re-

sulting impact of the domestic market will generally follow the process as shown

in Figure 1. The most important factor in moving the graph in Figure 1 is the size

of substitution effects between imported and domestic goods. If we can see these

movement in real life, only the imports increase caused by FTAs would affect the

market and all other exogenous factors affecting the market have not changed.


FIGURE 1. Impact of FTA on the domestic market

Consider beef as an example. If domestic consumers perceive imported

beef and domestic beef as perfectly homogeneous goods, and Korea’s import de-

mand for beef is so small that it does not affect international prices, the FTA ef-

fects of tariff reductions can be expressed as in Figure 2.

As shown in the left graph of Figure 2, Korea is a small country and

therefore has a fixed, fully elastic global supply function () at the price of im-

ported beef (). Since domestic and imported goods are perfectly substitutes in

domestic market, equilibrium quantity and price are determined in the market un-

less they are separated. Therefore, before the FTA takes effect, the import price

() of beef, which is subject to higher tariff, becomes the equilibrium price of

the domestic beef market (). The domestic production is determined at the inter-

section of this price and the supply function () of domestic beef, and the import

quantity () is determined at the intersection of this price and the ED function.

FIGURE 2. Tariff reduction effects (assuming perfect substitute and small country)


Let's look at the changes in the supply-demand when the FTA is con-

cluded and tariffs are reduced (or abolished). First, if the tariff is reduced as much

as , the import price of imported beef will fall from to

, and the price

of domestic beef will fall as much as the drop in import price. If the other con-

ditions are stable, the import volume will increase to on the fixed excess de-

mand function, and the domestic beef production will decrease to .

However, the reality of domestic beef market is different from this assumption.

First, since the import portion of beef is not negligible, it can not be said that the im-

port quantity of beef does not affect the price change of the global beef market.

Therefore, assuming Korea as a small country in the global market for beef could lead

to errors that would make the tariff reduction effect more significant than actual.

Unless Korea is a small country in the international market, the supply func-

tion of international beef () becomes upward-sloping as shown in the left graph in

Figure 3. If Korea increases beef from the global market, the price of beef in the

global market will also rise. As a result, the increase in imports due to tariff cuts

is reduced compared to the case of the small country assumption. The greater the

share of imports of beef in Korea, the steeper the slope of the supply function. As

a result, the effect of imports increase caused by tariff reduction becomes smaller.

The unrealistic assumption that imported and domestic beef are perfect

substitutes can lead to even greater errors in assessing the effects of tariff cuts.

In the domestic market, imported beef and Korean beef (Hanwoo) are in an in-

complete substitute relationship rather than a perfect substitute. In other words, do-

mestic consumers are somewhat heterogeneous in recognizing two different types

of beef in terms of taste, meat quality, marbling, and food safety. As a result,

there are separate markets for two beef. In the case of imported and domestic in-

complete substitutes, the impact of tariff cuts will be smaller than that of perfect

substitutes. Even though imports are increased, demand for domestic products does

not decrease accordingly. The weaker the substitute relationship between imported

and domestic products, the weaker the impact of the tariff cut.

The graph on the right side of Figure 3 shows that the demand curve of

imported beef in the incomplete substitute relationship is shifted to the left side,

but the movement is smaller than that of perfect substitute. As the demand curve

shifts slightly down compared to the case of complete substitute, the market price

of domestic beef also declines. Likewise, the decrease in domestic production will

be smaller than the increase in beef imports due to tariff cuts ( ≻ ).


FIGURE 3. Tariff reduction effects (assuming incomplete substitute and large country)

As a result, it is important to consider the degree of substitution in the

analysis of the impact of the tariff cut on beef. This can be seen through the

cross-elasticity of domestic and imported beef. Therefore, if the main purpose of

the study is to estimate the changes in the price and quantity of the individual

market and examine the changes in welfare of economic entities, rather than meas-

uring the effects on specific industries such as agriculture and livestock industry,

applying the partial equilibrium model considering the supply and demand factors

of the products can lead to more persuasive analysis results.

Ⅲ. Equilibrium displacement model

Considering the fact that the condition required for the direct payment of damages

is specified in Article 7, Paragraph 1 of the Special Act, a comparative static anal-

ysis is appropriate which can explain the factors that change the market equili-

brium between two specific item points.

It is necessary to establish the concept of “import contribution rate” as the

ratio between “price drop rate caused by increase of import due to FTA” and “real

price drop rate of agricultural product between two time points”. The definition

of “import contribution rate” described in this paper is the relative share of con-

tribution of the increase in imports from the FTA partner countries to the fall of


domestic prices.

We set up a partial equilibrium model that takes into account the supply

and demand system for each good to measure the imports contribution. As men-

tioned above, the partial equilibrium model is easy to quantitatively measure the

relationship between FTA implementation and price decline by identifying not on-

ly increased imports but also various other factors affecting the market price of

individual goods.

The equilibrium displacement model (EDM) is often used in empirical

studies to analyze policy effects because it is suitable for comparative static

analysis. EDM has the advantage of simulating changes in endogenous variables,

such as price and quantity, by changing exogenous factors of demand and supply

in individual markets. The theoretical model of the effects of increased imports on

domestic market prices following FTA implementation can be expressed as an

equation system composed of four functions as follows.

(1) : demand for imported good

(2) : demand for domestic good

(3) : supply for domestic good

(4) : market clearance condition

TABLE 1. Variables

variables description variables description

price of an imported good price of imported substitute

demand for domestic good income

price of a domestic good supply for domestic good

price of domestic substitute factor affecting demand other than price

factor affecting supply other than price price of input

demand for imported good

Using the rate of change of variables (

), we can rewrite equa-

tions (1) to (4) as follows.


(5)

(6)

(7)

(8)

The following is obtained by summarizing the equations (5) to (8) with

respect to .

(9)

The left side of the above equation represents the rate of change in do-

mestic prices. The first term on the right hand side indicates the degree to which

the change in imports contributed to the domestic price fluctuation. The second

term on the right side indicates the degree of change in income attributable to do-

mestic price fluctuations and the third term indicates the degree to which domestic

substitute price contribute to domestic price fluctuations. The remaining terms also

indicate the extent to which each variable contributed to domestic price volatility.

The equation above decomposes domestic price fluctuations by factors.

The imports contribution that this paper is interested in is the portion of the fluctu-

ation of imports to domestic price fluctuations. The extent to which fluctuations

in imports contribute to domestic price fluctuations are shown in the first term of

the right-hand side of Equation (9). Therefore, the ratio between actual change of

the domestic price and the first term of the right side is the import contribution

rate. This can be expressed as follows.

(10)


Figure 4 graphically illustrates the theoretical framework for measuring

the imports contribution of the individual market when various elasticity values are

given proactively. The graph shows the situation in which the equilibrium price

and the quantity fluctuate as the demand curve and the supply curve move in the

domestic market.

Let us first assume that imported goods (imperfect substitutes) are sold at

lower prices due to tariff cuts caused by the effects of FTAs. Demand for domes-

tic goods is reduced from to due to the fall in price of imported

commodity. As the demand declines, the equilibrium point shifts from to

in the domestic commodity market, and the market price of domestic commodity

falls from to . In the end, domestic production and consumption will also

decrease as import price drops due to tariff cuts.

However, as mentioned above, in addition to the direct effects such as tar-

iff cuts, implementation of the FTA also requires various investments to strengthen

marketing capabilities such as expansion of the domestic distribution network of

import-export companies and promotion and discounts of imported goods. In addi-

tion, as the consumption experience of imported products increases and the percep-

tion of domestic consumers increases, the market share of imported products may

gradually increase. In addition to the tariff reduction effect, the effect of this in-

direct FTA implementation can be ascertained through the increase in imports

from FTA partner countries.

FIGURE 4. Decomposition of price decrease


The elasticities derived from the import demand function suggest how

much the demand for domestic products decreases due to the increase in imports

from the FTA partner countries. That is, due to the increase in the volume of im-

ports, the domestic demand curve shifts from to . As the demand dimin-

ishes, the equilibrium of domestic market shifts from to , and the market

price of domestic commodities drops from to . After the FTA im-

plementation, the equilibrium quantity will also decrease due to the decrease in de-

mand for domestic products due to the increase in import volume.

However, this may not be a visible equilibrium in the actual market after

FTA implementation. This is because, besides the FTA implementation, there are

various factors that can change the demand and supply of the relevant commodity

market. Figure 4 shows an example where the supply curve shifts to the right-side.

This assumes that FTA implementation and supply increases only. If the crop of

the product is improved, or if the cultivation technique to increase productivity or

the introduction of new seed, the supply of domestic products increases, and the

equilibrium observed in reality becomes ′ . Therefore, the degree of contribution

of increase in imports from FTA partner countries to market price decline can be

estimated as .

Ⅳ. Empirical analysis

Using the import contribution rate formula, we estimate the contribution of US

beef to Hanwoo price decline in 2012. With the entry into force of the Korea-US

FTA in 2012, US beef imports as well as total beef imports have increased.

Reflecting the impact of increased beef imports, domestic beef prices and Korean

cattle prices have fallen to less than 90% of five year average. The US beef im-

port growth rate was 11.59% and the Korean beef price decline rate was 11.17%.

As a result, it met the criteria for direct payment program in 2013.

In order to measure the import contribution rate, we need values of

and in the equation (10) and the rate of the beef price

change. The above characters represent the own-price elasticity of import demand,

the cross-price elasticity of import demand, the own-price elasticity of domestic

demand, the cross-price elasticity of domestic demand, the price elasticity of do-


mestic supply, and change rate in beef import volume respectively.

To obtain the above parameters, we specify the following demand and

supply functions.

(11) ln ln ln ln

variables sources periods

domestic beef demand/populationNational Agricultural Cooperative FederationKOSTAT

1996~2012yearly

domestic beef consumer price/CPINational Agricultural Cooperative FederationBank of Korea

1996~2012yearly

US beef consumer price/CPIKorea International Trade AssociationBank of Korea

1996~2012yearly

GDP(real)/population

Bank of KoreaKOSTAT

1996~2012yearly

(12) ln ln ln ln ln


US beef demand/populationKorea International Trade AssociationKOSTAT

2008~2012quarterly

US beef consumer price/CPIKorea International Trade AssociationBank of Korea

2008~2012quarterly

Hanwoo price/CPINational Agricultural Cooperative FederationBank of Korea

2008~2012quarterly

imported beef consumer price/CPIKorea International Trade AssociationBank of Korea

2008~2012quarterly

GDP(real)/population

Bank of KoreaKOSTAT

2008~2012quarterly

(13) ln ln ln



Hanwoo beef supplyNational Agricultural Cooperative Federation

1996~2012yearly

wholesale price of domestic beef/CPINational Agricultural Cooperative FederationBank of Korea

1996~2012yearly

production cost/PPI

KOSTATBank of Korea

1996~2012yearly

The above equations are estimated to obtain the coefficients necessary for

the measurement of the import contribution rate. In the estimation process, the au-

tocorrelation detected and the problem was solved by using the AR process. The

estimation results are summarized below.

TABLE 2. Estimation results

coefficients values standard errors

-1.33 0.074

0.28 0.060

-0.65 0.146

0.32 0.120

0.64 0.051

Table 3 shows the estimates related to demand and supply elasticities

from the previous research carried out since 2000. There have been a number of

studies dealing with the domestic demand elasticities but few studies have esti-

mated the import demand elasticities.

TABLE 3. Elasticities from previous research

　　

domestic demand import demand supply

own-price elasticities

cross-price elasticities

own-price elasticities

cross-price elasticities

price elasticities

Jeong et al. (2006) -0.67 0.4~0.6 　　 0.60

Choi et al. (2006) -1.06 0.47 　　 0.49

Jeong et al. (2011) -1.06 　　　 1.11

Livestock Office of Agricultural Outlook Center

　　 -0.87 0.52 　

this study -1.33 0.28 -0.65 0.32 0.64


Jeong et al. (2006) estimates the elasticities of demand for domestic beef

using yearly data from 1970 to 2004. Choi et al. (2006) estimates the elasticities

of the main items in the process of studying the effects of the Korea-US FTA on

the domestic industry. The Livestock Office of Agricultural Outlook Center esti-

mates them using quarterly data from 1993 to 2003.

The estimates above have meanings of some random variables. Therefore,

by calculating the import contribution rate using properties of the distribution of

the elasticities, the interval of the results can be obtained and the probability that

the actual value is within the interval can be calculated. In the previous studies,

we do not have all the elements needed to calculate the import contribution rate.

Therefore, we estimated the 90% confidence interval of the import contribution

rate using the elasticities estimated in this study.

Using the elasticities and standard errors shown in Table 2, 10,000 ran-

dom elasticity combinations are extracted randomly by assuming a normal

distribution. Figure 5 shows the asymmetric distribution with a mean of 24.4. This

is because the import contribution rate formula is a nonlinear combination of the

elasticities. Using this result, we find that the mean of the import contribution rate

is 24.4% and the 90% confidence interval of the rate is 14.0% ~ 45.1%.

FIGURE 5. Distribution of the import contribution rate


Ⅴ. Conclusion and discussion

15 FTAs are currently in effect in Korea. The FTA implementation has the effect

of increasing the welfare of consumers by promoting international trade. It also

contributes to industrial development by increasing demand for highly competitive

industries. However, less competitive industries are exposed to intense competition

and face price declines and reduced output. Korea's agriculture is considered as

an industry with low competitiveness.

Various policies have been proposed to mitigate the damage faced by

agriculture. Among them, the Special Act was enacted on the basis of a

Korea-Chile FTA, aiming at improving the competitiveness of agriculture and

compensating the damage. The direct payment for damage is a program that com-

pensates a portion of the price drop if the price declines due to increased imports.

In the process of calculating the amount of compensation, the portion of the price

fall due to increase in imports is considered to be important. In 2013, it was de-

cided to reflect this part by the Committee for Supporting Farmers and Fishermen,

in the form of the import contribution rate in the calculation of direct payments.

The import contribution rate is defined as the portion of the impact of the

increase in imports to the price drop. To do this, it is necessary to disaggregate

the various components that make up the price decrease to identify the impact of

import increase. This paper uses EDM to decompose the elements that constitute

the price decline and measure the impact of the increase in imports. We propose

the method of estimating the import contribution rate as

In order to actually use the above income contribution derived from EDM,

various elasticity values are needed. Price elasticity of demand, cross-price elas-

ticity of demand, import price elasticity of demand, price elasticity of supply and

so on. There are two ways to obtain such elasticity. First, utilizing the results of

previous studies and second, estimating elasticity from the equations directly.

Concerns may arise from the fact that previous research results are based

on relatively old data, and there is concern that researchers’ subjectivity may be


involved in selecting various results. Direct estimation may raise doubts about the

lack of understanding of the commodity and the lack of objectivity of the result.

It is difficult to obtain satisfactory results by any methods.

Therefore, it is necessary to be careful in determining the elasticity to cal-

culate the import contribution rate. This is because the elasticity value can vary

greatly depending on the subject of the researcher. Therefore, the value should be

estimated by using updated data, but care should be taken to obtain reasonable es-

timates by consulting the relevant commodity experts.

REFERENCES

Arzac, E. R. and M. Wilkinson. 1979. “A Quarterly Econometric Model of United States

Livestock and Feed Grain Markets and Some of Its Policy Implications.” American

Journal of Agricultural Economics, 61, pp. 297-308. doi:10.2307/1239734

Brester, G. W. and J. M. Marsh. 1983. “A Statistical Model of the Primary and Derived

Market Levels in the U.S. Beef Industry.” Western Journal of Agricultural Economics.

pp. 34-49.

Choi, Sei-kyun, Yun-ski Kim. 2006. Evaluation of Impact of a Korea-Chile FTA on Korea

Agriculture. Korea Rural Economic Institute.

Choi, Sei-kyun, Yun-shik Kim, Yong-sun Lee, Joo-ho Song, Min-kuk Jeong, Gyu-dae Cho,

Young-su Cho, Dong-kyu Park, Byung-ryul Kim, Kyung-phil Kim, 2006. Korea-U.S.

FTA: Effects and Strategies in the Agricultural Sector. Korea Rural Economic Institute.

Choi, Sei-kyun, Tae-hun Kim, Dae-hee Chung. 2009. Evaluation of Compensation

Measures and Impacts of Implemented FTAs on Agricultural Sector. Korea Rural

Economic Institute.

Eor, Myoung-keun, Chung, Chung-gil, Kim, Bae-sung, Heo, Joo-nyung. 2004. Effects of

CJK FTA on Korean Agriculture. Korea Rural Economic Institute.

Farmer Service Center for FTA Implementation. 2013. “Annual Report on Agricultural

Products Subject to FTA Direct Payment.” Korea Rural Economic Institute.

Farmer Service Center for FTA Implementation, 2014, “Annual Report on Agricultural

Products Subject to FTA Direct Payment.” Korea Rural Economic Institute.

Freebairn, J. W. and G. C Rausser, 1975, “Effects of Changes in the Level of U.S Beef

Imports.” American Journal of Agricultural Economics. 57, pp. 676-688. doi:10.2307/

1238886

Hertel, Thomas and M. Tsigas, 1988, “Tax Policy and U.S Agriculture: A General

Equilibrium Analysis.” American Journal of Agricultural Economics. 70, pp. 289-302.

Jeong, Kyung-su, Byung-oh Lee, Jong-in Lee. 2006. “The Effect of Korea-US Free Trade

Agreement on Hanwoo Industry.” Korean Journal of Agricultural Management and

Policy. Vol. 33 No. 4: 1085-1095.

Jeoug, Min-kook, Duk Huh, Byung-joon Woo, Myoung-ki Lee, Hyun-joong Kim,

Hyung-woo Lee, Won-tae Kim. 2011. A Study on Improving a Livestock

Product-Distribution System for Stabilizing Inflation (Year 1 of 4). Korea Rural

Economic Institute.

Kenny, M. 1990. “CGE Modeling of Agricultural Policies.” Working Paper, Department of

Economics, Pennsylvanis State University.

Kim, Han-ho, 2004, “Analysis of the Impact of Korea-Canada FTA on Agricultural

Sector.” Research Institute of Agriculture and Life Sciences.


Kim, Yun-shik. 2006. “Measuring an FTA Impact in a Partial Equilibrium Model

Considering Substitution Effects: Application to Korea Beef Industry.” Korean Journal

of Agricultural Economics. 47(3), pp. 31-51.

Kulshreshtha, S. B. and A. G. Wilson. 1972. “An Open Econometric Model of the

Canadian Beef Cattle Sector.” American Journal of Agricultural Economics. 54, pp.

84-91. doi:10.2307/1237737

Park, Chang-won, Kyeong-soo Cheong, Suk-jin Cho. 2000. “Economic Effects of Market

Liberalization in Korean Livestock Industry.” Korean Journal of Agricultural

Economics. 41(3), 57-77.

Robinson, S., Kilkenny M., and Adelman I. 1989. “The Effect of Trade Liberalization in

Agriculture on the U.S. Economy: Projections to 1991.” In A. Stoeckel and D.

Vincent, eds. Macroeconomic Consequences of Farm Support Policies. Durhams, NC :

Duke University Press.

Shoven, John B., John Whalley. 1984. “Applied General Equilibrium Models of Taxation

and International Trade: An Introduction and Survey.” Journal of Economics Literature.

22, pp. 1007-1051.

Bank of Korea <http://www.bok.or.kr>.

Korea International Trade Association <http://www.kita.net>.

KOSTAT. <http://www.kostat.go.kr>.

National Agricultural Cooperative Federation <http://livestock.nonghyup.com>.




TRENDS IN SOUTH KOREA’S GRANTS-BASED AIDFOR AGRICULTURAL SECTORIN DEVELOPING COUNTRIES

LEE HYEJIN*

Keywords

agriculture, food security, official development assistance, Korea, ca-

pacity building

Abstract

Agriculture is the major income source in many developing countries.

Official development assistance (ODA) contributes to agricultural devel-

opment in those countries to alleviate poverty and hunger. Among the

significant ODA donors, the Republic of Korea holds a unique position

with its transformation from a recipient to a donor. The main objective of

this article is to examine Korea’s grants-based ODA disbursements to ag-

ricultural sectors for its contribution to agricultural development and food

security in its recipients. The data for analysis were collected from the

KOICA Statistics Service and OECD DAC Query Wizard for International

Development Statistics for agricultural sectors. Results showed Korea con-

tinued disbursing the largest share of its agricultural grants to Asia while

gradually shifting its investment to Africa. Other regions received rela-

tively small amounts of agricultural aid. However, within regional disburse-

ments to agricultural sectors, each region received distinct shares by aid

type, based on their needs and Korea’s national interest or aid policy.

For agricultural capacity-building, the analysis identified evolution of the

training program’s main focus over the last 25 years. This shift from tech-

nical capacity improvement to software one indicated Korea’s efforts to

better align its aid policy with international norms for aid effectiveness.

* Ph.D., Assistant Professor, Institute for International Development Cooperation, KonkukUniversity, Seoul, South Korea. e-mail: [email protected]

Journal of Rural Development 40(Speical Issue)126

I. Introduction

Rapid economic growth and increased agricultural productivity have contributed to

the decreasing global poverty and hunger. However, 767 million people or 10.7

% of the global population were estimated in extreme poverty, and 795 million

or 10.9 % undernourished in 2013 (FAO IFAD 2016; World Bank 2016). The ma-

jority of those in poverty and hunger live in developing countries. Across the

globe, South Asia and Sub-Saharan Africa are the two regions that suffer from the

severest poverty and hunger (IFPRI 2017). The extreme poverty rate, expressed as

the percentage of the population living below USD 1.90 per day in 2011 purchas-

ing power parity, was assessed 15.1 % in South Asia and 41.0 % in Sub-Saharan

Africa in 2013 (World Bank 2016). The two regions also recorded the highest

scores of Global Hunger Index (GHI) in 2016: 29.0 for South Asia and 30.1 for

Sub-Saharan Africa. The GHI scores between 20.0 and 34.9 indicate a serious lev-

el of hunger on a 100-point severity scale. The 2016 GHI score averaged across

the developing countries was 21.3 (IFPRI 2016).

To help mitigate poverty and hunger in developing countries, various

forms of global efforts have been made. Of them, food aid is one of the most

well-known forms of these efforts. It intends to improve food security and stim-

ulate economies of developing countries (Awokuse 2011; Murphy & McAfee

2005). According to the World Food Programme (WFP), Asia and Sub-Saharan

Africa received approximately 86 % of the total food aid in 2012, 23 % for Asia

and 63 % for Sub-Saharan Africa, respectively (World Food Programme n.d.). At

a national level, food aid enables developing countries to substitute for normal

spending on food imports, and generate extra foreign exchange. This extra foreign

exchange in turn can be used for non-food imports or repay foreign debts. At a

household level, it helps households sustain short-term food security, protect their

assets as a safety net, and insure against economic shocks (Tusiime, Renard, &

Smets 2013). However, its critics have raised questions about its contribution to

food security and economic growth in developing countries (Awokuse 2011; FAO

2006). The critics argue food aid subsidizes donors’ domestic interests rather than

assists recipient countries to improve their food security (FAO 2006). One reason

for this criticism comes from tied food aid, on which a donor places restrictions.

The restrictions may require the food to be obtained from the donor’s domestic

Trends in South Korea’s Grants-Based Aid for Agricultural Sector in Developing Countries 127

market, and the use of transportation and distributional services of the donor coun-

try contractors. With such arrangements, the donor country could capture a third

of all food aid resources (Awokuse 2011). Food aid, like other forms of foreign

aid, potentially encourages recipient governments to depend on the aid.

Consequently, this may discourage necessary policy reforms and create dis-

incentives for their agricultural development. Corrupt institutions also allow local

elites to benefit from food aid, instead of channelling it to the intended beneficia-

ries (FAO 2006). Given those issues of food aid, a more reliable option to im-

prove food security is to assist developing countries for building self-sufficient

agriculture.

Agriculture is the largest employer and major source of income for the

poor rural households in developing countries (FAO IFAD 2016). Growth in agri-

culture not only favours the poor directly but expands the poverty-reducing effects

to other sectors; it generates demands for other agricultural inputs and services,

and employs the landless poor (Kaya, Kaya, & Gunter 2013; Lynam, Beintema,

Roseboom, & Badiane 2016). Yet, developing countries face obstacles to invest

in agriculture, including budget shortage, government priorities shifted to other

sectors, unfavourable agricultural policies, changing global market, and depend-

ency on food imports and aid (Murphy & McAfee 2005).

Recognizing the obstacles and potential of agriculture, international donors

have invested their resources in agricultural sectors through the official develop-

ment assistance or ODA, which is bilateral and multilateral aid to promote eco-

nomic development and welfare of developing countries (OECD 2008). The collec-

tive ODA contributions to agricultural sectors peaked around from 1983 to 1986,

and stagnated through 2000. This downward trend in agricultural ODA was attrib-

utable to multiple reasons: high global food surpluses, low commodity prices, agri-

cultural aid fatigue, opposition from farm lobby groups, and changes in donor pol-

icies to social-sector investments (Kaya et al. 2013; Lynam et al. 2016). However,

beginning 2000, donors’ interest in agricultural development for food security

re-emerged due to the rising food prices and high-profile political commitment

with the United Nations Millennium Development Goals (Lynam et al. 2016).

As of September 2017, there are 30 members of the Organisation for

Economic Co-operation and Development (OECD) Development Assistance

Committee (DAC) (OECD n.d.-a). Of the 30 donor members, the Republic of

Korea (hereafter Korea) holds a unique position as the first country that success-


fully transitioned from an aid recipient to donor. The country became a member

of the OECD in 1996 and the OECD DAC in 2009 to be recognized as a sig-

nificant donor country (Chun, Munyi, & Lee 2010). Quantitatively, Korea in-

creased its ODA contribution to 0.14 % of gross national income or GNI in 2016

from 0.10 % in 2009. Qualitatively, Korea introduced a comprehensive ODA bill

in 2009 for a legal framework to guide the country’s ODA (Chun et al. 2010).

The Korea ODA comprises multilateral assistance, bilateral loans and grants. Of

the three, the Korea International Cooperation Agency (KOICA), established in

1991, provides grants that include transfers of cash, goods, and technical services

(KOICA 2011).

With the exceptional emergence of Korea as a new donor, many studies

have analysed the time-series data of the country’s overall ODA to compare it to

other major donor countries, identify determining factors of Korea ODA, or draw

policy implications among other research objectives (Choi 2010; Kim & Oh 2012;

Marx & Soares 2013). However, specific sectors and types of Korea ODA have

not been sufficiently explored. For this reason, the current article aims to examine

KOICA ODA with a particular focus on agriculture, forestry and fisheries (AFF)

as an aid sector, and AFF training programs as an aid type since the agency’s

establishment. To examine the sector and aid type, the data for analysis were col-

lected from the KOICA and OECD DAC statistics. This analysis intends to reflect

trends in Korea’s grants-based contribution to agricultural development in its recip-

ient countries.

II. Data Sources and Analysis

To investigate historical trends of Korea’s grants-based ODA to AFF, the time-ser-

ies data were collected from the KOICA Statistics Service, KOICA Annual

Reports and OECD DAC Query Wizard for International Development Statistics

(KOICA 2016, n.d.; OECD n.d.-b). The time period for the current study was set

from 1991 through 2015; KOICA was established in 1991, and the latest year for

the available data was 2015. The KOICA Statistics Service provided data on

KOICA ODA disbursements by region, aid type, sector, and other details. The

OECD data provided total Korea ODA to AFF including grants, loans and other


assistances. The OECD data for Korea ODA are assumed to share the same data

source as KOICA; Korea reports and submits its overall ODA data to OECD as

a member country (OECD n.d.-c). All KOICA and OECD data were analysed in

USD. Additionally, the AFF training programs were categorized into three based

on their main goal. This categorization allowed for the examination of thematic

and directional changes in those training programs during the past 25 years. The

findings from the data analysis are reported in the following section.

III. Results and Discussions

1. Overall trends of KOICA ODA disbursement since its establishment

For the regional KOICA ODA disbursement, Asia received the largest ODA or

40.9 % averaged across the years, followed by Africa (Figure 1).

Figure 1. KOICA’s ODA disbursements by region and year


In 1991 and 1992 however, Africa received larger shares than Asia did;

Africa received 24 % in 1991 and 22 % in 1992 while Asia 20 % and 18 %,

respectively. Interestingly, the category, others (for multi-county programs) re-

ceived the largest shares from 1991 to 1993. It is probable that KOICA disbursed

larger shares to the multi-county programs such as humanitarian aid while identi-

fying its strategic regions during the first years of its establishment.

Since 1994, Asia’s dominance has continued, as reflected in KOICA’s

2016 budget. The largest share or 45.6 % of KOICA’s 2016 budget was allocated

to Asia and Pacific Ocean (KOICA 2016). In comparison, Africa was allocated

with 31.7 %, Latin America and the Caribbean 11.5%, and Middle East Central

Asia 11.2 % (KOICA 2016). Although the geopolitical importance of Asia to

Korea remains strong, KOICA has gradually increased its ODA to Africa in recent

years. This increase resonates changes in Korean aid policies and its recognition

of greater aid needs to Africa (KOICA 2016). In line with this trend, Kalinowski

and Cho argue that the expansion to Africa reflects Korea’s resource diplomacy

to gain greater access to the continent’s natural resources, and follow China’s in-

crease in aid to Africa (Kalinowski & Cho 2012). Kim (2012) also makes a sim-

ilar argument that achieving resource security and promoting soft power are some

of the key factors for Korea’s Africa strategy (Kim 2012).

Between 2003 and 2007, Middle-East Asia received considerable shares of

KOICA disbursement, ranging from 21 % in 2007 to 39 % in 2004 (Figure 1).

These drastic increases coincided with the United States (US) invasion of Iraq in

2003. This suggested a temporary shift of Korean foreign policies in Middle-East

Asia as a close ally of the US in order to help stabilize the region. With increasing

efforts for reconstruction and peace-building in Middle-East Asia, the remaining

regions showed inverse trends with KOICA disbursements between 2003 and 2007

(Figure 1).

KOICA distributes its disbursement across seven aid sectors; health, edu-

cation, public administration, technology, environment and energy, AFF, emer-

gency relief, and others. Of those aid sectors, education received the largest share

followed by public administration and health, averaged across the years (Table 1).


Table 1. KOICA ODA disbursements by aid sector as % averaged from 1991 to 2015

Aidsector

Health EducationPublic

administrationTEE

§AFF ǂ

Emergency relief

Others #

Total

% 16.1 23.5 18.9 13.2 10.6 4.2 13.6 100

§: Technology, environment and energy, ǂ: Agriculture, forestry and fisheries, #: unclassified

Agriculture, forestry and fisheries received 10.6 %, only followed by

emergency relief (Table 1). Among the sub-sectors of AFF, agriculture received

averaged 83.7 %, fisheries 8.7 % and forestry 7.6 % (KOICA 2017). The Korean

agency continues prioritizing education, public administration and health as re-

flected in its 2016 budget allocations; public administration received 24 % of

KOICA total budget, education 22 % and health 20.4 % in 2016 (KOICA 2017).

2. Korea ODA to agriculture, forestry and fisheries

As partly shown in KOICA’s AFF ODA (Table 1), Korea did not distribute large

ODA disbursements to agricultural sectors. Of the total Korea ODA including

loans, subscriptions as well as grants, the AFF ODA ranged from the lowest 1.0

% in 2000 to the highest 15.3 % in 2012 (Table 2).

Table 2. Korea total ODA disbursements, Korea AFF-specific ODA disbursements and

shares of AFF sub-sectors as % of AFF ODA disbursements from 1991 to 2015

YearKorea total

ODA §Korea AFF

ODA §

% of AFF in Korea

total ODA

Sub-sectors of agriculture, forestry and fisheries (AFF)

% of agriculture in Korea AFF

ODA

% of forestry in Korea AFF

ODA

% of fisheries in Korea AFF

ODA

1991 136.0 1.6 1.2 - ǂ - -

1992 95.7 2.8 2.9 - - -

1993 59.8 3.4 5.7 - - -

1994 - - - - - -

1995 232.4 2.5 1.1 - - -

1996 - - - - - -

1997 220.0 6.3 2.9 - - -

1998 283.8 22.3 7.8 30.4 3.1 66.5


§: ODA disbursement in USD millions from the data source, OECD DAC Query Wizard

ǂ: Data unavailable

#: Average among available data

Two values for 1994 and 1996 in Table 3 were missing from the OECD

DAC Query Wizard. In 2003 and 2007, there was a drastic increase in Korea AFF

ODA compared to each 2002 and 2006 (Table 2). This might be Korea’s response,

either voluntary or peer-pressured, to the global food price crises around those

years, which affected many of the rural poor in developing countries (FAO 2009).

The segregated data for the three sub-sectors were only available from 1998 to

2015 (Table 2). Among the three sub-sectors, agriculture, forestry and fisheries,

agriculture received the largest share ranging from 30.4 % in 1998 to 96.8 % in

2012. The same trend was mentioned with KOICA’s grants-based ODA above. On

YearKorea total

ODA §Korea AFF

ODA §

% of AFF in Korea

total ODA

Sub-sectors of agriculture, forestry and fisheries (AFF)

% of agriculture in Korea AFF

ODA

% of forestry in Korea AFF

ODA

% of fisheries in Korea AFF

ODA

1999 365.7 4.2 1.1 61.6 14.6 23.7

2000 353.7 3.6 1.0 90.0 6.1 4.2

2001 264.3 5.2 2.0 69.8 13.9 16.2

2002 362.4 6.9 1.9 51.7 35.6 12.6

2003 415.0 47.2 11.4 74.0 4.3 21.7

2004 591.8 15.1 2.6 75.5 22.0 2.6

2005 712.8 44.7 6.3 88.2 10.4 1.3

2006 681.2 11.9 1.7 83.2 12.8 4.0

2007 1013.0 102.8 10.1 89.9 9.0 1.0

2008 1623.2 53.3 3.3 64.1 22.0 13.9

2009 1793.1 46.6 2.6 89.7 6.4 3.8

2010 1967.3 99.7 5.1 82.6 8.9 8.5

2011 1665.2 132.1 7.9 95.4 2.8 1.7

2012 1809.2 277.4 15.3 96.8 2.2 1.0

2013 2226.8 115.4 5.2 81.8 12.1 6.1

2014 2262.8 208.5 9.2 85.0 5.0 10.1

2015 2311.7 98.2 4.2 81.8 7.9 10.2

Average # 4.9 % 77.3 % 11.1 % 11.6 %

(continued)


average, agriculture received 77.3 %, forestry 11.1 % and fisheries 11.6 % of the

Korea total AFF ODA. Therefore, the grants, loans and multilateral assistances for

AFF mainly supported agriculture over forestry and fisheries. The dominance of

agriculture could be a result from the combination of the demands of the recipient

countries and Korean aid policies. Compared to agriculture in general, it can be

due to the lower appreciation of the two sectors for their contribution to food se-

curity, and smaller population sizes engaged in forestry and fisheries.

3. KOICA ODA to agriculture, forestry and fisheries

For KOICA’s AFF disbursements by region, Asia was a leading recipient with

50.5 % averaged across the years, followed by Africa with 29.4 % (Table 3).

Table 3. Shares of KOICA AFF disbursements by region as % averaged from 1991 to 2015

Region Asia AfricaLatin

AmericaMiddle-East

AsiaEastern Europe

and CIS §Oceania

Multilateral ǂ

Others Total

% 50.5 29.4 8.0 1.1 2.3 1.2 7.5 0 100

§: Commonwealth of Independent States, ǂ: UN agencies and other international organizations

In total, these two regions received about 80 % of KOICA AFF

disbursements. While Asia continued receiving larger AFF disbursements than

Africa, the gap between the two regions was closing in the recent two years.

Africa received 92 % of Asia’s AFF disbursement in 2014 and 90 % in 2015

(Figure 2). This trend may continue as the 2017 Korea ODA policy explicitly

mentions an overall ODA increase in Africa (Korea Official Development

Assistance 2016).


Figure 2. Disbursements to AFF as KOICA total, Asia and Africa from 1991 to 2015

The AFF disbursements were further dissected by aid type and region (Table 4).

Table 4. Shares by aid type and region as % of KOICA AFF disbursement averaged from

1991 to 2015

Aid typeRegion

Projecttype

Development consulting

Volunteer dispatch

Invitedtraining

Small grants §

Expert dispatch

PPP ǂ HA ǂ Total

Asia 60.5 5.0 15.6 9.3 0.3 1.0 8.1 0 100

Africa 52.2 4.8 18.1 16.5 1.4 0.5 6.5 0 100

Latin America

52.3 0 25.3 16.2 1.8 2.0 2.4 0 100

Middle-East Asia

13.3 0 2.0 35.1 47.3 0.9 1.4 0 100

Eastern Europe and

CIS #53.8 0 12.5 17.8 4.2 2.0 9.7 0 100

Oceania 28.7 0 19.6 36.7 14.2 0.8 0 0 100

Average ∫

43.5 1.6 15.5 21.9 11.5 1.2 4.7 0.0 100

§: Not exceeding USD 0.2 million per year

ǂ: PPP: Public-Private Partnership, HA: Humanitarian aid

#: Commonwealth of Independent States

∫: Average across region among available data

Multilateral cooperation was excluded as only applicable to multilateral organisations.

By aid type in AFF, the project type cooperation received the largest dis-

bursement on average, followed by the training programs and volunteer dispatch,


respectively. These distributions across the aid types in AFF were different from

the overall KOICA’s spending; across all aid types, KOICA as a whole disbursed

its funds in the order of project type (45.2 %), volunteer dispatch (17.6 %), and

training programs (10.5 %) averaged from 1991 to 2015 (Data not shown). Thus,

a noticeable difference in AFF was the higher shares of the AFF training

programs. Since a transfer of appropriate technology to the right targets can in-

crease agricultural productivity in a relatively short-term, the training programs

may be regarded more effective in agricultural growth and food security.

By region and aid type, Asia, Africa, Latin America, and Eastern

Europe and CIS (Commonwealth of Independent States) received the largest

AFF disbursement, or over 50 % for their project type cooperation (Table 4).

The project type cooperation is more comprehensive under a multiyear plan. It

may involve construction of agricultural infrastructure such as irrigation systems,

dams and roads, provide agricultural machinery, equipment and other inputs, and

deploy experts for consultancy. The project type as such requires substantial

funding compared to other aid types. Additionally, it produces tangible outcomes

comparatively in a short term. This can help convince the public for the ODA

expenditure, and increase international coordination with similar programs in the

same region.

In Middle-East Asia, small grants (not exceeding USD 0.2 million) re-

ceived the largest share of AFF ODA, and in Oceania the training programs.

While most of the regions received the considerable disbursements for the vol-

unteer dispatch ranging from 12.5 % to 25.3 %, Middle-East Asia did only 2

% (Table 4). This low level of the volunteer dispatch to Middle-East Asia was

likely due to the safety concerns and entry restrictions in some parts of this re-

gion compared to the others. Instead, Middle-East Asia received the large dis-

bursements for the small grants and training programs in AFF. Given that

Middle-East Asia received only 1.1 % of KOICA AFF on average by region

(Table 3), these two aid types were likely considered more cost-effective with-

out dispatching AFF experts or volunteers for safety concerns.


4. KOICA training programs for capacity-building in agriculture,

forestry and fisheries

The AFF training programs are designed to transfer appropriate AFF technology,

improve AFF research capacity in a short or medium term, and cultivate human

resources for agricultural growth in a long term. The training programs build

strong networks between a participating country and Korea as well.

On average, the AFF training programs in Oceania received the largest

share against its own regional AFF disbursement, 36.7 % whereas Asia did the

smallest, 9.3 % against Asia’s own (Table 4). Despite the smallest allocation to the

training programs, Asia was given the largest or second largest disbursement in the

absolute amount for the training programs, competing only with Africa (Figure 3).

Figure 3. Disbursements of AFF training programs by region from 1991 to 2015

Until 2006, Asia received the largest disbursement for the AFF training

programs. However, starting 2007 but 2008, Africa advanced Asia in the AFF

training funds. In 2012, KOICA spent almost twice on AFF training programs in

Africa against Asia (Table 5).


Table 5. Percentage of Africa’s AFF training program disbursement against Asia’s

Year 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

% 84.0 42.4 32.4 58.9 49.4 21.7 29.0 19.7 37.3 19.0 33.3 25.8 51.0

Year 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

% 25.4 12.9 45.3 107.1 84.8 146.3 146.5 134.5 199.6 139.5 131.3 128.7

Source: KOICA Statistics Service.

This trend is notable because the total AFF disbursement in Asia was always

larger than Africa. It indicates KOICA supported greater local needs or demands

for capacity improvement for agricultural growth in Africa. Or from a political angle,

the increase in the AFF training for Africa reflects Korea’s national interest in gaining

better access to export markets and natural resources in Africa. The training programs

invite trainees whom the recipient country and KOICA jointly select. Thus, inviting

high-level government officials from agricultural line ministries can initiate or

strengthen political ties between them. This is a reason some critics question effec-

tiveness of training programs, often one-off, short-term and mixed with political

intentions (de Rosa, Nadeau, Hernandez, Kafeero, & Zahiga 2016).

From 1991 to 2015, KOICA implemented total 501 AFF training pro-

grams (Figure 4). The AFF training programs were categorized based on their

main objective. This categorization allowed for the examination of changes in the

AFF training programs over the years. The first category is the management and

policy-oriented approach. Training programs that fell into this category emphasize

soft skills in policy formulation, management, leadership, or system building for

agricultural growth and food security. This category tends to invite policy-makers,

and high-ranking government officers and community leaders. The second category

is the production and technique-oriented approach. This category prioritizes im-

provement in agricultural skills and technology for production or processing. The

training programs in this category train working-level officers, technicians and

field researchers who are actually involved in technical operations of agriculture

and food. The third category covers the comprehensive and inclusive approach.

This category offers training programs for a broad rural development. These pro-

grams provide more comprehensive courses for rural community and social

development. As such, they cover a wide range of stakeholders including officials

at local or central governments, high or working levels, and community leaders

or members in rural communities.


Figure 4. Categorization of AFF training programs by three approaches and by two periods

Of the 501 AFF training programs, 220 programs fell into the managerial

and leadership-oriented approach, 160 into the production and technique-oriented

approach, and 121 into the comprehensive and inclusive approach (Figure 4).

Among the explanations for this distribution are; the managerial and leader-

ship-oriented programs had greater demands from the recipient countries; KOICA

considered capacity improvement in this area more critical than the other two; im-

planting the training programs in this category was more cost-effective; and it pro-

vided better opportunities for political networking.

To examine chronological changes in the main focus, the years were fur-

ther divided into the two periods, 1991 to 2006, and 2007 to 2015. The year of

2006 was selected for the two reasons: first, the 2005 Paris Declaration with em-

phasis on aid effectiveness and capacity building and second, KOICA’s 15th anni-

versary with its self-reflection and strategic rearrangement. For the first period, the

agency directed the training programs towards the production and technique-ori-

ented approach. Of the total 189 training programs during the first 15 years, 92

programs or 48.7 % covered the production and technique-oriented approach, 62

programs or 32.8 % the managerial and leadership-oriented approach, and 35 pro-

grams or 18.5 % the comprehensive and inclusive approach (Figure 4). The agri-

cultural sectors require hardware skills that increase productivity, decrease inputs

and operational costs, and improve market quality of produce. Those outcomes

from the technical training make a result-based evaluation of a training program

clearer. Therefore, when KOICA was pressured to show its aid effectiveness dur-


ing its early years, investment in the technical fields could appear more rewarding

than investment in longer-term impacts from improved software skills.

During the second period, or 2007 to 2015, total 312 training programs

were implemented. And the programs for the management and policy skills were

noticeably increased to 50.6 %. For the technical-skill, its share was 21.8 %, and

for the comprehensive and inclusive 27.6 % (Figure 4). This indicated a shift to-

ward software-skills building for agricultural development and food security. In a

short term, training for software skills can assist the trainees and their organ-

izations to better design and manage AFF programs. In a long term, they can

strengthen national AFF capacity through favorable policies and institution.

The number of the training programs with the comprehensive and in-

clusive approach increased during this second period. Integration of the ‘Saemaul

Undong’ or new village movement to the agency’s AFF program possibly con-

tributed to this increase. Saemaul Undong is the Korean rural development pro-

gram during the 1970s and 1980s. This program led to the successful increase in

agricultural productivity and reduction in the income gap between the urban and

rural areas of Korea. Accordingly, the model was promoted as an AFF training

program component. Although widely successful in Korea, introducing its own ru-

ral development model runs the risk; Korea’s own development model may fail

to consider contextual factors and challenges that are unique to rural regions in

developing countries (Chun et al. 2010). For instance, Lee and Lee (2014) and

Abafita et al. (2013) each identified differences in factors between Rwanda and

Korea, and Ethiopia and Korea for the successful implementation of the Korean

model from the social, political, economic, and cultural perspectives (Lee & Lee

2014; Abafita, Mitiku & Kim 2013). For Saemaul Undong to provide a useful

guideline over time, the model itself needs a transformation to be more relevant

to the current era and target areas (Kwon 2010).


Ⅳ. Conclusions

The ODA by nature is temporary and volatile. Yet, the international donor com-

munities continue evolving for better aid programs to mitigate poverty and hunger.

Of the different aid sectors, agriculture can be the engine of growth especially at

the early stages of economic development, and in the regions where reliance on

agriculture is high in economic terms. Thus, agricultural growth assisted by inter-

national aid provides an effective way to reduce poverty, enhance food security,

and accelerate social development. At the same time, donor aid policies are often

political choices, thus their aid policies may not mirror the needs and demands of

recipient countries for agricultural development and food security.

Of the new donors, Korea outstands for its unique development

experience. Many developing countries benchmark Korea for its transformation,

from a recipient to a donor, and from an agriculture-based to a knowledge-based

economy. For this reason, Korea ODA became the subject of this study to reflect

the historical trend and reality of Korea’s grants-based ODA to agricultural sectors

during the past 25 years.

While Asia was a leading recipient of KOICA’s AFF grants, the details

of AFF disbursements were distinct by region and aid type probably from the mix-

ture of agricultural needs and demands of different regions, and Korea’s national

interests. For the AFF training programs, their main objective appeared evolving

from the technical-oriented to software-oriented approach. This evolution indicates

that software capacity for good agricultural policies, effective leadership and man-

agement skills became as important as technical capacity for agricultural growth.

This shift may better align with Korea’s goal to sustain aid impacts in the long

run. Capacity development through training programs can be effective. But power

imbalance and political interests have a potential to abuse training programs partic-

ularly at the trainee selection stage. Besides, it is challenging to evaluate effective-

ness of the training programs, given the one-off and short-term nature of many

AFF training programs.

The current study explored part of Korea AFF ODA executed by KOICA

and its agricultural training programs. The results offer some valuable insights on

the trends of the specific sector and type of Korea’s grants-based ODA as many

studies have been carried out with the overall Korea ODA. Nevertheless, for a


comprehensive understanding of Korea AFF ODA, this study has limitations. First,

it did not fully analyze causes or relations of the trend changes made over time.

For instance, specifics of the investment increase in Africa need be identified with

such questions as; whether this increase in disbursements is concentrated only in

a few African countries, or simply more African countries are included; if then,

what the causes or criteria of the changes are; and how relevant the changes are

to their agricultural development. To investigate these questions, the AFF disburse-

ments could be dissected by each recipient country, its AFF status in economic

terms, or national AFF policies. Second, the AFF training programs can be further

analyzed by program duration, or gender ratio of participants. Such information

captures additional characteristics of the AFF training programs. Third, any

changes in bilateral loans and multilateral assistances for AFF should be examined

for a fuller picture of Korea’s contribution to agricultural development in its recip-

ient countries. If a country received smaller AFF grants from KOICA, yet larger

bilateral loans for its AFF sectors, an exclusive look at the grants can under-

estimate Korea’s contribution. Further research therefore would provide deeper in-

sights on Korea AFF ODA to better assist its recipient countries and conform to

international aid norms.

ACKNOWLEDGEMENTS

The author sincerely thanks Lee Haneul for the technical assistance for the data

collection and sorting at the KU Institute for International Development

Cooperation at Konkuk University, Seoul, South Korea.

CONFLICTS OF INTERESTS

The author declares no conflict of interests.

REFERENCES

Abafita, J., Mitiku, F., & Ryang, K. K. 2013. “Korea’s Saemaul Undong (New Village

Movement): A Model for Rural Development in Ethiopia?” Journal of the Korean

Society of International Agriculture, 25(3), 217-230. doi:10.12719/ksia.2013.25.3.217

Awokuse, T. O. 2011. “Food aid impacts on recipient developing countries: A review of

empirical methods and evidence.” Journal of International Development 23(4): 493-514.

doi:10.1002/jid.1680

Choi, J. W. 2010. “From a recipient to a donor state: Achievements and challenges of

Korea’s ODA.” International Review of Public Administration 15(3): 37-51.

doi:10.1080/12294659.2011.10805178

Chun, H. M., Munyi, E. N., & Lee, H. 2010. “South Korea as an emerging donor:

challenges and changes on its entering OECD/DAC.” Journal of International

Development 22(6): 788-802. doi:10.1002/jid.1723

de Rosa, C., Nadeau, A., Hernandez, E., Kafeero, F., & Zahiga, J. 2016. “Delivery

strategies to enhance the sustainability of training: lessons from the Food and

Agriculture Organization of the United Nations.” International Journal of Training and

Development 20(2): 164-176. doi:10.1111/ijtd.12074

FAO. 2006. The State of Food and Agriculture, 2006: Food Aid for Food Security? :

Food & Agriculture Org.

FAO. 2009. The state of food insecurity in the world: Economic crises-impacts and

lessons learned. Retrieved from http://www.fao.org/docrep/012/i0876e/i0876e00.htm

FAO IFAD. 2016. The State of Food Insecurity in the World 2015. Meeting the 2015

international hunger targets: taking stock of uneven progress. Food and Agriculture

Organization Publications, Rome.


IFPRI. 2016. 2016 Glober Hunger Index. Retrieved from http://ghi.ifpri.org/

IFPRI. 2017. 2017 Global Food Policy Report. Retrieved from

http://www.ifpri.org/publication/2017-global-food-policy-report

Kalinowski, T., & Cho, H. 2012. “Korea's search for a global role between hard economic

interests and soft power.” The European Journal of Development Research 24(2):

242-260. doi:10.1057/ejdr.2012.7

Kaya, O., Kaya, I., & Gunter, L. 2013. “Foreign aid and the quest for poverty reduction:

Is aid to agriculture effective?” Journal of Agricultural Economics 64(3): 583-596.

doi:10.1111/1477-9552.12023

Kim, E. M., & Oh, J. 2012. “Determinants of foreign aid: The case of South Korea.”

Journal of East Asian Studies 12(2): 251-274. doi:10.1017/s1598240800007852

Kim, S. 2013. Korea in Africa: a missing piece of the puzzle?. Korea. 13(15): 22.

KOICA. 2011. Korea International Cooperation Agency: 20 Years of KOICA. Retrieved

from http://www.koica.go.kr/upload/pr/annual/20anniversary_eng.pdf

KOICA. 2016. KOICA Brochure 2016. In Korea International Cooperation Agency (Ed.),

(pp. 7). Seongnam-si, South Korea: Korea International Cooperation Agency.

KOICA. 2017. KOICA's Mid-Term Sectoral Strategy 2016-2020. Retrieved from

http://www.koica.go.kr/

KOICA. n.d.. KOICA Statistics Service. Retrieved from http://stat.koica.go.kr

Korea Official Development Assistance. 2016. 2017 Korean Policies for International

Development Cooperation (In Korean). Retrieved from The Office for Government

Policy Coordination and Prime Minister's Secretariat of Korea

Kwon, H. J. 2010. Implications of Korea’s Saemaul Undong for International Development

Policy: A Structural Perspective.

Lee, E., & Lee, S. C. 2014. “Challenges of Saemaul ODA in African region.” The

Korean Journal of Local Government Studies. 18(2).

Lynam, J., Beintema, N. M., Roseboom, J., & Badiane, O. 2016. Agricultural research in

Africa: Investing in future harvests: Intl Food Policy Res Inst. doi:10.2499/97808962

92123

Marx, A., & Soares, J. 2013. “South Korea’s transition from recipient to DAC donor:

Assessing Korea’s development cooperation policy.” International Development Policy|

Revue internationale de politique de développement. 4(4.2): 107-142. doi:10.4000/pol

dev.1535

Murphy, S., & McAfee, K. 2005. US food aid: Time to get it right.

OECD. 2008. Is It ODA? Retrieved from https://www.oecd.org/dac/stats/documentupload

/IsitODA.pdf

OECD. n.d.-a. Development Co-operation Directorate (DCD-DAC): DAC members.

Retrieved from http://www.oecd.org/dac/dacmembers.htm

OECD. n.d.-b. OECD Query Wizard for International Development Statistics. Retrieved

from https://stats.oecd.org/qwids/


OECD. n.d.-c. OECD Development finance statistics: Frequently Asked Questions.

Retrieved from http://www.oecd.org/dac/stats/faq.htm

Tusiime, H. A., Renard, R., & Smets, L. 2013. “Food aid and household food security in

a conflict situation: Empirical evidence from Northern Uganda.” Food Policy. 43:

14-22. doi:10.1016/j.foodpol.2013.07.005

World Bank. 2016. Poverty and Shared Prosperity: Taking on Inequality. Retrieved from

http://www.worldbank.org/en/publication/poverty-and-shared-prosperity

World Food Programme. n.d.. Food Aid Information System: Quantity Reporting.

Retrieved from http://www.wfp.org/fais/quantity-reporting



145

Guidelines for Operating the Journal of Rural Development (Extract)

Established in November 2016

Amended in April 2017

Chapter 1. General Rules

Article 1 (Purpose) These guidelines aim to provide the details regarding the publication and

management of an academic journal, the Journal of Rural Development (hereafter JRD),

according to the Rules for Issuing Publications of the Korea Rural Economic Institute

(hereafter ‘Institute’).

Article 2 (Purpose of Publication) The JRD is published to achieve the following objectives.

1. Contribute to the academic development in the economy of agriculture, forestry, and

food and the rural sector in Korea

2. Promote the development and exchange of knowledge and information related to pol-

icies for the economy of agriculture, forestry, and food and the rural sector

3. Establish an academic foundation for the development of the economy of agriculture,

forestry, and food and the rural sector

Chapter 2. Operation of the Editorial Board

Article 3 (Purpose) The Editorial Board is established and operated to be in charge of all the

matters necessary to publish the JRD.

Article 4 (Composition and Term of Office) ① The Editorial Board consists of 10-20 mem-

bers in and outside the Institute, including the chairperson, and a secretary. The Director

of Planning & Coordination is an ex officio member.

② The President of the Institute appoints the chairperson among research directors or

outside experts, and can designate a distinguished overseas scholar in the related

field as co-chairperson.

③ The President of the Institute appoints employees of the Institute or outside experts as

board members, and appoints one of the people in charge of publishing as secretary.

④ The term of the chairperson and board members is two years, and they can be

reappointed. If they cannot perform their duties due to unavoidable circumstances,

146

they can be replaced during their term of office.

Article 5 (Roles of Editorial Board) The following are major roles of the Editorial Board:

1. Review of the rules for submission and examination of papers

2. Receipt of submitted papers

3. Decision as to whether a submitted paper is worthy of examination

4. Selection of referees and request for examination

5. Decision as to whether a submitted paper will be published

6. Matters related to editing and publication

7. Other matters necessary for publication and matters for which the President of the

Institute recognizes the necessity.

Chapter 3. Publication

Article 7 (Publication) The JRD is published in Korean four times a year, on March 21, June

21, September 21, and December 21. If many papers in English are submitted, a special

issue in English can be published once or twice a year.

Article 8 (Submission and Examination of Manuscript) ① A submitted manuscript should

comply with the examination guidelines of the JRD, and the Editorial Board decides

whether the manuscript will be published, according to the procedure specified in the

guidelines.

② The office in charge of publication performs other related tasks through discussion.

Article 9 (Manuscript Fees and Review Fees) ① If a manuscript submitted by an outside

author is approved for publication, the author can be paid a manuscript fee which the

Institute decided.

② If an outside referee reviews a submitted paper, he or she can be paid a review fee

which the Institute decided.

Article 10 (Publication Rights and Profit) The Institute holds the publication rights of a pa-

per published in the JRD, and the paper cannot be reproduced or reprinted in any lan-

guages without prior approval of the President of the Institute. Matters regarding the

transfer and lending of publication rights are decided by the JRD Editorial Board.

Article 14 (Order of Publication of Papers) ① The Editorial Board decides the order of pub-

lication of selected papers.

② If papers accepted for publication are too many to be published in one issue, some of

147

them can be published in the next issue.

Article 15 (Certificate of Acceptance) If a submitted paper is accepted for publication

through the final examination and the author requests its certification before publication,

a Certificate of Acceptance can be issued.

Chapter 4. Manuscript Submission

Article 16 (Qualifications for Manuscript Submission and Contents of Manuscript) ①

Anyone can submit his or her manuscript.

② Submitted manuscripts should be academic papers in Korean or English on topics in-

cluding the economy of agriculture, forestry, and food, rural society and welfare, re-

gional development, and international trade policy.

Article 17 (Length of Manuscript) ① The length of a manuscript should not exceed 20

A4-size pages (40 characters, 25 lines in Korean; 80 characters, 26 lines in English), in-

cluding an abstract and references, according to the format specified in the Manuscript

Preparation Instruction.

② Manuscripts should be prepared according to the attached Guidelines for Manuscript

Submission to the JRD.

Article 18 (Prohibition of Redundant Submission and Publication) A paper which was pub-

lished or is under examination in another academic journal at home or abroad should not

be submitted to the JRD. A manuscript which was disapproved for publication in the JRD

cannot be resubmitted to the journal. Also, a paper published in the JRD cannot be re-

produced in another publication discretionally.

Article 19 (Indication of Authorship) If a paper is written by two or more authors, the first

author should be indicated first among the authors, and the other authors should be list-

ed in order of their contribution to the writing of the paper.

Article 20 (Receipt of Manuscript) ① Manuscripts for the JRD are received all year round.

② The file of a manuscript should be sent to the following e-mail address: journal@

krei.re.kr.

③ Submitted manuscripts will not be returned.

Article 21 (Transfer of Copyright) The copyright of a manuscript accepted for publication is

transferred to the Institute, and the author of the manuscript shall sign and submit a

form. The Institute holds the rights to publish, distribute, or print the manuscript in an-

148

other medium.

Chapter 5. Examination

Article 23 (Appointment of Referees and Sending of Manuscript) ① Every submitted

manuscript should be checked through “Copy Killer,” the plagiarism checking system

operated by the Institute, and through “Paper Similarity Check” operated by the National

Research Foundation of Korea. Then, for every submitted manuscript, the chairperson

of the Editorial Board appoints three referees among the candidates recommended by

board members. The candidate referees must be experts in the field concerned

(amended April 2017).

② In selecting the referees, high regard should be given to fully consider the academic

speciality concerned. And to secure fairness in the examination process, an expert at

the same institution as the submitter must not be selected as a referee (amended

April 2017).

③ The personal information of the author cannot be disclosed in the manuscript under

examination.

Article 24 (Examination of Manuscript) ① The referee should disclose his or her assess-

ment of a submitted manuscript by marking it with one of the following four grades:

‘approve,’ ‘approve after revision,’ ‘reexamination after revision,’ or ‘disapprove.’ In ad-

dition, the referee must prepare a referee report and submit it to the Editorial Board by

a given date.

② If there arises a need to advise a revision of the manuscript, the referee must specifi-

cally state the contents to be revised.

③ If the referee wishes to grade the manuscript as ‘disapprove,’ he or she must state

the reason in detail.

Article 25 (Examination Criteria) ① The referee should review the manuscript according to

the following criteria:

1. Originality: Is the research topic or the method of analysis and approach new and orig-

inal?

2. Suitability of Research Method: Are the topic, methods of analysis, and approaches

appropriate?

3. Logic and Consistency of Reasoning: Are the composition and reasoning logical and

appropriate?

149

4. Feasibility and Objectivity of Analysis and Assessment: Is the analysis valid and based

on reliable and objective data?

5. Academic Contribution: Is the research expected to contribute to academic progress?

6. Accurateness: Are bibliographical materials and data correctly quoted and refer-

enced?

7. Completeness of Composition Elements: Does the manuscript have all the necessary

composition elements, such as keywords, abstract in Korean or English, introduction,

body, conclusion, and references?

8. Adherence to the Code of Ethics: Did the author observe the research ethics specified

in Chapter 6?

Article 26 (Management of Examination Results) ① If the referee’s assessment turns out

to be ‘reexamination after revision,’ then the Editorial Board should request the author

to amend the manuscript and submit a reply in response to the referee report. The Board

should ask for reexamination when it receives the reply and the revised manuscript.

② If the referee's assessment is ‘approve after revision,’ then the Board should request

the author to revise the manuscript and submit a reply in response to the referee report.

In this case, the revised manuscript shall be reviewed by the same referee.

③ If the referee gives a ‘disapprove’ grade, then the same referee should be exempt

from the reexamination task.

Article 27 (Appraisal of Examination Results) ① Referees should examine the manuscript

and grade the initial examination result into one of the following four categories:

1. Approve: good to be published as it is.

2. Approve after revision: needs a partial revision for publication.

3. Reexamination after revision: the same referee needs to reexamine after revision.

4. Disapprove: not suitable for publication.

② The final assessment of a manuscript should be made according to the grade table as

shown below. If the assessment falls into the category of ‘approve’ or ‘approve after

revision,’ then the manuscript will be published in the journal. However, if the ap-

praisal is ‘disapprove,’ then the Editorial Board should inform the author of the deci-

sion and the reason for the decision.

③ If the result of the first comprehensive examination is ‘reexamination after revision,’

then reexamination must proceed.

④ If the author of a manuscript rejects revision of his or her manuscript or a reply in re-

150

sponse to referees' opinions or does not submit a revised paper within 30 days for re-

examination, the reexamination result is 'disapprove.'

⑤ The result of a reexamination shall be classified into either one of the following

grades: ‘approve,’ ‘approve after revision,’ or ‘disapprove.’ Once the appraisal grade

is given at this stage, the decision will be final.

⑥ The Editorial Board makes a decision on publication of a manuscript by considering all

three referees' reports. The manuscripts to be considered for publication are the ones

whose total grade is marked as ‘approve’ or ‘approve after revision' as shown in the

grade table below.

⑦ If the comprehensive assessment of a manuscript is 'approve' or 'approve after re-

vision' but one referee grades the manuscript as 'disapprove,' then the Editorial Board

can request the author to make revisions to the manuscript. The Editorial Board can

also place the manuscript on the waiting list for the next issue.

⑧ If the manuscript that has passed the screening is found to have plagiarized all or part

of another person's manuscript, or has already been published in another academic

journal, then the Board should revoke the decision to publish the manuscript and han-

dle the matter according to Chapter 6.

※ Grade Table

Referee 1 Referee 2 Referee 3 Final Grade

Approve Approve Approve

ApproveApprove Approve Approve after revision

Approve Approve Reexam after revision

Approve Approve Disapprove

Approve after revision Approve after revision Approve

Approve

after revision

Approve after revision Approve after revision Approve after revision

Approve after revision Approve after revision Reexam after revision

Approve after revision Approve after revision Disapprove

Approve after revision Approve Reexam after revision

Approve after revision Approve Disapprove

Reexam after revision Reexam after revision Approve

Reexam

after revision

Reexam after revision Reexam after revision Approve after revision

Reexam after revision Reexam after revision Reexam after revision

Reexam after revision Reexam after revision Disapprove

Reexam after revision Approve Disapprove

Reexam after revision Approve after revision Disapprove

Disapprove Disapprove Disapprove

DisapproveDisapprove Disapprove Approve

Disapprove Disapprove Approve after revision

Disapprove Disapprove Reexam after revision

151

Article 28 (Presentation of Dissenting Opinion) ① The author who does not agree with the

referee report, including the request for revision and the reason for reexamination, can

express a dissenting opinion in writing to the Editorial Board. In doing so, the author has

to present sufficient reasons for the argument or empirical cases.

② The Editorial Board must review and deal with the author's contention. Also, the

Board can act as an anonymous mediator between the referee and the author. If the

gap between the differing views of the two remains unresolved, the Board shall

make a final decision as to the validity of the arguments.

③ The Editorial Board holds the right to dismiss any additional contention once the fi-

nal decision is made.

Article 29 (Protection of Personal Information) The personal information obtained from the

examination cannot be disclosed to anybody except the Editorial Board.

Article 30 (Other Issues of Concern) Decisions on issues that have not been touched upon

in the guidelines shall be made by the Editorial Board.

Chapter 6. Research Ethics

Article 31 (Purpose) This chapter aims to establish the standards of research ethics, which

authors who want to publish their papers, Editorial Board members, referees, and

Research Ethics Committee members should comply with, concerning the publication

of the JRD, the academic journal published by the Institute.

Article 32 (Subjects of Application) This chapter shall apply to an author who wants to pub-

lish his or her paper in the JRD (hereafter 'submitter'), Editorial Board members, refer-

ees, and Research Ethics Committee members.

Article 33 (Definition of Research Misconduct and Scope of Application) ① Research mis-

conduct (hereafter 'misconduct') includes fabrication, falsification, plagiarism, unrigh-

teous indication of authorship, and redundant publication as follows in proposing, per-

forming, or reviewing research, or in reporting research results. Misconduct does not in-

clude error, a minor mistake, or differences of opinion.

1. ‘Fabrication’ means making up data or research results and recording or reporting

them.

2. ‘Falsification’ refers to distorting research contents or results by manipulating research

materials, equipment, or processes, or by changing or omitting data or research results.

152

3. ‘Plagiarism’ means the appropriation of another person's ideas, logic, unique terms,

data, analytic systems and so on that are not general knowledge without appropriately

indicating the source, whether intentionally or not.

4. ‘Unrighteous indication of authorship’ refers to not bestowing, without good cause,

the authorship on a person who has made an academic or technical contribution to re-

search contents or results, or giving the authorship for the simple reason of showing

gratitude or courtesy to a person who has not done so.

5. ‘Redundant publication’ or ‘self-plagiarism’ means reusing all or part of one's pre-

vious work in a new work without adequate indication of the source, or using one's

past work beyond the amount socially accepted even with indication of the source.

6. Misconduct includes other actions far beyond the scope commonly permissible in

academia.

Article 34 (Standards of Research Ethics) ① Submitters should comply with the following

ethical standards.

1. Submitters should not commit research misconduct specified in Article 33.

2. (Quotation and Referencing) When citing a published academic material, submitters

should exactly indicate the fact according to the Guidelines for Manuscript

Submission to the JRD. An unpublished academic material can be cited only with con-

sent of the researcher who provided the information.

3. (Revision of Paper) Submitters should revise their papers according to the regulations

of the Editorial Board and submit the contents reflecting referees' opinions to the

Board.

4. Submitters should respect the opinion and examination result of the Editorial Board.

② The Editorial Board should comply with the following ethical standards.

1. (Board Members' Basic Duty) The Editorial Board should respect a submitter's per-

sonality and independence.

2. (Prohibition on Discrimination) The Editorial Board should fairly treat a paper sub-

mitted to the JRD based only on its quality and the rules for submission and examina-

tion of manuscripts, regardless of the submitter's gender, age, institution, and any

prejudice or private acquaintance.

3. Fair Request for Examination

a. The Editorial Board should request a referee with expertise in the field concerned

who can make a fair judgment to review a submitted paper.

153

b. When requesting the examination of the submitted paper, the Editorial Board

should provide the referee only with the content of the paper without the in-

formation that reveals the submitter's identity.

4. (Confidentiality) Editorial Board members should not disclose the information related

to a submitter or the content of a submitted paper to a person other than the referee,

nor should they use the content. However, the following cases are exceptions: with

the submitter's consent; for dealing with affairs regarding the assessment of the aca-

demic journal by the National Research Foundation of Korea; and according to the reg-

ulations of other legislations.

③ Referees should comply with the following ethical standards.

1. Sincere Examination

a. The referee should faithfully assess the paper which the Editorial Board sends,

within a period of time specified by the examination rules, and should notify the

Board of the assessment result.

b. If the referee cannot assess the content of the paper due to differences in specialty

or other personal reasons, he or she should immediately notify the Editorial Board

(or a board member) of the fact.

2. Fair Examination

a. The referee should fairly assess the paper according to the Examination Criteria of

Article 25.

b. If the referee grades the paper as ‘disapprove,’ he or she must state the reason

clearly.

3. Respect for the Submitter

a. The referee should respect the personality and independence of the submitter as

a professional intellectual.

b. Preparing a referee report, the referee should use respectful and polite expressions

and clarify his or her judgment on the paper. If the referee thinks that the paper

needs revision, he or she should also explain the reason.

4. (Confidentiality) The referee should keep confidentiality about the paper for

examination. The referee should not show or discuss it to or with another person ex-

cept when advice is essential for the appropriate assessment of the paper, and should

not disclose the content of the paper before its publication in the journal.

154

Article 35 (Bringing Up Violation of Regulations of Research Ethics) ① Concerning the

publication of the JRD, if doubt exists as to the violation of these regulations, anyone can

report the related matters to the chairperson or secretary of the Editorial Board.

Article 36 (Composition and Decision-making of Research Ethics Committee) ① If an is-

sue is raised according to the regulations of Section 1, Article 35, the chairperson shall

organize the Research Ethics Committee with five or more related experts recom-

mended by the Editorial Board.

Article 37 (Responsibilities and Rights of Research Ethics Committee) ① The Research

Ethics Committee has a responsibility to prove whether the regulations have been vio-

lated, and the author in question has a responsibility to prove his or her compliance with

the regulations.

Article 41 (Confidentiality about Subject of Investigation) People who participate in inves-

tigation and deliberation on whether the regulations have been violated, including

Research Ethics Committee members, should not reveal the content of the investigation

or the personal information of the author in question to the outside.

Article 42 (Disciplinary Measures) If the Research Ethics Committee judges the author to

have violated the regulations, the following disciplinary measures shall be applied.

① The author of the paper which was judged as plagiarism cannot submit a manuscript

to the JRD alone or jointly for a certain period of time.

② If plagiarism is judged after the publication of the paper, the paper will be officially re-

moved from the list of articles of the JRD.

③ The chairperson of the Editorial Board who received the report of the Ethics

Committee shall notify the author who violated the regulations of the facts of Sections

1 and 2. At the same time, the paper will be removed from the website of the

Institute, and this fact will be open to the public on the website.

④ Within 30 days after the completion of the work of Section 3, the chairperson of the

Editorial Board shall notify the National Research Foundation of Korea of the details

on the judgment of plagiarism and disciplinary measures.

⑤ Concerning the judgment of the violation of the regulations other than plagiarism,

disciplinary measures decided by the Research Ethics Committee will be applied.

155

Supplementary Rules

(Other Regulations) The President of the Institute decides matters not included in these

guidelines and the establishment and amendment of the guidelines through the deliber-

ation of the JRD Editorial Board.

Supplementary Rules (April 2017)

(Enforcement Date) These guidelines shall enter into force on the date of its approval from

the President of the Institute.

(Enforcement Date) These guidelines shall enter into force on the date of its approval from

the President of the Institute.

156

Guidelines for Manuscript Submission to JRD

1. The Journal of Rural Development (JRD) is an academic journal published by the Korea

Rural Economic Institute. The journal was officially registered to the Korea Research

Foundation (KRF) in 2005.

∙ Publication Frequency and Dates

- Four times a year (21st day of the following months: March, June, September,

December)

2. Submission Conditions, Topic, and Eligible Entries

∙ Submission Conditions: Anyone who agrees with the following terms and conditions

can submit his or her manuscript.

- The manuscript should not have been published in any other publication.

- The applicant should abide by the ethics code of the Institute.

- The manuscript can be open to the public in the form of a publication.

- The manuscript and the author’s profile can be registered to the KRF.

- The manuscript can be posted on the Institute’s website for public viewing in the form of a

PDF file.

∙ Topic: research and survey on agro-forestry economy and rural socio-economy.

∙ Eligible Entries: manuscripts which the Editorial Committee judges to have met 11s

examination criteria.

3. Manuscript Format and Style

∙ Length: less than 20 pages/A4, (12 pt., 80 columns × 25 lines)

∙ Please type your manuscript on MS-Word/Hancom office Hangul.

∙ Style: please refer to our website at http://www.krei.re.kr/web/eng in the order of

“Publication” → “Journal of Rural Development” → “Submission” → “JRD Style

Manual.”

∙ Annotation: When identifying the source of a quoted statement, please state the

source according to the Harvard (author-date) style of referencing. No comma is nec-

essary between the name of author and year.

<Example> (Fox et al. 1989; Choi 2004, 63-65).

157

∙ Bibliography Format

- The bibliography should be identified in the order of author’s name, year, title,

page/volume no./name of journal, publisher

- Double quotation is used to identify theses, booklets, seminar materials, and work-

ing papers.

- Book titles should be italicized.

<Example>

Fox, W.F., H.W. Herzog and A. M. Schlottman. 1989. “Metropolitan Fiscal Structure and

Migration.” Journal of Regional Science 29(2): 523-536.

∙ The following abbreviations should be avoided: ibid., loc cit., op. cit.

4. Submission

∙ Submissions are received all year round.

∙ Please submit your manuscript (or inquiries) to the person below.

JRD Editorial Office

Library and Publishing Team


Telephone: 82-61-820-2215

E-mail: [email protected]

∙ Manuscript Content Requirements

- Keywords

- Abstract: approximately 10 lines

The contents should include research purpose and method (2~3 lines), research re-

sults (4~5 lines), and implications or improvement suggestions (2~3 lines)

- Manuscript Title and Author’s Name(s)

- Correspondence: email address and telephone number

Documents

Korea Rural Economic Instituterepository.krei.re.kr/bitstream/2018.oak/22368/1/농촌... · 2019-01-16 · www. krei.re.kr Vol. 40 | Special Issue. Vol. 40 | Special Issue. Forest