An Exit Seminar: Sharing the Internship Experience at IRRI

7/29/2019 An Exit Seminar: Sharing the Internship Experience at IRRI

1/49

An Exit Seminar: Sharing theInternship Experience at IRRI

11.Mar.2013 6th SSD Division Seminar

Hogeun Park, Intern, SSDMS Candidate, Seoul National University

Supervisor: Dr. Taku W. TsusakaThanks: Dr.Val O. Pede

1


2/49

2

AGENDA1)Major Contents

The Linkage between Social Relationship and Behavioral

Spillover, The Case of Irrigated and Rainfed Rice Farmers in

Bohol,

Finding the Effect of Canal Irrigation on Farmers Altruism

and Intolerance, using the Method of Hierarchical LinearModeling,

Preliminary NMRice Simulation Study using IRRI-MICRA

Baseline Survey.

2) Additional Contents

Field experience in Bohol

Life in SSD, IRRI


3/49

3

The Linkage between Social Relationship and Behavioral Spillover,

The Case of Irrigated and Rainfed Rice Farmers in Bohol,

The focus of this paper is the spillover effect throughsocial relationship.

What kind of social relationship transmits

behavioral spillover

Combination of (1) behavioral game experiment,(2) household survey, and (3) spatial econometrictechniques.

Using Spatial econometric technique for

investigating social relationship spill over

effect

1


4/49

4



Endogenous Social Effect(or Spatial Lag Effect)

Yi = Behavior(e.g. altruistic behavior)

Yi

Xi

Xi = Individual Profile(e.g. Age)

Xj

Xj

XjXj

YjYj

Yj Yj

i = Residual

Exogenous Social Effect (orCross Effect or ContextualEffect)

Correlated Social Effect (orPerturbation Effect)

j

j

i

j

j

Effects of Social Neighbors

1


5/49

5



Social Weight Matrix

Spatial Lag Operator

WX =X =

x1x2x3

xn

x averaged over social neighbors for obs1

x averaged over social neighbors for obs 2x averaged over social neighbors for obs 3

x averaged over social neighbors for obs n

n observations

n

n

00

0

11

0

0

W

row standardization

The construction ofWwill be based on different social relationships

e.g.Kinship Friendship Frequency offace-to-faceinteraction

WorkRelationship

1


6/49

6



Dictator Game

P 20ID. P 0

Your partner is in the OtherRoom

P 20P 20

P 20

P 20

Only you receive 100 PesosYour partner does not.

How much do you transfer to your partner

if your partner is someone in your village?

The amount youkeep is your payoffof this game

The amount your partnerreceives is his payoff ofthis game

Total P100

Since your partners payoff is totally dependent on your altruism,

the transferred amount is interpreted as a measure of your altruistic behavior.

1


7/49



Public Goods Game (Two Rounds)

1

The game is played by groups of 4 people: You and 3 anonymouspartners.

Each member is given P100.

Contributing some amounts to the group. The total amountcontributed will be doubled, and the doubled amount will be

shared equally among all members, regardless of your contribution.

We consider two variables in the analysis Message Receipt Dummy

Free-riding Index (FRI)

Group Members

Average Contribution

Your Contribut

ion

Message

Check

Dummy

Indicator for peer pressure on you

Indicator for degree of awareness of own free-riding

The contributed amount is recorded as the result of the second round, and is in

terpreted as a measure of your contributory behavior to public goods in the presence ofmonitoring mechanism.

with the same partners as in the 1st round.Play the 2nd Round


8/49



Our Study Site

1

Figure by Barkada Tours

Irrigated area (IR) and adjacentRainfed area (RF)

Similar agro-ecological, hydrological, and cultural background.

IR

RF


9/49



Our dataset consists of primary data in the following

categories.

1

Agricultural and Socioeconomic Data (X) 4 crop seasons from 2009 to 2010

Age/Gender Dummy/Years of Schooling Latest Season

Asset/Field Area/Household Size/Household Female Ratio

4-season Average

Social Network Variables (W) Oct. to Dec. 2012

Different Criteria. Different types of social weight matrix

can be defined. Behavioral Game Results (Y)

Sep. 2011

290 randomly selected farmers

Irrigated (N= 144) & Rain-fed (N= 146)


10/49



Constructing Social Weight Matrices

1

Using the social network variables, we define the social weight matrices (W), i.e., whoare the social neighbors and who are not.

For each of the 3 Samples Sample 1: Whole (IR + RF) Sample 2: Irrigated (IR) Sample 3: Rainfed (RF)

How do we choose social neighbors??

Criterion 1: How often do you meet with this person? 1 ifevery week or more, 0 otherwise

Criterion 2: How often do you meet with this person? 1 ifevery day or more, 0 otherwise

Criterion 3: Whats the relationship with this person 1 if the answer is close relative, 0 otherwise

Criterion 4: Have you hired him/her for agricultural labor in the past 3 years? 1 ifyes, 0 otherwise

Criterion 5: Have you exchanged agricultural labor with this person in the past 3 years 1 ifyes,

0 otherwise

Criterion 6: Have you ever participated or will you participate in a wedding ceremony of this persons

family? 1 ifyes, 0 otherwise

We will examine the 6 criteria for choosing social neighbors, for each of the 3 samples.

In this paper, we symmetrize the social relationship (e.g., if farmer A says he meets farmer Bevery day, we assume farmer B meets farmer A every day).


11/49



Social-Spatial Model Identification

1

Below are the spatial models suggested for each case

by spatial diagnostic tests (LM tests).

Social Weight Social Weight 1 Social Weight 2 Social Weight 3 Social Weight 4 Social Weight 5 Social Weight 6

Description Meet every week Meet every day Close Relative Hired labor Exchanged labor Wedding Ceremony

Sample Whole IR RF Whole IR RF Whole IR RF Whole IR RF Whole IR RF Whole IR RF

Dictator

GameCross Cross Cross

Lag

&

Cross

Lag

&

Cross

Cross Cross Cross Cross Cross Cross Cross Cross

Lag

&

Cross

Cross Cross Cross Cross

Public Goods

Game, R1

Cross Cross Cross

Lag

&

Cross

Cross Cross Cross

Error

&

Cross

Cross Cross Cross Cross Cross Cross Cross Cross Cross Cross

Public Goods

Game, R2

Lag

&

Cross

Lag

&

Cross

Cross

Error

&

Cross

Lag

&

Cross

Cross

Error

&

Cross

Cross Cross Cross Cross Cross

Error

&

Cross

Lag

&

Cross

Cross Cross

Lag

&

Cross

Error

&

Cross


12/49



Social Spatial Regression Results (1)

1Statistical Significance:*** 1 %, ** 5%, * 10%, 15%.

Dictator GameSocial Weight Social Weight 1 Social Weight 2 Social Weight 3 Social Weight 4 Social Weight 5 Social Weight 6

Relationship Meet every week Meet every day Close Relative Hired labor Exchanged labor Wedding Ceremony


Model Cross Cross Cross

Lag

&

Cross

Lag

&

Cross

Cross Cross Cross Cross Cross Cross Cross Cross

Lag

&

Cross

Cross Cross Cross Cross

Endogenous

Social Effectn/a n/a n/a

0.220***(0.009)

0.270**(0.012)

n/a n/a n/a n/a n/a n/a n/a n/a0.291***(0.009)

n/a n/a n/a n/a

CorrelatedSocial Effect

n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a

Public Goods Game, Round 1Social Weight Social Weight 1 Social Weight 2 Social Weight 3 Social Weight 4 Social Weight 5 Social Weight 6



Model Cross Cross Cross

Lag

&

Cross

Cross Cross Cross

Error

&

Cross

Cross Cross Cross Cross Cross Cross Cross Cross Cross Cross

Endogenous

Social Effectn/a n/a n/a

0.091

(0.265)n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a

Correlated

Social Effectn/a n/a n/a n/a n/a n/a n/a

0.191

(0.113)n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a

While Meet every week (or more) relationship does not lead to the spillover ofaltruistic behavior, Meet every day relationship seems to do so.

As found in our previous study, no behavioral spillover is found in rainfed areas.


13/49



Social Spatial Regression Results (2)

1Statistical Significance:*** 1 %, ** 5%, * 10%, 15%.

Under the influence of monitoring, contributory behavior spills over through different types of socialrelationship, particularly in irrigated areas.

Peer pressure seems to effectively increase contribution, which is robust to different social weights.

Voluntary correction of contribution is also found in many cases.

Public Goods Game, Round 2Social

WeightSocial Weight 1 Social Weight 2 Social Weight 3 Social Weight 4 Social Weight 5 Social Weight 6



Model

Lag

&

Cross

Lag

&

Cross

Cross

Error

&

Cross

Lag

&

Cross

Cross

Error

&

Cross

Cross Cross Cross Cross Cross

Error

&

Cross

Lag

&

Cross

Cross Cross

Lag

&

Cross

Error

&

Cross

Endogenous

Social

Effect

0.222*(0.051)

0.348***(0.003)

n/a n/a 0.171**(0.048)

n/a n/a n/a n/a n/a n/a n/a n/a 0.124*(0.071)

n/a n/a 0.234**(0.017)

n/a

Correlated

Social

Effect

n/a n/a n/a0.150*(0.100)

n/a n/a0.154*(0.052)

n/a n/a n/a n/a n/a0.223***(0.004)

n/a n/a n/a n/a0.224

(0.232)

Message

Receipt

Dummy

7.433**(0.013)

6.622*(0.057)

9.178*(0.089)

7.000**(0.022)

4.349

(0.226)

10.918*(0.053)

7.890**(0.016)

8.105*(0.051)

10.604*(0.053)

9.560**(0.025)

6.851

(0.228)

11.654*(0.078)

12.65***(0.001)

13.03***(0.008)

12.178

(0.105)

7.388**(0.024)

6.690*(0.079)

5.338

(0.298)

Free-

RidingIndex

0.197**(0.034)

0.123

(0.315)

0.247

(0.109)

0.199**(0.034)

0.122

(0.349)

0.176

(0.232)

0.256**(0.012)

-0.058

(0.698)

0.231

(0.132)

0.162

(0.220)

-0.015

(0.944)

0.167

(0.349)

0.203

(0.107)

0.343**(0.022)

0.221

(0.314)

0.243**

(0.033)

0.081

(0.553)

0.207

(0.114)

MRD x

FRI

Interaction

-0.233

(0.256)

0.056

(0.809)

-0.668*(0.096)

-0.183

(0.384)

0.145

(0.556)

-0.739*(0.066)

-0.460*(0.059)

-0.179

(0.487)

-0.595

(0.158)

0.072

(0.811)

1.288***(0.007)

-0.622

(0.168)

-0.010

(0.971)

0.517

(0.118)

-0.355

(0.451)

-0.191

(0.403)

0.153

(0.564)

-0.520

(0.151)


14/49



Concluding Remarks

1

I. Altruistic behavior spills over through Meet every day relationship butnot through Meet every week relationship, indicating the role of frequentface-to face communication in the emergence of social norm, i.e., behaveas others behave.

II. On the other hand, LaborHiring relationship does not result in behavior

al spillover, while Labor Exchange does, which may reflect the fact thatthe former relationship is more or less businesslike whereas the latter isbased on a mutual cooperation mindset.

III. As found in our geographical neighborhood effect study, the spillover ofcontributory behavior drastically increases once monitoring system isenforced, implying the importance of such a system in implementing publicwork.

IV. It is also confirmed that behavioral spillover is not found in rainfed farmingsocieties, suggesting the role of irrigation management in the

emergence of social norm.

Fi di h Eff f C l I i i F Al i d I l


15/49

15

Finding the Effect of Canal Irrigation on Farmers Altruism and Intolerance,

using the Method of Hierarchical Linear Modeling,

To investigate the connection between managementof canal (gravity) irrigation and farmers social behavior

(1) measures social behavior through behavioral

game experiments

(2) estimates the effects of irrigation, neighborhood,as well as individual characteristics.

Combination of 1) behavioral game experiments and2) hierarchical level model

The availability of irrigation water in the villagedoes not only improve agricultural productivity but

also enhances social relationship among farmers

2

Fi di th Eff t f C l I i ti F Alt i d I t l


16/49

16



Theoretical Framework

Behavioral game experiments are designed so as toquantify participants social behavior under strategic

situations (Gintis 2003).

Employing dictator game and ultimatum game, whichare developed to explore altruistic and retaliatingbehaviors, respectively

2



17/49

17



Dictator Game

This game is intended to elicit participants fairness,generosity, or altruism (Hoffman et al., 1996).

2

?100 PHP is equivalent to 2.46 (USD) by Bloomberg currency data, as of 31 January 2013. The Philippines

GDP per capita is $2,370 (2011) as per World Bank data. Given these exchange rate and GDP per capita,100 PHP is considered sufficient to ensure incentive compatibility for the experiment purpose.



18/49

18



Ultimatum Game

This game is interpreted as an indicator of thereceivers retaliating behavior or unwillingness totolerate the level of distribution (Herbert et al., 2003).

2

?x x100 PHP is equivalent to 2.46 (USD) by Bloomberg currency data, as of 31 January 2013. The Philippines

GDP per capita is $2,370 (2011) as per World Bank data. Given these exchange rate and GDP per capita,100 PHP is considered sufficient to ensure incentive compatibility for the experiment purpose.



19/49

19



Results for Behavioral Game Experiments

2

Type of Anonymous

Partner(1) Irrigated

Sample

(N=131)(2) Rainfed

Sample

(N=114)(3) t-test for

mean difference

|(1)-(2)|Dictator Game

Someone in Senders

Purok33.97 27.81 6.16**(20.59) (19.04) [0.015]

Someone in Senders

Barangay32.06 27.11 4.96*(21.58) (18.28) [0.053]Ultimatum Game

Someone in Senders

Purok 24.43 34.83 10.40***(15.15) (19.61) [0.000]Someone in Senders

Barangay25.12 34.47 9.36***(16.47) (21.29) [0.000]

Finding the Effect of Canal Irrigation on Farmers Altr ism and Intolerance


20/49

20



HLM Methodology

2

While ANOVA and OLS analyses are commonlyused in quantitative assessments, care must be takenwhen the data are nested (Raudenbush and Byrk 1993).

Frog-Pond Theory;Robinson(1950) the problem ofcontextual effects

Reference: J. Kyle Roberts., An introduction to HLM with Rhttp://faculty.smu.edu/kyler/training/AERA_overheads.pdf

Finding the Effect of Canal Irrigation on Farmers Altruism and Intolerance


21/49

21



HLM Methodology

2

Our data set covers randomly selected 238 rice farmerswho reside in 3 municipalities and 18 barangays

Altruistic and retaliating behaviors arise from socialatmosphere; we try to differentiate individual effects

from barangay effects

Employing HLM to account for the barangay-level

characteristics that are expected to affect individuallevel social behaviors



22/49


using the Method of Hierarchical Linear Modeling,2

Level 1 (Household Level)Variable N Mean SD Min Max

Age 238 51.38 12.06 14 87Schooling Years 238 6.33 3.02 0 14Asset Holding (Log PhP) 238 10.61 1.09 6.21 13.31Household Size 238 5.93 2.32 1 12.5Parcel Size (ha) 238 1.45 1.02 0.12 8.12

Level 2 (Barangay Level)Variable N Mean SD Min Max

Irrigation Dummy 18 0.61 0.5 0 1Age 18 51.3 4.5 43.56 61Schooling Years 18 6.37 0.93 4.46 8Asset Holding (Log PhP) 18 10.57 0.52 9.44 11.53Household Size 18 5.99 1.1 4.65 8.76Parcel Size (ha) 18 1.31 0.46 0.58 2.19

Descriptive Statistics



23/49



= 00 + +

ICC(Intra Class Correlation) =0

2

(02 +

2)

Random

Coefficient St. Dev.Variance

Component d.f. 2 p-value ICC

Dictator GameIntercept 1, u0 5.830 33.989 17 38.817 0.002 0.085Level-1, r 19.079 364.008

Ultimatum GameINTRCPT1, u0 6.668 44.463 17 49.456


24/49



[Level-1 Equation]Yij = 0j + 1j (Ageij) + 2j(Schooling Yearsij) + 3j (Assetij) + 4j(Household Sizeij)+ 5j (Parcel Sizeij) + rij

[Level-2 Equation]0j = 00 + u0j, 1j = 10 + u1j, 2j = 20 + u2j, 3j = 30 + u3j, 4j = 40 + u4j, 5j = 50 +u5j

Estimates for Level-1 Equations

Game

Type

0

(Intercept 1)Age

Schooling

YearsAsset

Household

Size

Parcel

Size

Dictator 28.789*** -0.268*** 0.109 -0.658 0.143 0.375

Ultimatum 28.117*** -0.067 -0.578* -1.984* -0.427 0.797

*** p < 0.01, * p < 0.10 1

23



25/49



Yij = 00 + 01 (Irrigation Dummyj) + 02 (Agej) + 03 (Schooling Yearj) + 04 (Assetj) +05 (Household Sizej) + 06 (Parcel Sizej) + 10 (Ageij) + 20 (Schooling Yearij) +30 (Assetij) + 40 (Household Sizeij) + 50 (Parcel Sizeij) + u0j + u1j (Ageij) + u2j (Schooling yearij) + u3j (Assetij) + u4j (Household Sizeij) + u5j (Parcel Sizeij)+ rij

Game

Type

00

(Intercept 2)

Irrigation

DummyAge

Schooling

YearsAsset

Household

Size

Parcel

Size

Dictator 23.387*** 9.053* 0.166 -0.259 4.348* -0.724 6.087

Ultimatum 39.092*** -14.012*** -0.697** -1.124 -8.585*** 0.885 -4.964

*** p < 0.01, ** p


26/49



The result is highly suggestive of the significant socialeffects of canal irrigation schemes.

The positive effect on altruism and the negative effect

on retaliation indicate that the type of social interactionspromoted by the necessity for collective irrigationmanagement leads to inducing the accumulation ofgood social behavior among farmers.

One clue to validating the irrigation effect is to considerthe existence of TSAs (turnout service associations)in the irrigated communities

Concluding Remarks



27/49



TSA- private canal construction- purchasing machinery- providing micro credit

Compared with the rainfed, irrigated farmers are

exposed to more opportunities to meet and discuss

public arrangements with their neighbors

Dual role: to boost the rural economy throughincreased production, and to accumulate socialcapital among farmers.

Concluding Remarks


2


28/49



Anecdotal Information

( Inday Salaum )

Cultivated cassava before irrigation project

Cultivating Hybrid Rice twice a year

Three children- Crop science- Veterinary- Agronomy

Promoting children back to village foragriculture

Several neighbors children already backto village for their career

Irrigation and modern agricultural technology can prevent

brain drain from rural areas.


2


29/49



Limitation

Our behavioral game experiments were conducted in 2011 which

was after the construction of irrigation. This survey structure

prevents us from formulating a difference-in-difference

estimator that ensures a more proper impact assessment.

Preliminary NMRice Simulation Study using IRRI-MICRA Baseline Survey

3


30/49

30

Preliminary NMRice Simulation Study using IRRI MICRA Baseline Survey

3

Introduction

In 2011, IRRI in collaboration with MICRA conducted the baselinehousehold survey on rice farmers over two crop seasons (dry andwet) in four provinces of the Philippines, namely, Bohol, Bukidnon,Pangasinan, and Tarlac. 240 rice and corn farmers that had

irrigated land were randomly selected.

NMRice is being developed on the basis of solid agronomicsciences, it has not been empirically examined to what extent thetool can actually contribute to improving agricultural productivity

and profitability at farm level.


3


31/49

31


3

Methodology

By comparing FP with NM, the sample farmers are divided into twoor three groups (depending on the criterion) in accordance with theproximity of FP to NM

Nitrogen Quantity Applied: FP is defined as NM-Close if the FP quantity is 80-120% of the NM quantity, NM-Mid if 50-80% or 120-150%, and NM-Far if below 50% orabove 150%.

Phosphorus Pentoxide Quantity Applied: FP is defined as NM-Close if the FPquantity is 80-120% of the NM quantity, NM-Mid if 30-80% or 120-170%, andNM-Far if below 30% or above 170%.

Potassium Oxide Quantity Applied: FP is defined as NM-Close if the FP quantityis 80-120% fo the NM quantity, NM-Mid if 30-80% or 120%-170%, and NM-Far ifbelow 30% or above 170%.

Timing of the First Application: FP is defined as NM-Close if the FP timing iswithin 3 days of the NM timing, and NM-Far if the FP timing differs from the NMtiming by more than 3 days.


3


32/49

32


3

Nitrogen (N) Quantity Applied (kg/hectare)

Average(Standard Deviation)

Region FP NM SampleSize (FP-NM) P-Value

Bohol 25.87(20.80) 60.10(18.47) 48 -34.23*** 0.000

Bukidnon 26.04(14.01)

78.30(30.24) 16 -52.26*** 0.000

Pangasinan 130.30(106.54) 65.58(24.61) 22 64.72*** 0.013

Tarlac 94.97(103.00) 77.62(20.84) 51 7.35 0.237

WeightedAverage 68.38(86.94) 69.63(23.22) 137 -1.25 0.868FP and NM show the average values with the standard deviation in the parentheses. *** p< 0.01 ** p< 0.05


3


33/49

33


3

Phosphorus Pentoxide(P2O5) Quantity Applied (kg/hectare)


Bohol 22.96(20.41) 16.99(4.66) 48 5.97*** 0.046

Bukidnon 7.79(10.52) 20.08(8.16) 16 -12.29*** 0.001

Pangasinan 6.85(20.13) 18.40(6.07) 22 -11.55** 0.014

Tarlac 4.28(14.19) 19.64(5.26) 51 -15.36*** 0.000

WeightedAverage 11.65(19.07) 18.56(5.68) 137 -6.91*** 0.000


FP and NM show the average values with the standard deviation in the parentheses. *** p< 0.01 ** p< 0.05


3


34/49

34


3

Potassium Oxide (K2O)Quantity Applied (kg/hectare



Bohol 16.70(19.71) 16.99(21.74) 48 -0.29 0.918

Bukidnon 6.54(10.15) 20.08(66.64) 16 -13.54*** 0.000

Pangasinan 3.82(15.08) 17.75(7.19) 22 -13.93*** 0.001

Tarlac 2.06(12.07) 18.90(46.90) 51 -16.84*** 0.000

WeightedAverage 7.99(16.65) 18.18(6.42) 137 -10.19*** 0.000FP and NM show the average values with the standard deviation in the parentheses. *** p< 0.01 ** p< 0.05


3


35/49

35

y y g y

3

Frequency of Fertilizer Applications



Bohol 2.06(0.41) 2.21(0.67) 48 -0.15 0.164

Bukidnon 2.00(0.89) 2.50(0.63) 16 -0.50** 0.041

Pangasinan 1.78(0.52) 2.35(0.49) 23 -0.57*** 0.000

Tarlac 2.29(0.61) 2.51(0.51) 51 -0.22 0.062

WeightedAverage 2.09(0.50) 2.38(0.67) 138 -0.29*** 0.000FP and NM show the average values with the standard deviation in the parentheses. *** p< 0.01 ** p< 0.05


3


36/49

36

y y g y

3

Timing of the1st Application

Region FP Early Late SampleSize (FP-NMEarly) (FP-NMLate)

Bohol 10.85(6.39) 0.00(0.00) 13.83(0.81) 48 10.85*** -2.98 ***

Bukidnon 16.25(9.03) 5.25(6.15) 14.88(1.02) 16 11.00*** 1.37

Pangasinan 5.30(7.96) 4.70(5.99) 14.26(1.95) 23 0.60 -8.96***

Tarlac 12.02(3.15) 0.00(0.00) 13.06(1.71) 51 12.02*** -1.03**

WeightedAverage

10.99(6.92) 1.39(3.86) 13.74(1.54) 138 9.60*** -2.75***

NM Range

FP and NM show the average values with the standard deviation in the parentheses. *** p< 0.01 ** p< 0.05


3


37/49

37

y y g y

3

Whose Practice is Close to NM Recommendation?- Coefficient of Correlation bet. Grouping & Factors


Region SampleSize

PlotSize

CornRotation Age Gender

SchoolingYears

IrrigationCost

Bohol 48 -0.387*** -0.021 0.147 0.130 -0.062 0.386***

Bukidnon 16 0.190 -0.289 0.011 -0.372 0.236 0.400

Pangasinan 22 -0.181 -0.224 -0.284 n/a -0.111 -0.003

Tarlac 51 0.080 -0.157 0.072 0.030 -0.215 0.148

WeightedAverage137 -0.012 -0.223*** 0.015 -0.015 -0.126 0.125

n/a cannot be computed because at least one of the variables is constant*** p < 0.01, ** p


38/49

38

y y g y

3


Phosphorus Pentoxide(P2O5) Quantity Applied (kg/hectare)



39/49

39

3


Potassium Oxide (K2O) Quantity Applied (kg/hectare)



40/49

40

3


Timing of the 1st Application



41/49

41

3

Would NM Practice lead to Better Productivity?- Coefficient of Correlation bet. Grouping & Factors




42/49

42

3


Phosphorus pentoxide(P2O5) Quantity Applied (kg/hectare)



43/49

43

3


Potassium oxide (K2O)Nitrogen Quantity Applied (kg/hectare)



44/49

44

3


Timing Of the 1st Fertilizer Application



45/49

45

3

Regression Model, Profitability

VariablesStandardizedCoefficients

P-Value

FP-NM Proximity: N Quantity 0.178* 0.07FP-NM Proximity: P2O5 Quantity 0.029 0.893FP-NM Proximity: K2O Quantity -0.083 0.695FP-NM Proximity: Timing of

the 1st Application 0.227** 0.031Parcel Size 0.051 0.628Corn Rotation 0.079 0.479Age (Household Head) 0.146 0.162Gender Dummy (Household Head) 0.007 0.944Years of Schooling (Household Head) 0.037 0.73

Bukidnon Dummy 0.054 0.648Pangasinan Dummy -0.311** 0.018Tarlac Dummy -0.163 0.244Observations 121R2 0.179Adjusted R2 0.071F-Statistic (P-Value) 1.656 (0.076)

**: p < 0.05, *: p < 0.10


3


46/49

46

3

Regression Model, Yield

VariablesStandardizedCoefficients

P-Value

FP-NM Proximity: N Quantity 0.005 .964FP-NM Proximity: P2O5 Quantity -.066 .767FP-NM Proximity: K2O Quantity 0.265 .217FP-NM Proximity: Timing of

the 1st Application 0.043 .921Parcel Size -.052 .684Corn Rotation 0.243*** .256Age (Household Head) -.095 .629Gender Dummy (Household Head) -.126 .033Years of Schooling (Household Head) 0.004 .367

Bukidnon Dummy 0.098 .185Pangasinan Dummy -0.106 .970Tarlac Dummy 0.220 .413Observations 125R2 0.158Adjusted R2 0.047F-Statistic (P-Value) 1.418 (0.157)

**: p < 0.05, : p < 0.15


3


47/49

3

Quantity of nitrogen applied per area, is correlatedpositively with farm profitability and negatively withoverall fertilizer cost per area.

Timing of the first application, the FP-NM proximity isagain positively and highly significantly correlatedwith farm profitability on aggregate, and particularlyin Pangasinan.

NM-generated nitrogen quantity and timing of the firstapplication would be beneficial for improving farmprofitability

Concluding Remarks


48/49

48


49/49

Thank you for your attention

Documents

An Exit Seminar: Sharing the Internship Experience at IRRI