Growth Accounting for Sri Lanka’s Agriculture with Special Reference to Fertilizer and Non-Agricultural Prices: Do Policy

Reforms Affect Agricultural Development?

Running Title: Growth Accounting for Sri Lanka’s Agriculture

Mitoshi Yamaguchi Professor and COE Program Leader

e-mail : yamaguchi@econ.kobe-u.ac.jp Ph/Fax : +81-78-803-6828

And

M. S. SriGowri Sanker

COE Researcher e-mail : sgsanker@econ.kobe-u.ac.jp

Ph : +81-78-803-6872, Fax : +81-78-803-6869

Graduate School of Economics, Kobe University 2-1 Rokkodaicho, Nada-Ku, Kobe

JAPAN 657-8501

This Research is partially supported by the Grants-in-Aid for 21st Century COE Program of Graduate School of Economics, Kobe University, Kobe, Japan.

1

Growth Accounting for Sri Lanka’s Agriculture with Special Reference to Fertilizer and Non-Agricultural Prices: Do Policy

Reforms Affect Agricultural Development?

Abstract

The agricultural sector of Sri Lanka reacted sharply to the highly contentious policy reforms called Structural Adjustment Programs. We used a four sector general equilibrium model under growth accounting approach to find out the effect of the policy (exogenous) variable on the target (endogenous) variable. Here, we considered only the most important variables, and the overall results indicate that policy changes are favorable to overall agricultural development, although their impact on the domestic food sector is negative. The most serious negative determinant under the policy changes relates to fertilizer, and our study indicates that fertilizer prices considerably affect agricultural production; it especially has a negative impact on domestic food production. Secondly, this paper analyzed the impact of non-agricultural price and found that it positively helped the development of overall agriculture. Thirdly, agricultural exports increased under the new policy reforms and made large contributions to agricultural production.

Key Words: Structural Adjustment Policy (SAP); Sri Lanka’s Agricultural Sector; Domestic Food Sector; Four-Sector General Equilibrium Growth Accounting; Growth Rate Multiplier (GRM). JEL Classification: O11, O13, O41, Q18.

2

I. Introduction

Many Asian countries are dependent on an agricultural economy. Sri Lanka is a

country which is located in the Indian Ocean and has a tropical climate. Rural

agricultural development and raising food production so as to achieve self sufficiency

were the main objectives of almost all the governments since independence. Although

substantial amounts of resources were allocated towards agriculture and irrigation

infrastructures for this purpose, the rural agricultural economy of Sri Lanka struggled to

gain momentum. Various problems such as the use of low quality seed, continuing use

of old cultivation practices, insufficient and improper use of fertilizers and agricultural

chemicals, inadequate agricultural production credits, a low rate of productivity along

with the absence of reasonably priced production inputs, as well as the inadequacy of

well-organized farmer-centered marketing facilities, further amplified this problem.

Since about 60%-70% of Sri Lanka’s population live in rural areas and depend on

agriculture and related activities, an increase in labor through population growth, would

contribute significantly to employment and income development in rural areas.

Meanwhile, rice and other subsidiary food items make up the majority of imports. Any

reduction in these imports could not only help to remedy the foreign exchange

imbalance, but could also allow these resources to be allocated for the import of goods

for other much needed development activities. This peculiar general scenario needed

urgent remedial measures. Therefore, restructuring the agricultural sector with special

3

reference to the development of domestic agriculture has been a major policy issue of

successive governments since independence. Though there were many policy issues,

the most important one among them would be the Structural Adjustment Policies

(SAP), implemented in 1977. Due to the reasons mentioned above, it is essential to

estimate the effect of SAP on Sri Lanka’s Agriculture. In this paper, we define the effect

of SAP through following two actions. First is the performance of four policy variables

such as increase of agriculture exports (E1), decrease of food imports (M2), increase of

fertilizer price (PF) and non-agriculture price (PN), and second is the consequent change

in the economic structure by SAP using General Equilibrium Growth Accounting

approach where the latter is measured through the change in the Growth Rate

Multipliers(GRM) which is explained in the section 4. This combined effect gives us

the picture of overall effect of SAP and we obtained rather positive result which is

different from previous studies.

II. Structural Adjustment Policy in the context of Srilanka’s Agriculture

A. Evaluation of Structural Adjustment Policy

Although there are many frameworks to study the tendency of the Adjustment

Policy issues, we considered Sarris’s analytical framework targeting the Asian

economic structure as the most appropriate one. As Sarris (1990) discussed, the

majority of the adjustment programs came from a prevailing or expected decline in the

external balance, due to factors not likely to be inverted in the short-run and from an

4

external deficit which was not sustainable in the medium term. Furthermore, these

adjustments required a shift in domestic demand and changing supply and the

production structure to eliminate the external deficit. Since demand can be reduced

more easily and faster by changes in the money supply and public expenditures, it

happens to be the focus of the first attempts at correcting economic decline. However,

the improvements of the supply side are more difficult and slower to remedy. Therefore,

there is a tendency for such efforts to be associated with medium term structural

adjustment efforts. As the removal of the external inequality is the crucial focus of

adjustment, trade policies appear significantly in all adjustment programs and they

usually include two sets of measures, such as export promotion and import

liberalization.

It is notable from various countries, which implemented adjustment policies to

develop their damaged economies that the results of the SAPs have created several

internal controversies. The most important one is whether the results are due to the

policy reforms, or occurred because of other reasons. Such a question brings out the

issue of counterfactual analysis, which consists of constructing a scenario for the

economy that would have prevailed in the absence of the SAP. This type of scenario

should include controls for exogenous shocks unrelated to the policy reforms.

Comparison of the observed and the counterfactual values of the economic variables

would then indicate the differential impact of the SAP on the economy. The problem is

5

that the estimation of a detailed counterfactual path cannot be done in the absence of a

consistent multi-sector general equilibrium model (Sarris 1990). We consider that the

construction of such a model is rather difficult and a time-consuming task because of

the lack of proper and comprehensive data which is a constant difficulty in developing

countries such as Sri Lanka. Having understood this difficulty, especially in terms of Sri

Lanka’s economic structure, we proposed the General Equilibrium Growth Accounting

approach which captures the effect of changes in policy (exogenous) variables on target

(endogenous) variables, based on the initial framework as suggested by Sarris. For

simplicity, we concentrated on the effects of the two most important policy (exogenous)

variables (fertilizer and non-agricultural prices), with the support of another two policy

(exogenous) variables (agricultural export and food imports).

Let us now concentrate on the relevance of SAP to Sri Lanka which was among

the first developing countries to adopt SAPs in 1977. This program of economic policy

reforms were mainly designed and introduced by the World Bank. The major economic

policy reforms introduced in Sri Lanka included the provision of incentives to export

oriented sectors, a reduction in the protection provided to the import competing sectors,

exchange rate changes, and fiscal and monetary reforms. In addition, liberalization of

domestic factor and product markets from government intervention was carried out by

privatizing government owned enterprises, thus allowing the independent function of

market forces. Athukorala and Jayasuriya (1994), Bandara and Gunawardana (1989)

6

mainly studied the historical process of economic reforms in Sri Lanka, particularly in

relation to the macroeconomic effects. The impact of such policy reforms on the

domestic food sector was not evaluated, although the sector’s importance can not be

neglected in terms of contribution to GDP and importance to the labor market.

Therefore, we have considered the domestic food sector in more detail here and further

divided this sector into three sub sectors.

B. Agriculture and economic conditions prevailed during 1970-1994

The period of 1970-1977 could be considered as the pre-reform period, because

Sri Lanka followed a closed economic policy under which foreign exchange limitations

and restrictions on imports of food and agricultural inputs existed. During this period,

the government adopted a policy of food self-sufficiency under increased government

interventions in domestic factor and product markets. Many private business ventures

were put under government control and management, while vast areas of tea, rubber

and coconut cultivation land was nationalized under a land reform program1. Due to the

change in government in 1977, “a new economic reform policy” was introduced.

1. First Stage during 1978-1988: After the closed economy period, the new

government came to power in 1977 and implemented various policy reforms, aiming to

achieve accelerated economic growth, create employment opportunities, increase

capacity utilization, stimulate savings and investment, improve the balance of payments,

1 See Gunawardana, 1981.

7

and achieve international competitiveness (Athukorala and Jayasuriya 1994). In order

to achieve these, the government took several measures, the most important of which

are outlined below. First, a new tariff system was introduced in place of non-tariff

measures. Second, the exchange rate was unified and allowed to be determined by the

market, and exchange controls were removed. Third, Sri Lanka’s currency (Rupee) was

substantially devalued. Fourth, many public sector investment programs were

introduced and fifth, export processing zones were introduced. As we saw earlier, trade

liberalization was a major component of the policy reform package. Accordingly, the

introduction of this open economy policy also led to the elimination of most of the

controls. Major fiscal policy reforms also included the removal of food subsidies, the

introduction of a targeted food stamp scheme in 1978, and the reduction of fertilizer

subsidies. Government concessions on agricultural credit were also reduced (Lakshman

1994).

2. Second Stage during 1989-1994: In 1989, the leadership of the government changed,

and the same government implemented the “second series of policy reforms” for

various reasons. Macroeconomic instability, compounded by government

mismanagement of the domestic economy, the escalation of ethnic violence and

insurgency all blocked the progress of the “initial stage of incomplete reforms and

liberalization” during the 1977-1983 period (Dunham and Kelegama 1994). The first

stage of reforms caused many problems in certain sections of the community. The

8

social cost of the adjustment also forced the government to implement a converted

version of the policy in the second phase. This comprised two types of policy reforms

and initiatives; first, the technically important but low profile adjustments, and second,

high profile projects including the privatization of more public institutions. Thus, the

second phase placed further emphasis on export-oriented industrialization under a more

liberalized trade regime, and a major program of poverty alleviation. Also, the private

sector was allowed to carry out the import of fertilizers, the price of which was aligned

to the world market.2

III. Performance of major policy (exogenous) and target (endogenous) variables of

the study in the Sri Lankan context

As stated above, we observe two indices such as the change of Growth Rate

Multiplier (GRM) and the change of policy variables such as E1, M2, PF, and PN, and

used them as the principle variables to see the impact of the policy. Agricultural exports

increased under the policy reforms in Sri Lanka, and are considered to be the engine of

foreign exchange earning. The policy reforms also addressed this issue. Food imports

became open under the policy reforms, the impact of which was also widely felt by the

domestic food sector. Furthermore, fertilizer continued to play an important role under

the reforms. Gradually, as the subsidies were removed, the increase in usage and price

of fertilizer to the farmers became surprisingly high. Therefore, it is imperative to

2 See Dunham and Kelegama 1994 for detailed description of second wave policy reforms.

evaluate the impact of this. Non-agricultural sector performances are always portrayed

as a hurdle for agricultural development. Many of the policy reforms carried out in

other countries tried to use this sector to restructure the economy and neglected the

agricultural sector. So, in this paper we tried to see the impacts on the agricultural sector,

using the price of non-agriculture, fertilizer price, agricultural exports and food imports.

Table 1 and Figure 1 illustrate these trends clearly.

Table 1 Growth Rate of Major Target and Policy Variables

TargetVariables 1970-1974 1975-1979 1980-1984 1985-1989 1990-1996

Average1970-1996

GR(X1) 0.43 2.26 -0.53 -4.93 -1.11 -0.78

GR(X2) 4.02 8.72 2.89 -0.92 10.45 5.03

GR(X3) 3.76 6.89 1.12 0.78 8.73 4.26

GR(XA) 2.52 14.15 0.45 -2.31 7.35 4.43

GR(Cf) 1.35 4.23 -4.01 5.87 2.45 1.98

GR(P1) -11.61 1.66 17.79 -2.51 -40.93 -7.12

GR(P2) 12.01 14.24 30.72 4.26 32.61 18.77

GR(P3) 10.92 20.67 22.34 9.45 27.52 18.18

GR(GDP) 2.92 6.29 0.61 1.82 3.63 3.05

PolicyVariables 1970-1974 1975-1979 1980-1984 1985-1989 1990-1996

Average1970-1996

GR(E1) 0.31 26.94 0.09 -7.38 -2.41 3.51

GR(M2) -11.68 -6.94 -25.89 10.76 -1.87 -7.12

GR(PF) 5.73 -3.43 16.78 10.43 -3.62 5.18

GR(PN) 16.63 12.98 27.11 16.65 20.34 18.74

Note on Variable Explanation:

GR(X1) = Growth Rate of Agricultural Production from Sector 1 (Export Agriculture).

GR(X2) = Growth Rate of Agricultural Production from Sector 2 (Domestically Produced Import

Substitutable Agricultural Output).

GR(X3) = Growth Rate of Agricultural Production from Sector 3 (Domestically Produced and Consumed

Agricultural Output).

GR(XA) = Growth Rate of Aggregate Agricultural Production (Weighted Average of Production from three

Sectors).

GR(Cf) = Growth Rate of Food Consumption (Consumption only from Sectors 2 and 3).

GR(P1) = Growth Rate of Agricultural Price from Sector 1 (Export Agriculture).

9GR(P2) = Growth Rate of Agricultural Price from Sector 2 (Domestically Produced Import Substitutable

Agricultural Output).

GR(P3) = Growth Rate of Agricultural Price from Sector 3 (Domestically Produced and Consumed

Agricultural Output).

GR(GDP) = Growth Rate of Real GDP.

GR(E1) = Growth Rate of Agricultural Export.

GR(M2) = Growth Rate of Food Imports.

GR(PF) = Growth Rate of Fertilizer Price.

GR(PN) = Growth Rate of Non-Agricultural Price.

0.00

50.00

100.00

150.00

200.00

250.00

300.00

19

70

19

71

19

72

19

73

19

74

19

75

19

76

19

77

19

78

19

79

19

80

19

81

19

82

19

83

19

84

19

85

19

86

19

87

19

88

19

89

19

90

19

91

19

92

19

93

19

94

19

95

19

96

Year

Index

Total fertilizer Usage( Index, 1970=100) Fertilizer Price Index

Figure 1 Fertilizer Price and Usage Change Pattern from 1970 to 1996.

Table 1 and Figure 1 show the structure of the economy and its performance quite

clearly. The ethnic conflict in 1983 and the internal unrest in 1987 and 1988 contributed

to the decreasing trend in exports and food imports. Furthermore, the devaluation of Sri

Lankan currency Rupee under the policy reforms also contributed to the increases in

agricultural exports and non-agricultural prices, and the decrease in food imports. With

this brief introduction of the policy reform scenario, we used the following analytical

framework to evaluate the major impacts of these reforms on the agricultural sector.

10

11

IV. Structure of the analysis and model construction

So far there have been several studies dealing with the adjustment policy effects

on the economy of Sri Lanka. In two earlier works by Bandara (1989) and Cooray

(1998 and 1999), sub-sectors of the domestic food sector were not considered. We

divided the domestic food sector into three sub-sectors in this paper, and a three

sub-sectors model with Growth Rate Multiplier (GRM) approach is used to find the

major policy effects

3

3 For detailed information about the model variable, see discussion paper 0407 of

Yamaguchi, M and SriGowri Sanker (2004).

More specifically, here the economy was divided into two sectors such as agriculture

and non-agriculture. In order to evaluate the impact of plantation and other sectors, the

agricultural sector has been further divided into three sub-sectors. In our analytical

framework, the following assumptions are made. First, we assume that agriculture will

produce three products. The first one is agricultural exportable production (sector 1),

the second is import substitutable and domestically consumable production (sector 2)

and the final one is both domestically produced and consumable production (sector 3).

Second, we assume that aggregate agricultural production will depend on factors such

as land and capital, as well as variable factors such as labor and the imported input of

fertilizer. Here, the fertilizer price, which is considered as one of the important policy

variables in this study, is given for agriculture and will change under adjustment. Third,

we also assume that the price of the non-agricultural sector, which is another important

policy variable, will be determined by factors largely outside agriculture. This will

enable us to see the effect on the target (endogenous) variables.

The first framework of the model was initially done by Sarris in 1990, and we

made an almost completely different model by adding the following new equations to

remedy the inadequacies. First, Sarris did not specify the way to solve the equations in

order to fully capture the impact (effects) of the policy (exogenous) variables on target

(endogenous) variables. Second, the Sarris model also did not specify anything about

1

iven below.

the non-agricultural sector. Third, he neglected the domestic consumption of exportable

goods which is very impractical in the case of Sri Lanka. Consequently, our model has

been developed remedying these shortfalls. In our static model, we have 23 equations

which include agricultural and non-agricultural, 2 production functions, 3 consumption

functions, equations for income and equations for labor allocation in both sectors4.

From these 23 equations, we obtained the dynamic model which is reduced to 21

equations as shown in Appendix Table 1. Here, the model uses the General Equilibrium

Growth Accounting Approach5 to find the impact of 11 policy (exogenous) variables on

21 target (endogenous) variables. Since the focus is on fertilizer and non-agricultural

prices, only these major results are discussed here.

The dynamic model, as given in the Appendix Table 1, has the general form Ax=b

where A is a matrix of order (21 X 21) of structural parameters, x is the column vector

of rates of change of 21 target (endogenous) variables (X1, X2, X3, XA, C1, C2, C3, Cf,

P1, P2, P3, Pf, PA, CPI, DEF, LA, Y, GDP, E, XN, LN) and b is the column vector of

rates of change of 11 policy (exogenous) variables ( E1, M2, d, e, TA, TN, PF, PN, L,

N ,LA0)6. The details of these variables and parameters are g

4 These 23 equations are shown in the place following Appendix Table 1. Please see the

discussion paper 0407 for full description of the model, the variables, and their effects. 5 Papers among these studies are Yamaguchi and Binswanger (1975), Yamaguchi (1982) and

Yamaguchi and Kennedy (1983).

6 Detailed description of the exogenous and endogenous variables can be seen from the

Discussion paper 0407 of Yamaguchi and SriGowri Sanker (2004).

Table 2 Variables and Parameters used in the Model.

Ｄescription Data Ｓource

Target (Endogenous) Variables (21 variables):

1. Xi : Agricultural output of sector i, where i =1, 2, 3

2. X : Aggregate output of agricultural sector (sector 1, sector 2, and sector 3). A

A

3. C1 : Domestic Consumption of sector 1.

4. C2 : Domestic Consumption of sector 2.

5. C3 : Domestic Consumption of sector 3.

6. Cf : Food consumption from sectors 2 and 3.

7. Pi : Agricultural prices of three sub-sectors, where i = 1, 2, 3

8. Pf : Price of food consumption (sectors 2 and 3).

9. P : Agricultural price.

10. CPI : Consumer Price Index.

11. DEF : Deflator.

12. LA : Total agricultural labor force.

13. Y : Nominal GDP

14. GDP: Real GDP

15. E : Per capita income

2

16. XN : Non-agricultural output.

17. LN : Non-agricultural labor force.

Policy (Exogenous) Variable (11 variables):

18. E1: Exports of agricultural sector 1

19. M2: Food imports such as basic cereals that are perfect or near perfect substitutes.

20. d : Demand shifter of consumption (sector 1).

21. e : Demand shifter of consumption (food, sectors 2 and 3).

22. TA : Technical change in agriculture

23. TN : Technical change in non-agriculture.

24. PF : Fertilizer price.

25. P : Non-agricultural price. N

26. L : Total labor.

27. N : Population

28. LA0 : Initial value of agricultural labor.

Parameters

1. = Elasticity of production w.r.t Agriculture Labor. a

2. = Elasticity of production w.r.t Quantity of Fertilizer Used. b

3. η = Income Elasticity of Demand for Food.

4. ε = Price Elasticity of Demand for Food.

5. τ = Positive Elasticity of Transformation (CET Function).

6. = Price Elasticity of Cn 1.

7. = Income Elasticity of Cq 1. 3

8. σ = Elasticity of Substitution (CES Function).

9. s1 = C1/X1 = Ratio of Consumption from sector 1.

10. s2 = X2/C2 = Initial Self-Sufficiency Ratio of sector 2.

11. υ i = υ1

,υ 2,υ 3

Value of Shares of Each Agricultural Product in the Total

Value, . 3,2,1=i

12. μ A = Share of Agriculture in GDP.

13. Vf = Share of Food Products (Sectors 2 & 3) in the Total Consumer Budget.

14. λ 2 = Share of Sector 2 Agriculture to Total Consumption of Sectors 2 and 3.

15. ξ = Elasticity of Production (Non-Agriculture) w.r.t Non-Agriculture Labor.

16. γ 1 = Elasticity of TA w.r.t Agriculture Labor.

17. γ 2 = Elasticity of TN w.r.t Agriculture Labor.

18. γ 3 = Elasticity of L w.r.t Agriculture Labor.

19. = Share of Agriculture Labour Force to Total Labor (Ll A A/L).

20. = Share of Non-Agriculture Labour Force to Total Labor (LlB A/L)

Note:The data for this study was obtained from Annual Report of Central Bank of

Sri Lanka-Statistical Appendix, for various years, Socio-Economic Data of Sri Lanka

by Central Bank (various years) and HARTI Agrarian Data, through Time Series Data

of Sri Lanka Agriculture of Hector Kobbekaduwa Agrarian Research and Training

Institute (HARTI), Ministry of Agriculture, Government of Sri Lanka. Many of the

Parameters’ value were estimated and certain values were also obtained from

4

Central Bank and HARTI.

The inverse matrix of A displays the Growth Rate Multipliers (GRMs )7 which are

obtained by calculating the inverse of the above matrix of structural parameters. Here

we can see that GRMs are affected by SAP through the parametric values of A matrix.

These effect values are given in Appendix Table 2 which allows us to find the influence

of the exogenous variables on the endogenous variables. As an example,

element is

2,81 )( −A

∧∧

∂∂

2MC f (we write this as CfM2) which indicates by how much the

rate of change of aggregate consumption of food Cf changes due to (or effects) an

increase or decrease in the growth rate of import substitute M2. Similarly, we could

attribute to another policy (exogenous) variables, such as ( )2

1∧

∧

∂

∂

M

X =(X1M2) is the

relevant GRM, which shows how much percentage (%) of X1 would increase when M2

increases by 1%. From Figure2, we can see how SAP affected the GRM. These effects

are described in section 5.1.

In addition, the contributions of exogenous (policy) variables to the endogenous

(target) variables are calculated by multiplying the GRMs of each year interval by the

corresponding rates of change of the exogenous variables. The calculated values of

these contributions are given in Appendix Table 3. For example, CX1M2 =

5

7 For further details of the application of GRM, see Yamaguchi (1982), Yamaguchi and

Kennedy (1984), and Yamaguchi and Binswanger (1975).

( )() 2

2

1 MGRM

X∧

∧

∂

∂ , where CX1M2 is the contribution of the agricultural food imports M2

to the agricultural production for exports X1. This will explain how much is contributed

by the rate of change in food imports into the rate of change of agricultural production

for exports. Here, ( )2

1∧

∧

∂

∂

M

X =(X1M2) is the relevant GRM, which shows what

percentage (%) of X1 would increase when M2 increases by 1% as shown before.

Detailed discussion on these results is given in the following section. From Figure 3,

we can also see how SAP affected both the GRM and the growth rate of policy

variables. Concrete example of this phenomenon is explained at the end of section5.2

using the calculated values of GRM and contribution.

V. Discussion on results

A. Discussion on effects

Appendix Table 2 details the values of effects in detail in relation to this model

based on the GRMs. It would take considerable space to describe the performance of all

the effects, hence, only the principal effects which are mentioned in the earlier sections

of this paper and clearly describe the policy effects are discussed here8. As mentioned

6

8 See the discussion paper 0407 of Yamaguchi and SriGowri Sanker (2004) for detailed

analysis and the performance pattern of the entire exogenous and endogenous variables of

the model.

7

above, GRMs are obtained by calculating the inverse of the above matrix of structural

parameters.

Fertilizer price and non-agricultural price have completely opposite effects on the

economy. First, we focus on the fertilizer price. It can be seen that changes in the

fertilizer price have a much more negative effect on the domestic food sector (X2PF <0

and X3PF <0) than on X1 (X1PF <0), due to its usage pattern. For example, a 100%

increase in fertilizer prices would bring down production of X1, X2 and X3 by 1%, 10%,

and 10% in 1970-1974 to 3%, 16%, and 15% in 1990-1996 respectively. It clearly

shows the effect of the increase of fertilizer price under SAP. For example, XAPF

indicates the effect of fertilizer price PF on aggregate agricultural production, XA. In

this way, we can understand the effects of other policy variables on target variables as

explained in Figure 2 and in Appendix Table 2.

Effect on Export Agriculture Production (X1)

-0.10

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

70-74 75-79 80-84 85-89 90-96

X1PF X1PN X1E1 X1M2

Effect on Import Substitutable DomesticAgriculture Production (X2)

-0.30

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

0.50

70-74 75-79 80-84 85-89 90-96

X2PF X2PN X2E1 X2M2

Effect on Domestically Consumable DomesticAgriculture Production (X3)

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

0.50

70-74 75-79 80-84 85-89 90-96

X3PF X3PN X3E1 X3M2

Effect on Aggregate Agriculture Production(XA)

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

0.50

0.60

70-74 75-79 80-84 85-89 90-96

XAPF XAPN XAE1 XAM2

Effect on Export Agriculture Price (P1)

-2.00

-1.00

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

70-74 75-79 80-84 85-89 90-96

P1PF P1PN P1E1 P1M2

Efect on Import Substitutable DomesticAgriculture Price (P2)

-1.00

-0.50

0.00

0.50

1.00

1.50

2.00

70-74 75-79 80-84 85-89 90-96

P2PF P2PN P2E1 P2M2

Effect on Domestically Consumable DomesticAgriculture Price (P3)

-0.50

0.00

0.50

1.00

1.50

70-74 75-79 80-84 85-89 90-96

P3PF P3PN P3E1 P3M2

Effect on GDP (Real)

-0.05

0.00

0.05

0.10

0.15

0.20

70-74 75-79 80-84 85-89 90-96

GDPPF GDPPN GDPE1 GDPM2

Figure 2 Graphic Representations of Effects on Selected Target Variables.

Note: X1PF = Effect of fertilizer price PF on exportable agricultural production from sector 1 X1.

X1PN = Effect of non-agricultural price PN on exportable agricultural production from sector 1 X1.

8X1E1 = Effect of agricultural exports E1 on exportable agricultural production from sector 1 X1.

9

, due to the

chan

X1M2 = Effect of food imports M2 on exportable agricultural production from sector 1 X1.

Similarly other notations and trends can be understood.

The changing pattern of fertilizer prices always affects the prices of agricultural

products from all the three sectors (P1PF, P2PF, P3PF >0). Under the SAP, fertilizer

prices increased due to the reduction and removal of subsidies. This in turn, negatively

affected the production (X1PF, X2PF, X3PF <0), thus increasing the prices of

agricultural products. It severely affected agriculture after the policy reforms. This is

evident in the changing pattern of price indices of agriculture which registered an

alarming increase during 1990-1996 in comparison to that of 1970-1974 (e.g., P1PF

increased from about 0.79 to 1.32; P2PF increased from about 0.18 to 0.42; P3PF

increased from about 0.20 to 0.49). Hence, it could be concluded from this that a

considerably negative effect is felt on agricultural production and prices

ge in fertilizer prices.

On the other hand, we can also conclude that the expansion of the non-agricultural

sector does not adversely affect the agricultural sector. It is also worth mentioning that

the non-agricultural price (PN) decreased the agricultural exportable production (X1PN

<0) in the pre-reform period, but after the policy reforms, PN helped the production

from sector 1, X1 (X1PN >0. X1PN increases from about 0 to 0.03.). Positive effects

could be observed in the case of import substitute and domestic food sectors (X2PN >0,

X2PN increases from about 0.10 to 0.42. X3PN >0, X3PN increases from about 0.10 to

10

the policy reform stages in 1970-1979, it increased substantially in the

1990

y. However, the increase in fertilizer price has had a very

bad

0.40). Overall PN tends to increase the production of aggregate agriculture XA (XAPN

>0, XAPN increases from about 0.06 to 0.29). Although this effect was small in the

beginning of

-1996 period.

Increases in fertilizer price (PF ) decreases real GDP and decreases food

consumption (Cf ) although nominal GDP (Y ) increases because of the increases of P1,

P2, P3 (i.e., inflation due to the increase of fertilizer price). However, increases in

non-agricultural price (PN ) increases all the variables mentioned above, i.e., PN

increases nominal GDP (Y) and also increases real GDP (GDP) and increases food

consumption (Cf ) (see YPN>0, GDPPN>0 and CfPN>0 in Appendix Table 2).

Therefore, the expansion of the non-agricultural sector including PN does not always

adversely affect the econom

effect on Sri Lanka’s economy.

It is quite evident that the trend of agricultural exports E1 had a notable impact on

the agricultural production of both exportable and domestically produced and

consumed items (see X1E1, X2E1 and X3E1 in Figure 2 and in Appendix Table 2). The

effect in this regard is quite large compared to the two sectors, sectors 2 and 3. This is

expected under SAP but the effect of exports on sector 2 production is somewhat larger

than that of sector 3 (X2E1 > X3E1). Although these two effects are negative before

1975, the larger positive effects on sector 2 and 3 after 1975 clearly show that both

11

ll

agricultural production XA (XAM2<0) as shown in Figure 2 and in Appendix Table 2.

B.

learly shows

e percentage contribution of the policy variables to the target variables.

sectors, import substitute and domestic food production, are affected positively by

agricultural exports from the same period (1975-79). Nevertheless, the overall

agricultural production XA shows positive increasing effects since 1975-79 to 1980-84,

and again a declining trend until 1996 (see XAE1 in Figure 2 or Appendix Table 2). This

clearly illustrates the initial shift in the production of agricultural exports soon after the

implementation of the policy reforms and the decline in the later stages of the reforms

due to other exogenous factors affecting the exports and the production. Since the

opening of trade allowed the food imports M2, the negative effect of M2 became strong

in the domestic food production X2 and X3 (X2M2<0, X3M2<0) as well as on the overa

Discussion on contributions

Further changes on these effects can be clearly understood from analyzing the

results of contributions of the policy (exogenous) variables on target (endogenous)

variables, as given in graphically, in Figure 3 and in Appendix Table 3. The most

important variables, fertilizer and the non-agricultural prices, have also contributed

significantly to the agricultural output negatively and positively. Figure 3 c

th

12

Contribution of 4 Policy Variables toAggregate Agricultural Production (XA)

-6

-4

-2

0

2

4

6

8

10

12

14

16

70-74 75-79 80-84 85-89 90-96

Year

Perc

enta

ge C

ontr

ibuti

on

GR(XA) CXAPF CXAPN CXAE1 CXAM2

Contribution of 4 Policy Variables to ExportAgriculture Price (P1)

-50

-40

-30

-20

-10

0

10

20

30

70-74 75-79 80-84 85-89 90-96

Year

Perc

enta

ge C

ontr

ibution

GR(P1) CP1PF CP1PN CP1E1 CP1M2

Contribution of 4 Policy Variables to GDPGrowth (GDP-Real)

-2-101234567

70-74 75-79 80-84 85-89 90-96

Year

Perc

enta

ge C

ontr

ibuti

on

GR(GDP) CGDPPF CGDPPN

CGDPE1 CGDPM2

Contribution of 4 Policy Variables to ImportSubstitutable Domestic Agriculture Price (P2)

-10

0

10

20

30

40

70-74 75-79 80-84 85-89 90-96

Year

Perc

enta

ge C

ontr

ibuti

on

GR(P2) CP2PF CP2PN CP2E1 CP2M2

Contribution of 4 Target Variables toDomestically Consumable Domestic

Agriculture Price (P3)

-5

0

5

10

15

20

25

30

35

70-74 75-79 80-84 85-89 90-96

Year

Perc

enta

ge C

ontr

ibution

GR(P3) CP3PF CP3PN CP3E1 CP3M2

epresentation of Percentage Contribution of Policy Variables to

ution of fertilizer price PF to the performance of aggregate agricultural production XA.

Figure 3 Graphical R

the Target Variables.

Note: GR(XA) = Growth rate of aggregate agricultural production XA.

CXAPF= Contrib

CXAPN= Contribution of non-agricultural price PN to the performance of aggregate agricultural

production XA.

13

icy

varia

h

(CGDPPF <0, maximum negative contribution was -47.99% in the year of 1980-84).

CXAE= Contribution of agricultural exports to the performance of aggregate agricultural production XA.

CXAM2 = Contribution of food imports M2 to the performance of aggregate agricultural production XA.

Similarly, we can understand the other notations and trends. The growth rate of each

histogram (e.g., 2.52% of XA in 1970-74) is the sum of 4 contributions of 4 pol

bles in Figure 3, plus 7 other policy variables such as d, e, TA, TN, L, N and LA0.

First, as shown in the values of Figure 3 and Appendix Table 3, the fertilizer price

has about 235% of contribution to the decrease of the agricultural output (CXAPF

=-234.78%) in 1980-1994. The contribution of these policy variables is also shown in

the prices of the agricultural outputs XA in Figure 3. It is quite evident that the

contribution of fertilizer prices to the price of products of all the three sectors (CP1PF,

CP2PF, CP3PF ) is very large, reaching a high of 98.14% in sector 2 during 1985-89.

Mostly, the contribution of the fertilizer prices to P1, P2 and P3 was negative in the

beginning of the policy reform (1975-79) because of the decrease of fertilizer price

(-3.43%, see Table 1), but tends to increase later, thus increasing the prices of P1, P2

and P3. Since the small-scale farmers are often affected by this increase of PF, the

policy option here should be rationalizing the fertilizer subsidy based on the cultivation

size. Furthermore, the decrease in the growth rate of GDP was also evidenced here in

the period of 1985-89, following that of 1980-84. The negative contribution of the

decrease in agricultural exports to GDP (CGDPE1<0) is 48.20% in the period of

1985-89. Also, the fertilizer price increase negatively contributed to the GDP growt

14

e period (1975-1979). Also, the

large

the understanding of the performance of

It is noteworthy that the prices of products from sectors 2 and 3 (P2 and P3) are

also considerably affected by the non-agricultural price (PN) and the contribution from

this is quite strong (CP2PN and CP3PN are very large). This trend further endorses the

assumption that the development of the non-agricultural sector tends to increase its

prices, and this contributes positively to the agricultural products (CXAPN >0 in almost

all periods). The change of other policy variables given by E1 and M2 contributed to the

performance of agricultural output XA (CXAE1 and CXAM2). The biggest contribution of

exportable to agricultural output is in 1975-1979 just at the beginning of the policy

reform. Exports contributed almost 100% to the growth of XA (CXAE1 = 98.19 % in

1975-79). Also the biggest contribution of export to agricultural exportable price P1

(CP1E1) is 743% in 1975-1979. In the same way, E1 has the largest contribution to

other agricultural prices like P2 and P3 in the sam

st contribution of E1 to GDP is 65% in the same period.

However, the biggest contribution of import (owing to the decrease of import) to

agricultural output (CXAM2) is 88.89% in the 1980-1984. Also, the biggest contribution

of M2 to P1, P2 and P3 are in 1985-1989. Especially, the impact of M2 on P2, (i.e.,

importable and domestically produced) in 1985-1989 is the largest and decreased the

price to 137%. Consequently, inflation made GDP decrease in the same period. Hence,

the calculation of contribution further enriched

15

endo

eform (SAP) on the target variables. This gives

the overall picture of the SAP impact.

genous variables in relation to the GRM.

More specifically, Here we must know how much E1, M2 and PF increased or

decreased by SAPs. For PF, we can obtain 5.18% by calculating the growth rate of PF

from 1970 to 1996. In other words, the growth rate of PF would become zero if Sri

Lankan government controls the increase of PF. For E1 and M2, the situation is not so

simple, because of the influence of the devaluation of the Rupee9 on the increase of E1

and the decrease of M2. For simplicity, we assumed that the increase of E1 and decrease

of M2 are the same as the percentage of devaluation of the Rupee (i.e., 9.64%). Then,

we could obtain the growth rate of E1 as -6.13 (actual value is 3.51) %, and the growth

rate of M2 as 16.76 (actual value is 7.12) %. In other words, E1 increased and M2

decreased by 9.64%. Now, we can calculate how much these values (9.64, 9.64 and

5.18) for E1, M2 and PF contribute to the growth of XA by calculating (XAE1)*GR(E1),

(XAM2)*GR(M2) and (XAPF)*GR(PF). From our calculations, XAE1 = 0.44, XAM2 =

-0.04 and XAPF = -0.08, and the growth rates of E1, M2 and PF as 9.64, 9.64 and 5.18%.

Hence, the contribution of E1, M2 and PF (i.e., (XAE1)*GR(E1), (XAM2)*GR(M2) and

(XAPF)*GR(PF)) are calculated as 4.24, 0.39 and 0.41%. Therefore, the total

contribution of SAPs comes to 5.04% if we add these values together. This way we

could capture the effect of the policy r

9 Rupee is the name of the Sri Lankan local currency.

16

unting approach. We

obtai

of SAP on agriculture is favorable for both agricultural and economic

deve

VI. Overall conclusion

We constructed the model of AX = b, where b has the policy variables and X has the

target variables as explained in the previous sections. The effect of the policy variables

on the target variables were, therefore, able to be obtained through this model with an

appropriate approach as elaborated in the presiding sections. Empirical appropriateness

and suitability for policy evaluation is justified through this analysis. Since this study is

originally growth accounting for Sri Lanka’s agriculture, policy impact (the effect of

SAP) was evaluated through two methods such as the change of GRM and the change

of policy variables as explained above under this growth acco

ned the positive overall effect of Sap as summarized below.

The fertilizer prices that change under the policy adjustments have fairly large

negative effects on the domestic agricultural economy. The positive impact of the

non-agricultural price on agriculture is fairly large. Since the positive impact of export

earnings compensates for the negative impacts of food import and fertilizer price, the

overall impact

lopment.

In two earlier notable works by Bandara (1989) and Cooray (1998 and 1999), they

argued that the SAP reforms were detrimental to the agricultural sector and

concentrated only on the domestic food sector. We have carefully studied this fact

17

bined. Therefore, SAP had a positive effect, in total, on the Sri Lankan

econ

verall

agric

through dividing agriculture into three sub-sectors. It is proved from our study that SAP

had a comparatively bad effect on domestic food sectors, such as sector 2 and sector 3,

although SAP had a good effect on the exportable agriculture sector. The effects of M2

and PF were negative but the positive effect of E1 was larger than these two negative

effects com

omy.

As seen above, the negative fertilizer contribution, due to the increase of PF, was

considerably higher in the 3rd (1980-84) and 4th (1985-89) periods of SAP

implementation. Therefore, the Sri Lankan government advocated restoring the

fertilizer subsidy. The period of 1990-96 clearly shows that the decrease in the fertilizer

price positively contributed to agricultural development. Due to this policy, the o

ulture sector developed positively and GDP also recorded a positive trend.

Therefore, we have to control the fertilizer price in case of its escalation. Also, it

can be understood that an increase in non-agricultural price has a rather positive

contribution on agriculture too. This once again reinforces the importance of the growth

of non-agriculture for agriculture. However, we should not forget the importance of the

contribution of export of agriculture from sector 1 in the earlier periods, especially in

the period of 1975-79 as shown in Figure 3. The decrease in food import also

contributed positively to the growth of Sri Lankan agriculture. These increases in

exports and decreases in imports were made through the devaluation of the Rupee. In

18

valuation of local currency are extremely important for every

developing country.

short, we can reconfirm again that an effectively balanced combination of agricultural

and non-agricultural development, using the control of fertilizer price, the importance

of non-agricultural development, the importance of increasing exports and decreasing

imports, and the de

19

Macroeconomic Policies,

ic and Social

of Sri Lanka-Statistical

Socio-Economic Data of Sri Lanka. Colombo: Central Bank of Sri

ence

raining Institute. Ministry of Agriculture,

aper no. 6/89. Australia:

oach to the Sri Lankan Case” Regional Development

ysis and Planning:

References

Athukorala, Premachandra and Sisira, Jayasuriya. 1994.

Crises and Growth in Sri Lanka: 1969-90. World Bank.

Alexander, Sarris. 1990. “Guidelines for Monitoring the Impact of Structural

Adjustment Programs on the Agriculture Sector.” FAO Econom

Development Paper no. 95, Rome: Food and Agriculture Organization.

Central Bank. Various Years. Annual Report of Central Bank

Appendix. Colombo: Central Bank of Sri Lanka.

Central Bank. 2002.

Lanka

David, Dunham and Saman Kelegama. 1994. “The Second Wave of Liberalization in

Sri Lanka 1989-93: Reform and Governance”. Proceedings of International Confer

on Economic Liberalization of South Asia. Australian National University.

HARTI Agrarian Data. 2001. Time Series Data of Sri Lanka Agriculture. Colombo:

Hector Kobbekaduwa Agrarian Research and T

Lands and Forestry. Government of Sri Lanka.

Jayathilake, Bandara and Pemasiri J. Gunawardana. 1989. “Trade Policy Regimes and

Structural Changes: The Case of Sri Lanka.” Discussion P

School of Economics. La Trobe University.

Nawalage S. Cooray. 1998. “Macroeconomic Management in Developing Countries:

An Econometric Modelling Appr

Studies 4: 37-63.

Nawalage S. Cooray. 1999. “Econometrics Modelling for Policy Anal

Evolution and Application” Regional Development Studies 5: 27-40.

Pemasiri J. Gunawardana.1981. “Land Policy and Agrarian Change in Independent Sri

20

l

Japan: 1880-1965.” American Journal of Agricultural

Economic Development:

of Japan 1880-1970.” Canadian Journal of

iscussion Paper no. 0407. Kobe: Graduate

chool of Economics, Kobe University.

Lanka.” Sri Lanka Journal of Agrarian Studies 2: 27-43.

Weligamage D. Lakshman. 1994. “Structural Adjustment Policies in Sri Lanka:

Imbalances, Structural Disarticulation and Sustainability.” Proceedings of Internationa

Conference on Economic Liberalization of South Asia. Australian National University.

Yamaguchi, Mitoshi and Hans Binswanger. 1975. “The Role of Sectoral Technical

Change in Development of

Economics 57, no 2: 269-78.

Yamaguchi, Mitoshi. 1982. “The Sources of Japanese

1880-1970.” Economic Studies Quarterly 33, no 2: 126-46.

Yamaguchi, Mitoshi. and George Kennedy. 1984. “A Graphic Model of the Effects of

Sectoral Technical Change: The Case

Agricultural Economics 32, no 1: 71-92.

Yamaguchi, Mitoshi. and Srigowrisanker M. Sarama. 2004. “Empirical Analysis of Sri

Lanka’s Agriculture in Relation to Policy Reforms with General Equilibrium Growth

Accounting Approach (1970-1996).” D

S

Appendix Table 1 Dynamic Form of the Model

Equation Number Equations

(1) from equation A-10 where s11111 )1(∧∧∧

+−= CsEsX 1=C1/X1

(2) from equation A-11 where s222 )1( 22

∧∧∧

=−+ CMsXs 2=X2/C2

(3) from equation A-12 33

∧∧

= CX

(4) from equations A-8,A- 9 where υiiA

iPP ∑

=

∧∧

=3

1υ i = PiXi(P1X1+P2X2+P3X3)

(5) from equation A-22 + + ∧∧∧

+−= EqPnC 11 N d∧

+ NPn

(6) from equation A-4 )(111

1FAAAA PP

bbL

baT

bX

∧∧∧∧∧

−−

+−

+−

=

(7) from equation A-19 + A0 ∧∧∧∧

++= LTTL NAA 321 γγγ L

(8) from equation A-15 )( 22 ff PPCC∧∧∧∧

−−= σ

(9) from equation A-15 )( 33 ff PPCC∧∧∧∧

−−= σ

(10) from equations A-13, A-16 where, 3222 )1(∧∧∧

−+= PPP f λλ

λ2 = P2X2/(P2C2+P3C3)

(11) from equation A-17 + + )()( NfNf PPPYC∧∧∧∧∧

−−−= εη N e

(12) from equations A-5, A-18

))(1(11

111

1NNAFAAAA XPP

bbP

bL

baT

bY

∧∧∧∧∧∧∧

+−+⎥⎦⎤

⎢⎣⎡

−−

−+

−+

−= μμ

where Aμ = share of agriculture in GDP

(13) Nfff PPICP∧∧∧

−+= )1( νν

(14) NAAA PPFED∧∧∧

−+= )1( μμ

(15) from (6), (20) of Appendix Table 1

NAFAAAA XPPb

bLb

aTb

PDG∧∧∧∧∧∧

−+⎥⎦⎤

⎢⎣⎡ −

−+

−+

−= )1()(

1111 μμ

(16) from equation A-8 )( 11 AA PPXX∧∧∧∧

−+= τ

21

(17) from equation A-8 )( 22 AA PPXX∧∧∧∧

−+= τ

(18) from equation A-8 )( 33 AA PPXX∧∧∧∧

−+= τ

(19) from equation A-23 ∧∧∧

=−→= NEPDGNGDPE /

(20) from equation A-21 NNN LTX∧∧∧

+= ξ

(21) from equation A-20 where lNA LlLlL NA

∧∧∧

+= A=LA /L, lN=LN /L

From this Appendix Table 1, we can derive the model in the matrix form AX = b as

follows.

X1 X2 X3 XA C1 C2 C3 Cf P1 P2 P3 Pf PA CPI DEF LA Y GDP E XN LN

(1) 1 0 0 0 (-s1) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

(2) 0 s2 0 0 0 -1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

(3) 0 0 1 0 0 0 -1 0 0 0 0 0 0 0 0 0 0 0 0 0

(4) 0 0 0 0 0 0 0 0 v1 v2 v3 0 -1 0 0 0 0 0 0 0 0

(5) 0 0 0 0 1 0 0 0 (n) 0 0 0 0 0 0 0 0 (-q) 0 0

(6) 0 0 0 -1 0 0 0 0 0 0 0 0 b/1-b 0 0 a/1-b 0 0 0 0 0

(7) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0

(8) 0 0 0 0 0 1 0 -1 0 σ 0 (−σ) 0 0 0 0 0 0 0 0 0

(9) 0 0 0 0 0 0 1 -1 0 0 σ (−σ) 0 0 0 0 0 0 0 0 0

(10) 0 0 0 0 0 0 0 0 0 λ2 1−λ2 -1 0 0 0 0 0 0 0 0 0

(11) 0 0 0 0 0 0 0 -1 0 0 0 　(−ε) 0 0 0 0 η 0 0 0 0

(12) 0 0 0 0 0 0 0 0 0 0 0 0 μA/1-b 0 0 aμA/1-b -1 0 0 1-μΑ 0

(13) 0 0 0 0 0 0 0 0 0 0 0 　Vf 0 -1 0 0 0 0 0 0 0

(14) 0 0 0 0 0 0 0 0 0 0 0 0 μA 0 -1 0 0 0 0 0 0

(15) 0 0 0 0 0 0 0 0 0 0 0 0 bμA/1-b 0 0 aμA/1-b 0 -1 0 1-μΑ 0

(16) -1 0 0 1 0 0 0 0 τ 0 0 0 (−τ) 0 0 0 0 0 0 0 0

(17) 0 -1 0 1 0 0 0 0 0 τ 0 0 (−τ) 0 0 0 0 0 0 0 0

(18) 0 0 -1 1 0 0 0 0 0 0 τ 0 (−τ) 0 0 0 0 0 0 0 0

(19) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 -1 0 0

(20) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1

(21) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 l1 0 0 0 0 l2

1P

AL

NXNL

( ) 11 ˆ1 Es−

( ) FA PbbTb

ˆ1/ˆ1

1−+

−

( ) ( ) NAAAA PPbbT

bF ˆ1ˆ1/ˆ

1−+−+

−μμ

μ

( ) Nf PV ˆ1−

( ) NA P1−μ

N

NT

L

00

0

0

00

00

ζ−

2P3PfPAP

1X

2X3XAX

E

( ) FAAA PbbT

bˆ1/ˆ

1−+

−μμ

Y

FED

=

( ) 22ˆ1 Ms −

IPCˆ

PDG

∧∧∧∧

−−− NePP NN εη

∧∧∧∧

+++ 0321 ANA LLTT γγγ

∧

1C∧

2C∧

3C∧

fC

∧∧∧

++ NPnNd

22

These 21 equations in Appendix Table 1 are derived from the following 23 equation

(see discussion paper 0407 for more details).

---------------------------------------------------------------------------------------------------------

23

FAAA

FFAAA

F FAA

A FAA

FAAAA

ba XLTX = a,b >0 a+b <1 (A-1)

Max V (A-2) XPXP −=

=X (TAbbba bPPL −−− 1/11/11/1 )/() (A-3)

=X (TAbbbbba bPPL −−− 1/1/1/1 )/() (A-4)

V (A-5) bbbbbba bbPPLT −−−− −= 1//)1(1/11/1 )1()(

X (A-6) ∑=

−−−=3

1

)1/(/)1( )(i

iiA X τττττα

Max (A-7) ∑=

3

1iii XP

X i = 1, 2, 3 (A-8) ττα )/( AiAii PPX−=

P (A-9) ∑=

++−=3

1

)1/(11 )(i

iiA P τττα

X1 = E1 + C1 (A-10)

X2 + M2 = C2 (A-11)

X3 =C3 (A-12)

C (A-13) )1/(/)1(33

/)1(22 )( −−− += σσσσσσ ββ CCf

Min (P (A-14) )3322 CPC +

C i =2, 3 (A-15) σσβ −= )/( fiifi PPC

P (A-16) ∑=

−−=3

2

)1/(11 )(i

iif P σσσβ

Cf =f(N, Y, Pf, PN) = eN(Y/PN)η(Pf /PN )-ε ( e: demand shifter) (A-17)

Y = ( NNANNFFAA XPVYXPXPXP +=⇒+− ) (A-18)

LA = g(TA, TN, L) = LA0TAγ1TNγ2Lγ3 γ1 ,γ2 <0 γ3 >0 (A-1

9

XN = TNLNξ

-------------------------------------------------------------------------------------------------------

)

L = LA + LN (A-20)

(A-21)

qnN EPPdNC −= )/( 11 (d: demand shifter) (A-22)

E = GDP / N (A-23)

--

24

25

Effect on X1 70-74 75-79 80-84 85-89 90-96 Effect on X2 70-74 75-79 80-84 85-89 90-96

X1PF -0.01 -0.01 -0.02 -0.02 -0.03 X2PF -0.10 -0.11 -0.12 -0.14 -0.16

X1PN 0.00 0.00 0.00 0.01 0.03 X2PN 0.10 0.11 0.11 0.13 0.42

X1E1 0.75 0.75 0.74 0.71 0.65 X2E1 -0.02 0.18 0.28 0.29 0.28

X1M2 0.00 0.00 0.00 0.00 -0.01 X2M2 -0.02 -0.04 -0.06 -0.11 -0.20

Effect on X3 70-74 75-79 80-84 85-89 90-96 Effect on XA 70-74 75-79 80-84 85-89 90-96

X3PF -0.10 -0.11 -0.11 -0.13 -0.15 XAPF -0.06 -0.05 -0.06 -0.09 -0.12

X3PN 0.10 0.11 0.11 0.13 0.40 XAPN 0.06 0.04 0.05 0.08 0.29

X3E1 -0.02 0.17 0.27 0.27 0.27 XAE1 0.32 0.52 0.52 0.44 0.39

X3M2 0.00 0.00 -0.01 -0.05 -0.12 XAM2 0.00 -0.01 -0.02 -0.04 -0.11

Effect on Cf 70-74 75-79 80-84 85-89 90-96 Effect on P1 70-74 75-79 80-84 85-89 90-96

CfPF -0.08 -0.09 -0.09 -0.08 -0.05 P1PF 0.79 0.96 1.06 1.19 1.32

CfPN 0.08 0.09 0.09 0.07 0.12 P1PN 0.12 -0.04 -0.12 -0.26 -1.08

CfE1 -0.02 0.15 0.22 0.16 0.08 P1E1 5.78 4.47 3.52 3.07 2.66

CfM2 0.17 0.14 0.18 0.39 0.67 P1M2 -0.01 0.01 0.04 0.14 0.41

Effect on P2 70-74 75-79 80-84 85-89 90-96 Effect on P3 70-74 75-79 80-84 85-89 90-96

P2PF 0.18 0.29 0.37 0.40 0.42 P3PF 0.20 0.32 0.41 0.45 0.49

P2PN 0.81 0.70 0.62 0.59 1.56 P3PN 0.79 0.67 0.59 0.54 1.38

P2E1 0.59 0.67 0.49 0.30 0.21 P3E1 0.60 0.62 0.41 0.20 0.09

P2M2 -0.16 -0.25 -0.36 -0.54 -0.84 P3M2 0.02 -0.01 -0.05 -0.15 -0.35

Effect on Y 70-74 75-79 80-84 85-89 90-96 Effect on GDP 70-74 75-79 80-84 85-89 90-96

YPF 0.11 0.19 0.19 0.17 0.15 GDPPF -0.02 -0.02 -0.02 -0.02 -0.03

YPN 0.87 0.79 0.80 0.82 1.63 GDPPN 0.02 0.01 0.01 0.02 0.07

YE1 0.91 1.01 0.72 0.48 0.32 GDPE1 0.09 0.15 0.14 0.12 0.10

YM2 -0.01 -0.01 -0.02 -0.05 -0.09 GDPM2 0.00 0.00 0.00 -0.01 -0.03

Appendix Table 2: Effects of Exogenous Variables on Endogenous Variables

26

Year GR(X1) GR(X1) (%) CX1PF CX1PF (%) CX1PN CX1PN (%) CX1E1 CX1E1 (%) CX1M2 CX1M2 (%)1970-1974 0.43 100.00 -0.06 -13.55 -0.01 -3.30 0.23 53.89 -0.161975-1979 2.26 100.00 0.05 2.04 0.01 0.62 2.01 89.09 0.00 0.051980-1984 -0.53 100.00 -0.27 51.76 0.07 -13.05 0.07 -12.46 0.02 -3.621985-1989 -4.93 100.00 -0.22 4.48 0.09 -1.88 -5.21 105.58 -0.03 0.661990-1996 -1.11 100.00 0.17 -15.42 0.63 -56.77 -1.89 170.34 0.02 -1.80

Year GR(X2) GR(X2) (%) CX2PF CX2PF (%) CX2PN CX2PN (%) CX2E1 CX2E1 (%) CX2M2 CX2M2 (%)1970-1974 4.02 100.00 -0.58 -14.55 1.70 42.34 -0.01 -0.18 0.27 6.711975-1979 8.72 100.00 0.39 4.47 1.44 16.52 4.78 54.79 0.27 3.101980-1984 2.89 100.00 -2.00 -69.37 3.10 107.13 0.03 0.87 1.56 54.071985-1989 -0.92 100.00 -1.46 158.54 2.22 -241.74 -2.13 231.90 -1.13 123.321990-1996 10.45 100.00 0.99 9.49 10.10 96.63 -0.82 -7.88 0.40 3.81

Year GR(X3) GR(X3) (%) CX3PF CX3PF (%) CX3PN CX3PN (%) CX3E1 CX3E1 (%) CX3M2 CX3M2 (%)1970-1974 3.76 100.00 -0.57 -15.13 1.66 44.03 -0.01 -0.18 -0.05 -1.241975-1979 6.89 100.00 0.38 5.46 1.39 20.17 4.61 66.89 0.02 0.231980-1984 1.12 100.00 -1.92 -171.10 2.96 264.24 0.02 2.16 0.35 31.381985-1989 0.78 100.00 -1.38 -177.09 2.11 270.03 -2.02 -259.04 -0.50 -64.711990-1996 8.73 100.00 0.93 10.65 9.47 108.53 -0.77 -8.85 0.25 2.85

Year GR(XA) GR(XA) (%) CXAPFCXAPF

(%) CXAPN CXAPN (%) CXAE1 CXAE1 (%) CXAM2 CXAM2 (%)1970-1974 2.52 100.00 -0.35 -13.79 0.93 36.91 0.10 3.90 0.04 1.761975-1979 14.15 100.00 0.18 1.28 0.58 4.08 13.89 98.19 0.05 0.341980-1984 0.45 100.00 -1.06 -234.78 1.44 320.55 0.05 10.38 0.40 88.891985-1989 -2.31 100.00 -0.95 41.18 1.36 -58.70 -3.28 142.02 -0.48 20.631990-1996 7.35 100.00 0.71 9.71 6.93 94.31 -1.13 -15.38 0.22 3.00

Year GR(Cf) GR(Cf) (%) CCfPF CCfPF (%) CCfPN CCfPN (%) CCfE1 CCfE1 (%) CCfM2 CCfM2 (%)1970-1974 1.35 100.00 -0.47 -35.14 1.38 102.25 -0.01 -0.42 -1.98 -146.751975-1979 4.23 100.00 0.32 7.60 1.19 28.10 3.94 93.18 -0.99 -23.421980-1984 -4.01 100.00 -1.54 38.49 2.38 -59.44 0.02 -0.48 -4.75 118.561985-1989 5.87 100.00 -0.80 -13.66 1.22 20.84 -1.17 -19.99 4.22 71.881990-1996 2.45 100.00 0.27 11.14 2.78 113.50 -0.23 -9.26 -1.35 -55.23

Year GR(P1)GR(P1)

(%) CP1PF CP1PF (%) CP1PN CP1PN (%) CP1E1 CP1E1 (%) CP1M2 CP1M2 (%)1970-1974 -11.61 100.00 4.53 -39.05 2.08 -17.88 1.77 -15.28 0.10 -0.851975-1979 1.66 100.00 -3.29 -198.38 -0.48 -29.16 12.33 742.92 -0.04 -2.431980-1984 17.79 100.00 17.77 99.88 -3.39 -19.03 0.32 1.77 -0.94 -5.281985-1989 -2.51 100.00 12.43 -495.27 -4.35 173.47 -2.67 106.48 1.53 -60.951990-1996 -40.93 100.00 -8.00 19.54 -25.84 63.12 -7.71 18.83 -0.82 2.01

Year GR(P2) GR(P2) (%) CP2PF CP2PF (%) CP2PN CP2PN (%) CP2E1 CP2E1 (%) CP2M2 CP2M2 (%)1970-1974 12.01 100.00 1.03 8.55 13.51 112.48 0.18 1.52 1.90 15.821975-1979 14.24 100.00 -1.00 -7.05 9.03 63.40 7.96 55.87 1.75 12.321980-1984 30.72 100.00 6.23 20.29 16.79 54.67 0.04 0.14 9.35 30.441985-1989 4.26 100.00 4.18 98.14 9.85 231.33 -2.19 -51.52 -5.82 -136.531990-1996 32.61 100.00 -2.53 -7.76 37.28 114.33 -0.59 -1.82 1.70 5.22

Year GR(P3) GR(P3) (%) CP3PF CP3PF (%) CP3PN CP3PN (%) CP3E1 CP3E1 (%) CP3M2 CP3M2 (%)1970-1974 10.92 100.00 1.13 10.34 13.21 120.96 0.18 1.68 -0.21 -1.911975-1979 20.67 100.00 -1.10 -5.30 8.69 42.04 16.83 81.43 0.06 0.281980-1984 22.34 100.00 6.82 30.54 15.88 71.10 0.04 0.16 1.28 5.711985-1989 9.45 100.00 4.70 49.69 9.07 95.98 -1.44 -15.25 -1.62 -17.121990-1996 27.52 100.00 -2.94 -10.68 33.13 120.37 -0.25 -0.92 0.71 2.57

Year GR(GDP)GR(GDP)(%

) CGDPPFCGDPPF

(%) CGDPPNCGDPPN

(%) CGDPE1CGDPE1

(%) CGDPM2 CGDPM2(%)1970-1974 2.92 100.00 -0.10 -3.38 0.26 9.06 0.03 0.96 0.01 0.431975-1979 6.29 100.00 0.05 0.86 0.17 2.71 4.10 65.15 0.01 0.231980-1984 0.61 100.00 -0.29 -47.99 0.40 65.52 0.01 2.12 0.11 18.171985-1989 1.82 100.00 -0.26 -14.03 0.36 19.92 -0.88 -48.20 -0.13 -7.001990-1996 3.63 100.00 0.17 4.74 1.71 47.13 -0.28 -7.68 0.05 1.50

Appendix Table 3: Contribution of Exogenous Variables to the Endogenous Variables

Percentage Contribution of Exogenous Variables to Value of Production of Exportable Commodities

-0.00

(X1)

Percentage Contribution of Exogenous Variables to Value of Production of Import Substitute Food Commodity (X2)

Percentage Contribution of Exogenous Variables to Value of Production of Domestically Produced and Consumed Food Commodity (X3)

Percentage Contribution of Exogenous Variables to Average Price of Domestically Produced and Consumed Food Commodity (P3)

Percentage Contribution of Exogenous Variables to GDP

Percentage Contribution of Exogenous Variables to Agricultural Output(XA)

Percentage Contribution of Exogenous Variables to Food Consumption (Cf)

Percentage Contribution of Exogenous Variables to Average Price of Export Agricultural Output(P1)

Percentage Contribution of Exogenous Variables to Average Price of Import Substitute Food Commodity (P2)