Economies of Scale in Indonesian Rice Production: An Economic Analysis Using PATANAS Data

Abstract

インドネシアの稲作は中部ジャワや東部ジャワなどの主要生産地帯において，その零細性がゆえに国際競争力を失っている．競争力を回復するには，少なくとも，規模拡大のための政策が不可欠である．本論文は，稲作部門において規模の経済性が働きうるのか否かを，インドネシア農業省が作成したPATANASデータを用いて検証することを目的とする．そのことと頑強性を確保することを目的として，2つの異なる方法（生産の費用弾力性の測定，および，農家の利潤極大化行動により導出された利潤関数による測定）でこの仮説を検定する．分析の結果，規模の経済は，ほぼ全部の州において働くことが確認された．この結果は，平均経営規模が，零細で，ほとんどの稲作農家が農業機械を賃借している地域ですら達成されることを意味する．この結果は，また，ジャワ島ですら，経営規模の拡大の必要条件が満たされていることを示している．

1.  Introduction

Indonesia has achieved remarkable economic growth in the past decade. If Indonesia continues to grow at the same pace, the country could become a middle-income country by 2020. However, if Indonesia becomes an upper middle-income country, its agricultural sector would face structural adjustment problems similar to those experienced by Asian forerunners, Japan and South Korea. From the viewpoint of food security (Otsuka, 2013), rural-urban income disparity (ADB, 2012), and industrial sector development (Fukui, 2008), solving the structural adjustment problem of the agricultural sector is one of the most important policy issues for the Indonesian government.

The staple food of Indonesia, rice, has lost global competitiveness in the main producing areas of Central Java and East Java, due to small farm size. In order to recover rice competitiveness, indispensable policy measures to enlarge farm size are required (Antriyandarti, 2015).

Although achieving economies of scale is a necessary condition for enlargement of farm size, economies of scale might not be achieved in the Indonesian rice sector because most rice-growing farmers do not have their own agricultural machinery, but rent it from owners. Therefore, we need to examine whether economies of scale can be achieved in this sector.

Among the existing studies related to scale economies, Yamauchi (2014) investigates the technological progress of the Indonesian rice sectors, using unique data collected from wide areas, and finds a complementary relationship between land and machinery only in the outer islands, where the average farm size is larger than that of Java. Llewelyn and Williams (1996) examine the technical efficiency of multi-crop producing farms in Indonesia, not only rice, and some of the factors associated with inefficiency. They find that the majority of farms are technically and scale efficient, operating at constant returns to scale, while most farms that are scale inefficient are operating at decreasing returns to scale.

There are only two studies that focus on economies of scale in the Indonesian rice sector (Hidayah and Susanto, 2013; Hidayah et al., 2013), although these studies examine the scale economies of rice pro­duction in the Maluku province only, which is not a main rice producing area in Indonesia.

The empirical methodology used in this study is not new. No previous studies on the scale economies of the Indonesian rice sector (Llewelyn and Williams, 1996; Hidayah and Susanto, 2013; Hidayah et al., 2013; Yamauchi, 2014) have found scale economies in Java. These findings are significant when we consider structural adjustment policy for the Indonesian rice sector. The objective of this study is to investigate whether economies of scale can be achieved in the five main rice-growing provinces of Indonesia, using National Farmers’ Household Panel Survey (PATANAS¹) data collected by the Indonesian Ministry of Agriculture.

In order to achieve the study objective and robustness of testing, we examine the existence of scale economies via two different methods. First, we estimate the cost elasticity of production by regressing output on production cost, following Hayami and Kawagoe (1989). Second, we estimate the profit function derived from the maximization problem of the rice farmer, and test the existence of scale economies using the estimated parameters, following Lau and Yotopoulos (1971), Yotopoulos and Lau (1973), and Kako (1984).

The organization of the rest of this paper is as follows. We estimate the cost elasticity of output using cost and return data in Section 2. Section 3 investigates the existence of scale economies by using the parameters of land and capital stock estimated from profit function analysis. We conclude in Section 4.

2.  Cost elasticity of scale

We use the following regression equation to estimate the cost elasticity of scale (output elasticity), following Hayami and Kawagoe (1989).

  

$$ln{C}_{ij}= \alpha +\beta ln{Q}_{ij}+{\epsilon }_{ij}$$（1）

Here,

C = Production cost of paddy (IDR)²

Q = Production of paddy (kg)

ε = Random disturbance

i,j = Farm household, province

The production cost includes variable and fixed cost. The estimation results are shown in Table 1. According to Table 1, the estimated cost elasticity of output is less than 1 in all the provinces. In particular, in three of Java’s provinces, these values are significantly smaller than 1. This implies that economies of scale can be achieved in Java.

Table 1.  Estimation of Cost Elasticity of Rice Production in Indonesia

Variable	North Sumatra		West Java		Central Java
	Coefficient	SE	Coefficient	SE	Coefficient	SE
Q	0.9814***	0.0283	0.7063***	0.0302	0.6420***	0.0193
Constant	6.7905***	0.2209	9.1997***	0.2401	9.5035***	0.1353
Number of Obs.	215		439		622
F-Value	1201.22***		546.16***		1101.39***
R2	0.8494		0.5555		0.6398
F(β=1)	13.20***		159.66***		260.78***
Variable	East Java		South Sulawesi
	Coefficient	SE	Coefficient	SE
Q	0.6635***	0.0250	0.8841***	0.0363
Constant	9.1410***	0.1764	7.4846***	0.2721
Number of Obs.	586		295
F-Value	702.43***		592.90***
R2	0.5460		0.6693
F(β=1)	250.60***		20.19***

1) *Significant at 10% level; **Significant at 5% level; ***Significant at 1% level.

3.  Profit function approach

Next, we apply a profit function approach to examine economies of scale. For that purpose, we simultaneously estimate a Cobb–Douglas unit output price profit function and factor demand functions derived from conditions of profit maximization, following Lau and Yotopoulos (1971), Yotopoulos and Lau (1973), and Kako (1984). The profit function of the Cobb–Douglas form assumes that the production function also takes the Cobb–Douglas form. The profit function can be specified as equation 2.

  

$$\begin{array}{cc}ln{\pi }_{ij}=& ln{A}_{ij}^{*}+{\beta }_{T}ln{T}_{ij}+{\alpha }_{L}ln{L}_{ij}+{\alpha }_{S}ln{S}_{ij}+\\ \hfill & {\alpha }_{F}ln{F}_{ij}+{\beta }_{Tr}ln{Tr}_{ij}+{\epsilon }_{ij}\hfill \end{array}$$（2）

Here,

π = Real profit (IDR)

A* = Constant term in natural logarithms of profit function³

T = Farm size (ha)

L = Labor wage/rice price (IDR/HOK⁴/rice price)

S = Seed price/rice price (IDR/kg/rice price)

F = Fertilizer price/rice price (IDR/kg/rice price)

Tr = Tractor rent cost divided by rice price (IDR/rice price)

ε = Random disturbance

The following input factor demand functions are derived from profit maximization behavior. The input demand function is defined by equation 3.

  

$$-\frac{P_i{X}_i}{\pi }={\alpha }_i^{*}(i=L,S,F)$$（3）

Here, P_i is the price of input and X_i is the quantity of input. ${\alpha }_i^{*}$ denotes a parameter of input i which satisfies the conditions of profit maximization. However, there is some debate about the hypothesis of profit maximization in Java (Benjamin, 1992; Mulyo and Fukui, 2006).

Therefore, we estimate two types of models. One has a profit maximization restriction, and the other does not (α_i’). We statistically examine which type of model is more appropriate [H₀: ${\alpha }_i^{*}$=α_i’ for i=L, S, F (profit maximization)]. We calculate the F-ratio by using the estimated residuals of the restricted and unrestricted equations. The profit function and derived system of input demand are estimated simultaneously by applying the seemingly unrelated regression method.

We then test the null hypothesis of constant returns to scale, as follows:

  

$${\beta }_{Tr}^{*}+{\beta }_T^{*}=1$$（4）

Here, ${\beta }_{Tr}^{*}$ is the estimated parameter of the machinery, and ${\beta }_T^{*}$ is that of the land. Then, we examine the scale economies, according to equation (5).

  

$${\beta }_{Tr}^{*}+{\beta }_T^{*}\lessgtr 1$$（5）

The estimation results of profit functions and factor demand functions with restriction of profit maximization, as well as those without it, are shown in Table 2.

Table 2.  Cobb-Douglas profit and factor demand functions by seemingly unrelated regression

Variable Profit Function	North Sumatra		West Java		Central Java
	3 Restriction αi*=αi*’(i=L, S, F)		No Restriction		No Restriction
	Coefficient	SE	Coefficient	SE	Coefficient	SE
Constant	8.0188***	0.3546	8.1096***	0.4438	7.0193***	0.5056
Ln Farm size	1.0745***	0.0529	1.9580***	0.0415	1.3735***	0.0612
Ln Labor wage/rice price	–0.1645***	0.0458	–0.1103***	0.0084	–0.3639***	0.0770
Ln Seed price/rice price	–0.1703**	0.0960	–0.2389***	0.0599	–0.3438***	0.0854
Ln Fertilizer price/rice price	–0.2962**	0.1058	–0.1979***	0.0380	–0.9189***	0.1502
Ln Tractor rent cost/rice price	0.1358**	0.0767	0.1365**	0.0088	0.0495	0.1086
R2	0.7256		0.5866		0.4533
Factor Demand Function
Labor Demand	–0.1645***	0.0458	–0.0138***	0.0024	–0.0641***	0.0066
R2	0.8202		0.2727		0.4013
Seed Demand	–0.1703**	0.0960	–0.0521***	0.0016	–0.0106***	0.0059
R2	0.7361		0.067		0.2770
Fertilizer Demand	–0.2962**	0.1058	–0.0294***	0.0016	–0.0129***	0.0019
R2	0.6538		0.5160		0.3426
Number of Obs.	215		439		622
F-Value	0.31		4.71***		5.80***
Chi2(${\beta }_{Tr}^{*}$+${\beta }_T^{*}$=1)	13.75***		39.06***		5.09**
Variable Profit Function	East Java		South Sulawesi
	No Restriction		No Restriction
	Coefficient	SE	Coefficient	SE
Constant	5.6594***	0.4365	7.5131***	0.3997
Ln Farm size	1.7420***	0.0576	1.3926***	0.0422
Ln Labor wage/rice price	–0.0142	0.0726	–0.1007*	0.0538
Ln Seed price/rice price	–0.0313	0.0586	–0.1703***	0.0554
Ln Fertilizer price/rice price	–0.2927***	0.0978	–0.4154**	0.1679
Ln Tractor rent cost/rice price	0.4630***	0.0848	0.1453**	0.0787
R2	0.4393		0.6814
Factor Demand Function
Labor Demand	–0.0243***	0.0060	–0.0428***	0.0013
R2	0.1622		0.2684
Seed Demand	–0.0231***	0.0042	–0.0419***	0.0012
R2	0.1405		0.2414
Fertilizer Demand	–0.0244***	0.0020	–0.0204***	0.0056
R2	0.3284		0.2812
Number of Obs.	586		295
F-Value	8.78***		8.66***
Chi2(${\beta }_{Tr}^{*}$+${\beta }_T^{*}$=1)	29.54***		9.14***

1) *Significant at 10% level; **Significant at 5% level; ***Significant at 1% level.

From the F-values shown in Table 2, we find that the null hypothesis of profit maximization is rejected for all the provinces, except North Sumatra. This indicates that rice-growing farmers do not maximize their profits, with the exception of farmers in North Sumatra.

Therefore, the null hypothesis of constant returns to scale is rejected for all provinces, where the estimated sum of the parameters (${\beta }_{Tr}^{*}$+${\beta }_T^{*}$) is larger than 1. All provinces show the existence of scale economies.

These results are somewhat consistent with the results shown in section 2.

4． Conclusion

The Indonesian government is set to face structural adjustment problems for agriculture under the pressure of trade liberalization (Yonekura, 2014). In addition, current rice policies of Indonesia (e.g., import bans and input subsidies) are criticized by some international organizations, as well as Western economists, because these policies could potentially increase the budget deficit and lead to inefficient resource allocation (McCulloch, 2008; OECD, 2013; Warr, 2011).

Thus, the enlargement of farm size in the rice sector is one of the most important policy measures to cope with the problem of structural adjustment problems and trade liberalization in the agricultural sector. The achievement of economies of scale is a necessary condition to enlarge farm size in rice production.

From the analysis results, we found that economies of scale could be achieved in almost all provinces. The finding that economies of scale can be achieved in Java is inconsistent with Yamauchi (2014), who could not find a complementary relationship between land and machinery in the Java islands, where the average farm size is extremely small. Our findings suggest that if there were an increase in farm size, profit would increase more than proportionally to the increase of land and fixed capital, even in Java. This difference is caused by the definition of fixed capital. We assume the rental cost of machinery is also a fixed cost. However, Yamauchi (2014) does not.

Our findings imply that economies of scale can be achieved, even in areas where average farm size is very small, and where most farmers rent machinery. In addition, the findings suggest that a necessary condition for the enlargement of farm size has already been met in Java.

Acknowledgments

We appreciate that this work was financially supported by JSPS Grant-in-Aid Scientific Research (No. 26292118).

Notes

1  For details of PATANAS data, see Antriyandarti (2015).

2  IDR=Indonesian Rupiah.

3  A* is composed of the level of technology in production function, the parameters of variable inputs, and the ratio of marginal product of input to input price.

4  HOK=Workday of labor.

References

• Asian Development Bank (ADB) (2012) Asian Development Outlook 2012: Conflicting Rising Inequality in Asia. Manila: Asian Development Bank.
•  Antriyandarti,  E. (2015) Competitiveness and cost efficiency of rice farming in Indonesia. Journal of Rural Problems, 51(2), 74–85.
•  Benjamin,  D. (1992) Household composition, labor markets, and labor demand: Testing for separation in agricultural household models. Econometrica, 60(2), 287–322.
• Fukui, S. (2008) Economic development and agriculture. In: M. Takahashi and S. Fukui (Ed.) Theories and Policies in Economics Development. Keiso Shobo, 113–130 (in Japanese).
•  Hayami,  Y. and  Kawagoe,  T. (1989) Farm mechanization, scale economies and polarization. Journal of Development Economics, 31(2), 221–239.
•  Hidayah,  I.,  Hanani,  N.,  Anindita,  R. and  Setiawan,  B. (2013) Production and cost efficiency analysis using frontier stochastic approach, a case on paddy farming system with Integrated Plant and Resource Management (IPRM) approach in Buru District Maluku Province Indonesia. Journal of Economics and Sustainable Development, 4(1), 78–84.
•  Hidayah,  I. and  Susanto,  A. N. (2013) Economies of scale and allocative efficiency of rice farming at West Seram Regency, Maluku Province, Indonesia. Asian Economic and Financial Review, 3(5), 624–634.
•  Kako,  T. (1984) Efficiency and economies of scale in rice production: A case of Ishikari Area in Hokkaido. Journal of Rural Economics, 56(3), 151–162 (in Japanese).
•  Lau,  L. J. and  Yotopoulos,  P. A. (1971) A test for relative efficiency and application to Indian agriculture. The American Economic Review, 61(1), 94–109.
•  Llewelyn,  R. V. and  Williams,  J. R. (1996) Nonparametric analysis of technical, pure technical, and scale efficiencies for food crop production in East Java, Indonesia. Agricultural Economics, 15(2), 113–126.
•  McCulloch,  N. (2008) Rice prices and poverty in Indonesia. Bulletin of Indonesia Economic Studies, 44(1), 45–63.
•  Mulyo,  J. H. and  Fukui,  S. (2006) Testing the separation hypothesis in rural Java: Revisited. Journal of Rural Problems, 42(2), 224–236.
• Organisation for Economic Co-operation and Development (OECD) (2013) Agricultural Policy Monitoring and Evaluation 2013. OECD Countries and Emerging Economies. Paris: OECD Publishing. (http://dx.doi.org/10.1787/agr_pol-2013-en) [Accessed November 16, 2013].
•  Otsuka,  K. (2013) Food inequality, income inequality, and the changing comparative advantage in world agriculture. Agricultural Economics, 44(supplement), 7–18.
• Warr, P. (2011) Food security vs food self-sufficiency: The Indonesian case (Working Papers in Trade and Development 04). Canberra: Crawford School of Economics and Government, ANU College of Asia and the Pacific, The Australian National University.
• Yamauchi, F. (2014) Wage growth, landholding, and mechanization in agriculture: Evidence from Indonesia (World Bank Policy Research Working Paper No 6789). Washington D.C.: The World Bank. (http://www-wds.worldbank.org/external/default/WDSContentServer/WDSP/IB/2014/02/27/000158349_2014227112151/Rendered/PDF/WPS6789.pdf.) [Accessed May 11, 2015].
•  Yonekura,  H. (2014) Recent trend of agriculture and food policies in Indonesia: Food security and trade liberalization. Nogyo to Keizai, 80(2), 87–97 (in Japanese).
•  Yotopoulos,  P. A. and  Lau,  L. J. (1973) A test for relative economic efficiency: Some further results. The American Economic Review, 63(1), 214–223.