インドネシア稲作の競争力と費用効率性

Abstract

本稿は，インドネシア稲作部門の費用効率性が改善されることによって，どの程度，競争力が向上するかについて検証することを目的とする．そのために，まず，農業省のPATANASデータを用い，確率的フロンティア関数分析の手法により，主要なコメ生産地域5州における稲作経営の費用非効率性を測定する．さらに，計測された費用非効率性の値と，経営規模など他の関連変数とを，別途推計した国内資源費用比率（Domestic Resource Cost Ratio）に回帰することにより，費用非効率性への影響を推計する．その結果，費用効率性は，経営規模が小さな中部ジャワ，および，東部ジャワにおいて高い一方，これらの地域における国際競争力は低く輸入米と競争できないことが明らかとなった．また，費用非効率性，その他の関連変数とDRCの回帰分析の結果は，中部ジャワ，東部ジャワにおいて，費用非効率性の改善だけでは輸入米と競争可能な水準まで国際競争力を向上させることはできず，規模の拡大が必要であることを示した．以上の分析結果から，経営規模が零細で競争力の弱い中部ジャワや東部ジャワにおける競争力向上のためには，農家の規模拡が不可欠であると結論付けることができるのだが，ジャワの分散零細錯圃制度を考えると，その実現は容易ではないであろう．

1.  Introduction

(1)  Background

In Indonesia where per capita income rises up to USD 3,475 (World Bank, 2014), agriculture is one of the key strategic sectors identified by governments (OECD, 2012). In Indonesian agriculture, rice is the most important food crop. In the recent mid-term five years development plan, Indonesian government attaches importance to food security and plans to achieve the self-sufficiency of rice by strengthening rice policy, input subsidy program, and government procurement and reserve (Natawidjaja et al., 2012). However, domestic rice price has been higher than the international rice price since mid 2000’s except the duration of food crisis in 2008 and 2009 under the current protective rice policy. This suggests that if international trade of rice will be liberalized, self-sufficiency will not be achieved.

In addition, to alleviate the impact of high domestic price on the poor, Indonesian government has implemented a cheap rice distribution program for the poor (Raskin). However, the expenditures for this program and input subsidy program generated a large budget deficit for the government (OECD, 2013).

For the current Indonesian rice policies, which have serious problems as mentioned above, there are some debates.

Hadi et al. (2005), Natawidjaja et al. (2012), Irianta (2005), and Paasch et al. (2007) supported the implementation of protection policies because the policies provide a positive impact on improving the competitiveness and profitability of rice farming.

Fig. 1 

Domestic and World Rice Prices

Source: BULOG in Natawidjaja et al. (2012)

In contrast, international organization and Western economists criticized the policy packages, including import ban, rice procurement and stock reserve by BULOG, input subsidy program, and Raskin program. Moreover, they also forecasted that if the government continues these policies in the future, the global competitiveness of rice sectors in which small scale farms will be persistent under those policies will be declining; as a result, the government will have to increase expenditure for the rice sector and the poor (Timmer, 2004; McCulloch, 2008; OECD, 2012; OECD, 2013; Warr, 2005). Warr (2011) also suggested that a preferable strategy for achieving self-sufficiency is not to provide input subsidies but to promote productivity in the rice sector.

The fact that domestic rice price is always higher than world price implies that the Indonesian rice sector does not have global competitiveness. Therefore, the opinion that the current protective rice policies improved global competitiveness is not acceptable. McCulloch and Timmer (2008) asserted that reliance exclusively on domestic production results in much higher domestic rice prices, which increase poverty and make it harder for poor households, including farmers, to ensure their own food security.

If Indonesian Government makes a free trade agreement with ASEAN countries in the near future on schedule, it cannot help liberalizing the strict current protection policy, such as import ban (Yonekura, 2014). In such a situation, Indonesia will have to make efforts to enhance the productivity of rice if it tries to enjoy the benefits of Free Trade Agreements.

(2)  Objectives and significance of study

In order to enhance the productivity, in general, there are at least three measures. First is to improve technical and allocative efficiency. Second is to enlarge farm size to enjoy scale economy. Third is to develop new technologies that enhance productivity and profitability.

We focus on the possibility of first measure and investigate whether the Indonesian rice sector can improve the technical and allocative efficiency and whether it can obtain global competitiveness by improving the efficiency. Thus, we do not examine the scale economy due to the focus on global competitiveness of Indonesian rice farming.

As for the investigation of the competitiveness and efficiency of rice farming in Indonesia, many studies have been already conducted (Romdhon and Cahyadinata, 2004; Hidayah et al., 2013a, 2013b; Jamal and Dewi, 2009; Kusnadi et al., 2011; Makki et al., 2012; Effendy et al., 2013; Brazdik, 2006; Wikstrom and Fleyeh, 2013).

However, these works have serious problems to make a close investigation. Romdhon and Cahyadinata (2004), Hidayah et al. (2013a, 2013b) focused only on local rice production, such as provincial level; therefore, their findings cannot be extended to other areas in Indonesia. In addition, most studies use a production function model for estimating stochastic frontier function. However, production function approach has several weak points. First, it is difficult to estimate significant coefficients of inputs due to multi-collinearity. Second, if we use it, we cannot estimate the impacts of technical and allocative efficiency on the production cost.

Finally, most of them use the cross-sectional data; however, if we use frontier function approach, it is more desirable for us to use the pooled data (Kumbhakar et al., 2000). Thus, the relationship between the efficiency and competitiveness of rice has not been well explored in the literature, although there are many studies of Indonesian rice policy, which focus on the determinants of efficiency.

This paper makes two significant contributions. First, to the best of our knowledge, this is the first study that aims to examine the impacts of efficiency improvement on the global competitiveness of Indonesian rice. Second, this paper is among the few studies to measure the cost efficiency of Indonesian rice by applying a stochastic frontier approach not using the production function model but using a cost function model and by using PATANAS¹ data, which contain a rich pooled data of production cost in five provinces that are the main rice production areas in Indonesia.

(3)  Organization of paper

The organization of this paper is as follows. Section 2 explains the analytical framework and econometric model and presents our estimation results of technical inefficiency. Section 3 explains how to measure Domestic Resource Cost, which is an indicator of global competitiveness and estimates the effects of technical efficiency on DRC. Final section summarizes the major findings and draws policy implications.

2.  Estimation of technical inefficiency using stochastic frontier cost function

To estimate the technical inefficiency of rice production in Indonesia, we apply the stochastic frontier function approach. The word frontier may appropriately apply to each case because the function sets a limit to the range of possible observation. The amount by which a farmer lies below its production frontier and the amount by which it lies above its cost frontier can be regarded as the measures of inefficiency. All deviations from the frontier are assumed to be the result of technical inefficiency (Coelli et al., 2005).

The study does not use the DEA approach because this study uses information, assumptions and large samples. Statistical conclusions cannot be drawn when using nonparametric method. Parametric approach enters random error on the frontier, while the DEA approach does not incorporate random error. As a consequence, the stochastic frontier approach can consider the inefficiency factors such as land tenure, number of plot and off farm occupation. This is the strong point of stochastic frontier analysis comparison to DEA in which it takes into account measurement errors and other noise in the data (Kumbhakar et al., 2000; Coelli et al., 2005). This point is very important for studies of farm households in developing economy like Indonesia as data generally include measurement errors. We use a Cobb–Douglas stochastic frontier model because if we use translog model, we face the results will not be good.

This study follows the method of estimating a stochastic frontier cost function proposed by Aigner et al. (1977), Coelli et al. (2005) Kumbhakar et al. (2000), Debertin (1986) and Hidayah et al. (2013b). The cost function is specified as:

  

$$lnC={\alpha }_0+{\alpha }_{1}lnP{x}_1+{\alpha }_{2}lnP{x}_2+{\alpha }_{3}lnP{x}_3+{\alpha }_{4}lnP{x}_4+{\alpha }_{5}lnP{x}_5+{\alpha }_{6}lnP{x}_6+{\alpha }_{7}lnP{x}_7+{\alpha }_{8}lnP{x}_8+\left(V_i+U_i\right)$$（1）

where

C = Total production cost (IDR)

Px₁ = Land rent (IDR/ha)

Px₂ = Seed price (IDR/kg)

Px₃ = Labor wage (IDR/HOK)

Px₄ = Fertilizer price (IDR/kg)

Px₅ = Tractor rent cost (IDR/HOK)

Px₆ = Draft animal rent cost (IDR/HOK)

Px₇ = Irrigation cost (IDR/ha)

Px₈ = Production (kg)

U_i = Farmer specific characteristics related to cost inefficiency

V_i = Statistical disturbance term

Table 1 presented the input prices and production per plot in the surveyed area. The estimation result of Cobb–Douglas Frontier Cost Function shown by Table 2. We estimate the Cobb–Douglas frontier cost function with heteroscedasticity. The frontier model allows heteroscedasticity in either error term as a linier function of a set of covariates. We specify covariates for both variance of the U_i andV_i. The component of inefficiency and statistical disturbance term are heteroscedastic (Kumbhakar et al., 2000; Greene, 2005).

Table 1.  Input Prices and Production per Plot

Variable	North Sumatra	West Java	Central Java	East Java	South Sulawesi
Production/ha (kg)	5,519	6,535	6,182	6,513	5,199
Total Cost (IDR)	1,773,168	3,241,608	1,498,079	1,404,210	1,704,945
Production Cost/kg (IDR/kg)	816	986	1041	853	781
Land Rent (IDR/ha)	1,348,805	2,700,545	3,363,138	1,847,569	2,318,297
Seed Price (IDR/kg)	12,767	7,232	6,810	8,450	4,020
Labor Wage (IDR/HOK)	17,982	23,361	22,723	23,231	32,685
Fertilizer Price (IDR/kg)	1,713	1,740	1,611	1,577	1,606
Tractor Rent Cost (IDR/HOK)	247,853	327,517	189,485	158,599	305,444
Draft Animal Rent Cost (IDR/HOK)		169,107		66,578	158,403
Irrigation Cost (IDR/ha)	140,745	109,900	227,513	92,873	31,800

Source: PATANAS 2007–2012

1) IDR = Indonesian Rupiah

2) USD 1 = IDR 9,272 (in 2012)

3) HOK = Workday of labor

Table 2.  Estimation of Cobb–Douglas Frontier Cost Function

Variable	North Sumatra		West Java		Central Java
	Coefficient	SE	Coefficient	SE	Coefficient	SE
Land Rent	0.4448***	0.0689	0.2142*	0.1180	0.0476	0.0439
Seed	−0.0373	0.0568	−0.0275	0.0369	−0.0309	0.0356
Labor Wage	0.1589***	0.0307	0.1737***	0.0437	0.1109***	0.0341
Fertilizer	0.5841***	0.1171	0.1479	0.1150	0.2660***	0.0925
Tractor	0.1498**	0.0613	0.4328***	0.0486	0.4488***	0.0449
Draft Animal			0.1884***	0.0653
Irrigation	0.0117	0.0213	0.0508**	0.0225	0.0789***	0.0172
Production	0.7381***	0.0578	0.6159***	0.0347	0.4029***	0.0242
Constant	−4.7632	1.3099	−2.7791*	1.4678	1.5626*	0.8627
Number of obs.	215		439		622
Log likelihood	−46.4066		−235.8139		−375.4097
Sigma–square	0.2916		0.2474		0.4130
Average CE	0.4846		0.5881		0.6557
Variable	East Java		South Sulawesi
	Coefficient	SE	Coefficient	SE
Land Rent	0.1849**	0.0869	0.1595**	0.0721
Seed	0.0328	0.0366	0.1653***	0.0466
Labor Wage	0.1301***	0.0344	0.1874***	0.0340
Fertilizer	0.4173***	0.1379	0.0616	0.0985
Tractor	0.1625***	0.0421	0.1899***	0.0414
Draft Animal	0.2738**	0.1194	0.9928***	0.1250
Irrigation	0.0301	0.0248	0.1473***	0.0357
Production	0.4614***	0.0269	0.5551***	0.0382
Constant	−1.5129	0.9346	−12.363***	1.2742
Number of obs.	586		295
Log likelihood	−408.8667		−45.9371
Sigma–square	0.6279		0.2089
Average CE	0.6142		0.4382

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

The cost efficiency is the possible minimum cost ratio with specific inefficiency level toward actual total cost (Coelli et al., 2005; Kumbhakar et al., 2000). The cost efficiency index eU is calculated from the inverse of eU = q/f(x); q: actual cost, f(x): cost on the frontier function. Therefore, the cost efficiency is defined as the percentage achievement of production cost by best practice. Thus, the estimated cost efficiency indices calculated from frontier function of one province can be compared with the cost efficiency indices of another province.

Variable of land rent is significantly positive in North Sumatra, West Java, East Java, and South Sulawesi. The coefficient of seed price is significantly positive in South Sulawesi. The coefficient of labor wage is significantly positive in all regions. The increase in labor wage leads to an increase in the total cost. Fertilizer price are significant positive in all regions except in West Java and South Sulawesi. The tractor cost was also significantly positive in all regions. Similarly, the coefficients of draft animal are significantly positive in West Java, East Java and South Sulawesi. The coefficients of irrigation cost are significantly positive in West Java, Central Java and South Sulawesi.

The coefficient of production is significantly positive in all regions. These results suggest that an increase in production increases the total cost. Thus, the estimation results are mostly plausible and consistent with the theory.

The result of efficiency analysis shows that average cost efficiency is 0.4846 (North Sumatra), 0.5881 (West Java), 0.6557 (Central Java), 0.6142 (East Java), and 0.4382 (South Sulawesi), indicating that the rice farming system in Central Java has the highest cost efficiency.

To identify the determinants of cost inefficiency, we estimate the following regression equation.

  

$$\left|U_i\right|={\delta }_0+{\delta }_1{Z}_1+{\delta }_2{Z}_2+{\delta }_3{Z}_3+{\delta }_4{Z}_4+{\delta }_5{Z}_5+{\delta }_6{Z}_6+{\delta }_7{Z}_7+{\delta }_8{Z}_8+{\delta }_9{Z}_9+e$$（2）

Where

U_i = Cost inefficiency effects

Z₁ = Age of household head (years)

Z₂ = Education of household head (years)

Z₃ = Number of family member (person)

Z₄ = Kind of land (if paddy land with technical irrigation = 1; others = 0)

Z₅ = Location of land (if the same village with farmer’s house = 1; otherwise = 0)

Z₆ = Land tenure (if owned land = 1; others = 0)

Z₇ = Off Farm Occupation (if engage off farm activity = 1; otherwise = 0

Z₈ = Farm size of farmer (ha)

Z₉ = Number of Household’s Plot

e = Error term

The farm household characteristics used for estimation (2) are illustrated in Table 3. The result of the determinants of cost inefficiency is presented in Table 4. We apply linear regression model with White-robust standard error in the presence of heteroscedasticity (Baum, 2006). We explain only the results related to the discussion in the latter part. The age of household head has negative effect on inefficiency in Central Java and East Java. The older household head is more experienced and able to make efficient decision on rice farming. However, the estimated coefficient of age of household in South Sulawesi is significantly positive. This seemingly contradictive result can be explained if younger household head more adaptable to accept and implement the new innovation technology of rice farming that can achieve high efficiency. In Java, many young people prefer working in cities than in the agriculture sector.

Table 3.  Potential Determinants of Cost Inefficiency

Variable	North Sumatra	West Java	Central Java	East Java	South Sulawesi
Age: Age of Household Head (year)	44	47	50	48	45
Eduation: Education of household head (year)	7	7	8	7	8
Family Member: Number of family member (person)	5	4	4	4	5
Kind of Land: % of HH using paddy land with	79	81	80	80	79
technical irrigation: —Land lord (person)	153	294	479	492	206
—Tenant (person)	62	145	143	94	89
Location of Land: % of land is the same village with HH’s house	93	82	95	96	92
Land Tenure: % of land is HH’s owned land	71	67	77	84	70
—All owned (person)	140	259	429	437	181
—Mixed (owned and rent) (person)	60	141	134	96	89
—All rent (person)	25	39	59	53	25
Off-farm Occupation: % of HH engage off farm activity	79	67	83	77	75
Farm Size (ha)	0.396	0.572	0.344	0.349	0.43
Plot: Number of Household’s Plot	3	4	4	3	4

Source: PATANAS 2007–2012

Table 4.  Determinants of Cost Inefficiency

Variable	North Sumatra		West Java		Central Java
	Coefficient	SE	Coefficient	SE	Coefficient	SE
Age	0.0008	0.0014	−0.0002	0.0011	−0.0033**	0.0016
Education	−0.0011	0.0036	0.0003	0.0026	−0.0022	0.0045
Family Member	0.0144*	0.0087	0.0151*	0.0082	0.0029	0.0106
Kind of Land	0.0532	0.0379	−0.0550**	0.0247	−0.0260	0.0430
Location of Land	−0.0247	0.0481	−0.0043	0.0261	0.0047	0.0514
Land Tenure	0.1233***	0.0212	0.0508***	0.0195	0.1121***	0.0355
Off Farm Occupation	−0.0480	0.0308	−0.0178	0.0208	−0.0394	0.0363
Farm Size	−0.0167	0.0428	−0.0616***	0.0217	0.0154	0.0395
Plot	0.0107	0.0066	0.0011	0.0043	0.0134**	0.0062
Constant	0.2372	0.1186	0.4954***	0.0831	0.1213	0.1266
Number of obs.	215		439		622
F Calculated	10.5930***		16.7880***		29.43***
R2	0.2815		0.2905		0.3227
Variable	East Java		South Sulawesi
	Coefficient	SE	Coefficient	SE
Age	−0.0058*	0.0033	0.0024**	0.0011
Education	−0.0052*	0.0028	0.0025	0.0026
Family Member	0.0013	0.0229	0.0044	0.0059
Kind of Land	0.0458	0.0838	−0.0142	0.0124
Location of Land	−0.1923***	0.0693	0.0206	0.0271
Land Tenure	−0.1010	0.0786	0.0461**	0.0206
Off Farm Occupation	−0.0969	0.0864	−0.0205	0.0218
Farm Size	−0.1861**	0.0898	−0.1113***	0.0388
Plot	0.0144	0.0111	−0.0010	0.0027
Constant	0.8103***	0.2154	0.1972***	0.0751
Number of obs.	586		295
F Calculated	57.53***		11.96***
R2	0.4184		0.3809

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

The education of household head has negative influence on the inefficiency of rice farming in East Java. The result indicates that farmer who has higher education can manage rice farming more efficiently. The number of family member has positive effect on the inefficiency in North Sumatra and West Java. The household with larger family member will face the larger expenditure that make the household difficult to focus on rice farming. The coefficient of type of paddy land is significantly negative in West Java. If the farmers can utilize more developed irrigation system, inefficiency will be reduced. This is because farmers can manage paddy in a better manner if the irrigation system is improved (Mulwanyi et al., 2011).

The location of land also has negative influence on the inefficiency of rice farming in East Java. If the location of the farm land is the same village with the farmer’s house, the farmer can manage the rice farming more easily. The land ownership has positive effect on inefficiency in all regions, excluding East Java. This is highly related to the behavior of farmers in the study area where farmers who rent land have higher motivation and productivity compared to the owned land farmers. This finding is in accordance with that in Central Java by Fukui et al. (2002) that the production efficiency under tenancy land is not lower than that of the owned land. Off-farm occupation does not have an effect on the cost inefficiency of rice farming. The estimated coefficients of farm size are significantly negative in West Java, East Java and South Sulawesi. These show that the larger farm size will increase the efficiency of rice farming. The estimated coefficient of plot number in Central Java is significantly positive. This suggests the dispersion of small plot is one of the determinants of cost inefficiency.

3.  Measurement of DRC and estimation of impacts of technical efficiency on DRCs

This section aims at estimating the relationship between cost efficiency and global competitiveness. Therefore, we use DRC as an indicator of global competitiveness, as suggested by Sadoulet and de Janvry (1995). The DRC is the ratio of the cost in domestic resources and nontraded inputs (valued at their shadow prices) of producing the commodity domestically to the net foreign exchange earned or saved by producing the good domestically. DRC measures whether a commodity is more profitable when produced domestically or imported. DRC < 1 indicates that the commodity is more profitable when produced domestically; meanwhile, DRC > 1 indicates that it is less profitable to produce it domestically.

In order to estimate the relationship between the competitiveness of rice and the cost efficiency of rice farming, this study applies the linear regression analysis with the simple regression model as follow:

  

$$lnDR{C}_i=a+b_{1}ln\left(C{E}_i\right)+b_{2}ln\left(CP{I}_t\right)+b_{3}ln\left(E{R}_t\right)+b_{4}ln\left(F{S}_i\right)+e$$（3）

where

DRC_i = Domestic resource cost of farmer i

CE_i = Cost efficiency of farmer i

CPI_t = Consumer price index in year t

ER_t = The exchange rate of the USD to IDR (Indonesian Rupiah) in year t

FS_i = Farm size of farmer i (ha)

e = Error term

The data description and estimation result of the regression analysis between the competitiveness of rice and cost efficiency in rice farming are illustrated by Table 5 and 6.

Table 5.  Data Description of Indonesian Rice Competitiveness and Cost Efficiency of Rice Farming

Variable	North Sumatra				West Java				Central Java
	Mean	SD	Min	Max	Mean	SD	Min	Max	Mean	SD	Min	Max
DRC	0.811	0.203	–8.71	9.5	0.647	0.12	–2.07	5.2	1.635	0.88	–8.03	13.3
CE	0.485	0.294	0.092	0.92	0.588	0.30	0.039	0.99	0.656	0.39	0.04	0.97
Consumer Price Index	135.28	14.82	121	150.5	134.0	12.6	121	150.5	133.4	13.13	121	150.5
Exchange Rate (IDR)	9109	25.55	9085	9136	9196	263	8779	9678	9136	221.5	8779	9678
Farm size (ha)	0.396	0.278	0.08	1.4	0.572	0.47	0.042	3	0.344	0.313	0.012	3
Number of Obs.	215				439				622
Variable	East Java				South Sulawesi
	Mean	SD	Min	Max	Mean	SD	Min	Max
DRC	1.988	0.93	–7.22	11.4	0.823	0.3	–1.9	9.2
CE	0.614	0.17	0.05	0.94	0.438	0.2	0.11	0.87
Consumer Price Index	133	13.2	115.1	150.5	131	14	115.1	150.5
Exchange Rate (IDR)	9206	323	8779	10399	9312	460	8779	10399
Farm size (ha)	0.349	0.27	0.02	1.54	0.43	0.26	0.025	2
Number of Obs.	586				295

1) Farm size is endogenous variable

2) Endogeneity is untested due to data unavailability

Table 6.  Estimation Result of Indonesian Rice Competitiveness and Cost Efficiency of Rice Farming

Variable	North Sumatra		West Java		Central Java
	Coefficient	SE	Coefficient	SE	Coefficient	SE
lnCE	−0.6359***	0.0798	−1.0489***	0.07352	−0.5517***	0.0500
lnConsumer Price Index	−0.3154	0.3059	1.0293***	0.3745	−1.0947***	0.3035
lnExchange Rate	(omitted)		0.5509	1.3099	2.3074**	1.2435
lnFarm size	−0.0847*	0.0491	−0.1732***	0.0420	−0.137675***	0.0403
Constant	0.6094	1.5024	−10.8452	11.7707	−15.6576	11.1744
Number of Obs.	215		439		622
F Calculated	21.79***		64.01***		40.74***
R2	0.2365		0.3711		0.2089
Average DRC	0.8110		0.6468		1.6347
Average CE	0.4846		0.5881		0.6557
Variable	East Java		South Sulawesi
	Coefficient	SE	Coefficient	SE
lnCE	−0.6949***	0.0393	−0.9473***	0.0550
lnConsumer Price Index	0.1397	0.2534	0.5229**	0.2329
lnExchange Rate	2.6626***	0.7353	0.7038	0.5287
lnFarm size	−0.2175**	0.0318	−0.1582***	0.0373
Constant	−25.6627	6.9622	−9.7892*	5.4178
Number of Obs.	586		295
F Calculated	95.59***		79.68***
R2	0.3969		0.5236
Average DRC	1.9884		0.8228
Average CE	0.6142		0.4382

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

The cost efficiency is negatively significant in influencing the Domestic Resource Cost (DRC) in all regions at 1% level. The cost efficiency is negatively significant in influencing the Domestic Resource Cost (DRC) in all regions at 1% level. If the cost efficiency increases 10%, DRC will decrease by 6.35% in North Sumatra, 10.5% in West Java, 5.52% in Central Java, 6.95% in East Java, and 9.47% in South Sulawesi.

The price index is significantly affecting DRC in West Java, Central Java, and South Sulawesi. Price Index in Central Java is negatively affected the DRC. We do not have enough information to explain the estimation results. However, one possible interpretation is that the former effect is smaller than the latter effect in Central Java, because the DRC is greater than one there, while it is smaller than one in West Java and South Sulawesi where.

The variable of exchange rate is significantly positive to DRC in Central Java and East Java. Further, as the proxy of economy scale, the farm size is negatively significant to DRC in all regions, excluding West Java.

To improve competitiveness, the farm household should enlarge their farm size. However, the rice growing farmers’ behavior in rural Central Java and East Java show that usually not willing to sell their land (Iwamoto and Hartono, 2009). Farm size is not a new policy subject for Indonesia. The Basic Agrarian Law imposed a limit in the size of agricultural land holdings, with a minimum and maximum size of 2 and 20 ha, respectively. In general, the number of farmers producing food crops in Indonesia since then has increased while the size of land holding is small, particularly in Java. Enlargement of farm size without radical institutional changes is only possible on the outer Java island where the land is not as fertile as that of Java. Farm size in Java was much smaller (0.66 ha) than in the outer islands (>1.22 ha). After the colonization and transmigration program, farmers in Java continue the agricultural cultivation intensively. At the same time, the growth of the industrial sector and urban centers in Java has increased the demand for land. The accelerated conversion of paddy fields to non-agricultural uses has far reaching implications for the future performance of the agricultural sector in Indonesia This situation implies that the transaction cost of any land in Java is very costly (Anwar and Pakpahan, 1990).

DRCs of rice farming in North Sumatra, West Java, and South Sulawesi are less than 1. This finding is consistent with Mantau et al. (2014), stating that rice commodities in Sulawesi have comparative and competitive advantage. In contrast, the DRCs of rice farming in Central Java and East Java are larger than 1. This indicates that rice farming in these regions do not have a comparative advantage. In addition, even if the cost efficiency of rice farming in Central Java and East Java increase to 1, rice sectors in these areas will not obtain global competitiveness.

4.  Conclusion and recommendation

This paper aims to examine the effects of production efficiency improvement on the global competitiveness of rice sector in major rice growing areas of Indonesia. The main findings are mentioned below:

First, the production cost is positively influenced by the input prices, particularly the impacts of land rent, fertilizer price, and tractors cost on the production cost are the largest. This implies that the reduction of land rent, fertilizer price, and tractor cost are more effective in decreasing the production cost of rice farmers.

Second, the cost efficiency is the higher in Central Java and East Java of which the farm size is relatively small. The ownership of paddy land has a negative effect on cost efficiency in North Sumatra, West Java, Central Java, and South Sulawesi where share tenancy system is widely observed, but it does not have a significant effect on the cost efficiency in East Java. This is not consistent with that the results by Jamal and Dewi (2009) but consistent with those by Fukui et al. (2002). Moreover, this finding implies that land tenancy does not have negative effects on cost efficiency. The modern irrigation system has a positive effect on cost efficiency in West Java, Central Java and South Sulawesi but not in the other provinces.

Third, the rice farming in Central Java and East Java, which are the main rice producing areas do not have a comparative and competitive advantage.

Fourth, the competitiveness of rice is positively influenced by the cost efficiency of rice farming. If the cost efficiency increases by 10%, DRC will decrease by 6.35% (North Sumatra), 10.5% (West Java), 5.52% (Central Java), 6.95% (East Java), and 9.47% (South Sulawesi). However, even if rice farming in Central Java and East Java can achieve the best cost efficiency, rice sectors in these areas cannot obtain global competitiveness.

Fifth, the farm size has a positive effect on competitiveness.

From these findings, we can draw some policy implications. In order to achieve food security, which is one of the key strategic policies under the pressure of trade liberalization, Indonesian government will have to reform the current protective rice policies. The results of investigation in this paper imply that if it liberalizes the rice policy, rice farmers in the major rice growing areas, such as Central Java and East Java will have the largest negative effects because they do not have global competitiveness due to petty farm size. And the policies for the improvement of cost efficiency are not sufficient for them to achieve global competitiveness. For the purpose of achieving global competitiveness in Central Java and East Java, government needs to implement some policy mix, which aims not only to decrease the cost inefficiency but to facilitate rice farmers to enlarge the farm size. To improve the cost efficiency, improving the irrigation conditions and land consolidation are effective. To enlarge the farm size, facilitating transaction in land rental market will be more feasible than promoting land transaction through buying and selling in rural Central Java and East Java because rice growing farmers are usually not willing to sell their land (Iwamoto and Hartono, 2009).

However, the construction of an irrigation system and land consolidation is expensive, and cost performance must be examined. High transaction cost in the land rental market of rural Central Java and East Java where large surplus labor still exists, there may be obstacles to facilitate renting land.

Thus, it is not so easy for Indonesian government to devise effective policy measures for reducing the cost inefficiency and improving global competitiveness in the rice sector. It may have to do it to overcome this difficulty.

Acknowledgements

I sincerely thank Professor Seiichi Fukui for his comments, advice and research fund. The author is a grantee of DGHE Scholarship.

Note

1  PATANAS (National Farmer’s Panel) data is household survey, which is conducted by the Indonesian Government as a serial survey (time-series).

References

•  Aigner,  D.,  Lovell,  C. A. K. and  Schmidt,  P. (1977) Formulation and estimation of stochastic frontier production function models. Journal of Econometrics, 6, 21–37.
•  Anwar,  A. and  Pakpahan,  A. (1990) The problem of sawah–land conversion to non-agricultural uses in Indonesia. Indonesian Journal of Tropical Agriculture, 1(2), 101–108.
• Brazdik, F. (2006) Non-parametric analysis of technical efficiency: factors affecting efficiency of West Java rice farms (Working Paper Series 286). Prague: Charles University, Center for Economic Research and Graduate Education (CERGE) and Economics Institute (EI), Academy of Sciences of the Czech.
• Baum, C. F. (2006) An Introduction to Modern Econometrics Using Stata. Texas: Stata Press.
• Coelli, T. J., Rao, D. S. P., O’Donnel, C. J. and Battese, G. E. (2005) An Introduction to Efficiency and Productivity Analysis. New York: Springer.
• Debertin, D. L. (1986) Agricultural Production Economics. New York: MacMillan Publishing Company.
• Effendy,  H.  N.,  Setiawan,  B. and  Muhaimin,  A. W. (2013) Characteristic of farmers and technical ffficiency in cocoa farming at Sigi Regency–Indonesia with approach stochastic frontier production function. Journal of Economics and Sustainable Development, 4(14), 154–160.
•  Fukui,  S.,  Hartono,  S. and  Iwamoto,  N. (2002) Risk and rice farming intensification in Rural Java. The Japanese Journal of Rural Economics, 4, 32–43.
• Greene, W. (2005) Efficiency of Public Spending in Developing Countries: A Stochastic Frontier Approach. New York: Stern School of Business, New York University.
•  Hadi,  P. U. and  Wiryono,  B. (2005) Dampak kebijakan proteksi terhadap ekonomi beras di Indonesia (Impact of protection policies on rice economics in Indonesia). Jurnal Agro Ekonomi (Journal of Agro Economics), 23(2), 159–175.
•  Hidayah,  I. and  Susanto,  A. N. (2013a) 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.
•  Hidayah,  I.,  Hanani,  N.,  Anindita,  R. and  Setiawan,  B. (2013b) 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.
• Irianta, B. (2005) Analysis of the Impact of the Rice Import Tariff Policy on the Competitiveness and Profitability of Rice Farming Rice in Central Java, East Java and South Sulawesi Period 2002–2003. Thesis. Jakarta: Faculty of Economics, University of Indonesia.
• Iwamoto, N. and Hartono, S. (2009) Economic Structure and Social Institutions in Rural Java: A Case Study in Yogyakarta. Yogyakarta: Gadjah Mada University Press.
•  Jamal,  E. and  Dewi,  Y. A. (2009) Technical efficiency of land tenure contracts in West Java Province, Indonesia. Asian Journal of Agriculture and Development, 6(2), 21–33.
• Kumbhakar, S. C. and Lovell, Knox, C. A. (2000) Stochastic Frontier Analysis. Cambridge: Cambridge University Press.
•  Kusnadi,  N.,  Tinaprilla,  N. and  Purwoto,  A. (2011) Analisis efisiensi usaha tani padi di beberapa sentra produksi padi di Indonesia (Rice farming efficiency analysis in some rice producing areas in Indonesia). Jurnal Agro Ekonomi, 29(1), 25–48.
•  Makki,  M. F.,  Ferrianta,  Y., Rifiana and Suslinawati (2013) Impacts of climate change on productivity and efficiency paddy farms: empirical evidence on Tidal Swamp Land South Kalimantan Province Indonesia. Journal of Economics and Sustainable Development, 3(14), 66–72.
•  Mantau,  K.,  Harianto and Nuriantono,  N. (2014) Analysis of competitiveness of lowland rice farming in Indonesia: case study of Bolaang Mongondow District, North Sulawesi Province. Journal of Economics and International Finance, 6(4), 85–90.
•  McCulloch,  N. (2008) Rice prices and poverty in Indonesia. Bulletin of Indonesia Economic Studies, 44(1), 45–63.
•  McCulloch,  N. and  Timmer,  P. (2008) Rice policy in Indonesia: a special issue. Bulletin of Indonesia Economic Studies, 44(1), 33–44.
•  Mulwanyi,  A.,  Hutagaol,  P. and  Sinaga,  B. M. (2011) Impact of rice purchasing policy on welfare of both producers and consumers in Indonesia. Journal of International Society for Southeast Asian Agricultural Sciences, 17, 48–57.
• Natawidjaja, R. S. and Rum, I. A. (2012) Food security a situation and policy in Indonesia. (https://www.adelaide.edu.au/global-food/documents/food-security-situation-policy-indonesia.pdf.) [Accessed November 16, 2013].
• OECD (2012) OECD Review of Agricultural Policies: Indonesia 2012. OECD Publishing. (http://dx.doi.org/10.1787/9789264179011-en) [Accessed November 16, 2013].
• OECD (2013) Agricultural Policy Monitoring and Evaluation 2013. OECD Countries and Emerging Economies. OECD Publishing. (http://dx.doi.org/10.1787/agr_pol-2013-en) [Accessed November 16, 2013].
• Paasch, A., Garbers, F. and Hirsch, T. (2007) Trade policies and hunger: the impact of trade liberalization on the right to food of rice farming communities in Ghana, Honduras and Indonesia. Geneva: FIAN (FoodFirst Information and Action Network).
• Romdhon, M. M. and Cahyadinata (2004) The competitiveness and efficiency of rice farming system in North Bengkulu District, Bengkulu Province. Bengkulu: Faculty of Agriculture, University of Bengkulu.
• Sadoulet, E. and de Janvry, A. (1995) Quantitative Development Policy Analysis. Baltimore: The Johns Hopkins University Press.
• Timmer, C. P. (2004) Food security in Indonesia: current challenges and the long-run outlook (Working Paper 48). Center for Global Development.
•  Warr,  P. (2005) Food policy and poverty in Indonesia: a general equilibrium analysis. Australian Journal of Agricultural and Resource Economics, 49(4), 429–451.
• Warr, P. (2011) Food security vs food self-sufficiency: the Indonesia 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.
• Wikstrom, D. and Fleyeh (2013) Modified fixed effects estimation of technical inefficiency (Working Papers in transports, tourism, information technology and Microdata Analysis). Borlange, Sweden: Dalarna University.
• World Bank (2014) Data World Bank. (http://data.worldbank.org/indicator/NY.GDP.PCA.CD/countries/ID?display=graph) [Accessed November 6, 2014].
•  Yonekura,  H. (2014) Indonesia no saikin no seisaku doko: syokuryo anpo to boeki jiyuka (Recent trend of agriculture and food policies in Indonesia: Food security and trade liberalization). Nogyo-to Keizai, 80(2), 87–97.