4. Rice Production Systems

Paul Heytens

Rice is Indonesia's most important food crop. Harvested area, just under io million hectares per year, is roughly three times the area devoted to corn and eight times that planted to soybeans. Production of rice is almost entirely a smallholder activity and provides income and a staple food for perhaps 20 million households (almost ioo million individuals), which rarely operate more than two hectares of land and often less than one.

Rice is grown in a diversity of agroclimatic zones in Indonesia. In any particular region, individual farmers cultivate and manage rice and competing crops on land representative of several different cropping environments. For this reason, changes in price signals cannot be expected to have the same impact on all farmers in any particular region. For purposes of analyzing alternative rice policies, the wide diversity of Indonesia's rice environments must be averaged into a manageable number of representative rice production systems.

Wetland Rice Production Systems

Farmers were surveyed in five major rice-producing areas to provide a basis for characterizing the important production systems. The five areas (see Figure 4. i)-Kediri in East Java, Klaten in Central Java, Majalengka in West Java, Agam in West Sumatra, and Pinrang-Sidrap in South Sulawesi-are agroclimatically diverse and important rice producers. The survey approach used at each fieldsite consisted of a series of general exploratory interviews conducted in selected villages. On average, about twenty-five villages and seventy-five farmers from a representative sample of agroclimatic zones were visited per site.1 Tremendous diversity in climate, relative factor endowments, and available infrastructure was

observed. But the survey results can be generalized into the four representative systems found in Table _{4. i.}

The major land type used for rice production in Indonesia is _sawah or floodable, bunded cropland. _Sawah can be divided into irrigated and rainfed typologies. Controlled and timely applications of water, along with effective drainage, are prerequisites for achieving high yields with HYV seeds. Adequacy of water supply and control over its use, therefore, distinguish wetland rice production systems. Three major irrigated systems can be defined largely according to differences in control over water supplies. Water conditions on rainfed _sawah, although very different from those on irrigated land, are fairly similar throughout Indonesia so only one rainfed system is formally classified. Differences in water conditions, in turn, result in different uses of inputs, cropping intensities, and levels of productivity.

The irrigated and rainfed _sawah systems illustrated in Table _4.1 accounted for ₇₈ percent of cultivated area and go percent of total rice production in ₁₉₈₇ (the corresponding figures were ₉₄ and ₉₇ percent, respectively, for Java). These systems are found predominantly in lowland areas (those below _8oo meters). Rice production in Indonesia also occurs in dryland, swamp, and tidal/swamp environments. These environments have not been an important focus of government intensification programs to date and often are found in remote areas poorly serviced by transportation and marketing infrastructure. Consequently, production is characterized by tiny percentages of marketed output, low yields, and modest or zero levels of modern inputs. Table _4.2 provides a rough estimate of the shares of rice area and production contributed by each of the main rice systems.

Each of the systems is well represented throughout Indonesia. Even in seemingly homogeneous regions serviced by large-scale technical irrigation systems, such as the northern coastal area of West Java, variations in the level and timing of water deliveries, drainage, and system management result in the presence of several systems.

The economic and agronomic profiles of the four _sawah systems, described below, focus on the distinguishing features of crop production. There is considerable diversity even within systems, and descriptions depict only the most typical conditions. The main purpose of the summaries is to highlight the important influences on crop substitution and productivity in wetland rice production. In the rice-producing areas of Indonesia, farmers make basic agronomic and economic decisions for two or three crop seasons per year. As will be shown, the profitability of growing rice and competing crops varies widely among the production systems and seasons.

Budgets from the survey sites for rice and nonrice crops add a quantitative dimension to the analysis. In the budgets, input and output prices are allowed to vary by site and season to reflect local market prices. The appendix at the end of the chapter contains a summary of the values used in computing the budgets as well as a brief explanation of survey methodology.

Good-Control, Irrigated Sawah

The system with the best water control is also Indonesia's most productive. Good-control _sawah accounts for i 1 percent of cultivated area in rice and ₁₇ percent of total annual production. Examples of this system are concentrated in the rich, volcanic soil areas of lowland Java. Off-Java examples are found in the well-irrigated areas of South Sulawesi, Sumatra, and Bali, which together account for less than 20 percent of the total high-productivity area.

Good-control _sawah is characterized by at least two rice crops per year, often three, and technical or semitechnical irrigation with a reliable water source. 2₂ Local rainfall has little impact on cropping patterns, except when extreme drought affects reservoir levels and river flows. Further, because of the presence of technical and semitechnical irrigation with permanent controls up to secondary and tertiary canals, drainage problems rarely limit the production of nonrice _(palawija) crops.

Yields on good-control _sawah average over 6 tons of dry, unhusked rice _(gabah) per hectare. Land productivity often is highest in the dry seasons because ample irrigation water enables the rice plant to receive full benefit from the more favorable solar conditions prevailing in dryer periods. Drought stress, therefore, is rarely a problem in the dry seasons. Flooding and lodging can affect yields in the wet season when severe weather occurs.

In this stable and productive crop environment, farmers are able to exploit Green Revolution technologies. Adoption of improved, national variety seeds (e.g., Cisadane) or IR varieties (developed initially at the International Rice Research Institute) is almost universal. Fertilizer use is very high, ranging between 4oo and 6oo kilograms of urea and triple superphosphate (TSP) per hectare, and ammonium sulfate (ZA) and potassium chloride (KCI) are frequently applied. Both liquid and granular pesticides are used when conditions dictate. In areas where labor costs are high, such as in certain well-irrigated parts of West Java, West Sumatra, and South Sulawesi, mechanization of land preparation using tractors has occurred.

Farmers prefer to plant rice whenever possible (except to break pest cycles or to restore land fertility by fallowing). The figures in Table 4.3 indicate that rice is very profitable on good-control _sawah, and cash returns ranged from RP 700,000 to ₁.3 million per hectare in 1987, well in excess of prevailing land rents.

Table 4.4 shows that corn and soybeans, rice's major competitors in Java, are much less profitable than rice. This is particularly true in the dry seasons, when returns to rice are highest because of seasonally higher output prices and higher yields. Farmers in good-control areas on Java plant _palawija crops only for rotational reasons or when rice is not possible because of insufficient water. The crop substitution pattern in off-Java areas is similar, except that _palawija crops are much lower yielding and thus much less profitable than on Java. Farmers surveyed in Pinrang-Sidrap, South Sulawesi, and Again, West Sumatra, for example, often left _sawah fallow rather than plant _palawija crops when water was insufficient for rice.

In much of Java, villages with good-control _sawah are often targeted for either the government's sugarcane (TRI) or tobacco intensification programs (but rarely for both at the same time). If a village is chosen for either program, typically 33 percent (sugarcane) to ₅₀ percent (tobacco) of total _sawah must be set aside for a period ranging from two years for tobacco to several years for sugarcane. Among survey farmers, participation in either program is regarded as a form of forced cultivation.

Me level of rice profits seems excessive in comparison with prevailing land rents. During the survey, it appeared that the land market was in disequilibrium at most fieldsites, and survey respondents often had difficulty providing a current land rental rate when asked. Sawah land rents have been increasing in most survey villages but at the time of the survey had not caught up with the increases in rice profits, which have risen rapidly and steadily throughout the 1980s (see Chapter 6 for an estimate of the progression of rice profits since 1969).

Farmers designated to participate in the tobacco program in _Kabupaten Klaten, Central Java, for example, rent their land to a tobacco firm and do not cultivate or pay production costs. In 1987, the government offered farmers RP 7₈5 per kilogram of tobacco harvested from their land. With an average yield of 1.2 tons per hectare, gross cash returns to the farmer were Rp 942,000 per hectare. The opportunity cost of tobacco (grown from June until December) is one or two dry-season rice crops. Cash returns from only one dry-season rice crop in 1987 were about RP 400,000 per hectare higher than those from tobacco. Tobacco farmers have time to engage in other activities, but even if farmers were able to find agricultural work in lieu of cultivating rice, earned income from such activities over one rice season is unlikely to exceed RP 300,000 in Klaten. Off-farm work thus does not compensate fully the opportunity cost of not cultivating rice.

Participants in the sugarcane program pay production costs, but all management and cultivation is handled by a group leader, who receives a percentage of gross revenues. As Table 4.5 indicates, cash returns to sugarcane cultivation on _sawah in the three Java fieldsites ranged from Rp 522,ooo per hectare in Majalengka, West Java to Rp i million per hectare in Kediri, East Java. The opportunity cost of sugarcane on _{sawah is} three to five crops (at least two of which can be rice), depending on the farmers' ability to squeeze in rice crops before and after sugarcane cultivation. In a scenario of three lost rice crops, sugarcane harvested in August 1987 excluded a third-season rice crop in 1986 and wet- and second-season rice crops in 1987. Cash returns in the three Java fieldsites for these rice crops ranged from RP 2 to 3.5 million per hectare, a much higher return than that obtained from sugarcane cultivation. Even if farmers could find other

employment as agricultural laborers, rice farming remains by far the more lucrative alternative.

_{Moderate-Control, Irrigated} Sawah

Moderate-control _{Sawah is} also concentrated in Java, although not so heavily as good-control _Sawah. This second _Sawah system accounts for zi percent of total cultivated rice area and 27 percent of Indonesian rice production. The features of this system are similar to those of good-control _Sawah, but yields are lower and rice-cropping intensity is not so high because soil and water conditions are less favorable. Yields range between 4.5 and 6 tons of unhusked rice per hectare and typically fall in the dry seasons because of moderate drought stress and pests. Two rice crops per year are almost always feasible, and often, particularly on Java, a third, nonrice crop is grown. Three rice crops generally are not possible because of a lack of water in the third season. Drainage problems prevent cultivation of _palawija crops during the wet season.

The use of improved seeds is almost universal, and fertilizer application is high, ranging between 350 and 5oo kilograms total per hectare. Pesticide use is also high. Like its good-control counterpart, moderate-control _{Sawah is} located mainly in lowland areas, and mechanization of land preparation is common in regions where wages are high or land preparation needs to be done quickly.

Survey results revealed that rice grown on moderate-control _{Sawah is} also very profitable. Cash returns in the survey sites ranged from Rp 500,000 to 850,000 per hectare. Returns for rice were slightly higher in the dry season because of higher output prices and are much higher than those for competing crops. This is true even in Kediri, arguably the most profitable _palawija production area in Indonesia. A comparison of Tables 4.6 and 4.4 shows that rice maintained a comfortable profitability margin over both corn and soybeans in that East Java _kabupaten. Yields or prices of hybrid corn, the closest competitor, would have to rise roughly 2o percent to match the net profitability of dry-season rice in Kediri. Comparison of Tables 4.5 and 4.6 indicates that rice grown in moderate-control _Sawah in Kediri or Klaten is much more profitable for farmers than is sugarcane.

In Pinrang-Sidrap and Again, experience with _palawija crops is more limited. But there, too, rice is considerably more profitable than the nonrice alternatives. Many farmers in South Sulawesi stated that, because of low returns, they would rather leave their land fallow than grow _palawija. Nonrice crops were more in evidence in Kabupaten Again, West Sumatra, but very low yields caused net soybean returns there to be negative. Because nonrice crops are less profitable in all areas, they tend to be grown in the dry seasons only when water is insufficient for rice.

Poor-Control, Irrigated Sawah

Poor-control sawah is Indonesia's largest and most diverse rice system, accounting for.26 percent of cultivated rice area and 2'7 percent of production. This system is about equally divided between Java and the outer islands. Unlike the systems where water availability and delivery is highly controlled, poor-control, irrigated sawah is found at elevations above 8oo meters as well as in lowland areas. This system includes much of the picturesque terraced hillside rice paddies for which Indonesia is famous.

Poor-control sawah is characterized by one or two rice crops and one palawija crop per year. Yields are between 3.5 and 4.5 tons of unhusked rice per hectare, and irrigation water typically is provided by either a semitechnical or a simple irrigation system. Poor water control can lead to drought stress in the dry seasons and flooding in the wet season.

Use of modern inputs, such as HYV seeds and chemical fertilizers, is less common and less intensive than in the higher-productivity systems. In most cases, only urea and TSP are used, and levels rarely exceed 350 kilograms of total fertilizer per hectare. In some off-Java areas, including Agam, long-duration, traditional seed varieties are still planted, especially at higher elevations. These varieties mature in about five months, so farmers cannot grow two rice crops in a year. Farmers in Agam are aware of improved, shorter-duration seeds and in many cases have tried them, but prefer traditional varieties that offer the same or higher yields, better-tasting rice, natural pest resistance, and a moderate price premium over improved varieties.

The field surveys revealed that rice was reasonably profitable in 1987 (Table 4.7) on sawah where water is poorly controlled. Because profits were several hundred thousand rupiah lower than in the higher-productivity systems, palawija crops were more competitive. Nevertheless, rice remains more profitable on average (Tables 4.7 and 4.8), and nonrice crops typically are grown on poor-control sawah only when rice cultivation is not possible or too risky.

Drainage problems prevent cultivation of nonrice crops during the wet season in most of Indonesia's semitechnical and simple irrigation systems. This technical constraint alone eliminates between 5o and 75 percent of the potential crop substitution that could take place on poor-control sawah. To be able to plant crops such as corn or soybeans in the wet season, farmers would have to make substantial investments in removing bunds and raising land surfaces above the surrounding land to facilitate drainage (through this process, of course, the land would cease to be sawah). Farmers, understandably, are unwilling to undertake such an investment for potential returns lower than or equal to those from rice.

Sample farmers also perceived rice to be less risky than other crops because of the BULOG floor price for rice and the volatility of palawija

yields. Soybeans, in particular, suffer greatly from sporadic pest problems under tropical conditions. Various vegetable crops (e.g., peppers and green beans) provide much higher average profits than rice does, but successive vegetable crops cannot be grown without great risk of severe pest attacks. Given the short duration of most vegetable crops (one to two months), direct competition with rice can be avoided. Finally, survey farmers strongly preferred producing rice for home consumption rather than buying in the market. These additional points, along with higher average profitability, firmly entrench rice as the crop of choice on poor-control _sawah.

Rainfed Sawah

Rice area on rainfed _sawah accounts for zo percent of rice plantings and 18 percent of annual rice production in Indonesia. Rainfed rice is slightly more common in the outer islands but is also prevalent in Java. Like poor-control, irrigated _sawah, rainfed _sawah tends to be located in more remote areas that have less supporting infrastructure such as paved roads and agricultural extension.

From an agronomic perspective, the major distinguishing feature of rainfed _{sawah is} its dependence on rainfall for water. Since cultivation with HYV seeds is not successful without sufficient rainfall to flood the land for at least two months, rice production on rainfed lands is considerably riskier than that in any of the irrigated systems. Farmers must wait until the heavy monsoonal rains have started or risk drought stress at the crucial early stages of plant growth. But if they wait too long, they risk drought stress late in the crop cycle from an early cessation of the rains.

The lack of water control often reduces yields because of drought stress. Yields rarely exceed 4.5 tons of unhusked rice per hectare on rainfed land and are often much lower, particularly outside Java. Yields are reduced by lower fertilizer levels; applications generally are less than 3oo kilograms total per hectare. Pesticides tend to be used when necessary (and often when not) on all wetland rice, including rainfed rice. Improved seed varieties are also planted by most farmers in Java's rainfed areas, though less commonly so off Java.

Since the rainy season lasts on average about five months throughout much of Indonesia, only one rice crop is grown on about go percent of rainfed _sawah. Two rice crops are possible in areas with very long wet seasons (mostly located in Sumatra). Two rice crops also are possible in certain areas if the first rice crop is dry seeded with very short-maturing seed varieties (e.g., IR 36) just before the rains start _{(padi dodo rancah).} The spread of dry-seeded _sawah rice has been limited by low yields, high risk, and inappropriate soils, although the technique has been successfully introduced into several areas of Java. Scope for increasing _{padi gogo rancah} area in the future is limited, and such area has recently fallen since the practice was started in the early ig8os.

Table 4.9 indicates that rice cultivation is profitable on rainfed _sawah. In fact, profits are higher on average (but more variable) than those for the low-productivity, irrigated system. With the exception of dry-seeded rice, crop substitution is a moot issue on rainfed _sawah. Cultivation of nonrice crops is not feasible in the wet season because of drainage problems. Rice cannot be grown in the dry season because of a lack of water. Survey farmers in Klaten, Central Java, who grow dry-seeded rice, substitute

freely between rice and soybeans before the wet season, according to expected relative profitabilities, but this substitution represents a very small proportion of cultivated area and an even smaller proportion of total production in Klaten.

Conclusion

Indonesia's rice production environment is diverse. This chapter has characterized this diversity in terms of agronomic and economic differences. The aggregate response of wetland rice farmers to possible price and regulatory policy changes is examined in Chapter 7. For government policies that have a country-wide impact, such as the fertilizer subsidies and the floor price for rice, the diversity of rice systems causes farmers to have varying responses, even when they live in the same geographical region. The high degree of commercialization of rice production tends to make responses more homogeneous because large numbers of farmers are responding directly to market forces. But even in fully commercial rice operations, the diversity of soil types, water availability, farmers' skills, and alternative potential crop choices can lead to differing responses to uniform policy changes.

The budget-based analysis provided in this chapter indicates that rice production dominates in Indonesia's wetland arfeas. Rice offers much higher financial returns than do competing crops in all wetland rice systems. In addition, technical constraints limit the production of nonrice crops in the wet season in many environments. Because limited substitution possibilities are available on already existing rice land, most future growth in cultivated rice area will have to come from investments in irrigation infrastructure, new technologies made available from research, and changes in regulatory policy. This key result, which reflects both technical and economic constraints in rice farming, has important implications for the selection of policies to encourage increased rice production.

1

With the exception of sugarcane, the crop budgets are not based on statistical averages of sample farmers. Instead, the Stanford team created synthetic budgets to be representative of the particular crop and system in each of the fieldsites in the farm survey. The use of synthetic budgets reflects the Stanford Project's objective of gathering a broad spectrum of information concerning rice production and labor market conditions in major producing areas during a relatively short period of time.

2

Technical irrigation provides maximum control over water flows by using permanent canals, control structures, and measuring devices. The government typically controls water distribution up to the tertiary canals. Semitechnical systems have permanent canals but few controls or measuring devices, and the government generally controls only the source and the main canal. Simple systems have few permanent control and distribution structures and are usually farmer-managed.