9. Summary and Conclusions

Scott Pearson

This book sets out three alternative targets for Indonesian rice produc-tion: high (for output growth of about 4 percent per year), low (about i percent), and medium (about 2.5 percent). The study documents what policies would be needed to reach these targets and what the likely consequences of those policies would be for incomes and employment in rural Indonesia. Each target represents a rice strategy involving an in-terlinked set of policies with broad effects throughout the economy. The likely effects of each strategy are discussed in light of evidence from new surveys of rice production and the rural labor market.

Under the high-growth strategy, Indonesia would produce more rice than it consumes every year. In most years, this would result in a surplus that would have to be exported with subsidies. with the low-growth target, the country would gradually reduce its dependence on rice pro-duction, diversifying the rural economy into other crops and non-agricultural activities. Under the medium-growth scenario, Indonesia would maintain long-term self-sufficiency on trend, with production equaling consumption on average, and BULOG would use storage, trade, or both to stabilize rice prices from year to year.

The Past

A price policy environment leading to high profits and stable prices has induced substantial increases in rice production through the application of more variable inputs and on-farm investment in land preparation and other productive assets. In addition, government investment in irrigation and other rural infrastructure has allowed farmers to take full advantage of the latest rice varieties, and research and extension ensured that new techniques continued to become available. The global environment for Indonesian rice policy has been highly unstable so that Indonesian depen-dence on international rice trade entailed unusually large risks for the government and for rice producers and consumers. Furthermore, Indo-nesian imports have tended to raise the world price and exports to lower it, especially in the short run. This implies that any policies resulting in

dependence on world rice trade-consistently either exporting or im-porting rice-are risky. To help stabilize domestic prices with lower costs of domestic procurement and storage, however, Indonesia could trade small amounts intermittently.

Detailed new data for current rice production systems in Java, South Sulawesi, and West Sumatra show that rice is overwhelmingly the most profitable crop in these areas. Hence price policy adjustments are un-likely to induce many farmers to switch into or out of rice. A wide range of production techniques now coexist, following the underlying diversity in parcel size, topography, and agronomic conditions. Farmers alter the level and combination of inputs as their conditions change and new techniques become available. Input use, yields, and production levels, therefore, will be affected by the availability of irrigation and other infrastructure, along with appropriate technological packages for each type of farm.

The links among rice production levels, total employment, and wages in rural Indonesia are analyzed with new survey data from three sites on Java and one in South Sulawesi. These data show that changes in the demand for labor directly in rice production are no longer the principal influence on rural wage and employment levels. Instead, the rice sector contributes only a part, albeit an important part, of the total demand for labor. Rural incomes, which determine the demand for transportation, construction, small industry, and other labor-intensive rural activities, combine with urban wages and employment levels to set the demand for and return to unskilled labor in an integrated, competitive labor market. In the ig7os and early ig8os, steady increases in national income raised demand for labor in a wide range of industrial and service activities. These changes helped real rural wage rates to rise in the first half of the ig8os, narrowing the gap between high- and low-wage areas. This in-crease in labor costs, in turn, has led to the gradual shift away from traditional labor-hiring practices to the increasing use of contract labor and payment of wages in cash.

Rising real wages and other influences induced changes in rice tech-nology, rice production levels, and labor use in rice during the 1970s and ig8os. New technologies have raised rice output and farm incomes with-out changing the total quantity of labor used. This result provides further evidence that increases in rural and urban income, not increases in the amounts of labor used for rice production, have been the dominant source of higher employment levels in rural areas. Moreover, the spread of synchronized planting and harvesting to meet irrigation schedules or to allow integrated pest management strongly increase the seasonality of labor demand, thus enhancing the attractiveness of mechanization in many areas. This change, which reduces total employment in rice farm-ing, is already occurring in West Java and South Sulawesi.

The Near Future

Forward-looking simulations, using the evidence given above, can indi-cate what policies would be needed to reach alternative rice production targets, each one defining a separate rice strategy. These simulations quantify the increases in rice area, yields, and total output that can be expected from each policy instrument and show how those instruments fit together to form coherent strategies around each target level of growth.

A substantial burst of rice production, raising annual production by about i percent a year for five years, could be achieved through a gradual phaseout of the TRI sugar program, releasing about 30,ooo hectares of high-productivity land into rice every year for five years. This deregula-tion also would cause a direct increase in employment levels because employment in rice production and processing is higher than in sugar.

Continuing high levels of investments in irrigation and other rural infrastructure are needed to complement increases in rice output from the conversion of sugar lands. Although disaggregated data to show pre-cisely which lands could be upgraded most easily to higher-quality irriga-tion systems are not now available, a number of irrigation investments are under way or planned for REPELITA V. Relatively large increases in rice output can be achieved from either the rehabilitation of existing reser-voirs and canals or the construction of entirely new systems. More invest-ment in irrigation, along with continued investment in research and ex-tension services and improved pest management, together can add significantly to growth of rice production. This investment could bring annual rates of output growth from these sources to over 1 percent during REPELITA V (1989_-93).

To avoid falling below trend increases in consumption, both changes in regulatory policy (to bring sugar lands into rice cultivation) and increases in public investments (to increase the yields and cropping intensity of poorly- irrigated land) will be needed. If these are not undertaken, Indo-nesia would likely face either high-cost dependence on imported rice or increases in domestic rice prices. A 5 percent increase in the real price of rice could add perhaps another half percentage point of annual output growth. These enhanced price incentives almost certainly would impose much higher costs on the government and consumers than would chang-ing the regulatory policy for sugar or increasing public investment in irrigation and research. But additional price incentives appear to be nec-essary to maintain trend self-sufficiency during ₁99_1-g3.

If a phaseout of the TRI sugar program on Java, investments in irriga-tion and other infrastructure, and a 5 percent increase in the rice price in 1gg1 all occur, it should be possible to maintain trend self-sufficiency during the REPELITA V period. To produce a regular rice surplus, however, the amounts of investment or rice price hikes would need to be much higher and are unlikely to be feasible (given expected levels of government revenues) or desirable (given expected export prices relative to Indonesian policy prices). Consequently, the strategy of trend self-sufficiency, with a target growth rate of 2.5 percent per year, would be the lowest-cost, lowest-risk option.

The equity implications of adopting that strategy are tested by investi-gating the direct and indirect effects of the major policy instruments on rural incomes, employment, and wages. The new survey data on rural labor markets assist analysis of the effects of price policy, deregulation of sugar land, and investments in irrigation on the key parameters of the labor market in the major rice-producing regions. The results demon-strate that price policy has a significant potential to shift rural incomes because it affects the price of rice for all producers (thus changing their profits) and all consumers (thus influencing their real incomes).

Raising the rice price by 5 percent shifts income from urban centers into rural areas that are net producers of rice. But it also reduces the income of urban or rural residents who are net rice consumers, including most of the very poor. This direct effect would worsen income distribu-tion in all urban centers and in some rural regions. The majority of net consumers of rice would experience about a 1 percent decline in purchas-ing power if the rice price were increased by 5 percent. Deregulation of sugar lands and investment in irrigation, however, would have a positive effect on rural income and on income distribution. Both of these policies would have a large impact on employment levels because they would increase the intensities of labor and land use.

The estimated net income in rice production generated by the pro-posed rice strategy during the 1989_-93 planning period is RP 421 billion, about 1o percent of total income earned in rice farming on Java in 1987. Over half of this incremental income derives from irrigation investments, one-third from the phaseout of the TRI sugar policy, and one-eighth from price policy. The estimated net gains in rice employment throughout the five-year projection are 1.1 million labor years-about 15 percent of cur-rent rice employment and 3 to 5 percent of total unskilled jobs. About half of these new jobs would be associated with irrigation investment, one-third with price policy, and one-eighth with sugar policy. This estimate is overstated because it does not account for employment losses from reduc-tion of crops replaced by rice or from substitution of purchased inputs for labor in rice farming.

The negative equity impact on rice consumers would be tempered by two offsetting influences, both affecting the wage rate for unskilled labor. The demand by the rice sector for additional unskilled labor would have a direct upward influence on the wage rate; since rice production accounts for ₂₀ to 3₀ percent of total unskilled employment, this transmission effect could be as much as i or 2 percent of the wage rate. In addition, much of the increased returns to labor and land employed in rice produc-tion would be expended on consumption or investment, creating more indirect employment of unskilled labor in construction, transport, and other rural services. The likely magnitudes of these offsetting influences are not estimable with existing information. But they might compensate for a significant portion of the negative direct effects of the price change because most poor adult rice consumers are also unskilled laborers.

The Principal Empirical Results

The new data and analysis in this book provide clear evidence that the lowest-cost and lowest-risk rice strategy would be to aim for trend self-sufficiency at about 2.5 percent annual growth in production, achieved through a combination of regulatory reforms producing short-term gains, increases in production from higher public investment in irrigation and other infrastructure, and small increases in the real price of rice. Indo-nesia has a strong present comparative advantage in producing rice to prevent imports and the likelihood of maintaining a future comparative advantage in rice production.

Planning for a rice production growth rate of less than 2.5 percent annually is unwise because too much reliance would be placed on rice imports in an unstable international rice market. A low-growth strategy of i percent annually thus is inconsistent with the objective of national food security (and price stability). At the other extreme, a production growth target of 4 percent annually could be extremely expensive and may not be technically feasible. A high-growth target is inconsistent with the objec-tive of efficient resource allocation because the higher domestic rice prices required for its achievement would likely create the need for ex-port subsidies in most years.

Results from the field surveys show that rice cultivation is highly profit-able in both private and social terms. High levels of private profitability mean that policies designed to increase rice production have furthered the objective of rural income growth. Positive levels of social profitability imply that these policies also have been successful in allocating resources efficiently on an economy-wide basis. Relatively high profits in rice indi-cate that farmers are not likely to move out of rice into other food crops in the near term. In fact, technical constraints prohibit them from doing so in the rainy season. High levels of profitability also indicate that the indirect income and employment effects associated with expenditures and investments of rice profits are important for rural economic growth. Although rice production has grown rapidly during the past twenty years, aggregate employment in the rice sector has remained stable. Higher levels of employment stemming from rising yields, greater crop-ping intensities, and expansion of production area have been offset by changes in cultivation techniques (such as the shift from the anti-anti to the sickle for harvesting) and the gradual mechanization of certain tasks (such as land preparation and threshing).

Even though demand for labor in rice production in the aggregate has remained stable, real wages for unskilled labor in the rice sector rose rapidly in the first half of the ig8os. Hence policies designed to promote rice production-in the context of rapid growth of the entire Indonesian economy-generally have been consistent with the government's objec-tive of desirable income distribution. Growth in real wages for rural unskilled workers is a principal indicator of successful economic develop-ment for Indonesia.

Results from the field surveys indicate that the rural labor market is reasonably well integrated on Java. Unskilled workers face a wide range of job opportunities in both the agricultural and nonagricultural sectors of the economy and allocate their time between activities in response to differences in relative wages. Rural economic growth associated with a dynamic rice economy and higher levels of labor demand in rural and urban areas has improved the opportunities available to unskilled workers in the ig8os. Consequently, national income growth, rather than growth in rice production per se, is likely to be the main source of expanded job opportunities for rural unskilled laborers in the future.

Complementarities and Conflicts among Objectives and Policy Instruments

Three basic indicators of national welfare-aggregate income, income distribution, and food security-have been suggested as fundamental ob-jectives of government. These basic objectives often are in conflict with each other. Some instruments, notably price policy, create trade-offs among these objectives; for example, consumers, who buy more rice than they sell, are harmed by decisions to raise rice prices. But in general, the situation of rice in Indonesia has presented an exceptional, and very fortunate, case in which the three objectives of government policy have been served by a strategy of rice self-sufficiency.

This fortunate situation could change soon. An emerging trade-off be-tween the objectives of food security and efficient income growth is likely to be encountered if Indonesia pursues a strategy of rice self-sufficiency on trend. Increases in the rice price, needed to maintain self-sufficiency, could push domestic prices above the long-term c.i.f. trend and lead to

inefficient production. Higher prices also run the risk of causing inequita-ble income distributional effects for rice consumers. But any future trade-off between food security and equity would be lessened to the extent that higher rice prices were transmitted into higher unskilled wages and that increased rice incomes were expended on labor-intensive local goods and services.

Along with consumers' resistance to higher prices, the Indonesian gov-ernment faces other important obstacles in implementing policies to fur-ther its objectives. For example, the least expensive source of short-term increases in rice production is the deregulation of land use. This would result in shifts from sugar to rice production, providing higher profits and faster growth of national income, higher employment levels and better income distribution, and more food security. But it also would reduce the use of capacity and thus raise unit costs in Indonesia's large sugar mills, and it would induce greater sugar imports. Some mills probably would have to be shut, which would result in substantial conflict and con-troversy.

Investments in upgrading the quality of irrigation can have similar effects at the national level-raising aggregate income, creating em-ployment, and improving food security. These investments expanded rapidly with the increasing government income from oil in the 1970s, multiplying the benefits of that income many times over. But the constit-uency for rural investment is relatively diffuse. Despite its value for the nation as a whole, irrigation has fewer vocal advocates than do many other budgetary priorities. Hence irrigation expenditures fell even faster than other government spending when oil income declined in the ig8os, be-fore recovering near the end of the decade.

A critical policy decision is whether to use only domestic procurement and storage to stabilize prices, tying up capital in large food security stocks, or to supplement domestic stocks with frequent small amounts of imports and exports whenever prices seem favorable given local supply-demand balances. It is almost certain that the stocks-only policy (absolute self-sufficiency) would require higher rice prices, larger government sub-sidies, and more risk than the stocks-and-trade policy (trend self-sufficien-cy), but this is an important topic for future research.

The issues for future rice policy are abundantly clear. As this book documents, if Indonesia is to avoid dependence on regular rice imports in the iggos, it will be necessary to make a concerted effort to maintain prices that will provide positive incentives for farmers (while not placing unreasonable demands on consumers to adjust their food consumption patterns), to deregulate land use as much as possible, and to invest effi-ciently in irrigation and other rural infrastructure (and its maintenance) and in research for new rice varieties.

How might policymakers best implement a strategy of trend self-suffi-ciency for rice in Indonesia? Medium-term (five-year) planning calls for rice production gains of about 2.5 percent annually-enough to match the expected growth of rice consumption. During the REPELITA V period (1989_-93), two likely sources of annual rice output growth are irrigation investments (about i percent) and _gabah price policy (about o.5 percent). Because the impact of past public and private investment in irrigation, input distribution facilities, high-yielding seeds, and improved farming practices (the so-called trend effects) is included in the estimates of irriga-tion and rice price effects, a growth rate of 2.5 percent per year could be achieved only by deregulating TRI sugar land. This desirable policy change could add a full percentage point to growth in annual rice output for five years. Moreover, the deregulation of sugar land would further all three food policy objectives because it would result in an efficient growth of income, improve real incomes and employment, and enhance food security.

Short-term (year-to-year) planning for trend self-sufficiency calls for a flexible use of trade policy to balance unexpected domestic shortfalls or surpluses. Public storage of rice stocks in excess of levels needed for buffer stocking or for maintaining the food security stock (about 1 million metric tons) is generally more expensive than exporting, and public in-ventories maintained at levels so high as to avoid any imports even in severe drought years are far costlier than rice imports. With this knowl-edge in hand, Indonesian policymakers have exported surplus stocks at prices far below domestic costs (in 1985-86) and have supplemented domestic supplies with foreign procurement by recalling earlier loans of rice made to the Philippines and Vietnam (in 1988-89). Following an exceedingly good production year in 1989, Indonesia faced the happy prospect of having adequate domestic supplies both to meet all consump-tion needs and to rebuild food reserve stocks to desired levels. In this circumstance, Indonesia should be able to be self-sufficient in rice on average, while exporting unwanted surpluses or importing to fill tempo-rary deficits.

With this strategy of trend self-sufficiency in place, the Indonesian government could achieve increases in rice output that would improve food security by reducing or eliminating regular imports and generate continued growth in rural incomes with only a minor negative impact on the distribution of that income. But if these priorities are eclipsed by others, future food crises may be needed to bring rice production back into prominence.