Managing agricultural transformation : a case study of farm mechanization in Taiwan

This internal working paper is prepared for STAFF USE ONLY. The views expressed are not necessarily those of the World Bank. MANAGING AGRICULTURAL TRANSFORMATION A CASE STUDY OF FARM MECHANIZATION IN TAIWAN AGREP Division Working Paper No. 93 Prepared by: Ai Chin Wee (Consultant) under the direction of Graeme Donovan Economics and Policy Division Agriculture and Rural Development Department January 1982 Table of Contents I. Theoretical Framework . . . . . . * . . . . . . . . 1 II. Agriculture in Transttion . . . . . . ... . . . 7 III. Mechanization and Resource .Use in Agriculture . . 14 IV. The Supporting Packages . . . . . . . . . . . . . . 32 V. Summary and Conclusions . . . . . . . . . 48 Bibliography . . . . . ** ** ** 54 D i s c u s s a n t s . . . . . . . . . .. . . . 5 5 LIST OF TABLES Page 1. Taiwan: Growth Rates of Agricultural Output .in different phases of Agricultural Development, 1946-1974. 9 2. Comparative Growth and Employment: Agriculture vs Industry 9 3. Farm Income vs Non-Farm Income, 1952-1971 10 4. Categories of Rural-Based Labor 10 5. Changes in Proportion of Age Distribution of Agricultural 11 Workers 6. Taiwan: Growth Rates of Inputs in Different Phases of 15 Agricultural Development, 1946-1974 7. Taiwan: Sources of Human, Animal and Mechanical Energy in 19 Crop Production in Horsepower-Day equivalent units and percentages of total, 1952-75 8. Changes in Work Inputs per hectare of Rice, by Farming 20 Operations 9. Changes in Labor input and Machine Use in Taiwan Agriculture 22 10. Usage of Farm Machinery and Draft Animals, 1965-1979 24 11. Changes in Rice Production, 1950-1977 26 12. Ownership of Farm Machines 29 13. Custom Hiring of Farm Machines 29 14. Relative Contributions of Input and Productivity in 31 Agricultural production 15. Categories of Land ownership Before and After Land Reform, 36 1949-1961 16. Changes in Size of Landholdings Before and After Land Reform 36 1952-1960 17. Changes in Net Farm Income, 1950-1960 36 18. Percentage of Adoption of Recommended Farm Practices 39 19. Results of Rice Joint Farming Project in Taiwan, 1975-1978. 40 MANAGING AGRICULTURAL TRANSFORMATION A CASE STUDY OF FARM MECHANIZATION IN TAIWAN I. Theoretical Framework 1.01 The process of agricultural transformation in developing econ- omies may be characterized by a transition from subsistence farming to com- mercialized agriculture. The traditional farming system is typically a family enterprise which produces essentially for its own consumption, according to its needs and preferences and utilizing resources under its immediate command. Within the constraints of available land, labor, cap- ital, and limited access to technology, the triditional farm characteris- tically operates at low levels of capital investment and productivity, generating occasional surpluses to trade in local markets for non-farm goods and services. Within a stable local environment and limited consump- tion and production incentives, stable patteins of resource allocation emerge. These tend to approach the margin of efficiency under traditional technology and are perceptibly dominated by an attitude which tends to be conservative, emphasizing food security rather than untoward risktaking. 1.02 Over time, however, a multiplicity of factors starts to create strain on the delicately fostered balance of the traditional system. In- creased population increases food needs and man-land ratio. Requirements for non-farm goods leads to the growth of a non-agricultural sector. In the typical dual sector economy that emerges, the agricultural sector becomes plagued with the roles of providing food for the increased popula- tion and the urban sector, raw materials for domestic industry, capital to finance industry (through taxation and private transfer), employment at comparably remunerative wages, labor for industry, a domestic market for local fledgling manufactures, and coveted foreign exchange from export earnings and savings from (agricultural) import substitution. -2- 1.03 The transition from traditional to modern agriculture is anything but smooth. or automatic. A whole complex of issues must be tackled and resolved -- these range from the need for incentives to stimulate realloca- tion, research on improved technology to bring appropriate extension messages to the farmers, the facilities to organize the procurement and distribution of crucial new inputs, and the infrastructure to assemble, process, store, market and distribute the resultant agricultural production for its intended usages. 1.04 The process of development is one of continuous technological transition. The aim is to guide that transition in some rational, cost- effective manncr. One could venture into academic debate on whether, for instance, a "transformation approach" involving large-scale planned structural change might be preferred to a less traumatic "improvement approach" favouring diffusion of innovation by demonstration and extension and the injection of improved technology, inputs and credit into traditional farming systems. Whatever the preferred approach, one would invariably end up with a whole checklist of technical, financial, economic, institutional, social and political arrangements that affect and are affected by the process. A proliferation of parties, objectives and pro- grammes are involved, with whole series of iterative and adaptive actions and reactions, concerns and vested interests. On a more macro-economic level are the policies and programmes of Government and researchers, both public and private, and others, who have a mandate or vested interest in directing, experimenting or disseminating alternative improved techno- logies. On the institutional level, there are the efforts of public and private agencies which seek to supply inputs and services, or market the -3- outputs of the system. Then, at the operator level, there are the farmers, managers and farm-operators who try to achieve maximum or optimum returns from their inputs of labor and other production resources by organizing, reallocating and experimenting (to whatever level of risk their circumstances allow) with new production modes and opportunities as they became apparent to them over time. 1.05 While the overall long-term objective of agricultural development may be common to all participants in the process, each group may oftentimes pursue strategies which appear hardly complementary to that of the others. Thus, for instance, the risk-averse subsistence farmer is extremely cau- tious about adopting new cropping techniques-. The tenant farmer, uncer- tain about his tenure, prefers e.ploitative (as against conservationist) - measures which maximize rather than optimize returns. The field extension agent, typically a petty government employee (whose monthly paycheck will not benefit or suffer whatever production techniques are employed) offers disinterested advice which generally lags more than a few years behind latest research breakthroughs. The tractor sales agent employs utmost - persuasion on government bureaucrats to bring up sales (and commissions) nearer to monthly targets, (rather preferring not to hear disquieting claims by researchers that wanton removal of toposoil in the process of forest clearing would set the productive capacity of the soil back by so many years). The loan officer of the agricultural credit agency is generally more concerned about debt servicing arrangements than about whether particular equipment procured on loan carries adequate guarantees and spare parts for prompt servicing if required at short notice. Extensionists exhort, conservationists plead their cause, industrialists advocate the use of machines, politicians urge subsidization, planners pontificate, governments predicate and try to appear up-to-date, while historians and sociologists chart the course and ponder why farmers do what they do. That the process of bringing about technological change can be so multi-faceted comes as no surprise. The immediate decision-making environment, time dimension, degree of vested interest, incentives and intended outcomes are specific to each party and each has its place. 1.06 What changes eventuate? While researchers hypothesize about optimum approaches and various participants play out their roles, changes are already taking place at farm level and beyond, in situations where individuals can see the potential for private gains, or planners for social gains. Successful innovation depends on the extent to which various participants can achieve coherence in their efforts and can fine-tune their recommendations to the needs of the situation. Innovation occurs in stops and starts and farmers take their cue from proven successes of leaders. The peculiar mix of techniques that has the most chance of adoption usually has the following characteristics: (i) Complementarity - the parts of the package are technically com- plementary, supplementing and reinforcing their mutual impacts; (ii) Congruity and Compatibility - they are consistent with the human resource conditions into which they are introduced; (iii) Complexity - they do not cause excessive disturbances or re- orientation of farming operations to permit their introduction; (iv) Divisibility and Economy of Scale - the innovations are within farmers' means to own or use, whether individually or cooperatively; (v) Profitability - the marginal revenues exceed the marginal costs of the innovation to the investor; -5- (vi) Reliability - the usets should be convinced that the package is sufficiently reliable and predictable in the consequences of its usage. 1.07 Besides satisfying the requirements for initial adoption, the sustained acceptance of the package over time will depend on the specific effects on the "receiving environment," namely: (i) the direct effects on agricultural production, employment, income, and immediate linkage effects; and (ii) indirect effects such as pressures on the physical and institu- tional infrastructure of the region, in providing for supply of inputs, and usage and disposal of production; shifts in the distribution of income, consumption, wealth, behavior patterns and social influence among different groups in society. Hence, succeesful innovation has the elements of the provision of a 'suitable' package, the creation -f a suitably 'prepated' receiving envircnm,nt, and the ability to manage these over time, making the neces- sary adjustments in the light of resultant effects and consequence of the innovation. 1.08 These initiatives may come from the farmers, the extension service, the manufacturers, or the government. In essence, what is required is the need to induce innovation - among those who own or control the use of resources, to motivate them to want to change their methods of agricultural production. A blend of coercion and incentives might be employed, with different regimes weighting their approach towards one or the other. The instruments employed may include pricing, subsidy, taxa- tion, user and production incentives, supply of inputs, access to land, -6- credit and other resources. How each innovating environment responds to incentives varies tremendously; no single success can be wholly duplicated for another situation, yet it is instructive to see how the variables are managed in a successful test case. 1.09 In this example, the case is that of Taiwan, Republic of China. The study is of Taiwan's phenomenal success in agricultural mechanization in the last decade. From the 1960's, industrial development caused continding out migration of rural labor which in turn necessitated the adoption of labor - saving and capital - using technology. Farming became increasingly specialized and institutionalized. Land reform and the building of rural and irrigation infrastructure defined ownership rights and delineated more manageable boundaries and parcels of land. The organization of farming into groups was the next step - these groups took the form of formal "Irrigation Associations and Farmers' Associations", through which agricultural extension prograns were implemented; and Joint Farming groups which undertook group farming and cooperative usage of equipcent and facilities by farmers in the same locality. Meanwhile, farm mechanization received a great impetus from a variety of financial and technical measures instituted by the government. Manufacturers were quick to supply the needed inputs, and production of agricultural machines and tools accelerated in response to incentives. The success story of farm mechanization and agricultural transformation in Taiwan describes how, in a particular case, a coherent blend of Government and private initiatives has succeeded in anticipating, mobilizing, and enabling continuing adjustments in a changing development situation. These multidimensional but unidirectional interactions are surely relevant to those who are exploring the relevance of Taiwan's experience to agricultural transformation elsewhere. -7- II. Agriculture in Transition 2.01 Taiwan emerged from World War 11 with a predominantly. agricultural economy (essentially tenant-operated), a rapidly increasing population, and production levels at that of 1920. The agricultural sector employed over half of the population and generated 44% of the net domestic product. By 1975, these shares had fallen to 30% and 14% respectively. Even as the agricultural sector diminished relative to the non-agrcultural sector, gross domestic product was growing in real terms at 7.7% per annum from 1952 to 1967, and at 10% thereafter. Per capita income increased from $137 in 1951 to $845 in 1975. 2.02 Much of this general economic development is attributed to the dramatic growth performance of the agricultural sector itself. Three strategies for .agricultural development had been adopted in the past 30 years. In the early stage, the emphasis was on increased food production to meet domestic needs. This objective was met by more intensive use of land and labor and improved technology. The second phase emphasized diversification into cash crops (e.g., mushroom, asparagus, pineapple, etc.) which then led to the search for overseas markets in the third phase. 2.03 Agricultural development gained tremendous momentum during the period of post-war recovery and intensification. Within these years, Thorbecke and Karunasekera 1/ distinguished three subperiods of changing development: (i) the recovery period (1946-53); (ii) the growth period (1954-67); and (iii) the slowdown or stagnation period (1968-74), over which rates of growth registered 10.3%, 4.4% and 2.3% respectively. The whole period (1947-74), showed on average growth of 5.1Z p.a. Table 1 shows the growth performance of aggregate agricultural production as well 1/ Thorbecke, E. and Karunasekera, J., "Technological Change and Growth Performance in Taiwan Agriculture, 1946-75", Working Paper No. 176, -8- as for groups of products. The, figures are clearly indicative of the changing emphasis on different products anti the successful response to planned objectives. Remarkable growth rates were registered in Phase 1, especially for common crops, special crops and livestock. In Phase II, livestock continued to grow but much more emphasis was placed on diversification into higher value cash crops like fruits and vegetables, a trend which carried over into Phase II. Meanwhile, common food crops became less important as food self sufficiency was attained. In Phase III, aggregate agricultural production declined as common crops and special crops registered negative growth rates while rice production remained essentially constant. Cash crops, however, continued to grow. 2.04 Successful crop diversification had, by the early 1960s, led to export substitution and the search for market outlets.' This phase corresponded to the increasing tempo of industrial development which provided alternative job opportunities for agricultural labor in a spread of regional locations. By the mid-1960s, gains in agricultural productivity had enabled the transfer of investment resources to the non-agricultural sector. 2.05 The late 1960s and early 1970s marked a turning point in the economic transformation of Taiwan as industrial development continued to overtake growth rates in agriculture and, for the first time, caused a decline in the absolute numbers employed in agriculture. Table 2 shows the changing trends. 2.06 In addition, average farm incomes, traditionally less than half of non-farm incomes, could not keep pace with the latter and ratios decreased from 0.46 in 1960 to 0.38 in 1971. These figures are shown in Table 3. -9- Table 1: TAIWAN: GROWTH RATES OF AGRICULTURAL OUTPUT IN DIFFERENT PHASES OF AGRICULTURAL DEVELOPMENT, 1946-1974 (ANNUAL PERCENTAGE RATES) Agricultural Production Common Crops (Crop-& Rice Special Vege- Live- Phases Livestock) Total Only Crops Fruits tables stock I. 1947-53 10.3 8.1 8.8 17.4 1.2 5.7 17.0 Recovery II. 1954-67 4.4 3.2 3.0 2.9 14.2 5.9 6.3 Growth III. 1968-74 ' 2.3 -0.6 -0.1 -0.8 6.1 9.7 5.9 Slowdown IV. 1947-74 5.1 3.2 3.3 4.9 8.7 6.9 8.7 Source: Thorbecke and Karunasekara, op.cit. Table 2: COMPARATIVE GROWTH AND EMPLOYMENT: AGRICULTURE VS. INDUSTRY 1952 1966 1975 1. Value of Production (a) In N.T $ million Agriculture 16,979 29,730 36,448 Industry 8,324 33,752 93,497 (b) Average annual growth rates (%) Agriculture (4.1%) (2.3%) Industry (10.5%) (12.0%) 2. Labor Employed (a) In million persons Agriculture 1.53 1.71 1.65 Industry 0.38 0.94 1.96 (b) Average annual growth rates (%) Agriculture (0.8%) (-0.4%) Industry (6.7%) (8.5%) Source: H.H. Chen, "The Farm Labor Force of Taiwan: Problems and Prospects". -10- Table 3: FARM INCOME VS. NON-FARK INCOME Ya/La Ya/A Ya/La Yn/Ld Yn/Ln Year ($/ha) ($/pers.) ($/pers.) (ratio) 1952 5,973 3,316 7,706 0.43 1955 9,196 5,097 12,659 0.40 1960 19,016 10,029 21,976 0.46 1965 27,874 14,849 34,178 0.43 1970 36,168 19,583 48,233 0.41 1971 37,308 20,200 53,029 0.38 Notes: Y = income; L = labor; A = agricultural land; a - agricultural sector; n = non-agricultural sector. Sources: National Income of the Republic of China, DGBAS. Taiwan Statistical Data Book, CIECD. W.Y. Kuo, "Effects of Land Reform Agricultural Pricing Policy and Economic Growth on Multiple Crop Diversification in Taiwan", 1973. 2.07 With such wage differentials and fairly dispersed industrial development, farm workers gradually got drawn into jobs in the non-agricultural sector. Rural labor mobility surveys conducted in 1963 and 1968 distinguished 3 categories of the work-force in rural areas: commuters, seasonal workers and long-term employees. Table 4: CATEGORIES OF RURAL-BASED LABOR (%) 1963 1968 Male Female Male Female Commuters 23.5 35.0 45.7 54.3 Seasonal workers 61.0 41.1 27.6 21.8 (agriculture) Long-term employees 15.5 23.9 26.7 23.9 TOTAL 100.0 100.0 100.0 100.0 Source: H.H. Chen, "The Farm Labor Force of Taiwan", op.cit. -11- Table 4 shows that the proportion of male -commuters almost doubled in'this period while female commuters increased from 35% to 54%. These increases, as well as the increases in male long-term employees (from 15.5% to 26.7% in 1968) have been enabled by the decrease in the number of seasonal workers, both male and female. Thus, many male seasonal agricultural workers became full-time workers in non-farm jobs C'commuters') while the. rest became full-time farmers. About 20% of female seasonal agricultural workers became commuters in non-farm employment. These movements indicate increased occupational specialization. 2.08 While there was a net movement of labor out of agricultural employment, family farms showed no significant enlargement in size, averaging 1.05 ha in 1965 and 1.04 ha in 1978 (Tien, 1981). This is due, in part, to traditional attachment to land and reluctadce to sell, even though some or all of the siblings may have obtained outside jobs and may not have the time or interest to work the land. (From interview with Mr. Chang, Farming Group Leader, HsinWu township). In these cases, remaining family members (mainly elders and womenfolk) remain to work what they can on the family farm. Over the years, too, the average age of the residual farming population has increased, as seen in Table 5 below. Table 5: CHANGES IN PROPORTION OF AGE DISTRIBUTION OF AGRICULTURAL WORKERS Age Year Total 15-29 30-49 50 and above 1955 100 53.0 36.1 10.9 1965 100 40.7 45.2 14.1 1973 100 25.4 55.2 19.4 Source: Agriculture Census, quoted in Chen, H.H. "The Farm Labor Force of Taiwan: Problems and Prospects". -12- 2.09 The drastic reduction in the numbers of seasonal farm labor alone would indicate severe seasonal labor shortages, especially during peak harvesting and transplanting seasons. In such a competitive market, demand has led to dramatic increases of daily wages for hired farm labor of the following magnitudes. 1970 - NT $ 67 1975 - NT $194 1978 - NT $255, approximately a fourfold increase in eight years (Tien, 1981). It would appear that a classic case for agricultural mechanization has developed. 2.10 With these various developments, it becomes impossible to maintain a purely labor-intensive farming system. Labor preferences and - mobility indicate that conditions in Taiwan are favorable for extending farm size and accelerating farm mechanization. However, farm size appears to be fairly stable under traditional inheritance and value systems. Mechanization would have to proceed on a fairly divisible scale or require that farm-owners cooperate to take advantage of economies of scale. Whatever the option, (and given the low returns shown in Table 3 above) it is imperative that farmers should find it worth their while to invest in agriculture, through higher returns from assured product prices; higher yields from improved technology; lower production costs from cheaper or subsidized inputs; and assistance from a government that is convinced of the need to siphon profit back from industry to support the agricultural sector that fostered its growth. 2.11 Such were the critical issues that faced and continue to face Taiwan's planners since the 1950s - how could agricultural growth be -13- sustained and agricultural earnings improved to be sufficiently attractive to retain enough workes, owners and operators. Labor shortages and a need for high productivity inevitably necessitated a program for rapid mechanization and supporting packages, the implementation of which brought about the dramatic changes in resource use in Taiwanese agriculture since the 1950s. -l 4- III. Mechanization and Resourcb Use in Agriculture 3.01 The strategies for agricultural development adopted by the Government and the growth performance of agricultural production (discussed in paras. 2.02 and 2.03 above) are correlated with important changes in technology and in the pattern of resource use. The three phases of differential growth of agricultural output have been described in Table 1. Thorbecke and Karunasekera have also demonstrated how different input combinations have contributed to the growth performance in each of these phases (see Table 6). (i) Recovery period, 1946-53 3.02 The post-war period saw a very high annual growth rate for agricultural production of 10.3% with common crops, special cropt and livestock growing at 8.8, 17.4, and 17.07. respectively. The figures in Table 6 attribute this largely to a more intensive use of land and labor and use of current inputs. The annual increase in total workdays of agricultural labor was 6.4% while the actual number of workers rose by only 1.6% per annum. The average number of workdays of labor per agricultural worker rose from 111 in 1947 to 168 in 1953, at an average rate of 4.7%. 3.03 Similarly, the intensity of land cultivation (in terms of cropped area) registered a high annual growth rate of 3.8% while actual cultivated area increased by a mere 0.6% p.a. The multiple cropping index for the same period rose from 117 to 170. The value of current inputs used averaged a substantial 28% annual increase. 3.04 Hence, the recovery phase called land and labor resources into fuller utilization and intensified the application of both fixed capital and fertilizer inputs in agricultural production. -15- Table 6: TAIWAN: GROWTH RATES OF INPUTS IN DIFFERENT PHASES OF AGRICULTURAL DEVELOPMENT, 1946-74 (Annual Percentage Rates) Labor Land Fixed Capital Current Inputs Number Workdays Fertilizer in Total of per Cultivated Crop Total Machinery Total Nutrient Equivalent Phases 1/ Workdays Workers Worker Area Area Value Horsepower Value Chemical Organic (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) I. 1947-53 6.4 1.6 4.7 0.6 3.8 4.7 -- 26.7 n.a. n.a. Recovery II. 1954-67 1.7 0.4 1.3 0.2 0.9 3.5 77.0 7.3 5.7 3.3 Growth III. 1968-74 -0.9 -0.7 -0.2 0.1 -0.7 5.9 2/ 17.3 16.2 2/ 3.7 1.2 Slowdown IV. 1947-74 2.0 0.4 1.6 0.3 1.1 4.0 3/ 49.9 4/ 13.0 5/ 2.5 5/ -1.0 5/ 1/ The annual growth rates prevailing in each phase have been computed on the basis of three-year averages centered on the first and last year of each period, i.e., 1947 = average 1946-48, 1953 = 1952-54, etc. 2/ Average of 1968-72. 4/ Average of 1954-74. 3/ Average of 1947-72. 5/ Average of 1954-74. Source: From Thorbecke and Karunasekara, op. cit. -16- (ii) Sustained growth period: 1954-67 3.05 After the post-war exuberance the annual growth rate of agricultural production settled back to 4.4% per annum, with common crops and special crops growing at 3.2 and 2.9% respectively. Livestock settled down to a growth rate of 6.3% but high-value cash crops like fruits and vegetables registered high growth rates of 14.2 and 5.9% respectively. 3.06 Total labor inputs in workdays increased at 1.7% per annum, against a 0.4% increase in the number of agricultural workers, while workdays per worker continued to increase by 1.3% per annum. Actual per hectare labor input in agriculture increased from 281.2 workdays per year in 1952-56 to 331.2 workdays in 1967-71. Actual cultivated area increased by only 0.2% but cropped area increased by 0.9%. It is apparent that land and labor were being pushed to the margin of full utilization. The value- of current inputs grew by 7.3% and chemical fertilizers had markedly out stripped the use of organic nutrients in their application, especially to the new horticultural crops. 3.07 Over the same period, fixed capital inputs as measured by machinery horsepower increased by 77% per annum while the value of the same increased by 3.5%. These figures point to the use of fairly simple machines which undertook traditionally labor-intensive farm operations, and the scarcity of land that was available for acquisition and farm expansion. (iii) Agricultural slowdown: 1968-74 3.08 Agricultural production grew by a mere 2.3% per annum in this period with negative growth rates for common crops and special crops which were counterbalanced by fruits, vegetables and livestock growing at 6.1, 9.7 and 5.9% respectively. -17- 3.09 The agricultural labor force started to shrink for the first time in 1969 and the period saw a negative growth rate in total workdays of -0.9%. In actual terms, the per hectare labor input in agriculture dropped to 319.4 workdays in 1972-76. This was attributed to a decline in both the number of workers (-0.7%) and average number of workdays per worker (-0.2%). The impact of this is felt on the reduction of cropped area by -0.7% per annum, even as cultivated area increased marginally, by 0.1%. The continued out migration of farm labor brought about the adoption of innovations which were labor-saving and capital-using in nature. 3.10 Bigger and more sophisticated machines were increasingly employed to maintain the levels of agricultural production. Agricultural cooperatives were better placed to utilize more sophisticated and specialized machines. Machinery horsepower for this period increased by only 17.3% p.a. but the total value for 1968-72 increased at the rate of some 5.9% par annum. The growth rates in the use of current inprts had doubled from that of the last period with chemical fertilizers increasing at a rate three times that of organic fertilizer. 3.11- For the whole period (1947-74), the total workdays available to agriculture increased at an average 2.0% per annum, resulting from a mere 0.4% growth in the number of workers but fuller utilization as indicated by a 1.6% per annum increase in workdays per worker. 3.12 As with labor, land was also used more intensively. Cultivated area increased by 0.3% per annum but cropped area averaged 1.1% per annum. 3.13 Fixed capital contributed an average increase of machinery horsepower of 49.9% per annum (average of 1954-74) but a comparatively modest increase in total value of 4.0% per annum (average of 1947-72). -18- These figures attest to the suctess of mechanical innovation and the provision and uptake of locally-manufactured equipment at subsidized rates. 3.14 The intensity of agricultural production was also helped by the application of current inputs whose value registered a 13.0% annual increase iver 1974-72, with chemical fertilizers taking over the role of organic fertilizers at average growth rates of 2.5% and -1.0% per annum respectively. 1. Mechanization Trends 3.15 It is evident that land and labor constraints, as shown by negative growth rates of total workdays and cropped area (1968-74), can only be alleviated through mechanization. From a zero base in 1953, mechanization gradually gained momentum to represent, by 1967, 25% of total energy used in agriculture. The new technology substituted mainly for animal power from the 1950s and for human labor from the late 1960s. In the Slowdown Period (1968-74) an exodus of labor from agriculture further enhanced the economy's dependence on mechanical power and the reliance on human and animal labor accordingly diminished. By 1975, the proportions of mechanical, human and animal labor inputs were 58%, 35% and 7% respectively. The progression of these trends is shown in Table 7. 3.16 These observations are confirmed by comparative farm level data for inputs of human labor, animal labor and farm machines over the years. The data given in Table 8 below are for the first crops of rice only. The table shows substantial decrease in human labor inputs - these decreases were from land preparation, weeding and fertilization, the former due to the advent of machinery and the latter two, from the use of biochemical inputs. However, as a result of more intensive cultivation, -19- Table 7: TAIWAN: SOURCES OF HUMAN, ANIMAL AND MECHANICAL ENERGY IN CROP PRODUCTION IN HORSEPOWER-DAY EQUIVALENT UNITS AND PERCENTAGES OF TOTAL, 1952-75. Human Labor Animal Labor Mechanical Labor in Equivalent % of in Equivalent % of in Equivalent % of Total Energy in Horsepower Days Total Horsepower Days Total Horsepower Days Total Horsepower Days Year (million) Energy (million) Energy (million Energy (million) (1) (2) (3) (4) (5) (6) (7) 1952 24.20 67 11.94 33 -- -- 36.14 1975 28.20 35 6.12 7 -- -- 36.85 1953 24.70 67 12.15 33 0.01 0 37.25 1954 24.60 66 12.64 34 0.01 0 37.13 1955 24.30 65 12.82 35 0.02 0 38.11 1956 25.20 66 12.89 34 0.07 1 39.86 1957 26.90 67 12.89 32 0.25 1 41.17 1958 27.60 67 13.32 32 0.98 3 41.52 1959 27.50 66 13.04 31 1.76 4 41.80 1960 27.00 65 13.04 31 2.92 7 42.56 1961 26.70 63 12.94 30 5.21 12 44.67 1962 26.80 60 12.66 28 6.58 15 46.25 1963 27.50 59 12.17 26 8.13 16 48.28 1964 28.30 59 11.85 25 8.55 17 50.22 1965 30.10 60 11.57 23 14.84 26 56.90 1966 30.80 54 11.26 20 13.21 25 54.17 1967 30.40 56 10.56 19 17.12 29 57.75 1968 30.50 53 10.13 18 20.35 34 59.82 1969 29.90 50 9.57 16 24.19 39 62.38 1970 29.60 47 8.59 14 29.23 44 66.79 1971 29.90 45 7.66 11 32.45 48 68.35 1972 28.80 42 7.10 10 38.86 52 74.05 1973 28.80 39 6.39 9 46.02 56 80.61 1974 28.50 35 6.09 8 47.37 58 81.69 Source: Thorbecke and Karunasekara, op. cit., Table 11. For a description of the methodology underlying this table, see above source. -20- Table 8: CHANGES IN WORK INPUTS PER HECTARE OF RICE, BY FARMING OPERATIONS Human Labor Animal Labor Machine (Workday) (Workday) (hrs) Farming Operation 1926-7 1936-7 1967 a/ 1977 b/ 1926-7 1936-7 1967 a/ 1977 b/ 1967 a/ Land Preparation 16.6 23.3 11.0 8.7 14.6 18.7 7.4 1.7 5.3 Seedling nursery 0.9 6.5 3.7 3.9 0.8 0.1 Transplanting 7.2 11.5 12.1 10.7 0.3 0.1 Fertilizer 8.2 7.3 6.2 5.1 0.9 0.9 0.3 Field Weeding 20.8 30.4 23.6 7.9 0.4 Insect/Pest Control 2.6 1.1 7.2 6.5 Field Irrigation 8.0 9.9 12.5 11.5 Harvesting & Post Harvest 24.8 26.6 28.0 24.4 0.2 0.7 0.2 0.7 Others 7.3 0.1 3.1 9.3 1.7 0.6 0.2 Total 96.4 116.7 107.4 88.0 17.5 21.1 2.0 5.7 a/ Initial farm mechanization period. b/ Advanced farm mechanization period. Source: Lee, Chen & Chen, op. cit. -21- labor inputs in field irrigation and insect and pest control increased. Over 80% of animal labor is used in field 'preparation and the sharp decrease in total animal work workdays is directly attributed to increased mechanization. 3.17 Over the more general crop area, the same pattern is observed - see Table 9 below. In the 1960's farm machines replaced draft animals rather than human labor; hence human labor inputs were maintained at over 400 workdays per hectare of cultivated land and over 200 workdays for crop area. In fact labor inputs increased slightly due to the introduction of labor-intensive horticultural crops (mushroom, asparagus, tomatoes and other vegetables) and livestock enterprises. From 1969, increased mechanization enabled substantial reduction in labor inputs while animal labor inputs continued to decrease. 3.18 Thorbecke and Cunasekara have estimated the major determinants of mechanization with a recursive model using cross-sectional data covering 8 regions and 5 farm sizes. Mechanical energy per workday was positively correlated with: (i) labor costs (which increased with multiple cropping and use of hired labor but was negatively correlated with usage of family and animal labor). (ii) multiple cropping index, since increased number of farm enterprises required the tighter scheduling cf activities and greater timeliness of operations which mechanization enables; (iii) farm size, which, though positively correlated with mechanical energy used, was negatively related to both the use of animal labor and biochemicals per unit of crop area (indicating more intensive cultivation). -22- Table 9: CHANGES IN LABOR INPUT AND. MACHINE USE IN TAIWAN AGRICULTURE Per Hectare of Cultivated Land Per Hectare of Crop area Human Animal Machine Human Animal Machine Year Labor Labor Power Labor Labor Power (workday) (hours) (workday) (hours) 1961 408 28.4 - 182 12.7 - 1963 430 25.0 7.1 192 11.2 3.2 1965 408 18.8 5.8 202 9.3 2.8 1967 412 17.3 7.2 206 8.7 3.6 1969 355 13.1 9.3 180 6.6 4.7 1971 321 10.3 10.2 173 5.5 5.5 1973 289 9.2 16.6 161 5.1 9.3 1975 281 7.8 20.9 153 4.2 11.4 1977 272 6.3 25.6 150 3.5 14.1 Source: Provincial Department of Agriculture and Forestry, quoted in Lee, Chen & Chen, op. cit. -23- 3.19 The record of investmhnt in farm machinery (1965-1979) is shown in Table 10 and Chart 1. The more traditional equipment such as water pumps, power sprayers and rice threshers registered very steady increases. With the mechanization of land preparation,, there was a dramatic increase in the numbers of power tillers, averaging some 3,200 per year in 1965 - 1970, which brought about a concurrent decrease of about 19,000 heads of draft cattle per year. At this stage, machines were used mainly in land preparation. By 1967, per hectare labor ilput in land preparation was only 11 days compared to more than 20 days when done with draft animals. 3.20 Other labor-saving machines like grain dryers, transplanters and combine harvesters quickly followed as other farm operations in the peak planting and harvesting seasons became mechanized. In addition,*bigger - power tillers and four-wheeled tractors were increasingly employed. Power tiller capacity increased from 10 hp in 1965 to 20 hp. by the 1970s, even as the unit price dropped. Small tractors of 25 to 40 hp. are also used while big wheel tractors of 70 hp. capacity have been introduced for dryland farming. Tractor and tiller horsepower per ha. increased from 0.12 in 1965 to 0.89 in 1979. Mechanized tillage has reached the 85% level with power tillers substituting for draft cattle at the rate of one power tiller to 4 to 5 animals. 3.21 Expenditures on farm machinery in relation to annual capital investment in agriculture were NT 3 billion (or 10%) in 1965-70; NT 1.9 billion (or 19%) in 1975 and NT 2.4 billion (or 16%) in 1978. Most of these new machines operated on diesel. The consumption of diesel in the agricultural sector increased from 150,000 k1 in 1970 to 520,000 k1 in 1978, about one-quarter of the total national consumption. -24- Table 10: USAGE OF FARM MACHINERY AND DRAFT ANIMALS (units of machineriheads of animals) Machinery/Animal 1965 1970 1975 1976 1977 1978 1979 Power tiller 12,213 28,292 49,347 55,748 66,698 74,504 84,254 4-wheel tractor 425 539 1,380 1,718 1,879 2,296 2,894 Rice transplanter - 280 2,787 6,108 11,138 18,778 25,820 Power sprayed 4,489 17;820 37,874 37,489 45,582 44,917 52,554 Water pump 32,107 52,794 124,626 123,645 141 210 141,429 142,501 Rice power thresher n.a. n.a. 27,558 30,470 32,074 44,270 35,768 Grain dryer n.a. 198 2,419 9,269 18,147 21,544 25,032 Rice combine 150 20 2,010 2,811 3,930 7,300 10,569 Draft animals 370;370 275,007 195,770 188,748 122,144 106,052 95,111 Source: Tien, op. cit. -25- 2. Concurrent Changes in Agricultural Organization Improved technological packages 3.22 The high agricultural productivity that was enabled in part by mechanization would not have been possible without simultaneous development of an improved technology package that included improved crop varieties, seed multiplication and nursing facilities, improved water control systems, the production and supply of chemicals and fertilizers, and provision of extension and credit. Varietal selection 3.23 The example can be taken of rice, the major crop which has been cultivated over some 750,000 hectares annually, and which was, up to 1966, an important earner of foreign exchange. The annual area of rice has changed very little in the last 30 years. Yields and production, however, have both increased steadily at an average annual rate of 2.3% between 1950 and 1977. These changes are shown in Table 11 below. 3.24 Problems of severe blast and lodging were encountered when farmers attempted to apply high levels of fertilizer (which was increasingly available, and at lower prices) to older varieties like Taichung 65. Other lines have had to be developed which resist blast and yield best at high nitrogen levels. Fertilizer credit 3.25 The use of fertilizers gained widespread acceptance among farmers. through the demonstration of supply response. Government had the monopoly of fertilizer which was initially bartered for rice from the Japanese and only later was produced locally. The same barter system was extended down to the farmers who made partial payment for fertilizer at the time of -26- Table 11: CHANGES IN RICE PRODUCTION Year Planted Unit Total Area Yield Production (ha) (kg/ha7 ('000 MT) 1950 770,262 1,845 1,421 1955 750,739 2,151 1,615 1960 766,409 3,038 2,348 1965 772,918 3,038 2,348 1970 776,248 3,173 2,463 1977 777,609 3,406 2,649 Source: Lee, T. H. Chen, H. H. Chen Y.E., "Labor Absorption Taiwan Agriculture", in Labor Abs -27- procurement and settled the rembinder (with interest) at harvest time, both payments being made in padi. This system allowed the farmers credit on a major input, made against the promise of the new crop, and allowed Government leverage in adjusting desired levels of agricultural taxation and subsidy by varying the prevailing exchange rates of fertilizer for padi. Rice nurserying techniques 3.26 In propagating the new varieties and to achieve uniformity in timing it has been found economical to establish cooperative rice nurseries. A Government program was started in 1973 which subsidized the establishment of rice nursery centers to about 25% of investment costs. About 90 such centers have been set up, each supplying rice seedlings to 10 to 13 joint farming groups which collectively plant between 150-200 ha. under a common irrigation system. Joint farming groups can operate their own smaller rice nurseries. They are usually operated by a lead farmer who employs some fellow farmers to work with him and supplies rice seedlings to groups in his immediate locality. A total of 450 such nurseries have been set up. It is also common practice now to raise rice seedlings in seed boxes rather than in seedbeds, a practice that ties in with the use of rice transplanters. Ownership and cooperative use of farm machinery 3.27 More than 90% of farm machines are owned by individual farmers. These include smaller and cheaper items like manual sprayers, dusters, weeders, threshers, winnowers, transplanters and dryers. Larger machines. like tractors are either cooperatively owned or owned by agricultural cooperations such as the Taiwan Sugar Corporation. The 1975 agricultural -28- census showed that ownership of' machines was largely concentrated in the hands of farmers owning more than 1.0 ha. of land. Some illustrative figures are shown in Table 12 below. 3.28 Farmers are organized into groups covering 15 ha or more and undertake transplanting, spraying and harvesting of rice communally. The use of machines for these jobs is greatly facilitated by the implementation of the land consolidation program in irrigated areas by which individual plots were rearranged and consolidated into large rectangular fields. The machines may be owned by individual members who do custom work for other members at agreed rates. This work may be arranged by the owner himself or sometimes through Farmers' Associations. The extent of cooperative usage of farm machines is indicated in Table 13. Farmers who owned farm machines and over 2.5 ha of land spent most of the machines' working time on their own land. However, there were farmers with less than 2.5 ha who found it worth their while to own machines and do custom work for others for 64 to 88% of the time worked. On average, 52 to 63% of machine time for power tillers, rice transplanters and rice combines were spent on doing custom work on farms other than the machine-owners'. The exception to this is the power sprayer, a smaller and less costly piece of equipment, which is used for 78% of the time on the owner's own farm. In these various instances, full utilization of farm machinery was possible, inducing private 'investment in these inputs. 3.29 These various changes in agricultural practices at farm level have resulted in high cropping intensities of 170 to 190%, as indicated by the following figures: -29- Table 12: OWNERSHIP OF FARM MACHINES -- NO. OF FARMERS PER UNIT OF MACHINERY Under 0.5 - 1.0 - 1.5 - 2.0 - 3.0 ha Farm Size 0.5 ha. 1.0 ha 1.5 ha 2.0 ha. 3.0 ha & over Power tiller 56.8 15.9 7.8 6.0 4.4 3.3 Water pump 11.0 5.3 3.9 3.2 2.6 2.2 Power Sprayer 63.3 18.6 9.6 6.9 5.0 3.0 Power thresher 34.6 8.8 5.2 4.8 4.2 3.9 Source: Tien, op. cit. Table 13: CUSTOM HIRING OF FARM MACHINES (% of time worked) Power Rice Rice Tiller Transplanter Power Sprayer . Combine. Own For Own For Own For Own For Farm Others Farm Others Farm Others Farm Others Average 36.8 63.2 47.8 52.2 78.4 .21.6 47.7 52.3 Under 1 ha 17.3 82.7 10.9 89.1 39.2 60.8 15.9 84.1 1.0-2.5 ha 32.3 67.7 36.4 63.6 78.8 21.2 11.3 88.7 2.5-5.0 ha 58.5 41.5 79.5 20.5 86.9 13.1 76.0 24.0 5 ha & over 66.5 33.5 83.1 16.9 88.9 11.1 89.7 10.3 -30- Year Crop land !Cultivated area- Cropping intensity 1964 1,673,000 ha 882,000 ha 190% 1978 1,549,000 ha 918,000 ha 169% Increases in agricultural productivity have been measured by relative increases in total output and total input into agriculture. Table 14 shows that up to 1972, increases of total output were largely brought about by direct increases in inputs but in 1972-76, improved efficiency in production more than matched the contribution of factor inputs (which had increased by only 1.7% p.a) in the proportion of 54:46. 3.30 Duch of this has resulted from undertaking farming operations on a more economical and viable scale. Changes In cropping patterns and technological innovations have enabled increasing specialization. These changes have taken place within a carefully preserved framework of individual ownership of land and capital, while promoting the opportunities and economic incentives for pooling resources by sharing a rental. As with other inputs, success in the adoption of farm machinery has been dependent on maximum utilization through cooperative usage. Farmers in Taiwan are accustomed to exchanging labor, hiring draft animals, and sharing the use of farm tools. Building on this, Government has institutionalized various programs and measures to reorganize land distribution, supply improved inputs and infrastructure, coordinate individual and group actions, and reinforce cooperative ventures with legislation and financial support. These programs are discussed in the next section. -31- Table 14: RELATIVE CONTRIBUTIONS OF INPUTS AND PRODUCTIVITY IN AGRICULTURAL PRODUCTION Annual Growth Rates Relative Contribution by - Total Total Total Pro-* Output Input ductivity Input Productivity Period (1) (2) (3)=(l)-(2) (4)-(2)/(1) (5)-(3)/(1) 1946-51 10.2 7.8 2.4 76 24 1951-60 4.7 2.7 2.0 57 43 1961-72 4.2 3.2 1.0 76 24 1972-76 3.7 1.7 2.0 46 54 Source: JCRR Economic Digest Series, No. 21, quoted in Tien, op. cit. -32- IV. The Supporting Packages 5 4.01 The Taiwanese government, in a manner reminiscent of a benevolent patriarchal tradition, has historically assumed the major role in spearheading and guiding institutional development. Development needs are continually monitored, assessed, and incorporated into legislative, policy and regulatory measures that extend even into the regulation of interest rates and determining the allocation of bank credit among different priority sectors. The orderly development of agriculture had its roots in the early twentieth century with the creation of essential irrigation infrastructure, and thereafter charted its course in response to other agricultural needs and logical extensions of these development programs* It was this element of organization that permitted the eventual commercialization and mechanization of agriculture. Irrigation Infrastructure and Organization 4.02 The predominance of rice culture and the uneven distribution of rainfall made irrigation essential in Taiwan. The steep topography and short, swift rivers also made flood control necessary. In response to these needs, a widespread network of irrigation and drainage systems evolved through the efforts of local government authorities and farming cooperatives. 4.03 In 1921 the then Japanese colonial government (1896-1945) took steps to combine various cooperativeq and authorities that operated community and government canals to form irrigation cooperatives. These numbered 108 by 1937 and were further coalesced into a mere 38 cooperatives by 1942. The postwar expansion and improvement of irrigation facilities -33- led to further amalgamations ana land consolidation such that by 1975, a total of 16 Irrigation Associations remained to undertake the jurisdiction of some 440,000 ha., or 80% of total irrigated farmland in Taiwan. Under the Accelerated Rural Development Programs (started in 1972), government has provided subsidies up to 80 to 95% of the cost of improvement and repair of irrigation and drainage facilities, with matching funds from the farmers themselves. 4.04 The major functions of the Irrigation Associations include the construction and management of the facilities, regulation of water distribution and arbitration of disputes, collection of membership fees and loan repayments, and coordination with government in carrying out policies on land and water resources development. Each Irrigation Association has - charge of several working stations, managed by irrigation groups which operate 50 to 150 ha. each. An irrigation group comprises several irrigation teams of 10 to 15 members. Both group and team leaders are elected by the members. The groups meet to work out and monitor detailed plans which take into account water availability, types of crop and areas planted, planting schedules, irrigation methods, water requirements and delivery schedules, conveyance losses, and so on. Such detailed planning gives rise to the discipline of working together on cooperative projects. In addition, the development of irrigation facilities in Taiwan has eliminated boundaries and footpaths, increased cultivable land, and enabled multiple cropping and higher intensity of cultivation, with more dependable and higher yields from superior varieties, modern inputs and improved techniques. -34- Ownership of Production Resources - Land 4.05 The land reform program in Taiwdn was started in 1949 and conducted in three phases: (i) Rent Limitation Scheme; (ii) Sale of Public Land; and (iii) "Land to the Tiller" program. 4.06 The rent reduction program had two main thrusts. It legally limited the amount of rent to 37.5% of the major crop harvested (usually rice) and provided tenants with greater security on their lease contracts. Tenant farmers had been commonly charged more than 50% of the harvest in rent before the reform. In addition, new technology enabled farmers to diversify into cash crops on which rents were not levied, so farmers were able to acquire surplus cash. Rent control also caused an effective reduction in the market value of land of 20 to 30% within two years of its- enforcement. 4.07 These developments set the stage for the transfer of ownership program which was successfully implemented in 1953. The program started with the sale of public land to tenant farmers. Using land taken over from the Japanese colonial government and left in the lands of the Sugar Corporation as an example, land was parcelled into two to three hectare lots and sold to tenant farmers. These downstream procedures were thus established and followed by the third stage of transferring excess private tenanted land from landlords to tenants through compulsory government purchase as an intermediate step. A total of 71% of the total area of public and private tenant-operated lands changed hands in this manner. The landlords were compensated with certificates which matured in twenty installments and could be redeemed for padi or cash, thus providing them a regular income. -35- The effect of land rbform on land ownership is seen in Table 15: Apart from change in ownership status, a,second important effect. of land reform was on the reduced size of land holdings, with the major redistribution from the 3 to 10 ha. category into smaller farm sizes. After land reform, 52% of all holdings came within the 1 to 3 ha. category. The changes are shown in Table 16 below. 4.08 With the confidence in ownership of land and accessibility to new techniques, farmers have not hesitated to add agricultural inputs, even with borrowed capital. Economic incentives have induced the farmers to adopt new methods and tools in production. In the decade 1950 to 1960, net farm incomes increased by 178%, enabling wage incomes to increase by 208%, while rent payments declined to 90%, and farmers now incurred net loans - with iterest payments which increased 400% and taxes and fees that had increased some 300% [See table 17]. These changed indicated the radical transformation of traditional tenant agriculture into a largely owner-operated, and increasingly commercial undertaking. While land reform was highly instrumental in promoting farmers' incentives, progress also has to be attributed to the presence of a well-organized institutional framework through which extension and farm inputs could be'channelled. This was provided by the Farmers' Association network in Taiwan. Farmers' Associations (F.A.'s) 4.09 Farmers' organizations in Taiwan include Farmers' Associations, Irrigation Associations, Fishermen's Associations, Forest Protection Associations and Fruit Cooperatives. In this section, we shall discuss the far-reaching effect that the F.A.'s had in promoting the application of modern agricultural technology. -36- Table 15: CATEGORIES OF LAND OWNERSHIP BEFORE AND AFTER LAN]5 REFORM BEFORE REFORM (1949) AFTER REFORM (1961) CHANGE.(1949-61) Families % Families % % Owners 244,378 36 500,268 65 205 Tenants 239,939 39 109,215 14 46 Part owners 156,558 25 186,519 21 106 Total 620,875 100 776,002 100 125 Source: JCRR, Rural Economics Division, quoted in H.H. Chen "Impact of Land Reform on Economy" Table 16: CHANGES IN SIZE OF LANDHOLDINGS BEFORE AND AFTER LAND REFORM BEFORE REFORM AFTER REFORM CHANGE SIZE OF HOLDING .1 1952-60 Ha % Ha % % Below 0.5 ha 67,511 10 80,320 11 119 0.5 - 1.0 ha 102,577 15 165,873 22 162 1.0 - 3.0 ha 227,890 33 380,028 52 167 3.0 - 10.0 ha 175,064 26 111,613 15 39 Over 10.0 ha 108,108 16 111,613 15 39 Total 681,154 100 737,834 100 108 Source: JCRR, quoted in H.H. Chen, "Impact of Land Reform on Economy" Table 17: CHANGES IN NET FARM INCOME 1950 1960 CHANGE NT 7 NT % T73.O Net farm income 2,252 100 4,003 100 178 Total wage income 878 39 1,830 46 208 Total rent payment 975 43 878 22 90 Total interest payment 259 12 877 22 339 Taxes and fees 140 6 418 10 299 Source: JCRR, quoted in H.H. Chen, op.cit. -37- 4.10 In the period 1900-1943, a plurality of organizations existed in Taiwan. There were the Farmers' Organizations (FAs) started since 1900 as spontaneous associations of farmers with the goal of "guaranteeing the right to till the land and to strive for rent reduction". At that time they were mainly dominated by nine big commercial farmers. 4.11 Other than the FAs, there were the Agricultural Cooperatives which were responsible for credit, supplies, marketing, warehousing and processing facilities. Then there were the Small Agricultural Units (SAUs), comprising smaller groups of farmers who undertook cooperative work like rice nurserying, as well as being used as the point of contact for extension work done by the FAs. Between the FAs and the Agricultural Cooperatives, the Japanese government employed, by 1930, some 40;000 people, second in size only to the government bureaucracy itself. Of these, 13,000 were extension workers (at the rate of 32 farm households per worker) and 9,000 were agricultural advisors attached to the SAUs at the village level. 4.12 The FAs and Cooperatives were merged in 1944-46 and later evolved into-a federated system of multipurpose cooperative organizations, responsible for promoting farmers' interest, knowledge and skills, increasing farm production and income, improving living conditions, and developing the rural economy. In 1950, the Joint Commission for Agricultural Development (present-day Council for Agricultural Planning and Development) moved to reorganize FAs to enable popular participation and control by bona fide farmers, while non-farmers/businessmen were relegated Associate member status. FAs were formed at village, township, country and -38- provincial levels. The villageFAs were essentially the former SAUs with 200 to 300 members each, representing 85 io 95% of the village community. 4.13 From 1952, an agricultural extension department was established in each Township FA. (Other sections were established for Rural Credit, Marketing and Supply, and Livestock Insurance). Each Township FA had 3 to 4 farm extension advisors (1:1, 200 farm families); and 1 to 2 4-H Club Organizers who would appoint local outstanding farm women as volunteers to assist them. By the end of 1974, the Extension service employed 548 extension workers. Assisted by 3,900 voluntary local leaders, and working through some 5,300 study groups at farm level, they have provided technical services and trained some 133,000 farmers. Effective extension services have stimulated farmers' demand for credit and farm supplies. 4.14 The FA's, for instance, arrange to receive fertilizer supplies from the Food Bureau at FA warehouses, and then arrange for transportation to farmgate. The transportation charges used to be subsidized by government, but have g'-adually devolved on to farmers who buy the fertilizer. For the past 7 years, government had also contributed an across-the-board subsidy for fertilizer plants. 4.15 The FAs help with purchase of farm machinery as well. The extension worker helps the farmer to file application for loans and subsidies (up to 20% of the cost, depending on type) which are submitted to the Provincial Department of Agriculture and Forestry. The farmer would then bring proof of purchase to the County FA which arranges to reimburse the subsidized portion of the purchase price directly to the seller concerned. The FA guarantees repayment of loans to the Land Bank and Cooperative bank and levies handling charges at 2 to 5% of the loan -39- amount. Some 40 Township FAs are' provided with farm machinery w6rkshops for repair services which farmers pay for., while members of the other 232 Township FAs take their repairs to local private workshops. The close contact that the FAs maintain with the farming community, and their role in arranging purchases and subsidy on farm machinery has been a major factor in the propagation of new techniques, among farmers in joint farming groups. 4.16 By the 1960s, subdivision of holdings had become severe enough to inhibit small farmers from fully adopting technological packages. Half the landholdings were less than 1.0 hectare. A survey by the National Taiwan University in 1964 found the following: Table 18: PERCENTAGE OF ADOPTION OF RECOMMENDED FARM PRACTICES CROP: RICE Recommended Practice Percentage of Adoption Seed disinfection before sowing 62.0 Thin Sowing 78.7 Improved nursery beds 8.0 Pest Control 62.3 Deep ploughing 34.0 Straight transplanting with close spacing 86.1 Fertilizer application to recommended levels 45.1 Source: Hong, P.P., 1979. -40- It was also found that using all the recommended practices could raise rice output by some 15 to 30%. The 'idea of joint farming for rice was introduced at farmer discussion groups organized by local extension workers. Twenty to thirty farmers would pool resources and cultivate a block of 12 to 15 ha of land as one farm, raising the same variety and using bigger and better farm equipment with division of duties for specialized functions. In 1964 the Rice Joint Farming Project was initiated with 100 groups, comprising 2,325 farm families, cultivating blocks of 10 to 20 ha each. By 1966, the project had expanded to 296 groups with 7,291 rice farmers participating. Part-time farmers could entrust farming operations to project farmers and pay for their services at agreed rates. Specialization led to greater efficiency through division of labor and investment in equipment. A "with" and "without-project" comparison of production, costs and profit yielded the following: Table 19: RESULTS OF RICE JOINT FARMING PROJECT IN TAIWAN, 1975-1978 Unit Production Production Cost Crop With Without A x 100 (NT per 100 kg grain) Year Season Project Project B With Without C 100 Project Project D (A) (B)* (C) (D) 1975 1st 4,925 4,379 112 512 582 88 2nd 3,692 3,309 112 712 803 89 1976 1st 5,144 4,760 108 531 591 90 2nd 3,918 3,408 115 696 820 85 1977 1st 4,920 4,654 106 582 636 92 2nd 3,866 3,418 113 707 825 86 1978 1st 4,752 4,511 104 612 667 92 2nd 3,498 3,111 127 867 1,015 85 Source: Hong, P.F., Op. Cit. -41- Farmers in each locality are also encouraged to buy the same models of machinery to facilitate repair and service arrangements. In Hsin Wu for instance, a typical farming group comprised fifteen members, ten of whom were working in off-farm jobs (construction, industry, etc). The remaining five full-time farmers worked the total of 16.4 ha. and owned between them four tractors (of varying sizes, typically 18 hop.), three transplanters and one combine harvester. The group had agreed on rates at which the operators charged absentee farmers for work done on the latter's plots. These were quoted as follows (May 1981): Job Done NT $ Ploughing and levelling 8,000/ha Transplanting 4,500/ha Spraying 1,400/ha Fertilizing 500/day x 2 days/ha Harvesting 8,000/ha Farm machinery was purchased at subsidized interest rates through the Farmers' Association under which the group was formed; servicing and spare parts supply was reported to be prompt and efficient, and both full-time and part-time farmers found the systeT extremely convenient and workable. An informal secondary level organization has also evolved among 11 farmer groups; each unit calls on the others to help out where necessary. Farmers were cooperative and willing to participate due to the shortage of labor. 4.18 In 1970, joint farming operations were introduced for vegetables, fruit and livestock and different types of farm enterprises. By 1974, with the support of the Accelerated Rural Development Programme, the rice joint cultivation programme was extended to cover the entire paddy fields in each township as a "specialized agricultural area" to promote total adoption of improved rice production techniques. With such institutional -"42- support and group participation, farmers were poised to venture into new crops and practices which the individual, acting alone, would be hesitant to risk getting into. As a group, farm planning and record keeping became important and increased management efficiency was thus enabled. The other advantages of joint farming were that farmers, both large and small, gained access to new techniques and higher production and incomes; labor efficiency was improved; land belonging to part-time farmers and absentee owners (engaged in non-farm jobs) was cultivated by those who remained, for a fee; government assistance in organizing rural infrastructure, extension and training became better coordinated with local level organization; and economies of scale were realized from expanded farm operations. 4.19 In planning the use of farm machinery, for instance, the Council- for Agricultural Planning and Development (CAPD) assesses how many units of different types of machines could be fully utilized in the area. Farmers in the groups decide what they can buy, with the assurance that the machines would be fully utilized in their own farms and also in doing custom work for cooperating farmers at agreed rates. The farmer-owner provides the labor to operate the machine. CAPD assists him to get loans through the local Farmers' Association, together with any--subsidy-a.may-be- currently determined by the Farm Mechanization Promotion Committee. Accelerated Farm Mechanization Program 4.20 Farm mechanization in Taiwan started as far back as 1956 when power tillers were first used for land preparation, gradually replacing draft cattle. Over the years, it became apparent that mechanization had - come to stay and government accordingly gave it high priority. Following the new Agricultural Policy of 1969-70, a four-year Farm Mechanization -43- Plan was formulated and a Farm Mechanization Promotion Committee was established, under the Ministry of Economic Affairs in 1971. The Committee had the responsibility of mechanizing 450,000 hectares of farm land through the installation of 120,000 sets of farm machinery, including power tillers, rice transplanters, combines, reapers, grain dryers, pumps, sprayers and other farm machines. A loan fund was set up with contributions from JCRR (13%), Ministry of Finance (37%) and the three agricultural Banks (50%). The Agricultural Mechanization Fund was started in July 1978 with an annual NT $ 1 billion allocation from government, which by 1981 had grown to NT $4 billion. From July 1978 the Farm Mechanization Plan was extended to cover machinery requirements in fisheries, forestry and livestock, 4.21 Loans were either disbursed to borrowers directly, through the Provincial Food Bureau, or Township FAs (which earned a handling charge of 1.25%). The Farm Mechanization Promotion Committee decided which items of agricultural machinery should qualify for subsidized loans at 8-1/2% interest per annum. Subsidies were also given on the purchase price of equipment - for example, 15% and more on power tillers and 20 to 25% on rice transplanters, grain dryers and rice comi)ines. Over time, and depending on the rate of uptake of these machines in relation to implementation targets, the rate of subsidy is reduced. At present only the rice combine and grain dryer are subsidized to 10% of their purchase prices. Government also provides a subsidy of about 30% on charges for electricity used by irrigation pumps and grain dryers. 4.22 A Farm Mechanization Fund Loan Program for FY 1979-82 was initiated in July 1978. The Government provides NT $4 billion at the rate -44- of a billion a year to the Fund!, with an equal matching contribution from three agricultural banks and township Farmers' Associations. Farmers pay an interest rate of 8.5% p.a. for credit .over a seven-year period. Loans are used to purchase farm machines which have passed the government's inspection and performance tests. 4.23 The provision of subsidy on farm machines is also contingent on such machines having passed the inspection and performance tests. This requirement helps accelerate the adoption of new machines and complements the policy of supporting local machinery manufacturers. Manufacturers can apply for low. interest investment loans, subsidized to 6% p.a. A tax exemption on 2.5% of their annual total sales is proposed for those manufacturers who set up Research and Development Departmetts to intensify these activities. Government's support has resulted in the rapid takeoff of local manufacturing industry. Power tillers were first imported in the 1950s; since 1975, all have been locally produced except for some 10% of components which are still imported. Rice transplanters, combines andgrain dryers were first imported from Japan in 1970, and by 1975-76, local manufactured had started; today, they are almost wholly locally- manufactured except for some 20% of component parts. Simpler machines like power sprayers, waterpumps and rice threshers have been domestically manufactured since 1969. Bigger tractors continue to be imported from USA, Japan and Europe though there are medium term plans to produce 33-50 h.p. tractors based on Japanese models. On average, the Industrial Development Bureau estimates that less than 10% of agricultural machinery is imported. A local manufacturer estimates that there are about 100 private small -45- workshops around the island that make cheaper reproductions of Japanese machinery with varying exactness, for purely'local applications. 4.24 An extensive and competitive network of service centers operate to provide spare parts and prompt after-sales service to farmers. In addition to these facilities, Farmers' Associations in major townships have established Agricultural Mechanization Promotion Centres (AMPCS). Their functions are to maintain and repair farm machines owned by individual farmers, arrange custom work for farmers who do not have such machines and also to make full use of the operating units; assist farmers to complete bank procedures when applying for machinery loans; and to carry out programs that train the farmers to operate different kinds of machinery and carry out minor repairs like fixing the fuel and ignition systems of an engine, and replacing chains, belts and gaskets. 4.25 The Government plans to reorganize the manufacturing industry and eventually sponsor five major manufacturing concerns which will supply all the farm machines required locally and seek to expand sales overseas. Already agricultural machinery specialists have been sent to countries in --- Southeast Asia, Latin America and Africa to promote Taiwanese manufactures and assess potential overseas demand. It is recognized that Japanese manufacturers have preceded them and that for the moment, Taiwanese machinery has no great price advantage over Japanese exports since their lower wage cost is counterbalanced by Japan's larger output and lower unit cost advantage. However, with Government support, five years' tax holidays for manufacturers, and other instruments, Taiwan's machinery manufacturers anticipate a systematic penetration of user-markets overseas. Indeed, machinery exports from Taiwan increased from US$ 16 million in 1976 to over US$ 45 million in 1979. -46- Role of JCRR/CAPD 4.26 The Joint Commission on Rural Reconstruction or JCRR (now called the Council for Agricultural Planning and Development or CAPD) has played a key role in providing technical assistance and financial aid for agricultural and rural development in Taiwan. Sponsored by U.S. aid funds, JCRR moved from Mainland China to Taiwan in August 1949. JCRR is concerned with both technological improvements and institutional changes and readjustments. The former includes the propagation of new crop varieties, cultural techniques, irrigation, and the use of new inputs. The latter includes the improvement of marketing facilities, development of the agricultural credit system, the reorganization of Farmers' Associations, and training of rural youth. JCRR has also been instrumental in' implementing projects on farm mechanization, livestock, slope-land development, food processing and export, working closely with various other sponsoring and local cooperating agencies. Grants and loans are provided to farmers, agricultural institutions and local governments to carry out agricultural reconstruction, reasearch and extension work. Emphasis is placed on projects which are demonstrative or innovative in nature, the financing of which involves extended risks. Once such activities can be made financially viable, they will be referred to Agricultural Banks or Township Farmers' Associations for regular financing. Before FY-66 JCRR was funded by the Counterpart Fund and from FY-67 by the Sino-American Economic and Social Development Fund. Of the average annual appropriations of NT $405 million, half was used for loans. In accordance with national. priorities, the bulk of lending was made to livestock and fisheries project (30% and 24%, respectively) in FY67-73; from FY-74, emphasis shifted to -47- programs to accelerate farm mechanization and rural development. The role of JCRR in assessing the need for, and promoting the uptake of farm machinery in particular localities has been discussed in paragraph 4.19 above. -48- V. SUMMARY AND CONCLUSIONS 5.01 The record achievements in the agricultural development of Taiwan have been enabled by a blend of economic circumstances, private initiatives and Government instrumentality. The exuberance of post-World War II recovery started to tail off in the 1960s as agricultural growth rates declined drastically (see Table 1). Industrial development gained momentum and caused a severe drain of workers from the agricultural sector due to the higher wages offered (Tables 2 and 3). The residual agricultural labor force became characterized by more part-time workers and higher average age of farmers (Tables 4 and 5). Government monitored these changes and sought to make agriculture a more viable and worthwhile occupation, sustained by improved technology and supported financial incentives. Its remarkable success in maintaining food production, diversifying into expanded cash crop production by mobilizing capital, research and development, extension efforta, infrastructure development, and organizing farm groups towards these ends emphasize the need for a comprehensive and coherent multi-pronged approach to agricultural development. 5.02 We have analyzed, as a specific case in point, the farm mechanization program,'a major factor in the rapid transformation of Taiwanese agriculture from a traditional, labor-intensive, subsistence occupation and way of life for the majority of the population, into a modern, dynamic, capital-intensive, labor-efficient agricultural industry operated by less than 30% of the population. The mechanization program as it developed in Taiwan had the elements of an "improvement" approach. There was no massive injection of capital at any point ih time nor any -49- attempt to radically reorganize' the ownership or management of agriculture. That mechanization proceeded so rapidly and farmers and other private entrepreneurs responded so positively to Government programs and incentives indicates the sensitivity with which the mechanization strategy had been evolved. Initiated in response to acute shortage of agricultural labor and decline productivity, the mechanization program gradually developed its own momentum and required other ancillary reorganizations of farm practices and institutional arrangements to enable the full potential benefits of mechanization to be realized. In the short run, the program sought to bring about a transfer of technology to alleviate operating constraints in agriculture as the rest of the economy developed. In the longer run, the program derived its continuity from the complemehtary and - enabling policies and institutional arrangements that provided suitable infrastructure and working environment for the innovations to take root. 5.03 The mechanization program had elements which fulfilled the following conditions for its successful adoption: (i) Complementarity (of other Supporting Programs) (a) New, improved technological packages were continually tested and propagated - these included seed and varietal improvements; breeding to enable higher cropping intensities and handling by machines; increase in sophistication of cultivation and handling techniques; supply and maintenance of essential inputs like fertilizer machines, fuel, services and accessibility of these various inputs to farmers in all localities; -50- (b) Ongoing institutional' reform and development, including Land Reform, provision of Irrigation infrastructure and organization, institutional credit subsidized through Farmers' Associations; Organization of Joint Farming Groups, Farm Mechanization Promotion program; and the promotional and planning activities of JCRR/CAPD; (c) Government support for manufacturing, distribution and servicing of farm machinery together with subsidized credit to farmers for their purchase of machinery and training programs for their usage; (d) Channelling of extension, credit, supply and marketing functions through the Farmers' Associations which farmers find abcessible and convenient to refer to and operate through; FAs in turn, being locally-constituted and on-site, are well placed to offer advice and monitor the use of loans and facilities by their farmer-members. (ii) Congruity and Compatibility (with the human resource conditions) (a) Mechanization was selectively implemented according to felt needs and resource bottlenecks. For instance, in the 1960s, power tillers replaced draft animals and reduced labor requirements for land preparation. With the agglomeration of farming units, transplanters, harvesters and dryers were introduced to achieve timeliness of operations; -51- (b) Joint farming operati6ns succeeded in large part due to the tradition of cooperative effort, 4xchange of labor, sharing of farm tools, and participation in maintaining common-user facilities (irrigation, etc.); (c) A relatively monetized economy and familiarity with financial management was advantageous for the introduction of a systcn of paid contractual services, and for persuading farmers of the economics of organized farming. (iii) Complexity The simplest operations were mechanized first. Water pumps, sprayers and rice threshers were familiar equipment. Power tillers were then introduced and more sophisticated.equipment at a later stake when the need became apparent. Training and demonstration was emphasized at all stages of implementation. Farmers did not appear to be overwhelmed by the size or complexity of the machines they were taught to adopt. (iv) Divisibility or Economy of Scale The case has often been made that equipment should be simplified and made economical (and affordable) for the individual farmer. Experiences of cooperatively-owned assets have often been unhappy because shared responsibility is generally diffuse. In the example of joint farming groups, the equipment is individually-owned and operated but put to work for the group. Rather than compromise on the size and efficiency of particular equipment, machinery can be -52- developed to optimum specifications.in different size categories and the adjustment takes place in terms of number of such units that are needed to work a give area efficiently. (v) Profitability Farmers have to be convinced of the profitability of an investment before they will undertake it. It may be argued that government's subsidy on selected machinery distorts the real costs of the - investment. However, these subsidies are carefully managed and apply for limited duration. In any case, the farmer-owner bases his investment decision on private costs. Despite declining subsidies, machinery sales have accelerated as the innovations take root after the introduction period. Farmers have found it economical to own equipment for their own use or in the case of larger items, found it profitable to hire out their services. Also, fairly predictable yields (contingent on adherence to prescribed technology packages) and known marketing outlets help individual farmers assess returns to their investments. (vi) Reliability Extension, field demonstrations, field trips organized for farmers by FAs, and the known successes of innovative counterparts have been convincing enough to stimulate sales of farm machinery and equipment. Availability of spares and service workshops in convenient locations instill confidence in the purchase of known trade names. Proximity to Japan and frequent trips arranged for FA' leaders and outstanding farmers testify to the results that could be attainable with high levels of mechanization. Farmers are confident in the reliability of machines and promise of worthwhile returns to their investment and application. -53- 5.04 The above factors helked to foster rapid mechanization of agriculture in Taiwan. The sustained impact of mechanization is enabled by direct, individual ownership of production assets in such a way that farming is run as a business operation, and not just as a traditional occupation. Farm investments have to be seen to be profitable since opportunities for non-agricultural employment present very real opportunity costs for those engaged in agriculture. Ownership of land and capital allows for tradeability of assets. Good accessibility to resources, information, training, credit and other facilities allow mobility between occupations. Private efforts at farm modernization have been found successful and profitable. Government programs have wisely not attempted to shoulder accumulating burdens of subsidy beyond the desired tihnovation - stages. The careful balance of government encouragement and facilitating measures and the ultimate test of private profitability in a relatively unencumbered market situation is probably the best prescription for an approach to agricultural modernization. -54- ,BIBLIOGRAPHY Chen, Hsi-huang, "The Farm Labor Force of Taiwan: Problems and Prospects", Rural Economics Division, Joint Commission on Rural Reconstruction. Chen, Hsi-huang, "Impact of Land Reform on Economy". Hong, Pi-feng, "A Strategy on Food Production through Organized Farming. Groups in Integrated Rural Cooperative Development", Dept. of Farmers' Services, Council for Agricultural Planning and Development, Taiwan ROC., June 1979. Kuo, Wanyong, "Effects of Land Reform, Agricultural Pricing Policy and Economic Growth on Multiple-Crop Diversification in Taiwan", Graduate Institute of Economics, National Taiwan University, Taiwan ROC., November 1973. Lee, T.H., Chen H.H., Chen Y.E., "Labor Absorption in Taiwan Agriculture", in Labor Absorption in Agriculture - The East Asian Experience, Asian Regional Team for Employment.Promotion, International Labor Office, Bangkok, July 1980. Peng, Tien-Song, "Taiwan Agriculture and its Mechanization", Taiwan Agricultural Machinery Guide, 1981, Taipei, Taiwan ROC., February 1981. Taiwan ROC., Joint Commission on Rural Reconstruction, Agricultural Credit Division, "Agricultural Credit Program and Operation in Taiwan, Republic of China". June 1978. Thorbecke, E. & "Technological change and Growth Performance & Karunasekera.J in Taiwan Agriculture, 1946-1975", Working Paper No. 176, Department of Economics, College of Arts and Sciences, Cornell University. Wu, Torng-Chuang, "Agricultural Development Planning in Taiwan, Republiz of China", Council for Agricultural Planning and Development, Taiwan. -55- DISCUSSANTS Discussions were held with officials of: (1) Association of Agricultural Machinery Manufacturers (2) Council for Agricultural Planning and Development CAPD, 37, Nanttai Road, Taipei (3) Hsin Wu Farmers' Association (4) Industrial Development Bureau 109 Hankow Street, Taipei