Regional economic integration would force domestic crops to compete with imported commodities. Nonetheless, increased foreign direct investment would provide more non-farm job opportunities and promote labour migration. While this stimulates structural reform in the farm economy by shifting from traditional crops to more competitive varieties on the international market, more non-farm job opportunities would ensure farmers rely less on agriculture. In fact, reliance on non-farm work will be deepened in an area with more job opportunities, a larger population, less agricultural land, and the complexity of external extension such as on the island of Java in Indonesia. Rising non-farm income will contribute to increased farm household income and provide labour to labour intensive industries. On the other hand, it could decrease the productivity, profitability and competitiveness of agriculture as a result of delays to technological innovation and farmland consolidation. The major objective of this study is to investigate how the expansion of off-farm job opportunities affects the farm development strategy of individual farm households.
Study area and methodology
'J' village in Majalengka Regency, West Java Province was selected as the study area. The study focused on two hamlets where, previously, Yokoyama (1995) conducted rural surveys from 1989 to 1990. In 1989 to 1990, 76 per cent of total households (297 households) in the two hamlets were farm households (227 households). Of the total farm households, more than half (126) received a part of their income from non-agricultural work. This implies that even during the previous study period, non-agricultural jobs had a significant importance in the household economies. The average operational land area of the surveyed households was 0.38 ha, which was less than the average operational land area of Java (0.47 ha, as of 1993). The previous study also reported that although no significant non-agricultural industries existed in the study village, many villagers engaged in non-agricultural jobs in the surrounding areas, such as at brickyards or seasonal migrant work in large cities during the dry season. Therefore, it was considered rational to select the same village to study in order to identify how the recent economic and social changes had affected the non-agricultural jobs and farm household economies.
'J' village is located in a hilly area approximately 250 km from Jakarta. The altitude of the village area is around 50- 127m. According to census data (BPS, 2005), the total area of 'J' village is around 299 ha including 175 ha of paddy land, 60 ha of upland fields, 62 ha of residential area, and 2 ha of forest and other land use. The population totals 2,013 and the number of households is 609. The main crops are rice and vegetables. For the study 167 farm households in the two hamlets (368 households) in the village were interviewed during the survey period (October - November 2007). Data was collected pertaining to: 1) household income; 2) measurements for an increase in income; 3) the importance of household expenditure; 4) measurements for farm economic development; and 5) future plan of rice production. As for 2), 3) and 4), the respondents were requested to evaluate the importance of the options by selecting their answers from 'very important', 'important', 'medium', 'not very important' and 'not important at all', which were converted into scores of +2, +1, 0, -1 and -2 respectively.
Results and discussions
Present situations of the hamlets
Farming households (167 households) made up just 45 per cent of the total households (368 households) in the two surveyed hamlets. This means that non-agricultural jobs play a more significant role in the village than in the previous study period. The preponderance of non-farm householders was day workers on construction sites as well as small-scale traders. The majority of non-agricultural jobs undertaken by the farming householders were seasonal migrant workers on construction sites in Jakarta; or Bandung, the West Java provincial capital; during the dry season. The variety of non-agricultural jobs remained almost the same as the previous study conducted by Yokoyama (1995). It was quite remarkable that around 20 migrant workers were found to work in the Middle East and Malaysia in the present study. Most farmers planted rice in their paddy fields once a year in the rainy season and planted upland crops such as soybean in the dry season. Mixed cropping of several varieties of vegetables (cucumber, eggplant, snap bean, chilli, etc.) was a common practice in upland fields. The current agricultural situation was almost same as the previous study by Yokoyama (1995). It is noteworthy that more paddy fields are now fallow during the dry season than in 1989-1990.
Household income structure
Average annual household income of all surveyed households (167 households) was 2,554 thousand rupiah per capita (1 US$ = 9,111 rupiah as of October 2007), slightly higher than the poverty line in Indonesia (approximately 2,000 thousand rupiah, as of 2007) but 65 per cent of the surveyed households earned less than the poverty line (Table 1). Non-agricultural income such as daily construction work in urban areas and small-scale trading formed 69 per cent of total household income, which implies that non-agricultural work plays an important role in the household economy (Table 2). Though rice is the major crop in the village, rice income forms only 9 per cent of total household income. Vegetables are recognized as high-value crops in Indonesia and contribute to farm income. However, income from vegetable production forms just 12 per cent, slightly higher than rice income.
Table 1 Annual per capita income of the respondents
Source: Calculated from interview survey results
Table 2 Source of household income
Measurements of income increase
Regarding their current income, 80 households, nearly half of the respondents, answered 'very unsatisfied' or 'unsatisfied', which outnumbered the respondents who answered 'very satisfied' or 'satisfied' (Table 3). In the survey, respondents were requested to evaluate the possible measurements of an increase in income (rice production, vegetable production, other crop production, hired farm labour, animal production and aquaculture, and non-agricultural work) by scoring the respective options as 'very important' (+2 points), 'important' (+1), 'neither important nor not important' (0), 'not very important' (-1), 'not important at all' (-2). Rice production showed the highest average scores of all the respondents, followed by vegetable production and non-agricultural work. The average score of the other options was less than 1 (Table 4).
Using an interval scale, Pearson's correlation coefficient between the scores of the respective options and the share of rice income, vegetable income and non-agricultural income in the total household income was calculated. It showed a positive significant correlation between 'rice income share' and 'rice production', 'vegetable income share' and 'vegetable production', and 'non-agricultural income share' and 'non-agricultural work'. The results suggested that the respondents hoped to develop their major jobs from which they earned most of their household income.
Priority of household expenditure
The average score of the evaluation results on the priority of household expenditure by the respondents was calculated. 'Investment in agriculture' showed the highest score, followed by 'Food', 'Medical expenses', 'Education' and 'Investment in non-agricultural work' (Table 5). This suggests that in spite of its minor share in household income, respondents still believe that investment in agriculture is important. The correlation coefficient between per capita household income and the scores of the respective expenditures showed that 'Durable goods' and 'Investment in non-agricultural work' had positive significant correlations with per capita household income. Considering the situation that non-agricultural income formed a major share of household income, it is suggested that respondents prefer the expenditure for durable goods such as motorcycles and electrical appliances, and investment in non-agricultural work rather than for investment in agriculture. This, in turn, would result in more reliance on non-agricultural income.
Measurements of farm economic improvement
The average score of the evaluation results on the importance of measurements to improve the farm economy was calculated. Technological improvements such as 'increased irrigation water usage', 'adopting improved varieties' and 'increased fertilizer/pesticide use' showed higher scores, more than 1 (Table 6). This suggests that the respondents were willing to adopt new technologies. On the other hand, changing cropping patterns such as 'convert from rice production to other crops' and 'convert other crops to rice' showed negative scores. Some of the respondents cited water availability and physiological conditions such as soil characteristics as reasons for the difficulty in changing cropping pattern.
The options for production area expansion received lower scores than options for technological improvements. This reflects the agricultural conditions of the study village, which has limited land area and limited area expansion. In the hamlets surveyed, several farmers who had sold their land in the previous five years were noted. Land leased between residents was also a common practice in the study area. However, the price of farm land (14 to 47 million rupiah per hectare) and land lease fee (1.0 to 2.5 million rupiah per hectare per year), which was very expensive compared to household income, were suggested as constraints to production area expansion. Rice area expansion received a higher score than vegetable area expansion. Comparing the same crops, land purchase received a higher score than land leasing as a measurement of area expansion. Several respondents commented that they would prefer to purchase land rather than to lease land due to its value as an asset. However, it was usually difficult for farmers to accrue enough capital to purchase land.
Farmers' attitude to rice production
A total of 104 out of 106 farm households that produce rice confirmed they would continue rice production even if they could earn sufficient income to purchase rice for self consumption. The number of respondents who suggested 'To individually secure rice for self consumption' (80 respondents) as a reason of their intention to continue rice production exceeded those who suggested 'To sell the produce' (65 respondents). Annual per capita rice consumption in Indonesia is decreasing year by year thanks to steady economic growth. However, it remains at 113 kg per capita per year (milled rice, calculated by the authors based on FAOSTAT data) and the procurement of rice is still a critical issue for most small-scale farmers. On the other hand, the producer price of rice occasionally plummeted due to cheap imports and oversupply around harvest time. It is unreasonable for farmers to expect high profits from rice sales. However, for low-income households, rice self sufficiency remains a low risk option to alleviate poverty. This could be the rationale behind the respondents' preference to persevere with rice production.
The impact of the farm economic structure on farmers' attitudes to development strategies
The respondents were classified into four groups according to differences in their attitude to farm development strategies.
Table 7 shows the results of a Mann-Whitney test, which analysed the differences among the groups regarding their preference to certain farm development strategies (except cropping pattern changes, which showed negative average scores). Group 1 is the farmers who rely on neither rice nor vegetable production as a source of household income. A significant difference was observed between Group 1 and the other three groups concerning the preference of 'rice area expansion by land purchase'. The score of Group 1 for this option was significantly higher than the other groups. The reason for this was the strong preference for rice self sufficiency, which was indicated in the previous section, and the availability of capital for land purchases thanks to higher income from non-agricultural work. On the other hand, for technological improvements such as 'Increased fertilizer/pesticide use' and 'Tractor use', the other groups, which rely more on agriculture, showed significantly higher scores than Group 1. This result suggests that the positive attitude of Group 1 to rice area expansion might not coincide with productivity improvements. It was assumed that Group 1 wanted to hold farm land as an asset rather than a means of rice production.
A significant difference was observed between Group 2, which relies on vegetable production, and Group 3, which relies on rice production, as a source of income, regarding 'other crop expansion through land leasing' and 'other crop expansion by land purchase'. Group 2 showed significantly higher scores for both options. This suggested that vegetable producers, which usually can earn more income than rice producers, have greater intention for area expansion. No significant difference was observed in terms of technological improvements, except for 'Increased irrigation water use'.
The numbers in the table above show the difference in average score was significant at the 5 per cent level and the average score of the group in the left-hand column is more than that of the group in the top row.
Though the farmers in the study village rely heavily on non-agricultural jobs as a source of household income, the survey results showed that most of the farmers who produce rice and vegetables, major crops in the area, still hoped to develop their agriculture and increase agricultural income. These farmers have a keen interest in technological improvements. On the other hand, it seemed difficult to diversify the current cropping pattern drastically due to land and water constraints. Therefore, it would be useful for farmers to introduce new technologies that enable yield increases and crop quality improvement while maintaining the current cropping patterns in order to improve their farm economic conditions. Farmers who rely heavily on non-agricultural income are interested in expanding rice production. However, their interest stemmed from the value of farmland as an asset. At the time of survey, no difference was observed between rice farmers and vegetable farmers regarding their intent for technological improvements except water use.
Further globalization will reduce the profitability of rice production in Indonesia, which is currently protected by various support schemes. On the other hand, if Indonesian farmers could succeed in producing high quality vegetables, which can meet the demand of international consumers, farmers would increase income through exports. The major vegetables produced in the study villages are tropical vegetables such as cucumber, eggplant, long bean, chilli, etc. It is difficult to produce temperate vegetables such as cabbage and potato, which are the major export vegetables from Indonesia. Eggplant is the third most exported vegetable after cabbage and potato. Eggplant is a popular vegetable in the study village but due to the low quality and uncompetitive price, all produce is traded on the domestic market. Conversely, the Indonesian Government has redoubled its efforts to produce exportable vegetables such as cabbage, which can be cropped in lowland areas (author's personnel communication from a researcher in the Vegetable Research Institute, Ministry of Agriculture, Government of Indonesia). It would be possible for the study villages to produce exportable vegetables through the adaptation of new varieties, quality improvements and lower production costs. It is important for policy planners to consider the farmers' willingness to develop their current farming practices as well as the current global agricultural situation and then implement policies that contribute to rural development using globalization as a precious opportunity.