The accuracy of the information presented in this article is the sole responsibility of the authors. Various policy instruments have supported these projects in efforts to promote the growth of the agricultural sector and the economy. Sevele (1983) argued that motivation is one of the most important social constraints to agricultural development in Tonga.
Developments that ignore them are less likely to be successful and sustainable, both from the point of view of the planners and the beneficiaries. The head of the household was therefore chosen as the interviewee on behalf of the household. One of the major concerns in the analysis of multi-objective decision-making problems is the presence of incompatible or conflicting multiple objectives.
Land Tenure of Sample Households
Paired comparisons and rating scales yielded similar rankings for the objectives in both studies, with the former being easier to administer, as respondents are only asked which of two alternatives they prefer. The lower bound 0 indicates the lack of importance of the objective while the upper bound refers to the maximum possible importance value. A large proportion of the total number of tax allotments (65 percent) and agricultural tax allotments (46 percent) are located on the main island of Tongatapu.
An important feature of the surveyed households was that almost all obtained income from more than one source during the period considered.
Variation in income among villages
Row A is the percentage of households that receive income from different sources; Row B is the total cash received (T$) from each source; C is the percentage of total income from that source. LSD values p < 0.005).
Household Cash Expenditure
Goal Levels
Minimise risks
The ultimate goal is to make enough time for leisure and is therefore included in the Goal Programming model. The demand for leisure was treated in much the same way as the demand for basic food and cash consumption, and included it in the target constraints. The amount of free time (for 2 units of work) is limited by the amount of productive work to be done and the total time available for the division between farm work and free time.
The objective in limiting the goal is to minimize the lack of fulfillment of monthly leisure requirements.
The Combined Objective Function
Validation
Policies for Development
Relatively few agricultural farms remain to be distributed so that unless the government changes policy, future population growth will produce a marked increase in the number of landless households. Important issues that policymakers need to address include what constitutes a "sufficient" size of ownership and how to ensure efficiency and equity in land redistribution. The goal programming model was used to explore the effects of change in the amount of land available to Tongan households.
The significant changes also occurred in the farm plan and the squash area was mostly affected with 89 percent reduction at 1.67 ha and was totally eliminated at the allotment size of 1.11 ha. As shown in Table 16, the reduction of available arable land was associated with a reduction of fallow land and increasing use of fertilizers to improve soil fertility. Because improvements in input and output marketing affect prices, it is appropriate to evaluate them in the model.
The effect of changes in the price of output for squash and yam on the farm's operating plan showed significant variations in the responses. However, a 40 percent decrease in the price of export squash (market 2) will cause only a 3 percent reduction in the area of squash, but an increase in yam 2 of about 12 percent, and an increase in total cropped area of 0, 4 percent. In comparison, markets 3 and 4 showed that changes in the price of yam have a more significant impact on the three main farm types.
As shown in Tables 18 and 19, an increase in the price of yam to T$1.90/kg would generate an increased cash surplus of about 17 percent (progressive), 39 percent (emergent), and at least 50 percent for marginal farm households. The result shows that improved market export opportunities for yams and cassava (market 5) would dramatically increase cash surpluses at the end of April by more than 140 percent for new farmers in the three main groups, with the highest increase of more than 200 percent in Ha'apai the group. The variation between the archipelagos is attributed to the variation in the initial market constraints in each base model.
The basic models assume that the technologies are constant in the three regions and farm types.
Effects of Labour Policies
The results also show effects where farmers place a higher value on leisure time or off-farm employment compared to time spent on agricultural activities. In Tonga, most farming households have one or more adults who work off the farm at least part of the year. Such off-farm income can be an important factor in family well-being.
Off-farm income, which includes paid employment, remittances and business income, constitutes a significant portion of household cash income. Off-farm income is important not only for reducing sole dependence on subsistence agriculture, but also for providing capital to finance farm improvement activities. It is therefore important to explore the impacts of off-farm income levels on goal attainment and reinvestment in agriculture for different types of farms.
The impact on farm plans for the three main farm types in Tongatapu for different levels of off-farm income was investigated and is summarized in Table 24. In contrast to the marginal farm (Farm3), some of the cash requirements are for necessities, church. and social obligations were not fulfilled. The responses of these marginal farmers to off-farm income imply what other authors (Hau'ofa and Ward, 1980; Sisifa et al., 1993) have noted in that they have "target household income".
This means that since the family's labor supply exceeds the current needs of the farm, there is an opportunity for off-farm employment. Off-farm income provides farmers with a more secure income, allowing them to invest more in profitable agricultural ventures.
Effects of Credit policies
Modelling the significance of church and social obligations
In contrast, in Church 2, where there is more emphasis on surplus cash (0.1) and less emphasis on church liabilities (0.0001), there was a smaller effect on total crop area but an increased cash surplus of 8 percent. These differences are attributed to the variation in income levels between the three groups and the variation in the weights for each farm type in the initial models. More emphasis on the achievement of social goals showed differences between production plans for the three farm types, with decreases in excess cash by 44 percent (marginal), 41 percent (emergent) and only 7 percent in progressive.
The data shows that increasing the weights for social and church obligations from 0.0001 to 0.001 would reduce excess cash by 15 percent and 45 percent, respectively. As noted in the model development phase, the three business types have different leisure needs. The effects of varying the weights of the leisure goals relative to the excess money show that there is more weight on the.
The opposite effects occur when the weighting is reversed, as indicated in the Leisure 2 column of the table. Under this scenario, progressive farms would reduce root crop production in favor of growing more pumpkins, contributing to a 5% increase in crops and a 15% increase in surplus cash. Marginal farmers with the goal of maximizing profits would not only increase their cultivated area by 15 percent, but also have a threefold increase in their surplus cash.
Variability in the responses expressed by changes in production plans and excess cash production is observed among the three main types of farms where behavioral assumptions differ. Progressive farmers who give more priority to maximizing surplus money in favor of church, social and leisure were less affected compared to the other two types of farms.
Aggregation for macro-policy consideration
Under the assumption that the representative farm "sufficiently" represents other holdings, the eight representative models developed in this study were used to evaluate the results of a proposed development program. This was done by multiplying each representative farm plan and its associated financial results by the number of holdings it represents to obtain a regional aggregate of production, activity levels and financial data (eg surplus cash at the end of the year). The project benefits were estimated for each island group based on information from the 1993 Land Use Crop Survey and the categorization into the 3 main farm types.
If the project was implemented correctly, it can generate an additional cash surplus of more than 3 million. 34;-" means that the model did not select the activity. Cropped area values are rounded to 2 decimal places.
Summary and Conclusions
These self-managing farmer organizations could eventually take the form of cooperatives, depending on the key limiting factors in providing these different inputs and services in the local village community. Increasing export production from the agricultural sector is considered the key factor for the growth of the Tongan economy. The increasing commercialization of semi-subsistence farms is a key requirement for increasing production with limited land area.
In Tonga where markets for land and capital are not well developed, pricing instruments appear to have limited influence on resource allocation. Since a large part of the consumers of agricultural products are within the agricultural sector, it will be difficult to increase the domestic market quickly. Increasing the export of agricultural products will therefore be an important strategy for expanding the market for the subsistence farmer.
However, the sustainable success of the agricultural sector depends on how long niche markets for these crops remain open and profitable for Tongan farmers, and on the success of efforts to diversify agricultural products and export to other markets. It also requires improving the capacity of semi-subsistence farmers to be able to manage the uncertainties and risks of increasing commercialization. To introduce new products and new inputs, it is necessary to improve the dissemination of new knowledge through diffusion.
Exploring the relationship between farmers' goals and the phenomenon of pasture degradation in the beef production systems of central Brazil. Smallholder Agriculture in Tonga: Report of the South Pacific Smallholder Project in Tonga 1984-85, South Pacific Smallholder Project, University of New England.
