For several years, starting in about 2004, the Thailand Village Fund was, by any standard, the largest single microcredit scheme in the world. It was put in place very quickly, and to this day it remains the single most important source of credit for poor Thai household Given its size, it is important to ask whether the Village Fund has had a measurable impact on spending and income.
We are able to address this issue using data from the large nationally representative Socioeconomic Surveys of 2002 and 2004, undertaken just one and three years after the VF was launched. In the absence of random assignment, we were obliged to use quasi-experimental methods to quantify the effect of the VF on outcome variables. Both the cross-sectional and rural panel data estimates produce highly defensible results: the Village Fund does appear to have had an impact, raising current expenditures by a statistically significant 3.5%, and income by 1.4%, and durable goods ownership by 3–5% points, in the early years of the Fund's operation. The results of the propensity score matching model are very similar to those based on the rural panel, which suggests that our estimates are quite robust.
Our interpretation of these findings is that the VF has indeed had a moderate impact on household spending, and also (but to a lesser extent) on household income. We thus find a more modest effect than that observed bywho find that household spending in 2003 increased by more than the amounts borrowed from the VF. Although their methodological approach is similar, their sample is relatively small, and limited to four provinces in central and northeastern Thailand, which may explain the difference in the findings.
The small effect of microcredit on incomes is not what the designers of the Fund had envisaged; instead they had expected that the VF would boost household-level income from non-traditional sources. It appears that the short-term nature of the VF loans makes them suitable for farmers – they allow for the financing of inputs during a crop cycle – and for consumer credit, but are not sufficiently long-term (or perhaps large) to be very useful for most of the other remunerative activities that households might initiate.
Our findings differ from those of who apply propensity score matching on a region-by-region basis to the households in the (rural) panel component of the 2004 Socioeconomic Survey, and also apply double differences to this group. They “find that the VF program does not have a positive impact on alleviating the country's poverty. The lack of such an effect on poverty is the product of its insignificant impacts on income and expenditure” Our propensity score equation includes more variables than does the Chandoevwit and Ashakul model, and uses all the 2004 data, not just the (rural) panel component. And our fixed-effects panel estimators are more complete than the double differences that they use, in that we control for a wide variety of household variables. It does appear, however, that the results of the impact evaluations of the VF are sensitive to the choice of model.
Our third interesting finding is that there are complementarities between VF and BAAC loans; while VF borrowing influences consumption, BAAC loans help drive income. Loans from these two sources have different specializations. This has an important practical implication: The BAAC should be slow to withdraw from village-level lending, even if it is tempted to do so by a perception that the VF can fill the gap.
Our other important finding is that VF lending appears to be pro-poor. It is well known that the way that the VF is targeted – disproportionately to small villages, which tend to be less affluent – in effect helps the poor. What is less obvious is our finding that the impact of a given VF loan on spending and income is stronger for poor than for rich borrowers. These two effects help explain the enduring popularity of the Village Fund in much of rural Thailand, and the goodwill it created in such areas toward the Thaksin government and its descendants (including the Yingluck government elected in July 2011).
Our results do not allow one to make a judgment about the desirability of the VF, which would require additional information about the full costs of the program, and an evaluation of its sustainability. We note, however, that the proportion of Thai households that borrow from the VF has remained steady, at about 28% of the total, from 2007 through 2011, so the Village Fund has clearly endured using data from a survey of Village Funds undertaken in 2010, argue that the economic benefits, if they are of the same magnitude as those reported here, outweigh the economic costs. It would be valuable to determine whether the impact of the VF weakens over time, a finding that is common elsewhere This would be particularly desirable, given the importance of the Thai experiment with large-scale microcredit, and rising interest within Thailand ab