This study aims to propose ex-post quantitative policy evaluation models. To suggest an alternative evaluation model and interpretation of the evaluation results, this study conducts an empirical analysis on three different public programs in a field ...
This study aims to propose ex-post quantitative policy evaluation models. To suggest an alternative evaluation model and interpretation of the evaluation results, this study conducts an empirical analysis on three different public programs in a field of rural and regional development. Even though previous studies have been investigated through versatile quantitative evaluation methods, they require sophisticated statistical data and highly professional knowledge for the result interpretation. On the other hand, this research tries to fill the academic and practical vacuum by supporting the ease of constructing the evaluation data and the utilization of the results from an a posteriori point of view.
The evaluation models that this study propose consider limited evaluation resources and data that the evaluators face in an actual evaluation practice. Therefore, the models utilize quasi-experimental and non-experimental evaluation design rather than true experimental design. Among the three empirical analyses included in this research, the first one applies the Heckman Selection Model to non-experimental design, which focuses on selection bias. On the other hand, the second and third ones are based on quasi-experimental evaluation design. Thus, they concentrate on statistical homogeneity between treatment and control groups, using a combination of the Binary Logit Model, the Extended Blinder-Oaxaca Decomposition Technique, and another integration between the Spatial Econometric Model and the Blinder-Oaxaca Decomposition Technique.
First of all, the first analysis evaluates the effects of the Sixth-order Industrialization of Agriculture on the increase of farm household income. This analysis uses the Heckman Selection Model as an evaluation method considering the policy stage and available data of the sixth-order industrialization, Especially in order to examine the effectiveness of the current sixth-order industrialization policy direction, the analysis uses a correction term of the selection bias as a concept of a counterfactual control group.
According to the evaluation results of the first essay, it seems that under the current agricultural and rural environment in Korea, the strategy of comprehensive combination has a negative impact on the increase of the farm household income. This is because relatively simple single-bonded sixth-order industrialization appears to be more effective in terms of income increase, in spite of the strategic emphasis on comprehensive sixth-order industrialization which combines 1x2x3 in the sub-dimension. The estimated negative correction term, lambda, provides significant evidence for this interpretation. In other words, when controlling the major explanatory variables affecting the conjunction of economic activities and the farm income, it is shown that the income is likely to be decreased by selecting the comprehensive combination instead of the single bond type.
The second empirical analysis estimates the economic impact of rural tourism on the farm household in the case of the Traditional Theme Village Project. The evaluation model consists of both the Binary Logit Model and the Extended Decomposition Technique. The application of the extended version enhances the statistical validity instead of ordinary Blinder-Oaxaca Decomposition Technique for logit analysis.
The second analysis positively evaluates the impact of the project on generating the off-farm income of farm households. Both the cross-sectional and the longitudinal results show the positive residual effect and negative endowment effect. The statistical significance of endowment coefficients also appears high. First, the positive residual effect can be interpreted to mean that the project contributes to the creation of the off-farm income in a participatory group, assuming the residual part after controlling other factors affecting off-farm income in the regression model as the program net-effect. However, the negative endowment effect suggests that, for the participatory group, the controlled characteristics in the regression model would have reduced the probability of generating off-farm income without policy intervention.
The last essay analyzes the impact of integrated rural development on the improvement of settlement conditions in rural areas. As with this program, the regional project that invests in ameliorating the settlement environment is more difficult to assign randomly to the treatment or the control group than the project at an individual and household level. It is not only because the number of available samples are limited, but also because the accumulated local characteristics in the long-term form the uniqueness of the region. The third analysis thus employs the spatial econometric model in conjunction with traditional Blinder-Oaxaca Decomposition Technique based on the linear regression result. In addition, this essay transforms the qualitative level of settlement into a quantitative evaluation indicator through multivariate analysis.
Comparing the participatory and non-participatory areas, the participated in the project show the positive effects of improving the rural settlement condition. The positive residual effects argue that the project positively influences the settlement environment in the participatory areas. On the other hand, the endowment effect represents a negative direction. This result implies that if the project is not implemented in the participating regions, the likelihood that the endowed resources in the regions will enhance the settlement conditions would be lower than those in the non-project regions. However, the results of the comparison of characteristics before and after the project in participatory areas are different from the comparison between the participant and non-participant areas above. For this reason, if it is confined to the changes of the participatory areas without cross-sectional comparison, the possibility of improving the settlement environment of the participant after the project can be lower than before. The cause of the difference can be attributable to both internal and external factors, such as lack of endogenous regional development capacity without governmental intervention and the changes of rural environment over time.
Policy evaluation is a modern way for citizens who live in representative democracies to search for their political identities. It is not the only way, but it is obviously one of the significant alternatives. The retrospective examination of the rationale of public investment is a primary means of materializing common values projected onto public policy. This study is the first step to demonstrate the policy effects by strengthening the interpretation of selection bias and applying the decomposition techniques in various ways. The alternative evaluation model and interpretation proposed by this paper are therefore expected to contribute to carrying out social values through policy for the long haul.