In this paper, scale efficiencies and relative efficiencies of R&D projects in Industrial Technology Program, sponsored by Ministry of Trade, Industry and Energy, Korea, are calculated and compared. For the process, various DEA (Data Envelopment Analysis) models are adopted as major techniques. For DEA, two stage input oriented models are utilized for calculating the efficiencies. Next, the calculated efficiencies are grouped according to their subprograms (Industrial Material, IT Fusion, Nano Fusion, Energy Resources, and Resources Technology) and recipient types (Public Enterprise, Large Enterprise, Medium Enterprise, Small Enterprise, Lab., Univ., and etc.) respectively. Then various subprograms and recipient types are compared in terms of scale efficiencies (CCR models) and relative efficiencies (BCC models). In addition, the correlation between the 1st stage relative efficiencies and the 2nd stage relative efficiencies is calculated, from which the causal relationship between them can be inferred. Statistical analysis shows that the amount of input, in general, should increase in order to be scale efficient (CCR models) regardless of the subprograms and recipient types, that the 1st and 2nd stage relative efficiencies are different in terms of the programs and recipient types (BCC models), and that there is no significant correlation between the 1st stage relative efficiencies and the 2nd stage relative efficiencies. However, the results should be used only as reference because the goal each and every subprogram has is different and the situation each and every recipient type faces is different. In addition, the causal link between the 1st stage relative efficiencies and the 2nd relative efficiencies is not considered, which, in turn, is the limitation of this paper.

When evaluating effectiveness of a program, there is a tendency to simply compare the performances of the treated before and after the program or to compare the differences in the performances of the treated and the untreated before-after the program. However, these ways of evaluating effectiveness have problems because they can’t account for environmental changes affecting the treated and/or effects coming from the differences between the treated and the untreated. Therefore, in this paper, panel data analysis (fixed effects model) is suggested as a means to overcome these problems and is utilized to evaluate the effectiveness of fusion technology program conducted by Ministry of Trade, Industry and Energy, Korea. As a result, it turns out that the program has definitely positive impacts on the beneficiary in terms of sales, R&D expenditure, and employment.