경제위기 이후 대기업의 연구개발투자에 대한 조세지원제도의 효과 분석

Tax incentives aim at achieving specific national economic objectives, one of which is to encourage firms to invest in R&D. This study examines the impact of changes in tax incentives, R&D tax credit and reserves for R&D, on the major firms' R&D investments after the Asian economic crisis. The tax incentives for major firms' R&D have diminished since the economic crisis because Korean tax authority thought that large companies had motivation and ability enough to make R&D investment without government's tax incentives. However, the tax policy for R&D was changed to give more incentives to large companies from 2008. In spite of a lot of money for tax incentives and the big change of tax policy, there is not a study that analyzes the effectiveness of tax incentives for R&D after the Asian economic crisis. There are four changes in tax incentives from 2000 to 2007. In 2000, major firms can choose the larger amount as tax credit between 5% of R&D expenses and 50% of R&D expenses which exceed the average expenses occurred during the immediately preceding four business years. However, only 50% of the excess of expenses incurred for R&D is applied as tax credit rate in 2001, so tax incentives for R&D is curtailed from 2000 to 2001. From 2003, R&D tax credit rate is decreased from 50% to 40%. In 2006, tax incentive is boosted contrary to the previous years. Tax credit rate on the R&D expenditures incurred to commission small and medium-sized enterprises or universities to conduct R&D increases from 40% to 50%. In 2007, reserves for R&D is abolished. In short, tax incentives are cut down in (2001∼2002), (2003∼2005), and (2007), but are expanded in (2006). This study hypothesizes that R&D expenditures of major firms would increase(decrease) as tax incentives for R&D expand(cut down). This study uses 1,512 firm-year observations collected from Korean stock market over 2000-2007 period with multiple regression analysis. Sample firms are only large companies which have experiences to invest in R&D activities. The sample of prior studies included medium-sized companies and large companies that invested no money into R&D, which might induce misleading results. To analyze the effect of changes in tax incentives for R&D on the firms' R&D expenditure, control variables such as R&D expenditure of the previous year, internal financing(measured by cash flow from operating), external financing(measured by net cash flow from financing), debt ratio, Tobin's q ratio, profitability(measured by return on asset), size, principal stakeholders' equity ratio, and industry dummy variables are incorporated into the research model. In this study, I find that the four changes in tax incentives from 2000 to 2007 don't have a significant predicted effect on the large firms' R&D investment. The coefficients of and are negative as predicted, but not significant. The coefficient of is also positive as predicted, but not significant. The coefficient of is statistically significant, but its sign is different from the predicted. R&D investment during 2003-2005 period significantly increases relative to 2001-2002 period in spite of the 10% cut in tax credit rate for R&D of 2003. These results imply that the expansion (curtailment) in tax incentives for R&D doesn't increase(decrease) the R&D spending of major firms after the Asian economic crisis. To investigate the unexpected results further, I examine if the tax burden of an individual firm affects the relation between the changes in tax incentives and R&D investment. I classify firms into a low and a high tax burden group based on the effective tax rate. Expansion in tax incentives significantly increases the R&D investment in the high tax burden group. In the low tax burden group, however, expansion(cut) in tax incentives doesn't increase(decrease) the R&D investment. Instead, cut in tax incentives of 2003 leads to the increase of R&D expenditure in the low tax burden group. These results imply that the firms in the low tax burden group don't care about the additional tax burden due to the reduction in tax incentives and keep investing in R&D because their existing tax burden is sufficiently low. In addition, I reinvestigate the study of Lee et al.(2000) which analyzes the effect of changes in tax incentives on the firms' R&D spendings from 1987 to 1993. Their sample includes not only large companies that don't invest in R&D at all but also medium-sized firms. I reinvestigate their study using only large companies that have experiences to invest in R&D from 1992 to 1994. According to the result, the change of 1993 that cuts down the tax credit rate of R&D expenses from 10% to 5% and raises the tax credit rate of the excess R&D expenses from 10% to 25% doesn't increase significantly R&D spendings, which is not consistent with the result of Lee et al.(2000). However, the change of 1994 that raises the tax credit rate of the excess R&D expenses from 25% to 50% and does away with the use of the tax credit rate of R&D expenses significantly decrease R&D expenditures of major firms, which is consistent with the result of Lee et al.(2000) The result of this study that firms increase consistently their R&D investment even though the curtailments in tax incentives in 2000s provides tax authority with an empirical validity of its claim that majority firms have sufficient motivation and ability to invest in R&D without the government's incentives or supports. However, this study has some limitations. R&D spending of a firm may be affected by macro economic situation, so it is necessary to incorporate variables which reflect macro economic situation. Even though the research model of this study includes control variables such as profitability or q ratio, it may not be sufficient. In addition, there may be time gap between changes in tax policy and their fruits. Time gap is not taken into account in this study. However, firms can have time to adjust their R&D investment according to the changes of tax policy because discussions about changes in tax policy usually starts in the middle of the previous year. Thus, the effect of time gap on the result may be small.