Do Consumer Subsidies Increase Productivity? Evidence from Swedish Service Firms

Abstract: We suggest that "ROT-avdraget" has had a positive effect on industry productivity in the business sector affected by the policy and that this effect is divided into two statistically proven parts; the average firm is becoming more productive and the most productive firms gain market share. Industry and firm level productivity is estimated on a large panel data sample of firms using Levinsohn and Petrin's semi parametric model. By utilizing a natural experimental design where the companies affected by the subsidy is evaluated against a control group operating in a closely related business sector, we are able to perform several difference in difference regressions to test the relationship between a demand increasing ad valorem subsidy and productivity in the business sector providing these subsidized services. Testing for robustness using a second control group yields a positive point estimate of the true effect on productivity of the policy as well, but this is not statistically significant at a reasonable level. Finally we correct the standard errors for biases related to potential serial correlation in the dependent variable as well as possible heteroskedasticity and conclude that these corrections do not change our results.