Tax incentives for innovation in the context of macroeconomic stability: an analysis of causality

Abstract

The article deals with the topical issue of R&D tax incentives and their impact on the level of innovation development and macroeconomic stability. The research is based on causality analysis and estimation of the strength, time lags and directions of mutual influence of R&D tax incentives and macro indicators. Systematization literary sources and approaches for solving this problem indicates that R&D tax incentives are studied in fragments in the context of macroeconomic stability. The research's main purpose is to improve the methodological bases of substantiation of the choice of relevant instruments of innovation stimulation considering causal relations of R&D tax incentives and macro indicators. The paper presents the results of dynamic analysis of R&D tax incentives in 13 European countries, for which OECD statistics for 2007-2017 are freely available. The significance, strength, and nature of the relationship between these indicators and the following macro indicators are determined: the level of the country's innovation development, the share of investment in GDP (in general and in the corporate sector in particular), net international investment position, the share of the business sector in the cost structure of R&D. Pearson and Spearman correlation coefficients were calculated depending on the variable subordination to the law of normal distribution (verified by the Shapiro – Wilk test) on the admissible calculation interval taking into account time lags from 0 years to 3 years. The causality of the studied indicators was established using the Granger causality test. The calculations are important for the prioritization of instruments for the implementation of innovation support. The highest priority should be given to the establishment of tax incentives for R&D, as this tool's impact on all studied macro indicators in most countries was direct. Its effect was manifested in the shortest possible time (with a lag of 0-3 years). The second priority should be given to setting hidden rates of business tax subsidies on R&D, as this indicator's impact on most of the studied indicators was statistically significant and direct with a time lag of 0–3 years. The paper substantiates the inefficiency of direct public financial support, as the impact of this indicator on most of the analyzed macro-indicators was reversed with a lag of 0–2 years. Thus, it is more expedient for the state to help entrepreneurs by providing tax benefits to provide innovation development and macro stability than through direct reimbursement of costs. Moreover, lag regression models were built for those countries where identified links were the most important (Belgium, Denmark, the Netherlands, and the Czech Republic). They take into account inflation rates and interest rates on long-term liabilities and the number of labour resources in the country as control variables.