Behavioural additionality of China’s business-tax-to-VAT reform on firms’ internal and external R&D investment

ABSTRACT

This paper investigates the unintended effect of the business-tax-to-VAT (B2V) reform in China on manufacturing firms’ internal and external R&D behaviour. The B2V reform was implemented in the service sectors to replace business tax with value-added tax. It creates tax incentives for downstream manufacturing firms to increase external R&D outsourcing, as purchasing R&D and technology services becomes tax-deductible after the reform. Using a large administrative firm-level dataset, we found that downstream manufacturing firms that were previously more closely linked to upstream R&D and technology services are more affected by the B2V reform, and tend to increase external R&D outsourcing and reduce internal R&D expenditures. The unintended behavioural additionality is mainly driven by small and medium-sized firms and firms with low R&D capabilities which usually bear high costs of conducting internal R&D and rely more on external R&D services. Our study implies the possibility of leveraging tax incentive measures to encourage firms to seek external knowledge and technologies.

KEYWORDS:

• Business-tax-to-VAT reform
• internal R&D
• external R&D
• behavioural additionality

JEL CLASSIFICATION:

• O32
• O38
• H32
• H25

Acknowledgments

Xiaoyong Dai acknowledges the financial support from the National Natural Science Foundation of China (No. 71804140) and the Humanities and Social Science Foundation of the Ministry of Education of China [No. 18YJC790021]. Zhiqi Zhao acknowledges financial support from the National Natural Science Foundation of China (No. 72003123) and the Science and Technology Commission of Shanghai Municipality (No. 20692191900).

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ More specifically, transportation services include transportation by land (excluding railroad transport), waterways, air, and pipeline; technology services include technology transfers and licencing; cultural and creative services include not only design, advertising, and intellectual property-related services, but also the transfer of trademarks, copyright, and business goodwill services.

² The B2V reform was implemented in a gradual process across various regions in China. Our identification relies on exploiting the cross-industry, cross-time variations in tax incentives.

³ The reason that we do not include $Backward$ in the estimation is that the measure is industry-specific, while we have already controlled firm-specific fixed effects.

⁴ The measure was calculated for all manufacturing industries at the two-digit industry code level. According to the National Industry Classification of China, the sample covers 30 manufacturing industries. For the measure of backward linkage to R&D services, the mean average is 1.097 with a standard error of 0.693. Industries with high levels of backward linkage to R&D services include telecom machinery, instrumentation manufacturing, automobile industry, medicine, special purpose equipment, chemical products, and so on.

⁵ The reasons that we use the data during the period 2010–2015 is twofold. The first is that the dataset does not contain information regarding firms’ external R&D expenditures before 2010. The second is to isolate the impact of the VAT reform, which was initialized in 2004 and completed in 2009..

⁶ Firms that did not report either internal or external R&D expenditures during the entire sample period are excluded from our sample, as those R&D-inactive firms do not involve in R&D activities and thus will not be affected by the B2V reform in terms of R&D behaviour.