The Chinese government has been using strong fiscal stimuli to encourage investment. While these fiscal policies, such as investment tax credits, often encourage firm investment, we find that investment tax incentives may generate an unintended reduction of firms’ innovation. Moreover, the crowding-out effect is non-monotonic in the level of financial constraints.

This table provides evidence on the effect of the 2004 VAT reform on the total number of patents applied for by a firm in year t+1 and eventually successfully granted [ln(1+Patt+1)]. NE is an indicator equal to one for the three provinces in the northeastern part of China and zero otherwise. Eind is an indicator equal to one for those six industries targeted by the reform and zero otherwise. Post2004 is also an indicator, which is equal to one for the years 2005–2007 and zero for years 2001–2004. Column (1) estimates the basic impact of the 2004 VAT reform on firm innovation. Column (2) controls for a set of firm-level time-variant variables including the logarithm of total assets, firm’s age, the return on the assets ratio defined as operating income divided by the book value of total assets, the proportion of capital owned by the foreign sector, and the proportion of capital owned by the state. The average values of these controls in Columns (3) and (4) are calculated across all firms in the pre-reform period. Column (3) interacts the Post2004 dummy with the pre-reform average controls. Column (4) fully controls for the interactions between year dummies and the pre-reform average controls. Standard errors clustered at the firm-level are in parentheses.

We investigate the heterogeneous effects of the VAT reform in terms of firms’ financial constraints. We follow Hadlock and Pierce (2010) to construct the size-and-age (SA) index to measure the extent to which each firm is financially constrained. A firm with a higher SA index is financially more constrained. Although originally estimated based on data for U.S. firms, it has been argued that the SA index is a good proxy for firms’ level of financial constraints in various contexts [see Hadlock and Pierce (2010) and Berkowitz, Lin, and Ma (2015)]. Previous research also confirms that both smaller firms and younger firms face tighter financial constraints in China, suggesting that the SA index can continue to be a reasonable proxy in China [see Cull and Xu (2005) and Liu and Mao (2019)]. In addition, we verify that our findings are robust under alternative proxies such as firm size and the cash flow ratio, which have been used in Liu and Mao (2019).

This table presents the heterogeneous effects of the 2004 VAT reform on the total number of patents applied for by a firm in year t+1 that were eventually successfully granted [ln(1+Patt+1)] by the financial constraint index (SA index). The average SA financial constraint index (Hadlock and Pierce, 2010) measured in the pre-reform years are used to divide our sample into deciles and define the top three deciles as tight, the middle four deciles as intermediate, and the bottom three deciles as loose financial constraints. Intermediate and Loose are two dummy variables based on the SA index to identify the groups of firms within the intermediate level (in the middle four deciles) and the loose level (in the bottom three deciles) of financial constraints, respectively. Robust standard errors clustered at the firm-level are reported in parentheses.

Finally, we show that the non-monotonic effects of the reform are not limited to innovation — firms’ labor costs (defined as the log value of total wages) are also non-monotonically (in the level of financial constraints) affected by the VAT reform as predicted by the model. In addition, we find that the effect of the reform on fixed asset expenditure is insignificant, consistent with our finding that the effect of the reform on fixed asset expenditure depends on the parameters of firms’ production functions. Specifically, as a firm upgrades its technology, the crowding out effect on fixed asset expenditure due to the additional cost associated with advanced technologies and the increase in firm’s capital intensity coexist; these two forces act in opposite directions.

This table provides evidence on the heterogeneous effects of the 2004 VAT reform on firm-level labor costs and the fixed investment rate in terms of financial constraints proxied by the SA index. The dependent variables are the log value of total wage (in 1998 Renminbi) in Columns (1)–(3) and the gross fixed asset investment in a given year normalized by the lagged net fixed asset stock are in Columns (4)–(6). The three subsamples with different levels of financial constraints are constructed as in Table 2. Robust standard errors clustered at the firm-level are reported in parentheses.

Our research (Ke et al., 2019) suggests that investment tax credits may have unintended negative effects on firms’ innovation, thereby deepening our understanding about the role of financial constraints in this type of policy. Our findings on the crowding-out effects of investment tax credits on innovation are relevant for policymakers for designing optimal fiscal policies to improve innovation and boost economic growth. This is especially important for China — even with rapid economic growth, China has long been criticized for insufficient innovation. Promoting healthy economic growth motivated by innovation is crucial for long-term stable economic growth.

Note 2: Specifically, , and x represents the interaction among , , and . When we increase x from zero to one, dy = (1+y)ln(1+y). The change in the number of patents (dy) from its mean value (0.0944) during 2002–2008 is then equal to 0.0082×(1+0.0944) = 0.0090, which accounts for 9.51 percent of the mean value of the number of patents.

(Shaowei Ke, assistant professor of economics at the University of Michigan; Yao Lu, associate professor in finance in the School of Economics and Management at Tsinghua University; Xinzheng Shi, associate professor in economics in the School of Economics and Management at Tsinghua University; Yeqing Zhang, PhD candidate in finance in the School of Economics and Management at Tsinghua University)

References