DeepDyve requires Javascript to function. Please enable Javascript on your browser to continue.
Tax incentives, adjustment costs, and R&D investment in China Tax incentives, adjustment costs, and R&D investment in China
Li, Wanfu; Du, Jing
2016-10-01 00:00:00
China Journal of aCC ounting StudieS , 2016 Vol . 4, no . 4, 433–455 http://dx.doi.org/10.1080/21697213.2016.1252088 Tax incentives, adjustment costs, and R&D investment in China* a,b c Wanfu Li and Jing Du a b Center for accounting, f inance and institutions, Sun Yat-sen university, guangzhou, China; School of a ccounting, nanjing university of f inance & economics, nanjing, China; Sun Yat-sen Business School, Sun Yat-sen university, guangzhou, China ABSTRACT KEYWORDS r&d investment; tax Tax incentives for firms’ research and development (R&D) activities incentives; adjustment costs have been widely used to solve the “market failure” problem and to increase firms’ R&D investment. However, there is no consensus on whether the incentive effects of R&D tax policies are effective. This study empirically analyses the moderating effect of the adjustment costs of R&D on the incentive effects of R&D tax policies in China. The results show that tax incentives policies stimulate firms’ R&D investment. However, the incentive effect of tax incentives weakens as adjustment costs increase. When the adjustment cost is greater than a critical value (0.012), the tax incentive effect of R&D disappears. About 93% of Chinese firms have adjustment costs lower than this critical value, which suggests that China’s R&D tax incentives policy is generally effective. This study also finds that the incentive effect of tax policy on R&D investment is more significant for non-state-owned firms than for state-owned firms. 1. Introduction In the current global environment, economic development and technological innovation are important drivers of economic growth in China. Research and development (R&D) is the most important factor in technological innovation. R&D activities not only produce new technologies, they also enhance the absorption and innovation capability of existing tech- nologies. Thus, a firm can increase and maintain its competitive advantage by promoting R&D. As An, Shi, and Ludovico (2006) point out, sustainable economic development in China relies not only on increasing manufacturing capability, but also on the ability of enterprises to innovate. As the basis for technological innovation, R&D should receive more attention. In fact, all developed countries and a growing number of developing countries have increased R&D investment in recent decades. However, due to R&D’s characteristics of strong externalities and public goods, many firms invest less than the socially optimal amount in R&D (Arrow, 1962). Thus, governments must encourage firms’ R&D behaviour. In general, governments use either direct financial CONTACT Jing d u dujing001@126.com *Paper accepted by Kangtao Ye. © 2016 a ccounting Society of China 434 W. LI AND J. DU subsidies or tax incentives to promote R&D investment. Both methods have advantages and disadvantages. Direct financial subsidies have a fixed budget and the amount invested can be determined in advance. However, some studies suggest that direct subsidies have a substitutional role in firms’ R&D investment (Lichtenberg, 1984, 1988b; Wallsten, 2000; Busom, 2000; Yao & Zhang, 2001). In contrast, tax incentives are driven by market forces, and have been widely adopted by governments. Many Western countries used tax incentives in the late 1970s and 1980s to stimulate R&D in firms. Since its reform and opening up period, China has introduced some R&D-related tax policies. We note here that different countries have different tax incentives policies. For example, Chapter 3.a of “Audit Techniques Guide: Credit for Increasing Research Activities” published in June, 2005, states that a firm’s new R&D expenses in America can be calculated according to the given provisions, and that 20% can be used directly as deductions from taxable income. In Section 2.2 of “SR&ED Investment Tax Credit Policy” issued in December, 2014, the Canadian government establishes a R&D tax credit rate of 15% for tax years that end after 2013 or 20% for tax years that end before 2014 generally, and the enhanced tax credit rate for Canadian-controlled private corporations at 35%. The tax law in China regulates that R&D expenses can be additionally deducted by 50% from the total taxable income amount. Obviously, different countries have different R&D tax incentives and applicable conditions. Therefore, previous research conclusions may not apply in China. As an emerging market, China has its own problems, such as a weak legal protection environment, serious government intervention, and concentrated share ownership (Allen, Qian, & Qian, 2005; Feng, 2004; Pan, Dai, Wu, & Liu, 2009; Qin & Shao, 2015). Although the theoretical effect of tax incentives is positive in China and developed countries, it is possible that firms in China are not able to either respond to these tax incentives policies or take advantage of them positively. Previous studies of R&D tax incentives have not yet reached a consensus on their effects. Some researchers believe that tax incentives can lead to a substantial increase in firms’ R&D investment (Baily & Lawrence, 1992; Hall, 1993; Mamuneas & Nadiri, 1996), whereas others have reached the opposite conclusion (Broadway, Chua, & Flatters, 1995; Griffith, Sandler, & Van Reenen, 1995; Mansfield, 1986). Some studies suggest that R&D tax policies do not fully and effectively stimulate R&D investment in the personal sector (Li, 2007; Wu, 2003). However, studies that focus on macro-level analysis find that government tax policies have positive effects on firms’ R&D investment (Wang, 2011; Zhu & Xu, 2003). Furthermore, many researchers argue that even if governments’ tax incentive policies encourage firms’ R&D investment, firms may still struggle with high adjustment costs (as defined in footnote 3) (Bernstein & Nadiri, 1982; Hall, Griliches, & Hausman, 1986; Hall & Hayashi, 1988; Himmelberg & Petersen, 1994). When a firm increases its R&D investment, in addition to the cost for buying new equipment, it may also need to spend resources on physical installation and commissioning of new equipment (e.g., Gould, 1968; Mussa, 1977). When expanding the scale of its R&D, a firm must reorganise and rearrange its R&D activities or recruit new R&D workers to meet the new needs, which incur certain costs (Groth & Khan, 2010). And also, investing in R&D workers’ training is necessary to fully and ec ffi iently exploit high-tech R&D facilities (Cooper & Haltiwanger, 2006). These types of costs prevent firms from costlessly adjusting the level of R&D capital. Consistent with other studies (e.g., Berger, 1993; Mansfield, 1986), the r&d tax incentives policies mentioned in this study are forms of income tax credit. CHINA JOURNAL OF ACCOUNTING STUDIES 435 Professional R&D workers, as the main source of intellectual capital, have high wages, which typically account for more than 50% of the R&D budget (Hall, 1993). If a firm pays for R&D staff with its own earnings, the resulting reduction in cash flow will hinder the manager’s ability to make a profitable business in the future. Moreover, if a firm incurs debts to pay these wages in a timely manner to avoid reductions in the initiative of R&D workers, then the firm has to pay interests or even give creditors some other benefits, such as luxurious entertainment, extravagant feasts, and costly gifts. In addition, the sensitivity of firms’ R&D investment to price may be low due to the long-term and uncertain nature of many R&D projects (Hall, 1993). Therefore, high adjustment costs have a significantly negative effect on R&D investment (e.g., Dixit & Pyndick, 1994; Groth & Khan, 2010; Hall, 1993), and thus the effectiveness of R&D tax policies. This negative effect may be larger when firms are faced with great uncertainty in the future (Hall, 1993). Thus, the incentive effects of R&D tax policies deserve further discussion. Existing studies tend to ignore the effect of adjustment costs, and few studies discuss the interaction of R&D tax incentives and adjustment costs. Furthermore, the nature and effectiveness of R&D tax incentives policies vary across countries, and this may lead to different studies reaching different or even opposite conclusions. When firms have difficulty in meeting adjustment costs, tax incentives do not significantly stimulate R&D activities (Hall, 1993). China’s rapid economic development, low per capita income, weak legal protection, and imperfect tax regulation reflect the common characteristics of emerging markets. Therefore, China can be seen as a typical representative of emerging markets in the global economy. However, few studies (especially micro-level studies) of R&D tax incentives in China have been based on empirical evidence. Therefore, in this study we use a sample of Chinese companies to discuss the effects of R&D tax incentives in a global economy, and to provide suggestions for effective tax-related R&D incentives for emerging markets. Our sample consists of high-tech listed companies from the 2007 to 2012 period. We discuss the effect of R&D tax incentives in China and find that the incentive effect of tax policies on R&D investment depends on the size of the adjustment costs. Specifically, when adjustment costs are larger than a certain critical value, the incentive effect ceases to exist. The results of our estimation and analysis of the critical value of adjustment costs under China’s existing R&D tax incentives policies show that China’s existing R&D tax incentives are, on the whole, effective. This incentive effect is statistically larger in non-state-owned enterprises than in state-owned enterprises. This study makes three main contributions. Firstly, with the consideration of the role of adjustment costs in R&D investment, we contribute to the debate about the incentive effects of R&D tax policies which has existed for a long time (e.g., Bloom, Griffith, & Van Reenen, 2002; Griffith et al., 1995; Guellec & Van Pottelsberghe, 2003). Many recent studies have explored the incentive effects of tax policies, but they have not reached consistent conclu- sions. Unlike previous studies, this paper considers the role of adjustment costs in R&D tax incentives and shows why it has been so difficult for researchers to reach a unanimous conclusion on the effect of R&D tax incentives. Specifically, our study based on the evidence from China indicates that the effects of R&D tax incentives gradually weaken as adjustment as an emerging economy, China has an income tax system similar to that found in other developed economies, although the rules and regulations in China remain at the establishment and improvement stage. f or example, income tax in China and many developed economies such as america, Japan and Korea is calculated using a proportional tax based on pre-tax income, despite different tax rates in each country. 436 W. LI AND J. DU costs increase. Only if adjustment costs are less than firms’ affordable critical values, can tax incentives policies effectively stimulate R&D. Secondly, we provide some empirical evidence about the incentive effects of R&D tax policies among firms with different types of share ownership, which enriches the literatures about share ownership and economic consequences of tax incentives. Prior studies about the incentive effects of R&D tax policies (e.g., Bloom et al., 2002; Cooper & Haltiwanger, 2006; Hall & Van Reenen, 2000) rarely discuss the role of property rights. As an emerging capital market, China has some problems in its market environment, such as weak legal protection, severe government intervention, and concentrated share ownership (Allen et al., 2005; Feng, 2004; Pan et al., 2009; Qin & Shao, 2015). Meanwhile, the complex principal-agent relationship in state-owned firms not only provides insufficient incentive, it also leads to ineffective supervision (Wu, 2009). We find that the effect of tax incentives on R&D investment is rela- tively weak in state-owned firms, indicating that through affecting the motivation of state- owned firms to use R&D tax incentive policies to reduce their costs, state ownership plays an important role in technological innovation. Therefore, our study supplements and expands the findings of Wu (2009), which suggest that state ownership brings heavier tax burdens to firms without an incentive tax rate than to firms with an incentive tax rate. Thirdly, this study contributes to enriching empirical research on the effects of policies of indirect government subsidies. Prior literature on government subsidies in China mainly provides empirical evidence relating to direct funding (e.g., Li, Tang, & Zuo, 2013; Xie, Tang, & Lu, 2009), but the economic impact of indirect subsidies is seldom discussed. Potentially, it is because indirect subsidies, which are ex post incentives, have no obvious advantages in terms of the speed of driving firms’ R&D investment when compared with direct government subsidies. However, we find that the existing tax incentives policies still significantly stimulate R&D overall with consideration of adjustment costs impeding firms’ R&D investment. This suggests that tax-oriented incentive policies for R&D are beneficial to promote technical progress. Therefore, this paper contributes to the literature focusing on indirect government subsidies, such as tax refunds or allowable deductions from pre-tax profit. Also, it provides a reference for governments setting policies regarding technical progress and indirect subsidies. 2. Literature review The role of technical progress in modern economies has become one of the core questions of the new economic growth theory, and this has led to increasing research on R&D activities (Aghion & Howitt, 1992; Grossman & Helpman, 1991; Romer, 1990). Modern economic theory suggests that technical progress and knowledge accumulation are important factors in eco- nomic growth (Romer, 1990). R&D is the main source of new technologies and knowledge (An, Zhou, & Pi, 2009) and also helps enterprises to make better use of existing external knowledge, to improve their capacity for digestion and absorption, and to increase their knowledge capital. Thus, R&D can indirectly promote technological innovation (Romer, 1990). Previous studies show that the spillover effect of knowledge may lead a firm to invest less than the socially optimal amount in R&D (Arrow, 1962). Therefore, to stimulate R&D investment, governments must create incentives. At present, direct subsidies and tax credits are the most widely-used policies. Lichtenberg (1984) believes that government subsidies for R&D may substitute for firms’ R&D spending. However, Zhu and Xu (2003) argue that CHINA JOURNAL OF ACCOUNTING STUDIES 437 subsidies and tax credits are complementary and that improving one will increase the ee ff ct of the other. They find that when both incentives exist, tax credits dominate. In recent dec - ades, various measures have been taken to increase firms’ R&D investment activities, even in developed countries, and firms are making use of R&D tax incentives policies (Brown, 1999). Many recent studies have explored the effects of tax incentive policies, but the results have been inconclusive. Hall (1993, 1995) argues that tax incentives have a positive effect on firms’ R&D investment, and this conclusion is supported by Eisner, Albert, and Sullivan (1984), Berger (1993), Baily and Lawrence (1992), Mamuneas and Nadiri (1996), Bloom et al. (2002), and Guellec and Van Pottelsberghe (2003). Furthermore, Hall and Van Reenen (2000) find that an increase of one US dollar in tax credits encourages firms to increase R&D invest - ment by one dollar. Other research shows that R&D investment increases by more than one dollar when tax credits increase by one dollar (Bernstein, 1986). However, although many studies suggest that tax incentives promote R&D investment, some studies have reached the opposite conclusion. For example, Mansfield (1986) and Griffith et al. (1995) argue that tax incentives do not promote a substantial increase in R&D input. Moreover, using a marginal effective tax rate model to analyse the incentive effect of tax incentives for firms over a five-year tax-free period, Broadway et al. (1995) find that this tax rule actually punishes firms, and creates a loss during the period. Estache and Gaspar (1995) argue that reductions in average effective tax rates are the result of tax arbitrage and tax evasion rather than tax incentives. They find that the abuse of tax incentives distorts the tax system. Some studies using data from China have found that R&D tax policies have a negative ee ff ct on incentives. For example, Wu ( 2003) and Li (2007) hold that existing R&D tax policies cannot fully stimulate firms’ R&D investment. Although Zhu and Xu (2003) and Wang (2011) suggest that R&D tax incentives policies in China have positive effects, these studies focus on a macro-level analysis; they do not conduct micro-level evaluations and do not discuss adjustment costs. However, some studies show that adjustment costs cannot be ignored when firms change R&D capital stock (e.g., Hall, 1993; Romer, 2006). Lucas (1967) points out that firms that expand capital stock are likely to face considerable adjustment costs. This conclusion is supported by other studies (e.g., Shapiro,1986; Lichtenberg, 1988a; Cooper & Haltiwanger, 2006). Moreover, some researchers find that the scale of a firm is not restricted when adjustment costs do not exist (e.g., Abel, 1983; Caballero, 1991). Pindyck (1993) argues that a firm’s scale is decided by adjustment costs only, that its profit is a linear function of capital stock when the firm is in constant returns to scale, and that its growth is described by a demand curve of infinite elasticity. In this case, adjustment costs are the only factor that limits a firm’s scale (Pindyck, 1993). If marginal profit is more than the unit cost of capital, the expansion of a firm’s scale may be unlimited (Pindyck, 1993). At the same time, Pindyck (1993) finds that many small firms will enter the market if entering the market is free, because these small firms spend less on R&D and have few adjustment costs. Thus, entering or exiting the market may be easier for small firms. As for the relationship a djustment costs are the costs firms face when they change the level of investment, including the costs for installation, training workers to operate new machines, reorganising the process, etc. (groth & Khan, 2010; romer, 2006). t hese types of costs prevent firms from costlessly adjusting the level of capital. ( abel & eberly, 1994, 1998; dixit & Pyndick, 1994; groth & Khan, 2010). if adjustment costs are high and difficult to change in the short term, they will have an inhibitive effect on firms’ r&d , and thus diminish greatly the incentive effects of a tax policy on r&d (hall, 1993). 438 W. LI AND J. DU between a firm’s investment and adjustment costs, Abel and Eberly (1994, 1998) and Dixit and Pyndick (1994) agree that adjustment costs may prevent a firm from frequently changing its existing capital stock. Groth and Khan (2010) argue that adjustment costs can cause business investment inertia, which leads to failure to make timely adjustments and correc- tions in response to changes in the external environment. Christiano and Todd (1996) and Edge (2000) find that adjustment costs delay business investment plans and prevent a firm from changing an existing investment plan. Recently, many studies have analysed the rela- tionship between adjustment costs and business investment. Cooper and Haltiwanger (2006) point out that adjustment costs are important in the evaluation of investment-driven tax incentives policies. However, many researchers have ignored the role of adjustment costs when discussing Chinese investment issues, such as investment-driven tax policies in China. The foregoing review shows that there is no consensus on the relationship between R&D and tax incentives or on the effect of adjustment costs on R&D investment. Few studies of the effect of tax incentives policy on R&D investment consider adjustment costs. Moreover, conclusions based on empirical research in developed countries do not necessarily apply to developing countries such as China. The Chinese government has issued a series of policies on R&D tax incentives. In February 2006, the state council issued the “National Outline of Medium-term and Long-term Science and Technology Development Plan from 2006 to 2020,” and several supporting policies for the implementation of this plan. These policies allow firms to deduct an amount of taxable income equal to 150% of their actual R&D expenses in the incurred year. In addition, the new “Enterprise Income Tax Law in the People’s Republic of China,” issued on March 16, 2007, includes the tax policies mentioned above, demonstrat- ing the Chinese government’s commitment to R&D tax incentives. However, as an emerging capital market, China has some problems in its market environment, such as weak legal protection, severe government intervention, and concentrated shareholdings (Allen et al., 2005; Feng, 2004; Pan et al., 2009; Qin & Shao, 2015). This kind of market environment is likely to lead to a lack of supervision and incentives, particularly in state-owned firms ( Wu, 2009). Thus, managers in state-owned firms have less motivation to take advantage of R&D tax incentives policies to reduce their firms’ actual tax burden. However, convincing empirical evidence, especially micro-level evidence from China, is rare for resolving the questions of whether firms increase their R&D investment under the Chinese system and whether the incentive effects of tax policies are different for firms with different types of share ownership. 3. Institutional background Since the end of the last century, when the Central Government and the State Council of China declared its intention to increase the input into science and technology by multi-chan- nel and multi-level financing system according to the policy “Decision of on Accelerating the Progress of Science and Technology” in 1995, the scope of China’s R&D tax incentives policies has constantly expanded. More recently, the implementation management of tax incentives has been increasingly standardised with a series of amendments. Tax incentives policies have been stable since 2006. In 2006, China’s Ministry of Finance and State Administration of Taxation issued a fiscal and taxation provision, known as [2006] No. 88. In addition to granting firms a 100% deduction of actual technology development fees, the provision allows another 50% of the actual technology development fee to be deducted CHINA JOURNAL OF ACCOUNTING STUDIES 439 before income tax. The provision’s purpose is to encourage R&D activities in firms by allowing firms to combine tax benefits with technology development activities. Compared with pre- 2006 tax policies, the post-2006 policies offer incentive tax treatments to firms. Although [2006] No. 88 lost its efficacy according to the policy “Decision of the Ministry of Finance about the Announcement on the Revocation and Failure of Financial Regulations or Documents (eleventh)”, the State Administration of Taxation issued a provision known as [2008] No. 116 to stimulate R&D investment. This provision has been in place since 2008 and reflects the government’s continuing focus on R&D activities. Comparing [2006] No. 88 with [2008] No. 116, we find that there are no substantive differences with regard to R&D tax incentives, although the latter provides a more detailed explanation for the calculation of the additional R&D deduction. Although these provisions reflect the importance of R&D activities to the Chinese government, few studies have examined whether they have sub- stantial incentive effects on firms’ R&D activities (e.g., Wang, 2011). However, Wang (2011) did not discuss the role of adjustment costs. Furthermore, although the Chinese government offers a series of R&D tax policies to stimulate R&D, existing studies find that firms that increase their R&D investment inevitably face adjustment costs. For example, if the installation of new equipment leads to a complete change in the structure of production, a firm must retrain workers in the new production processes (Cooper & Haltiwanger, 2006). An increase in R&D workers incurs recruitment fees. Caballero, Engel, and Haltiwanger (1997) point out that adjusting employment is costly. Firms must also bear the costs of nonflexible salary contracts. Hall (1993) believes that R&D workers, as the main carriers of firms’ knowledge capital, receive high salaries, which account for more than 50% of a typical R&D budget. Although the degree of capital market devel- opment varies between countries, all firms inevitably face adjustment costs when they take advantage of R&D tax incentives policies. Therefore, firms in China’s emerging capital market have adjustment costs associated with R&D investment. It is necessary to consider whether China’s R&D tax incentives policies are effective after adjustment costs have been taken into account. This study analyses how adjustment costs interact with the incentive effects of R&D tax policies in China’s institutional environment. Our study supplements existing research by focusing on adjustment costs when discussing the incentive effects of R&D tax policies. Moreover, this study contributes to the literature on R&D tax incentives and capital invest- ment by providing new empirical evidence for the effects of R&D tax incentives, and offering practical and flexible suggestions for the construction of tax systems in emerging markets. 4. Theoretical analysis Modern economic growth theory indicates that technical progress and knowledge accu- mulation are important components of economic growth (An et al., 2009). R&D not only generates new knowledge and fresh information that promote firms’ innovation, it can also improve firms’ use of external existing knowledge and their ability to digest and absorb information, promote firms’ stock of knowledge, and thereby indirectly increase technolog - ical innovation (Romer, 1990). However, R&D activities have positive externalities, and as a result firms or investors do not receive all of the benefits of R&D activities, which lead to the “market failure” problem (Arrow, 1962). Therefore, when the social marginal benefit of 440 W. LI AND J. DU investment is higher than the firm’s benefit, the government should consider firms’ benefits from R&D activities in light of the social benefits. Through direct subsidies or tax incentives, governments can encourage firms to invest in R&D. Such policies reduce firms’ R&D costs and enhance their core competitiveness in the market (Grossman & Helpman, 1991). Because direct subsidies are used only in a few industries, they are not effective in moti- vating an overall increase in R&D investment for those firms which can’t employ the policies of direct subsidies (Tang, Lu, & Li, 2008). In contrast, tax incentive policies allow firms to choose R&D projects according to their own development strategies and market demand, and thereby allow them to make full use of their power and the market (Li, 2007). As a result, governments often adopt tax incentive policies aimed at reducing firms’ R&D costs to stim- ulate R&D investment (Brown, 1999; Wang, 2011). However, firms that change their R&D capital stock may generate corresponding adjustment costs (Cooper & Haltiwanger, 2006; Groth & Khan, 2010). In such cases, tax policies do not necessarily lead to the expected incentive effects because high adjustment costs may negatively influence R&D investment. Romer (2006) argues that if adjustment costs are increased to infinity and capital stock increased to infinity, lower interest rates will finitely increase investments. As a consequence, a firm’s capital stock will gradually achieve a new ideal level. Although a firm benefits from tax incentives policies, it must still pay for equipment installation, staff training, and produc - tion restructuring when it increases R&D investment. In addition, some R&D spending is so viscid that it hinders firms from increasing investment (Hall, 1993). For example, once firms increase the number of R&D workers, they will have to maintain these higher staff wages for a long time. The responsiveness of R&D investment to tax incentives may be low or even be unobservable due to such adjustment costs (Cooper & Haltiwanger, 2006). Thus, adjustment costs cannot be ignored when discussing tax incentive effects on R&D investment. This study argues that firms engaged in R&D investment activities can afford adjustment costs just within a certain range, as R&D incentive effects produced by tax policies will decrease with an increase in adjustment costs. Specifically, firms with small adjustment costs will have positive responses to a tax incentives policy for R&D investment. However, firms faced with unaffordable adjustment costs may not be willing to increase investment in R&D, even if tax incentives policies are in place. Therefore, whether tax incentives significantly affect R&D investment depends on weighing the strength of tax incentives against adjust - ment costs. We predict that tax incentives policies for R&D investment in Chinese firms will only have significant incentive effects if the corresponding adjustment costs are affordable. Based on the above discussion, we propose the following hypothesis: Hypothesis 1: The strength of the existing R&D tax incentives in Chinese firms overcomes the barriers created by adjustment costs, ceteris paribus. 5. Empirical analysis 5.1. Variable design and empirical model 5.1.1. Variable design This study discusses the effect of adjustment costs on tax incentives for R&D investment. Therefore, the dependent variable is change in R&D investment. The independent variables include B-index, which measures the effect of tax incentives, adjustment costs, and the CHINA JOURNAL OF ACCOUNTING STUDIES 441 interaction between them. Other factors that may affect R&D investment are used as control variables. (1) Change_rd: R&D investment includes funds, staff, creativity, and other informa- tion. However, creativity and information lack objective and quantifiable criteria. Furthermore, creativity is mainly generated by R&D workers. Investment in R&D processes mainly consists of R&D expenses, and funds play a key role in a firm’s R&D investment. Therefore, we choose the first-order difference in R&D invest - ment to measure high-tech listed firms’ changes in R&D investment, scaled by operational revenue. (2) B-index: Warda (1994, 1996, 2001, 2006) constructs the B-index as a measure of the effectof R&D tax incentives. This measure has been adopted by many researchers (e.g., Czarnitzki, Hanel, & Rosa, 2011; Ernst, Richter, & Riedel, 2014; Guellec & Van Pottelsberghe, 2003). Specifically, Warda equates the B-index to ATC /(1 − t), where ATC (After-tax Cost) represents the net costs equal to one unit of R&D investment minus tax incentives, and t represents the corporate income tax rate. When the deduction rate of R&D investment before tax is v, ATC = 1 − vt. Ast is the corporate income tax rate, 1 − t is the total after-tax cost of one unit of R&D expenditure. Thus, the B-index equals (1 − vt)/(1 − t), which represents the ratio of net costs calculated by one unit of R&D expenditure minus tax incentives to total after-tax costs. Therefore, a larger B-index corresponds to a less effective tax incentive. We adopt a negative B-index for the convenience of observing the role of tax incentives on R&D investment. (3) Adjustment: In general, as the introduction of new equipment and methods can lead to changes in a firm’s entire R&D structure, firms need to pay for the training or retraining of R&D workers (Cooper & Haltiwanger, 2006). Therefore, adjustment costs and R&D investment are positively correlated. Firms with a large R&D capital stock face small adjustment costs. Firms accumulate extensive experience during the capital investment process. Thus, they pay less when adding another unit of R&D investment. Summers (1981) uses (a/2)(I /R ) R to measure adjustment costs, t t t where a is a parameter that is greater than 0, R represents R&D capital stock in period t, and I represents a firm’s R&D investment in period t. Other research- ers, such as Polder and Verick (2004), Pindyck and Rotemberg (1983), Hassett and Hubbard (1996), and Hall (1993, 2004), adopt a similar approach to measure adjust- ment costs. Although the development level of the capital market varies between countries, the main features of adjustment costs are common, and are strongly related to R&D activities. China, as an emerging capital market, has a distinctive institutional environment. However, both China and other countries are faced with the same basic problems in technical innovation. Hence, following Summers (1981), we use Chinese capital market data to re-evaluate parameter a under the current Chinese environment. The OLS estimation model is as follows: I ∕R = c + 1∕a ⋅ (q − 1)+ e t t t t where c is a constant, q is the value of Tobin’s q in t period, and e is a random distur- t t bance term. By re-evaluating parameter a under the current Chinese environment, we get an estimated value of a equal to14, which is significantly different from 0 at 442 W. LI AND J. DU the 1% significance level. And then, we take adjustment costs per unit of capital, which is equal to (a/2)(I /R ) , as one variable of interest in this paper. t t (4) B_index × Adjustment: Although R&D tax incentives policies reduce the price of R&D, tax incentives for R&D activities may not have an incentive effect due to high adjustment costs. When the savings created by R&D tax incentives policies are the same for two firms, the firm with the larger adjustment costs will invest less in R&D. When firms engage in R&D activities, they need to balance tax incentives and adjustment costs. To increase the R&D investment of firms it is necessary to consider the interaction effect. Hence, this study analyses the combined effects of adjustment costs and R&D tax incentives on R&D. In addition to adjustment costs and tax incentives, we control some variables which may affect a firm’s R&D investment, such as firm growth, debt-paying ability, firm size, firm age, the nature of the industry (e.g., Chen & Gupta, 2010; Hall, 2002; Huang & Chen, 2011; Khanna & Iansiti, 1997; Liu & Liu, 2007). The details of these variables are discussed below. (1) Size: Firms with different sizes have different advantages in technical innovation. Generally, a small firm has behavioural advantages, such as flexibility, whereas a large firm has more material capacity. Pavitt, Robson, and Townsend (1987) find a U-shaped relationship between the effect of R&D spending and firm size. Cohen and Levin (1989) finds that increases in R&D spending are proportional to the size of the firm above a certain critical value. Lee (2002) believes that firm size affects R&D investment indirectly through unobserved technical competitiveness. Some studies suggest that the effect of firm size on R&D investment is uncertain. For example, Chen and Gupta (2010) argue that although a large firm can allocate a larger budget for R&D, large firms are in general more mature and thus less likely to increase their R&D investments. However, other studies suggest that firm size and R&D investment are not related. For example, Jefferson, Bai, Guan, and Yu (2006) use a panel dataset of Chinese medium-sized manufacturing firms to show that after controlling for industry effects the intensity of R&D expenditure is not significantly affected by firm size or the degree of market concentration. An et al. (2006) use a questionnaire sent to manufacturing industries in Jiangsu, China to show that the trends in R&D intensities in small, medium, and large firms have a V-shaped structure. In a study based on the top 100 firms in the Chinese electronic information industry in the 2000 to 2003 period, Wang (2005) shows a significantly positive relationship between the scale of R&D expenditure and firm size, but not between the intensity of R&D expenditure and firm size. In this study, we use firm size, which is equal to the natural logarithm of total assets, as a control variable in our model. (2) Cash: A firm planning to implement R&D requires adequate cash to purchase equip - ment, employ R&D workers, and adjust its production process. Cash flow directly determines a firm’s available budget for R&D investment. Huang and Chen (2011) argue that cash flow affects R&D investment, and that financing costs for R&D are associated with cash flow. Therefore, we include cash flow in our model, which is equal to the cash flow generated from operating activities scaled by total assets. CHINA JOURNAL OF ACCOUNTING STUDIES 443 (3) Leverage: Generally, as a firm’s debts increase, the interest payments increase. Hence, an increase in interest rates can increase the cost of capital, which hinders afirm’s ability to increase its R&D investment. Chen and Gupta (2010) believe that a high asset–liability ratio may limit corporate discretionary spending to the point that a firm cannot implement its R&D investment plan. Hall (2002) suggests that although investment in innovation can generate intangible capital, such as highly specialised human capital, banks and other creditors generally prefer to mortgage physical assets. Thus, when creditors have to choose between providing a loan to R&D projects or to plant and equipment building, they are likely to choose the latter. In addition, firms cannot obtain the full benefits of R&D investment due to the high risks and positive externalities of R&D investment. Therefore, creditors are unwilling to lend or require very high interest for R&D investments. As a result, it is difficult for firms to finance R&D investment through loans, and only R&D pro - jects with readily available funds can be implemented. Bhagat and Welch (1995) also believes that firms with high financial leverage should reduce R&D spending. Huang and Chen (2011) argue that corporate debt has a negative effect on R&D investment. In this study, the level of corporate debt is controlled for in the model, which is equal to the ratio of total liabilities to total assets. (4) Age: At every stage of development firms must choose the operating conditions that will sustain their development and counter their challenges. Generally, a firm in its birth period increases market share by providing customers with new prod- ucts or services. Managers are willing to increase R&D investment in this period because they have a strong sense of innovation and competition. In the growth stage, the continuous expansion of firms may lead to resource shortages, and firms will likely have difficulty financing R&D investment. In the mature period, firms own extensive capital reserves through pre-development and visionary entrepreneurs may opt to invest vigorously in technical innovation to avoid a recession caused by the absence of new revenue growth points. In the recession period, firms’ sense of innovation and competition wanes, management ideas and management modes recede, and productivity is low. This declining business performance leads to sig- nificantly inadequate input into R&D. Huang and Chen (2011) find that firm age has a significant effect on R&D investment. Based on their finding, we control for firm age in our empirical analysis, which is equal to the years since a firm was founded. (5) ROA: Although a firm with weak profitability has considerable innovative motiva- tion, its R&D investment is insufficient due to financial constraints. Moreover, the long cycle of innovation investment requires continuous financial support. Thus, to support innovation and investment, firms require profitability to be steadily improving (Zhang, Liu, & Wang, 2012). Khanna and Iansiti (1997) find that firms with more abundant resources are more inclined to invest in technological innovation. Zhao, Su, and Zou (2006) argue that firms with a substantial market share and high performance are strongly motivated to invest in R&D. Therefore, this study controls for business performance, equal to the net profit divided by total assets. (6) Tenure: R&D provides firms with great future benefits and helps to maintain and improve their core competitiveness. Therefore, R&D is conducive to a firm’s long- term development. A r fi m that is steadily developing gives its executives more ben - efits and a sense of accomplishment. However, executives only enjoy the benefits 444 W. LI AND J. DU of R&D investments if they stay with a firm for a long time. This time lag between R&D investment and profits may reduce the enthusiasm of executives for R&D investment. Liu and Liu (2007) also believe that executives who stay longer in their positions are more likely to benefit from R&D investment and are therefore more likely to invest in R&D. Therefore, a positive correlation exists between R&D invest- ment and executives’ tenure. We use executive tenure, which is equal to the number of years an executive has worked in a firm, as a control variable in our model. (7) Industry: In addition to the control variables above, we also control for the effect of industry factors on R&D investment. Although the sample used in this study consists of high-tech industries, there are variations between types of high-tech industries. According to “The Notice of High-tech Industries Statistics Categories” ([2002] No. 33) and “Listed Companies Industries Classification Guideline in 2001,” high-tech industries include chemical materials and chemical products manufactur- ing, chemical fibre manufacturing, electronics manufacturing, medical equipment manufacturing, aerospace and aviation manufacturing, instrument manufacturing, pharmaceutical and biological manufacturing, and information technology. 5.1.2. Empirical model This study builds on existing research to construct a regression model with the aforemen- tioned variables. The regression model is as follows: Change_rd = C + B_index + Adjustment + B_index × Adjustment i,t 1 i,t 2 i,t−1 3 i,t i,t−1 (1) + Controls + k i,t i,t The measurements of the intensity of R&D investment and the effect of R&D tax incentives are based on the measurement of R&D expenses. The measurement of adjustment costs is related to the amount of R&D investment. We do not use the intensity of R&D investment directly as the dependent variable, as there could be in this case a natural mechanical cor- relation between the dependent and independent variables. Following previous studies (Czarnitzki et al., 2011; Ernst et al., 2014; Guellec & Van Pottelsberghe, 2003; Warda, 2006), this study uses the B-index to measure the effect of R&D tax incentives policies on firms and the change in R&D investment as the dependent variable. Our research design avoids some of the problems related to mechanical correlation as the B-index is based on the pre-tax deduction rate and corporate income tax rate. 5.2. Empirical analysis 5.2.1. Sample selection and data As the “Enterprise Accounting Standards” have been in effect since January 1, 2007, our sample consists of companies in high-tech industries listed on the Shanghai and Shenzhen stock exchanges during the period 2007 to 2012. The R&D investment data are obtained from the annual reports of each firm. We use listed companies in high-tech industries as they in general require more R&D than non-high-tech firms. Furthermore, Wu (2006) finds that the output elasticity of R&D in high-tech industries is significantly greater than in CHINA JOURNAL OF ACCOUNTING STUDIES 445 non-high-tech industries. He believes that R&D in high-tech industries is an important factor in productivity growth (Wu, 2006). Chinese listed companies have been disclosing R&D data in their annual reports since 2002. Related research shows that in annual reports written before 2007, firms’ R&D invest - ment data are mainly disclosed in the category “cash payment relating to other operating activities” in the notes on the statement of cash flow (Xu & Tang, 2012). In the process of checking the listed companies in high-tech industries in the 2007 to 2012 period, we find that the disclosures of R&D data after 2007 are mainly under the category “management expenses.” The categories that reflect R&D investment include R&D expenses, technology research expenses, technology development expenses, scientific research expenses, ad con - sulting, and technology development expenses. By manually collecting data and deleting missing values, we obtain a final sample of 2092 R&D investment observations disclosed in “management expenses,” 860 R&D investment observations in “cash payment relating to other operating activities,” and 681 R&D investment observations disclosed in both subjects. In our empirical analysis, we calculate the amount of R&D investment in each period using R&D investment data from both subjects, as R&D investment includes R&D expenses for staff salaries and other fees, and R&D laboratories, equipment, and other capital-oriented spending. High-tech listed firms’ annual reports are obtained from the information disclosure web - site (www.cninfo.com.cn) designated by the China Securities Regulatory Commission. Other data are from the CSMAR and Wind databases. In the empirical analysis, we winsorize each continuous variable at the 1% level to mitigate the effect of outliers. 5.2.2. Descriptive statistics Table 1 reports the descriptive statistics of the variables. The median of Change_rd is 0.001, indicating that the change in R&D expenditure is less than 0.001 in nearly half of the firms. The B_index mean of −0.89 and its standard deviation of 0.05, shows that considerable dif- ferences exist in the effect of tax incentives on different firms in our sample. Tenure has a mean of 1.75, indicating that the executives’ tenure in high-tech listed firms is short, with an average of only 1.75 years. The average firm foundation age is 11.80 years, and the longest is only 24 years, suggesting that Chinese high-tech listed firms are young. We analyse the correlations between these variables. The results are shown in Table 2. The correlation coefficient between the change in R&D investment and B_index is 0.060 and significant at the 5% level. Initially, the result shows that a greater B_index is related to more R&D investments; that is, tax incentives policies have a positive effect on R&D investment. However, the first-order lagged adjustment costs and R&D investment are significantly neg - atively correlated, showing that r fi st-order lagged adjustment costs impede R&D investment. Table 1. d escriptive statistics. Variable Mean Median Max Min Std. Change_rd 0.01 0.00 0.10 −0.04 0.02 B_index −0.89 −0.91 −0.75 −1.00 0.05 Adjustment 0.00 0.00 0.07 0.00 0.01 Size 21.21 21.16 23.92 18.16 0.99 Leverage 0.42 0.38 2.73 0.03 0.35 Cash 0.05 0.04 0.27 −0.20 0.08 Age 11.80 11.00 24.00 2.00 4.84 ROA 0.04 0.05 0.28 −0.34 0.08 Tenure 1.75 1.00 12.00 0.00 1.89 446 W. LI AND J. DU Table 2. Pearson correlation coefficients between variables. Variable 1 2 3 4 5 6 7 8 9 1. Change_rd 1.000 ** 2. B_index 0.060 1.000 *** 3. Adjustment −0.103 −0.041 1.000 * *** ** 4. Size −0.046 −0.102 −0.059 1.000 *** *** *** * 5. Leverage −0.137 0.120 −0.153 −0.039 1.000 * ** *** *** 6. Cash 0.043 0.033 0.063 0.060 −0.169 1.000 *** *** *** *** *** 7. Age −0.127 0.083 −0.120 0.164 0.339 −0.031 1.000 *** ** *** *** *** *** *** 8. ROA 0.093 −0.041 0.143 0.073 −0.403 0.321 −0.179 1.000 *** *** *** *** 9. Tenure −0.006 −0.115 0.069 0.088 −0.082 −0.004 0.033 0.011 1.000 *** ** * notes: , , and indicate statistical significance at the 1, 5, and 10% levels, respectively. CHINA JOURNAL OF ACCOUNTING STUDIES 447 Table 3. a djustment costs, tax incentives, and r&d investment. (1) Do not consider adjustment costs (2) Consider adjustment costs Variable Coefficient t value Coefficient t value *** *** Intercept 0.056 4.16 0.065 4.72 Size 0.000 0.18 0.000 0.27 *** *** Leverage −0.005 −3.02 −0.005 −2.94 Cash 0.009 1.44 0.009 1.41 *** *** Age −0.000 −3.79 −0.000 −3.80 ROA 0.005 0.85 0.006 0.88 Tenure −0.000 −0.58 −0.000 −0.77 *** *** B_index 0.042 3.72 0.053 4.53 *** Adjustment −3.917 −3.17 *** B_index × Adjustment −4.443 −3.22 Industry Yes Yes obs. 1,603 1,602 a dj. R 0.078 0.084 Prob. > F 0.000 0.000 notes: t he dependent variable is Change_rd. ***, **, and * indicate statistical significance at the 1, 5, and 10% levels, respectively. Although the correlation coefficients between the control variables are partly significant at the level of 1% level, the variance inflation factors of these control variables are less than 10, indicating that our regressions do not have a serious multicollinearity problem. 5.2.3. Empirical results The regression results in Table 3 show that the estimated coefficient of B_index is always significantly different from 0 at the 1% level, whether we consider adjustment costs or not. When we consider adjustment costs, the estimated coefficient of Adjustment is −3.917 and is significant at the 1% level. Moreover, the estimated coefficient of the interaction, B_index × Adjustment, is significantly negative at the 1% level. Overall, all three independent variables have significant explanatory powers. In addition, the marginal effect U of B_index on the change in R&D investment can be expressed as U = Change_rd∕ B_index = 0.053 − 4.443 Adjustment where U should be greater than 0 if tax incentives policies have a positive effect on R&D investment, because a larger B_index implies a greater strength of tax incentives. Furthermore, incentive effect U decreases as the first-order lagged adjustment costs increase. When the first-order lagged adjustment costs are equal to the critical value of 0.012 (=0.053/4.443, equivalent to 93% quantile), U is equal to 0. Accordingly, when the first-order lagged adjust - ment costs are less than the critical value of 0.012 (Adjustment < 0.012), U is larger than 0, indicating that tax incentives have a positive incentive effect on R&D investment. When the first-order lagged adjustment costs are greater than the critical value of 0.012, U is less than 0, indicating that tax incentives have a negative incentive effect on R&D investment. Thus the effect of tax incentives on R&D investment depends on whether the adjustment costs exceed the critical value. Based on the critical value of 0.012, which corresponds to the 93% quantile of Adjustment, we can infer that nearly 93% of the firms in our sample are affected by tax incentives, which means that U is greater than 0, and that the current R&D tax policies in China are effective. This conclusion supports hypothesis 1, which predicts that the incen- tive effects of R&D tax policies gradually weaken with an increase in adjustment costs. Overall, 448 W. LI AND J. DU the current R&D tax policies in China are strong enough to overcome the constraints imposed by adjustment costs and are effective. 5.3. Robustness tests 5.3.1. Excluding direct government subsidies In practice, Chinese firms engaged in R&D activities may obtain direct subsidies from the government in addition to benefits from tax incentive policies. Therefore, firms may increase their investments in R&D due to direct government subsidies rather than tax incentives policies. We exclude from our sample firms that receive subsidies from the Chinese govern- ment, in order to eliminate the effect of direct government subsidies. After re-running the regression model, we find that the results (see Table 4) remain robust. 5.3.2. Other control variables Two important factors, namely growth opportunity and investment opportunity, are not controlled for in our original model as only a few studies have focused on these two factors. However, given the potential effects of growth and investment opportunities on R&D invest - ment, we attempt to control for these two factors in our robustness tests. The results (see Table 4) remain unchanged. 5.3.3. Firms with higher R&D investment If a firm’s R&D expenditure is small, then the corresponding benefits from R&D tax incentive policies will also be relatively small. In this case, a firm has no motivation to take advantage of the tax policies. To eliminate this bias, we exclude from our sample firms with R&D invest - ments less than the median. We re-run the regression analysis using the sample of high-in- vestment firms; the results (see Table 5) support our original findings. 5.3.4. Excluding firms suffering losses Currently, R&D tax incentive policies in China stipulate that firms can hold, in addition to a 100% deduction for R&D expenses, an extra 50% deduction of R&D expenses that occur in the same year. Accordingly, if a firm suffers a loss in one year and does not generate taxable income, it cannot benefit from R&D tax incentive policies that year as it does not have a taxable income. Therefore, we exclude from our sample firms that suffer losses. Then we re-run the regression analysis; the results (see Table 5) remain substantially unchanged. 5.3.5. Endogeneity In the research design, we not only adopt the change in R&D investment (Change_rd) as the dependent variable, we also use the B-index (B_index) to measure the effect of R&D tax incentives. Thus, we avoid a mechanical correlation between the variables that may be caused by directly using R&D investment as the dependent variable and the amount of tax incentives calculated by deduction rate as the explanatory variable. To further reduce poten- tial endogeneity, we test the extent of potential endogeneity using the method proposed by Baum, Schaffer, and Stillman (2003, 2007). They estimate their regression model using an endogenous variable as an exogenous variable and then test the corresponding orthogonal conditions. The statistic is subject to the chi-square distribution, and the degree of freedom is the number of variables in the alternative test. The null hypothesis is that one endogenous variable can be considered an exogenous variable. In this study, we test the extent of CHINA JOURNAL OF ACCOUNTING STUDIES 449 Table 4. t ax incentives, adjustment costs, and r&d investment: robustness tests i. (3) Sample excluding firms that receive direct subsidies (4) Control Tobin’s q (5) Control MB Variable Coefficient t value Coefficient t value Coefficient t value *** *** *** Intercept 0.058 3.36 0.059 3.77 0.056 3.51 Size 0.001 0.88 0.000 0.65 0.001 0.81 *** *** Leverage −0.001 −0.79 −0.005 −3.06 −0.005 −3.15 Cash 0.011 1.37 0.008 1.31 0.008 1.28 *** *** *** Age −0.001 −3.52 −0.000 −3.85 −0.000 −3.76 ROA 0.003 0.47 0.004 0.62 0.003 0.44 Tenure 0.000 0.36 −0.000 −0.75 −0.000 −0.73 *** *** *** B_index 0.058 3.77 0.052 4.39 0.051 4.35 * *** *** Adjustment −2.577 −1.81 −3.968 −3.20 −3.965 −3.20 * *** *** B_index × Adjustment −3.087 −1.90 −4.498 −3.26 −4.493 −3.26 Tobin’q 0.000 0.86 MB 0.000 1.12 Industry Yes Yes Yes obs. 772 1,602 1,602 a dj. R 0.075 0.084 0.084 Prob. > F 0.000 0.000 0.000 notes: t he dependent variable is Change_rd. ***, **, and * indicate statistical significance at the 1, 5, and 10% levels, respectively. Table 5. t ax incentives, adjustment costs, and r&d investment: robustness tests ii. (6) Sample of firms with higher than (7) Sample excluding firms suffering median R&D investment losses Variable Coefficient t value Coefficient t value *** *** Intercept 0.143 5.29 0.063 4.19 *** Size −0.004 −3.15 0.000 0.77 *** Leverage −0.004 −0.92 −0.006 −2.90 Cash 0.021 1.66 0.001 0.17 * *** Age −0.000 −1.67 −0.000 −3.78 *** ROA −0.005 −0.28 0.028 2.75 ** Tenure −0.001 −2.07 −0.000 −0.81 ** *** B_index 0.043 2.02 0.057 4.63 *** *** Adjustment −4.470 −2.78 −4.152 −3.36 *** *** B_index × Adjustment −4.800 −2.68 −4.727 −3.43 Industry Yes Yes obs. 845 1,405 a dj. R 0.112 0.102 Prob. > F 0.000 0.000 notes: t he dependent variable is Change_rd. ***, **, and * indicate statistical significance at the 1, 5, and 10% levels, respectively. endogeneity in the independent variables and find that we cannot reject the null hypothesis significantly at the 1% level. Therefore, we do not consider the endogeneity problem as a serious concern. Thus, the estimated coefficients in the regression model are acceptable. 5.4. Further discussion The above analysis shows that the incentive effects of R&D tax policies decrease with an increase in adjustment costs. Overall, the effect of R&D tax incentives in China overcomes the constraints imposed by adjustment costs and has a positive effect on R&D investment. However, these incentive effects may be significantly different for firms with different 450 W. LI AND J. DU property rights. Given the large state-owned economic sector in China, this is an important issue. Therefore, we analyse the relation between property rights and R&D tax incentive effects. We believe that a tax incentives policy will have different incentive effects on the R&D of firms with different types of share ownership. First, the government is the principal and controlling shareholder of state-owned firms. The government benefits from both firms’ profits and from collected taxes. Therefore, both taxes and profits are forms of wealth con- trolled by the government stakeholder. In non-state-owned firms, only the after-tax profits are real wealth that belongs to the principal (that is, the private controller). As a result, the two principals have different motivations to avoid taxes. Second, in addition to value max - imisation, state-owned firms have social goals, such as easing employment pressure, increas - ing state revenue, and maintaining social stability. Paying more taxes ensures the sustainability of the financial revenue needed to achieve those goals. For managers in state-owned firms, taxes are the measure of operating performance and are vital for their compensation and political promotion. Third, the nature of property rights affects the costs of tax planning in financial reports and thus the behaviour of tax planners. Wang, Wang, and Peng (2010) point out that judging whether a firm’s decline in performance is caused by tax planning or mis- management is difficult for outside investors due to the existence of asymmetric information. Under serious information asymmetry, outside investors mistakenly evaluate firms’ perfor - mances as a result of tax planning, which leads to higher financial reporting costs. The tax planning of state-owned firms often contributes to high financial reporting costs, because state-owned firms are more likely have severe information asymmetry (Scholes, Wolfson, Erickson, Maydew, & Shevlin, 2005), which results in a lack of motivation to consider R&D tax incentives policies in tax planning. Similarly, the complex principal-agent relationship in state-owned firms not only provides insufficient incentive, it also leads to ineffective supervision (Wu, 2009). In addition, the shareholders of state-owned firms in China are all citizens in this country. Chinese govern- ment exercises the rights of shareholders on behalf of the citizens. However, the government prefers to focus on its administrative goals, such as much employment and social stability, which are likely to differ from shareholders’ benefits (He, 1998). Thus, the government has less incentive to monitor managers of state-owned firms effectively, which is prone to incur problems of insider control and moral hazard. A firm’s manager is likely to use the controlled resources to seek personal benefits at the expense of stakeholders’ interests, which may cause long-term insufficiency in R&D investment for state-owned firms (Liu & Liu, 2007). On the basis of the above analysis, we infer that state-owned firms may lack the motivation to use R&D tax incentive policies to reduce their costs. As a result, the effect of tax incentives on R&D investment is relatively weak in state-owned firms. Furthermore, the critical value of adjustment costs is low for these firms. If this inference is satisfied, then we should observe a lower coefficient of the B_index in a sample of state-owned firms. At the same time, we can obtain a low critical value of adjustment costs by calculating the equation U = 0. To better investigate the difference in tax incentive effects on R&D investment in state- and non-state- owned firms, this study divides the entire sample into two subsamples and re-runs the regression model. Table 6 shows that the coefficients of the B_index in the two subsamples are significantly positive, but the coefficient of the B_index for state-owned firms is less than the coefficient for non-state-owned firms (i.e., 0.048 < 0.059). The difference between the two coefficients is statistically significant. The critical value of Adjustment for state-owned CHINA JOURNAL OF ACCOUNTING STUDIES 451 Table 6. t ax incentives, adjustment costs, and r&d investment: State-owned vs. non-state-owned. (8) State-owned (9) Non-state-owned Variable Coefficient t value Coefficient t value ** *** Intercept 0.039 2.29 0.076 3.40 Size 0.001 1.30 0.000 0.04 ** Leverage −0.004 −1.27 −0.005 −2.45 Cash 0.016 1.88 0.006 0.69 ** ** Age −0.000 −2.56 −0.000 −2.57 ROA 0.003 0.35 0.003 0.33 Tenure 0.000 0.36 −0.000 −0.91 *** *** B_index 0.048 3.55 0.059 3.23 * ** Adjustment −3.475 −1.67 −4.111 −2.56 * *** B_index × Adjustment −3.911 −1.69 −4.663 −2.60 Industry Yes Yes obs. 695 907 a dj. R 0.063 0.079 Prob. > F 0.000 0.000 ** diff. of B_index 0.011 (State-owned vs. non-state-owned) diff. of Adjustment −0.636 (State-owned vs. non-state-owned) *** diff. of B_index × Adjustment −0.751 (State-owned vs. non-state-owned) notes: t he dependent variable is Change_rd. ***, **, and * indicate statistical significance at the 1, 5, and 10% levels, respectively. firms is less than for non-state-owned firms (i.e., 0.012 < 0.013), implying that R&D tax incen- tives policies do not have an incentive effect (i.e., U = 0) when state-owned firms have an Adjustment value of 0.012. However, this incentive effect does not disappear until Adjustment is larger than 0.013 in non-state-owned firms. These results support our inference that R&D tax incentives policies have a greater incentive effect on non-state-owned than on state- owned firms. 6. Conclusions As the source of technological innovation, R&D is the key to achieving and maintaining a competitive advantage in science and technology. Tax incentives to stimulate the R&D activ- ities of firms are an important tool for governments trying to solve the “market failure” of R&D investment. However, previous studies of the incentive effects of tax policies on R&D investment have not reached a consensus. Overall, existing research on R&D tax incentives policy ignores the effects of adjustment costs on R&D investment, particularly in emerging and transitional developing countries such as China. This study of China considers the effects of adjustment costs on the incentive effects of R&D tax policies. We find that R&D tax incentives stimulate firms to invest in R&D. However, the incentive effects decrease gradually with the increase in adjustment costs. Tax incentive effects vanish when adjustment costs per unit of capital exceed the critical value of 0.011. Currently, nearly 93% of firms in China have adjustment costs that are less than this critical value, implying that R&D tax incentives policies in China are generally effective. Furthermore, we find that the incentive effects are significantly different for firms with different types of property rights. Specifically, incentive effects are more significant in non-state-owned firms than in state-owned firms and the critical value of adjustment costs in the former is higher. 452 W. LI AND J. DU Using a sample of Chinese firms, this study demonstrates why previous studies have been unable to come to consistent conclusions about the effects of R&D tax incentives. Our study enriches the existing literature on adjustment costs, R&D investment (e.g., Cooper & Haltiwanger, 2006; David & Hall, 2000), R&D tax incentives, and innovation. We provide a micro-level analysis of the incentive effects of R&D tax policies in China. Our findings can be used to improve the evaluation and optimisation of R&D investment incentive policies for firms. This study also provides a reference for decision makers trying to promote independent R&D of firms and the social progress of science and technology. Acknowledgements We appreciate the valuable comments of Prof. Zhiyuan Liu and the other participants at the 2013 Conference of the China Journal of Accounting Studies in Guangzhou. We are especially grateful for the detailed and constructive suggestions of the anonymous referees. The authors gratefully acknowl- edge the financial support of the Key Project of the National Natural Science Foundation of China (71332004), the General Project of the National Natural Science Foundation of China (71472047, 71272198, 71572038), the Young Scholar Research Project of the Ministry of Education (13YJC630080), the Project of the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), the A-level Social Science Research Project of the Department of Education in Fujian province (JA13047S), and the Social Science Planning Project in Fujian province (2014B022), the Training Program for Accounting Masters in Ministry of Finance of China ([2016]15), and the project of special fund of basic scientific research business expenses from Sun Yat-sen University (16wkjc01). References Abel, A.B. (1983). Optimal investment under uncertainty. American Economic Review, 73, 228–233. Abel, A.B., & Eberly, J.C. (1994). A unified model of investment under uncertainty. American Economic Review, 84, 1369–1384. Abel, A.B., & Eberly, J.C. (1998). The mix and scale of factors with irreversibility and fixed costs of investment. Carnegie-Rochester Conference Series on Public Policy, 48, 101–135. Aghion, P., & Howitt, P. (1992). A model of growth through creative destruction. Econometrica, 60, 323–351. Allen, F., Qian, J., & Qian, M. (2005). Law, finance, and economic growth in China. Journal of Financial Economics, 77, 57–116. Arrow, K. (1962). The economic implication of learning by doing. Review of Economic Studies, 29, 155– An, T.L., Shi, H., & Ludovico, A. (2006). An observation and empirical study of R&D behavior of Chinese manufacturing firms: Based on a survey of the manufacturing firms in Jiangsu province. Economic Research Journal, 41 (2), 21–30 (In Chinese). An, T.L., Zhou, S.D., & Pi, J.C. (2009). The stimulating effects of R&D subsidies on independent innovation of Chinese enterprises. Economic Research Journal, 44 (10), 87–98 (In Chinese). Baily, M., & Lawrence, R. (1992). Tax incentives for R&D: What do the data tell us? Washington, DC: Study Commissioned by the Council on Research and Technology. Baum, C.F., Schaffer, M.E., & Stillman, S. (2003). Instrument variables and GMM: Estimation and testing. Stata Journal, 3(1), 1–31. Baum, C.F., Schaffer, M.E., & Stillman, S. (2007). Enhanced routines for instrumental variables/GMM estimation and testing. Stata Journal, 7, 465–506. Bhagat, S., & Welch, I. (1995). Corporate research & development investments, international comparisons. Journal of Accounting and Economics, 19, 443–470. Berger, P. (1993). Explicit and implicit effects of the R&D tax credit. Journal of Accounting Research, 31, 131–171. CHINA JOURNAL OF ACCOUNTING STUDIES 453 Bernstein, J.I., & Nadiri, M.I. (1982). Financing and investment in plant and equipment and research and development (NBER Working Paper). Cambridge: National Bureau of Economic Research. Bernstein, J.I. (1986). The effect of direct and indirect tax incentives on Canadian industrial R&D expenditures. Canadian Public Policy, 12, 438–448. Bloom, N., Griffith, R., & Van Reenen, J. (2002). Do R&D tax credits work? Evidence from a panel of countries 1979–1997. Journal of Public Economics, 85 (1), 1–31. Broadway, R.W., Chua, D., & Flatters, F. (1995). Investment incentives and the corporate tax system in Malaysia. Oxford: Oxford University Press. Busom, I. (2000). An empirical evaluation of the effects of R&D subsidies. Economics of Innovation and New Technology, 9, 111–148. Brown, V. (1999). Innovation battle. Beijing: China Machine Press. Caballero, R.J. (1991). On the sign of the investment uncertainty relationship. American Economic Review, 81, 279–288. Caballero, R.J., Engel, E.M.R.A., & Haltiwanger, J. (1997). Aggregate employment dynamics: Building from microeconomic evidence. American Economic Review, 87, 115–137. Chen, M.C., & Gupta, S. (2010). The incentive effects of R&D tax credits: An empirical examination in an emerging economy (SSRN Working Paper). Taipei: National Chengchi University. Christiano, L., & Todd, R. (1996). Time to plan and aggregate fluctuations. Federal Reserve Bank of Minneapolis Quarterly Review, 20, 14–27. Cohen, W.M., & Levin, R. (1989). Empirical studies of innovation and market structure. Handbook of Industrial Organization, 2, 1059–1107. Cooper, R.W., & Haltiwanger, H. (2006). On the nature of capital adjustment costs. Review of Economic Studies, 73, 611–633. Czarnitzki, D., Hanel, P., & Rosa, J.M. (2011). Evaluating the impact of R&D tax credits on innovation: A microeconometric study on Canadian firms. Research Policy, 40, 217–229. David, P.A., & Hall, B.H. (2000). Heart of darkness: Modeling public–private funding interactions inside the R&D black box. Research Policy, 29, 1165–1183. Dixit, A., & Pyndick, R.S. (1994). Investment under uncertainty. Princeton, NJ: Princeton University Press. Edge, R. (2000). The effect of monetary policy on residential and structures investment under differential project planning and completion times (Discussion Paper). Washington, DC: Federal Reserve Board. Eisner, R., Albert, S., & Sullivan, M. (1984). The new incremental tax credit for R&D: Incentive or disincentive? National Tax Journal, 37, 171–185. Ernst, C., Richter, K., & Riedel, N. (2014). Corporate taxation and the quality of research and development. International Tax and Public Finance, 21, 694–719. Estache, A., & Gaspar, V. (1995). Why tax incentives do not promote investment in Brazil. Oxford: Oxford University Press. Feng, G. (2004). Double principal-agent theory: Another analysis framework of listed companies’ governance. Economic Research Journal, 39 (12), 16–25 (In Chinese). Gould, J.P. (1968). Adjustment costs in the theory of investment of the firm. The Review of Economic Studies, 35, 47–55. Grossman, G.M., & Helpman, E. (1991). Innovation and growth in the global economy. Cambridge, MA: MIT Press. Groth, C., & Khan, H. (2010). Investment adjustment costs: An empirical assessment. Journal of Money, Credit and Banking, 42, 1469–1494. Griffith, R., Sandler, D., & Van Reenen, J. (1995). Tax incentives for R&D. Fiscal Studies, 16, 21–44. Guellec, D., & Van Pottelsberghe, B. (2003). The impact of public R&D expenditures on business R&D. Economics of Innovation and New Technologies, 12, 225–244. Hall, B.H., Griliches, Z., & Hausman, J.A. (1986). Patents and R&D: Is there a lag? International Economic Review, 27, 265–284. Hall, B.H., & Hayashi, F. (1988). Research and development as an investment (NBER Working Paper). Cambridge: National Bureau of Economic Research. Hall, B.H. (1993). R&D tax policy during the eighties: Success or failure? Tax Policy and the Economy, 7, 1–36. 454 W. LI AND J. DU Hall, B.H. (1995). Effectiveness of research and experimentation tax credits: Critical literature review and research design. Washington, DC: Congress of the United States, Report to Office of Technology Assessment. Hall, B.H., & Van Reenen, J. (2000). How effective are fiscal incentives for R&D? A review of the evidence. Research Policy, 29, 449–469. Hall, B.H. (2002). The financing of research and development (NBER Working Paper). Cambridge: National Bureau of Economic Research. Hall, R.E. (2004). Measuring factor adjustment costs. Quarterly Journal of Economics, 119, 899–927. Hassett, K.A., & Hubbard, R.G. (1996). Tax policy and investment (NBER Working Paper). Cambridge: National Bureau of Economic Research. He, J. (1998). An empirical study on listed firms’ corporate governance structure. Economic Research Journal, 33 (5), 50–57 (In Chinese). Himmelberg, C.P., & Petersen, B.C. (1994). R&D and internal finance: A panel study of small firms in high-tech industries. The Review of Economics and Statistics, 76, 38–51. Huang, J., & Chen, X.Y. (2011). Business group and R&D investment: An analysis from the perspective of knowledge spillover and internal capital market. Economic Research Journal, 46 (6), 80–92 (In Chinese). Jefferson, G.H., Bai, H., Guan, X., & Yu, X. (2006). R&D performance in Chinese industry. Economics of Innovation and New Technology, 15, 345–366. Khanna, T., & Iansiti, M. (1997). Firm asymmetries and sequential R&D: Theory and evidence from the mainframe computer industry. Management Science, 43, 405–421. Lee, C.Y. (2002). Industry R&D intensity distributions: Regularities and underlying determinants. Journal of Evolutionary Economics, 12, 307–341. Lichtenberg, F.R. (1984). The relationship between federal contract R&D and company R&D. American Economic Review, 74, 73–78. Lichtenberg, F.R. (1988a). Estimation of the internal adjustment costs model using longitudinal establishment data. Review of Economics and Statistics, 70, 421–430. Lichtenberg, F.R. (1988b). The private R&D investment response to federal design and technical competitions. American Economic Review, 78, 550–559. Lucas, R.E. (1967). Adjustment costs and the theory of supply. The Journal of Political Economy, 75, 321–334. Li, H.D., Tang, Y.J., & Zuo, J.J. (2013). Does innovation depend on your own money or others’ money? Journal of Financial Research, (2), 170–183 (In Chinese). Li, L.Q. (2007). The effectiveness of current tax incentives on enterprises’ R&D investment in China. China Soft Science Magazine, (7), 115–120 (In Chinese). Liu, W., & Liu, Y.G. (2007). The executive term and R&D expenses of listing corporation in China. Management World, (1), 128–136 (In Chinese). Mamuneas, T.P., & Nadiri, M.I. (1996). Public R&D policies and cost behavior of the US manufacturing industries. Journal of Public Economics, 63, 57–81. Mansel fi d, E. ( 1986). The R&D tax credit and other technology policy issues. American Economic Review, 76, 190–194. Mussa, M. (1977). External and internal adjustment costs and the theory of aggregate and firm investment. Economica, 174, 163–178. Pan, Y., Dai, Y., Wu, C., & Liu, J. (2009). Social capital, political connections and corporate investment decision. Economic Research Journal, 44 (11), 82–94 (In Chinese). Pavitt, K., Robson, M., & Townsend, J. (1987). The size distribution of innovating firms in the UK: 1945– 1983. Journal of Industrial Economics, 35, 273–288. Pindyck, R.S., & Rotemberg, J.J. (1983). Dynamic factor demands and the ee ff cts of energy price shocks. American Economic Review, 73, 1066–1079. Pindyck, R.S. (1993). A note on competitive investment under uncertainty. American Economic Review, 83, 273–277. Polder, M., & Verick, S. (2004). Dynamics of labor and capital adjustment—A comparison of Germany and the Netherlands (IZA Discussion Paper). Bonn: Institute for the Study of Labor. CHINA JOURNAL OF ACCOUNTING STUDIES 455 Qin, J., & Shao, X. (2015). Cross-listings, government intervention, and efficiency of capital allocation. Economic Research Journal, 50 (6), 117–130 (In Chinese). Romer, P. (1990). Endogenous technological change. Journal of Political Economy, 98, 71–102. Romer, D. (2006). Advanced macroeconomics (3rd ed.). New York, NY: The McGraw-Hill Company. Scholes, M.S., Wolfson, M.A., Erickson, M.M., Maydew, E.L., & Shevlin, T.J. (2005). Taxes and business strategy: A planning approach. New Jersey, NJ: Pearson Prentice Hall. Shapiro, M.D. (1986). The dynamic demand for capital and labor. Quarterly Journal of Economics, 101, 513–542. Summers, L.H. (1981). Taxation and corporate investment: A q-theory approach. Brookings Papers on Economic Activity, 12, 67–140. Tang, Q.Q., Lu, S.S., & Li, Y.D. (2008). Enterprises as innovative entities and government roles in R&D subsidies. China Soft Science, (6), 88–98 (In Chinese). Wallsten, S.J. (2000). The effects of government-industry R&D programs on private R&D: The case of the small business innovation research program. The RAND Journal of Economics, 31, 82–100. Warda, J. (1994). Canadian R&D tax treatment: An international comparison. Ottawa: Report, the Conference Board of Canada. Warda, J. (1996). Measuring the value of R&D tax provisions, s fi cal measures to promote R&D and innovation (pp. 9–22). Paris: OECD. Warda, J. (2001). Measuring the value of R&D tax treatment in OECD Countries. STI Review, 27, 186–211. Warda, J. (2006). Tax treatment of business investments in intellectual assets: An international comparison (OECD Science, Technology and Industry Working Papers). Paris: OECD. Wang, Y.T., Wang, L.L., & Peng, Y. (2010). Ownership nature of ultimate controller, debt-related tax shields and capital structure. Economic Research Journal, 45 (9), 122–136 (In Chinese). Wang, R.F. (2005). The internal determinants of firm R&D expenditures: Evidence from top 100 companies of Chinese electronic & information industry. Studies in Science of Science, 23 (2), 225–231 (In Chinese). Wang, J. (2011). The measure of government R&D tax incentive intensity and its effect verification in China. Science Research Management, 32 (9), 157–164 (In Chinese). Wu, L.S. (2009). State ownership, tax incentives, and corporate tax burdens. Economic Research Journal, 44 (1), 109–120 (In Chinese). Wu, X.B. (2003). Impact of revenue stimulation on R&D investment: Empirical analysis and policy implement selection. R&D Management, 15 (1), 36–41 (In Chinese). Wu, Y.B. (2006). R&D and productivity: An empirical study on Chinese manufacturing industry. Economic Research Journal, 41 (11), 60–71 (In Chinese). Xie, W.M., Tang, Q.Q., & Lu, S.S. (2009). Governments’ R&D subsidies, firms’ R&D investment and self- dependent innovation. Journal of Financial Research, (6), 86–99 (In Chinese). Xu, X., & Tang, Q.Q. (2012). R&D, technology in draught and industry changes of main business: An empirical research on Chinese manufacturing listed companies. Journal of Financial Research, (10), 193–206 (In Chinese). Yao, Y., & Zhang, Q. (2001). An analysis of technological efficiency of Chinese industrial firm. Economic Research Journal, 36 (1), 66–79 (In Chinese). Zhu, P.F., & Xu, W.M. (2003). On the impact of government’s S&T incentive policy on the R&D input and its patent output of large and medium-sized Industrial enterprises in Shanghai. Economic Research Journal, 38 (6), 45–53 (In Chinese). Zhao, F.M., Su, S.A., & Zou, S.G. (2006). Analysis on the impacts of S&T investment of Chinese government to that of large and medium-sized industrial enterprises. R&D Management, 18 (2), 78–84 (In Chinese). Zhang, X.Z., Liu, Z.Y., & Wang, J. (2012). The analysis of the relationship between firm size and R&D investment based on the firm’s profitability. Studies in Science of Science, 30 (2), 265–274 (In Chinese).
http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png
China Journal of Accounting Studies
Taylor & Francis
http://www.deepdyve.com/lp/taylor-francis/tax-incentives-adjustment-costs-and-r-d-investment-in-china-FUHF03dgRL
