Tax Burden and Corporate Investment Efficiency

Abstract

Using A-share listed companies in Shanghai and Shenzhen from 2015 to 2021 as the research sample, a fixed-effects model was used to examine the effect of the reduction of corporate tax burden on investment efficiency under the tax reduction policy, as well as the role of tax avoidance and financing constraints in the mechanism. The results of the study show that the reduction of tax burden can effectively improve the efficiency of corporate investment, and this positive effect is reflected in the alleviation of corporate under-investment and discouragement of over-investment. The paper also analyses the mechanism through which tax burden affects the efficiency of corporate investment, and finds that tax reduction can discourage inefficient investment by reducing corporate tax avoidance and alleviating corporate financing constraints. In further analysis, it is found that the effect of tax cuts on investment efficiency is more significant in the sample of non-state enterprises, low corporate governance and low marketisation. The findings of the study support the positive significance of the current tax reduction policy. We provide a reference of tax reduction benefits to curb tax avoidance behavior, and provide a basis for relevant policy departments to further accelerate the implementation of tax reduction policies.

1. Introduction

At present, China is in a phase of intensifying downward economic pressure and slowing economic growth momentum, and the demand for liquidity within companies is also increasing. In order to develop and grow new dynamic energy and promote structural reform on the supply side, China has, in recent years, been committed to giving new impetus to the development of enterprises by reducing taxes and fees for the benefit of enterprises. While deepening the “camp reform” tax system reform, promoting the synergy between general tax reduction and structural tax reduction is an important measure to reduce taxes and fees in the new phase. According to data from the National Bureau of Statistics, the sum of VAT and corporate income tax, which are the main tax expenditures of enterprises in China, decreased by 4.17% of GDP year-on-year in 2019, and, even though GDP growth will slow down under the impact of the new pneumonia epidemic in 2020, the proportion of VAT and corporate income tax to GDP still decreased by 9.21% year-on-year, indicating that China’s tax and fee reduction has achieved remarkable results.

Past papers have shown that tax cuts have a significant impact on improving the efficiency of government spending and improving the welfare of residents [1]. However, the objectives of tax reduction policies are not only macroeconomic considerations, but also the desire to promote the development of the real economy. By reducing the tax burden on enterprises, it serves to encourage investment and expand domestic demand, promote the upgrading of industrial structure, further improving the anti-risk capacity and competitiveness of enterprises, and increasing the value of enterprises. It has been shown that tax burden reduction has a catalytic effect on enterprise R&D investment, innovation output and investment scale [2,3,4,5]. So, the research questions of this paper include: Does the tax burden have a significant impact on investment efficiency [6], which has an important role in the transformation of the economic development of enterprises?

The contribution of this paper is reflected in the following two aspects: First, the relationship between tax burden and investment efficiency is studied in the context of large-scale tax reduction policies. Studies have focused more on the positive effect of tax incentives on enterprises’ R&D investment, ignoring the important impact of tax burden on investment efficiency, which plays an important role in transforming enterprises’ economic development mode and improving enterprise value. This paper further provides empirical evidence for the positive impact of tax reduction policies by conducting a theoretical analysis and empirical test on the mechanism of the role between tax burden and investment efficiency. Secondly, the study of the impact of tax burden reduction on micro enterprises is extended from the perspective of tax avoidance behavior. By verifying the role played by tax avoidance behavior in the impact of tax burden on investment efficiency, we provide a reference for enterprises to make full use of tax reduction benefits to curb tax avoidance behavior, and thus alleviate internal information asymmetry, and provide a basis for relevant policy departments to further accelerate the implementation of tax reduction policies.

2. Theoretical Analysis and Research Hypothesis

Previous studies have shown that tax burden reduction has an impact on enterprise research and development investment [7], investment scale and innovation output, etc., all have a promoting effect.

Tax Burden and Investment Efficiency

The impact of the tax burden on investment efficiency is reflected in three main areas: policy regime, regulators and financing.

First, from the perspective of policies and systems, the tax reduction policy aims to stimulate market vitality and reduce the burden on enterprises. At present, our country is in the stage of economic growth slowing down. Corporate taxation is an important measure for the government to interfere in corporate activities. The future direction of tax reduction policy will have a very obvious impact on corporate investment. In China, the “Replacing the Business Tax with A Value-Added Tax” policy, which has achieved remarkable results, has promoted the improvement of enterprise investment efficiency by reducing the degree of financing constraints (Qian Xiaodong, 2018); the accelerated depreciation policy for fixed assets has significantly promoted the investment in fixed assets of pilot enterprises by reducing the cash flow pressure of enterprises [8]. Since 2015, with the continuous deepening of the structural reform on the supply side, the “cutting overcapacity, de-stocking, de-leveraging, reducing costs and improving weak links” has achieved remarkable results. As part of the “cost reduction”, tax reduction and fee reduction have played a direct and important role in reducing the burden on enterprises and promoting the transformation of old and new kinetic energy of enterprises (<Tax Reduction and Fee Reduction: giving enterprises a greater sense of “access”>, Economic Daily, 23 May 2017). At the beginning of 2018, the Government Work Report set a tax and fee reduction target of RMB 1.1 trillion in scale, and has since introduced a series of measures to promote the development of the real economy and support scientific and technological innovation, bringing the scale of tax and fee reduction to more than RMB 1.3 trillion for the year (Larger tax cuts necessary and feasible, State Council, 10 October 2018). The deepening of the VAT reform and the revision of the personal tax law have become a major leap forward in tax reform, with significant tax reduction effects. From Figure 1, it can be seen that the tax burden of enterprises in China began to decline significantly in 2015, and has been showing a significant downward trend since then. The decline was smaller in 2020, due to the impact of the new crown pneumonia epidemic, and reached 9.33% in 2021 afterwards. In March 2022, the Government Work Report further pointed to increase the scale of tax cuts and implement new tax cuts. The tax cuts are aimed at keeping more cash flow in enterprises, enabling companies to develop, expanding social aggregate demand and easing weak demand during the economic downturn. The tax expenditure savings provide more incentive for companies to invest, in which case they can use the tax expenditure savings to improve the under-investment caused by the shortage of funds, thus increasing investment efficiency.

Second, from the perspective of the company’s managers, the additional cash flows generated by the lower tax burden can be used to motivate managers to fully exploit their talents to make scientific investment decisions. Managers and the exercise of their competencies are important factors affecting investment efficiency [9,10,11,12]. Based on the principal-agent theory, it is believed that, due to the principal–agent conflict between managers and shareholders, in the process of corporate decision-making, managers will use their own information advantages to invest in projects that are beneficial to them but will damage corporate value, such as over-investment caused by investing in projects with negative net present value, or under-investment caused by abandoning projects with positive net present value. However, with the development of sociology and psychology, managers will consider their future social reputation when making investment decisions [13], and a decline in performance, or even a financial crisis, as a result of inefficient investment may cause more serious damage to the manager. Therefore, compared to the role of an agent, managers prefer to take on the role of a steward when they are incentivized with tax savings, and by making full use of their talents [14], they can rationalize the allocation of corporate resources and improve investment efficiency.

Thirdly, from the perspective of corporate financing, a decrease in corporate tax burden and an increase in corporate cash flows are positive signals for external investors, in which case investors’ willingness to invest is enhanced and companies are more likely to obtain financing. In terms of financing sources, on the one hand, the increase in cash flow brought about by the tax cut can effectively alleviate the endogenous financing constraint of enterprises, increase the level of corporate profits and improve the liquidity position of enterprises [15]. On the other hand, the improvement of corporate profit level and liquidity position is important for stimulating corporate R&D and innovation [16,17,18], “deleveraging” and “cost reduction” [19], and enhancing firm productivity have significant positive effects; the above effects can enhance the long-term competitiveness of firms and provide growth momentum for their development [20]. Based on signaling theory, a company’s future growth prospects can increase the confidence of external investors, which in turn makes it easier to attract and obtain external financing. For non-state enterprises in particular, the existence of strong financing constraints leads to higher financing costs, compared to state-owned enterprises which can borrow at will at a set interest rate. Tax cuts increase their internal financing, while also providing them with greater possibilities to alleviate their external financing constraints.

In summary, tax cuts and their positive signals to investors can alleviate corporate under-investment, while the incentive of tax expenditure savings to managers can alleviate over-investment and, taken together, enhance corporate investment efficiency. Based on this, this paper proposes the following hypothesis.

Hypothesis 1 (H1). 

A lower tax burden helps to improve the efficiency of corporate investment.

3. Research Design

3.1. Samples and Data

This paper selects the companies listed on the Shanghai and Shenzhen A-share main board from 2015 to 2021 as the initial sample, and filters and processes the data according to the following criteria: (1) Excluding listed companies that were ST and PT during the sample period; (2) Considering the special characteristics of the financial industry, the financial statements of these companies were different from other industries, so all listed companies in the financial industry were excluded; (3) Excluding the main; (4) In order to calculate the residuals of Richardson’s investment efficiency model, we need to use the data from the previous period, i.e., the sample used in this paper is actually from 2014–2021; (5) In order to reduce the impact of outliers on the regression results, all variables are winsorize shrunken in 1% and 99% of the quartiles. After the above treatments, we finally selected 1357 listed companies with a total of 6042 observations. The financial and market data used in the article were obtained from the CSMAR database, except for the marketability index. This article uses stata15.1 to process the data and test the regression model.

3.2. Variable Definition

3.2.1. Explained Variables (Inefficient Investment)

This paper draws on the research methods of Yuan Zhizhu et al. (2020) and Zhao Na et al. (2019) to measure investment efficiency, using the model constructed by Richardson (2006), as shown in Model (1) [21]. By conducting OLS regressions on Model (1) and calculating the residuals, when the residuals are greater than 0, the actual investment is higher than the optimal value, i.e., over-investment, and when they are less than 0, the actual investment is lower than the optimal value, i.e., under-investment. The residuals of Model (1) are taken as the absolute value of the explanatory variable Inveffi, which indicates the absolute amount of deviation of the actual investment scale from the optimal scale, i.e., the degree of inefficient investment: the larger the value, the less efficient the investment. Meanwhile, this paper uses Over_Inv and Under_Inv to represent the degree of over-investment and under-investment of the company, respectively, to explore the impact of tax burden on the investment efficiency of the company, respectively.

$$\begin{gathered} In{v}_{i,t}={\beta }_0+{\beta }_{1}Taxburde{n}_{i,t}+{\beta }_{2}Tobin{Q}_{i,t-1}+{\beta }_{3}Le{v}_{i,t-1}++{\beta }_{4}Cas{h}_{i,t-1}+{\beta }_{5}Siz{e}_{i,t-1}+{\beta }_{6}Ag{e}_{i,t-1} \\ +{\beta }_{7}Re{t}_{i,t-1}+{\beta }_{8}In{v}_{i,t-1}{\displaystyle \sum }Year+{\displaystyle \sum }Industry+{\epsilon }_{i,t} \end{gathered}$$

(1)

where ${\mathrm{Inv}}_{\mathrm{i},\mathrm{t}}$ represents the new investment expenditure of Company I in Year t, which is equal to the ratio of the difference between the cash paid for the new construction of fixed assets, intangible assets and other long-term assets in Year t, and the cash recovered from the disposal of fixed assets, intangible assets and other long-term assets to the total assets at the beginning of the year; ${\mathrm{Taxburden}}_{\mathrm{i},\mathrm{t}}$ is the ratio of the difference between the various taxes paid and the tax rebates received to its operating income; ${\mathrm{Tobin}\mathrm{Q}}_{\mathrm{i},\mathrm{t}-1}$ is the Tobin Q value of Company I in Year t−1, which is equal to the ratio of market value to total assets and represents the investment opportunities faced by Company i; ${\mathrm{Lev}}_{\mathrm{i},\mathrm{t}-1}$ represents the financial leverage of Company I in Year t−1, which is equal to the ratio of total liabilities to total assets; ${\mathrm{Cash}}_{\mathrm{i},\mathrm{t}-1}$ represents Company I’s monetary cash holdings in Year t−1, which is equal to monetary cash divided by total assets; ${\mathrm{Size}}_{\mathrm{i},\mathrm{t}-1}$ represents the size of Company I in Year t−1, which is equal to the natural logarithm of total assets; ${\mathrm{Age}}_{\mathrm{i},\mathrm{t}-1}$ represents the natural logarithm of Company I’s IPO years up to the end of t−1; ${\mathrm{Ret}}_{\mathrm{i},\mathrm{t}-1}$ represents the share yield of Company I in Year t−1. In addition, the model controls for year and industry dummy variables.

3.2.2. Explanatory Variables (Tax Burden)

Drawing on the methodology of Liu Jun (2014) and Deng Liping (2020) [22], this paper measures a company’s tax burden as the ratio of the difference between the various taxes paid and the tax rebates received to its operating income, with a larger value indicating a higher level of tax burden borne by the company.

3.2.3. Grouping Variables

In conjunction with the above analysis, this paper conducts group tests on the impact of tax burden on investment efficiency in terms of the nature of property rights, the level of corporate governance and the degree of marketisation, respectively.

3.2.4. Controlling Variables

Referring to studies on investment efficiency [23], this paper introduces the following control variables: firm size (Size), growth rate of operating income (Growth), financial leverage (Lev), years of listing (Age), corporate free cash flow (Fcf), cash flow from operating activities (Cfo), stock return (Rate), proportion of independent directors (Dr), dual role (Dual) and proportion of top shareholder ownership (Top 1). Finally, to control for the effects of year and industry, year and industry dummy variables are added to the model in this paper.

Table 1. Variable Definitions.

Variable Type	Variable Name	Variable Definition
Explained variables	Inveffi	Taking the absolute value of the residuals of Richardson investment efficiency model
	Over_Inv	Degree of over-investment, taking the positive residual of Richardson investment efficiency model
	Under_Inv	Degree of under-investment, taking the absolute value of the negative residual of Richardson’s investment efficiency model
Explanatory Variables	Taxburden	Tax burden, the ratio of the difference between the various taxes paid and the tax refunds received to operating income
Intermediate variables	BTD	Extent of tax avoidance as measured by accounting-tax differences
Control Variables	Size	Natural logarithm of the company’s total assets
	Growth	Growth rate of the company’s gross operating income
	Lev	Ratio of total liabilities to total assets of the Company
	Age	Natural logarithm of the number of years the company has been listed
	Fcf	EBIT + depreciation and amortization—increase in working capital—capital expenditure
	Cfo	Ratio of net cash flows from operating activities to total assets of the Company
	Rate	Return on equity, considering annual equity return on reinvestment of cash dividends
	Dr	Percentage of independent directors, independent directors as a proportion of the number of directors on the board
	Dual	Dummy variable, 1 if both chairman and managing director, 0 otherwise
	Top 1	Percentage of shareholding of the largest shareholder
	Year	Annual dummy variables
	Industry	Industry dummy variable
Grouping Variables	SOE	Nature of ownership of listed companies, with state-owned enterprises taking a value of 1, otherwise 0
	GOVER	Indicators of the level of corporate governance
	GOV	Marketization Index of Fan Gang

provides a description of the definitions of the key variables.

3.3. Model Setting

Based on the above analysis, this paper develops Model (2) to empirically test the relationship between tax burden and investment efficiency, i.e., to test hypothesis 1.

$$Inveff{i}_{i,t}={\beta }_0+{\beta }_{1}Taxburde{n}_{i,t}+{\beta }_{2}Control{s}_{i,t}+{\displaystyle \sum }Year+{\displaystyle \sum }Industry+{\epsilon }_{i,t}$$

(2)

In Model (2), the explanatory variable Inveffi represents the firm’s investment efficiency, measured as in Model (1); the explanatory variable Taxburden represents the level of the firm’s tax burden. In addition to adding the control variables described above, this paper uses the method of adding year and industry fixed effects to the regression model to avoid the effect of omitted variables. Also, the paper uses a Hausman test to select panel fixed effects or random effects, with a result of 257.41, thus rejecting the original hypothesis and using a fixed effects model. The robustness of the results is initially ensured by including robust standard errors for the clustering of firms in the regression model.

In Model (2), we focus on the coefficient of tax burden (Taxburden) i.e., the value. When significantly positive, it means that, as the level of tax burden of the company decreases, the level of inefficient investment is suppressed and investment efficiency will be improved, i.e., a lower tax burden has a significant positive contribution to the investment efficiency of the company.

3.4. Descriptive Statistical Analysis

Table 2. Descriptive statistics of variables.

	VAR	Mean Value	Median	Standard Deviation	Minimum Value	Maximum	Sample Size
Explained variables	Inveffi	0.035	0.021	0.044	0.000	0.292	6042
	Over_Inv	0.046	0.021	0.069	0.000	0.397	2358
	Under_Inv	0.028	0.021	0.027	0.001	0.147	3684
Explanatory variable	Taxburden	0.070	0.051	0.071	−0.053	0.359	6042
Control variable	Size	22.980	22.820	1.389	20.010	26.650	6042
	Growth	0.149	0.083	0.421	−0.546	2.789	6042
	Lev	0.485	0.490	0.199	0.086	0.902	6042
	Age	2.649	2.890	0.663	0.693	3.296	6042
	Fcf	0.011	0.019	0.092	−0.336	0.240	6042
	Cfo	0.052	0.050	0.068	−0.162	0.253	6042
	Rate	0.058	−0.035	0.435	−0.531	2.410	6042
	Dr	0.374	0.364	0.055	0.300	0.571	6042
	Dual	1.829	2.000	0.376	1.000	2.000	6042
	Top1	0.368	0.354	0.150	0.088	0.767	6042
Grouping Variable	SOE	0.612	1.000	0.487	0.000	1.000	6042
	GOVER	−0.419	−0.617	0.771	−1.797	2.082	5662
	GOV	8.525	9.140	1.964	−1.770	12.630	6042

is a descriptive statistic of the variables in this paper, in which the mean value of the interpreted variable’s ineffective investment level (Inveffi) is 0.035, the median value is 0.021, the maximum value and the minimum value are 0.292 and 0, respectively, and the standard deviation is 0.044. In this sample, the proportion of over-investment sample is 39.03%, and the proportion of under-investment sample is 60.97%, which indicates that under-investment exists generally in the listed companies in Shanghai and Shenzhen A-shares. The mean value of the over-investment sample is 0.046, while the mean value of the under-investment sample is 0.028, indicating that the problem of over-investment is more serious among the listed companies in Shanghai and Shenzhen A-shares in China, although the under-investment sample accounts for a relatively large proportion. The mean and median of the explanatory variables of this paper, Taxburden, are 0.070 and 0.051, respectively, indicating that the average tax burden of listed companies in the sample accounts for 7.1% of business income. Also noteworthy are the indicators of the level of corporate governance and the degree of marketability. The mean value of corporate governance level is −0.419, the median value is −0.617 and the standard deviation is 0.771, indicating that the current governance level of A-share listed companies in China is generally low, while there is a large gap; the mean value of marketization degree is 8.525, the median value is 9.140 and the standard deviation is 1.964, indicating that the development level of marketization degree in China varies from region to region, and there is still much room for improvement in some regions. The results of the descriptive statistics of other control variables are generally consistent with the existing studies.

4. Empirical Results and Analysis

4.1. The Impact of Tax Burden on the Efficiency of Corporate Investment

Table 3. The Impact of Tax Burden on Investment Efficiency.

VAR	(1)	(2)	(3)
	Inveffi	Under_Inv	Over_Inv
Taxburden	0.048 ***	0.025 ***	0.102 ***
	(2.591)	(2.832)	(3.179)
Size	0.028 ***	−0.002 ***	−0.000
	(7.714)	(−3.280)	(−0.022)
Growth	0.022 ***	0.004 **	0.055 ***
	(8.250)	(2.476)	(10.072)
Lev	−0.002	−0.006 *	0.017
	(−0.202)	(−1.876)	(1.253)
Age	−0.004	−0.005 ***	−0.008 **
	(−0.575)	(−5.452)	(−2.526)
Fcf	−0.025 ***	−0.017 ***	−0.020
	(−3.039)	(−2.630)	(−1.162)
Cfo	−0.019 **	−0.018 **	−0.039
	(−2.050)	(−2.564)	(−1.629)
Rate	0.000	−0.002	0.001
	(0.115)	(−1.476)	(0.254)
Dr	−0.000	0.007	−0.032
	(−0.008)	(0.874)	(−1.155)
Dual	−0.001	0.000	−0.007 *
	(−0.295)	(0.148)	(−1.843)
Top1	0.007	−0.004	−0.003
	(0.430)	(−1.066)	(−0.220)
Constant	−0.586 ***	0.092 ***	0.092 **
	(−7.407)	(7.465)	(2.094)
Annual Effect	Yes	Yes	Yes
Industry Effect	Yes	Yes	Yes
Obs	6042	3684	2358
R-squared	0.157	0.108	0.252

Note: ***, **, * denote estimated coefficients significant at the 1%, 5% and 10% levels, and values in brackets are t-statistics or F-statistics for the corresponding coefficients.

presents the regression results of the impact of tax burden on the efficiency of business investment. In particular, column (1) is a full sample regression. From the results in column (1), it can be seen that the coefficient of Taxburden is 0.048 and significant at the 1% level, indicating that when the tax burden of enterprises is reduced, their inefficient investment phenomenon is significantly suppressed, i.e., investment efficiency is improved, and there is a significant positive promotion effect of tax reduction on the investment efficiency of enterprises, and hypothesis 1 is verified; in terms of economic significance, when enterprises reduce of tax reaches 1% of business income, investment efficiency will increase by 4.8%. Columns (2) and (3) of Table 3 present sub-sample regressions of the impact of tax burden on under- and over-investment. The coefficient of Taxburden in column (2) is 0.025 and is significant at the 1% level, indicating that a lower tax burden can significantly improve under-investment within a firm, which has positive implications for further promoting R&D and innovation. The regression results for the over-investment sample in column (3) show that the coefficient of Taxburden is 0.102 and significant at the 1% level, indicating that when the tax burden is reduced, the over-investment phenomenon is also significantly suppressed. This may be due to the fact that a reduced tax burden increases the amount of cash retained by the firm, which can be used to motivate managers in addition to R&D and innovation, thereby alleviating agency problems and reducing over-investment. Together, these results suggest that a reduction in the tax burden is effective in improving the efficiency of corporate investment, and that this positive effect is reflected in both the alleviation of under-investment and the suppression of over-investment.

4.2. Robustness Test

To ensure the reliability of the study findings, the following robustness tests are conducted in this paper.

4.2.1. Replacing the Measurement of Investment Efficiency

To mitigate the impact of variable measurement bias on the study findings, this paper replaces the explanatory variable measure to re-run the regression of Model (2). Firstly, the model residuals are recalculated by replacing TobinQ in the Richardson model with the growth rate of operating income (Growth), referring to the methods of Pan Yue et al. (2020) and Liu Huilong et al. (2014) [24]; secondly, the investment efficiency is measured again, drawing on Biddle et al. (2009) and Chen et al. (2011) [25].

Among them, The Biddle model uses a one-year lagged growth rate of operating income (Growth) to regress the actual scale of investment (Invest); the Chen model adds a dummy variable NEG and a cross product of NEG and growth rate of operating income (Growth) to the Biddle model. NEG is defined as follows: when the growth rate of business revenue is less than 0, NEG takes 1; otherwise, it takes 0. The absolute value of the residuals of the model is taken to indicate the level of inefficient investment by regressing by year and by industry.

Table 4. Regression Results of Replacement Investment Efficiency Measures.

	Richardson Model Based on Growth Rate of Operating Income	Biddle Model	Chen Model
VAR	(1)	(2)	(3)
	Inveffi	Inveffi	Inveffi
Taxburden	0.062 ***	0.047 ***	0.048 ***
	(3.182)	(3.368)	(3.462)
Controls	Yes	Yes	Yes
Annual Effect	Yes	Yes	Yes
Industry Effect	Yes	Yes	Yes
Constant	−0.605 ***	−0.093	−0.099 *
	(−6.781)	(−1.616)	(−1.717)
Obs	5964	5934	5934
R-squared	0.171	0.074	0.077

Note: ***, * denote estimated coefficients significant at the 1% and 10% levels, and values in brackets are t-statistics or F-statistics for the corresponding coefficients.

shows the results of the regression of Model (2) after replacing the measures of the explanatory variables. It can be seen that the coefficient of tax burden (Taxburden) is significantly positive at the 1% level for the Richardson model, based on the growth rate of operating income and the investment efficiency calculated with the Biddle and Chen models [26], indicating that tax burden reduction can effectively alleviate the inefficient investment of enterprises, i.e., improve the investment efficiency, and the conclusion of this paper is still robust.

4.2.2. Endogeneity Test

As potential omitted variables and measurement errors may lead to endogeneity problems, thus affecting the reliability of the findings, the instrumental variable method is an effective means to address the endogeneity problem. In this paper, we refer to the methods of Li et al. (2018) and Liu Gae-hao et al. (2020) and use the mean value of tax burden of enterprises in the same industry in the same city as an instrumental variable (IV) for tax burden. In terms of exogeneity, the average tax burden of same-city same-industry enterprises is mainly influenced by macro-level factors, such as, the tax system, industry characteristics and the intensity of tax administration in the location, but not by individual enterprise characteristics at the micro-level, thus satisfying the exogeneity condition of the instrumental variable; in terms of correlation, the individual tax burden of enterprises is related to the average tax burden of the city and industry they belong to, thus satisfying the correlation condition.

The results of the two-stage regression of the instrumental variables are shown in

Table 5. Endogeneity Test: Instrumental Variable Approach.

VAR	(1)	(2)	(3)
	First Stage	Second Stage
	Taxburden	Inveffi	Inveffi
IV	1.022 ***
	(114.477)
Taxburden		0.043 ***	0.043 ***
		(3.490)	(3.490)
Controls	Yes	Yes	Yes
Annual Effect	Yes	Yes	Yes
Industry Effect	Yes	Yes	Yes
Constant	0.022 ***	0.098 ***	0.098 ***
	(2.726)	(8.526)	(8.526)
Obs	6042	6042	6042
R-squared	0.837	0.153	0.153
Weak identification test	13,105

Note: *** denote estimated coefficients significant at the 1% level, and values in brackets are t-statistics or F-statistics for the corresponding coefficients.

, where column (1) shows the results of the first stage regression, with the coefficient of the instrumental variable (IV) being 1.022 and significant at the 1% level, indicating that the instrumental variable is highly correlated with the tax burden (Taxburden). The bottom of the table reports the Cragg–Donald Wald test F-value for the weak instrumental variable test, which is 13,105, greater than the Stock–Yogo threshold [27], thus rejecting the original hypothesis of a weak instrumental variable and further confirming the validity of the instrumental variable. Column (2) is the result of the second stage regression; the coefficient of tax burden is 0.043 and significant at 1%, indicating that a lower tax burden can significantly curb the level of inefficient investment in listed companies, i.e., improve investment efficiency. Column (3) shows the results of the parsimonious regression. The regression results in Table 5 show that the conclusions of this paper remain robust after accounting for omitted variables and measurement error issues.

5. Mechanism and Heterogeneity Analysis

5.1. Mechanism Analysis

5.1.1. Degree of Tax Avoidance

Broadly speaking, tax avoidance refers to any behavior that reduces a company’s explicit tax obligations. Tax avoidance by companies achieves the objective of retaining cash flows and after-tax income in the company [28,29,30,31]. Thus, when faced with excessive tax burdens, companies tend to choose to alleviate financial pressure through tax avoidance practices. However, at the same time, tax avoidance is often hidden and complex, which can increase the financial complexity and create risks for companies. On the one hand, tax avoidance activities are monitored by the tax authorities. When tax avoidance is discovered by the tax authorities, there is a risk of high penalties and damage to the company’s reputation. On the other hand, companies often face serious problems of low transparency, if they are unable to clarify the financial complications resulting from tax avoidance to various investors and analysts [32]. The concealment and complexity of tax avoidance practices can exacerbate internal and external information asymmetries in companies [33], which can lead to agency problems and damage shareholder value. Studies show that tax avoidance is positively related to share price collapse [34], agency costs and debt costs [26,35], while it is positively related to firm investment efficiency and operating performance [36] and firm value are negatively correlated.

Information asymmetry and agency problems are important determinants of investment efficiency. Adverse selection and moral hazards caused by information asymmetry will cause inefficient allocation of corporate capital [37,38]. When there is serious information asymmetry within the company, managers with information advantages tend to adopt opportunistic behavior to maximize their own interests, which may result in managers’ investment decisions not being the optimal investment level and low investment efficiency. Agency problems can also lead to inefficient corporate investment [39]. This is mainly because the interests of managers and shareholders are inconsistent when the two powers are separated, so managers tend to choose projects that can bring private benefits but have negative NPV, which results in over-investment. Managers may also reject projects with high risk but positive NPV in order to improve their irreplaceability in the company and avoid damage to reputation if the projects fail, resulting in under-investment.

Based on the above analysis, the company’s behavior of increasing cash flow by reducing operating costs through tax avoidance will aggravate the information asymmetry and agency problems within the company, which will result in inefficient investment [40]. Based on the background of the tax reduction and fee reduction policy, the internal free cash flow of the company increases correspondingly when the tax burden of the company decreases, and the motivation of tax avoidance of the company weakens. Compared with tax avoidance with complicated operations and high risks, companies are more inclined to reduce their tax burdens and tax costs by virtue of the burden-reducing effect brought by policy advantages. Therefore, when companies can mitigate tax avoidance by reducing the burden of policy advantages, information asymmetry and agency problems are suppressed to a certain extent, corporate information transparency is improved, and investment efficiency is enhanced.

To provide a mechanistic test of the extent of tax avoidance, this paper uses two methods to measure the extent of tax avoidance (TA), namely accounting-tax differences and their variants. First, drawing on Desai and Dharmapala (2006), Chan et al. (2016) and Wei et al. (2020) [30,41], the accounting-tax difference (BTD) is used to measure the extent of corporate tax avoidance. The reason for choosing accounting-tax differences as a measure is that Chan et al. (2010) found that accounting-tax differences of Chinese listed companies have a significant positive correlation with tax audit adjustments (TAAs), issued by the tax authorities. Therefore, it is generally applicable and reasonable to use accounting-tax differences as an important indicator for tax authorities to determine whether tax avoidance exists in listed companies. The specific calculation method is as follows: BTD = (pre-tax accounting profit − taxable income)/total assets at the end of the period, where taxable income = (income tax expense − deferred income tax expense)/nominal income tax rate, and the larger the BTD, the higher the likelihood that the firm is using BTD for tax avoidance behavior. Secondly, the degree of corporate tax avoidance measured by the difference between accounting book and actual tax liabilities (DDBTD), calculated using the fixed-effects residual method, is also widely used [42], which is shown in Model (3), where TACC is the total accrued profit, which is numerically equal to (net profit − net cash flows from operating activities)/total assets, and DDBTD is equal to the sum of ${\mu }_i\mathrm{and}{\epsilon }_{i,t}$ in model (3).

$$BT{D}_{i,t}={\alpha }_{1}TAC{C}_{i,t-1}+{\mu }_i+{\epsilon }_{i,t}$$

(3)

In order to test the mechanism of tax avoidance in tax reduction to improve investment efficiency, Models (4) and (5) are developed in this paper. In particular, Model (4) is used to test whether a lower tax burden can significantly inhibit corporate tax avoidance, i.e., the first step of the mechanism test. When the coefficient ${\beta }_1$ of Taxburden in Model (4) is significantly positive, it indicates that tax avoidance is reduced as the tax burden faced by the company decreases, and tax avoidance is mitigated. Adding both the tax burden (Taxburden) and the extent of tax avoidance (TA) to the regression model, i.e., Model (5), is the second step in the mechanism test. In the Model (5), the coefficient ${\beta }_1$ is for tax burden (Taxburden), and ${\beta }_2$ is for the degree of tax avoidance (BTD). Combined with Model (4), when both ${\beta }_1$ and ${\beta }_2$ are significantly positive, it indicates that the degree of tax avoidance plays a mechanism role in tax burden affecting investment efficiency, i.e., a lower tax burden can improve investment efficiency by reducing the degree of tax avoidance of firms.

$$T{A}_{\mathrm{i},\mathrm{t}}={\beta }_0+{\beta }_{1}Taxburde{n}_{i,t}+{\beta }_{2}Control{s}_{i,t}+{\displaystyle \sum }Year+{\displaystyle \sum }Industry+{\epsilon }_{i,t}$$

(4)

$${\mathrm{Inveffi}}_{i,t}={\beta }_0+{\beta }_{1}Taxburde{n}_{i,t}+{\beta }_{2}T{A}_{i,t}+{\beta }_{3}Control{s}_{i,t}+{\displaystyle \sum }Year+{\displaystyle \sum }Industry+{\epsilon }_{i,t}$$

(5)

Table 6. A Mechanistic Test of the Extent of Tax Avoidance.

VAR	(1)	(2)	(3)	(4)
	BTD	Inveffi	DDBTD	Inveffi
Taxburden	0.099 ***	0.044 **	0.055 ***	0.045 **
	(3.933)	(2.352)	(3.418)	(2.394)
BTD		0.060 ***
		(3.005)
DDBTD				0.095 ***
				(2.861)
Constant	−0.437 ***	−0.575 ***	−0.314 ***	−0.571 ***
	(−5.481)	(−6.765)	(−5.531)	(−6.720)
Controls	Yes	Yes	Yes	Yes
Annual Effect	Yes	Yes	Yes	Yes
Industry Effect	Yes	Yes	Yes	Yes
Obs	6042	6042	6042	6042
R-squared	0.283	0.165	0.670	0.165

Note: ***, ** denote estimated coefficients significant at the 1% and 5% levels, and values in brackets are t-statistics or F-statistics for the corresponding coefficients.

presents the regression results for the mechanism test on the extent of tax avoidance. Of these, columns (1) and (2) show the regression results using the BTD measure of the extent of tax avoidance. The coefficient of Taxburden in column (1) is 0.112 and is significant at the 1% level, indicating that when the tax burden decreases, there is a corresponding decrease in corporate tax avoidance. This may be due to the fact that a lower tax burden increases a firm’s internal cash flow and, in order to avoid the risk of tax avoidance being detected, it will spontaneously reduce the behavior of keeping more cash in the firm through tax avoidance. The coefficients of tax burden and tax avoidance (BTD) in column (2) are also positive and significant at the 1% level, indicating that the mechanism of tax avoidance is verified, i.e., the reduction of tax burden alleviates information asymmetry and agency problems to a certain extent by discouraging corporate tax avoidance, and improves corporate information transparency, thus promoting investment efficiency. Columns (3) and (4) show the regression results of the degree of tax avoidance measured using DDBTD. Among them, the signs of the coefficients of tax burden (Taxburden) and degree of tax avoidance (BTD) are in line with expectations and significant, indicating that the mechanism test for the degree of tax avoidance is somewhat robust.

5.1.2. Financing Constraints

According to the above research results, we can find that reducing the tax burden of enterprises is helpful to improve the investment efficiency of enterprises. Faulkender and Petersen (2012) also show that financially constrained firms increase their investments after they experience tax cuts due to the American Jobs Creation Act (AJCA) [43]. These increased investments are probably positive NPV projects and, thus, the findings of Faulkender and Petersen (2012) are consistent with the idea that tax reduction can reduce under-investment of positive NPV projects. However, it is still necessary for us to further analyze through what channels the reduction of tax revenue affects the investment efficiency of enterprises. In a recent study, Dou, Johnson and Wu (2022) show that financially constrained firms compete less aggressively in the product market and become less distressed after they experience tax cuts due to the American Jobs Creation Act (AJCA) [44]. Their findings imply that additional tax burden can lead to the amplification of financial constraint through the mechanism of product market competition, rendering firms making less efficient investment decisions. The product market competition channel helps magnify the impact of the changes in tax burden. This amplification effect can be essential, because it can offset the opposite effects coming from the free cash flow channel (e.g., Jensen and Meckling, 1976). In addition, the reduction of corporate tax burden and the increase of free cash flow effectively alleviate the internal financing constraints of enterprises, and “de-leverage” and “cost reduction” of enterprises [19,45]. It [20,46] has a significant and positive effect on the company’s long-term competitiveness. This will send good information to the outside world. Based on the signaling theory, the company’s future growth prospects can improve the confidence of external investors, thus, it is easier to attract and obtain external financing. Generally speaking, enterprises can improve their own profit level and corporate liquidity by easing financing constraints [15]. According to the existing research, the improvement of corporate profit level and liquidity can stimulate R&D and innovation [17], therefore, it can alleviate the situation of insufficient investment of enterprises and improve the investment efficiency of enterprises. Finally, the relaxation of financial constraints can help firms maintain their human capital, especially key talents, such as innovators and managers who are specialized in making efficient investment decisions (e.g., Grieser and Liu, 2019 [47]; Dou, Ji, Reibstein, and Wu, 2021) [48], which can help improve firms’ investment efficiency in general.

To sum up, the reduction of corporate tax burden helps to alleviate the financing constraints faced by enterprises, thus achieving the purpose of improving the investment efficiency of enterprises.

In this paper, the SA index is chosen for the following reasons: Firstly, the variables used to calculate the SA index are relatively exogenous and more objective; secondly, the SA index is relatively robust, and its measure of the degree of financing constraints is consistent with the results of the WW index and cash-flow sensitivity. The specific construction formula is: SA = −0.737 × Size + 0.043 × Size² − 0.04 × Age, where Size is the natural logarithm of the firm’s total assets, and Age is the firm’s year of establishment. The SA index calculated in this paper is positive, and the larger the value, the lower the degree of financing constraints faced by the firm. The SA index is used to replace the tax avoidance (TA) in Models (4) and (5), respectively, to test the mechanism effect of financing constraints. Meanwhile, to alleviate the endogeneity problem of the variables, this paper does not use the natural logarithm of total assets (Size) as a control variable, but uses the natural logarithm of total market capitalization (Logvalue) to measure firm size added to the regression model [49], and the regression results are shown in

Table 7. The Mechanism Test of Financing Constraints.

VAR	(1)	(2)
	SA	Inveffi
Taxburden	−0.620 ***	0.042 ***
	(−4.502)	(3.661)
SA		−0.003 **
		(−2.403)
Constant	−26.507 ***	0.004
	(−145.742)	(0.103)
Controls	Yes	Yes
Annual Effect	Yes	Yes
Industry Effect	Yes	Yes
Obs	5507	5507
R-squared	0.932	0.158

Note: ***, ** denote estimated coefficients significant at the 1% and 5% levels, and values in brackets are t-statistics or F-statistics for the corresponding coefficients.

.

The coefficient of Taxburden in column (1) of Table 7 is −0.620 and is significant at the level of 1%, indicating that when the tax burden of an enterprise is reduced, the degree of financing constraint of the enterprise is reduced, i.e., tax reduction increases endogenous financing by keeping more cash in enterprises, while attracting exogenous financing by sending positive signals to the outside through improving liquidity position, increasing profit level, etc. To a certain extent, the purpose of the tax cut is to alleviate the financing constraint of the enterprises by means of improving the liquidity position, increasing the profit level, etc. The results are shown in column (2) of Table 7, where both the tax burden (Taxburden) and the degree of financing constraint (SA) are added to the model. The coefficient of Taxburden in column (2) is significantly positive, while the coefficient of SA is significantly negative, indicating that tax cuts can improve investment efficiency by easing corporate financing, and the mechanism effect of financing constraints is verified.

6. Research Conclusions

This paper takes the main board listed companies in Shanghai and Shenzhen A-shares from 2013–2019 as a sample, based on the current policy background of tax reduction and fee reduction in China, to explore the impact of tax burden reduction on corporate investment efficiency and the mediating effect of the degree of tax avoidance, and further analyzes the different effects of the nature of property rights, the level of corporate governance and the heterogeneity of marketization on investment efficiency. The findings show that: (1) The reduction of corporate tax burden significantly improves investment efficiency under the favorable impetus of tax reduction policies; the above basic findings still hold after robustness tests such as replacement of variable measures and instrumental variables approach; (2) The reduction in tax burden has led to more cash staying in the enterprise, thus reducing the operating costs of enterprises through riskier tax avoidance practices, and the reduction in tax avoidance practices can effectively improve the transparency of information within the enterprise and alleviate the agency problem, thus improving investment efficiency; the reduction in tax burden has also effectively improved the phenomenon of inefficient investment by alleviating the financing constraints of enterprises; and (3) Compared with non-SOEs, SOEs have more policy preferences, obvious resource advantages and less financing constraints, so this effect of tax reduction to improve corporate investment efficiency is more obvious in non-SOEs, while information asymmetry and agency problems are more obvious in the sample with lower corporate governance than those with higher corporate governance, so the reduction of tax burden and tax avoidance can alleviate the agency problem to a greater extent, thus improving investment efficiency significantly more than in the sample with higher corporate governance. Secondly, the effect of tax reduction on corporate investment efficiency is more pronounced in less market-oriented regions than in more market-oriented regions with more developed legal systems and more mature capital markets.

The findings of this paper have the following implications for the further in-depth implementation of the tax reduction policy and the positive impact of the tax reduction effect: (1) Following the economic policy trend to for a company to achieve its own development is an inevitable step in the current downward phase of China’s economy. The findings of this paper further confirm the positive role of the tax reduction policy in China’s current development stage, and provide a better understanding of the positive effect of tax reduction and fee reduction empirical evidence for a better understanding of the positive effects of tax cuts and fee reductions, and provides a reference for companies to take advantage of the tax cut policy to expand their investments and promote their long-term healthy development; (2) The tax reduction policy can effectively increase social demand and promote economic recovery and development. As well, as private enterprises are a strong contributor to the promotion of China’s economic development, achieving tax cuts for private enterprises is a powerful means of accelerating the promotion of economic growth. Therefore, for non-state-owned enterprises facing strong financing constraints, tax reduction policies can be relatively favorable, for example, reducing the tax rate of the industry and granting phased tax concessions, so as to reduce the tax burden of non-state-owned enterprises, improve their “sense of access” to tax reduction policies, and give non-state-owned enterprises an impetus for development.