Government Intervention and Corporate Investment Efficiency: Evidence from China

Apart from agency conflicts and information asymmetry between managers and shareholders, potential investors or creditors, which prevent companies from making optimal investment decisions, government intervention is another form of friction, especially common in the economic transitional settings, distorts corporate investment behavior and leads to investment inefficiency. Chinese investment system reform in 2004 aims to restrict government intervention on corporate investment and causes an exogenous shock to firm’s investment environment. In the quasi-natural experiment, difference-in-differences analysis shows that investment efficiency promotes after the investment system reform and the result is robust to an alternative model specification and placebo test. Further analysis shows that the improvement of investment efficiency concentrates among non-SOEs. The findings indicate that the investment system reform in China has alleviated the government intervention in corporate investment and improved the firm’s investment efficiency as well.


Introduction
Market frictions such as agency conflicts and information asymmetry have long recognized as the suspects causing firms to deviate from their optimal investment level [1].A large literature expands on these frictions, of which empirical evidence supportive is extensive [2] [3].Yet, there still exist other frictions, such as government intervention that can lead to suboptimal corporate investment [4] [5].Although government intervention in business is not unique in China, the Chinese setting is particularly interesting for the following reasons.Firstly, the Chinese government can play a vital role in the domestic business activities through its majority ownership in state-owned enterprises (SOEs).The goal of these enterprises is not only to maximize their shareholders' wealth, but contains lots of political purposes.Although private companies and foreign-invested firms (non-SOEs) are growing in number and have pulled a large part of Chinese economic growth after the Economic Reform and Opening Up in 1978 [6], SOEs still dominate the country's business activities.Furthermore, the government can also maintain its control on both SOEs and non-SOEs through appointing political connected executives.Firms with political connections usually have to adjust their business decisions to cater to the government's goals.Chen et al. ( 2011) [4] used government ownership and political background of the firm's top executives (political connections) as the proxies of government intervention and found that SOEs, especially those with political connected executives, have lower investment efficiency.While Deng et al. ( 2017) [5] focused on the economic stimulus package (ESP) led by China's government during the 2008 global financial crisis period, which provides positive shock to the supply of external finance, together with the government intervention, leading to government-intervened firms overinvested and their investment efficiency declined after the ESP.Both the empirical findings suggest that government intervention plays a negative role in the corporate investment efficiency.This paper aims to extend the empirical research on government intervention in corporate investment from the opposite site, basing on an event that is likely to have alleviated the impact of government intervention in corporate investment, namely "the Decision of the State Council on Investment System Reform" ("Decision") in China in 2004.Before the Economic Reform and Opening Up in 1978, when Chinese economy was primarily centrally planned, the government largely controlled the country's economic output by allocating resources and determining market prices.SOEs and quasi-SOEs (collective enterprises) occupied in almost every sector, while private enterprises and foreign-invested firms were nearly nonexistent [7].Simply put, the enterprises in China before 1978 were merely the appendage to the Chinese government, and the investment policies they made just reflected the will of the government.
From 1978 to 2004, with the continuous deepening of the economic reform, non-SOEs had become involved in sectors that had previously been monopolized by SOEs, and the government had been trying to implement a series of market-oriented investment reform policies.Unfortunately, these policies had little effect on the restriction for the government intervention in corporate investment decision.
The investment system reform or the "Decision" announced in 2004 clearly pointed out for the first time that companies should be the primary investment-decision-makers, which meant that the government should return the in- First, after the investment system reform, the corporate investment efficiency promotes.The reform can remedy the distortion of investment behavior regardless of overinvestment or underinvestment, through restricting the "grabbing hand" of the government [9].Moreover, the effect of the reform on the firm's investment efficiency varies across SOEs and non-SOEs.I find that the promotion of the corporate investment efficiency concentrates among non-SOEs, and I postulate that even if returning the investment-decision rights to the market, SOEs still have burden of catering to the political purposes.Thus, the investment system reform has little impact on SOEs.
This paper makes two contributions to the literature.First, this paper provides empirical evidence to the research on corporate investment efficiency.Previous studies in this area are mainly focused on the agency conflicts and information asymmetry between managers and shareholders or potential investors in the mature markets.However, in a transitional economy, the government intervention is another vital friction that leads to firm's suboptimal investment policy.Reform and Opening Up.The investment system reform in 2004 is a part of Economic Reform, actually.This paper identifies government intervention as another friction to explain the distortion of the corporate investment behavior and how the investment system reform can reverse it, which in fact illustrates the effect of Chinese Economic Reform as well.
The remainder of this paper proceeds as follows.Section 2 provides theoretical framework and hypothesis development.Section 3 describes the research design.Section 4 reports the empirical results, robustness tests and heterogeneity test.Section 5 concludes this paper.

Theoretical Framework and Hypothesis Development
When  [4] are not the first one who bring forward the impact of government intervention in corporate investment.We can trace back to the "grabbing hand" theory, established in a famous book, "The grabbing hand: government pathologies and their cures", wrote by Shleifer and Vishny (1998) [9].They used the "grabbing hand" to describe the government intervention in the firm's business activities and their model established in chapter 9 shows that the politicians tend to convince of the managers to overinvest because they can obtain political benefits from the overinvestment, while the managers dislike overinvestment for it hurts the benefits of the shareholders.
Therefore, the equilibrium of the firm's investment level depends on the bargain ability between the politicians and managers.In this sense, the "grabbing hand" of the government, which interfere in the business activities, usually distorts the corporate behavior and leads to suboptimal decision.
If there is a policy that is designed to restrict the "grabbing hand" on the business activities, the distortion of the corporate investment can be mitigated.Meanwhile, SOEs in China usually have the mission to help the government to accomplish social and political goals such as employment, regional development, social stability, etc., which distorts corporate investment behavior and leads to investment inefficiency.Even after the investment system reform, SOEs still have to cater to government's goals.Thus, I propose another hypothesis: H2.The promotion of investment efficiency mainly concentrates among non-SOEs after the investment system reform.

Methodology and Model
Unlike the existent empirical research on the impacts of government intervention in corporate investment efficiency, which examine the relationship directly, this paper stands at the opposite site to find out whether the corporate investment efficiency has improved when restricting or removing the government intervention.If the restriction or the removal can help to promote the corporate investment efficiency, it means the government intervention does play a negative role in the firm's investment efficiency.This paper uses the reform of investment system in China in 2004 as a quasi-natural experiment setting to examine the relationship between government intervention and corporate investment efficiency.The difference-in-differences (DiD) regression model is illustrated as: t Time is a dummy variable, which equals to1 after 2004q3 (when the investment system reform was announced) and equals to 0 before 2004q3.

Reform
is also a dummy variable, which equals to 1 when affected by the investment system reform, and otherwise, equals to 0. 3 α is the DiD coefficient which is my primary interest and measures the effect of the reform.The vector , , 1 i j t X − includes control variables.I use firm size (Size), le- verage (Leverage) and cash equivalents (Cash) to control the financial characteristics of the firms.All the controls are a quarter prior to the investment starting date and cluster the error term , , i j t µ at the two-digit industry level following the China Securities Regulatory Commission industry codes 1 .Table 1 reports the variable definitions.

Sample and Data
The investment system reform was proposed on July 16 th , 2004, thus the sample period in this paper starts from 2002q1 to 2007q3, ten quarters before the reform (2002q1~2004q2) and thirteen quarters after the reform (2004q3-2007q3).The

Main Results
I study the government intervention in corporate investment efficiency by estimating Equation (1) and Table 2 presents the coefficient estimates.Column (1) and ( 2) have tested how the reform affected firm's overinvestment while column (3) and (4) have tested the impact of the reform on firm's underinvestment.
In overinvestment situation, if the DiD coefficient is positive, it means the reform cannot alleviate the investment distortion and the corporate investment efficiency even lowers after the reform; if the DiD coefficient is negative, the reform does restrict the government intervention in corporate investment and the investment efficiency promotes.In underinvestment situation, the positive Y. Huang  1) after the PSM.The dependent variable is the investment residual calculated from Richardson (2006), which is the proxy of the firm's investment efficiency.The key explanatory that examine the impact of the reform on the firm's investment efficiency is Reform*Time.Reform*Time is a dummy equal to 1 when firm affected by the reform.Column

Robustness Test
Figure 1.The trend of the average investment efficiency.
2) I include firm fixed effects ( i γ ) and time fixed effects ( t δ ) instead of the dummy variables , , i j t Reform and t Time .Moreover, I also conduct a placebo test to ensure that the result is driven by the investment system reform rather than unobservable characteristics.As an illustration, I use the quarter from 2002q1 to 2004 q2 as the sample in the placebo test and redo the analysis, for which the Equation ( 1) is done five quarters before the actual reform (2003q2).
The results are reported in Table 3.The sign of DiD coefficients calculated from panel model are similar to the standard DiD model, while this pattern is not significant in the placebot test.

Heterogeneity Effect of the Reform between SOEs and Non-SOEs
Even if affected by the reform, SOEs still have burden of catering to the political purposes, which may lead to insignificant result among SOEs.I separated the treated firm based on the nature of firm ownership.A firm is classified as an SOE if it is controlled by the government or government institutions, while A firm is classified as a non-SOE if it is controlled by the shareholders that are unrelated to the government or government institutions.The data of ownership structure comes from CSMAR.The indicator variable D SOE equals to 1 when the firm is SOE, and D Non-SOE equals to 1 when the firm is non-SOE.The results of heterogeneity effect test are reported in Table 4.

Conclusion
This paper investigates the relationship between government intervention and corporate investment efficiency.In the quasi-natural experiment settings that provide an exogenous shock to firm's investment environment and restrict government intervention, the difference-in-differences analysis shows that the investment efficiency promotes after the investment system reform.The result is   [8].The key explanatory that examine the impact of the reform on the firm's investment efficiency is also Reform*Time.Reform*Time is a dummy equal to 1 when firm affected by the reform.Column Chen et al. (2011) [4] andDeng et al. (2017) [5] have provided evidence to this strand of literature.This paper extends the strand of literature by standing on the opposite site: if the removal of the government intervention in corporate investment can promotes the firm's investment efficiency, it means the government intervention has negative impact on the firm's investment efficiency.This paper finds the similar evidence to Chen et al. (2011)[4] or Deng et al. (2017)[5].Second, this paper enriches the extant literature on the Chinese Economic Journal of Service Science and Management Investment system reform announced in 2004 in China, clearly points out for the first time that companies should be the primary investment-decision-makers and proclaimed that the government should not interfere in the corporate investment process.Thus, after the announcement of the reform, we should observe promotion of corporate investment efficiency.So I conduct an empirical test of the following hypothesis: H1.Corporate investment efficiency promotes after the investment system Y.Huang DOI: 10.4236/jssm.2019.123018271 Journal of Service Science and Management reform.

( 1 )
and (2) reports the results in overinvestment situation while column (3) and (4) reports the results in underinvestment situation.I report p-value in parentheses.The standard errors are clustered at two-digit industry level.*, ** and *** denote statistical significance at 10%, 5% and 1% levels, respectively.DiD coefficient means the reform does work while the negative DiD coefficient means the reform fails.The main result shows that when restricting or removing the government intervention, the corporate investment efficiency promote regardless of overinvestment or underinvestment, which illustrates that the government intervention does play a negative role in the firm's investment efficiency.Moreover, I present the trend of average investment efficiency from 2003q1 to 2007q3 in Figure 1.It is clearly demonstrated in the Figure 1 that regardless of overinvestment or underinvestment, the investment residuals are becoming close to the zero after 2004q3, which means the average investment efficiency increase after the investment reform.

( 1 )
and (3) are the regression results of Equation (2), while Column (2) and (4) are the regression results of Equation (1) using the sample from 2002q1 to 2004q2.Column (1) and (3) include year and firm fixed effect, while Column (2) and (4) include dummy variables reform and time.All specification include size, leverage and cash.I report p-value in parentheses.The standard errors are clustered at two-digit industry level.*, ** and *** denote statistical significance at 10%, 5% and 1% levels, respectively.
[1]talk about what drives the firms' deviate from its optimal investment level, agency conflicts and information asymmetry are always the top two factors that come to our minds[1].Chen et al.
[4]11)[4]opened a new chapter that government intervention is also a vital friction that leads to the firm's suboptimal investment behavior.They used government ownership and political connections as the proxies of government intervention to figure out that this kind of friction play a negative role in the corporate investment efficiency.Deng et al. (2017) [5] used economic stimulus package (ESP) led by China's government during the 2008 global financial crisis period as the quasi-natural experiment and found out that government-intervened firm overinvested after the ESP.Nevertheless, Chen et al. (2011)
Because all the firms in China would be affected by the reform, I use the manufacture firms listed in China as the treated firms.As for the matched firms, I use the manufacture firms listed in America, which was not affected by the Chinese investment system reform.The financial data of Chinese firms comes from the China Stock Market and Accounting Research database (CSMAR), while the financial data of American firms comes from the COMPUSTAT database.I exclude firm-quarter observations for which information is not available and winsorize all the continuous variables at the 1% in each tail.Moreover, I exclude observations with negative assets, capital expenditures (which use to calculate the firm's investment efficiency) and cash, as well as observations with leverage larger than one.Before PSM, I have to unify the currency units between CHY and USD.The exchange rate data also comes from CSMAR.In addition, it is necessary to match the four-digit SIC industry code with the two-digit China Securities Regulatory Commission industry codes before merging the listed manufacturing firm's data from CSMAR with listed manufacturing firm's data from COMPUSTAT.
This table presents the DiD regression results of Equation (

1 )
In the robustness test, I use a panel model instead of the standard DiD model to estimate the DiD coefficient and the model is presented as:

Table 3 .
Richardson (2006)his table presents the robustness tests.The dependent variable is also the investment residual calculated fromRichardson (2006)

Table 4 .
[8]hardson (2006)ect test.This table presents the heterogeneity effect results test using the standard DiD model or Equation(1).The dependent variable is investment residual calculated fromRichardson (2006)[8].The key explanatory that examine the impact of the reform on the firm's investment efficiency is Reform*Time.Reform*Time is a dummy equal to 1 when firm affected by the reform.In Column (1), I separate treated firm based on their ownership structure in overinvestment situation, while column (2) reports the results in underinvestment situation.All specification includes reform, time, size, leverage and cash.I report p-value in parentheses.The standard errors are clustered at two-digit industry level.*, ** and *** denote statistical significance at 10%, 5% and 1% levels, respectively.