Sage Journals: Discover world-class research

Abstract

In the process of issuing and using local government debt, local officials may be corrupted by self-interest into inappropriately increasing the bond supply, leading to excessive expansion of local government debt and an increase in debt risk. Adopting the perspective of government investment efficiency, this paper employs a piecemeal causal steps approach to analyze the influence of local official corruption on local government debt. The results from panel data on 30 provinces in China show that the local official corruption triggers the expansion of local government debt. A mediation analysis reveals that the effect operates by impairing government investment efficiency. An analysis of regional heterogeneity further shows that the impact is weaker or even disappears in developed in comparison to underdeveloped regions, which further indicates that the anti-corruption campaign supports local government debt management, especially in developed regions. This paper not only explains the relationship between local official corruption and local government debt, but also provides new ideas for local government debt management.

Keywords

local official corruption local government debt government investment efficiency mediating effect regional heterogeneity

Introduction

Public debt has always played an important role in economic development by providing financial support. However, since 2008, public debt has risen sharply globally, and some countries in Europe have even seen debt crises, which have seriously harmed growth (Sarangi & El-Ahmadieh, 2017). In 2020, countries issued large amounts of public debt in response to the COVID-19 pandemic, which further inflated their debt (Institute of International Finance, 2021). According to the publications of the Institute of International Finance, the COVID-19 pandemic added $24 trillion to global debt in 2020, bringing it to a record $281 trillion (355% of global GDP), an increase of 35 percentage points over 2019. This debt growth rate even exceeds that of 2008, the year of the subprime crisis. A $12 trillion increase in public debt accounts for half of this increase in global debt (Institute of International Finance, 2021). The scale of public debt has reached a record high, and the consequent risks, such as reductions in public expenditure efficiency (Teles & Mussolini, 2014) and in fiscal sustainability (Mahmah & Kandil, 2019), have also started to emerge at the country level. These risks jeopardize the long-term development of national economies (Chudik et al., 2018). As the largest socialist country among emerging economies, China’s local government debt balance was approximately 30.47 trillion yuan by 2021 (Ministry of Finance of the People’s Republic of China, 2022), with its debt issuance increasing by 63% year-on-year. At the end of 2020, the debt ratio of local governments in China reached 93.6% (Ministry of Finance of the People’s Republic of China 2021), close to the 100% debt risk warning line (National People’s Congress of the People’s Republic of China, 2015). This expansion of local government debt has also slowed China’s economic development (Cang & Matt, 2017). Since 2014, Chinese government authorities have launched a variety of policies and regulations to prevent and defuse financial risks. However, due to the imbalance between local government revenue and expenditures and the imperfect regulation of local government debt, the health of public finance stability has suffered greatly (Lam & Wang, 2018). The risks brought about by the increasing scale of local government debt still plague the Chinese authorities, and the problem of debt scale management cannot be ignored.

Moreover, in the process of issuing and using local government debt, local government officials may engage in corruption for their own benefit and inappropriately increase borrowing, resulting in excessive expansion of government debt scale. This has threatened both the economic and social stability of China (Pei, 2007). China’s political performance appraisal system characterized by “GDP worship” (F. Su et al., 2012) and the promotion tournament model (Li & Zhou, 2005) provide a “hotbed” for corruption among local government officials. According to rent-seeking theory and agency theory, corrupt officials engage in rent-seeking and rent creation in the process of issuing and using debt to advance their personal interests and political performance (by fostering rapid short-term GDP growth). This can lead to moral hazard and inappropriate debt issuance and use. Such practices not only directly expand the scale of debt but also reduce the investment efficiency of debt funding, create funding gaps, prompt corrupt officials to borrow for rapid access to funds, and indirectly lead to further expansion of the debt scale. This increases the risk associated with debt and endangers the stability of the entire economy. Therefore, to provide new ideas for local government debt management, this paper explores, in the context of the Chinese political and economic system, how local official corruption affects the scale of local government debt.

Existing research on the influencing factors of the scale of local government debt has approached the question from fiscal (Abubakar, 2020) and taxation (Ukeme & Ifayemi, 2020) perspectives, but few studies have paid attention to the impact of government official corruption on the scale of local government debt and its associated mechanism. Second, most research on corruption and macroeconomics focuses on the relationship between corruption and economic growth (Gründler & Potrafke, 2019), investment (Brada et al., 2019), and so on. Only some scholars have focused on the relationship between corruption and public debt (Benfratello et al., 2018; Liu et al., 2017). However, these studies use mostly international data to explore the relationship between corruption and public debt at the national level, ignoring the regional characteristics of local government debt dynamics under China’s economic system and not considering the country’s anticorruption campaign in 2012, which was primarily successful in exposing corrupt officials (Pei, 2018). In addition, existing research suggests that government official corruption reduces the efficiency of government investment (Haque & Kneller, 2008), leading to an increase in public spending (Swaleheen et al., 2019). And one way to meet public spending is to issue public debt (Bhandari et al., 2017). Does government investment efficiency represent a channel of the impact of corruption on public debt? Based on this question, in the context of the current surge in global debt, the continuous expansion of the local government debt scale in China, and the anti-corruption campaign, the current study focuses on 30 Chinese provinces to explore the impact of local government official corruption on local government debt scale through the mechanism of government investment efficiency. It further analyzes whether there are differences in the influence of local government official corruption on the local government debt scale in regions with different economic development levels in China.

The main contributions of this paper are as follows:

(1) In the context of the Chinese political and economic system, the local government official corruption is identified as a factor affecting the scale of local government debt. Most of the existing research on the relationship between corruption and public debt has been conducted at the national level (Benfratello et al., 2018). The few studies focused at the regional level within countries are based on the economic structure of capitalist countries such as the United States (Liu et al., 2017). This paper is based on the organizational structure of China’s socialist M-form multi-layer-multi-regional economy (Qian & Xu, 1993) and analyzes the relationship between corruption and debt at the regional level.

(2) The article explores government investment efficiency as the path by which local official corruption affects the local government debt scale. Existing studies on the relationship between corruption and public debt pay little attention to the channel linking them (Baklouti & Boujelbene, 2021; Ivanyna et al., 2018). While research suggests that corruption impacts government investment efficiency (Haque & Kneller, 2008), government investment also affects the scale of public debt (Belguith & Omrane, 2017).

(3) The article distinguishes differences in the impact of local official corruption on local government debt scale by level of economic development in China. Existing research analyses national-level differences in public debt in countries at different levels of economic development (Ibrahim, 2020). However, the possible impacts of different economic systems across countries are ignored. This study explores the effect heterogeneity by economic development level under China’s economic system, offering a meaningful supplement to existing research.

The paper is organized as follows: Section II discusses the existing research about corruption, government debt, and government investment efficiency. Section III provides a theoretical analysis of the path whereby local official corruption impacts the local government debt scale and presents the key hypothesis. Section IV describes the variables, data, and model. Section V presents the results derived from the hypothesis and a heterogeneity analysis by region. Section VI summarizes the main conclusions and highlights policy implications. We present a flowchart to show the research methodology as shown in Figure 1.

Figure 1.

Flowchart of the research methodology.

Literature Review

This paper summarizes the literature by focusing on three main aspects: corruption indicators, the relationship between corruption and government investment efficiency, and the relationship between corruption and public debt.

At present, there are three main ways to measure corruption: the International Country Risk Guide (ICRG) index, the World Governance Indicators (WGI), and Transparency International’s (TI’s) Corruption Perception Index (CPI). The CPI is the most widely used (Paldam, 2021; Potrafke, 2019). This is because the ICRG measure reflects political risk brought about by corruption rather than directly reflecting the degree of corruption (Ko & Samajdar, 2010). Some of the components of the WGI are made up of unobservable parts, and there are limitations to its applicability (Qu, 2019). The CPI, by contrast, measures general “public sector” corruption and provides a general or overall assessment of country risk, competitiveness, and governance (Ko & Samajdar, 2010). Therefore, it is suitable for wide application. However, the CPI also has some limitations: TI measures corruption at the national level, not within countries. Corruption as measured by the CPI refers to “the abuse of public office for private gain” (Transparency International, 2012), while according to studies, corruption in China includes not only traditional abuse of public power for personal gain, but also malfeasance (Wederman, 2004). As a result, the CPI is not suitable for studies on corruption across the regions within China. Some studies have proposed measuring corruption by the number of corruption cases (Glaeser & Saks, 2006) and the number of peoples involved in corruption (Kim et al., 2018).

Most studies on corruption and investment efficiency focus on the effect of corruption on market investment efficiency. Analyses at the enterprise level show that corruption will reduce the efficiency of capital investment (Brown et al., 2019). From the government perspective, Haque and Kneller (2008) directly study the relationship between corruption and the efficiency of public investment and find that in countries with a high degree of public corruption, the efficiency of public investment seems to decrease. In addition, more scholars have incorporated economic development into their research, and have found that in countries with high corruption, large-scale investment does not promote economic growth, because the efficiency of investment is damaged (Lim, 2019). In addition, many studies approaching the question from the angle of global anti-corruption suggest that anti-corruption improves the investment efficiency of enterprises (Kong et al., 2020) and greatly enhances their total factor productivity, especially for state-owned enterprises (Guo et al., 2021). This is because anti-corruption efforts have eased enterprises’ overinvestment and underinvestment problems (Imamah, 2020).

Some studies focus on the relationship between corruption and public debt, but incorporate other factors, such as the shadow economy (Cooray et al., 2017), fiscal gaps (Ivanyna et al., 2018), and so on. The results show that corruption prevents public debt from playing its appropriate role in economic growth. Furthermore, a few scholars directly study the impact of corruption on public debt and draw the same conclusion: corruption increases public debt (Benfratello et al., 2018; Liu et al., 2017). However, most of these studies have been performed at the national level.

In summary, existing studies on the relationship between corruption and public debt take a variety of perspectives, but most of them are based on the CPI at the cross-national level. Regional differences within countries are not considered because the CPI does not apply to research conducted within countries or regions. In addition, a few studies on the relationship between corruption and public debt have been carried out at the within-country regional level but ignore the influence mechanism linking corruption and local government debt, and the role of government investment efficiency as a possible impact path between the two requires further discussion. In this vein, this paper chooses the number of corruption cases per 10,000 public officials to measure corruption and takes 30 provinces in China as its sample to analyze the influence of corruption on the local government debt scale. In addition, we explore the mechanism or impact path from the perspective of government investment efficiency.

Hypothesis Development

Direct Impact of Local Official Corruption on Local Government Debt

Existing studies have shown that there are two effects of corruption: one is the “greasing-the-wheel” mechanism reflected in lubricant theory (Huntington, 2006; Leff, 1964), and the other is the “sanding-the-wheel” mechanism articulated by rent-seeking theory. Under the “greasing-the-wheel” mechanism, corruption is an insurance mechanism against potential losses caused by ineffective policies (Kong et al., 2020). The most effective companies tend to offer the most bribes to win projects, save on time costs, and achieve an efficient allocation of resources, thereby improving market efficiency and promoting economic growth (Beck & Maher, 1986; Lui, 1985). Under the “sanding-the-wheel” mechanism, corruption allocates social resources to rent-seeking activities rather than production (Shleifer & Vishny, 1993), leading to a decline in the productivity of economic agents and undermining economic development (Yamarik & Redmon, 2017).

In view of the above two competing theories, how does corruption affect local government debt in China? Does rent-seeking theory or lubricant theory better explain this relationship in China?

Studies have shown that in the early stage of China’s economic reform, the development of non-state-owned enterprises depended largely on using corruption to avoid or undermine discriminatory policies (He, 2000), so the “greasing-the-wheel” mechanism of corruption played a strong role in promoting economic development and was conducive to controlling the debt scale. However, with the development of China’s economy, public administration education and ethics education have received great attention (Hu & Zhang, 2020; Sun et al., 2021), which has promoted the development of appropriate public management. The ineffectiveness of government policies has been greatly reduced, so the “greasing-the-wheel” mechanism is less likely to be operative, while the “sanding-the-wheel” mechanism is more likely to damage the economy, in turn leading to an expansion of debt. Therefore, we argue that corruption among local officials in China inflates local government debt.

Next, we specifically discuss how corruption expands debt. China’s political performance appraisal system characterized by “GDP worship” (F. Su et al., 2012) and the promotion tournament model (Li & Zhou, 2005) provide a “hotbed” for public corruption. This system subordinates public administration to politics (Young et al., 2020), promoting corrupt behavior in the process of debt management among officials to improve their political performance. Corrupt officials, in order to achieve impressive political performance and rapid GDP growth in the short term in hopes of garnering recognition from the central government and promotion, may abuse their position and seek to boost GDP through inappropriately “lavish” public investments. They may be more inclined to obtain funds by borrowing (Blanchard & Shleifer, 2001). At the same time, corruption can worsen local governance, producing debt management loopholes and excessive debt, leading to the expansion of the local government debt scale.

Based on the promotion tournament model (Li & Zhou, 2005), China’s political performance appraisal system and the associated “GDP worship” emphasize the following aspects. First, the performance appraisal system encourages officials to focus their work on economic development during their tenure (Y. Chen et al., 2005), ignoring future developments. Second, there is an information asymmetry between the central government and local governments. As a result, senior officials are unable to accurately assess the economic performance of subordinate officials. GDP growth is often seen as a signal of good performance on the part of an official (Whiting, 2006). Third, this appraisal system promotes GDP competition among local government officials (Li & Zhou, 2005). These characteristics make officials inclined to seek regional GDP increases in the short term to secure political promotions, which may corrupt their decision-making about public affairs to advance their own interests.

Under the “GDP worship” system, to be promoted, corrupt officials may abuse their positions to invest in “vanity” projects or “lavish” public projects and stimulate short-term false prosperity in GDP, even resulting in overinvestment. While project investment plays a crucial role in driving GDP growth, this short-sighted behavior can put enormous pressure on government finances. Under China’s decentralized fiscal system, expanding government debt can provide rapid financing in the short term (Kalim, 2023) to alleviate fiscal pressures. In term, enormous fiscal pressures may lead to excessive debt expansion. Moreover, when there is high corruption among local government officials, the degree of corruption in a region deepens, decreasing governance capacity and quality (Abada & Ngwu, 2019). Furthermore, government debt management may be jeopardized, and supervision of the issuance and use of debt may be poor. There may be inappropriate issuance of bonds, inflating local government debt.

In summary, under the performance appraisal system with GDP at its core, in areas with high corruption, local officials may seek to advance their own promotion prospects and improve their political performance during their tenure through short-term, rapid GDP growth, using the convenience of their positions to overinvest and overborrow. At the same time, such an increase in corruption worsens local governance and opens loopholes in debt management, inappropriately inflating local government debt. Based on the above analysis, this paper proposes the first hypothesis:

Hypothesis 1: Local official corruption increases local government debt.

Indirect Impact of Local Official Corruption on Local Government Debt Through Government Investment Efficiency

Local official corruption not only has a negative impact on local economic development (Gründler & Potrafke, 2019) but also affects local government investment efficiency (Haque & Kneller, 2008), which includes economic investment and social investment efficiency. Economic investment concerns mainly investment in fixed assets by the government to perform its economic function. Social investment includes investment in infrastructure, education, health care, ecology, and other aspects, whereby the government fulfills its social and cultural functions. Local officials, for their own benefit, might inappropriately increase government investment to stimulate local economic development. This behavior could lead to a rise in government investment costs and distort market resource allocation, thus reducing government investment efficiency (Liu et al., 2017) and raising the level of investment needed to achieve the original expected goals, giving rise to a larger expansion of local government debt.

First, local official corruption may raise government investment costs, resulting in a reduction in government investment efficiency. According to rent-seeking theory (Krueger, 1974) and bureaucratic budget maximization theory (Mitchell & Niskanen, 1971), officials seek to maximize public budget expenditure, which provides the conditions for corruption. For example, in government investment projects, the central government allocates earmarked funds to local governments in the form of transfer payments. Under multiple allocations of project funds, officials may carry out rent-creating and rent-seeking activities, using their position to manipulate loopholes in regulations to put some of the public money into their own pockets and choosing to invest in inappropriate projects to drive the economy (Lambsdorff, 2003). Both behaviors can cause the original fund allocations to be insufficient. In such a case, it is easy to produce a poor-quality, “jerry-built” project (Nieto-Morales & Ríos, 2022), which wastes investment resources and harms investment efficiency. To compensate for such project and achieve the same goals as expected, local governments may require a transfusion of government investment in project construction, further raising government investment costs and lowering investment efficiency. One way to quickly address the problem of rising investment costs is to borrow, resulting in an expansion of government debt.

Second, local official corruption can distort the allocation of market resources, reducing government investment efficiency. Under the decentralized fiscal system in China, local governments control local public resources. Corrupt officials can abuse the information asymmetry between the public and government, through their allocation of public resources to conduct rent-seeking (S. Su & Ni, 2023). Corrupt officials who obtain gains can offer permits and licenses for enterprises and social programs (Meza & Pérez-Chiqués, 2021). Under such corrupt behavior, “bad money drives out good “ and influential enterprises may try to bribe officials to receive preferential benefits, disorting the competitive environment and market resource allocation. Government investment efficiency is thus constantly eroded by corruption, even bringing about economic losses. For the development of the economy, local governments must take responsibility for economic losses. This increases financial pressure on the government, which further expands the local government debt scale.

In summary, government investment is conducive to promoting local economic development. In recent years, the state has increased investment (Liang et al., 2017). However, in regions with high corruption, government investment efficiency is damaged due to the local official corruption. With a proliferation of “jerry-built” projects, original fund allocations may be insufficient to complete project construction, resulting in large funding gaps. Furthermore, given China’s imperfect market supervision, corrupt officials are more inclined to seek rents and accept bribes, which distorts market resource allocation, impairs investment efficiency and yields economic losses. To fill funding gaps and make up for economic losses, corrupt officials choose to inflate local government debt. Based on the above analysis, this paper proposes another set of hypotheses:

Hypothesis 2a: Corruption decreases local government investment efficiency;

Hypothesis 2b: Through the decrease the government investment efficiency, corruption expands the scale of government debt; that is, the government investment efficiency has a mediating effect on the influence of corruption on local government debt.

This paper explains the impact of local official corruption on the local government debt scale through both direct and indirect channels. The research framework is shown in Figure 2.

Figure 2.

Research framework.

Research Design

Variables

Dependent Variable

The dependent variable in this paper is the scale of local government debt (DEBT). This paper uses the balance of urban investment bonds and local government bonds to measure local government debt in China’s provinces (Feng, 2013) as these instruments are the main components of local government debt and their sum can reflect government debt situation to some extent.

Independent Variable

Local official corruption (CORR): Based on prior studies (Glaeser & Saks, 2006; Liu et al., 2017), this paper measures the degree of corruption by the number of registered cases of corruption per 10,000 public officials in each province. The number of registered cases of corruption refers to the number of acts, including corruption and bribery, conflicts of interest, and malfeasance of local public officials.¹ Because of the covertness of corruption, it is difficult to obtain relevant data before a corruption case is announced. Therefore, we use the number of corruption cases reported in the China Inspection Yearbook as a measure of corruption.

How do we interpret the degree of corruption in a region based on this measurement? Existing research suggests that the number of registered cases is more directly reflective of anti-corruption more than corruption (Xu & Yano, 2017). This is because corruption is covert (Amoah et al., 2022), and it is difficult to judge how much corruption is taking place without being discovered. Under China’s anti-corruption campaign, many corrupt practices have been exposed to the public, so the number of corruption cases reflects the achievements of China’s anticorruption efforts. The greater the number of corruption cases, the better are the results of the fight against corruption—that is, the greater is the effort to fight corruption (Zhang et al., 2019). In such an environment, the criminal cost of corruption and psychological pressure faced by corrupt local government officials increase greatly, forcing government officials to reject corruption in the region (J. Deng, 2018). Therefore, anti-corruption and corruption are two opposite concepts; that is, the number of corruption cases directly measures the degree of anti-corruption, and when it is used to measure the degree of corruption, it is a reverse indicator; that is, the greater the value is, the lower the degree of corruption.

Mediating Variable

Government investment efficiency (IE): This paper uses the Data Envelopment Analysis (DEA) model to calculate government investment efficiency (IMF, 2015). The indicators selected fall into two categories: economic and social investment, corresponding to public construction and social services investment indicators (IMF, 2015). All indicators are shown in Table 1:

Table 1.
Indicators for Calculation of Government Investment Efficiency Through the DEA Model.

Items Indicators Description Unit

Economic investment Per capita investment in fixed assets Yuan

Social investment Per capita expenditure for General public services Yuan

Per capita expenditure for education Yuan

Per capita expenditure for science and technology Yuan

Per capita expenditure for Health care Yuan

Per capita expenditure for social security and employment Yuan

Per capita expenditure for energy conservation and environment protection Yuan

Economic output^a Strength of economic development Per capita GDP Yuan

Capabilities of economic development Fiscal revenue/ GDP

Stability of economic development Reciprocal of the coefficient of variation of GDP growth rate

Rationality of economic development 1—Industrial Structure Theil Index

Outward-orientation of economic development Total imports and exports/GDP

Social output General public services output Road network density^b km/100 km²

Per capita area of paved roads m²

Number of public buses and trolley buses per 10,000 persons Unit

Length of city sewage pipes 10,000 km

Per capita electricity consumption kwh

Education output Per capita education years^c Year

Average school enrolment per 100,000 population of higher education Person

Science and technology output Number of patent applications Pieces

Full-time equivalent of R&D personnel Man-year

Health care output Number of beds in health care institutions per 1,000 persons Bed

Health technical personnel in health care institutions per 1,000 persons Person

Social security output Number of employed persons in urban units 10,000 persons

Urban-rural income gap^d

Ecological environment output Per capita area of parks and green land m²

Common industrial solid wasteIntegrate reused rate %

GDP output per sulfur dioxide emissions 10,000 yuan/kg

a
The indicators of economic output are mainly measured by referring to Mlachila et al. (2017). ^bRoad network density is formed by highway density and railway density, which is the sum of Length of Highways and Length of Railways divided by the Urban Area. ^cPer capita education years = 6 × Primary + 9 × Junior +12 × Senior +16 × College, where, Primary, Junior, Senior and College represent the proportion of the population aged 6 years and above with primary, junior, senior and college education, respectively. ^dUrban-rural income gap is the reciprocal of the ratio of the Per Capita Disposable Income of Urban Households to the Per Capita Disposable Income of Rural Households.

To eliminate differences in the selected indicator units, all the indicators are standardized, and then the nondimensionalized data of the indicators are obtained. In addition, this paper equally weights the descriptions to synthesize the corresponding indicators (IMF, 2015). As a result, 11 input indicators and 7 output indicators are obtained to calculate the government investment efficiency of each province.

Control Variables

This paper introduces the following six control variables based on previous studies: (1) The GDP growth rate (GGDP) is represented by the annual GDP growth rate of each province (C. Chen et al., 2017). (2) Government investment (INV) is measured by fixed asset investment (Chatterjee et al., 2017). (3) Fiscal expenditure (FE) is expressed by the amount of fiscal expenditure announced by local governments (Cai, 2017). (4) Fiscal decentralization (FD) is measured by the ratio of per capita local government fiscal revenue to per capita central government fiscal revenue in the same year (Liu et al., 2017). (5) The marketization level (MAR) is measured by the proportion of the tertiary industry in total GDP (Todd & Yates, 2020). (6) The financial development level (FIN) is measured by the ratio of the increase in financial sector gross product to total GDP (Huang et al., 2018).

All variables in this paper are summarized in Table 2.

Table 2.
Variable Definitions and Descriptions.

Type Variable Abbreviation Definition

Independent variable Local government debt DEBT Local government debt Outstanding

Dependent variable Local official corruption CORR Number of corruption case/number of public officials

Mediating variable Government investment efficiency IE Calculated though DEA model

Control variable GDP growth rate GGDP (GDP of this year—GDP of last year)/GDP of last year

Government investment INV Total investment in fixed asset

Fiscal expenditure FE Fiscal expenditure

Fiscal decentralization FD Per capita local government fiscal revenue/per capita central government fiscal revenue

Marketization MAR Tertiary industry in the GDP

Financial development FIN Increase in financial sector gross product/GDP

Data

This paper takes 34 provinces, municipalities and autonomous regions in China as the research objects to analyze the influence of local official corruption on local government debt. Regarding data collection, since data on corruption cases were been published only until 2017 and the starting year of the data on local government debt in the Wind database is 2010, the initial sample interval is from 2010 to 2017, and the data are at annual frequency.

The data on local government debt come from the Wind database. The corruption data are collected and sorted from the National Statistical Yearbook of China and the work statements of procuratorates of each province. The government investment efficiency index data are obtained from the website of the National Bureau of Statistics, Wind database, China Statistical Yearbook, China Statistical Yearbook on Science and Technology and China Population Census. The data used to measure fixed asset investment, fiscal expenditure, fiscal decentralization, financial development and marketization are all from the Wind database.

Note that the number of registered corruption cases announced in 2012 and 2017 is the sum of the cases for the past 5 years. Therefore, in this paper, the registered corruption cases in 2012 and 2017 are calculated as the cases from the past 5 years minus those from the previous 4 years. The number of public officials is represented by the number of employed persons in public management, social security, and social organizations (United Nations Office on Drugs and Crime, 2004), and the data are derived from the Wind database.

Due to a serious lack of data for the Tibet, Taiwan, Hong Kong, and Macau, they are excluded from the data compilation, and 30 provinces are retained in the sample. For small amounts of missing data, we adopt multiple imputation (Rubin, 1996). In addition, to make the empirical test results more robust and significant, we conduct logarithmic processing on all variables except the GDP growth rate and finally obtain 240 panel data observations of 30 provinces from 2010 to 2017.

Model

To test the direct impact of corruption on the scale of local government debt and the indirect impact via government investment efficiency, this paper uses the piecemeal causal steps approach (Baron & Kenny, 1986). The econometric model is as follows:

$\begin{matrix} DEB T_{i, t} = α_{0} + α_{1} COR R_{i, t} + α_{2} GGD P_{i, t} + α_{3} IN V_{i, t} \\ + α_{4} F E_{i, t} + α_{5} F D_{i, t} + α_{6} MA R_{i, t} + α_{7} FI N_{i, t} + ε_{i, t} \end{matrix}$ (1)

$\begin{matrix} I E_{i, t} = β_{0} + β_{1} COR R_{i, t} + β_{2} GGD P_{i, t} + β_{3} IN V_{i, t} \\ + β_{4} F E_{i, t} + β_{5} F D_{i, t} + β_{6} MA R_{i, t} + β_{7} FI N_{i, t} + ε_{i, t} \end{matrix}$ (2)

$\begin{matrix} DEB T_{i, t} = λ_{0} + λ_{1} COR R_{i, t} + λ_{2} I E_{i, t} + λ_{3} GGD P_{i, t} \\ + λ_{4} IN V_{i, t} + λ_{5} F E_{i, t} \\ + λ_{6} F D_{i, t} + λ_{7} MA R_{i, t} + λ_{9} FI N_{i, t} + ε_{i, t} \end{matrix}$ (3)

In the model, i represents each province, t represents time, and ε represents the residual.

First step, Model (1) is used to test the direct impact of local official corruption on the local government debt scale. If $α_{1}$ is significantly negative, it indicates that corruption increases local government debt and thus that H1 is verified.

Second step, Model (2) is used to test the influence of local official corruption on government investment efficiency. If $β_{1}$ is significantly positive, it indicates that corruption reduces government investment efficiency and thus that H2a is verified.

Third step, Model (3) is used to test the mediating (indirect) effect of local official corruption on the local government debt scale through government investment efficiency. If $λ_{2}$ is significant, then there is a mediating effect. If $λ_{1}$ is not significant, then there is a complete mediating effect of government investment efficiency, while if $λ_{1}$ is significant, there is a partial mediating effect, and H2b is verified (Baron & Kenny, 1986).

Empirical Analyses

Basic Statistics

The descriptive statistics of the variables are shown in Table 3, offering a preliminary understanding of the debt and corruption situation of 30 provinces in China.

Table 3.
Result of Descriptive Statistical Analysis.

Variables Units M SD Minimum Maximum

DEBT 100 million yuan 2,462.571 3,445.45 10 23,343.68

CORR 22.604 6.603 7.66 42.04

IE 0.729 0.189 0.28 1

GGDP % 0.116 0.07 −0.22 0.27

INV 100 million yuan 15,348.361 10,923.472 1,016.9 55,202.719

FE 100 million yuan 4,154.232 2,269.15 557.53 15,043.09

FD 3.549816 2.623425 1.158271 13.35832

URB 55.981 12.705 33.81 89.6

MAR 44.194 9.411 28.6 80.56

According to the descriptive statistical analysis shown in Table 2, the average local government debt in China is 2,462.571, while the minimum is 10, the maximum is 23,343.68, and the standard deviation is 3,445.45, indicating that the debt size varies greatly in different regions in China.

Similarly, based on their means, maximums, minimums, and standard deviations, it can be concluded that the local official corruption and government investment efficiency also have great differences in different regions and years, which may be caused by differences in geographical location and economic development.

Results Analysis

Based on the static panel data model, in this paper, we use the panel feasible generalized least square method (FGLS) to effectively solve the problems of heteroscedasticity and sequence dependence in processing and analyzing the data (Ogbeifun & Shobande, 2020). All the regression results in this paper are obtained with this FGLS model. Table 4 shows the empirical test results of the three steps of the model.

Table 4.
Regression Results of the Influence of Local Official Corruption on Local Government Debt.

Variable First step Second step Third step

(1) (2) (3) (4) (5)

DEBT DEBT DEBT IE DEBT

CORR −1.287* (−3.42) −0.694* (−3.83) 0.185* (3.39) −0.441* (−2.61)

IE −1.366* (−6.98)

GGDP −4.684* (−5.73) −4.816* (−6.06) 1.948* (8.15) −2.153* (−2.63)

INV 1.301* (8.07) 1.417* (8.89) −0.248* (−5.19) 1.077* (7.03)

FE 0.190 (0.87) 0.0578 (0.27) 0.493* (7.66) 0.732* (3.36)

FD −0.450* (−3.75) −0.562*** (−4.67) 0.0630* (1.74) −0.476* (−4.31)

MAR 3.120* (6.53) 2.668* (5.58) 0.402* (2.80) 3.217* (7.26)

FIN 1.392* (6.63) 1.537* (7.41) −0.323* (−5.19) 1.095* (5.49)

Cons 10.61* (9.12) −20.23* (−12.12) −16.52* (−8.74) −3.873* (−6.82) −21.81* (−11.59)

N 240 240 240 240 240

Wald 11.69* 1,237.34* 1,327.60* 181.01* 1,645.81***

Note. t Statistics in parentheses.
*
p < .1. ***p < .01.

In Table 4, the baseline results of Models (1) and (2) show a significant relationship between the local official corruption and the selected control variables and the local government debt scale.

Model (3) tests for a direct influence of local official corruption on the local government debt scale and is the first step in the mediating effect test. As seen from the regression results, CORR and DEBT have a significantly negative correlation ( $α_{1} = - 0.694$ , $p < 0.01$ ). Because CORR is a reverse indicator, the results show that the local official corruption has a significant positive effect on the local government debt scale. This supports Hypothesis 1.

Model (4) shows the influence of local official corruption on government investment efficiency. As seen from the results, the coefficient of CORR is significantly positive ( $β_{1} = 0.185$ , $p < 0.01$ ); as CORR is a reverse indicator, this indicates that local official corruption has a negative effect on government investment efficiency. That is, the higher the corruption, the lower is government investment efficiency. This provides evidence consistent with Hypothesis 2a.

This outcome may be because corrupt officials, to shore up their personal promotion prospects and interests, inappropriately increase government investment to improve their economic performance, thereby decreasing government investment efficiency.

Model (5) tests the influence of local official corruption on the local government debt scale through government investment efficiency. The results show a significant negative correlation between IE and DEBT ( $λ_{2} = - 1.366$ , $p < 0.01$ ), indicating that there is a mediating effect. H2b is preliminarily verified. There is a significant negative correlation between CORR and DEBT ( $λ_{1} = - 0.441$ , $p < 0.01$ ), indicating that the mediating effect of government investment efficiency is partial. This provides strong evidence in favor of Hypothesis 2b. That is, corruption expands the scale of government debt by affecting government investment efficiency.

This effect could arise because corrupt officials increase inappropriate investments to increase their personal income, reducing government investment efficiency, and driving large gaps in government fiscal. The local government would in turn enlarge local government debt to relieve the pressure from these fiscal gaps.

Heterogeneity Analysis

The analysis of the descriptive statistical results shows great differences in each variable across regions, which may be caused by differences in the level of economic development.

Therefore, to further test the regional differences in the influence of local official corruption on local government debt, this paper divides the sample into three subsamples in accordance with the scheme employed by China’s National People’s Congress: eastern, central, and western regions. The regression results are shown in Table 5.

Table 5.
Regional Test Results of the Influence of Corruption on the Local Government Debt.

Eastern region Central region Western region

(1) (2) (3) (4) (5) (6) (7) (8) (9)

DEBT IE DEBT DEBT IE DEBT DEBT IE DEBT

CORR 0.431 (1.28) −0.146* (−1.78) 0.287 (0.86) −0.746* (−3.00) −0.065 (−0.93) −0.870* (−4.13) −1.009* (−3.03) 0.209 (2.17) −0.655 (−2.18)

IE −0.987 (−2.32) −1.898* (−5.71) −1.695* (−4.76)

GGDP −5.712* (−4.35) 1.374* (4.31) −4.355* (−3.11) −2.632 (−2.26) 0.791 (2.40) −1.131 (−1.11) −4.235* (−2.65) 2.957* (6.41) 0.779 (0.45)

INV 2.478* (11.08) −0.459* (−8.45) 2.025* (6.93) 0.0954 (0.25) −0.479* (−4.36) −0.813 (−2.24) 1.374*** (3.59) −0.209* (−1.89) 1.019* (2.98)

FE −1.226* (−4.52) 0.563*** (8.55) −0.670* (−1.88) 2.626* (4.67) 0.498* (3.12) 3.571* (7.11) 0.234 (0.48) 0.562* (3.99) 1.187 (2.52)

FRD −0.346 (−1.40) −0.0373 (−0.62) −0.383 (−1.59) −0.123 (−0.53) −0.384* (−5.91) −0.852* (−3.68) −0.0296 (−0.08) 0.0338 (0.32) 0.0276 (0.09)

MAR 6.403* (6.00) −1.448* (−5.59) 4.973* (4.13) 2.561*** (3.69) −0.325* (−1.65) 1.944* (3.27) 2.802* (2.79) 0.535* (1.85) 3.709* (4.15)

FIN 1.583* (3.28) 0.264 (2.25) 1.843* (3.82) 0.920* (3.32) −0.027 (−0.34) 0.869* (3.72) 1.604* (3.41) −0.365* (−2.68) 0.985 (2.29)

Cons −34.32* (−7.37) 4.868* (4.30) −29.52* (−5.93) −23.96* (−7.41) 1.816 (1.98) −20.51* (−7.35) −17.63* (−5.31) −5.446* (−5.68) −26.87* (−7.71)

N 88 88 88 80 80 80 72 72 72

Wald 655.91 102.73 701.31 614.24 107.78 897.20 561.73 180.87 761.38

Note. t Statistics in parentheses.
*
p < .1. p < .05. p < .01.

As seen from Table 5, for the eastern region, the coefficient between CORR and DEBT is positive ( $α_{1} = 0.431$ )but fails the statistical significance test. In addition, since Model (1) is the first step of the mediating effect test, the mediating effect of government investment efficiency fails the test. That is, government investment efficiency does not play a mediating role in the influence of corruption on local government debt.

For the central region, there is a significant negative correlation between CORR and DEBT ( $α_{1} = - 0.746$ , $p < 0.01$ ), indicating that corruption promotes the growth of local government debt. However, in the second step in the mediating effect test in model (5), corruption and government investment efficiency are not significantly negatively correlated. Thus, the mediating effect of the government investment efficiency does not pass the significance test.

For western China, there is a significant negative correlation between CORR and DEBT ( $α_{1} = - 1.009$ , $p < 0.01$ ), indicating that corruption promotes the growth of local government debt. The mediating effect of IE also passes the test ( $β_{1} = 0.209$ , $p < 0.05$ and $λ_{2} = - 1.695$ , $p < 0.01$ ) and is identified as a partial mediating effect, indicating that corruption expands local government debt through government investment efficiency. However, compared with the significance of the results for the full sample, the significance of CORR in Model (9) is reduced ( $λ_{1} = - 0.655$ , $p < 0.05$ ).

The comparison of the results for the three regions is as follows.

(1) The coefficient of corruption in Model (4) is greater than that in Model (7) ( $- 0.746 > - 1.009$ ), but because CORR is a reverse indicator, the effect of local official corruption on the local government debt scale in the western region is larger than that in the central region. And then contact the result of eastern region, it can be found that this effect grows from the eastern to the central and western regions, with no direct effect in the former region but a direct effect detected in the latter regions.

(2) Only for the western region is there a mediating effect of government investment efficiency; that is, corruption in the western region expands local government debt through the channel of the government investment efficiency. And there is no the mediating effect in the eastern region and central region, which shows the mediating effect of the government investment efficiency is gradually enhanced, from no mediating effect in eastern and central regions to the partial mediating effect in western region.

This result may be because the economic development of the eastern region is the highest among the three regions, followed by the central region, and the western region.² This could make the implementation of the debt risk prevention and the anti-corruption campaign in the eastern region more efficient than that in the central and western regions. Corruption and debt are thus well controlled and the direct and indirect effects of corruption on local government debt disappear. In contrast, the level of economic development in the central region is intermediate, and corruption has only a direct impact on local government debt. The western region is backward in terms of economic development, and its implementation of policies is poor. Therefore, the direct influence of corruption on the local government debt scale and the indirect influence through investment efficiency still exist.

Overall, the above results indicate that China’s debt risk prevention and anti-corruption campaign have achieved good results. Nonetheless, in the eastern and central regions, the influence of corruption on local government debt is weaker, and the mediating effect of government investment efficiency is suppressed.

Robustness Tests

To ensure the robustness of the empirical test results, we conduct robustness tests on the following two aspects: (1) the direct impact of corruption on local government debt and (2) the impact path of government investment efficiency. That is, we test whether the mediating effect of government investment efficiency in the link between corruption and local government debt is robust. This paper adopts the approach of employing an alternative empirical method to carry out the robustness tests.

Robustness Test of the Direct Effect

For the robustness test on the direct effect of corruption on local government debt, we adopt the following methods:

(1) Mixed effects model (MOLS). We test for a positive correlation between corruption and the local government debt scale without considering individual and cross-sectional variation.

(2) Lag phase model (LP). To avoid endogeneity from a reverse causal relationship between corruption and local government debt, whereby an increase in local government debt aggravates corruption, we implement a one-stage lag test on corruption.

(3) Fixed effect model (FE). To avoid the influence of regional differences on local government debt, we further test the positive correlation between corruption and local government debt through the fixed effect model.

(4) Quantile regression model (Q30, Q50, and Q70). To detect whether the positive correlation between corruption and local government debt is significant only in a certain quantile, we adopt quantile regression repeated 1,000 times to test whether the relationship holds at each quantile.

All the test results are summarized in Table 6.

Table 6.
Robustness Test Results of the Direct Effects.

Variable MOLS LP FE Q30 Q50 Q70

(1) (2) (4) (5) (6) (7)

CORR −0.694 (−3.76) −0.494* (−2.81) −0.838* (−3.55) −0.720 (−2.12) −0.716* (−3.37) −0.846* (−4.49)

Control Yes Yes Yes Yes Yes Yes

Cons −16.52* (−8.60) −18.35* (−10.18) −18.26* (−5.77) −15.81* (−4.46) −15.67* (−7.06) −15.88* (−8.09)

N 240 210 240 240 240 240

F (Wald) 183.34* 1,097.85* 1,722.98*

Adj R² (Pseudo R²) .842 .6150 .6421 .6473

Note. CORR in LP model is CORR_t₋₁. t statistics in parentheses.

p < .05. p < .01.

For the robustness test of the direct effect in Table 6, the results of the mixed effect model ( $α_{1} = - 0.694$ , $p < 0.01$ ), lag phase model ( $α_{1} = - 0.494$ , $p < 0.01$ ), fixed effect model ( $α_{1} = - 0.838$ , $p < 0.01$ ), and quantile regression model (Q30: $α_{1} = - 0.720$ , $p < 0.05$ ; Q50: $α_{1} = - 0.716$ , $p < 0.01$ ; Q70: $α_{1} = - 0.846$ , $p < 0.01$ ) all support the empirical results above; that is, greater corruption expands the local government debt scale, and the relationship between them is robust.

Robustness Test of the Mediating (Indirect) Effect

To ensure the robustness of the mediating effect of government investment efficiency, we adopt two methods: (1) The first is the Sobel Test (Sobel, 1982). The summary of the Sobel, Goodman-1, and Goodman-2 results is shown in Table 7.

Table 7.
Sobel Test Results of the Mediating Effect.

Variable (1) (2) (3)

DEBT IE DEBT

CORR −0.694 (−3.76) 0.185* (3.34) −1.366* (−6.85)

IE −0.744* (−3.45)

Control Yes Yes Yes

Cons −16.519* (−8.60) −3.873 (−6.71) −21.811* (−11.37)

N 240 240 240

F 183.34* 25.00* 198.01*

Adj R² .842 .413 .868

Sobel −0.252* (−2.999)

Goodman-1 (Aroian) −0.252* (−2.973)

Goodman-2 −0.252* (−3.025)

Ratio of indirect to direct effect 0.364

Note. t Statistics in parentheses.
*
p < .01.

In Table 7, the Sobel, Goodman-1, and Goodman-2 results show that the mediating effect exists and is significant ( $t = - 0.252$ and $p < 0.01$ ); government investment efficiency plays a partial mediating effect with a proportion of the total effect of 36.4%.

(2) The second method is the bootstrap test (Hayes, 2017). A total of 1,000 bootstrap tests are performed, and a summary of the results is shown in Table 8.

Table 8.
Bootstrap Test Results of the Mediating Effect.

Observed coef. Bias SE [95% CI]

−.252* (−3.09) 0.001 0.082 [−0.414, −0.099] (P)

[−0.420, −0.106] (BC)

Note. t statistics in parentheses. P = percentile confidence interval; BC = bias-corrected confidence interval.
*
p < .01.

From Table 8, the mediating effect of government investment efficiency is 0.252 significantly ( $p < 0.01$ ). The percentile confidence interval is [−0.414, −0.099] at the 95% confidence level, and the bias-corrected confidence interval is [−0.420, −0.106]. Neither interval contains zero, so the mediating effect of government investment efficiency in the impact of corruption on local government debt is significant (Preacher & Hayes, 2004).

The above two methods show that the test results of the mediating effect are consistent with the previous research results. Therefore, the robustness of the mediating effect is verified.

Conclusion

In the context of debt risk prevention and the anti-corruption campaign in China, this paper takes 30 provinces as the research object, selects panel data from 2010 to 2017 as the sample, and discusses the direct effect of corruption on local government debt and the indirect effect from the perspective of government investment efficiency.

Discussion

The results show the following. (1) Local official corruption directly increases the local government debt. This conclusion is consistent with the conclusions of research on capitalist countries such as the United States (Liu et al., 2017). This paper finds that under China’s socialist political and economic system, corrupt officials abuse their power to overinvest for their own benefit. At the same time, local official corruption worsens local governance and debt management, which promotes excessive expansion of local government debt.

(2) Local official corruption reduces government investment efficiency and further affects the local government debt scale. Existing research suggests that corruption reduces the efficiency of government investment (Haque & Kneller, 2008) and that more government investment also increases public debt (Belguith & Omrane, 2017). In a further analysis of government investment, this paper finds that through corruption, local officials advance their own interests by inappropriately increasing government investment. This leads to a rise in the cost of government investment and a distortion of public resource allocation, which reduces the efficiency of government investment, increases fiscal gaps, and ultimately inflates local government debt.

(3) Through the analysis of regional heterogeneity, it is found that the direct and indirect effect are gradually rise from the eastern to central and western regions in China. In other words, the higher the region’s economic development, the weaker is the impact of corruption on local government debt (see Note 2). This finding is a meaningful complement to existing country-level studies and is broadly consistent with the view that corruption has a weaker impact on public debt in developed or high-income countries than in developing or low-income countries (Ibrahim, 2020). This is because in developed region, corruption and debt are better controlled and more transparent, which reduces the mediation role of government investment efficiency and thus weakens the direct effect and the mediating effects.

Policy Implications

Based on the conclusions of this paper and China’s economic environment, we argue that the local government debt scale can be controlled in the following ways:

(1) Improving the political performance appraisal system in China. The current view of political performance that prioritizes “GDP worship” is a cause of local official corruption. Therefore, the political performance appraisal system should consider both present and future development to ensure that government officials do not engage in corruption to enhance their promotion prospects and economic achievements and create a false prosperity in the local economy by increasing local government debt.

(2) Improving government investment efficiency. There are two main aspects of this implication. First, the administration of government investment project examination and approval should be strengthened. In the construction process for government investment projects, supervision and management should be carried out continuously. Relevant departments and individuals involved in the project examination and approval should be strictly supervised to prevent corruption so that the expansion of local government debt can be restrained to some extent. Second, the transparency of government information should be improved. This means that the relevant financial budget, cash flow, and investment funds of the government should be supervised by higher-level authorities and the public. This will increase the criminal costs and psychological pressure associated with corruption among government officials. Dual supervision can force government officials to comply with all normative procedures and refuse corruption (Wei et al., 2021), making the fiscal expenditure budget more reasonable and reducing unnecessary expenditure, so that the expansion of debt can be curbed to a certain extent.

(3) Increasing the efforts on policy implementation in less developed regions. For these regions, the efforts on policy implementation should be appropriately increased. In addition, we emphasize promoting debt risk prevention and the anti-corruption campaign. In more economically backward regions, local governments may put more effort into promoting economic development and ignore anticorruption and debt risk control, but the influence of corruption on debt still exists. In the long run, this will lead to economic losses and slow growth. Therefore, in regions with lower economic development, more attention should be paid to policy implementation to better promote local development.

Limitations and Future Directions

The research explores the mechanism whereby local government official corruption impacts local government debt and presents certain theoretical contributions and policy implications for controlling the local government debt scale in China. However, there are still many limitations that can be explored in future research. First, this paper explores the role of local government official corruption in local government debt growth at the regional level under China’s socialist system, and future research can be extended to other socialist countries. Second, this paper analyzes the mechanism whereby local official corruption affects local government debt, namely, via government investment efficiency, which is a component of fiscal expenditure; further research can discuss channels related to fiscal revenue, such as taxation (Yaru & Raji, 2022). Finally, the local government debt scale is measured in this paper by local government bonds and urban investment bonds. Under China’s political and economic system, local government contingent debt is also an important component of government debt, although there is controversy surrounding contingent debt statistics. With the deepening of research, contingent debt can be considered in the scope of the debt scale.

Items	Indicators	Description	Unit
Economic investment	Per capita investment in fixed assets		Yuan
Social investment	Per capita expenditure for General public services		Yuan
Per capita expenditure for education		Yuan
Per capita expenditure for science and technology		Yuan
Per capita expenditure for Health care		Yuan
Per capita expenditure for social security and employment		Yuan
Per capita expenditure for energy conservation and environment protection		Yuan
Economic output^a	Strength of economic development	Per capita GDP	Yuan
Capabilities of economic development	Fiscal revenue/ GDP
Stability of economic development	Reciprocal of the coefficient of variation of GDP growth rate
Rationality of economic development	1—Industrial Structure Theil Index
Outward-orientation of economic development	Total imports and exports/GDP
Social output	General public services output	Road network density^b	km/100 km²
Per capita area of paved roads	m²
Number of public buses and trolley buses per 10,000 persons	Unit
Length of city sewage pipes	10,000 km
Per capita electricity consumption	kwh
Education output	Per capita education years^c	Year
Average school enrolment per 100,000 population of higher education	Person
Science and technology output	Number of patent applications	Pieces
Full-time equivalent of R&D personnel	Man-year
Health care output	Number of beds in health care institutions per 1,000 persons	Bed
Health technical personnel in health care institutions per 1,000 persons	Person
Social security output	Number of employed persons in urban units	10,000 persons
Urban-rural income gap^d
Ecological environment output	Per capita area of parks and green land	m²
Common industrial solid wasteIntegrate reused rate	%
GDP output per sulfur dioxide emissions	10,000 yuan/kg

Type	Variable	Abbreviation	Definition
Independent variable	Local government debt	DEBT	Local government debt Outstanding
Dependent variable	Local official corruption	CORR	Number of corruption case/number of public officials
Mediating variable	Government investment efficiency	IE	Calculated though DEA model
Control variable	GDP growth rate	GGDP	(GDP of this year—GDP of last year)/GDP of last year
Government investment	INV	Total investment in fixed asset
Fiscal expenditure	FE	Fiscal expenditure
Fiscal decentralization	FD	Per capita local government fiscal revenue/per capita central government fiscal revenue
Marketization	MAR	Tertiary industry in the GDP
Financial development	FIN	Increase in financial sector gross product/GDP

Variables	Units	M	SD	Minimum	Maximum
DEBT	100 million yuan	2,462.571	3,445.45	10	23,343.68
CORR		22.604	6.603	7.66	42.04
IE		0.729	0.189	0.28	1
GGDP	%	0.116	0.07	−0.22	0.27
INV	100 million yuan	15,348.361	10,923.472	1,016.9	55,202.719
FE	100 million yuan	4,154.232	2,269.15	557.53	15,043.09
FD		3.549816	2.623425	1.158271	13.35832
URB		55.981	12.705	33.81	89.6
MAR		44.194	9.411	28.6	80.56

Variable	First step	Second step	Third step
CORR	−1.287*** (−3.42)		−0.694*** (−3.83)	0.185*** (3.39)	−0.441*** (−2.61)
IE					−1.366*** (−6.98)
GGDP		−4.684*** (−5.73)	−4.816*** (−6.06)	1.948*** (8.15)	−2.153*** (−2.63)
INV		1.301*** (8.07)	1.417*** (8.89)	−0.248*** (−5.19)	1.077*** (7.03)
FE		0.190 (0.87)	0.0578 (0.27)	0.493*** (7.66)	0.732*** (3.36)
FD		−0.450*** (−3.75)	−0.562*** (−4.67)	0.0630* (1.74)	−0.476*** (−4.31)
MAR		3.120*** (6.53)	2.668*** (5.58)	0.402*** (2.80)	3.217*** (7.26)
FIN		1.392*** (6.63)	1.537*** (7.41)	−0.323*** (−5.19)	1.095*** (5.49)
Cons	10.61*** (9.12)	−20.23*** (−12.12)	−16.52*** (−8.74)	−3.873*** (−6.82)	−21.81*** (−11.59)
N	240	240	240	240	240
Wald	11.69***	1,237.34***	1,327.60***	181.01***	1,645.81***

	Eastern region	Central region	Western region
CORR	0.431 (1.28)	−0.146* (−1.78)	0.287 (0.86)	−0.746*** (−3.00)	−0.065 (−0.93)	−0.870*** (−4.13)	−1.009*** (−3.03)	0.209** (2.17)	−0.655** (−2.18)
IE			−0.987** (−2.32)			−1.898*** (−5.71)			−1.695*** (−4.76)
GGDP	−5.712*** (−4.35)	1.374*** (4.31)	−4.355*** (−3.11)	−2.632** (−2.26)	0.791** (2.40)	−1.131 (−1.11)	−4.235*** (−2.65)	2.957*** (6.41)	0.779 (0.45)
INV	2.478*** (11.08)	−0.459*** (−8.45)	2.025*** (6.93)	0.0954 (0.25)	−0.479*** (−4.36)	−0.813** (−2.24)	1.374*** (3.59)	−0.209* (−1.89)	1.019*** (2.98)
FE	−1.226*** (−4.52)	0.563*** (8.55)	−0.670* (−1.88)	2.626*** (4.67)	0.498*** (3.12)	3.571*** (7.11)	0.234 (0.48)	0.562*** (3.99)	1.187** (2.52)
FRD	−0.346 (−1.40)	−0.0373 (−0.62)	−0.383 (−1.59)	−0.123 (−0.53)	−0.384*** (−5.91)	−0.852*** (−3.68)	−0.0296 (−0.08)	0.0338 (0.32)	0.0276 (0.09)
MAR	6.403*** (6.00)	−1.448*** (−5.59)	4.973*** (4.13)	2.561*** (3.69)	−0.325* (−1.65)	1.944*** (3.27)	2.802*** (2.79)	0.535* (1.85)	3.709*** (4.15)
FIN	1.583*** (3.28)	0.264** (2.25)	1.843*** (3.82)	0.920*** (3.32)	−0.027 (−0.34)	0.869*** (3.72)	1.604*** (3.41)	−0.365*** (−2.68)	0.985** (2.29)
Cons	−34.32*** (−7.37)	4.868*** (4.30)	−29.52*** (−5.93)	−23.96*** (−7.41)	1.816** (1.98)	−20.51*** (−7.35)	−17.63*** (−5.31)	−5.446*** (−5.68)	−26.87*** (−7.71)
N	88	88	88	80	80	80	72	72	72
Wald	655.91	102.73	701.31	614.24	107.78	897.20	561.73	180.87	761.38

Variable	MOLS	LP	FE	Q30	Q50	Q70
CORR	−0.694*** (−3.76)	−0.494*** (−2.81)	−0.838*** (−3.55)	−0.720** (−2.12)	−0.716*** (−3.37)	−0.846*** (−4.49)
Control	Yes	Yes	Yes	Yes	Yes	Yes
Cons	−16.52*** (−8.60)	−18.35*** (−10.18)	−18.26*** (−5.77)	−15.81*** (−4.46)	−15.67*** (−7.06)	−15.88*** (−8.09)
N	240	210	240	240	240	240
F (Wald)	183.34***	1,097.85***	1,722.98***
Adj R² (Pseudo R²)	.842			.6150	.6421	.6473

Variable	(1)	(2)	(3)
CORR	−0.694*** (−3.76)	0.185*** (3.34)	−1.366*** (−6.85)
IE			−0.744*** (−3.45)
Control	Yes	Yes	Yes
Cons	−16.519*** (−8.60)	−3.873 (−6.71)	−21.811*** (−11.37)
N	240	240	240
F	183.34***	25.00***	198.01***
Adj R²	.842	.413	.868
Sobel	−0.252*** (−2.999)
Goodman-1 (Aroian)	−0.252*** (−2.973)
Goodman-2	−0.252*** (−3.025)
Ratio of indirect to direct effect	0.364

Observed coef.	Bias	SE	[95% CI]
−.252*** (−3.09)	0.001	0.082	[−0.414, −0.099]	(P)
			[−0.420, −0.106]	(BC)

Footnotes

We thank editors for excellent editorial guidance and anonymous reviewers for their helpful comments.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research,authorship,and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research,authorship,and/or publication of this article: This work was supported by the National Natural Science Foundation of China (NSFC) under Grant [number 71971169] and Shaanxi Province Philosophy and Social Science Research Special Youth Project under Grant [number 2023QN0165].

ORCID iD

Yang Li

References

Abada

I. M.

Ngwu

E. C.

(2019). Corruption, governance, and Nigeria’s uncivil society, 1999-2016. Análise Social, 54(2312), 386–408. https://doi.org/10.31447/AS00032573.2019231.07

Abubakar

A. B.

(2020). Does fiscal tightening (loosening) reduce public debt? African Development Review, 32(4), 528–539. https://doi.org/10.1111/1467-8268.12458

Amoah

Asiama

R. K.

Korle

Kwablah

(2022). Corruption: Is it a bane to renewable energy consumption in Africa? Energy Policy, 163, 112854. https://doi.org/10.1016/j.enpol.2022.112854

Baklouti

Boujelbene

(2021). Corruption, democracy, and public debt: A case of the Arab countries. Journal of the Knowledge Economy, 177, 111–133. https://doi.org/10.1007/s11127-018-0592-7

Baron

R. M.

Kenny

D. A.

(1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173. https://doi.org/10.1037/0022-3514.51.6.1173

Beck

P. J.

Maher

M. W.

(1986). A comparison of bribery and bidding in thin markets. Economics Letters, 20(1), 1–5. https://doi.org/10.1016/0165-1765(86)90068-6

Belguith

Omrane

(2017). Macroeconomic determinants of public debt growth: A case study for Tunisia. Theoretical & Applied Economics, 24(4), 161–168. http://store.ectap.ro/articole/1314.pdf

Benfratello

Del Monte

Pennacchio

(2018). Corruption and public debt: A cross-country analysis. Applied Economics Letters, 25(5), 340–344. https://doi.org/10.1080/13504851.2017.1321831

Bhandari

Evans

Golosov

Sargent

T. J.

(2017). Fiscal policy and debt management with incomplete markets. The Quarterly Journal of Economics, 132(2), 617–663. https://doi.org/10.1093/qje/qjw041

10.

Blanchard

Shleifer

(2001). Federalism with and without political centralization: China versus Russia. IMF Staff Papers, 48(1), 171–179. https://doi.org/10.2307/4621694

11.

Brada

J. C.

Drabek

Mendez

J. A.

Perez

M. F.

(2019). National levels of corruption and foreign direct investment. Journal of Comparative Economics, 47(1), 31–49. https://doi.org/10.1016/j.jce.2018.10.005

12.

Brown

N. C.

Smith

J. D.

White

R. M.

Zutter

C. J.

(2019). Political corruption and firm value in the US: Do rents and monitoring matter? Journal of Business Ethics, 168(4), 1–17. https://doi.org/10.1007/s10551-019-04181-0

13.

Cai

(2017). Nonlinear analysis of economic growth, public debt and policy tools. Asian Economic and Financial Review, 7(1), 99.

14.

Cang

W. M.

Matt

Z. M.

(2017). Will rising debt in China lead to a hard landing? International Journal of Economics and Finance, 9(9), 60–69. http://ccsenet.org/journal/index.php/ijef/article/view/69961/38090

15.

Central Commission for Discipline Inspection. (2021). A work report on the Fifth Plenary Session of the 19th CPC Central Commission for Discipline Inspection. Retrieved March 15, 2021, from http://www.ccdi.gov.cn/xxgk/hyzl/202103/t20210315_237935.html

16.

Chatterjee

Gibson

Rioja

(2017). Optimal public debt redux. Journal of Economic Dynamics and Control, 83, 162–174. https://doi.org/10.1016/j.jedc.2017.08.005

17.

Chen

Yao

Lin

(2017). Optimal government investment and public debt in an economic growth model. China Economic Review, 45, 257–278. https://doi.org/10.1016/j.chieco.2016.08.005

18.

Chen

Zhou

L. A.

(2005). Relative performance evaluation and the turnover of provincial leaders in China. Economics Letters, 88(3), 421–425. https://doi.org/10.1016/j.econlet.2005.05.003

19.

Chudik

Mohaddes

Pesaran

M. H.

Raissi

(2018). Rising public debt to GDP can harm economic growth. Economic Letter, 13(3), 1–4. https://www.dallasfed.org/research/~/media/documents/research/eclett/2018/el1803.pdf

20.

Cooray

Dzhumashev

Schneider

(2017). How does corruption affect public debt? An empirical analysis. World Development, 90, 115–127. https://doi.org/10.1016/j.worlddev.2016.08.020

21.

Deng

(2018). The National Supervision Commission: A new anti-corruption model in China. International Journal of Law, Crime and Justice, 52, 58–73. https://doi.org/10.1016/j.ijlcj.2017.09.005

22.

Deng

Zhang

(2020). The new economy in China: An intercity comparison. Sage Open, 10(4), 2158244020977870. https://doi.org/10.1177/2158244020977870

23.

Feng

(2013). Local government debt and municipal bonds in China: Problems and a framework of rules. The Copenhagen Journal of Asian Studies, 31(2), 23–53. https://doi.org/10.22439/cjas.v31i2.4332

24.

Glaeser

E. L.

Saks

R. E.

(2006). Corruption in america. Journal of public Economics, 90(6–7), 1053–1072. https://doi.org/10.1016/j.jpubeco.2005.08.007

25.

Gründler

Potrafke

(2019). Corruption and economic growth: New empirical evidence. European Journal of Political Economy, 60, 101810. https://doi.org/10.1016/j.ejpoleco.2019.08.001

26.

Guo

Wang

Yang

(2021). China’s anti-corruption shock and resource reallocation in the energy industry. Energy Economics, 96, 105182. https://doi.org/10.1016/j.eneco.2021.105182

27.

Haque

M. E.

Kneller

(2008). Public investment and growth: The role of corruption (Discussion paper series 98). Centre for Growth and Business Cycle Research. https://hummedia.manchester.ac.uk/schools/soss/cgbcr/discussionpapers/dpcgbcr98.pdf

28.

Hayes

A. F.

(2017). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford Publications. https://www.guilford.com/books/Introduction-to-Mediation-Moderation-and-Conditional-Process-Analysis/Andrew-Hayes/9781462549030

29.

(2000). Corruption and anti-corruption in reform China. Communist and Post-Communist Studies, 33(2), 243–270. https://doi.org/10.1016/S0967-067X(00)00006-4

30.

Zhang

(2020). Incorporating emergency management into public administration education: The case of China. Journal of Public Affairs Education, 26(2), 228–249. https://doi.org/10.1080/15236803.2019.1702364

31.

Huang

Panizza

Varghese

(2018). Does public debt crowd out corporate investment? (No. BOOK). Graduate Institute of International and Development Studies, International Economics Department. http://repository.graduateinstitute.ch/record/296011/files/HEIDWP08-2018.pdf

32.

Huntington

S. P.

(2006). Political order in changing societies. Yale University Press. https://yalebooks.yale.edu/book/9780300116205/political-order-in-changing-societies/

33.

Ibrahim

(2020). Corruption, public debt and economic growth–evidence from developing countries. International Journal of Development Issues, 20(1), 24–37. https://doi.org/10.1108/IJDI-12-2019-0208

34.

Imamah

Lin

T. J.

Suhadak

Hung

J. H.

(2020). The moderating effect of an anti-corruption campaign on the relation between political connections and investment behavior-the case of Chinese listed firms. Australasian Accounting, Business and Finance Journal, 14(1), 14–32. https://doi.org/10.14453/aabfj.v14i1.3

35.

Institute of International Finance. (2021). Global Debt Monitor: COVID Drives Debt Surge - Stabilization Ahead?. Retrieved February 17, 2021, from https://www.iif.com/Research/Capital-Flows-and-Debt/Global-Debt-Monitor

36.

International Monetary Fund. (2015). Making Public Investment More Efficient. Staff Report. International Monetary Fund: Washington, DC. https://www.imf.org/-/media/Websites/IMF/imported-full-text-pdf/external/np/pp/eng/2015/_061115.ashx

37.

Ivanyna

Mourmouras

Rangazas

(2018). The macroeconomics of corruption. Springer Texts in Business and Economics. https://doi.org/10.1007/978-3-319-68666-0

38.

Kalim

(2023). Sukuk as an alternative source of public debt. In Mehdi Khosrow-Pour (Ed.) In Research anthology on macroeconomics and the achievement of global stability (pp. 1225–1239). IGI Global. https://doi.org/10.4018/978-1-6684-7460-0.ch066

39.

Kim

D. S.

Tarzia

(2018). Value of corruption in China: Evidence from anti-corruption investigation. Economics Letters, 164, 112–116. https://doi.org/10.1016/j.econlet.2018.01.021

40.

Samajdar

(2010). Evaluation of international corruption indexes: Should we believe them or not?. Social Science Journal, 47(3), 508–540. https://doi.org/10.1016/j.soscij.2010.03.001

41.

Kong

Tao

Wang

(2020). China’s anti-corruption campaign and firm productivity: Evidence from a quasi-natural experiment. China Economic Review, 63, 101535. https://doi.org/10.1016/j.chieco.2020.101535

42.

Krueger

A. O.

(1974). The political economy of the rent-seeking society. The American Economic Review, 64(3), 291–303. https://www.jstor.org/stable/1808883

43.

Lam

W. R.

Wang

(2018). China’s Local Government Bond Market. International Monetary Fund.

44.

Lambsdorff

J. G.

(2003). How corruption affects productivity. Kyklos, 56(4), 457–474. https://doi.org/10.1046/j.0023-5962.2003.00233.x

45.

Leff

N. H.

(1964). Economic development through bureaucratic corruption. American Behavioral Scientist, 8(3), 8–14. https://doi.org/10.1177/000276426400800303

46.

Liang

Shi

Wang

(2017). Local government debt and firm leverage: Evidence from China. Asian Economic Policy Review, 12(2), 210–232. https://doi.org/10.1111/aepr.12176

47.

Zhou

L. A.

(2005). Political turnover and economic performance: The incentive role of personnel control in China. Journal of Public Economics, 89(9–10), 1743–1762. https://doi.org/10.1016/j.jpubeco.2004.06.009

48.

Lim

K. Y.

(2019). Modelling the dynamics of corruption and unemployment with heterogeneous labour. Economic Modelling, 79, 98–117. https://doi.org/10.1016/j.econmod.2018.10.004

49.

Liu

Moldogaziev

T. T.

Mikesell

J. L.

(2017). Corruption and state and local government debt expansion. Public Administration Review, 77(5), 681–690. https://doi.org/10.1111/puar.12711

50.

Lui

F. T.

(1985). An equilibrium queuing model of bribery. Journal of Political Economy, 93(4), 760–781. https://doi.org/10.1086/261329

51.

Mahmah

A. E.

Kandil

(2019). Fiscal sustainability challenges in the new normal of low oil prices: Empirical evidence from GCC countries. International Journal of Development Issues, 18(1), 109–134. https://doi.org/10.1108/IJDI-02-2018-0033

52.

Meza

Pérez-Chiqués

(2021). Corruption consolidation in local governments: A grounded analytical framework. Public Administration, 99(3), 530–546. https://doi.org/10.1111/padm.12698

53.

Ministry of Finance of the People’s Republic of China. (2022). Local government bond issuance and debt balance in December 2021. Retrieved February 9, from, 2022. http://yss.mof.gov.cn/zhuantilanmu/dfzgl/sjtj/202202/t20220209_3786613.htm

54.

Ministry of Finance of the People’s Republic of China. (2021). Transcript of the State Council’s regular policy briefings on strengthening and improving management of local government special bonds. Retrieved December 16, from, 2021. https://www.mof.gov.cn/zhengwuxinxi/caizhengxinwen/202112/t20211216_3775680.htm

55.

Mitchell

W. C.

Niskanen

W. A.

(1971). Bureaucracy and representative government. American Political Science Association, 68(4), 1775. https://doi.org/10.2307/1959997

56.

Mlachila

Tapsoba

S. J.

(2017). A quality of growth index for developing countries: A proposal. Social Indicators Research, 134(2), 675–710. https://doi.org/10.1007/s11205-016-1439-6

57.

National People’s Congress of the People’s Republic of China. (2015). The Standing Committee of the National People’s Congress approved a ceiling of 16 trillion yuan for local government debt in 2015. Retrieved August 29, From, 2022http://www.npc.gov.cn/zgrdw/npc/xinwen/2015-08/31/content_1945486.htm

58.

Nieto-Morales

Ríos

(2022). Human resource management as a tool to control corruption: Evidence from Mexican municipal governments. Public Administration, 100(4), 1019–1036. https://doi.org/10.1111/padm.12782

59.

Ogbeifun

Shobande

(2020). Debt sustainability and the fiscal reaction function: Evidence from MIST countries. Future Business Journal, 6(1), 1–8. https://doi.org/10.1186/s43093-020-00037-6

60.

Paldam

(2021). The transition of corruption institutions and dynamics. European Journal of Political Economy, 67, 101952. https://doi.org/10.1016/j.ejpoleco.2020.101952

61.

Pei

(2007). Corruption threatens China’s future (Vol. 55, pp. 1–8). Carnegie Endowment for International Peace. http://carnegieendowment.org/files/pb55_pei_china_corruption_final.pdf

62.

Pei

(2018). How Not to fight corruption: Lessons from China. Daedalus, 147(3), 216–230. https://doi.org/10.1162/daed_a_00512

63.

Potrafke

(2019). Electoral cycles in perceived corruption: International empirical evidence. Journal of Comparative Economics, 47(1), 215–224. https://doi.org/10.1016/j.jce.2018.11.003

64.

Preacher

K. J.

Hayes

A. F.

(2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods Instruments & Computers, 36(4), 717–731. https://doi.org/10.3758/BF03206553

65.

Qian

(1993). Why China’s economic reforms differ: The M-form hierarchy and entry/expansion of the non-state sector. Economics of Transition, 1(2), 135–170. https://doi.org/10.7551/mitpress/9780262534246.003.0011

66.

Slagter

Sylwester

Doiron

(2019). Explaining the standard errors of corruption perception indices. Journal of Comparative Economics, 47(4), 907–920. https://doi.org/10.1016/j.jce.2019.07.003

67.

Rubin

D. B.

(1996). Multiple imputation after 18+ years. Journal of the American statistical Association, 91(434), 473–489. https://doi.org/10.2307/2291635

68.

Sarangi

El-Ahmadieh

(2017). Fiscal policy response to public debt in the Arab region. Economic and Social Commission for Western Asia.

69.

Shleifer

Vishny

R. W.

(1993). Corruption. The Quarterly Journal of Economics, 108(3), 599–617. https://doi.org/10.2307/2118402

70.

Sobel

M. E.

(1982). Asymptotic confidence intervals for indirect effects in structural equation models. Sociological Methodology, 13, 290–312. https://doi.org/10.2307/270723

71.

Tao

(2012). Local officials’ incentives and China’s economic growth: Tournament thesis reexamined and alternative explanatory framework. China & World Economy, 20(4), 1–18. https://doi.org/10.1111/j.1749-124X.2012.01292.x

72.

(2023). Corruption in social relations: Bribe-taking, corruptibility, and corruption risks. Public Administration, 101(2), 481–499. https://doi.org/10.1111/padm.12795

73.

Sun

Bwanali

T. R.

Berry

(2021). Internationalization of public administration education in China: International students’ perspective. Journal of Public Affairs Education, 27(1), 72–95. https://doi.org/10.1080/15236803.2019.1699984

74.

Swaleheen

Ben Ali

M. S.

Temimi

(2019). Corruption and public spending on education and health. Applied Economics Letters, 26(4), 321–325. https://doi.org/10.1080/13504851.2018.1468549

75.

Teles

V. K.

Mussolini

C. C.

(2014). Public debt and the limits of fiscal policy to increase economic growth. European Economic Review, 66, 1–15. https://doi.org/10.1016/j.euroecorev.2013.11.003

76.

Todd

Yates

(2020). Public debt and democratic statecraft in nineteenth-century france. In Barreyre

Delalande

(Eds.) A world of public debts (pp. 79–106). Palgrave Macmillan. https://doi.org/10.1007/978-3-030-48794-2_4

77.

Transparency International. (2012). Corruption perceptions index 2012. Working Papers, 45(9), 17986A–17986B. https://doi.org/10.1111/j.1467-6346.2008.01935.x

78.

Ukeme

A. N.

Ifayemi

M. O.

(2020). Effect of tax revenue on public debt and capital expenditure in Nigeria. Journal of Taxation and Economic Development, 19(2), 82–103. https://journal.citn.org/home/journal/57

79.

United Nations Office on Drugs and Crime. (2004). United Nations convention against corruption. Retrieved October 31, from, 2022 https://www.unodc.org/documents/treaties/UNCAC/Publications/Convention/08-50026_E.pdf

80.

Wederman

(2004). The intensification of corruption in China. The China Quarterly, 180, 895–921. https://doi.org/10.1017/S0305741004000670

81.

Wei

Guo

(2021). Dancing on a tightrope: The reputation management of local governments in response to public protests in China. Public Administration, 99(3), 547–562. https://doi.org/10.1111/padm.12699

82.

Whiting

S. H.

(2006). Power and wealth in rural China: The political economy of institutional change. Cambridge University Press.

83.

Yano

(2017). How does anti-corruption affect corporate innovation? Evidence from recent anti-corruption efforts in China. Journal of Comparative Economics, 45(3), 498–519. https://doi.org/10.1016/j.jce.2016.10.001

84.

Yamarik

Redmon

(2017). Economic freedom and corruption: New cross-country panel data evidence. Journal of Private Enterprise, 32(2), 17–44. https://ssrn.com/abstract=3232261

85.

Yaru

M. A.

Raji

A. S.

(2022). Corruption, governance and tax revenue performance in Sub-Saharan Africa. African Journal of Economic Review, 10(1), 234–253. https://www.ajol.info/index.php/ajer/article/view/219361/207019

86.

Young

S. L.

Wiley

K. K.

Searing

E. A.

(2020). “Squandered in real time”: How public management theory underestimated the public administration–politics dichotomy. The American Review of Public Administration, 50(6–7), 480–488. https://doi.org/10.1177/0275074020941669

87.

Zhang

Song

Tan

Xia

Zhang

Jiang

Xiong

Cheng

Zhang

(2019). Anti-corruption efforts, public perception of corruption, and government credibility in the field of real estate: An empirical analysis based on twelve provinces in China. Cities, 90, 64–73. https://doi.org/10.1016/j.cities.2019.01.042

Impact of Local Official Corruption on Local Government Debt in China: The Mediating Role of Government Investment Efficiency

Abstract

Keywords

Introduction

Literature Review

Hypothesis Development

Direct Impact of Local Official Corruption on Local Government Debt

Indirect Impact of Local Official Corruption on Local Government Debt Through Government Investment Efficiency

Research Design

Variables

Dependent Variable

Independent Variable

Mediating Variable

Control Variables

Data

Model

Empirical Analyses

Basic Statistics

Results Analysis

Heterogeneity Analysis

Robustness Tests

Robustness Test of the Direct Effect

Robustness Test of the Mediating (Indirect) Effect

Conclusion

Discussion

Policy Implications

Limitations and Future Directions

Footnotes

Declaration of Conflicting Interests

Funding

ORCID iD

References