Economics Letters 224 (2023) 110999

Contents lists available at ScienceDirect

Economics Letters
journal homepage: www.elsevier.com/locate/ecolet

The role of political connections in bad times: Evidence from the

COVID-19 pandemic
∗
Hengxu Liu a , , Wenxi Zhao b
a
School of Accounting, Zhongnan University of Economics and Law, Wuhan, Hubei, China
b
School of Public Administration, Zhongnan University of Economics and Law, Wuhan, Hubei, China

article info a b s t r a c t

Article history: This study investigates the relationship between political connections and firm financial performance
Received 8 September 2022 during the COVID-19 pandemic. Using a difference-in-differences methodology, we found that po-
Received in revised form 18 January 2023 litically connected enterprises paid more taxes, employed more employees, and suffered financial
Accepted 22 January 2023
performance. This study enriches the literature on the impact of COVID-19 on enterprises and provides
Available online 4 February 2023
suggestions for regulators.
JEL classification: © 2023 Elsevier B.V. All rights reserved.
G32
G34

Keywords:
COVID-19
Political connections
Corporate governance
Firm financial performance

1. Introduction which has an endogenous problem. We use the exogenous impact

of the COVID-19 pandemic to fill this gap.
The sudden outbreak of the COVID-19 pandemic led to pro- Few studies have examined the impact of political connec-
duction shutdowns in many organizations, increasing liquidity tions on firm behaviour during crises. On the one hand, political
risk and adversely affecting the capital market (Hu and Zhang, connections may act as a ‘‘helping hand’’, providing access to
2021; Ren et al., 2021; Shen et al., 2020). Enhancing the immunity financing and preferential access to government bailouts, ulti-
of enterprises during the pandemic is a topic worthy of further mately enhancing corporate value. On the other hand, political
research into how best organizations can survive in the COVID- connections may act as a ‘‘grabbing hand’’. Politicians are mo-
19 pandemic. We investigate the role of political connections in tivated to intervene in the operation of politically connected
bad times. enterprises to make them undertake the political objectives of
Political connections are widespread in global capital mar- expanding employment, actively paying taxes and maintaining
kets (Faccio, 2006). At present, there is no consensus on how stability. During the period of COVID-19, the unemployment rate
political connections affect financial performance. Some argue became higher and the financial pressure on local governments
that political connections bring a range of benefits, including became greater. To maintain social stability and ease the financial
bank loans, government contracts, government bailouts, avoid- pressure on local governments, politicians will urge politically
ance of regulatory penalties, and increased firm value (Brogaard connected enterprises to hire more labour and pay more taxes,
et al., 2021; Amore and Bennedsen, 2013). However, political thus damaging the value of enterprises. Because some politically
connections also have some drawbacks, including reduced ac- connected enterprises may be affected by the above two as-
counting quality, increased information risk, reduced cash hold- pects during the crisis, we constructed a difference-in-differences
ing value, increased insider opportunism, and harmed corporate model to investigate the impact of political connections on firm
value (Jagolinzer et al., 2020; La Rocca et al., 2022; Saeed et al., value using the COVID-19 pandemic as an exogenous shock. The
2016; Shleifer and Vishny, 1998). Existing studies usually use results show that political connections undermine the value of
static research settings to test the impact of political connections, enterprises. This study shows the cost that companies pay to
maintain politically connected, especially during crises.
∗ Corresponding author. The Chinese capital market provides a suitable environment
E-mail addresses: [email protected] (H. Liu), [email protected] to analyse this issue for several reasons. First, COVID-19 broke
(W. Zhao). out in Wuhan City. Second, political connections are common in

https://doi.org/10.1016/j.econlet.2023.110999
0165-1765/© 2023 Elsevier B.V. All rights reserved.
H. Liu and W. Zhao Economics Letters 224 (2023) 110999

Table 1 2021). We explore the mechanism of how political connections

Descriptive analysis. affect the value of firms during COVID-19, and enrich the research
Variable Mean SD Min P50 Max on the factors that influence enterprise value.
TobinQ 2.005 1.289 0.861 1.604 8.803
PC 0.265 0.441 0 0 1 2. Research design
Size 22.10 1.116 20.07 21.98 25.68
Lev 0.387 0.188 0.055 0.379 0.855
Growth 0.368 0.491 0.003 0.221 3.387 2.1. Sample selection and data sources
Separation 4.505 7.135 0 0.139 29.60
Large 30.77 13.41 8.730 29.26 68.76 This study considers the quarterly financial indicators of A-
Inv 0.349 1.501 0 0.0310 12.55
share listed enterprises from the second quarter of 2019 to the
third quarter of 2020 as the research object. The missing samples
Table 2 of ST enterprises, the financial sector, state-owned enterprises,
Main results. and related variables were excluded, resulting in a total of 12,689
(1) (2) sample observations. All continuous variables are winsorised by
PC*COVID −0.106*** −0.107*** the upper and lower 1%. Financial data were obtained from the
(−3.10) (−3.12) CSMAR. Political connections data were collected manually by us.
Size −0.402***
(−2.95)
Lev −0.271 2.2. Methodology
(−1.19)
Growth −0.025* Drawing on Garel and Petit-Romec (2021), this study used a
(−1.91) difference-in-differences model to test the effect of firm value.
Separation 0.011*
(1.91) TobinQi,t = α + β1 PC ∗ COVIDi,t + θ Controlsi,t + ξi,t (1)
Large −0.006
(−1.49) i denotes the firm and t denotes the quarter. We use TobinQ
Inv 0.006 to measure firm value. PC is a dummy variable that we de-
(0.95)
Constant 1.942*** 11.057***
fine as 1 if the chairman or general manager of the private
(145.80) (3.74) firm is a former or current government official, NPC deputy,
Firm FE YES party representative, or CPPCC member, and 0 otherwise. COVID
Quarter FE YES is a dummy variable defined as 1 for 2020 and 0 for previ-
Observations 12689 12689
ous years. Following the literature, the control variables include
adj_R2 0.101 0.114
enterprise size (Size), financial leverage (Lev), return on invest-
Note: ***, **, and * are significant at 1%, 5% and 10% levels ment (Inv), revenue growth (Growth), degree of separation of
respectively.
powers (Separation), and the shareholding of the first largest
shareholder (Large). We control for both individual fixed effects
and time-fixed effects.
China due to the imperfect legal system. The Chinese govern-
ment launched an unprecedented anti-corruption campaign in 3. Empirical results
2013, which suppressed political connections, making China an
interesting and useful research case. 3.1. Descriptive analysis
The main contributions of this paper are as follows: First,
on the controversial issue of how political connections affect Table 1 reports the descriptive statistics of the main variables.
corporate performance, we chose the COVID-19 outbreak as a During the sample period, 26.5% of private firms had political
quasi-natural experiment to provide new evidence to solve en- connections, indicating that political connections are common
dogeneity. Existing studies usually use static research settings to in China. The minimum value of firm value is 0.861 and the
test the impact of political connections, that is, to compare the maximum value is 8.803, indicating that the market value of
cross-sectional changes in economic outcomes between compa- private firms in China varies widely.
nies with and without political connections (Pan and Tian, 2020).
we use the COVID-19 outbreak to alleviate endogenous prob- 3.2. Empirical results
lems. This study thus contributes to the debate on the enterprise
value effects of political connections. Secondly, we discussed the The first two columns in Table 2 show the results of the
impact mechanism of political connections on enterprise value baseline regression. Column (1) controls time and individual fixed
during the COVID-19 period from two aspects of tax and employ- effects. The coefficient of PC * COVID is −0.106, which is signifi-
ment, which enriched the research on the politically related costs cantly negative at the level of 1%. Column (2) has added control
of enterprises. Most of the research on political connections of variables. The coefficient of PC * COVID is −0.107, which is sig-
enterprises focused on the benefits brought to enterprises by po- nificantly negative at the level of 1%. This shows that during the
litical connections, while there is less on the direct costs brought period of COVID-19, political connections damaged the enterprise
to enterprises by political connections. Our study is one of the few value by 8.3% (0.107/1.289).
that illustrates the social costs. Our research shows that political
connections hinder and damage the value of enterprises, high- 3.3. Robustness tests
lighting the distorting effect of political connections on the actual
operation of enterprises. Third, the study enriches the literature Table 3 reports the results of the robustness tests. To ensure
on the COVID-19 pandemic and its consequences on firms from the robustness of the study’s findings, the following tests were
the perspective of political connections. Existing studies have conducted in this study:
studied the impact of COVID-19 from the perspectives of financial First, a parallel trend test was performed. As shown in Fig. 1,
flexibility (Fahlenbrach et al., 2021), environmental responsibility before COVID-19, there is no significant difference between the
(Garel and Petit-Romec, 2021), and corporate culture (Li et al., experimental and control groups, which met the balance trend
2
H. Liu and W. Zhao Economics Letters 224 (2023) 110999

Table 3
Robustness tests.
(1) (2) (3) (4) (5) (6)
Placebo Wuhan Medicine Period PSM-DID Roa
PC*COVID −0.005 −0.111*** −0.103*** −0.123*** −0.106*** −0.004**
(−0.14) (−3.19) (−3.16) (−3.10) (−3.09) (−2.02)
Size −0.587*** −0.408*** −0.416*** −0.362*** −0.403*** −0.042***
(−2.99) (−2.99) (−3.03) (−2.66) (−2.94) (−12.10)
Lev −0.433 −0.273 −0.208 −0.188 −0.266 0.153***
(−1.51) (−1.19) (−0.88) (−0.75) (−1.16) (15.03)
Growth −0.037* −0.024* −0.028** −0.032** −0.024* 0.001
(−1.86) (−1.83) (−2.11) (−2.35) (−1.82) (0.72)
Separation 0.010 0.011* 0.004 0.005 0.011* 0.001**
(1.60) (1.87) (0.77) (0.79) (1.90) (2.38)
Large −0.000 −0.006 −0.006 −0.006 −0.006 −0.000
(−0.04) (−1.50) (−1.40) (−1.10) (−1.48) (−1.15)
Inv 0.008 0.006 0.005 0.007 0.005 0.001**
(1.16) (0.93) (0.89) (1.29) (0.77) (2.14)
Constant 15.640*** 11.188*** 11.332*** 10.137*** 11.064*** −0.005
(3.67) (3.78) (3.80) (3.45) (3.72) (−0.06)
Firm & QuarterFE YES
Observations 12689 12516 11750 10063 12626 12326
adj_R2 0.063 0.114 0.111 0.115 0.113 0.141

Note: ***, **, and * are significant at 1%, 5% and 10% levels respectively.

Fig. 1. Parallel trend test.

test. Second, we conducted a placebo test. We assumed that this paper uses the event study method for analysis. We choose
COVID-19 broke out in the third quarter of 2019. Third, we January 23, 2020, as the event date, which is the time when
excluded firms in Wuhan because the epidemic pattern in Wuhan Wuhan was closed. This paper uses the market model method
is different from that in other cities. Fourth, we excluded the to measure the cumulative excess return rate of stocks:
pharmaceutical industry because the industry most directly re-
lated to COVID-19 is significantly different from other industries. Ri,t = α0 + α1 RM ,t + ε (2)
Fifth, we excluded the first quarter of 2020 because it is the most
Ri,t represents the individual stock yield of the company i on
serious period of COVID-19, and the Chinese government took
the trading day t regarding the reinvestment of cash dividends,
strict prevention measures. Sixth, we did the PSM-DID test. We
and RM ,t represents the market yield on the trading day t. This
used the propensity score matching method to screen the control
paper takes the [−150, −30] window period before the event
group to reduce the problem of sample selection bias. Propensity
date as the estimation period, and uses model (2) to estimate the
scores are assigned by the Logit model based on tangible asset li-
risk-free rate of return of each company.
ability ratio (Ltv), management expense ratio (Agency), return on
assets (Roa), etc. The kernel matching method was used to match ARi,t = Ri,t − α
ˆ0 − αˆ1 RM ,t (3)
the propensity scores of the samples in the treatment group and
the control group, and the matched samples were regressed. Each In this paper, the cumulative excess return (Car) of the com-
covariate of the matched pre-treatment group and the control pany on three trading days before the event date and three
group met the balance test. Due to space limitations, the balance trading days after the event date, a total of seven trading days, is
test is not reported. Seventh, we used Roa to measure enterprise selected as a measure of the change in the enterprise stock price.
value. Eighth, to alleviate the potential endogenous problems, This paper establishes the following multiple regression model
3
H. Liu and W. Zhao Economics Letters 224 (2023) 110999

Table 4 rate and taxation) and maintaining social stability are two key
Event study. performance evaluation indicators that determine the promotion
CAR of politicians (Edin, 2003). These two indicators are easier to
PC −0.009** measure (Kerr, 1975), and more likely to get the attention of
(−2.06)
superiors. During the period of COVID-19, the unemployment
size −0.004*
(−1.88) rate remained high, and the local government was faced with
lev −0.012 huge financial pressure. The tax revenue and employment rate
(−0.84) also became the key indicators of official promotion evaluation
Profit 0.002* indicators during the crisis. Therefore, politicians may intervene
(1.76)
Em 0.000
in the operation of politically connected enterprises to make them
(0.78) bear the responsibility of expanding employment, actively paying
Pe −0.000* taxes and maintaining social stability. To verify this mechanism,
(−1.96) we use the corporate income tax rate to measure the degree of
Constant 0.090*
corporate tax payment and the number of employees divided
(1.68)
Ind & City YES by the total assets to measure the number of employees. The
N 324 empirical results are shown in Table 5. The coefficient of PC *
adj_R2 0.183 COVID is significantly positive, indicating that private enterprises
Note: ***, **, and * are significant at 1%, 5% and with political connections hire more employees and pay more
10% levels respectively. taxes.

Table 5 4. Research conclusions

Mechanism analysis.
(1) (2) The outbreak of COVID-19 caused severe damage to the cap-
Taxes Employees
ital market. We explored the impact of political connections
PC*COVID 0.498** 0.199*** on corporate behaviour during COVID-19. We found that po-
(2.09) (3.26)
Size −0.156 2.105***
litically connected enterprises paid more taxes and hired more
(−0.36) (18.75) employees, and suffered financial performance.
Lev −0.387 −2.222***
(−0.30) (−6.80) Funding
Growth −0.008 −0.023
(−0.06) (−0.69)
Separation 0.005 −0.042*** This research did not receive any specific grant from funding
(0.15) (−4.72) agencies in the public, commercial, or not-for-profit sectors.
Large 0.011 0.048***
(0.33) (5.64)
Inv 0.004 0.016 Declaration of competing interest
(0.09) (1.46)
Constant 3.524 −42.661*** The authors declare that they have no known competing finan-
(0.37) (−17.48)
cial interests or personal relationships that could have appeared
Firm FE YES
Quarter FE YES to influence the work reported in this paper.
Observations 12106 12106
adj_R2 0.001 0.055 Data availability
Note: ***, **, and * are significant at 1%, 5% and 10% levels
respectively. Data will be made available on request.

References
for analysis:
Amore, M.D., Bennedsen, M., 2013. The value of local political connections in a
Cari,t = α + β1 PCi,t + θ Controlsi,t + ξi,t (4) low-corruption environment. J. Financ. Econ. 110, 387–402.
Brogaard, J., Denes, M., Duchin, R., 2021. Political influence and the renegotiation
Control variables mainly include the Growth rate of total profit of government contracts. Rev. Financ. Stud. 34, 3095–3137.
(Profit), Enterprise value multiple (Em), PE ratio (Pe), etc. We Edin, M., 2003. State capacity and local agent control in China: CCP cadre
control both city and industrial effects. The empirical results management from a township perspective. China Q. 173, 35–52.
are shown in Table 4. There is a significant negative correlation Faccio, M., 2006. Politically connected firms. Am. Econ. Rev. 96, 369–386.
between political connections and cumulative excess return (t Fahlenbrach, R., Rageth, K., Stulz, R.M., 2021. How valuable is financial flexibility
statistic is 2.06, which is statistically significant at the 5% level). when revenue stops? Evidence from the COVID-19 crisis. Rev. Financ. Stud.
34, 5474–5521.
The empirical results show that during the period of COVID-19,
Garel, A., Petit-Romec, A., 2021. Investor rewards to environmental re-
political connections have damaged the value of enterprises. sponsibility: Evidence from the COVID-19 crisis. J. Corp. Finance 68,
101948.
3.4. Possible channels Hu, S., Zhang, Y., 2021. COVID-19 pandemic and firm performance: Cross-country
evidence. Int. Rev. Econ. Finance 74, 365–372.
Maintaining social stability and financial pressure is the key Jagolinzer, A.D., Larcker, D.F., Ormazabal, G., Taylor, D.J., 2020. Political
goal of the government during the COVID-19 period, and politi- connections and the informativeness of insider trades. J. Finance 75,
1833–1876.
cians are motivated to achieve the assessment goal by intervening
Kerr, S., 1975. On the folly of rewarding A, while hoping for B. Acad. Manag. J.
in politically connected enterprises. In China’s political system, 18, 769–783.
politicians are appointed rather than elected. Relative to social La Rocca, M., Fasano, F., Cappa, F., Neha, N., 2022. The relationship between
welfare indicators (such as improving education and providing political connections and firm performance: An empirical analysis in Europe.
health care), maintaining local economic growth (GDP growth Finance Res. Lett. 49, 103157.

4
H. Liu and W. Zhao Economics Letters 224 (2023) 110999

Li, K., Liu, X., Mai, F., Zhang, T., 2021. The role of corporate culture in bad Saeed, A., Belghitar, Y., Clark, E., 2016. Do political connections affect firm
times: Evidence from the COVID-19 pandemic. J. Financ. Quant. Anal. 56, performance? Evidence from a developing country. Emerg. Mark. Finance
2545–2583. Trade 52, 1876–1891.
Pan, X., Tian, G.G., 2020. Political connections and corporate investments: Shen, H., Fu, M., Pan, H., Yu, Z., Chen, Y., 2020. The impact of the COVID-
Evidence from the recent anti-corruption campaign in China. J. Bank. Financ. 19 pandemic on firm performance. Emerg. Mark. Finance Trade 56,
119, 105108. 2213–2230.
Ren, Z., Zhang, X., Zhang, Z., 2021. New evidence on COVID-19 and firm Shleifer, A., Vishny, R.W., 1998. The Grabbing Hand: Government Pathologies
performance. Econ. Anal. Policy 72, 213–225. and their Cures. Harvard University Press.