Local official turnover, ownership, and firm cash holdings: Insights from an emerging market

• Date: 01 September 2019
• Published date: 01 September 2019
• DOI: http://doi.org/10.1111/eufm.12194

DOI: 10.1111/eufm.12194

ORIGINAL ARTICLE

Local official turnover, ownership, and firm cash

holdings: Insights from an emerging market

Xiaoran Ni

Department of Finance, School of

Economics and Wang Yanan Institute for

Studies in Economics (WISE), Xiamen

University, 361005, China

Email: nxr@xmu.edu.cn

Abstract

Using a hand-collected dataset of city-level local official

turnover in China, I find that average cash holdings of listed

firms decrease significantly upon turnover of city heads, and

this effect concentrates in privately owned enterprises. Such

effects are more pronounced in firms located in cities with

lower government quality. I also find that local official

turnover leads to decreases in equity issuance for privately

owned enterprises but not for state-owned enterprises,

which largely explains our primary findings. Overall, this

paper reveals that the cash policy of privately owned

enterprises is sensitive to local official turnover in an

emerging market.

KEYWORDS

cash holdings, emerging market, local official turnover, ownership

structure

JEL CLASSIFICATION

G18, G32, G38

I thank John Doukas (the Editor), two anonymous referees, Chong’en Bai, Weixing Cai, Hui Dong, Xiaosheng Ju, Yue Li,

Shangkun Liang, Ajid Ur Rehman, Guoman She, Sandy Suardi, K. C. John Wei, Sirui Yin, Xiaofeng Zhao, and participants

at the 2016 Frontiers of Business Research in China International Symposium (Beijing), 2016 Chinese Young Economists

Conference (Beijing), 2016 China Finance Review International Conference (Shanghai), 2016 Chinese Finance Annual

Meeting (Dalian), 2017 European Financial Management Symposium (Xiamen), 2017 China Accounting and Finance

Conference (Beijing), and seminar participants at Guangdong University of Finance and Economics for their helpful

comments. I especially thank David Ji for his efforts on earlier versions of the paper. All errors are my own.

Eur Financ Manag. 2018;1–34. wileyonlinelibrary.com/journal/eufm © 2018 John Wiley & Sons, Ltd.

|

1

1013Eur Financ Manag. 2019;25:1013–1046. wileyonlinelibrary.com/journal/eufm © 2018 John Wiley & Sons, Ltd.

1

|

INTRODUCTION

Cash is of vital importance to firms. As a method to ease financial frictions, cash holdings determine

whether businesses can take advantage of investment opportunities. However, cash is vulnerable to

expropriation by entrenched managers and local officials, especially in economies with powerful

governments and less developed legal systems (Caprio, Faccio, & McConnell, 2013; Chen, Li, Xiao, &

Zou, 2014; Feng & Johansson, 2014; Stulz, 2005). In such economies, turnover of government officials

can significantly affect government policies (Mikesell, 1978; Rogoff & Sibert, 1988). Although a vast

number of studies investigate the economic impact of turnover of government officials, they focus

primarily on the uncertainty of election outcomes (e.g., Atanassov, Julio, & Leng, 2015; Julio & Yook,

2012). It remains largely unexplored whether government official turnover can affect firm behavior in

ways other than creating uncertainty prior to elections. To fill this gap, I utilize a hand-collected dataset

from China and investigate the effect of local official turnover on firm cash holdings.

I take advantage of frequent city-level local official turnover in China, where a higher-level

authority changes local leaders partially based on their economic performance (Cull, Xu, Yang, Zhou,

& Zhu, 2017; Xu, 2011). At the city level, many officials serve for very short periods of time.

1

The

inherent short-term horizon of this multilayer hierarchy motivates newly appointed officials to make a

rapid impact to be selected for promotion (Gao, Long, & Xu, 2016). Since it takes time to transform

investment projects into gross domestic product (GDP) growth, such officials tend to initiate image

projects immediately after assuming office in the radical manner of ‘new brooms sweep clean’to

increase their likelihood of promotion (Cao, Julio, Leng, & Zhou, 2015; Du, Lu, & Tao, 2015).

2

Theoretically, the effect of local official turnover on firm cash holdings is ambiguous. It is possible

that turnover of local officials may reduce cash holdings. I consider two views regarding the role of

local governments: the grabbing hand view and the helping hand view. The former view conjectures

that newly appointed officials with promotion incentives may expropriate firms. Local officials may

enlist firms, especially those with a significant amount of cash, to provide funding for public projects,

which can be unprofitable for the firms but helpful to local governments (Du et al., 2015).

3

After the

original official leaves, firms lose their established political connections and become vulnerable to

expropriation (Feng & Johansson, 2014). Therefore, they take action to minimize such risk when the

cost of doing so is relatively small (Li, Maydew, Willis, & Xu, 2014). Not only can actual political

expropriation have an effect, but the prospect of political expropriation can also cause firms to

strategically shelter their cash (Caprio et al., 2013; Smith, 2016). Therefore, the grabbing hand view

predicts that newly appointed officials may decrease firm cash holdings.

The helping hand view suggests that newly appointed officials with promotion incentives may

support firms. Local governments can assist firms in various ways, such as providing subsidies (Lim,

Wang, & Zeng, 2018) and information (Cull et al., 2017), helping them secure loans from debtors

(Faccio, 2010) and implementing bailouts (Faccio, Masulis, & McConnell, 2006). If firms can

successfully develop connections with new officials, they can obtain subsidies and financial support

1

At the city level, although an official's nominal term is typically 5 years, many local officials leave for other positions

within a term, and some serve for more than one term. During our sample period, 29.5% of officials serve for their position

for only 1 or 2 years; 69.9% of them serve their position for less than 5 years. Sources: http://www.economist.com/news/

china/21610316-weighing-economic-impact-anti-corruption-campaign-anti-graft-anti-growth

2

The Chinese equivalent for ‘new brooms sweep clean’is ‘a new official usually does something unconventional and

sensational.’

3

See Shleifer and Vishny (1994), Fisman (2001), Svensson (2003), Khwaja and Mian (2005), Stulz (2005), Claessens,

Feijen, and Laeven (2008) for more discussion on this issue.

2

|

NI

f

Q

J

t

r

d

C

r

t

r

h

e

m

F

g

2

I

s

a

h

s

t

t

I

m

e

c

T

t

p

e

g

t

r

r

m

i

4

t

s

d

N

N

2

|

NI

1014

e

o

l

&

s

t

s

,

n

t

l

,

e

a

m

e

o

e

f

s

y

,

e

o

e

l

o

w

y

,

s

n

t

n

I

from the government. Consequently, firms face weaker financial constraints and hold less cash (Han &

Qiu, 2007; Opler, Pinkowitz, Stulz, & Williamson, 1999).

Conversely, local official turnover can mechanically increase cash holdings. In agreement with

Julio and Yook (2012), policy uncertainty associated with turnover of government leaders can enhance

the value of waiting. As a result, firms reduce investment expenditure until policy uncertainty is

resolved. Therefore, cash holdings increase.

To test these predictions, I collect the personal information of city heads in China.

4

I then match this

dataset with information on listed non-financial firms for the period 2002–2013. It is noteworthy that in

China, local officials are appointed by higher-level authorities (party committees or governments)

rather than elected competitively. These officials can be promoted or replaced at any time. Therefore,

turnover of local officials in China is largely immune from lobbying.

I begin by examining the effects of local official turnover on cash holdings. On average, firms

reduce their cash holdings by 3.1% when local official turnover occurs. I then investigate

heterogeneous effects between state-owned enterprises (henceforth SOEs) and privately owned

enterprises (henceforth POEs). SOEs are less concerned with political expropriation risk, because their

managers are semi-officials in the bureaucratic system (Chen, Jiang, Ljungqvist, Lu, & Zhou, 2016;

Feng & Johansson, 2014; Li & Zhou, 2005). In contrast, POEs are generally less connected with local

governments, more financially constrained (Allen, Qian, & Qian, 2005; Feng, Johansson, & Zhang,

2015), and more likely to become targets of rent seeking (Shleifer & Vishny, 1994). Therefore,

I conjecture that either the grabbing hand or the helping hand effects of local governments, if any,

should be more pronounced in POEs. Indeed, I find that the negative effects concentrate in POEs; the

average cash holdings of POEs decrease by 6.5% upon turnover of city heads, while the average cash

holdings of SOEs do not change significantly.

My research design assumes that local official turnover triggers changes in cash holdings. I take

steps to validate this assumption. To begin with, I conduct a dynamic test around the time when

turnover of city heads occurs. I find that cash holdings begin to decrease only after local official

turnover, but not before. Since the results do not provide sufficient evidence of a causal relation,

I conduct several additional tests. To address the concern that firms experiencing local official turnover

may be substantially different from other firms, I construct a propensity-score-matched sample to

examine whether the primary findings remain valid. Additionally, omitted variables may lead to

changes in both local official turnover and cash holdings, which could bias the estimation results.

Therefore, I employ an instrumental variable (IV) approach. I identify two IVs for local official

turnover, the city head's age in the preceding year and the percentage of other cities in the same

province that experience turnover of a city head in the current year. A two-stage least square (2SLS)

estimation is conducted using these two instruments. In addition, I employ the dynamic panel system

general method of moment (GMM) estimation to check the causal relationship between local official

turnover and cash holdings. Both the 2SLS and GMM estimations yield a negative and significant

relation between local official turnover and POE cash holdings. In addition, the primary results are

robust to multiple econometric specifications, various subsamples, and alternative definitions of the

main variables.

Next, I explore how the effect revealed above differs among subsamples. Using city-level

indicators of government quality provided by the World Bank Survey (Chen et al., 2014; Mako, 2006),

4

City head refers to the secretary of the prefectural Chinese Communist Party Committee. As party chief, the secretary is

the most powerful political leader in the city, outranking the mayor. The ‘secretary’here is not the counterpart of the

secretary in other countries who do office work such as answering phone calls and arranging meetings. See section 2 for a

detailed description.

NI

|

3NI

|

3

I

1015