Politically Connected Private Equity and Employment

• Date: 01 April 2017
• Author: HUNG‐CHIA HSU,MARA FACCIO
• DOI: http://doi.org/10.1111/jofi.12483
• Published date: 01 April 2017

THE JOURNAL OF FINANCE •VOL. LXXII, NO. 2 •APRIL 2017

Politically Connected Private Equity

and Employment

MARA FACCIO and HUNG-CHIA HSU∗

ABSTRACT

We investigate the employment consequences of private equity buyouts. We find ev-

idence of higher job creation, on average, at the establishments operated by targets

of politically connected private equity firms than at those operated by targets of

nonconnected private equity firms. Consistent with an exchange of favors story, es-

tablishments operated by targets of politically connected private equity firms increase

employment more during election years and in states with high levels of corruption.

In additional analyses, we provide evidence of specific benefits experienced by tar-

get firms from their political connections. Our results are robust to tests designed to

mitigate selection concerns.

DURING THE 2012 PRESIDENTIAL ELECTION, the controversy around Mitt Rom-

ney’s participation in Bain Capital fuelled the debate concerning the role of

private equity firms. While lawmakers and labor unions often depicted pri-

vate equity firms as rent-seeking investors that make profits at the expense

of target firms’ workers, others offered a different perspective.1According to a

Wall Street Journal article, for example, the White House played a crucial role

during the election by encouraging the Carlyle Group, a politically connected

private equity firm, to take a majority stake in a Philadelphia oil refinery,

allegedly “saving 850 unionized jobs.”2Recent academic studies that reexam-

ine the effect private equity has on employment at target firms conclude that

∗Mara Faccio is with the Krannert School of Management, Purdue University. Hung-Chia Hsu

is with the WaltonCollege of Business, University of Arkansas. We thank two anonymous referees;

Kate Holland, Joel Houston, Byoung Hwang, Mitch Johnston, April Knill, Maria-TeresaMarchica,

Roberto Mura, Bill O’Brien, Michael Roberts (the Editor), Stefano Rossi, Jin Xu, Deniz Yavuz,

and Maxim Zagonov; seminar participants at Florida State University, Georgia State Univer-

sity, National Chengchi University, National Chiao-Tung University, National Taiwan University,

Tsinghua University, the University of Arkansas, the University of Memphis, and the University

of South Florida; and the Ph.D. students at Purdue University for comments. This research was

conducted while Hsu was visiting the Finance Area at Purdue University, whose hospitality and

support are gratefully acknowledged. The authors have no conflicts of interest to declare.

1Matt Taibbi, “Greed and Debt: The True Story of Mitt Romney and Bain Capital,” Rolling

Stone, August 29, 2012, http://www.Rollingstone.com/politics/news/greed-and-debt-the-true-story-

of-mitt-romney-and-bain-capital-20120829.

2Mark Maremont, “White House Workedwith Buyout Firm to Save Plant,” The Wall Street Jour-

nal, August 21, 2012, http://www.wsj.com/articles/SB1000087239639044371370457760328133059

7966.

DOI: 10.1111/jofi.12483

539

540 The Journal of Finance R

while private equity buyouts appear to have only modest impacts on employ-

ment (Davis et al. (2014)), they positively affect target firm workers’ long-term

career paths (Agrawal and Tambe (2015)). In this paper, we link the controversy

about political involvement in the private equity industry to the recent litera-

ture by investigating the relation between private equity ownership, political

connections, and employment.

To investigate our research question, we combine data from several sources.

We first identify 3,748 targets acquired by private equity buyout firms over

the 1990 to 2012 period from Capital IQ. Next, we employ the National Es-

tablishment Time-Series (NETS) database to track the employment, size, and

growth of 20,073 establishments (i.e., U.S. subsidiaries, branches, and plants)

operated by our sample targets, before and after the buyout. We then supple-

ment the NETS data with target and private equity firms’ data from Capital

IQ and Preqin. We also use Capital IQ to gather biographical information on

the individuals affiliated with the private equity firms. We use these data to

identify the political connections of each private equity firm. We classify a pri-

vate equity firm as politically connected if a general partner, board member,

or top employee had a major political position (e.g., served as a senator, mem-

ber of Congress, governor, U.S. president, etc.) or reported an affiliation3with

another person in one of those roles. Based on this definition, 14.14% of the

(1,690) private equity firms in our sample are politically connected, and 33.75%

(44.16%) of the target firms (establishments) in our sample are acquired by a

buyout syndicate/club that includes at least one politically connected private

equity firm.

Our establishment-level analysis shows that employment at existing estab-

lishments increases by 1.24% per year, on average, following a buyout by a

politically connected private equity firm. By contrast, employment at existing

establishments on average increases by a more modest 0.33% per year, on av-

erage, following a buyout by a nonconnected private equity firm. Thus, at least

on average, buyouts by politically connected private equity firms are associated

with higher job creation (or less job destruction) at the existing establishments

operated by their target firms.

A caveat in the interpretation of these results, of course, relates to endogene-

ity. One specific concern is that buyout decisions are not random, as employ-

ment dynamics may vary across establishments and targets for reasons that

are unrelated to the political connections of their private equity acquirers. For

example, non–politically connected private equity firms could invest in targets

that are growing at slower rates than the targets of politically connected pri-

vate equity firms. To control for time-invariant establishment-specific charac-

teristics that may be correlated with omitted explanatory variables, we exploit

the panel dimension of our sample and employ fixed effects specifications. In

these specifications, we compare the employment dynamics of a given estab-

lishment before and after the private equity buyout. We further augment these

3Political affiliations represent instances in which the person in question worked for a politician

or a government agency, as opposed to holding a political office herself.

Politically Connected Private Equity and Employment 541

specifications with controls to reflect establishment and target firm attributes

that might change around the time of the buyout (e.g., establishment-level

size and growth, target firm leverage, tangibility, and profitability). While we

recognize that we cannot control for all factors that might change around the

time of the buyout, this test attempts to mitigate the sources of confounding

variation.

A second endogeneity concern is that the assignment to treatment and con-

trol groups is itself not random. Politically connected private equity firms differ

from their nonconnected peers on a number of dimensions, and those differ-

ences might explain the differences in the labor dynamics observed. For ex-

ample, the documented differences in the labor strategies of target firms may

reflect differences in local market conditions, buyout type, and/or the size of the

private equity firms rather than the political connections of the private equity

acquirers. To mitigate this concern, we employ a propensity score matching

analysis in which we compare the employment dynamics at establishments op-

erated by targets backed by politically connected private equity firms to those

of establishments operated by similar peers backed by nonconnected private

equity firms. Peers are selected from (i) the same buyout year and state if such

a match exists, or (ii) the same buyout type, and then are further matched on

observable establishment-level characteristics and private equity firm size. We

find that establishments operated by targets of politically connected private eq-

uity firms experience significantly higher employment growth after the buyout

compared with establishments operated by otherwise similar targets. While

we again acknowledge that the assignment to treatment and control groups is

not random, these results mitigate the scope for alternative interpretations.

To further address selection concerns, we use placebo tests to compare real-

ized buyouts with failed buyouts, before and after the buyout attempt. If the

aforementioned results were due solely to selection, then the employment dy-

namics observed among targets of successful takeovers should also be observed

among targets of failed takeovers. Because the decision to withdraw a takeover

is itself endogenous, to understand the nature of these failed transactions and

reduce remaining endogeneity concerns we focus on two specific subsamples

of failed takeovers. The first subsample consists of transactions that presum-

ably would have had a positive net present value (NPV) for the target had the

deal gone through. The second subsample comprises transactions that were

cancelled by the target firm or by a third party. These transactions presum-

ably would have had a positive NPV for the private equity bidder had the

transaction gone through. We do not find evidence of systematic job creation

in either subsample of failed transactions subsequent to the buyout announce-

ment; rather, the opposite is true. In addition, the difference-in-difference-in-

differences (DDD) results provide evidence consistent with the idea that es-

tablishments operated by targets of politically connected private equity firms

increase employment significantly more following successful buyouts.

This evidence begs the question of why politically connected private equity

firms boost employment at their targets. Perhaps some private equity firms

boost employment to maintain good public relations and/or good relations with