Income Smoothing in Family‐Controlled Companies: Evidence from Italy

• Date: 01 November 2011
• Author: Annalisa Prencipe,Pietro Mazzola,Sasson Bar‐Yosef,Lorenzo Pozza
• Published date: 01 November 2011
• DOI: http://doi.org/10.1111/j.1467-8683.2011.00856.x

Income Smoothing in Family-Controlled

Companies: Evidence from Italy

Annalisa Prencipe,* Sasson Bar-Yosef, Pietro Mazzola, and

Lorenzo Pozza

ABSTRACT

Manuscript Type: Empirical

Research Question/Issue: This paperfocuses on the relationship between one of the main corporate governance dimensions

– ownership structure – and income smoothing. The paper investigates whether family-controlled companies differ from

non-family-controlledcompanies with respect to income smoothing. Due to different incentives of management and owner

investment horizons, we hypothesize that income smoothing is less likely among family-controlled companies than among

non-family-controlled companies. Additionally, we hypothesize that among family-controlled f‌irms income smoothing is

less likely when CEO and BoardChairman are members of the controlling family.Various def‌initions of “familycontrol” are

applied. A sample of Italian listed companies is used for the empirical analysis.

Research Findings/Insights: We f‌ind evidence that income smoothing is less likely among family-controlled companies

than non-family-controlled companies. Moreover, among family-controlled companies, income smoothing is less likely for

f‌irms whose CEO and Board Chairman are members of the controlling family.

Theoretical/Academic Implications: This paper f‌ills a gap in the literature, suggesting that not only the level of ownership

concentration or insider ownership but also the nature of the dominant shareholder (family versus non-family) should be

considered when addressing the motivations for income smoothing. Furthermore, our f‌indings indicate that agency theory

and stewardship theory are complementary in explaining the role played by family control in income smoothing decisions.

While in non-family-controlled companies the traditional owner-manager agency problems tend to prevail and motivate

income smoothing, in family-controlled companies such agency issues become less relevant and a stewardship attitude

emerges, rendering income smoothing less likely.

Practitioner/Policy Implications: This study is of interest to f‌inancial statement users, including analysts and investors, as

it shows that different company types (e.g., family versus non-family) have a different attitude towards income smoothing.

In particular, these results aid users in interpreting the company’s reported prof‌itability and its potential variability. The

conclusions also are of interest to auditors when evaluating the reliability of the reported income of companies characterized

by various ownership structures.

Keywords: Corporate Governance, Income Smoothing, Family Companies, Ownership Structure, Italian Companies

INTRODUCTION

It is well known thatmanagers tend to smooth income – i.e.,

to reduce the variabilityof the earnings reported in public

f‌inancial statements– in order to gain personal or contractual

benef‌its.1This paper focuses on the issue of whether one of

the main elements of corporate governance – ownership

structure – is associated with income smoothing. In parti-

cular, the paper tests whether income smoothing is less likely

among companies whose ownership is classif‌ied as family-

controlled than among non-family-controlled companies.

Current literature on the issue of income smoothing pro-

vides evidence of the fact that listed companies do indeed

smooth income (e.g., Buckmaster, 2001; Chaney, Jeter, &

Lewis, 1998; Moses, 1987; Ronen & Sadan, 1981; Young,

1998). Income smoothing aims at moderating year-to-year

income f‌luctuations by shifting earnings from peak years to

the less successful periods, thereby reducing income volatil-

ity (Copeland, 1968). Income smoothing can be achieved

through different methods, e.g., through real transactions or

through the adoption of accounting policies that reduce

income-stream variability (Fudenberg & Tirole, 1995).

*Address for correspondence: AnnalisaPrencipe, Department of Accounting, Bocconi

University, Via Roentgen 1, 20136 Milan, Italy. Tel: +39 02 58362574; Fax: +39 02

58362561; E-mail: annalisa.prencipe@unibocconi.it

529

Corporate Governance: An International Review, 2011, 19(6): 529–546

© 2011 Blackwell Publishing Ltd

doi:10.1111/j.1467-8683.2011.00856.x

While there exists strong empirical evidence to support

income-smoothing behavior among widely-held companies

(e.g., Buckmaster, 2001; Chaney et al., 1998; Moses, 1987;

Ronen & Sadan, 1981; Young, 1998), the association between

ownership structure and income smoothing practices has

been overlooked. This paper expands on prior research in

that it focuses on whether listed family-controlled compa-

nies differ from non-family-controlled companies with

respect to income smoothing.

It is hypothesized that income smoothing is less likely to

occur among family-controlled companies than non-family-

controlled companies. This hypothesis is developed on the

premise that, based on both agency and stewardship theo-

ries, there exist profound differences between family- and

non-family-controlled f‌irms, as documented in the litera-

ture. In particular, the focus is placed on the different types

of relationship between executives and shareholders

(Brunello, Graziano, & Parigi, 2003; Crisci & Tarizzo, 1995;

Miller & Le Breton-Miller, 2006; Volpin, 2002) and on the

different owners’ investment horizons (Anderson, Mansi, &

Reeb, 2003; Miller & Le Breton-Miller, 2006).

The empirical analysis is based on a sample of Italian

listed companies. The results support the hypothesis that

family-controlledcompanies are associated with less income

smoothing than non-family-controlled companies. The

results are robust to various def‌initions of family-control. In

addition, the results indicate that among family-controlled

companies income smoothing is less likelyfor f‌irms in which

both the CEO and the Board Chairman positions are occu-

pied by members of the dominant family.

Our f‌indings suggest that current knowledge related to

the income-smoothing phenomenon is incomplete, and that

one needs to draw a distinction between characteristics of

corporate ownership that manifest different management

attitudes towards income smoothing. These results are

important because a signif‌icant number of companies

around the globe are classif‌ied as family-controlled (Burkart,

Panunzi, & Shleifer, 2003; La Porta, Lopez-de-Silanes, &

Shleifer, 1999); therefore, understanding income smoothing

in such companies might lead to a better understanding of

the economics of such companies. The main research impli-

cation of the current study is that ownership structure and

ownership characteristics need to be considered when

addressing the impacts, effects, and implications of compa-

nies’ income-smoothing practices.

The paper is structured as follows. The next section

reviews prior studies on ownership structure and income

smoothing and develops our hypotheses. In the third

section, a description of the sample and the research design

is provided. The empirical results are presented and dis-

cussed in the fourth section, followed by conclusions.

INCOME SMOOTHING, OWNERSHIP

STRUCTURE AND HYPOTHESES

DEVELOPMENT

Prior studies have posited several motivations for income

smoothing, some of which are related to capital-market

incentives. In particular, some studies propose that since

current earnings are used as predictors of future income,

managers have the incentive to smooth income as a signal-

ling device for communicating private information to the

market (e.g., Chaney & Lewis, 1995; Hunt, Moyer, & Shevlin,

1996; Ronen & Sadan, 1981). Other studies suggest that the

primary motivation for income smoothing is to reduce per-

ceived risk by investors and, as a consequence, reduce the

corporate cost of capital (e.g., Barth, Landsman, & Wahlen,

1995; Gebhardt, Lee, & Swaminathan, 2001; Wang & Will-

iams, 1994). Yet other explanations for income smoothing

have been related to accounting numbers-based manage-

ment compensation contracts (e.g., Lambert, 1984; Trueman

& Titman, 1988). Political costs are also considered to be an

important motivation for income smoothing (Cahan, 1992;

Godfrey & Jones, 1999; Watts & Zimmerman, 1986; Wong,

1988). In fact, income smoothing diverts attention from

“excessively” high or low income that might attract adverse

political attention. Also, managers’ compensation and con-

cerns over job security have been considered as another

motivation for income smoothing. For example, Fudenberg

and Tirole (1995) predict that income smoothing occurs

because managers boost reported income in bad times in

order to raise the probability of keeping their jobs and

decrease reported income in good times since “information

decay” causesgood current performance to be weighted less

in future performance. Similarly, Amihud and Lev (1981)

suggest that risk-averse managers tend to engage in transac-

tions such as conglomerate mergers in order to stabilize the

f‌irm’s income stream and reduce “employment risk.”

Most prior income-smoothing studies focus on widely-

held public companies, with empirical evidence that largely

conf‌irms that managers of such companies smooth income

for market or contractual motivations. However, income-

smoothing incentives in ownership settings other than those

that are widely-held have so far received scant attention.

Indeed, the relationship between ownership structure and

income smoothing has been addressed in the past by only a

few studies, producing non-univocal results. Among these,

Smith (1976) distinguishes “manager f‌irms” from “owner

f‌irms” based on ownership concentration (using a threshold

of 5 per cent to 10 per cent of the voting stocks for the

classif‌ication),2showing that the proportion of smoothing

f‌irms to total f‌irms is larger for the manager group than it is

for the owner group. The author argues that the results are

the consequence of the fact that owners of relatively large

shares of stock typically have access to more complete infor-

mation and constrain any income manipulation that manag-

ers might attempt, including income smoothing. In a similar

vein, Carlson and Bathala (1997) show that higher insider

ownership (i.e., above 10 per cent) decreases the probability

of income smoothing. They suggest that when managers are

also owners of the company, they are less likely to use

income smoothing as a job-preserving strategy. Also, as the

ownership of insiders increases, companies tend to fall into

the “owner f‌irms” category; therefore, as suggested by

Smith (1976), income smoothing tends to decrease. Different

results were reported by Moses (1987), however, who f‌inds

no signif‌icant relationship between insider ownership and

income-smoothing behavior. Similarly, in a more recent

study, Che-Ahmad and Mansor (2009) f‌ind no signif‌icant

relationship between management shareholdings and

income smoothing among Malaysian-listed f‌irms.

530 CORPORATE GOVERNANCE

Volume 19 Number 6 November 2011 © 2011 Blackwell Publishing Ltd