Why do firms heap dividends? New evidence on the key role of firm characteristics

• Published date: 01 July 2023
• Author: German Rubio,Keith Jakob,Augusto Castillo,Jorge Niño
• Date: 01 July 2023
• DOI: http://doi.org/10.1002/jcaf.22612

Received:  October Accepted:  December 

DOI: ./jcaf.

RESEARCH ARTICLE

Why do firms heap dividends? New evidence on the key role

of firm characteristics

German Rubio1Keith Jakob2Augusto Castillo3Jorge Niño

Facultad de Administracióny Economía,

Universidad Diego Portales, Santiago,

Chile

College of Business, University of

Montana, Missoula, Montana, USA

Escuela de Negocios, Universidad Adolfo

Ibañez, Santiago, Chile

Correspondence

Keith Jakob, Department of Accounting

and Finance, University of Montana, 

Campus Drive, Missoula, MT , USA.

Email: keith.jakob@business.umt.edu

Abstract

In this paper we explore how a cognitive bias known as heaping (or rounding to

a higher level) influences the dividend policy of companies. Recent articles have

shown that dividend size and the level of several information uncertainty vari-

ables help to explain changes in the likelihood of rounding dividends. Using data

from the US that covers the – period, we verify that the previous litera-

ture overlooked the key role played by a groupof variables denominated as firm

characteristics. These variables have alreadyproven to have a role in the decision

of paying dividends or not. We also show that other attributes,such as being reg-

ular payers of dividends, impact the likelihoodof rounding dividends. Finally, we

verify that the propensity to heap dividends has been changing over time, even

after controlling for changes in the dividend size, in the level of the informa-

tion uncertainty variables and in the status of the firm characteristic variables,

opening space to further research.

KEYWORDS

cognitive bias, dividends, firm characteristics, heaping

1 INTRODUCTION

Companies have been distributing dividends to stock-

holders for at least four centuries (Baskin, ). The

managerial decisions revolving around paying out div-

idends are important for both the corporation and its

investors. The selected magnitude of the dividend distri-

bution influences several relevant corporate metrics such

as the firm’s capital structure, the earnings reinvestment

rate or plowback, the firm’s future growth potential, the

dividend yield, and the overall taxation of investors. Most

popular stock valuation techniques also rely heavily on

dividends as a key input into the analyses. There is a

large body of empirical and theoretical academic literature

that investigates the approach and intentions of corpo-

rate dividend payments; however, the intricacies of the

firm’s dividend issuance decision still deserve additional

consideration.

Turner () argues that people underuse those digits

that are not multiples of the divisor of the base of the num-

ber system while overusing or heaping at those digits that

are multiples of the divisors of the base of the number sys-

tem. According to Turner, people estimate or round to a

heaped measure when they are unsure of the correct value,

but they use more accurate numerical values when they

can. The most common heaped values are the largest divi-

sors of the base ten numbering system which are ten, five

and two.

There is a recent and growing line of academic research

that examines the heaping heuristic within a corporate

accounting and finance framework (Bamber et al., ;

Dechow & You,; Herrmann & Thomas, ). Jakob &

Nam () study the distribution of cash dividends in the

US and report that % of dividends are rounded or clus-

tered at specific intervals consistent with heaping in a base

ten numbering system. They report that the probability of

80 ©  Wiley Periodicals LLC.J Corp Account Finance. ;:–.wileyonlinelibrary.com/journal/jcaf

RUBIO   .81

rounding to a heaped value is significantly and positively

related to the size of the dividend distributed. They also

report that heaping is significantly related to several mea-

sures of the level of information uncertainty faced by the

management of the company. Nam et al. ()examine

the rounding of dividends phenomenon in Australia and

concur that the probability of heaping is related to infor-

mation uncertainty.Castillo et al. () examine dividend

payout policy in four Latin American countries with sig-

nificantly different currency magnitudes. They verify that

the magnitude or relative strength of the country-specific

currency significantly influences the likelihood and char-

acteristics of heaping observed in the dividends for each

country. Finally, Jakob et al. () study dividend pay-

ments from four key European markets(Germany, France,

the UK, and Switzerland). They report that the changes of

currency in Germany and France (they both adopted the

Euro in January ) significantly influenced the likeli-

hood of heaping observed in dividends paid both in the

short and in the long run.

In this study, we extend the academic literature that

examines how cognitive biases influence corporate deci-

sions regarding dividend allotment. In particular, the

objective of the paper is to expand and improve the

methodology used in previous studies to analyze the

motives or reasons why managers heap dividends. We

accomplish this goal by incorporating new information

uncertainty variables into the analysis as well as by also

adding financial characteristics of the firm as additional

explanatory variables.

We deepen the analysis further by distinguishing

between companies that are regular payers of dividends,

and those who are not regular payers. We observe that

regular payers of dividends have a lower likelihood of sys-

tematically rounding dividends, and we also verify that the

difference in likelihood of heaping between regular pay-

ers and non-regular payers has been increasing over the

last three decades. Finally, we verify that, for three of the

four intervals considered, the propensity to heap dividends

has been decreasing over time, even after controlling for

changes in the dividend size, in the level of the informa-

tion uncertainty variables and in the status of the firm

characteristic variables.

2LITERATURE REVIEW

The vast majority of the ensuing dividend policy liter-

ature can be classified into categories of either rational

or behavioral dividend payment theories. In the area

of rational theories, there are recognized studies such

as: Bhattacharya (), Easterbrook (), Karpavičius

(), Miller and Modigliani (), etc.

Several behaviorally based dividend payment theories

attempt to explain observed corporate dividend policy

based on managerial and investor biases. Long ()pro-

vides evidence that investors’ interest in or demand for

dividends varies across time. Baker and Wurgler ()

postulate that the managerial dividend payment decision

is driven by prevailing investor sentiment regarding div-

idend payers. They proclaim that the dividend choice is

used by managers to attract investors by paying dividends

when investors prefer payers, and by not paying dividends

when investors put a stock price premium on nonpayers.

Several additional articles look at how certain psycho-

logical biases of investors can influence dividend payout

policy. The bird-in-hand argument (e.g., Gordon, ;

Lintner, ) proposes that since investors must real-

ize wealth to be able to consume, they, therefore, prefer

cash dividends over capital gains. However, this bird-in-

hand argument is theoretically contested by Miller and

Modigliani (). Black (), and Shefrin and Statman

() suggest that dividends are used as primarily an

investor self-control mechanism. In the absence of divi-

dends, shareholders are forced to liquidateor sell shares for

the purpose of consumption. In this psychologically based

explanation, dividends are used as a tool to help investors

regulate and control their level of consumption.

In another line of behavioral research, studies of cor-

porate managers associate managerial over-optimism or

overconfidence with both dividend policy and investment

policy decisions (e.g., Deshmukh et al., ;Malmendier

&Tate,). This research indicates that managerial

overconfidence impacts the overall level of dividend distri-

butions as well as the market response to announcements

regarding changes in dividend payments. In our paper,

we study how heaping as another managerial behavioral

bias impacts the corporate dividend payment decision. We

argue that firm managers will be affected by this often-

observed behavioral heuristic and its use by managers

will directly influence their choice of corporate dividend

payouts.

A well-documented bias or human error is to round

numbers even though precise results are preferred. This

numerical rounding heuristic, known as heaping, is par-

ticularly noticeable and documented in census or demo-

graphic surveys where the age in years is sought (e.g.,

Myers, ). The primary heaping literature finds obser-

vations clustered at intervalsof ten, f ive, and twowhich are

the principal divisors of the base ten numbering system.

As an extension to the survey data rounding literature,

several researchers ask whether this particular kind of

cognitive bias or other similar heuristics systematically

influence the actors within the financial markets as well as

firm management and their corporate decision making. An

extensive line of literatureexamines rounding or clustering