BEAUTY PREMIUM AND MARRIAGE PREMIUM IN SEARCH EQUILIBRIUM: THEORY AND EMPIRICAL TEST

• Author: Francis Kiraly,John Wildman,Roberto Bonilla
• Date: 01 May 2019
• DOI: http://doi.org/10.1111/iere.12370
• Published date: 01 May 2019

INTERNATIONAL ECONOMIC REVIEW

Vol. 60, No. 2, May 2019 DOI: 10.1111/iere.12370

BEAUTY PREMIUM AND MARRIAGE PREMIUM IN SEARCH EQUILIBRIUM:

THEORY AND EMPIRICAL TEST∗

BYROBERTO BONILLA,FRANCIS KIRALY,AND JOHN WILDMAN1

Newcastle University, U.K.; Newcastle University, U.K.; Newcastle University, U.K.

We propose a theoretical explanation for the so-called beauty premium. Our approach relies entirely on

search frictions and the fact that physical appearance plays an important role in attracting a marriage partner. We

analyze the interaction between frictional labor and marriage markets, making use of what we label constrained

job search. The optimal strategy entails different reservation wages for different men, and we establish the

existence of a search equilibrium characterized by a beauty premium and/or marriage premium. Predicted

profiles of premia allow for potential falsification tests and point to relevant empirical evidence in support of

our theory.

1. INTRODUCTION

There is widespread evidence that labor market outcomes are influenced by more than just

productivity. Anthropometric characteristics such as beauty, height, and—to some extent—

weight also appear to have an effect on employment and wages. In particular, individuals

perceived as having attractive physical attributes tend to earn higher wages. In the literature,

this earnings gap is referred to as the “beauty premium,” and has been the subject of extensive

empirical research.

We offer an explanation for the existence of the male beauty premium as an equilibrium

outcome. Our theoretical model incorporates two crucial features of labor and marriage mar-

kets. First, both markets are characterized by search frictions: It takes time, effort, and luck

to find a suitable job or marital partner. Second, matching in the marriage market seems to

involve multidimensional preferences that reflect an implicit trade-off between anthropometric

and socioeconomic characteristics.

The key message of our article is that labor market decisions and outcomes (including various

types of wage premia) may be influenced by expectations and behavior in the marriage market,

and vice versa.

To capture this interdependence, we construct a simple equilibrium search model where the

two frictional markets are interlinked.2Single men are heterogeneous in terms of their physical

appearance: In the eyes of all women, some men are more attractive than others. We consider

a two-sided search scenario, where men and women look for each other, and unemployed men

search for jobs knowing that earnings, together with physical attributes, determine whether or

not they can form marriage partnerships.

∗Manuscript received January 2014; revised September 2017.

1We are especially grateful to the Editor, Guido Menzio, for many valuable suggestions. We also thank two anony-

mous referees for their very helpful comments. Melvyn Coles, Peter Dolton, Daniel Hamermesh, Adrian Masters,

Alberto Trejos, and Mich Tvede, as well as participants at the 2014 EALE Conference (Ljubljana) and the 2015 Search

and Matching (SaM) Annual Meeting (Aix-en-Provence), also contributed with many useful remarks. In addition, R.

Bonilla would like to thank CESifo and Universidad de Oviedo for logistical support. Please address correspondence to:

Roberto Bonilla, Newcastle University Business School, Newcastle University, 5 Barrack Road, Newcastle-upon-Tyne,

Tyne ad Wear NE1 4SE, U.K. Phone: +44 7507064459. E-mail: roberto.bonilla@ncl.ac.uk.

2The literature on interlinked frictional markets is sparse. For two very interesting recent papers, see Kaplan and

Menzio (2016) and Rupert and Wasmer (2012).

851

C

(2018) by the Economics Department of the University of Pennsylvania and the Osaka University Institute of Social

and Economic Research Association

852 BONILLA,KIRALY,AND WILDMAN

In our model, physical appearance does not affect men’s options in the labor market. Crucially,

however, it does affect their job search decision in that market, as their marriage prospects are

influenced by both their looks and wages. We call this type of decision problem constrained

search, and analyze it in detail.

If women regard physical characteristics and wages as substitutes and rank men in the same

way, we show that there exists an equilibrium in which less attractive men find it optimal to

accept jobs that pay lower wages than those of their more attractive rivals. The intuition is

straightforward and it stems from the frictional nature of the labor market: A less attractive

man needs a relatively high wage to be accepted by a woman but such a well-paid job may

be too difficult to find and so he settles for a lower wage. As a consequence, more attractive

men (single and married) will earn, on average, higher wages than less attractive (single and

married) men.

Our approach incorporates some aspects of the marriage market the importance of which

was stressed by Chiappori et al. (2012) in the context of assortative matching. They argue that

the standard marriage matching framework is too narrow, in the sense that it overlooks the role

played by uncertainty and it restricts attention to one-dimensional preferences. That is, it tends

to play down the role of search frictions and, in the main, it ignores preferences toward multiple

characteristics and the potential trade-offs this implies.3

We consider explicitly the random nature of matching and include preferences over multidi-

mensional (anthropometric and socioeconomic) characteristics. Indeed, our results are driven

entirely by the very existence of such search frictions and the perceived trade-off between phys-

ical attributes and wages. The only other papers we are aware of that consider marital matching

with multiple attributes in a search equilibrium framework are Coles and Francesconi (2011,

2017), who provide fascinating insights into some possible effects of equal opportunities for

women.

Interestingly, the strategies that give rise to the beauty premium can also account for the

so-called “marriage premium”: the puzzling empirical fact that, on average, married men earn

higher wages than single men.4We are the first to establish a theoretical link between the two

types of premia as equilibrium outcomes.

The potential market equilibria, each characterized by specific patterns of beauty premium

and marriage premium, allow for an extensive empirical test of our theoretical predictions.

Using U.K. data, we estimate the two types of wage differentials across male workers who

differ in terms of anthropometric characteristics. Following the literature, we use height and

weight as proxies for physical attractiveness, and the estimates seem to support our theory.

The empirical analysis we carry out is the very first attempt at finding evidence for the search-

theoretic approach to beauty premium and marriage premium.5

Hamermesh (2011) offers a stimulating survey of the literature on beauty premium. Hamer-

mesh and Biddle (1994) find that individuals with below-average attractiveness earn 9% less

than the “average-looking” ones, while the wage of individuals with above-average looks is

5% higher. These results are obtained after controlling for educational attainment and experi-

ence. Persico et al. (2004) attempt to quantify the so-called height premium and observe that

increasing height at age 16 by 1 inch increases adult wages by 2.6%, on average. In two fairly

recent studies using U.K. data, Case and Paxson (2008) and Case at al. (2009) find that the

3Chiappori et al. (2012) consider marital matching along multidimensional characteristics and reduce it to a matching

problem with preferences captured by a one-dimensional index. Using PSID data on married couples, they find an

interesting trade-off between anthropometric and socioeconomic factors affecting marital outcomes: Men compensate

1.3 additional units of body mass index (BMI) with a 1% increase in wages.

4Numerous studies report that, after controlling for education and other characteristics, the male marriage premium

is consistently around 10% or above, while such marital wage differentials are considerably smaller for women, and

their sign varies. For excellent surveys of the empirical literature, see Daniel (1995) or Grossbard-Shechtman and

Neuman (2003).

5In their survey chapter, Ponthieux and Meurs (2014) stress the need for empirical studies aimed at testing the search

theoretical approach to marital wage inequalities.