Obesity and labour market success in Finland: The difference between having a high BMI and being fat

https://doi.org/10.1016/j.ehb.2009.01.008Get rights and content

Abstract

This paper examines the relationship between obesity and labour market success in Finland, using various indicators of individual body composition along with body mass index (BMI). Weight, height, fat mass and waist circumference are measured by health professionals. We find that only waist circumference has a negative association with wages for women, whereas no obesity measure is significant in the linear wage models for men. However, all measures of obesity are negatively associated with women's employment probability and fat mass is negatively associated with men's employment probability. We also find that the use of categories for waist circumference and fat mass has a substantial influence on the results. For example, the category for high fat mass is associated with roughly 5.5% lower wages for men. All in all, the results indicate that in the absence of measures of body composition, there is a risk that labour market penalties associated with obesity are measured with bias.

Introduction

Many studies document the negative effects of obesity on labour market success measured by wages and employment (Atella et al., 2008, Baum and Ford, 2004, Brunello and D’Hombres, 2007, Cawley, 2004, Cawley, 2007, Cawley and Danziger, 2005, Conley and Glauber, 2006, Garcia and Quintana-Domeque, 2006, Han et al., in press, Härkönen, 2007, Morris, 2007, Kennedy and Garcia, 1994, Sarlio-Lähteenkorva et al., 2004). For reasons of data availability, this literature has used body mass index (BMI) as a measure of obesity.1 However, it is difficult to determine whether the labour market penalties for obesity are due to discrimination or health reasons such as the limited ability to work. One reason for this is that BMI blurs the distinction between fat and fat-free mass such as muscle and bone (Burkhauser and Cawley, 2008). In the medical literature, BMI alone is not considered to be a valid measure of obesity nor a sufficient predictor of obesity-related health outcomes (Burkhauser and Cawley, 2008, Yusuf et al., 2005).

This paper re-examines the relationship between obesity, wages and employment, using indicators of individual body composition along with BMI. The indicators we use are fat mass expressed as kilograms of fat and waist circumference. Waist circumference distinguishes individuals who have a high fat mass with the bulk of the fat concentrated around the waist compared with those with a lot of fat that is more evenly distributed around the body. A large waist circumference in relation to height may be interpreted by employers as a non-attractive physical appearance, which has been found to be associated with lower earnings (Hamermesh and Biddle, 1994). The reason for this is that the fat concentrated around the waist is more visible than fat in general. Moreover, waist circumference measures fat that is visceral (i.e. around the internal organs), which is especially harmful to health (Kopelman, 2000). Therefore, it is challenging to distinguish different channels of influence even with the use of measures of body composition.

We use data from the “Health 2000 in Finland” data set, a cross-section of about 8000 people above the age of 30. (Aromaa and Koskinen, 2004, provide a description of the data set.2) This data set contains information on individual fat mass measurements obtained from an eight-polar bioelectrical impedance analysis, which is performed by running a small constant current through the body (Scharfetter et al., 2004). Resistance, or impedance, is higher in fat than in other types of tissue, which makes it possible to calculate the proportion of fat mass in the body.

An advantage of the data is that it contains information not only on fat mass but also on other measures of obesity. The few existing studies of fat mass and labour market success (Burkhauser and Cawley, 2008, Heineck, 2007, Wada and Tekin, 2007) have almost exclusively relied on prediction equations for the measures of body composition as a function of electrical resistance as well as height and weight and some other variables.3 This approach is not as accurate as one based on actual measurements.4

Annual individual wage data originating from the Finnish tax authorities have been linked to the Health 2000 data set, using the personal identification number that every person residing in Finland has.5 This is another advantage over most of the earlier studies in this field of research, because almost all of them have used survey-based information on earnings that is prone to non-response and reporting bias.

Nonetheless, our data has two shortcomings. First, we are using a cross-sectional data. Thus, we cannot estimate fixed effects models that would account for unobservable heterogeneity at the individual level. Second, we lack a valid instrument.6 Hence, we cannot estimate causal effects and address the possibility that the obesity measures may be endogenous (Averett and Korenman, 1996, Cawley, 2004, Wada and Tekin, 2007). Accordingly, this paper documents associations or correlations between different measures of obesity and labour market success.7 The Finnish evidence is of broader interest, because the prevalence of overweight and obesity has increased rapidly among both men and women during the past few decades (Böckerman et al., 2008). The increase in obesity in Finland has not been as rapid as in USA. Despite this, the share of obese adults is higher in Finland than in other Nordic countries (Audretsch and DiOrio, 2007).

Section snippets

Data and empirical approach

The “Health 2000” population survey data set was collected in order to give a comprehensive picture of the health and functional ability of the working-age and old-age Finnish population. The data set is a random sample of 10,000 adults from the entire country, and the information was collected between September 2000 and June 2001 by means of personal interviews, telephone interviews, and professional health examinations. Importantly, all measures of obesity originate from professional health

Wages

In contrast to many other studies for women BMI and the square of BMI are insignificant (Table 3, Columns 1 and 5).17

Conclusions

There is increasing concern among medical researchers about the reliability of BMI as a measure of obesity because of its inability to distinguish between body fat and fat-free mass. Recently, economists have also recognized these concerns and started utilizing alternative measures to study the effects of obesity on economic outcomes (Burkhauser and Cawley, 2008, Heineck, 2007, Wada and Tekin, 2007). The obesity measures used in this paper are BMI, weight, fat mass and waist circumference. An

Acknowledgements

The authors would like to thank John Cawley, five anonymous referees and John Komlos (Editor) for valuable comments that have greatly improved the paper. An earlier version of this paper circulated under the title “The effect of obesity on wages and employment: The difference between having a high BMI and being fat”. Earlier versions were presented at the EALE Conference, Oslo, and at the Nordic Summer Institute in Empirical Labour Economics, Helsinki. We are grateful to the seminar audiences

References (40)

  • S. Averett et al.

    The economic reality of the beauty myth

    Journal of Human Resources

    (1996)
  • C. Baum et al.

    The wage effects of obesity: a longitudinal study

    Health Economics

    (2004)
  • G. Brunello et al.

    Does body weight affect wages? Evidence from Europe

    Economics and Human Biology

    (2007)
  • P. Böckerman et al.

    Unemployment and self-assessed health: evidence from panel data

    Health Economics

    (2009)
  • J. Cawley

    The impact of obesity on wages

    Journal of Human Resources

    (2004)
  • J. Cawley

    The labor market impact of obesity

  • J. Cawley et al.

    Morbid obesity and the transition from welfare to work

    Journal of Policy Analysis and Management

    (2005)
  • J. Cawley et al.

    A comparison of the relationship between obesity and earnings in the U.S. and Germany

    Schmollers Jahrbuch

    (2005)
  • D. Conley et al.

    Gender, body mass and socioeconomic status: new evidence from the PSID

  • J. Fox et al.

    Generalized collinearity diagnostics

    Journal of the American Statistical Association

    (1992)
  • Cited by (0)

    View full text