Elsevier

Labour Economics

Volume 14, Issue 3, June 2007, Pages 413-433
Labour Economics

The impact of obesity on employment

https://doi.org/10.1016/j.labeco.2006.02.008Get rights and content

Abstract

Using data from two rounds of the Health Survey for England I investigate the impact of obesity on employment. I use three approaches: a univariate probit model; propensity score matching; and IV regression using a recursive bivariate probit model. Conditional on a comprehensive set of covariates, the findings show that obesity has a statistically significant and negative effect on employment in both males and females. In males the endogeneity of obesity does not significantly affect the estimates, and the magnitude of effect is similar across the three methods. In females, failure to account for endogeneity leads to underestimation of the negative impact of obesity on employment.

Introduction

Obesity is a rapidly growing health problem that affects an increasing number of countries worldwide (WHO, 1998).1 In the United States, over a quarter of all adults are obese (HHSD, 2001), while in England and many other European countries the prevalence of obesity is also rising to epidemic proportions. In 1980 six per cent of males and eight per cent of females in England were obese; by 2003 the prevalence had trebled to 21 per cent and 24 per cent, respectively (Department of Health, 2003). The worldwide growth in obesity is a serious cause for concern because as well as being a debilitating condition in its own right, obesity is an important risk factor for a number of major diseases including coronary heart disease, type II diabetes, osteoarthritis, hypertension and stroke (NHLBI, 1998). In the United States obesity is the second leading cause of preventable disease and death next to smoking (HHSD, 2001), while in England seven per cent of all deaths are attributable to obesity (House of Commons Health Committee, 2004).

As well as imposing large morbidity and mortality costs obesity imposes a substantial financial burden. Studies for Australia (Segal et al., 1994), Canada (Katzmarzyk and Janssen, 2004), England (House of Commons Health Committee, 2004), France (Detournay et al., 2000), New Zealand (Swinburn et al., 1997) and the United States (Wolf and Colditz, 1998) report that obesity accounts for between 1% and 8% of national health expenditure. The effect on employers is also considerable; the economic cost of obesity to business in the United States is more than $12 billion per annum (Thompson et al., 1998), and in England there are in excess of 15 million days of medically certified sickness absence from work due to obesity each year (House of Commons Health Committee, 2004). Obesity may have other consequences that affect economic outcomes. For example, obese individuals are more likely to suffer from social stigmatisation and discrimination (NHLBI, 1998), which have been documented in a variety of settings including health care and the labour market (WHO, 1998).

The aim of this paper is to examine the impact of obesity on employment. The analysis is conducted using data from two rounds of the Health Survey for England. I use three approaches: a univariate probit model in which employment measured as a binary variable is regressed against obesity, also measured as a binary variable, plus a comprehensive set of individual and area covariates; propensity score matching using a variety of matching methods; and instrumental variables (IV) regression using a recursive bivariate probit model. I instrument individual obesity using an area level obesity measure: the prevalence of obesity in the area in which the respondent lives. This is positively correlated with individual obesity and is plausibly not itself correlated with the error term in the employment equation.

Several studies have previously analysed the effect of obesity on employment. In the first of two British studies Sargent and Blanchflower (1994) used the 1981 round of the National Child Development Study (NCDS) to examine the impact of obesity on the labour market outcomes of young British adults. They report the impact of obesity (defined as a BMI at the 90th percentile of the sample distribution or greater, and at the 99th percentile or greater) at age 16 years on unemployment at age 23 years, controlling for race, social class, region of residence and ability. Using a logit model the authors report that obesity has an insignificant effect on unemployment. Harper (2000) used a later (1991) round of the NCDS to estimate the impact of obesity, plus general physical appearance, attractiveness and height at age 23 years on unemployment at age 33 years also using a logit model. Obesity was defined by a BMI in the 80–89th percentiles and the 90–100th percentiles of the sample distribution. As with the earlier study the results showed that obesity had an insignificant effect on unemployment for both males and females.

Using a 1994 sample of Finnish adults, Sarlio-Lahteenkorva and Lahelma (1999) used a logit model to analyse the impact of current obesity (measured as a BMI of 30 kg/m2 or more) on current employment and long term unemployment, defined as being unemployed for two years or more in the previous five year period. Controlling for age, educational attainment, region of residence and limiting longstanding illness they found in females that obesity has a significant and positive impact on long-term unemployment, and an insignificant effect on current employment. For males, obesity had an insignificant effect on both employment measures.

None of these studies investigated the endogeneity of obesity and employment. Both Sargent and Blanchflower and Harper regress current employment against lagged obesity, which may deal with simultaneity, but will not correct for omitted variables that affect both obesity and employment (an example is time preference, which might affect human capital and tastes for work, and hence employment, and also obesity). Cawley (2000a) addresses the endogeneity problem using an IV regression approach. He analyses the impact of BMI on wages and employment in a sample taken from the US National Longitudinal Survey of Youth panel over the period 1981 to 1998. In the employment analysis, probit and IV probit models are estimated, regressing employment against BMI plus covariates. The instruments for BMI are the BMI of a biological child aged six to nine years plus interactions of this with the child's age and gender. The analysis is restricted to females who have borne children. In the probit model the BMI coefficient is not statistically significant. In the IV probit model BMI is found to have a positive effect on employment that is statistically significant at the 10% level.

This paper makes a number of contributions to the literature. First, it investigates the relationship between obesity and employment in England using a number of methods, in order to obtain a better understanding of the mechanisms by which obesity affects employment. Second, it considers males and females of normal working age rather than young adults, who have been the focus of previous British studies. Third, the data are more recent than those in earlier British studies, covering the period 1997–1998. This is useful because as the prevalence of obesity increases over time the impact of obesity on employment may change and older studies will become out of date. Additionally, since the election of the Labour government in May 1997 obesity has become an increasingly prominent policy issue and there have been a number of recent developments to address the health and social effects (Department of Health, 2004). The results presented in this study provide a baseline to assess the impact of these policies on the employment prospects of the obese.

The rest of the paper is structured as follows. Section 2 discusses the model underpinning the analysis and the econometric framework. Section 3 discusses the data and variables used. The estimation results are in Section 4 and Section 5 concludes.

Section snippets

Preliminaries

There are four reasons why obesity and employment may be correlated.

  • (1)

    Obesity causes unemployment. This might arise for two reasons. First, obesity can be a debilitating health condition (NHLBI, 1998). Therefore, all else equal, the obese are likely to be less productive than the non-obese, and therefore less likely to be employed. Second, there may be discrimination against the obese. Based on Balsa and McGuire (2003) this might arise for three reasons. First, there may be prejudice by

Data sources

I use pooled data from two rounds (1997 and 1998) of the Health Survey for England (HSE). The HSE is a nationally representative survey of individuals aged two years and over living in England. A new sample is drawn each year and respondents are interviewed on a range of core topics including demographic and socioeconomic indicators, general health and psychosocial indicators, and use of health services. Additionally, there is a follow up visit by a nurse at which various physiological

Results

The proportion of the sample in each obesity category is in Table 1. Only 34% of males and 43% of females are in the healthy category, while 17% and 19%, respectively, are obese. In males the overweight category has the highest proportion of the sample (46%), while in females the healthy category is the largest (43%). Table 1 also shows the percentage of the sample in each category that is employed. In all categories there are more employed males than females. In males the highest proportion

Concluding remarks

In this paper I investigate the impact of obesity on employment in England using three different methods and a dataset containing a rich set of variables likely to affect employment. Assuming plausibly that the instrument is valid I identify the causal effect of obesity on employment from other factors that might cause the two variables to be correlated.

In the IV models I find that obesity has a statistically significant and negative impact on employment in both males and females. The impact of

Acknowledgements

I would like to thank seminar participants at LSE Health and Social Care, the Department of Health, the Tanaka Business School and the Statistical Advisory Service, Imperial College London for their input into the paper. I would also like to thank the Editor and two anonymous referees for their constructive criticism and helpful suggestions.

References (50)

  • Burke, M.A., Heiland, F., 2005. Social Dynamics of Obesity. Florida State University working paper,...
  • Burke, M.A., Heiland, F., 2005. The strength of social interactions and obesity among women. Florida State University...
  • D. Card

    Using geographic variation in college proximity to estimate the return to schooling

  • Cawley, J., 2000a. Body weight and women's labor market outcomes. Working paper no. 7841, National Bureau of Economic...
  • J. Cawley

    An instrumental variables approach to measuring the effect of body weight on employment disability

    Health Services Research

    (2000)
  • J. Cawley

    The impact of obesity on wages

    Journal of Human Resources

    (2004)
  • J. Currie et al.

    Welfare and child health: the link between AFDC participation and birth weight

    American Economic Review

    (1993)
  • Department of Health

    Health Survey for England, 2001: A Survey Carried Out on Behalf of the Department of Health

    (2003)
  • Department of Health

    Choosing Health: Making Healthy Choices Easier

    (2004)
  • B. Detournay et al.

    Obesity morbidity and health care costs in France: an analysis of the 1991–1992 Medical Care Household Survey

    International Journal of Obesity and Related Metabolic Disorders

    (2000)
  • DTLR [Department of Transport, Local Government and the Regions]

    Measuring Multiple Deprivation at the Small Area Level: The Indices of Deprivation 2000

    (2000)
  • M. Everett

    Let an overweight person call on your best customers? Fat chance

    Sales and Marketing Management

    (1990)
  • H. Gravelle et al.

    A model of supply and demand influences on the use of health care: implications for deriving a ‘needs-based’ capitation formula

    Health Economics

    (2003)
  • B. Harper

    Beauty, stature and the labour market: a British cohort study

    Oxford Bulletin of Economics and Statistics

    (2000)
  • HHSD [Department of Health and Human Services, Public Health Service]

    The Surgeon General's Call to Action to Prevent and Decrease Overweight and Obesity

    (2001)
  • Cited by (193)

    View all citing articles on Scopus
    View full text