Health, income and relative deprivation: Evidence from the BHPS☆
Introduction
This paper is motivated by the growing literature on the relationship between health, income and the distribution of income. There is considerable evidence that income is a determinant of individual health (Ettner, 1996, Pritchett and Summers, 1997, Subramanian and Kawachi, 2004), this is often referred to as the absolute income hypothesis. However, there are a number of indirect mechanisms that may or may not be important determinants of health (Deaton, 2003). Indirect effects may be due to income inequality, relative deprivation, or other equally complex pathways. These pathways are often, unhelpfully, grouped together under the heading of the relative income hypothesis. The impact of such indirect mechanisms is still controversial, while evidence of ill-health due to the influence of hierarchy and inequality has been documented (Marmot et al., 1978, Sapolsky, 1993, Subramanian and Kawachi, 2004), there are a number of studies that do not support these findings (see Wilkinson and Pickett (2006) for a recent review).
This paper focuses on the impact of income and of relative deprivation on individual health. The impact of relative deprivation is one version of the relative income hypothesis. We use the definition of relative deprivation of Wagstaff and van Doorslaer (2000) which is defined as the difference between an individual's income and the income of individuals in their reference group. This differs from a hypothesis based on rank, where only position matters, however, for estimation the two may be indistinguishable due to the high collinearity between measures of income difference and rank (Subramanian and Kawachi, 2004). Another form of the relative income hypothesis is the income inequality hypothesis which focuses on overall income inequality, this measure is not considered in this paper, although the findings in this literature, as part of the wider relative income hypothesis, are important for motivating our paper.
A measure of relative deprivation is included as an extra regressor in models of the determinants of self-assessed health (SAH) and psychological well-being. Panel data allows us to control for individual heterogeneity which may be confounding the relationship. Semiparametric estimation is used to allow for a flexible functional form with respect to income. This will prevent the coefficient on relative deprivation being contaminated by higher order non-linearity in the relationship between health and income which is not captured by the usual logarithmic transformation.
The results from simple OLS models indicate that income and relative deprivation are both key determinants of health. However, as we progress through more sophisticated estimation strategies we find that the impact of income remains, while that of relative deprivation disappears. Only in robust semiparametric models of psychological well-being for men do we find that relative deprivation is statistically significant.
Section snippets
The literature
Economics has a long tradition of incorporating relative income and status into models of individual utility maximisation. Duesenberry (1949) hypothesised that individuals’ desire for self-esteem makes them emulate the consumption patterns of those who are on a higher socioeconomic rung of the ladder. Frank, 1985a, Frank, 1985b, Frank, 1997 considered the impact of positional goods and subjective well-being. Empirical work in economics has also started to investigate issues of interdependent
Methods
This paper specifically focuses on the estimation of the deprivation hypothesis: that individual health is affected by the difference between the individual's income and the income of individuals higher up the income distribution.2 Estimating the models raises a number of challenges. Firstly, it may be that individuals with high incomes have good health due
Data
The models are estimated using 11 waves of data from the British Household Panel Survey (BHPS), with the initial sample having been collected in 1991. It is a longitudinal sample of over 5000 households, containing over 10,000 adult individuals.
The longitudinal nature of the BHPS requires that individuals be re-interviewed each year. All individuals who were interviewed at wave one were re-interviewed (unless they dropped out, e.g., because of death or lack of cooperation) and these are known
Results
As our interest lies on the role of income and relative deprivation we only focus on these results.16, 17
The relationship between income and self-reported health
The literature on the impact of health on income is diverse and often controversial (Frijters et al., 2005). However, a non-linear relationship between health and income is often reported in the literature (Ettner, 1996), and the results in Table 3, Table 4 are largely in line with these results. Concentrating on results from the UK there is a large body of evidence in favour of income increasing health at a decreasing rate. Gravelle and Sutton (2006) and Benzeval et al. (2001) using the
Conclusion
This paper has made a number of contributions to the literature. Firstly, we have provided further evidence on the relationship between health and income. We find clear income effects for SAH for both men and women, and effects of income on GHQ for women. However, these effects are small in magnitude. The results for SAH do conform with the general literature, although this paper has highlighted a number of problems which we outline below.
Secondly, we have demonstrated that the specification of
Acknowledgements
The authors would like to acknowledge the comments of the referees as well as participants at the Health, Econometrics and Data Group, York. John Wildman would like to acknowledge the funding of the Leverhulme Trust. All the opinions expressed in this work are those of the authors.
References (76)
- et al.
Healthy, wealthy and wise? Tests for the direct causal paths between health and socioeconomic status
Journal of Econometrics
(2003) - et al.
Income and health: the time dimension
Social Science and Medicine
(2001) - et al.
Well-being over time in Britain and the USA
Journal of Public Economics
(2004) Pay reference standards and pay satisfaction: what do workers evaluate their pay against?
Social Science Research
(2004)- et al.
Satisfaction and comparison income
Journal of Public Economics
(1996) - et al.
The reliability of self-assessed health status
Journal of Health Economics
(2002) - et al.
Equity and equality in health and health care
Journal of Health Economics
(1993) - et al.
Income and health: what is the nature of the relationship?
Social Science and Medicine
(1999) New evidence on the relationship between income and health
Journal of Health Economics
(1996)- et al.
The causal effect of income on health: evidence from German reunification
Journal of Health Economics
(2005)
Money and mental well-being: a longitudinal study of medium sized lottery wins
Journal of Health Economics
Adaptation and scale of reference bias in self-assessments of quality of life
Journal of Health Economics
Relative income, race and mortality
Journal of Health Economics
Endocrinology alfresco: psychoendocrine studies or wild baboons
Recent Progress in Hormone Research
When information dominates comparison; learning from Russian subjective panel data
Journal of Public Economics
Aggregation and the measurement of income inequality: effects on morbidity
Social Science and Medicine
Does inequality in self-assessed health predict inequality in survival by income? Evidence from Swedish data
Social Science and Medicine
Income-related inequalities in health: some international comparisons
Journal of Health Economics
Modelling health, income and income inequality: the impact of income inequality on health and health inequality
Journal of Health Economics
Income inequality and population health: a review and explanation of the evidence
Social Science and Medicine
Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations
The Review of Economic Studies
Understanding the relationship between income and health: how much can be gleaned from cross-sectional data?
Social Policy and Administration
Problems with instrumental variables estimation when the correlation between the instruments and the endogenous explanatory variable is weak
Journal of the American Statistical Association
Measuring Health: A Review of Quality of Life Measurement Scales
Neuroeconomics: how neurosciences can inform Economics
Journal of Economic Literature
Social Ethical Index Numbers
Inequality-aversion and Income Mobility: A Direct Test
Unhappiness and unemployment
Economic Journal
The dynamics of health in the British Household Panel Survey
Journal of Applied Econometrics
Health, inequality and economic development
Journal of Economic Literature
Income, Saving and The Theory of Consumer Behaviour
Income-related reporting heterogeneity in self-assessed health: evidence from France
Health Economics
Income and well-being: an empirical analysis of the comparison income effect
Journal of Public Economics
How important is methodology for the estimates of the determinants of happiness
Economic Journal
The demand for unobservable and other nonpositional goods
American Economic Review
Cited by (162)
Partner's contribution to housework and couple bonding activities among Korean working women: Heterogeneity by socioeconomic status
2024, Women's Studies International ForumThe effect of housing price inequality on mental health
2023, Labour EconomicsOn a tendency in health economics to dwell on income inequality and underestimate social stress
2023, Economics and Human BiologyA tale of tails: Nonlinear effects of ability rank on adolescent psychological well-being
2023, Social Science ResearchGender heterogeneity in self-reported hypertension
2021, Economics and Human Biology
- ☆
The BHPS data were supplied by the Economic and Social Research Council Research Centre on Micro-Social Change, British Household Panel Survey. Colchester, Essex: The Data Archive, SN: 3953. Neither the original collectors of the data nor the Archive bear any responsibility for the analysis or interpretations presented here.