How About No | Course Paper

Are the healthier wealthier or the wealthier healthier?

The European Evidence

This book reviews the literature on the association between socioeconomic status (SES

hereafter) and health status (HS hereafter) with special emphasis on the older labour force for

a number of European Union countries, namely, Denmark, Finland, France, Greece, the

Netherlands and the UK. The main focus of the review is to gather the evidence on (i) the

relationship between income inequality or poverty and physical or mental health, (ii) the

relationship between individual socio-economic status indicators, including income and

occupational status, and the individual health status and (iii) the interrelationship between SES,

physical and mental health and labour market participation. In particular, this review focuses

on the evidence regarding the older workforce – that is the 50 to 65 age group, and, when

possible, their implications in terms of retirement behaviour are also explored. This review

casts light on those aspects that can reasonably be thought of as being important components

of the association between SES and HS but are not explored by researchers in some of the

European countries. For instance, despite the concern with the ageing population in Europe,

the repercussions of the early exit from the labour market in light of the SES-HS nexus have

not been examined in all countries. If such knowledge gaps do exist, then one also has also to

address the question of whether this is due to data limitations, methodological difficulties or

simply due to a lack of interest.

The starting point is a discussion the reasons why the relationship between HS and SES is so

important to quantify and show that there is no consensus among researchers on the

mechanisms underlying the HS-SES relationship. This relationship is so complex that it raises

important methodological difficulties. These problems are discussed in detail. In the

subsequent chapters important studied are summarized and their findings contrasted and

discussed.

1. Why is it important to investigate the relationship between HS and SES?

2

Economic inequalities are a common characteristic of market economies (Atkinson et al, 1995,

Atkinson, 1996). As such, their sources as well as their consequences are of a major concern

not only to academia but also to policy makers. The identification of inequality sources (the

sources of inequality) is more than a requirement if correction mechanisms are to be designed

in order to reduce economic inequalities. Indeed, while the theoretical debate is on whether

redistributive mechanisms are likely to reduce market efficiency or not, policy makers seem

to agree upon the idea that redistribution should compensate at least for differences in initial

endowments, that is, in inequality factors which are out of individuals’ control (Piketty, 1997).

To illustrate this point, suppose that part of the population is endowed with poor health due to

either genetic factors or poor early childhood environment. It is very likely that the

performance of this part of the population in the labour market will be less successful than

that of healthy individuals and hence the level of earnings and wealth accumulation over the

lifetime will be lower (Kaestner & Corman, 1995, Currie and Hyson, 1999, Smith, 1999).

Thus, health endowments could influence one’s socioeconomic status. This means that one

way of reducing economic inequalities could be, for instance, a policy aiming at equalising

access to health care. Indeed, policy makers could target economic inequalities per se but they

could also aim at reducing health inequalities which, in turn, would increase the earning

power of the particular section of the population. However, the design of a combined but also

well-balanced policy is possible only if the association between health and socioeconomic

inequalities is well understood.

From an economic point of view, inequality is not necessarily bad. Some inequality in

outcomes might provide individuals with incentives to perform better. In contrast, inequality

in opportunities may result in persistent inequalities in outcomes. As such, it might have

remarkable economic and social implications. First, the redistribution mechanisms it might

require could turn out to be very costly (Bourguignon, 1999). Second, most of the time,

inequality in opportunities is considered to be an indicator of social progress, that is, of how

the outcome of economic growth is shared among the members of the society (Atkinson et al,

2001). Third, it might have serious repercussions on the social environment in terms of crime

level, educational inequality and, in general, in terms of individuals’ well-being in society

(Waildman & Andrew, 2001, Frey & Stutzer, 2002). Indeed, suppose that part of the

population suffers from poor socioeconomic status due to either the absence of adequate

3

wealth or due to a poor earning power in the labour market. Individuals from this part of the

population are very likely to adopt risky behaviour such as smoking, alcohol consumption or

sedentary lifestyle. They are also most likely to be less aware of the importance of hygiene or

preventive health care (Ruhm & Black, 2002). Moreover, depending on the institutional

framework of the health care system, they might have limited access to health care or have

access to only low-quality care. This suggests that socioeconomic status may influence one’s

health status. It also implies that one way of reducing health inequality could consist, for

example, in designing redistributive mechanisms to reduce economic inequality. This

connection between health inequality and economic inequalities, in turn, implies that any

policy aiming at reducing the former should also affect the latter.

Furthermore, governments intervene specifically in the health sector either directly through

the provision of funding, or indirectly through regulation, and the policies underlying such

intervention must balance efficiency concerns with equity, a conflict which, in turn, has

serious implications on the performance of the health sector. On the one hand, the importance

of the health sector is such that any wasteful use of resources can be a serious burden on the

public sector and to the economy as a whole (European Competitiveness Report, 2004,

Chapter 3). On the other hand, health has a significant role in determining economic growth

via its impact on labour productivity (Bloom, Canning and Jamison, 2004).

However, the health sector also affects the competitiveness of the overall economy. This

influence may occur via two channels. The first is the effect of tax rates and insurance

contributions on labour costs. The second is the effect of taxation and national and/or

occupational insurance schemes on job mobility and hence on labour market flexibility. This

highlights the role of institutions. For instance, problems of supplier induced demand are very

likely to result in over-consumption of healthcare services and hence to result in unnecessarily

high increases of labour costs relative to the health-related increases in productivity. Likewise,

job mobility is likely to depend on the method of health care funding. Workers’ propensity to

move jobs is certainly related to their fear of losing insurance coverage or of facing higher

insurance prices or lower benefits (Holtz-Eakin, 1994; Gruber, 1998).

4

A further reason why the policy implications of the association between SES and HS should

be investigated is the issue of ageing. Any improvement of the health of the adult population

increases the size of the population and hence influences the dependency ratio (Bloom,

Canning and Jamison, 2004). However, whether this influence is positive or negative depends

on the part of the adult population that mainly experiences the improvements in health. Indeed,

the dependency ratio will decrease only if health improvements result in an increase of the

available labour force. In contrast, if improvements in health are mainly experienced by those

adults who have past retirement age, due, for instance, to an increase in life expectancy, this

may actually cause the dependency ratio to rise. This is clearly relevant given the current

trends towards an ageing population in Europe. A relatively stable or even declining

workforce in most EU states has to finance the healthcare for increasing numbers of elderly

citizens. The increase in the numbers of retired people also creates an increasing burden on

countries’ pension systems, since fewer workers are contributing per retirees claiming. It is

therefore important to identify the determinants of labour market participation behaviour of

the older workforce. Economic theory suggests socioeconomic status is one of these

determinants, but there is also evidence that health is one of the main factors that affects the

choice of a retirement date (McGarry, 2002. See also Mein et al, 2000).

One important implication of this is that if both the health and labour market participation

decisions of the older workforce are severely affected by poor socioeconomic conditions, then

higher economic inequalities may turn out to require a substantial increase in public

expenditure on the older workforce. Furthermore, not only does early retirement imply a

lower financial contribution from these individuals to the social security system, but it is also

the case that low health status, together with a poor socioeconomic status, might result in an

increase of transfer income that must be distributed to such individuals, which, in turn, affects

the cost of welfare provision. The size of the impact of the socioeconomic and occupational

differences in health is therefore crucial, not only for the understanding of the development of

age-related diseases and disability, but also for how these may be avoided by controlling

contributory socioeconomic factors. There is indeed clear evidence that in many countries,

health care expenditures on the elderly are now growing much faster than GDP (Fuchs, 1998).

Of course, such policies may, in turn, require changes to the level of targeted spending and

volume of protective legislation.

5

Because of the dynamic nature of the association between HS and SES, its importance is very

likely to be age specific. Therefore, its implications for the older workforce cannot be inferred

from general analyses, but rather from studies that specifically focus on the older workforce.

First, it might simply be the case that only after a reasonably long working career, do the

health consequences of a poor socioeconomic status become really disabling (Budetti et al,

2001). Second, there are cohort effects in the sense that the current health status of the older

workforce, as compared to that of younger individuals, is the outcome of decisions that have

been made in a different economic and institutional environment (Deaton & Paxson, 2001).

Third, neither the health nor the socioeconomic status of children is the outcome of their own

decisions or behaviour. In contrast, as individuals become older, they make decisions which

condition their health as well as their socioeconomic status (Smith, 1999, Venti & Wise,

2002). Consequently, both early childhood factors, which are out of individuals’ control, and

the sequential decision process they have adopted throughout their life cycle determine

individuals’ health and socioeconomic status at the later stages of working life (Smith, 1999,

Grossman, 1972).

2. The Received Wisdom on Health Inequalities

Evidence on socio-economic differences in mortality can be traced back to the XVIIIth

century. But empirical research in this area shows that such inequalities are still very high

nowadays (e.g. Couffinhal et al., 2005). This justifies the large body of literature aiming at

identifying the main determinants of socio-economic inequalities in health. Historically, the

origin of socio-economic inequalities in health was attributed to individuals’ living conditions

and habits and only in the twentieth century have researchers considered the role of working

conditions and access to health care, all these factors being potential explanations of why the

poor are endowed with poorer health than the rest of the population. In the early eighties,

however, the publication of the Black report (Black et al., 1982) cast doubt on this “absolute

poverty” model (Wilkinson, 1986). Not only did social inequality in mortality not decrease in

Great Britain between 1931 and 1981, but it even increased, despite the improvement of

general living conditions as well as working conditions and despite the establishment in 1948

of the National Health Service, which offered British citizens equal access to health care. In

addition, socio-economic inequalities in health did clearly not reduce to contrasting the

6

contrast between the poor and to the wealthy or manual to and non-manual workers. It

became clear, indeed, that there is a real social gradient in health which cannot be solely

explained by material living conditions, the risk of death being continuously decreasing along

the social hierarchy.

One of the hypotheses that the Black report had favoured is that of selection: inequalities in

mortality are not due to the effect of socio-economic status on health, but are rather due to the

effect of health on socio-economic status and social mobility. A number of studies, however,

have shown that only a limited part of inequality in mortality is explained by such an effect.

For instance, Fox & Goldblatt (1986) as well as Marmot (1986) have shown that differences

in mortality are of the same extent whether one considers individuals’ occupation at their time

of death or the one they held many years prior to death. Actually, the debate among

researchers is not clear cut yet: While there is agreement that health status has an effect on the

probability of withdrawal from the labour market, the results regarding the effect of health on

wages are still mixed (Currie & Madrian, 1999).

The lack of clear environmental explanations of why there are health status differences across

social classes has induced researchers to consider behavioural explanations instead. Studies of

the European Working Group on Socio-Economic Inequalities in Health (Mackenbach et al.,

1997; Kunst, 2000) have shown that while EU member countries differ in the extent of socio–

economic inequality in mortality the latter also increases with the social gradient in life styles

(Alcohol consumption, smoking, dietary habits, sedentary life style, etc). However, the

conclusions of the Whitehall study (Marmot, 1986) again cast doubt on the importance of

these behavioural explanations. Indeed, it shows that health status differences between British

civil servants remain after a number of lifestyle factors are controlled for. Likewise, in a study

by Marmot (2000), only the third of the variance of mortality due to coronary diseases has

been shown to be attributable to smoking, cholesterol, blood pressure, sedentary lifestyles and

height.

Interestingly enough, a number of researchers have explored the idea that the effects of

environmental and/or behavioural factors cumulate over one’s life cycle. Only by accounting

for these cumulative effects would one be able to explain both health and socio-economic

statuses (Blane, 1999). In this view, the causes of health differences at adult age should be

searched for in individuals’ early childhood environment. Health during childhood has

important consequences on individuals’ working careers (Wadsworth, 1986). Likewise, the

living conditions of children or even intra-uterine environment have important consequences

for their health at adult age (Smith, 1999; Wadsworth, 1999).

It is probably the failure of the dominant model to result in convincing explanations of the

social gradient in health that led a number of social epidemiologists to engage in a different

research path, exploring the so-called social determinants of health (Marmot & Wilkinson,

1999; Berkman & Kawachi, 2000). The idea is that it is not the absolute standard of living

which matters for individuals’ health, but rather the relative level. To be more specific,

feelings of hierarchical domination, loss of autonomy, etc are important sources of stress and

mental bad-being. Generalisation of this hypothesis implies that, depending on how societies

are socially structured, lack of social cohesion might induce psycho–social stress which, in

turn, has important consequences on mental health, cardio-vascular diseases and, more

generally, on the sensitivity of individuals to illnesses.

A remarkable feature of the literature on the association between health and socio-economic