Trends in the association between average income, poverty and income inequality and life expectancy in Spain
Introduction
Several recent studies have shown that inequality in income is an important predictor of mortality and health status in relatively wealthy societies. Although some research in the late seventies had already shown the importance of this association (Lynch & Kaplan, 1997) the study of the association between income inequality and mortality came to be widely known in the fields of sociology and health beginning in the mid eighties, with the work published by Wilkinson. These studies showed important associations between inequalities in income and differences in life expectancy among industrialised countries (Wilkinson (1986), Wilkinson (1990), Judge (1992a), Wilkinson (1992b)).
Judge et al. (Judge, 1995; Judge, Mulligan, & Benzeval, 1998) made a critical analysis of these studies based on international comparisons and noted that many biases had not been taken into account when interpreting their results, for example, the lack of data comparability; the fact that income was not adjusted for taxes, benefits and household size; the lack of a theory or a rational basis for the measure of income inequality selected; and the failure to control for other factors that might be interacting with income. However, several studies comparing the 50 states and the metropolitan areas of the US, in which the same source of data was used, and which controlled for several confounding variables and took into account household size and the differences in taxes among the different states, have confirmed Wilkinson's findings in detecting an important relation between various measures of inequality and different health indicators (Kaplan, Lynch, Pamuk, Cohen, & Balfour, 1996a; Kaplan, Pamuk, Lynch, Cohen, & Balfour, 1996b; Kennedy, Kawachi, & Pothrow-Stih, 1996; Kennedy, Kawachi, Glass, & Prothrow-Stith, 1998; Lynch et al., 1998; Ross et al., 2000). Other studies comparing different countries have also found an association between income inequality and premature mortality (Van Doorslaer et al., 1997; Lobmayer & Wilkinson, 2000).
In other cases, these findings have not been confirmed. For example, some authors have observed that changes in the magnitude of socio-economic differences in mortality in the Nordic countries have not been associated with changes in income distribution in the population (Martikainen & Valkonen, 1999). Likewise, a study using data from different Spanish regions found no relation between inequalities in income and the prevalence of long-term disability (Regidor, Navarro, Domı́nguez, & Rodrı́guez, 1997). Similarly, in Canada no relation was found between income inequality and mortality when comparing the different provinces and metropolitan areas (Ross et al., 2000).
Despite these discrepancies in the empirical evidence on this subject, and even though many studies come from a single country—the United States—it is surprising that a wide theoretical debate exists about the possible mechanisms that mediate the relation between income inequality and mortality (Muntaner & Lynch, 1999; Wilkinson, 1999; Lynch, Davey Smith, Kaplan, & House, 2000; Baum, 2000; Wilkinson, 2000; Lynch, Due, Muntaner, & Davey Smith, 2000). Some authors explain this relation in terms of a perception of relative disadvantage and the psychological consequences of inequality, while others focus on the structural causes of inequalities, such as the systematic underinvestment across a wide range of human, physical, health and social resources.
This debate is even more surprising because, as some authors have pointed out, research on the relation between income distribution and health is in its infancy, so there are many issues yet to be explored (Lynch & Kaplan, 1997). With the exception of one study (Chiang, 1999) there is little empirical confirmation of the original hypothesis according to which variations in life expectancy in the more developed countries are progressively more dissociated from absolute income level and increasingly associated with measures of income inequality (Wilkinson, 1994; Wilkinson, 1996). Likewise, despite the fact that many studies have found the association between socio-economic status and mortality to be less strong for women than for men, few studies of inequalities in income and health have studied men and women separately. Consequently, the present study was undertaken to test the original hypothesis in Spain, by analysing the association between average income and life expectancy and between different measures of income inequality and life expectancy in 1980 and 1990, using the 17 regions of Spain as the unit of analysis, and analysing men and women separately.
Section snippets
Source of data
The measures of average income and income inequality were obtained from the study “Inequality and Poverty in Spain” (Martı́n-Guzman et al., 1996). The investigators in that study used information contained in the Survey of Family Budgets (EPF) to estimate average income, income inequality and measures of poverty in Spain's 17 regions. In this study, we used the indicators calculated from the EPFs of 1980–81 and 1990–91.
The EPF is carried out periodically by the National Statistics Institute;
Results
Table 1 shows the median and range for average income, measures of income inequality and measures of relative poverty in 1980 and 1990, respectively. The median of the measures of income inequality and of relative poverty decreased between 1980 and 1990. For the measures of income inequality the range was wider in 1990 than it was in 1980 because the maximum value was higher; in contrast, the range became narrower for the measures of relative poverty because both the lowest value and the
Discussion
The results of this study show that in men no statistically significant association between average income and life expectancy was found in either the first or second period, whereas in women the association seen in 1980 between life expectancy and average income adjusted for measures of income inequality decreased or disappeared in 1990. No statistically significant relation was found in either men or women between the measures of income inequality and life expectancy after adjusting for
References (49)
- et al.
Income inequality and population health
Social Science & Medicine
(1998) The patterning of health by social position in contemporary BritainDirections for sociological research
Social Science & Medicine
(1986)- et al.
Income-related in healthSome international comparisons
Journal of Health and Economics
(1997) On the measurement of inequality
Journal of Economic Theory
(1970)Social capital, economic capital and powerFurther issues for a public health agenda
Journal of Epidemiology and Community Health
(2000)- et al.
What is the lag between income inequality and health status
Journal of Epidemiology and Community Health
(2000) - et al.
East–West mortality divide and its potential explanationsProposed research agenda
British Medical Journal
(1996) - et al.
Widening social inequalities in mortalityThe case of Barcelona, a southern European city
Journal of Epidemiology and Community Health
(1997) Economic transition and changing relation between income inequality and mortality in TaiwanRegression analysis
British Medical Journal
(1999)- et al.
Intravenous administration among heroin users in treatment in Spain. Geographical variations and other related factors
International Journal of Epidemiology
(1994)
The impact of drug-related deaths on mortality among adults in Madrid
American Journal of Public Health
La heroı́na y la cocaı́naAspectos de la polı́tica de salud pública en la década de los noventa
Quadern Caps
Income distribution and life expectancyA critical appraisal
British Medical Journal
Income and mortality in the United States [letter]
British Medical Journal
Inequality in income and mortality in the United StatesAnalysis of mortality and potential pathways
British Medical Journal
The relationship of income inequality to mortalityDoes the choice of indicator matter?
Social Science & Medicine
Income distribution, socio-economic status, and self-rated health in the United StatesMultilevel analysis
British Medical Journal
Income distribution and mortalityCross sectional ecological study of the Robin Hood index in the United States
British Medical Journal
Cited by (44)
Spatiotemporal trends in life expectancy and impacts of economic growth and air pollution in 134 countries: A Bayesian modeling study
2022, Social Science and MedicineCitation Excerpt :They have largely focused on inequality, and showed that economic inequality produces health inequalities related to LE both within and between countries (Marmot et al., 2012; World Health Organization, 2008). Strong evidence for the existence of important gradients in health outcomes, such as LE, by economic conditions has been identified both in developing and developed countries (Regidor et al., 2003; Babones, 2008; Rasella et al., 2013; Chetty et al., 2016); indeed, in most countries. These studies have potentially important consequences for how we think about public health outcomes (e.g., LE) and economic development.
Campania and cancer mortality: An inseparable pair? The role of environmental quality and socio-economic deprivation
2021, Social Science and MedicineCitation Excerpt :A branch of the literature argues that the probability of diagnosing cancer at an early stage is significantly lower in economically deprived areas, since residents of these areas resort to screening procedures less frequently and are less willing to seek medical care in view of potential cancer symptoms (Catalano et al., 2003; Bennett et al., 2011). More broadly, many studies have focused on the relationship between deprivation (cultural, material, relational, etc.) and health (Regidor et al., 2003; Mackenbach, 2006; Ministry of Health, 2011), finding that more deprived areas generally feature more health-related problems. Among others, Azamjah et al. (2019) evaluated the trend of breast cancer mortality rate (CMR) during the 1990–2015 period in a cross-country perspective.
The role of geographic scale in testing the income inequality hypothesis as an explanation of health disparities
2012, Social Science and MedicineCitation Excerpt :Other factors could contribute to the gender difference as well. Regidor (2003) concluded that life expectancy in developed countries has become more dissociated from average income and more associated with income inequality among women and surmised that the absence of such a relation in men may be due to the impact of premature AIDS-related mortality. Gold (2004) found that higher income inequality is associated with closer pregnancy spacing among a welfare-recipient population, which might be another pathway through which gender is related to the IIH as close pregnancy spacing is related to repeated teen pregnancy among teen girls and often associated with poor social and economic outcomes.
Measuring inequalities in health: What do we know? What do we need to know?
2012, Health PolicyCitation Excerpt :When analysing equity in outcomes at the individual level, different health outcome measures have been used in the literature, ranging from subjective measures of health such as self-assessed health (e.g., [42]) and quasi-objective indicators of health status such as the SF-36 physical functioning score (e.g., [43,44]), indicators of specific illnesses such as coronary heart disease (e.g., [45]), limiting longstanding illness (e.g., [46]), BMI (e.g., [106]), and mental health problems [47], to more objective measures of health such as biological markers [48]. At an aggregate level, health population variables such as life expectancy [49], QALYs (e.g. [50,51]), DFLE [108] have been used in the literature. However, given the data requirements for the analysis of population level data, most of the literature measuring socioeconomic inequality in health relies on individual health indicators.
Multilevel analysis of income, income inequalities and health in Spain
2012, Social Science and MedicineCitation Excerpt :Among the normative inequality indices, the Atkinson one does not show statistical significance for either dependent variable in this study, regardless of the degree of aversion to inequality considered. This differs from research in the US (Zheng, 2009) and it is partially in line with a previous Spanish study (Regidor et al., 2003). Further, our study is the first in the literature to test the RIH with two social welfare indices derived from the Dalton–Atkinson–Sen approach.