Printer Friendly

Health, social insurance and income.

Abstract Health, a form of human capital, can be defined by longevity and physical wellbeing. Social policy decisions require an understanding of the factors that contribute to the creation of health inequalities. To learn more about socioeconomic variables and health capital, this paper examines the relationship between three key variables: health, social insurance, and income, for the Swedish population. Using a randomized research survey design, data from 3,600 participants of a larger Swedish study, conducted in 2005, was analyzed. A linear model of Three Stage Least Squares was chosen to correct for simultaneous bias in the Health, Social Insurance, and Income (HSI) Model. Findings confirm the importance of socioeconomic, behavioral and environmental factors in explaining health inequalities. The results clearly show men, educated people, nonsmokers, individuals that exercise and youngsters possess higher health status than other people. The dependency on social insurance is mainly caused by poor health; a higher degree of social insurance dependency was offset by income increases due to age and higher professional level.

Keywords Health. Structure. Equality. Efficiency. Social insurance. Income

JEL C10. D60. I00. J30

Introduction

Health inequalities continue to be a persistent and growing policy problem in most countries. A fundamental requirement in the development of a social policy to counteract or control such differences is a good understanding of the factors contributing to the creation of health inequalities (Jones and Rice 2005). Health policy decisions deal with, among other things, resource allocation of healthcare services and other activities which may influence health and behaviors of institutions and individuals. Health, as a form of human capital, can be defined by longevity andphysical wellbeing. This paper aims to determine the relationship between three key variables: health, social insurance dependency, and income (HSI), for the Swedish population. Further analysis focuses on population characteristics in relation to health, in an attempt to determine the significance of social and environmental factors in explaining differences in health status.

Background

Nearly 25% of the potential European labor force is dependent on allowances and other benefits from the welfare state. Unemployment, sick leave, and early pensions are the main benefits. Of these, sick leave and early pension correspond to states of incapacity, due to the lack of health and/or the ability to adjust to work demands. Sadly, this situation has deteriorated over the past 30 years (Ferraz-Nunes 1996) because (1) dependency on the welfare state contributes to rigidity in the labor market and places a heavy financial burden on the working population, (2) budget deficits cause disturbances in the economy because individuals loose their capacity as members of the labor force and society and (3) an increased demand on healthcare may occur due to poor health related to such a dependency. Thus, there is a need to increase our knowledge about the relationship between socioeconomic variables and health capital. This relationship is always explained by interdependent relations among difference variables and, thus, there is a structural problem.

Methods

Sample and Setting

A randomized research survey design was used to elicit data from 3,600 individuals who participated in a wider survey encompassing the entire population of Sweden (Ferraz-Nunes 2006). The initial 2005 survey included questions comprising several fields to acquire information about the perceptions of society members.

Model

It is assumed a structural problem exists when the factors influencing health, social insurance dependency, and income are investigated simultaneously. Therefore, a linear model of Three Stage Least Squares was chosen to correct for the simultaneous bias (Chamberlain 1984). This decision may cause one to be cautious about the interpretation of the size of the parameters. The HSI model is summarized by the following equations:

HEALTH = [f.sub.1] (age, gender, education, smoke, exercise) (1)

SOCINSUR = [f.sub.2] (health, gender, education) (2)

INCOME = [f.sub.3] (age, profession, social insurance dependency). (3)

Health Status

The EQ-5D is an instrument of a general measure of health related quality of life (HRQL), the level of health status. Using the EQ-5D, health status (HEALTH) corresponds to the value of time trade-off (TTO) tariff according to the EQD5-questionnaried used. In this analysis, the values of the TTO-tariff are based on a large-scale British survey (Dolan et al. 1996). The HRQL dimension corresponds to values that assign weights to different health states on a scale from zero (dead) to one (wellness) (Kaplan 1996). This can be used to generate descriptive data, as a simple method to elicit a person's own rating of his current health status and as a preference based generic index (Williams 1995). The EQ-5D index is based on a five-dimensional set (mobility, self-care, usual activities, pain/discomfort, anxiety/ depression), where each dimension has three levels of severity with the ability to generate a total of the 243 different health states. Unconscious and dead are added to these, making a grand total of 245 different health states. For an exhaustive explanation of the state of the art in EQ-5D, refer to Brooks et al. (2003). The index value ranges from zero to one.

Other Variables

Social insurance dependency (SOCINSUR) corresponds to the degree of dependency of cash benefits due to sick leave, unemployment or early pension, for every individual, where 100 indicates 100% financial dependency of social insurance and zero reflects no dependency. Household income (INCOME) corresponds to the annual income measured in thousand Swedish crowns (SEK). A single American dollar (USD) is about 7 SEK. Age (AGE) is the actual age of an individual. Gender (GENDER) is indicated by two values: one for women and zero for men. Professional groups (PROF) qualify individual into one of four different groups, where 1 = unskilled blue collar and very small business, 2 = specialized worker and white collar workers without leading tasks, 3 = managers with higher education demanded, and 4 = industrialist, administrators, and academic professionals. Education (EDUC) refers to the level of academic preparation, where 0 = meets less than basic requirements to 10 = meets doctoral degree or similar requirements. Smoking habits (SMOKE) reflect the number of cigarettes smoked, where 0 = no cigarettes and 10 = more than a packet (20 cigarettes) a day. Exercise (EXERCISE) corresponds to how often an individual exercises, where 0 = no exercise and 10 = exercise every week.

Hypotheses and Motivation

Equations 1 and 2 form the main question in this paper and determine how some major factors may explain difference in health at the population level. It is hypothesized that more smoking and less exercise will correlate to a lower level of health status (Kenkel and Decicca 2000). Health deteriorates with higher age and produces a negative correlation between the two factors. Earlier studies indicate women report more health problems than men (Ferraz-Nunes 2006). Thus, it ishypothesized that there is a negative relationship between gender and health status. Some researchers suggest health status is positively correlated with education level (Fuchs 1982; Grossman 1972). In a recent study, Lleras-Muney (2002) concluded education has a clear, causal and positive effect on health.

Equation 2 presents the dependent variable as the degree of social insurance dependency (SOCINSUR). Social insurance is divided into three parts corresponding to sick leave, unemployment benefits and early retirement. Sick leave should be related to poor health. Unemployment is not necessarily connected to poor health; however, there is high probability, at least, persistent states of unemployment may negatively influence health status. Early retirement is due to poor health or disability by definition, often connected to persistent unemployment and older age. In Sweden, coverage is almost universal and intends to substitute decreased income due to work absence. The main reasons for this situation are (1) poor health, (2) lack of appropriate job or (3) both. The worse the state of health status (HEALTH), the likelihood of social insurance dependency should increase. Holding level of health status constant, it is hypothesized that women influence social insurance dependency to a higher degree than men (GENDER), because women may have a weaker position in the labor market than men do. However, it is unclear in which degree this difference will be observed. Higher education (EDUC) may contribute to higher level of professional state and better opportunities in the labor market. However, this is unclear for some parts of the population, particularly those younger than age 40. Education should be a kind of prevention to avoid social insurance dependency. Regardless, there are some indications, grounded in public debate and official statistics, young academics are unemployed while the age to enter early retirement has dramatically decreased recently.

Equation 3 documents the general hypothesis in labor economics that household income (INCOME) is positively correlated with age (AGE) and a higher professional level (PROF) and negatively correlated with social insurance dependency. It should be a surprise and simultaneously difficult to explain, if the type of correlation between the explanatory variables and the dependent variable should be different.

Results

Study findings are based on the sample of Swedish population reports gathered in 2005. These results (Table 1) confirm the importance of socioeconomic and behavioral factors in explaining health inequalities. The estimation of parameters in the HIS model clearly show most of the hypotheses are confirmed. All the parameters in Eq. 1 are statistically significant, indicating a high probability that the explanatory variables influence corresponds to health status in the estimation. Women seem to report a considerable lower level of health status than men, ceteris paribus. The same can be concluded for the other variables; all parameters existence clearly affects health.
Table 1 Results of estimates of the HSI-model

Variables      HEALTH (Eq.1)     SOCINSUR (Eq.2)      INCOME (Eq.3)

           Parameter  P Value  Parameter  P Value  Parameter  P Value

Constant     .86       .0001     8.9       .0001    3.8        .0001
AGE         -.001      .0009                         .028      .003
GENDER      -.046      .0002     -.15      .10
PROF                                                 .511      .0005
EDUC         .008      .0100      .056     .02
SMOKE       -.006      .0010
EXERCISE     .008      .005
HEALTH                          -9.8       .0001
SOCINSUR                                           -1.18       .0001

System weighted [R.sup.2]:.34


Poor health is very important in explaining social insurance dependency. Labor market inability to achieve a reasonable level of stable employment produces negative health and, thus, contributes to a decrease in health capital. The level of education is positively correlated with social insurance dependency, which may reflect the high level of unemployed younger people with higher levels of education. This HSI model cannot affirmatively answer to the question of a strong relationship between gender and social insurance dependency. Parameters are statistically significant to at least the 2% level (P value) with the exception of gender which is significant at the 10% level.

Equation 3 does not create any surprises. It may be seen, to a certain degree, as a validation test of the data used in this paper. Income seems to increase with age and level of professional skills, which is known in the human capital research field, but this relationship may be offset when the degree social insurance dependence increases.

Specific study results are summarized with the estimation of parameters, described below.

1. health status is higher among the following classifications:

(a) men,

(b) lower age groups,

(c) individuals who smoke less,

(d) individuals who exercise more,

(e) individuals in higher educational levels;

2. social insurance dependency is higher among the following classifications:

(a) individuals with lower level of health status,

(b) individuals with higher levels of education,

(c) women, probably;

3. income is higher among the following classifications:

(a) older individuals (increasing age),

(b) individuals with higher professional level,

(c) individuals with lower social insurance dependency.

Discussion and Key Messages

The results of the estimation based of Swedish data confirm that health inequalities are grounded by education, gender, age, and habits. On the other hand, poor health is a fundamental argument for explaining social insurance dependency. These two aspects are important when discussing how results in economic research may influence public policies in health and resource allocation.

Social insurance assists those old, in poverty, with disabilities or unemployed in correcting market failures. The success of achieving those goals has been considerable in Sweden. However, transfer policies are also failing, creating a dependency of social insurance which may be considered, to some extend, some of the negative external effects of the policy programs. The results in this paper show a correlation between social insurance and low income. This question is particularly important because more than 20% of the potential working force is dependent on the social insurance transfer payments to live.

When considering the importance of labor market and individual behavior in improving health status and decreasing health inequalities, these two aspects cannot be separated: social groups who report lowest levels in health status are the same as those who possess weakest conditions in labor markets and any successful policy that aims to increase health capital in society has to positively influence the social conditions of weaker groups, including labor market conditions, thus reducing health inequalities. It is difficult to determine which policies are more cost effective to achieve these two goals.

The results in this study may show, in order to be efficient, one must consider the indirect effects on health capital. Hauk et al. (2002) suggest variations in health production functions must be considered when addressing variations in healthcare quality and health services utilization. Individuals may share the same health problems and use the same level and quality of health services, yet their health outcomes may vary. This variation is due to the determinants of health which are important in explaining different health production functions. Perhaps, not all health inequalities are avoidable; however, much work has yet to be done for policy developments that may reduce avoidable health inequalities and improve health capital.

Health services do not control differences in health between individuals who are dependent on social economic environment and lifestyle choices. If the goal in a society is more health capital and less difference in health among individuals, the focus on health policy must shift outside of healthcare to include areas of resource distribution formulas by assigning larger weights to efficient health and investments in areas such as the labor market, housing and social networks.

References

Brooks, R., Rabin, R., & Charro, F. (Eds.). (2003). Measurement and valuation of health status using EQ-5D: A European perspective. Netherlands: Kluwer.

Chamberlain, G. (1984). Panel data. In J. Eckman, & E. Leamer (Eds.), Handbook of econometrics (vol. II, (pp. 1247-1318)). Amsterdam: North-Holland.

Dolan, P., Gudex, C., Kind, P., & Williams, A. (1996). The time trade-off: results from a general population study. Health Economics, 5, 141-154.

Ferraz-Nunes, J. (1996). The limits of the welfare state. Advances in Economic Research, 2(4), 434-443.

Ferraz-Nunes, J. (2006). Samre Halsa for Kvinnor och Arbetare. In L. Weibull (Ed.), Du Stora Nya Varld (pp. 231-240). Sweden: SOM-Rapport nr 29 Goteborg University.

Fuchs, V. (1982). Time preference and health. In V. Fuchs (Ed.), Economic aspects of health (pp. 93-120). Chicago: The University of Chicago Press.

Grossman, M. (1972). On the concept of health capital and the demand for health. Journal of Politicla Economy, 80, 223-255.

Hauk, K., Shaw, R., & Smith, P. (2002). Reducing avoidable inequalities in health: a new criterion for setting health care capitation payments. Health Economics, 11, 667-677.

Jones, A., & Rice, N. (2005). Using longitudinal data to investigate socioeconomic inequality in health. In P. Smith, & L. Ginelly (Eds.), Health policy and economics (pp. 88-120). Berkshire, England: McGraw-Hill.

Kaplan, R. (1996). Utility assessment for estimating quality-adjusted life years. In F. Sloan (Ed.), Valuing health care (pp. 31-59). New York USA: CUP.

Kenkel, D., & Decicca, P. (2000). Putting out the fires: Will higher taxes reduce youth to smoke? Working Paper 00-3, Aarhus School of Business, Denmark.

Lleras-Muney, A. (2002). The relationship between education and adult mortality in the USA. Working Paper 8986, National Bureau of Economic Research, Washington.

Williams, A. (1995). The measurement and valuation of health. Discussion Paper 136, Centre for Health Economics, York, UK.

J. Ferraz Nunes

Goteborgs universitet, Gothenburg, Sweden

e-mail: jose.ferraz-nunes@socmed.gu.se
COPYRIGHT 2008 Atlantic Economic Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2008 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Nunes, Jose Ferraz
Publication:International Advances in Economic Research
Article Type:Report
Geographic Code:4EUSW
Date:Aug 1, 2008
Words:2625
Previous Article:The Spanish tomato export sector of the Almeria region: an econometric approach.
Next Article:Estimating the percentage of the US population without health insurance: an ecological approach.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters