An inquiry into the relationship between education and health

By Anne Franklin,2014-08-17 16:33
7 views 0
An inquiry into the relationship between education and health

    2008 Social Capital Global Network

    Workshop on Social Capital and Health

    Draft abstract

    10-11 October 2008 Paris

    Institute for research and information in health economics

IRDES 10 rue Vauvenargues 75018 Paris Tél. : +33(0)1 53 93 43 00 Fax : +33(0)1 53 93 4350

    - 2 -

Karen S. Cook, (Stanford University)

Irena Stepanikova, (University of South Carolina)


    In light of the proliferating public and academic discourse about the benefits of trust, an empirically grounded approach is needed to evaluate evidence on trust and its outcomes. This study examines evidence on the role of trust in health care access, utilization and quality. We focus on interpersonal trust between individual users and providers of health care and on generalized forms of trust in physicians and medical institutions. We conducted our empirical review by searching academic databases for qualitative studies and quantitative studies using statistical controls that reported a relationship between trust and one or more health care outcomes. Most of the studies yielded by our review addressed interpersonal trust between a physician and a patient. Evidence was especially strong for linkages between interpersonal trust and the quality of communication, patient satisfaction, adherence, access to services, and health status. Several studies in this category were prospective, suggesting potential causal relationships between trust and its outcomes. Evidence on health care outcomes of institutional trust was mainly cross-sectional, making it less clear whether aspects of health care linked to institutional trust are its consequences or its antecedents. Evidence on general trust in physicians was presented in relatively few studies. We propose a model of trust and health care outcomes based on our review of the published results and discuss potential reasons for the existence of linkages in this model.

2008 Social Capital Global Network II Workshop on Social Capital and Health Paris 10-11 October 2008 IRDES - OECD

    - 3 -

Eline Aas mailto:

Tor Iversen mailto:


    In this paper we want to analyse the impact of socio-economic status, through health related quality of life and survival on the intensity of treatment of colorectal cancer treatment. Given the stage of advancement, there are two opposite effects: Individuals with a high level of education live longer after the time of diagnosis. Increased survival implies the opportunity to receive more treatment, thus the cost of treatment will increase. On the other hand, if low socio-economic status is positively related with general health condition, we would expect an increase in the intensity of treatment; like number of re-hospitalization, complications and co-morbidity. In this study we define socio-economic status by gender, education and labour market participation. In addition, marital status will be included as one measure of social capital.

2008 Social Capital Global Network II Workshop on Social Capital and Health Paris 10-11 October 2008 IRDES - OECD

    - 4 -

Paul Dourgnon, Senior research fellow, (IRDES) mailto:

Michel Grignon (McMaster University , Ontario) mailto:

Florence Jusot, Senior research fellow (University Paris-Dauphine, Legos and IRDES) mailto:



    The purpose of this paper is to investigate the link between country of origin and religion and access to care, as it is mediated via social capital.


    French and international literature and previous research show the followings:

    1. There is a clear link between social capital and health, even though any causal relationship

    is hardly proven. The link between Social Capital and Health services consumption has

    been rarely investigated, except by Laporte et. Al. (Laporte, Nauenberg, Shen, 2007) who

    that Social Capital had a significant impact on GP services utilization.

    2. On average, immigrants have less access to social capital resources through social or

    linguistic isolation or access specific community social capital with differential returns (Kao,

    Lindsay, 2007)

    3. Religion and country of origin (where parents were born) seems to play an important role in

    access to social capital among immigrant communities and visible minorities notably as a

    source of shared values.

    4. There is a link between immigration status and health (Jusot, Silva, Dourgnon, Sermet,

    2007), but it is complex and multi-directional: country of origin socioeconomic and sanitary

    situation effect; migration selection effect (the healthy immigrant effect means that on

    average immigrants are in better health when they arrive in the country) ,socioeconomic

    situation and isolation in the arrival country effect. These can confound the impact of living

    in a foreign environment with lower access to social capital.

    5. Ethnological approaches show heterogeneity across religious groups in the same country

    in utilization of health services in France (Faizang, 2001).

2008 Social Capital Global Network II Workshop on Social Capital and Health Paris 10-11 October 2008 IRDES - OECD

    - 5 -

    As a result, we will investigate the simultaneous impact of origin, religion, and social capital on utilization of health care services.

Data and methods

    ; Data

    The analysis will be based on a unique dataset, the 2006 wave of the Survey on Health, Health Care and Insurance run in France. It is a general population survey linked to administrative files from the main national sickness funds, comprised of exhaustive information on reimbursed health services consumption within the year.

Our analysis will use the following data :

- Social Capital measured by membership in informal organizations;

    - Religion (religious affiliation, frequency of religious attendance);

    - National origin and migration status: the data allow to distinguish new immigrants and

    ―second generation immigrants‖ meaning those born in France from migrant parents, and

    their countries of origin.

    - Health: Self Assessed Health, Lifestyle (alcohol, cigarette);

    - Health Care Consumption: inpatient and outpatient care (primary care, secondary care,

    preventive care);

    - Educational attainment;

    - Standard socio-demographic factors (occupation, income, age and gender).

; Analytic strategy

We will estimate simultaneously :

    ; how Social capital is determined by Origin and migration status, Religious affiliation,

    the type of neighbourhood, education, income, work status

    ; and How Health Care Consumption is affected by Social Capital, Health Status,

    Insurance Status, Education, Income.

    A simultaneous estimation strategy (two-stage least squares or probit) will allow taking into account the direct effect of origin and religion on utilization and its indirect effect (the portion that is mediated by social capital). We will address the following question: what is the share of social capital (and interactions between migrant or minority populations with the broader society) in heterogeneity of health care utilization across immigration or minority status. 2008 Social Capital Global Network II Workshop on Social Capital and Health Paris 10-11 October 2008 IRDES - OECD

    - 6 -

Audrey Laporte, Assistant Professor of Health Economics at the University of Toronto mailto:

    Eric Nauenberg, University of Toronto mailto:

    Leilei Shen, University of Toronto mailto:


    This paper provides results from ongoing research that seeks to better understand the dynamic between social capital and health care utilization. Initial cross-sectional (2001) results of our work in Canada suggests that Individual and community social capital operate differently with the former enabling better access to health care services while the latter serving as a substitute for these services. This begs the question of why these different levels of social capital behave very differently vis-à-vis health care utilization.

    To further this work, we have obtained longitudinal data through the Ontario Ministry of Health and Long Term Care for 2006 linked back to the initial (2001) Canadian Community Health Survey (CCHS waves 1.1 and 1.2) and Canadian Census responses on social capital. We estimate two-part models of utilization for annual GP visits, hospital emergency department visits, hospitalized days, home care visits, and prescriptions filled with special attention paid to the impact of community and individual social capital. We control for the effect of diet, substance abuse (smoking, alcohol consumption), immigrant status, community migration levels, baseline health status (# of chronic conditions), a regular source of primary care, income, education, and labour force participation. Individual social capital (ISC) is measured in two components: a four-point likert scale regarding how connected a person is to their community and on the frequency of religious service attendance. Community Social Capital (CSC) is measured at the metropolitan level using employment levels in religious and community-based organizations [NAICS code 813XX]a.k.a., Petris Index). In the next phase of research, we plan to engage the public through focus groups and questions inserted in population-based surveys to determine which factors (e.g., education) and issues mediate the relationship between social capital and health care utilization.

    The initial cross-sectional results showed that a 1% increase in the Petris CSC index produced a 2% decrease in overall GP visits with the impact concentrated in the population 65+, The impact from increased ISC was in the opposite direction and also smaller in magnitude than for CSC. The results of the quantile regression indicated that the effect of CSC is most prominent in the middle ranges of utilization while that of ISC is mostly at the lower end. We anticipate that the longitudinal analysis will produce results with larger magnitudes but maintain the opposing relationships observed. The focus groups and survey work undertaken will then help answer why the two forms of SC behave differently and describe the different mechanisms through which each likely operates.

2008 Social Capital Global Network II Workshop on Social Capital and Health Paris 10-11 October 2008 IRDES - OECD

    - 7 -

Richard M. Scheffler, PhD mailto:

Brent Fulton, PhD mailto:

Timothy T. Brown, PhD mailto:



    Area-level social capital has been found in numerous studies to be related to the health of both adults and children. However, it is currently unknown whether the level of parental education modifies the relationship between social capital and children‘s health and it is unknown whether parental education moderates any effects that social capital may have on children‘s educational outcomes.


    To determine whether social capital is positively associated with the health and educational outcomes of children and whether these associations are moderated by parental education levels

Data and Methods

We use five waves of the nationally representative Early Childhood Longitudinal Study

    Kindergarten Class of 1998-1999 (ECLS-K), which includes 11,820 children who were followed between kindergarten and fifth grade. The data includes information on parental demographics, including educational level, as well as numerous health measures including general health status and behavioral health status. It also includes standardized mathematics and reading achievement scores. We use a first-differenced regression model to estimate the association between changes in social capital (interacted with parent‘s education) and changes in children‘s health and educational outcomes.


    The results of this study will inform the research community on how social capital is related to children‘s health and educational achievement, and whether this relationship is moderated by

    parental education levels. This information will provide a better understanding of the pathways through which social capital works.

    2008 Social Capital Global Network II Workshop on Social Capital and Health Paris 10-11 October 2008 IRDES - OECD

    - 8 -

Kelly Noonan, Rider University and National Bureau of Economic Research

Nancy Reichman, Robert Wood Johnson Medical School

Hope Corman, Rider University and National Bureau of Economic Research mailto:


    The purpose of this paper is to estimate the effects of poor infant health on parental social interactions and the effects of social interactions on the health of parents. Having a child with a serious health problem imposes significant time and monetary costs on the child‘s parents. One aspect of these costs is the potential reduction of social interactions. This is a loss since social interactions directly provide utility and have been shown to increase mental and physical health. Social interaction can also produce alternatives to market-purchased child care services and other forms of instrumental support that can be especially important for low-income families. In addition, the loss of social interactions by the parents may adversely affect their health. Prior studies of the effects of social interaction on health have been criticized for ignoring endogeneity of social interactions. In this paper, we will estimate IV models using poor infant health as an instrument. Poor infant health can be measured as a random event and if it is causal on social interaction it could be a good instrument for social interaction.

    The data that will be used come from the Fragile Families and Child Wellbeing survey (FFCWB). The FFCWB is a longitudinal survey of about 5,000 new births to mostly new parents in 20 U.S. cities during the years 1998-2000. Both the mothers and fathers are interviewed at the time of the birth and followed up 1, 3, and 5 years later. Poor infant health is measured in alternative ways to reflect both serious and random health problems. Social interactions include religious affiliation, frequency of religious attendance, and visiting with relatives, as well as participation in church groups, service clubs, political organizations, union or other work-related groups, community groups, and organizations working with children. In addition, there are data on whether parents voted in the 2000 election. Parents‘ mental and physical health are measured at all follow-up

    waves. Data on income, education and other demographics are included in the analyses.

2008 Social Capital Global Network II Workshop on Social Capital and Health Paris 10-11 October 2008 IRDES - OECD

    - 9 -

Bryony Hoskins, European Commission CRELL mailto:

Beatrice D'hombres European Commission CRELL mailto : Beatrice.D‘


     1In the European context Active Citizenship has been promoted with in the Education and

    Training Lisbon Strategy as a tool to support the continuation of democracy, human rights and greater social inclusion. In this regard CRELL has developed a composite indicator to measure to this phenomenon built from European Social Survey 2002 data (Hoskins et al 2006). This composite indicator was then used by the European Commission in the 2007 Progress Report on the Education and Training Lisbon Strategy (European Commission 2007). The results of this composite indicator showed that the levels of active citizenship are higher across the North of Europe and lowest within Eastern and Southern European Countries.

    Active citizenship has focused its emphasis on the benefit of participation towards social cohesion. We contribute to this debate by examining the correlation between active citizenship and well-being. More precisely, we aim to investigate whether active citizenship is a determinant of well-being, once we control for other country characteristics. To that end, we carry out a cross-country analysis and use the active citizenship indicator recently developed by Hoskins et al. (2006). We

    will attempt to deal with the endogeneity problem using instrumental variable estimators.

     1 Active citizenship is defined as, ‗Participation in civil society, community and/or political life, characterized by mutual respect and non-violence and in accordance with human rights and democracy‘. (Hoskins, 2006).

2008 Social Capital Global Network II Workshop on Social Capital and Health Paris 10-11 October 2008 IRDES - OECD

    - 10 -

     1, 2,Nicolas Sirven Research fellow Mailto:

     1, Thierry DebrandSenior Research fellow mailto:


    The empirical literature on social capital and health indicates that individuals involved in social activities (associations, clubs, etc.) suffer less from cognitive impairments than their counterparts. It is often assumed that involvement in voluntary associations has positive effects on respondents‘ mental health, thought it may also be that causality is reversed (healthier people are more prone to participate in social activities), or more assuredly, that both variables influence each other. In this study, we aimed at testing the causal influence of social participation on a form of mental health deprivation. An instrumental variable approach helped in distinguishing the direct effect of individuals‘ investment in social capital on the incidence of cognitive impairments. Using longitudinal micro-data from the two first waves of SHARE, we find evidence that ―becoming alone in the household‖ between 2004 and 2006 is potentially a good instrument to be used. It

    thus seems that a change in household structure modifies the incentives to join social groups, and does not significantly alter memory or fluency that could lead to cognitive deprivation. Accounting for incentives to join a social group (changes in household structure) suggests that the influence of social capital was under-estimated in the usual binomial model.

Key words: Social Capital, Cognitive Impairments, SHARE data

JEL Classification: I1, Z13

     1 IRDES, 10, rue Vauvenargues 75018 Paris (France) Tél/Fax : +33(0)1 53 93 43 64 2 CSC-VH1, St Edmund‘s College, University of Cambridge, CB1 OBN (England). Tel; +44(0)1223728763; Fax: +44(0)1223762822

2008 Social Capital Global Network II Workshop on Social Capital and Health Paris 10-11 October 2008 IRDES - OECD

Report this document

For any questions or suggestions please email