Contact:
Professor Margaret Whitehead
Tel/Fax: +44 (0)1948 664529
Click to email

Funded by:
The National Institute of Public Health and
The Swedish Council of Social Research.

British Researchers:
Professor Margaret Whitehead
Professor Ken Judge
Ms. Michaela Benzeval
Ms. Susanna Shouls

Duration of Research:
January 1997 - December 1998

Swedish Researchers:
Professor Finn Diderichsen
Professor Sven E O Hort
Dr. Maria Danielsson
Department of Public
Health Sciences
Karolinska Institute
S-172 83 Sundbyberg
Sweden

Background
Evidence about the existence and causes of health inequalities is both universal and compelling. In contrast, knowledge about the most cost-effective ways of reducing them or of ameliorating their effects is much more scarce. The purpose of this Anglo-Swedish collaborative study is to redress this imbalance in scientific understanding. It aims to do so by identifying ways in which social welfare policies influence the health-related circumstances of unemployed people and lone parents in Britain and Sweden.

The primary focus is on social welfare policies because of their potential impact on health inequalities. While there is little research evaluating the impact of specific health care interventions, there is a virtual absence of studies assessing the impact of social welfare policies. Unemployed people and lone parents have been chosen as groups with clear evidence of health deficits compared to the population as a whole and because many of them are dependent on social welfare. Britain and Sweden have been selected for comparative analysis because both have experienced substantial changes in economic and social policies during recent years that might be expected to have an impact on health inequalities. At the same time, the two countries differ in terms of their social coherence and welfare cultures and in the severity of the health inequalities they experience. These circumstances and characteristics create the possibility of identifying and evaluating a 'natural experiment' with real potential for policy learning.

Aims and Objectives
The project has three aims:

• to develop appropriate methods for cross-national comparative policy analysis in relation to health inequalities;
• to use these methods to evaluate ways in which social welfare policies ameliorate and/or exacerbate the health and related circumstances of disadvantaged groups;
• to assess the wider implications for policy development in Britain and Sweden.

Study Design
The study employs both quantitative and qualitative methods, carried out by British and Swedish researchers in a collaborative initiative. First, household survey data about adults aged 16-64 will facilitate comparisons of the health status and the living conditions of unemployed people and lone parents in both Britain and Sweden. Changes will be assessed between countries and over time from 1979 to 1994 using the General Household Survey in Britain and the Survey of Living Conditions in Sweden, which are broadly comparable. Secondly, these data will be used to identify those living conditions potentially amenable to policy interventions that are associated with differences in health inequalities. Finally, a combination of analyses of official documents and statistics and interviews with key informants will be employed to monitor changes in the economic and social policies that are thought to influence the health of unemployed people and lone parents. The more ambitious goal is to assess the extent to which differences in both the substantive content of social welfare policies and users' experiences of them might help to account for variations in health inequalities between the countries and/or over time.

Policy Implications
In both Britain and Sweden, there is a growing concern about health inequalities and emerging signs that policy makers are looking for guidance from the research community about the most effective ways in which public policies could be modified to ameliorate them. This project will not provide definitive answers, but it will provide informed guidance to those with policy responsibilities and influence who are persuaded of the need to develop a practical agenda to reduce social variations in health. A central goal of the project therefore is to establish good lines of communication with, and to disseminate findings to, as many policy and practitioner communities as possible in both Britain and Sweden and beyond.

Project Summary
The UK government, like those elsewhere in Europe, is looking to the scientific community for methodologies through which to assess the health impact of policies to tackle health inequalities. But the complex issues involved in policy-impact analysis are holding back the development of appropriate methodologies.

The project aimed to improve the methodology for investigating the impact of social welfare policies on health inequalities and to test this improved methodology in a comparative study of Britain and Sweden. It was a collaborative undertaking, between UK researchers and researchers at the Karolinska Institute in Stockholm. The methodology for policy impact analysis built on earlier work by one of the co-applicants, Finn Diderichsen. His model mapped the points where policy interventions could be targeted to break into the causal chain running from the social structure to health inequalities. The model was used to analyse social welfare policies and socio-economic conditions in Britain and Sweden from 1979 to 1995/96 and examine their consequences for the health and social circumstances of two vulnerable groups: lone parents and those at risk of and experiencing unemployment.

Key findings

• The project developed and refined a conceptual framework for studying the impact of policy on inequalities in health, which encompasses the social context as well as individual-level pathways. The framework identifies four key policy entry points:

1. policies which target the determinants of an individual's social position (e.g. policies to tackle disadvantage in childhood and low educational achievement);
2. policies which influence differential exposure to the risk factors associated with social position (e.g. by breaking the link between unemployment/lone parenthood and poor housing and low income);
3. policies which influence the differential effects of being exposed to the risk factors (because the size of the effect of a risk factor is often amplified by the presence of other risk factors and conditions);
4. policies influencing the differential consequences of ill health (e.g. through rehabilitative and community care services, through workplace policies and welfare benefits for those with long-term illnesses).

• A statistical methodology was developed so the framework could be used to analyse policy effects in countries with contrasting welfare systems. The methodology is based on the sufficient-component cause model developed within epidemiology. This makes the joint effect of multiple exposures empirically testable with a meaningful etiological interpretation. It may be used both for empirical analysis of natural experiments (as in this study) and as a framework for using existing empirical findings for health inequality impact assessment.
• New findings on the health of lone mothers. Although the inequalities in health between lone and couple mothers are of a similar magnitude in Britain and Sweden, the pathways leading to the observed health disadvantage of lone mothers are very different in the two countries. Overall, in Britain around 50% of the health disadvantage of lone mothers is accounted for by the mediating factors of poverty and joblessness, whereas in Sweden these factors only account for between 3% and 13% of the health gap. These can be traced to different policy interventions - a finding with implications for strategies to achieve the government's targets for reducing health inequalities.
• New findings on the social consequences of chronic sickness. There were higher employment levels, and lower economic inactivity levels, in Sweden than in Britain both for those with and without a limiting long-term illness. There were also greater socio-economic differentials in inactivity rates among men with limiting long-term illnesses in Britain. The evidence suggested that Swedish employment protection and social security policies to help those with chronic illnesses get and keep jobs were more effective than those in Britain.
• The conceptual framework has been taken up by the Rockerfeller Foundation and Swedish International Development Authority as a helpful tool for developing countries and has already been used within the Global Health Equity Initiative. It has also been used by Swedish national agencies (the National Public Health Commission, the National Public Health Institute and the Federation of County Councils) in their work on health inequality impact assessment.