Health inequalities – a challenge for the social investment welfare state

Social investment policy has become a central response to the demographic and economic challenges facing European welfare states. This focus on investment in human capabilities and their efficient use is, however, challenged by health inequalities where education, health and employment are increasingly linked. This paper outlines the main principles of social investment policies (learning, activation and protection) and links them to a conceptual model of health inequalities and the policy entry-points tackling them by addressing the processes of social stratification, differential exposure and vulnerability as well as differential consequences of illness. It illustrates, with reference to selected empirical studies from the Nordic countries, how the balance between the elements of social investment policies might be adjusted, resources allocated differently and policies supplemented by more direct investments in health so as to enable social investments to tackle the health divide.


INTRODUCTION
Social investment welfare policies are developed to sustain an economy where knowledge is considered a major driver of productivity and economic growth (Esping- Andersen et al., 2002).The knowledge-based economy thus rests on a skilled, flexible and healthy labor force, which can easily adapt to the constantly changing needs of the economy but also be the motor of those changes (Morel et al., 2012).Investment in education and health is, however, challenged by growing inequalities in health where education, health and therefore also employment are increasingly linked to each other.Health has not been a major issue in the literature on social investment policies, and this paper looks at how an analysis of mechanisms generating health inequalities might identify policy entry-points where social investment policies can be adjusted to tackle the health divide.
European welfare states are also facing demographic changes with an ageing population due to low mortality and fertility, single parenthood, etc.Those out of paid work that previously primarily were homeworking are now on various types of social benefits (Dølvik, 2015).
Facilitating access to the labor market by groups that traditionally have been left outside is therefore a priority in many countries.

SOCIAL INVESTMENT POLICIES
Already in the late 1990s -before the most recent economic crisis -several attempts had been made to redefine the principles, goals and instruments of the welfare state.A common thinking of these attempts has been to develop policies that are more "preparing than repairing" the workforce facing the rising demands, and transforming social policies from being mainly passive and "compensatory" to be more active and preventive.The notion of social investment has been used as a unifying term to describe these many different attempts (Morel et al., 2012).
This social investment perspective is often described as based on three principles.First is the principle of learning as a pillar on which the economy and development of societies can be built in the future.This principle leads to policy attention to human capital development, starting with focus on early child development among pre-school children and on equal opportunity and high quality of school education.This leads to the promotion of social spending designed to break the intergenerational cycle of poverty and increase social mobility.Day care for preschool children that also low-income single parents can afford, and schools of good quality even in underprivileged neighborhoods is regarded as essential elements.Lifelong learning and retraining for adults with jobs and for those in need of a new job is another part of this principle.
Secondly, there is a principle of activation among adults with problems in the labor market.This involves active labor market policies and vocational rehabilitation, including programs to restore workability by improving physical and mental functioning, and finding better matches between the functional limitations of employees and job demands.However, this policy often entails cuts in unemployment benefits, and tax benefits for employed, motivated by the principle of strengthening economic incentives -in order to "make working pay".
Thirdly, there is a principle of protection to avoid poverty and social exclusion.A constantly changing economy demands a very flexible, mobile and adaptive labor force, but that is difficult to achieve without a high level of security when leaving a job -a policy of "flexicurity".Protection shall also ensure a low level of inequality as inequality may be an obstacle for economic growth in addition to the many ethical arguments against it.
These principles for a social policy as a productive factor and investment rather than a spending have been embraced and promoted by international organizations like OECD and EU, most prominently by the European Council in the Lisbon Agenda in 2000.The reality, however, in many European countries is still a social policy mainly pursuing passive compensating insurance and service provision (Morel et al., 2012) often combined with post-crisis austerity that has turned out in some cases to have devastating effects on equality, growth -and on population health (Karaneikolos et al., 2013).
Criticisms of the social investment approach has been raised by authors like Cantillon (2013) and Pentilon (2013) arguing that "social risks that are likely to induce poverty (for example, unemployment, illness and disability, early school dropout, low levels of education and divorce) are socially stratified within all welfare states, including the Scandinavian".If these persisting inequalities are not taken into account, the social investment policies may fail in their ambition to enhance growth and limit inequalities.Many authors have therefore emphasized that social investment welfare policies shall be seen as package where the balance between the three elements is critical (Morel et al., 2012).

INEQUALITY AND GROWTH -IN ECONOMY AND IN HEALTH
The Nordic countries, Netherlands and to some extent even the UK are countries where social investment policies have been pursued more than elsewhere, but with large differences in the balance between the three elements of learning, activation and protection.The Nordic countries have compared to other countries put strong emphasis on learning and protection.Universal educational systems of high quality as well as a high level of flexibility, mobility and security on the labor market were for many years probably essential for ensuring the unique combination of economic growth and a relatively low (albeit recently increasing) level of economic inequality (Dølvik, 2015).See figure 1a.This article is downloaded from www.idunn.no.© 2016 Author(s).This is an Open Access article distributed under the terms of the Creative Commons CC-BY-NC 4.0 License (http://creativecommons.org/licenses/by-nc/4.0/).Sweden in particular has a long tradition of human investment policies dating back to Alva and Gunnar Myrdal's writings in the 1930s on how to improve the quantity and quality of the population (Myrdal & Myrdal, 1934).A renewed interest in recent years, now in the shape of local social investment funds in Swedish municipalities, has focused on preventing young people dropping out and failing in school, and on the labour market (Hultkrantz, 2015).
Several studies have shown that the Nordic countries are not, as some economic theories would argue, exceptions from a rule that inequality is necessary for growth, but that the opposite relationships exist -that equality boost growth (Ostry et al., 2014).The mechanisms behind that relationship may not be fully understood, but education and other forms for social investment in human capital seem to be a critical mediating link (Dølvik, 2015;Ostry, 2014).Investing not only in education but also in health has more recently been introduced in the discussion about social investment policies in Europe (McKee, 2010;MacDaid et al., 2015;WHO, 2015).

EDUCATION, HEALTH AND EMPLOYMENT ARE INTERDEPENDENT
It is a central finding and forceful argument in favor of social investment policies that education has a strong impact on employment chances.Employment in jobs with a high level of skill discretion will in addition develop new skills and it has been shown to benefit health as well.Good mental health and workability is necessary for fulfilling mental and social demands in today's labor market and therefore has a strong impact on employability.Education, employment and health are therefore increasingly interdependent.See figure 2.
This article is downloaded from www.idunn.no.© 2016 Author(s).This is an Open Access article distributed under the terms of the Creative Commons CC-BY-NC 4.0 License (http://creativecommons.org/licenses/by-nc/4.0/).

THE NORDIC WELFARE PARADOX OF HEALTH INEQUALITIES
Social inequalities in health is about the strong linkage between education and health -two factors with strong influence on employability.If the causal relationships between education, health and employment develop into a virtuous circle it will be important for the ambition of combining economic growth with limited economic inequality.A social investment welfare state is therefore challenged by inequalities in health and has to avoid that the triangle develops into a vicious circle of growing health inequalities.Numerous studies conducted over the last 40 years have documented, for some health outcomes and periods, large growing social inequalities in health in Europe, including the Nordic countries (Marmot, 2013;Diderichsen et al., 2012;Dahl et al., 2014).So while Figure 1a illustrated that it has been possible to combine equality and growth in economy, figure 1b illustrates that the same is not true for another important welfare dimension: health.Figure Figure 2 illustrates the virtuous -or vicious -circle between education, health and employment (arrows represent effects).The figure also illustrates how the three policy elements of learning, activation and protection might interfere in this process.The "learning" element is about using resources for human capital development by providing education and training that might feedback positively into the system, since a well-educated population is a driver of social and economic development.Societies with large inequalities will have difficulties in optimizing educational policies (Ostry, 2014).Using resources for "activation" promotes employment that will contribute to individual and collective resources."Protection" protects against poverty and social exclusion when employment opportunities are limited for various reasons, such as a lack of available jobs or limited workability.Labor market policies play an important role here (Reeves, 2014;McAllister, 2015).In figure 2, we have added a fourth investment-policy element: investment in health for reasons described below.
This article is downloaded from www.idunn.no.© 2016 Author(s).This is an Open Access article distributed under the terms of the Creative Commons CC-BY-NC 4.0 License (http://creativecommons.org/licenses/by-nc/4.0/).1b shows that over five decades there has been a gradually stronger association between education and mortality with Norway 1 as an example.The fact that the Nordic countries show inequalities in morbidity and mortality that are equal to, or larger than, England and Southern Europe, which have much larger social and economic inequalities, is striking in view of the many years of social investment policies and relatively free and equal access to health care (Mackenbach, 2012).Other studies have shown that in particular mental morbidity among those out of work is increasing more than among employed (Andersen et al., 2016;Blomquist et al., 2014), which indicates that the association between being out of work and ill health is getting stronger.
That is a challenge for the social investment policies for many reasons.The efforts of raising educational levels in schools have met obstacles in terms of raising mental problems among school children.Children with weak cognitive development when starting school will have difficulties when facing learning demands and that will in turn increase the risk of mental disorders that further increase difficulties in meeting the academic demands (Gustafsson et al., 2010).Rehabilitation studies have shown that the chance of getting back to work after a period of illness is strongly influenced by the level of education (Thielen et al., 2015;MacAllister, 2015).The added effects of low education and health problems (in particular mental disorders) on workability may represent an overwhelming obstacle for returning to work.

THE POLICY ENTRY POINTS TO TACKLE THE HEALTH DIVIDE
The fact that those countries who have implemented social investments policies for a longer time and more comprehensively than other countries still have large and in some cases growing health inequalities raises some questions: Are there inherent problems in the social investment policy that does not take care of health inequalities, or are imbalances between the three policy elements and their implementation responsible?Do health inequalities constitute an obstacle that interferes with harvesting the full benefits of a social investment policy?Are there other effective policies that have been implemented (sufficiently)?
To be able to address these questions it is necessary to disentangle the different mechanisms and policy entry points by which social policies can influence health inequalities.
Health inequalities are generated by mechanisms where human development and social position through mediating and interacting causes are linked to health outcomes (Diderichsen et al., 2012).Figure 3 illustrates the most important mechanisms (I-V), and four policy entry points (A-D).Arrows indicate cause-effect relationships that however might be modified both by individual factors such as social position and by contextual conditions created by macroeconomic conditions and policies.
1. Inequality in life expectancy can also be measured by a Gini-coefficient.But since length of life have upper limits (very few are getting older than 105 years) declining Gini-coefficient of length of life will nearly always follow an increasing life expectancy and is therefore of limited interest.The socioeconomic differences in mortality and morbidity are what have attracted both scientific and political interest.
This article is downloaded from www.idunn.no.© 2016 Author(s).This is an Open Access article distributed under the terms of the Creative Commons CC-BY-NC 4.0 License (http://creativecommons.org/licenses/by-nc/4.0/).A: Policy entry point (A) acts on the effect that early childhood development in terms of cognitive, emotional and social development have on the social position on the labor market an individual will attain as an adult, i.e. social stratification -(arrow I in figure 3).This is where the learning element of social investment policies work.Access to and quality in preschool child care and the quality of compulsory school education can modify some of the effects of genes and parents' resources, but to what extent that will happen depends strongly on the financing, quality and accessibility of those institutions.Children growing up with fewer resources due to low-educated, unemployed or sick parents tend to cluster in low income neighborhoods.In many cities, this social segregation is growing, and the ongoing privatization of schools, for instance, in Sweden and Denmark, threatens to make segregation between children from resourceful and less privileged background even stronger.If not resource allocation to daycare centers and schools are made proportional with variations in need for high quality education, the social investment function to secure equal opportunities and labor market preparedness cannot be fulfilled.The structural fra- mes within which the educational efforts can be done is therefore of major importance (Stigendal & Östergren, 2013).
Even when a high level of equality of opportunity is ensured, and policies promoting a high degree of social mobility are in place, they do not ensure equality of outcome in terms of employment, living conditions and health.Many children and adolescents will not be successful in education and will not have the necessary qualifications to compete for wellpaid healthy jobs, and the subsequent life conditions that are decisive for their health.How their health will develop will therefor depend on actions in the second policy entry point (B): B: A large proportion of inequalities in health are generated by differential exposure to causes of disease in living and working conditions as well as in health behaviors (arrow II in figure 3).Social policies have a strong impact here as they influence income levels among families, child poverty, and housing standards that can be afforded (Stigendal & Østergren, 2013).Income levels also have an impact on health behaviors, in particular, dietary patterns and obesity, but economic stress also influences smoking and alcohol consumption.More than half of the social inequalities in mortality in Denmark and Norway are, for example, related to tobacco and alcohol use (Strand, 2014;Diderichsen, 2012), and recent studies have shown that access to unhealthy consumption and drugs are important mediators of the health effects of social risks generated by economic crisis and austerity (Karaneikolos, 2013).Society's ability to protect the population against the marketing and health effects of unhealthy diets, tobacco and alcohol is essential for preventing health effects of adverse socioeconomic conditions.The protection element of social investment policies are thus not only about protection against poverty and social exclusion but also about protection against health risks in consumption and environmental risks.Without this aspect the potential of social investment policies for employment and economic growth will not be fully exploited.The positive news is that a large amount of research has shown that many feasible public health policies are highly cost-effective (McDaid et al., 2015) and many have a differential effectiveness that benefit people with short education and low income (Diderichsen et al., 2012).
C: The strength of the effect on health of being exposed to a cause of disease depends on the exposure to other interacting causes.That is why some people are more vulnerable -or less resilient -than others to the effects of adverse conditions.People with low education tend to be exposed to many risk factors over their life course like child poverty, the stress of unemployment and poverty in adulthood, unhealthy living with smoking and obesity, etc.Some of these risk factors interact with each other -i.e. the health effect of one cause is stronger in the presence of another cause (Nordahl et al., 2014).Stress and high blood pressure make people more vulnerable to the effect of smoking, adverse childhood conditions make people more vulnerable to crises later in life, etc.This clustering and interaction of risk factors among low educated and unemployed generates a differential vulnerability (III in figure 3).Social investment policies and its protection element play in this context an important role, for example, by reducing adverse conditions for pregnant mothers and children (e.g.reducing child poverty, promoting early child development), that may reduce vulnerability to the health effects of exposures later in life.The Nordic countries have indeed achieved low levels of child poverty and high levels of child care coverage, but inequalities in quality of child care may still exist.
This article is downloaded from www.idunn.no.© 2016 Author(s).This is an Open Access article distributed under the terms of the Creative Commons CC-BY-NC 4.0 License (http://creativecommons.org/licenses/by-nc/4.0/).D: Health inequalities are not only about different risks of getting ill, but also about different consequences of being ill in terms of survival, disability and participation -i.e.differential consequences (IV in figure 3).Social investment policies -both the activation and protection dimension are relevant here.Activation policies do not only involve instruments for retraining and education as well as economic incentives but also, as earlier mentioned, rehabilitation of functional abilities as well as adapting working demands to fit functional limitations, and thereby achieve better workability.Positive results of these efforts are, however, not striking when looking at the aggregate figures and trends.Many countries including the Nordic countries and UK have over decades experienced increasing numbers of people leaving the labor market because of reduced workability (MacAllister et al., 2015).The evidence of what are cost-effective methods of rehabilitation is also rather limited (Howard-Wilsher, 2016) and the health care structure for long term rehabilitation demands close collaboration between primary care, hospitals, municipalities, employers etc.But many health care reforms, marketization and new economic incentives have promoted a fragmentation rather than stronger collaboration within health care systems.Care for those with chronic illnesses and often multi-morbidities have been suffering from those changes (Dahlgren, 2014).
Psychiatric diagnoses are prevalent among people who have been granted disability pensions in recent years.Eligibility criteria for disability pension have been made stricter in many countries, which means that an increasing number of people are not eligible, but, on the other hand, neither are they able to meet the demands of the labor market.Prevalence of common mental disorders increases therefore among people on non-health related means tested social benefits and unemployment benefits (Andersen, 2016).The very controversial WCA-reform in England, where disability pensions and other benefits that have been granted can now be withdrawn after reassessment of disability seems only to have resulted in few returning to work, but worsening mental health for many (Barret et al., 2016).
The protection dimension is finally relevant due to the fact that ill health, depending on policies pursued, may lead to marginalization, social exclusion and poverty.It is in particular serious mental and addiction disorders that tend to have these serious long-term social consequences.But even for a much larger group of people with different kinds of limiting illnesses it has, at least in Sweden, been found that the cuts in benefits and protection have increased the poverty rates in those groups (Falck et al., 2013).

CONCLUSION
The sustainability of modern welfare states depends on a high employment rate and strong human capital i.e. a large, well-educated and healthy workforce.Social inequalities in health is about the strong linkage between education, health and employment-a triangle in which positive causal relations are crucial for combining economic growth with limited economic inequality.A social investment welfare state is therefore challenged by inequalities in health and has to avoid that the triangle develops into a vicious circle of growing health inequalities.The above analysis has led us to the following conclusions for how the three elements of social investment policies can contribute to tackling the health divide: This article is downloaded from www.idunn.no.© 2016 Author(s).This is an Open Access article distributed under the terms of the Creative Commons CC-BY-NC 4.0 License (http://creativecommons.org/licenses/by-nc/4.0/).
The learning element is crucial for creating equality of opportunities but to what extent it succeeds depends on whether resources and quality of institutions to support early child development and school education are allocated according to need.Interventions to promote cognitive development in preschools and tackle common mental problems among school children and adolescents are important.
Creating equal opportunities and social mobility does not ensure equal outcomes and population groups that fail in an increasingly meritocratic system might face adverse conditions in periods of growing inequalities.Economic redistribution is therefore essential to prevent health inequalities.
The policy of protection is essential to prevent child poverty and the increased vulnerability that creates increased health effects of later exposures.The increased child poverty among risk groups such as non-western immigrants should therefore be prevented.
The increased focus on activation policies including ambitions to "make working pay" should be balanced to avoid growing income inequalities.A growing number of people with mental problems seem not to qualify for disability benefits due to stricter eligibility criteria, but, they have, on the other hand, difficulties meeting the demands of the labor market, and will often end up with means-tested benefits.Access to more flexible work demands is needed for this group.
These points illustrate that a careful balance between the three policy elements of learning, activation and protection is important for health equity.
The growing mortality differentials in the Nordic countries illustrates that without a strong protection against the health-damaging consumption of tobacco, alcohol and drugs the health effects of adverse social conditions are much stronger, in particular, among more vulnerable low educated or unemployed groups.

Figure 1
Figure 1 b): Life expectancy at birth (years) and difference between and low and high educated in mortality (deaths per 10 000 age adjusted).Norway 1960-2010.Source: Strand, 2015.

Figure 2 :
Figure 2: Social investment policies (learning, activation and protection) can be illustrated by its effect on virtuous circle of education, health and employment.