Lars Fredrik Andersson, Håkan A. Bengtsson, Stefan Fina, Bastian Heider Unequal Sweden Regional socio-economic disparities in Sweden

FRIEDRICH-EBERT-STIFTUNG – POLITICS FOR EUROPE Europe needs social democracy! Why do we need Europe? Can we demonstrate to European citizens the opportunities offered by social politics and a strong social democracy in Europe? This is the aim of the new Friedrich-Ebert-Stiftung project“Politics for Europe”. It shows that European integration can be done in a democratic, economic and socially balanced way and with a reliable foreign policy. The following issues will be particularly important: – Democratic Europe – Social and ecological transformation – Economic and social policy in Europe – Foreign and security policy in Europe We focus on these issues in our events and publications. We provide impetus and offer advice to decision-makers from politics and trade unions. Our aim is to drive the debate on the future of Europe forward and to develop specific proposals to shape central policy areas. With this publication series we want to engage you in the debate on the“Politics for Europe”! About this publication Internationally, Sweden is still associated with a growth model that secures a high level of social equality via a strong universal welfare state. However, this perception overlooks a distinct rise in inequalities over the past 30 years. Successive marketbased reforms and structural change have created distinct regional socioeconomic disparities. This leads to the observation that there a four Swedens. Inputs for overall growth and employment stem from a minority of regions. Municipalities in the lagging areas are increasingly under severe pressure to uphold attractive levels of high-quality public services for their populations. Neither previous ongoing economic growth nor central government transfers to municipalities and EU regional funds have helped to balance the regional disparities and inequalities. On the contrary, they may have deepened the divide. A new approach for regional development is needed. About the Authors Lars Fredrik Andersson is associate professor for economic history at the Swe‑ dish Umeå University. Håkan A. Bengtsson is director of the Swedish Arenagruppen to which the think tank Arena Idé belongs. Stefan Fina is a Geography Professor at RWTH Aachen University and Head of the Geoinformation and Monitoring section at ILS – Research Institute of Regional and Urban Development Dortmund. Bastian Heider is an economic geographer and co-manages the Geoinformation and Monitoring section at ILS – Research Institute of Regional and Urban Development Dortmund. Responsible for this publication within the FES Dr. Philipp Fink, Director FES Nordic Countries. Additional information and supplementary materials are published at https://fes.de/unequal-sweden With the financial support of the European Parliament. The present report does not represent the European Parliament's views.

Lars Fredrik Andersson, Håkan A. Bengtsson, Stefan Fina, Bastian Heider Unequal Sweden Regional socio-economic disparities in Sweden FOREWORD 2 1. THE INEQUALITY OF LIVING STANDARDS IN SWEDEN 5 1.1 Spatial divergence through structural change 5 1.2 What is spatial inequality about today? 7 2. SWEDEN TODAY: GROWING DISPARITIES IN AN AGEING COUNTRY 8 2.1 Methodology 8 2.2 Four Swedens 10 3. NEW POLICIES FOR EQUALITY OF LIVING CONDITIONS AND SOCIAL COHESION 15 3.1 Limitations of current spatial policy 15 3.2 Recommendations for local-based action 16 3.3 Recommended actions at national level 16 3.4 Conclusion 17 ANNEX A  Indicator documentation 19 ANNEX B  Methodological notes 20 ANNEX C  Indicator value ranges 21 Literature 22 List of figures and tables 23

FRIEDRICH-EBERT-STIFTUNG – POLITICS FOR EUROPE 2 FOREWORD Similar to the other Nordic countries, Sweden is still seen as the archetypal equal and consensual society with a strong and redistributive welfare state, a high degree of job security, low income and wealth inequality and a renowned high-tech export sector. It comes as no surprise that Sweden together with its Nordic neighbours ranks near the top of many lists that measure equality and quality of life in various countries. However, a closer look reveals an uneven picture of society and the economy in Sweden. As in other European countries, economic, digital and ecological transitions have led to structural change. The capability to adapt to these changes in society and the economy is unequally spread, with some regions benefiting from change and others falling behind. There is more and more evidence that social inequalities are increasingly linked to regional disparities. It appears to be difficult to fight inequalities without addressing the regional divide. The research shows that Sweden has developed an uneven path in terms of employment. Ca. 95 per cent of employment growth has taken place in metropolitan and university cities – areas responsible for most of the country’s growth. Hence, Sweden is an unequal country in which the inhabitants are often faced with different living circumstances based on their place of residence. The cluster analysis undertaken in this report by Stefan Fina and his team at the Research Institute for Regional and Urban Development(ILS) Dortmund in collaboration with our Swedish partners at Arena Idé shows that in terms of living conditions, economic indicators and social well-being, Sweden can be divided into four distinct regions, what we call the“Four Swedens”. Whilst the majority of people live in dynamic urban regions, more than 30 per cent of Swedes live in areas that do not reach the same level of welfare. Moreover, more than 1 in 10 Swedes live in areas categorised as peripheral with low ­population density, higher rates of unemployment and poverty, higher dependency ratios and lower income levels as well as a lower provision of public services. Although in comparison to other countries and societies these levels of inequalities may seem slight, disparities in Sweden are at the centre of the political debate and can be seen to contribute to a rising dissatisfaction with the country’s democratic, political, and social institutions and conventions. The success of the Swedish development model and the ability of public institutions to guarantee equal living standards

Foreword 3 and equal opportunities for individuals crucially depends on the way in which non-urban areas and small cities can be integrated into the development strategy. There is a vicious circle that needs to be diffused: on the one hand, some“forgotten areas” are falling behind in terms of economic activity, causing highly skilled people to move away. This in turn results in worsening infrastructure and public services. On the other hand, highly urbanised areas, where economic activity is concentrated, attract more and more people, resulting in increased competition for jobs, higher living and housing costs and higher risk of social exclusion. Can we think of a development model that offers equal opportunities and high standards of living regardless of one’s place of residence? The challenge we want to highlight, for national as well as for European policymakers is that it is impossible to provide opportunities and equality for all individuals regardless of their economic and social background, unless regional inequalities are addressed. The findings of this report underline the need to overhaul the economic and social policies on a national as well as on the EU level. The authors point to the importance of an equal level of welfare provision throughout the country. To achieve that they suggest changing the way regional disparities are thought of. Policies should be directed towards investing in people and not in administrative structures. Without the intervention of the public sector, no opportunities are going to be generated. It is not only a matter of placing a stronger emphasis on the needs of lagging regions, but also rather the need to understand economic development as sustainable over time only if all areas develop and attain higher levels of well-being. The same approach should be taken at the EU level; the example of Sweden shows the need to adjust the scope of EUcohesion policies and understand that many other European policies can help address social and regional divides; the EU green deal, the EU strategy for the rights of the child, the EU gender equality strategy, to name just a few. Regional and structural policies should be more intertwined with other policy programmes such as research and development, innovation, and industrial policy. A broader approach towards economic and social well-being needs to be followed with the EU addressing the issue of social and economic inequalities in all their dimensions. Possible social and economic push and pull factors of regional development should be considered in the programme and policy designs. Rather than focussing on the spatial concentration of growth and employment effects, the aim should be to attain a more balanced growth picture by forging links between dynamic growth centres and the lagging regions. This study, which was written with the support of the progressive Swedish thinktank Arena Idé, is part of a joint FES and FEPS project on regional socioeconomic disparities in five EU member states(Finland, Sweden, Estonia, Italy and Romania). The findings of the national disparity studies form the basis for a European analysis aiming to put forward proposals to reform of the EU approach to regional policy and enhance the EU ability to contrast the cohesion policies. Local development and well-being in all areas of a country is not only a goal for economic policy, rather it is a matter of strengthening democracy and ensuring opportunities and participation for all. Growing spatial inequalities in many EU-member states have been fuelling the rise of antidemocratic movements and forces, questioning the respective democratic and political institutions. To diffuse rising dissatisfaction, the EU-member states and EU institutions need to address these inequalities and to cater for a more even development path. DR. PHILIPP FINK Director Friedrich-Ebert-Stiftung Nordic Countries Office DR. DAVID RINALDI Director of Studies and Policy Foundation for European Progressive Studies

FRIEDRICH-EBERT-STIFTUNG – POLITICS FOR EUROPE 4

The inequality of living standards in Sweden 5 1 THE INEQUALITY OF LIVING STANDARDS IN SWEDEN For long, a recognition of Sweden as the role model of universal welfare capitalism has prevailed. In recent decades a series of market reforms and cuts in taxes and social spending have, however, generated the most decisive rise of in‑ equality in the Nordic context. In this report, we examine consequences for spatial inequalities and present a policy recommendation for inclusive growth. The last three decades provide in two principle regards a transformation of the mixed economic model that emerged after WWII. The first element is the transformation of policy. After WWII, a social democratic welfare capitalist model evolved in Sweden. To promote high and equal standards of living, a mixed economic policy nationally and free trade policy internationally was combined with a universal system of welfare spending. The mixed economy with both strong public and private sectors directed high levels of investment to promote growth and structural change. To utilise the comparative advantages of the national economy, a free trade policy was targeted to enhance economic growth. To both achieve growth and mitigate the consequences of structural change, the rise of social spending and public supply of welfare services were seen as measures to meet both efficiency and equality targets respectively. The goal to reach full employment, high growth and price stability was however challenged with the oil crisis in the 1970s. After a decade of struggle between continuity and change, the model was largely reformed during and after the economic crisis in the 1990s. Inspired by the monetarist school of thought, the goal to reach price stability replaced employment as the key macroeconomic target. The mixed economy model was ­replaced with a market-based model. Public services were either privatised or managed by a quasi-market-based pub‑ lic management system. Social insurance was cut and pri‑ vate insurance initiatives became increasingly important for labour-income losses due to sickness, unemployment, and retirement. A strict budget policy was adopted to cut public debt by surplus targets throughout the business cycle after the 1990s crisis. Monetary policy governed by experts, and not active fiscal policy, was adopted as the key mechanism to reach macroeconomic price stability. The second element is the transformation of spatial value creation. Since the crisis of the 1990s, the macro economic situation has improved in Sweden. Real per capita incomes have increased by two per cent annually – a higher figure than both the EU and the Organisation of Economic Development Employment has risen by one per cent annually – outpacing population growth. With the exception of a decline in productivity growth in last two decades, the per‑ formance of the Swedish economy is fairly good, also from a historical perspective. In an international comparison, Sweden is doing well(OECD 2021). However, the gains of the last 30 years of economic prosperity have not been shared equally across the country. Most of the growth in employment has taken place in the metropolitan regions, foremost in the capital Stockholm. In other parts of the country, structural economic problems have led to declining employment. Whereas stagnating regions are facing problems related to internal migration, an ageing population and public deficits, the growing metropolitan regions are suffering from their success by facing problems related to housing, public infrastructure and segregation. The spatial dimensions of inequality today are evident from a long array of governmental reports, commissions and EU initiatives(OECD 2020; Widuto 2019; SOU 2017: 1; SOU 2020: 8). Although initiatives to mitigate spatial-related economic problems are considered, the situation has far from improved over the past decade. As will be shown in the present report, Sweden is not doing well across the nation. 1.1 SPATIAL DIVERGENCE THROUGH STRUCTURAL CHANGE Although regional disparities in economic development are far from a new phenomenon, the diverging trends we have seen in the last 30 years or so are largely recognised as a major concern for social and economic cohesion in developed economies. In Sweden, the issue of regional disparity was first recognised in the policy context in the 1960s, when the decline in the resource-based primary sector(forestry, ag‑ riculture, and fishing) led to economic stagnation foremost in the northern inland regions(Nilsson 2012). Declining employment led to internal migration, primarily of young individuals to growing cities in the coastal areas, manufacturingbased regions in central parts of Sweden and to metropolitan areas(Andersson et al. 2018). The expansion of employment in the manufacturing sector up till the late 1960s created balanced development except for in the resource-based regions in northern Sweden. In

FRIEDRICH-EBERT-STIFTUNG – POLITICS FOR EUROPE 6 most growing regions, employment growth continued even after manufacturing employment gradually started to decline in the 1970s. The expansion of welfare services created work in schooling, hospitals, care homes and other publicly financed work, making the regional economic development balanced in relation to the population. However, the crises in basic metal manufacturing in the late 1970s and early 1980s resulted in massive layoffs in the metropolitan regions of Malmö and Göteborg along with other cities dominated by basic manufacturing. For the metropolitan areas, a gradual transition into other lines of manufacturing and services slowly balanced the situation on the labour market(Andersson et al. 2018). In regions with a weaker manufacturing base and less diversified service sectors lost jobs were never replaced. but also led to a transition from rental to owner-occupied flats. The privatisation of publicly owned/controlled housing stock and cutbacks in public support have further reduced the supply of affordable rental housing. Unequal housing has grown into a major concern over the last two decades. Outside the metropolitan areas, employment growth has turned out to be much slower. A weaker labour market went together with a decline in population figures for many regions. When considering the development in population figures over the last 30 years, the picture of this development becomes clearer. Since 1990, the total population in Sweden has grown by 1.7 million inhabitants. Approximatively 95 per cent of this population growth has taken place in metropolitan and university cities or in closely affiliated regions(SCB 2021a). At the national level, the structural problems in basic manufacturing industries emerging in the 1970s and 1980s were temporarly balanced by an offensive devaluation policy and expanded public service sector employment. All efforts to mitigate the structural problems came to an end with the crisis in the 1990s. The crisis can be seen as a turning point of the post-war period. As indicated above, the crisis signals a shift from mixed economy policy to market-based policy and cuts in social insurance(Hultcrantz et al. 2011). In the short-term, the severe economic crisis led to major public deficits and layoffs in the publicly financed service sector. Employment in manufacturing declined and the decrease in domestic demand led to a crisis in the private servi‑ ces sector. A grave recession ensued. From a long-term perspective, the crisis signals the beginning of greater spatial imbalances. The export-oriented manufacturing sectors expanded rapidly after massive investments in new technologies, making demand for employment less strong that previously. The public sector faced major deficits and was unable to expand employment for many years. While the losses in manufacturing employment were shared fairly equally across the country, job growth in the private services sector during the recovery from the 1990s crisis was highly concentrated in the metropolitan areas(Andersson et al. 2018). Mainly, two major categories of work expanded rapidly. One the one hand, qualified work in business, finance, and technical services became increasingly important. On the other hand, less qualified work in personal services, retail, restaurants, and transport also increased. The growth of both qualified and less qualified employment led to a polarisation on the labour market in metropolitan areas, most evident in Stockholm(Åberg 2013). The unequal growth in employment spatially has further led to a difficult housing situation, especially for low-income households in metropolitan areas. An over-heated housing market in metropolitan areas has created upward pressure on prices and made homeownership less affordable. In the rental sector, negotiated rents have kept down rising housing costs, Close to half of all municipalities has lost population in the last 30 years – in total a drop of around 200,000 inhabitants(SCB 2021a). In total, some are lost. The stagnating regions face a situation with over-sized infrastructure and public facilities. The decline of the population has reinforced the ongoing demographic process of population ageing as most elderly individuals stay. Instead of schooling, elderly homes are in high demand. As seen from the population figure of the stagnating municipalities, the share of population in retirement age is equal to 27 per cent, a figure that is 6 per cent higher than the municipalities with a growing population on average(SCB 2021a). This spatial shift in population has far-reaching concequences for the Swedish model of welfare service provision, where welfare services(schooling, healthcare, elderly care) are managed and provided at the municipal level. Although based on the attractive local-democratic ideas of local independence and responsibility, the economic and social necessities of nationwide equal welfare provision is increasingly difficult to achieve. One reason is that an older population comes with a less economically active population. This had pronounced effects on the tax base of local governments. Less employment implies lower levels of income tax. The latter is the basis for tax revenue of municipalities/regions. To mitigate the financial consequences of structural imbalances, a financial equalisation system by municipalities is in operation. Despite the ambition to offer the same public services for the same taxes, there is a systematic gap in revenues and expenses. We find that there is a perfect negative linear correlation between tax rates today and population growth over the last 30 years (see figure 1). For inhabitants living in municipalities with the lowest population growth the tax rate averages 35 per cent, while for the ones with the highest population growth, the tax rate is 29 per cent(SCB 2021b). There are few reasons to believe that services are better in municipalities suffering most from financial deficits and cuts in services. In stagnating regions, the elderly are offered less welfare provision at a higher cost than in growing areas. Although inter-municipal equalisation has been in operation, the structural differences are not fully compensated, as shown by the perfect linear relation between population growth and tax rate.

The inequality of living standards in Sweden 7 Compared to many other European countries, local selfgovernance and state governance by municipalities in Sweden is extensive. Municipalities in Sweden have a wide range of responsibilities, inclusing welfare provision such as childcare, schooling elderly care. The municipalities provide welfare services such as education as well as child and elderly care. They also provide technical services. In addition, munici‑ palities play a key role in societal development; they are responsible for physical planning, housing, infrastructure and community development. They also have supervisory and control duties across various areas, such as the environment and food security. Furthermore, municipalities are important employers; in 2017, the municipalities had 700,000 employees(SOU 2020: 8). To meet the demand for statistics at a local level, Eurostat maintains a system of Local Adminis‑ trative Units(LAUs) compatible with NUTS. These LAUs are the building blocks of the NUTS and comprise the municipalities and communities of the European Union, with the lowest LAU level being LAU level 2, that of municipalities or equivalent units in the 28 EU Member States(Eurostat 2021). A symmetrical model of tasks and responsiblities is applied irrespectively of the great heterogeneity in size and structure. Faced with changing demographics and migration towards metropolitan areas, there are concerns to what extent citizens in Sweden are equally supported by welfare in relation to the local taxation rate. The last three decades of population growth(and related change in age composition), have led to increasing differentiation of local taxes. In the municipalities facing a declining population, the tax rate is significantly higher than in municipalities with a growing population. As shown in the scatterplot below(Fig. 1), there is a perfect negative linear relationship between last 30 years of population growth and local tax rate in 2020. Figure 1 Local tax rate(2020) and population growth(1990–2020) 36 34 Taxe rate (%) 32 30 28 –1 0 1 2 3 Population growth(%) Source: Own illustration, SCB(2021 a, b). 1.2 WHAT IS SPATIAL INEQUALITY ABOUT TODAY? The spatial inequality of today comprises a division between stagnating and growing regions. At both ends of the distribution, the unequal spatial situation is creating strong tensions. One of the most obvious is that between the growing metropolitan areas – the place where most of the new employment is created – and the declining remote areas – where old jobs are destroyed but not replaced. In a perfect world without any barriers in costs or asymmetries in information, individuals would freely move to find a place to satisfy their needs and improve their well-being. However, there are massive barriers and asymmetries that are closing the doors, barring individuals from the wealth created in today’s economy. Where you are born, raised and live are having an increasingly decisive impact on the prospects of life and well-being in Sweden. In stagnating regions where individuals face difficulties in finding work, the option of staying presents a sizable risk of being less well-off. Financial deficits despite high taxation rates make foremost the elderly in stagnated regions suffer from limited welfare provision. The alternative of internal migration – to move from stagnating to growing labour markets – is not without difficulties either. We have seen a tremendous expansion of labour markets in the metropolitan areas in Sweden over the last 30 years(Andersson et al. 2018). Expansion has however been slowly adopted in the planning of housing and infrastructure. One of the major barriers to entering metropolitan labour market is the shortage of affordable housing. Although the growing spatial imbalance is the outcome of the last 30 years of development, the policy response is lagging. The regional policy of today is not purposely designed to tackle the great divide of stagnating and growing regions. If anything, the regional policy was turned into a growth policy in the 1990s in conjunction with EU-membership(Nilsson 2012). Hence, it has become a regional poli‑ cy for achieving national growth goals. In the rhetoric of the growth policy, it aims to be spatially tailored in its design to make the most of the competitive advantages that each region’s resources hold. In practice, the growth policy may potentially work in regions that already show dynamic growth, whereas stagnating regions receive less support, if any, by attempts to create a competitive and innovative local economy from the top down. Many of the ideas guiding the policies are not based on evidence-based research, but on buzzwords, quasi-scientific terms or unrealistic comparisons with successful regions. A series of quantitative evaluations based on causal design has found no impact of Sweden’s current regional growth policy. On the regional level, there are no examples where a stagnating region has been turned into a growing one thanks to well-designed growth policy. For obvious reasons, the spatial restructuring related to the fundamental change from production of goods to services – where jobs in manufacturing decline in all regions and jobs in services emerge in metropolitian areas – is a spatial transformation that is not mitigated by a few projects funded through EU-structural funds. Any policy aiming to turn the vehicle around is unrealistic (Normark/Svensson 2019). Instead, it should support adaption to the ongoing structural change of the economy and find ways of helping all individuals equally to make the best of their lives and take an active part in the value created in the economy. After reviewing the current situation of Swedish regions in the following chapter, we return to our ambition of advancing a proposal to improve the current situation and to make economic development more balanced in the future.

FRIEDRICH-EBERT-STIFTUNG – POLITICS FOR EUROPE 8 2 SWEDEN TODAY: GROWING DISPARITIES IN AN AGEING COUNTRY The assessment of growing inequalities described in the previous section contradicts, to a certain degree, international comparisons where Sweden frequently ranks amongst the best countries in the world. A case in point is the“Better life” reporting of the OECD: The regional disparity index for eight subnational regions(Riksomraden) lists Sweden as the 12th best out of 30 assessed OECD countries. When applied to 21 subnational regions(Län), Sweden reaches the pole position: On this level of comparison, the regional disparity index in Sweden is the lowest of all OECD countries(OECD 2020: 55). The reason for this contrasting assessment is that the regional disparity index informs about overall disparities but has limited explanatory value for the socio-economic framing conditions for Sweden’s future development. A more comprehensive set of indicators, as employed for the analysis presented in this section on the level of 290 Swedisch municipalities, delivers additional insights. Key findings are: – Sweden’s population dynamics constitute diverging development paths for a demographically ageing country under 21st century transformation pressures. Without in-migration from abroad, Sweden would be a shrinking country. – Most of the new arrivals move into larger cities, just like many(predominantly young) Swedes from rural and peripheral areas.(Sub-)urbanisation pressures and disparities between urbanised and rural areas follow. – Prosperity and better living conditions in Sweden are clearly associated with the proximity to urban economies. The best living conditions can be localised in municipalities in comfortable commuting distance to the most lucrative urban labour markets. – Exceptions from this general picture can be found in a few large but sparsely populated municipalities in the very North with high-income opportunities in mining and tourism. – Living conditions in large cities(Stockholm, Göteborg, Malmö, Uppsala, Linköping, Örebro, Västerås and others) are more diverse and exposed to new risks of social polarisation. Disadvantaged households and unqualified people are confronted with rising living costs and a transforming labour market. Job and income opportunities are increasingly linked to high education and qualification levels that divide the workforce and put pressure on people to adapt. – Rural and peripheral areas struggle with the changing demand for existing infrastructure and services. Ageing populations require more health services. Out-migration of families leads to a surplus of housing and problems on the real estate market. Oversized infrastructure is costly to maintain and downsize/adapt to changing population numbers. On top of the vicious cycle of demographic change that shrinking municipalities experience, continued mechanisation in agriculture and automation in industry places limits upon job opportunities – continued out-migration ensues. The key results show that Sweden faces a range of challenges. New concepts of strategic structural and social policies are needed to manage diverging development trends – despite the fact that inequality in Sweden is currently relatively low compared to other countries. Long-term(e.g. climate change, global integration) and short-term transformation needs(e.g. the corona pandemic or the 2008 financial crises) have the potential to deepen disparities in the future. In this context, studies on regional economic resilience argue for “place-sensitive distributed development policies” that carefully reflect on potential effects of national and regional policies not only where they are implemented, but also in terms of their effects on neighbouring regions(Iammarino et al. 2019; Pontarollo/Serpieri 2020: 163). Based on this viewpoint, the analysis of socioeconomic disparities in this report focuses on an integrated assessment of Sweden’s framing conditions for future development. It discusses current spatial variations of strengths and weaknesses in the light of future risks and challenges for the country. Selected indicators inform on(1) economy, employment and the labour market;(2) educational opportunities and life chances;(3) prosperity and health;(4) state action and partici‑ pation; and(5) internal migration patterns. 2.1 METHODOLOGY The strengths and weaknesses of Sweden’s geography are diverse. Indicators used to capture their spatial variation and differences were chosen for their explanatory power for selected topics. They stand as proxies for unequal developments that can be associated with geographical framing

Sweden today: Growing disparities in an ageing country 9 conditions and interpreted in comparison with developments elsewhere in the country. Next to the choice of indicators, the spatial granularity for input data is important. Values for indicators can more clearly be attributed to the policy environment if the area of observation represents the sphere of influence for political action and governance accurately. Even though national and state policies as well as local decisions always interact to some degree, values for the municipal level show more informative value in this context than overarching administrative levels where data is aggregated and resulting averages can lead to a blurring of spatial patterns. The novelty of this report is the integrated analysis of a comprehensive set of indicators on the municipal level in a geostatistical procedure known as cluster analysis. Single indicator maps are combined into areas with similar strengths and weaknesses in comparison to the national average. The resulting map informs about a spatial typology of disparities in Sweden, the so-called Disparity Map of Sweden. It is important to read the map in conjunction with statistical information that characterises a cluster. Moreover, a brief textual interpretation portrays the visible spatial patterns with a view towards explanatory factors. The following indicators and indicator groups were used in the generation of the Disparity Map of Sweden 1 : 1. Employment rate, demographic dependency ratio (Economy, employment and labour market): Employment is the foundation of economic activities. Higher rates demonstrate a successful match between the job opportunities a region has to offer and the skill levels and preferences of the local and regional workforce. Employed people usually generate the funds for dependant people through income and social insurance contributions. The demographic dependency ratio indicates the ratio of dependant people to working age people. Higher values point towards higher demands of dependent people and higher pressure on private and public funds to support them. High values are frequently an implication of demographic ageing and out-migration of working-age people. 2. At-risk-of-poverty rate for children, highly qualified workforce(Educational opportunities and life chances): An upbringing in poverty is a significant burden for children. Many studies show that concentrations of poverty can lead to negative milieu-specific effects for the realisation of life chances. The share of employees in the highly qualified workforce emphasises the importance of education in this respect. Higher values show where more people have the prerequisites to compete on an increasingly competitive labour market if matching job opportunities exist. The current match, however, is only part of the picture. Higher education levels are also associated with higher potentials for personal development and reorientiation on a transforming labour market. 1 in bold: indicator name; in italics: topic group 3. Average monthly gross salaries, house prices, medical and health-related employment, gender pay gap(Prosperity and health): Salaries are fundamental for covering the cost of living. Insufficient income leads to exclusion and pressure on families and/or the welfare state to cover living costs for dependent people. House prices stand for the variation of living costs across the country. Higher house prices not only show where the real estate market is not as affordable as elsewhere. Higher costs for housing are frequently associated with other higher expenditures in more expensive regions. The share of employees in medical and health-related employment is used as a proxy for the importance of the health sector in a region. Higher shares can be constituted by high demand, for example in regions with a high share of elderly people, or high demand for specialised services in comparison to other form of employment in a region. Regional variations in the gender pay gap show where women’s salaries deviate from men’s average income. Higher values(i.e. women earn less than men) can frequently be found in highly qualified jobs where issues of gender equality are not regulated. 4. Voter turnout, investments per capita(State action and participation): The share of people who vote at national elections shows people’s interest in democratic participation. Higher shares are frequently attributed to higher levels of education and wealth. Affluent and educated people are more likely to vote. Certain“hot” topics and the specific appeal of personalities can also motivate people to vote. This can also be seen as a positive contribution to participation. Investments per capita signify the capacity of local government to invest in infrastructure and services for the local population and economy, including funds from supraregional institutions that are successfully directed into the area. Higher levels are associated with higher quality infrastructure and services that are more likely to match the demands of local users. 5. Internal migration balance(Migration): The balance of in- and out-migration can be interpreted as an earlywarning indication of spatial mismatches between people’s expectations for the realisation of life chances on the one hand, and the significance of deficiencies that motivates migration on the other. Demand and supply of infrastructure, stability of the labour market, and many cultural and societal inequalities are associated with migration patterns and the resulting population base. In this context, internal migration can be interpreted as an expression of locational preferences and the perception for desired living conditions in the Swedish population. Figure 2 shows the resulting spatial typology in the national disparity map of Sweden. The clusters are semantically framed with labels derived from the interpretation of indicator values and additional information on the geography of their delineation. Table 1 gives a summary overview of indicators that characterise the single spatial types. Arrows are used to symbolise the mathematical value of indicator values(very high:  ↑ ; high:  ↗ ; average:  o ; low:  ↘ ; very low:  ↓ ).

FRIEDRICH-EBERT-STIFTUNG – POLITICS FOR EUROPE 10 In some cases, high values stand for a positive locational factor(i.e. high values for average incomes, high employment rates), in others they are rather negative for life chances (i.e. high at-risk-of-poverty rates or high values for home loans). For this reason an additional colour background (shades of green= rather positive or very positive; light grey: average; shades of red= rather negative or very negative) is used to indicate the assessment of values in terms of strengths or weaknesses of a region – always to be interpreted in comparison to the national averages. The combination of the disparity map and its constituting statistical values aims to help interpretation. An interactive web map allows further investigation of values for all input variables and their combined effect in the disparity map: https://fes. de/unequal-sweden. 2.2 FOUR SWEDENS The disparity map shows that Sweden can be differentiated into four spatial types with distinct socioeconomic advantages and disadvantages. The colours in the legend use associative colours on purpose: shades of green show areas that currently fare better in the overall assessment and seem to be better prepared for the challenges of the future – at least for the majority of people. The ochre colour shows areas that frequently have indicator values close to the national average. The violet colour is used to map out areas with a majority of negative indicator values – areas in need of dedicated policy attention. Based on this colour interpretation the map shows basically three settings for disparities: average, above average, and below average. 1. Average: Sweden’s solid middle(ochre colour in figure 2) is home to 2.14 million people(20.9 per cent of the total Swedish population of 10.2 million inhabitants) and 121 municipalities(41.7 per cent of a total of 290 munici‑ palities). In the South it contains mainly inland regions, in the North it covers sparsely populated but large municipalities. This cluster surrounds the above average clusters which hints towards the decisive role of distance to the main business centres as a driver of socioecono‑ mic disparities. This spatial type is characterised by a majority of indicator values close to the national average (see table 2). comparatively less qualified labour from new in-migration compete for lower-income jobs. Rising living costs are a critical driver of social polarisation, gentrification, and displacement of disadvantaged households. Successive suburbanisation pressure leads to increasing commuter numbers that put stress on overstretched and automobile-dependant transport systems. These two spatial types are inhabited by 6.9 million people(67.1 per cent of the Swedish population) in 77 municipalities (26.6 per cent of all Swedish municipalities). 3. Below average: The periphery of Sweden(violet colour in figure 2) shows clear signs of the locational disadvanges. The long distance to the lucrative urban labour markets and limited business opportunities in the area itself drive an out-migration that threatens the viability of the remaining locational framing conditions. Automation and continued mechanisation in agriculture and industrial processes add to the limited prospects for young people in this cluster. Fewer people return after life phases in education and early careers in the large cities. The remaining population is therefore older on average and in need of stretched health services. Vacancies on the housing market and oversized infrastructure due to a shrinking population base become problematic for local government to manage. There are two exceptions in this cluster where municipalities closer to large dynamic cities show similar values(areas around Helsingborg north of Malmö and Eskiltuna west of Stockholm). These areas suffer from similar problems due other disadvantageous locational factors. In total, the sparsely populated Swedish periphery is populated by 1.33 million people (13.0 per cent of the Swedish population) in 82 municipalities(28.3 per cent of all Swedish municipalities). This summary characterises the patterns of disparities at a glance. The definition of such spatial types also lends itself to evaluate structural and social policies in the future. For this purpose, table 2 shows the current bandwidths of indicator values accompanied by the name of the respective municipalities with minimum and maximum values within each cluster. 2. Above average: The majority of Swedes live in better-off affluent commuter belts and prospering city regions with high risk of social exclusion(light and darker green in figure 2); also large but sparsely populated mining areas in the North. Affluent commuter belts and prospering city regions complement each other in terms of their functional relationship between a dynamic economic centre and municipalities within commuting distance to the centres. These suburban or exurban commuter belts differ in terms of indicator values: Some show higher quality of life in terms of wealth distribution, local living conditions, but also limited affordability for low-income households. In addition, highly urbanised areas in the core of these two clusters experience a rising risk of social exclusion. A larger share of unemployed people and

Sweden today: Growing disparities in an ageing country 11 Figure 2 The Swedish disparity map Luleå Östersund Umeå Örnsköldsvik Sundsvall Karlstad Falun Gävle Örebro Uppsala Västerås Eskilstuna Stockholm Linköping Göteborg Jönköping Cities(inhabitants) ! City > 500,000 ! City > 100,000 ! City < 100,000(selection) motorway and primary road system Halmstad Helsingborg Malmö 0 100 km Växjö Kalmar Karlskrona Source: Own illustration. Data: SCB Statistikdatabasen Sweden. Cluster Affluent commuter belts and mining areas of the North Low population density Prospering city regions with high risk of social exclusion Sweden´s solid middle The periphery High population density Geodata: Geofabrik GmbH, OpenStreetMap Contributors 2018

FRIEDRICH-EBERT-STIFTUNG – POLITICS FOR EUROPE 12 Table 1 Spatial typology of socioeconomic disparities in Sweden Characterisation Indicator assessment Affluent commuter belts(A) and mining areas of the North(B) (44 municipalities; 1.56 mio. inhabitants) Spatial delineation Subcluster A: Sweden’s most prosperous regions are located in the suburban commuter belts of the major metropolitan business centres of the South. They are home to a highly qualified and high-income workforce with top employment rates and the highest level of politi‑ cal participation(expressed by the voter turnout) in comparison to the Swedish average. The affluence of these regions has follow-on effects, for example the relatively low risk that children will experience poverty, high government investment rates and a low ratio of people who depend on others to support them(dependency ratio). It is likely that foreign and internal in-migration into urban centres generates suburbanisation pressures(“spillover effect”) into commuting areas where traditional family patterns prevail. High gender pay gaps and very high prices on the real estate market support this interpretation. Especially the latter have a selective effect on social milieus that can afford to live here. Employment rate: 85.1% ↑ Dependency ratio: 65.6% ↘ Highly Qualified: 34.1% ↑ Child poverty risk: 9.7% ↓ Gender pay gap: 74.5% ↓ Income: 32,589 SEK ↑ Healthcare 2 : 36.6 per o 100 inh. House prices(in SEK ↑ 1,000): 4,367 Investments 3 : 8,267 SEK ↗ Voter turnout: 90.0% ↑ Migration: –0.8 inh. ↗ per 1,000 Subcluster B: Similar values of affluence and prosperity can be found in sparsely populated large municipalities of the North. High quality labour markets in mining and tourism are the main explanatory factors for above-average local living conditions in these areas. Prospering city regions with high risk of social exlcusion (43 municipalities; 5.3 mio. inhabitants) This cluster includes the largest and most dynamic cities in Sweden (Stockholm, Göteborg, Malmö, Uppsala, Linköping, Örebro, Västerås and others). These are the most populated areas(average of 123,000 inhabitants) in comparison to all other clusters(average of 20,000 inhabitants). Business opportunities and services are excellent but not accessible to everyone: Overall employment rates are rather low. Parts of the urban society are not part of the workforce. This has follow-on effects for children in some milieus where the child poverty risk can be quite high(overall in this cluster: average in comparison to the Swedish average). Unemployed and disadvantaged households are exposed to rather high living costs as expressed by the house prices. The Swedish popu‑ lation in this cluster fluctuates but is stable in balance, the dependency ratio is lowest: Population growth in the cities is driven by new arrivals from outside of Sweden. Main challenges are integration of people into a transforming labour market and policies to control rising living costs as drivers of social exclusion. Employment rate: 79.1% ↘ Dependency ratio: 57.8% ↓ Highly Qualified: 34.1% ↑ Child poverty risk: 19.7% o Gender pay gap: 81.2% ↗ Income: 31,554 SEK ↗ Healthcare: 46.2 per ↑ 100 inh. House prices(in SEK ↗ 1,000): 3,633 Investments: 6,495 SEK o Voter turnout: 86.8% o Migration: –0.8 inh. ↗ per 1,000 > 2“Healthcare” stands for the indicator“employees in the medical and health sector per 100 inhabitants over 65 years of age”. 3“Investments” stands for the indicator“investment per capita by local government/municipality in SEK 1,000(average of 2016–2019)”.

Sweden today: Growing disparities in an ageing country 13 Value key: very high values:  ↑  high values:  ↗  average values:  o  low values:  ↘  very low values:  ↓ How to interpret: very positive   positive  average  negative  very negative  Characterisation Sweden’s solid middle (121 municipalities; 2.14 mio. inhabitants) This cluster has the highest number of average values across all selected indicators of this study. It comprises vast areas of Sweden with ca. 40% of all 290 municipalities and about 20% of the population. Demographic ageing and out-migration are the main challenges. Statistically, comparatively many people depend on the remaining workforce, arguably elderly. Many young and qualified people are leaving the area in search for jobs in the cities, frequently relocating to the affluent commuter belts in a later phase in life(“family suburbanisation”). The municipalities in this cluster are therefore struggling to stabilse the population with new arrivals from abroad and changes to the social setup: Comparatively low house prices stand for limited attractivity of some regions within the municipalities in this cluster, an average number of children live at risk of poverty, income and government investments are below average. Municipali‑ ties in this cluster clearly show early symptoms of the upcoming threats of demographic change in need of policies for structural support. Indicator assessment Spatial delineation Employment rate: 81.6% ↗ Dependency ratio: 73.2% ↗ Highly Qualified: 22.3% ↘ Child poverty risk: 20.1% o Gender pay gap: 77.9 % o Income: 30,666 SEK o Healthcare: 32.9 per ↘ 100 inh. House prices(in SEK ↘ 1,000): 1,754 Investments: 5,563 SEK o Voter turnout: 87.7% o Migration: –2.9 inh. ↘ per 1,000 The periphery (82 municipalities; 1.33 mio. inhabitants) In comparison to Sweden’s solid middle, the majority of municipalities in this cluster belong to the low-density periphery of Sweden. They show advanced symptoms of the development threats of demographic change under 21st century transformation pressures. Areas far away from the cities have consistently lower values than the Swedish average for all indicators except the gender pay gap and house prices. Out-migration of the Swedish population to other places drives a vicious cycle: People have a comparatively high risk of unemployment and limited chances for reorientation on the labour market due to lower qualification levels and the lack of business opportunities. Incomes are low. Government budgets are stretched with necessary investments to maintain oversized infrastructure. Many children live under the risk of poverty. Participation levels, as expressed by the voter turnout, are also below average. This cluster is in special need in terms of policy incentives, e.g. to support digital business opportunities and strategic investments in supporting infrastructure services. Two exceptions in more densely populated vicinities of larger cities complement this cluster: The area around Helsingborg north of Malmö and Eskilstuna west of Stockholm show socioeconomic disadvantages similar to the regions in the periphery. Employment rate: 77.4% ↘ Dependency ratio: 74.3% ↗ Highly qualified: 19.2% ↘ Child poverty risk: 29.0% ↑ Gender pay gap: 81.7%  ↗ Income: 30,126 SEK ↘ Healthcare: 31.6 per ↘ 100 inh. House prices(in SEK ↓ 1,000): 1,171 Investments: 4,963 SEK ↘ Voter turnout: 85.1% ↘ Migration: –30.2 inh. ↓ per 1,000 Source: Own illustration. Data: SCB Statistikdatabasen Sweden.

FRIEDRICH-EBERT-STIFTUNG – POLITICS FOR EUROPE 14 Table 2 Bandwidth of indicator values for the spatial types Indicator Value Affluent commuter belts and mining areas of the North Prospering city regions with high risk of social exlcusion Sweden’s solid middle The periphery Employment rate(%) Demographic dependency ratio(%) Highly qualified workforce(%) At-risk of poverty rate (age 0–19)(%) Monthly gross salary (SEK) Min. Max. Min. Max. Min. Max. Min. Max. Min. 79.5(Danderyd) 89.1(Habo) 52.5(Mölndal) 78.9(Tjörn) 19.7(Gällivare) 54.8(Danderyd) 5.0(Täby) 16.0 Boden) 30,200(Kiruna) Max. 36,900(Sollentuna) Gender pay gap(%) Employees in the medical and health sector per 100 inh. over 65 years of age Min. Max. Min. Max. 64.7(Danderyd) 81.5(Lidköping) 23.9(Danderyd) 50.3(Mölndal) House prices (in SEK 1,000) Min. Max. 1,273(Boden) 11,209(Danderyd) Investment per capita by local government/munici‑ pality in SEK 1,000 (average of 2016–2019) Voter turnout(%) Overall internal net migration balance (per 1,000 inh.) Min. Max. 1,972(Knivsta) 29,472(Gällivare) Min. Max. Min. 82.3(Gällivare) 93.9(Lomma) –75.6(Boden) Max. 119.7(Trosa) 67.8(Malmö) 84.0(Åre) 43.3(Sundbyberg) 71.9(Eksjö) 22.6(Nässjö) 55.8(Lund) 11.0(Solna) 37.0(Malmö) 29,700(Norrköping) 34,300 Väsby; Solna) 73.3(Upplands-Bro) 87.3(Åre) 36.2(Sigtuna) 66.2(Umeå) 1,528(Falköping) 8,874(Solna) 2,128(Gävle) 11,953(Skövde) 77.5(Södertälje) 89.9(Karlskrona) −29.0(Sigtuna) 34.1(Åre) 76.0(Degerfors) 86.4(Krokom) 60.2(Svalöv) 92.6(Borgholm) 15.9(Vansbro) 34.0(Höganäs) 12.0(Mörbylånga) 30.0 Köping) 28,600 Lycksele) 33,900(Höganäs) 71.3(Grums) 86.5(Borgholm) 20.4(Båstad) 49.8(Lycksele) 443(Bjurholm) 3,915(Båstad) 1,218(Heby) 14,277(Älmhult) 83.6(Nordanstig) 91.7(Mörbylånga) −49.4(Arvidsjaur) 65.9(Gnesta) 66.4(Eda) 85.5(Sorsele) 57.9(Helsingborg) 95.5(Pajala) 14.4(Filipstad) 32.2(Helsingborg) 19.0(Åsele) 45.0(Högsby) 28,200(Åsele) 32,300( Helsingborg) 72.3(Klippan) 92.7(Sorsele) 21.9(Strömstad) 42.7 Eskilstuna) 330(Åsele) 4,042 borg) 1,028(Laxå) 13,007 stad) 72.8(Haparanda) 88.5(Vingåker) −126.0 berg) 34.5(Ockelbo) Source: Own illustration. Data: SCB Statistikdatabasen Sweden.

New policies for equality of living conditions and social cohesion 15 3 NEW POLICIES FOR EQUALITY OF LIVING CONDITIONS AND SOCIAL COHESION After reviewing the spatial configuration of the economic and social aspects of inequality, we find that inequality of life and well-being is not achieved though the current regional growth policy or the present policy of regional equalisation/ redistribution. The last 30 years of growing economic prosperity – during which the total national income almost doubled – has largely created wealth in the growing metropolitan areas. Almost half of all municipalities or local regions (LAU2) have faced a decreasing population base. In contrast, the dynamic and growing metropolitan areas, university ­cities including their hinterlands have expanded tremendously. Not since the eve of the industrial breakthrough is the spatial dynamics of regional divergence stronger than the development observed in the last decades. The spatial dynamics of rapidly growing labour markets in foremost metropolitan areas with strong service sectors stand in strong contrast with the stagnating regions with jobs being lost in manufacturing. The tension created between the growing and stagnating regions is becoming stronger and hamper social cohesion and creates inequality in well-being. As a result, the spatial polarisation in Sweden is contributing to the country’s political polarisation(Lindell/ Pelling 2021). Currently, the stagnating regions are struggling with a shrinking population, over-sized and obsolete infrastructure, ageing populations and the loss of jobs and tax revenues to provide citizens with income and welfare. Growing regions are expanding at a pace where difficulties in adapting to the new spatial dynamics create problems with housing, as manifested by over-crowding, poor housing conditions, high real estate prices and incommensurable waiting time for rental apartments. Lagging investments in physical infrastructure make access to work and welfare provision time-consuming and stressful with a loss of well-being. Polarisation in the labour market reinforces the trend of housing segregation within urban areas, organised criminal networks and incidents of gang violence. The unbalanced regional economic growth creates structural problems of major concern in both growing and stagnating regions. A policy aimed at accommodating the current spatial econo‑ mic dynamics needs to address both growth and decline to provide opportunities for all individuals equally, and to make the best of individuals’ lives and make it possible for all to take an active part in the value created in the economy of today and in the future. 3.1 LIMITATIONS OF CURRENT SPATIAL POLICY The regional policy of today is not purposely designed to tackle the great divergence between stagnating and growing regions. Neither are the current more specific problems facing stagnating and growing regions addressed in a coherent manner to achieve equal opportunities and improve living conditions for all individuals. The regional policy of today suffers from two major misconceptions. Almost all expansion of employment in the last few decades has taken place in the service sector in metropolitan areas, university cities and related areas. The expanding labour market has stimulated rapidly growing housing demand, where close to 95 per cent of the entire population growth(1.7 million) between 1990 and 2020 has taken place in metropolitan areas, university cities and related(commuting) regions(Statistics Sweden, Population). The growing demand has largely been met by a neoliberal housing policy. Since the early 1990s, investment has been left to the market, although every citizen’s right to housing was maintained, at least in rhetorical terms. The shortage in supply together with low interest rates led to a surge in real estate prices, and the transition of rentals into owner-occupied flats. Housing became increasing difficult to access and afford, while polarisation in the labour market between high- and low-skilled works expanded simultaneously. However, rather than demanding greater social cohesion, the current debate on the housing market is largely dominated by an even stronger neoliberal push towards market reforms in rental housing. Any policy to mitigate the over-heated metropolitan areas’ problem of shortages in housing, segregation and social unrest has however proved, if anything, ineffective. Fostering a booming market without accommodating the social and economic consequences of growth is the first misconception. At the other end of the distribution are the regions suffering from job losses in manufacturing, without any significant creation of jobs in the service sector. Stagnating regions are burdened by less work and lower incomes. The traditional regional policy aiming at offering support to firms willing to invest in declining regions was replaced in the 1990s by a growth policy inspired by buzzwords and quasi-scientific ideas of creating competitive regional environments of growth (Nilsson 2012). Without more of a fundamental understanding of the mechanism of growth dynamics in the service sec-

FRIEDRICH-EBERT-STIFTUNG – POLITICS FOR EUROPE 16 tor, most initiatives fell short. The stagnation was meaningfully changed by the growth policy initiative, but neither were the structural problems of declining population and ageing population addressed in a coherent manner. Although intermunicipality redistribution of funding for asymmetrical incomes and costs were in operation, we find in this report that inhabitants in stagnating municipalities still faced a higher tax rate but received a lower level of welfare services. The insufficiencies of inter-municipality redistribution have received substantial attention in the national debate on regional disparities. The debate has largely taken place in the media, but also became a concern in governmental reports. There is a growing consensus that stagnating regions suffer from a combination of higher taxation and lower welfare provision. Adapting to stagnating development without accommodating the social and economic consequences of decline is the second misconception. Any policy aimed at turning the tide is far from realistic. It is time to realise that we need a spatial-based social and economic coherency policy that helps us adapt to the ongoing structural change of the economy. We need a policy aimed at helping all individuals equally to make the best of their lives and take an active part in the value created in the economy. The next section will outline the point of departure for such a policy initiative. 3.2 RECOMMENDATIONS FOR LOCAL-BASED ACTION The spatial analysis of social and economic disparities illustrates a number of challenges specific to the clusters identified. To provide more locally focussed recommendations for action, in addition to the aforementioned general recommendations, more specific recommendations are put forward. To improve access to expanding labour markets, investments in infrastructure to expand local labour markets are crucial. Greater inclusion of migrants may be stimulated by a more active and educational integration policy. And to reduce the impact of social and economic disparities in childhood, greater compensatory measures should be applied in schooling. For the booming labour markets related to booming mining, energy, and basic metal industries, we recommend stronger national support to enhance the benefits of economic activities in sparsely populated areas. To improve the situation in areas with declining labour markets, where many municipalities face both increasing demand for services and welfare, but also a more difficult financial situation, a number of measures may be considered. We recommend stronger mechanisms to mitigate the local financial deficits related to ageing populations. For regions and individuals, the difficulties in accessing labour and employment for different skills need to be facilitated by national support. Investments in sustainable energy production is an opportunity to strengthen the local economy and reduce emissions nationally. For more dense regions with low employment rates, we recommend a more active education and labour market policy. Table 3 Summary of spatial-based recommendations for action Affluent commuter belts and mining areas of the North – An active housing policy to enhance the supply of affordable housing – Expansion of the not-for-profit housing segment – Investment in infrastructure to expand local labour markets and commuting areas – Stronger national support to enhance the benefits of booming industries in the North Prospering city regions with high risk of social exclusion – Improve employment through support for education and on-the-job training – Make access to housing more equal by enhancing the supply and expanding the non-profit segment – Apply a more active integration policy to stimulate higher employment among foreign-born people – Reduce disparities in childhood by making schooling more compensatory Sweden’s solid middle – Facilitate access to greater labour market regions – Apply stronger mechanisms to mitigate the local financial deficits linked to ageing populations – Offer national support to attract the skills and labour to welfare and other services – Support investments in sustainable energy production The periphery – Impose broader mechanisms to alleviate the financial burden of ageing populations – Provide stronger support to digital business opportunities – Support investment in sustainable energy production – Support redemption of housing – Improve employment through support for education and on-the-job training 3.3 RECOMMENDED ACTIONS AT NATIONAL LEVEL The locally focussed policy recommendations should be seen as part of a greater national and international policy with regard to spatial effects of policy action. From a national perspective, one of the most decisive policy areas from a spatial perspective is the public budget. As most of the budget is related to welfare, the distribution of revenues and expendi‑ ture in that area is of vital importance for individual wellbeing over the life cycle. 1. Make taxation independent of place of residence! We argue that welfare provision based on the idea of universalism is fundamental for equality and efficiency reasons. To achieve a universal welfare regime, a basic requirement is that all citizens are offered equal quality in welfare provision. To achieve this over-arching target of welfare policy from a spatial perspective, taxation also needs to be equally independent of place of residence. Securing equal welfare provision also demands that structural differences at the expenditure side are addressed in a coherent manner. Still, funding of welfare at the regional level has proved untenable historically. To address

New policies for equality of living conditions and social cohesion 17 spatial inequality in welfare, a socially coherent policy will attain: – Equal taxation – Citizens are taxed equally by place of residence – Equal welfare – Citizens are offered equal welfare services net of spatial structural differences 2. Strengthen wealth and income redistribution! A stronger redistributing element of taxation would mitigate the structural differences between growing and stagnating areas, but also reduce segregation within the growing areas. In the last 30 years the redistributive element of taxation has been largely reduced by lowering taxation of capital and by lowering taxation of toplabour incomes. Imposing equal taxation of labour and capital income and restoring the taxation of property and wealth are decisive measures both for reducing urban segregation and the regional divergence in disposable incomes. To address income inequality from a spatial dimension, a socially coherent policy will attain: – Equal labour and income taxation – Progressive taxation of income 3. Increase public investment! Public investment policy needs to tackle the lack of investment in physical infrastructure and housing. To overcome the housing shortage in the expanding urban areas, support for investment in good, affordable, and more accessible housing facilities is a necessity. Although such a policy will benefit the urban space as such, we need to consider equal opportunities to individuals willing to move into the expanding labour markets. An insurmountable barrier to entry is not the key to resolve segregation or the difficulties to move into growing regions. To mitigate urban segregation, and create opportunities for entering the metropolitan labour markets, a social coherent policy will attain: – Public investment in infrastructure – Public support/incentives for housing investments make adaptation to changing spatial economic structures less tense, also in Sweden. From an economic-spatial perspective, the European Regional Development Fund(ERDF) could play a major role in improving social and economic cohesion in Sweden. Over the past six years, the ERDF spent a total of€7 billion, of which most was attributed to regional competition, environmental protection, sustainability, and research and development. Of the total EU funding in Sweden, social cohesion programmes encompass€0.5 billion(EU Fonder 2021). In a review of the spending over the recent budget cycle, it was concluded that the administration of funds works efficiently. Internal targets of projects are achieved to a large extent, but the impact on regional economic development was highly limited and any long-term effect difficult to find. The EU policies directed at regional and structural development have, if anything, had a limited capacity to reduce regional disparities in a country. The focus on growth policy rather than social cohesion, is questionable given the direction of investments and the poor outcomes. To reduce the barriers of adaptation to the ongoing structural changes in the economy, investment in human capital offers greater potential. A socially coherent policy will attain: – Investment in human capital – Social cohesion programmes in segregated areas For the next generation of EU citizens, we need a coherent social and economic policy that aims at helping all individuals make the most of their lives and take an active part in the value created in the economy. In that regard, the Next Ge­ neration EU programme offers opportunities to mitigate the individual cost and to increase the opportunities of adapting to the structural changes in the economy by making access to work, housing, education, and welfare an equal right for everyone. 4. Increase public employment in lagging and segregated areas! The localisation of publicly financed work, such as government agencies, have to a greater extent than previously been attributed to metropolitan areas. With localisation of universities into different regions outside the metropolitan areas, prospects of growth have become attainable. To go beyond universities and locate a greater share of public sector jobs outside the metropolitan areas is a decisive measure to offer stagnating labour markets greater diversity. A socially coherent po‑ licy will attain: – Localisation of public sector jobs in stagnating regions – Localisation of public sector jobs in segregated urban areas 5. EU programmes should focus on social cohesion! From an EU-perspective, the ambition to reduce the spatial structural imbalances within the EU is a key aspect of EU-cooperation. The EU budget has great potential to 3.4 CONCLUSION The transition from mixed capitalism and universial welfare towards monetaristic and neoliberal policy in recent decades has been reinforced. We have seen growing spatial inequality in the last three decades – which breakes a historical trend of convergence in income levels. The most important reason arises from the structural shift from manufacturing to services. Almost all expansion of the labour market has taken place in growing metropolitan areas, university cities and related areas where a shortage of housing, segregation and integration are major concerns today. Job losses, stagnating populations and financial difficulties have emerged as major concerns of local policy in other parts of the country. One important exception is mining – basic metal industries in the North that benefit from the transition to the green economy – battery factories, fossil-free steel.

FRIEDRICH-EBERT-STIFTUNG – POLITICS FOR EUROPE 18 Like Finland, we have a strong local-government level with 290 municipalities in 24 counties. However, the symmetry in governance has problems with regards to adressing assymetry in regional development – inter-municipal equalisation is not enough. To address the rising spatial inequality, a new policy needs to be formulated that should support the adaption to the ongoing structural transformation of the economy, and finds ways of helping all individuals make the most of their lives equally and take an active part in the value created in the economy. At the local level, the expanding labour market regions require greater investment in infrastructure, a more active and educational integration policy, and an active housing supply policy. In declining labour market regions, stronger mechanisms to mitigate the local financial deficits linked to ageing populations is needed. Investment in sustainable energy production is an opportunity to strengthen the local economy and reduce emissions. The national policy level needs to address spatial inequality by imposing equality in taxation and welfare provision, national support for investment in local infrastructure and housing. The local presence of national government needs to become stronger. On the EU level, investment in the green transition is needed. Human capital and supporting greater social cohesion in segregated areas are more effective measures to reduce the barriers of adaptation to the ongoing structural changes in the economy than structural funds supporting business.

ANNEX A 19 ANNEX A: Indicator documentation Indicator Employment rate Demographic dependency ratio Highly qualified workforce At-risk-of-poverty rate (age 0–19) Monthly gross salary Gender pay gap Medical and health related employment House prices Investment per capita Voter turnout Overall internal net migration balance Definition Source Gainfully employed persons as a percentage of the population of the municipality in the age group 20–64(2018) Statistikdatabasen(SCB) 2020a: Gainfully employed persons as a percentage of the population of the municipality in the age group 20–64(2018), https://www.statistikdatabasen.scb.se/pxweb/en/ssd/START__AM__AM0207__AM0207J/RAMSForvInt04/ (21.03.2021). Share of dependents aged zero to 14 and over the age of 65, compared with the t­otal population aged 15 to 64(2019) in per cent SCB 2020b: Population by region, marital status, age and sex. Year 1968–2020, https://www.statistikdatabasen.scb.se/pxweb/en/ssd/ START__BE__BE0101__BE0101A/BefolkningNy/(26.04.2021). Share of people with post-secondary ­education of all inhabitants(age 16+) (2018) in per cent SCB 2020c: Population 16–74 years of age by region, highest level of education, age and sex. Year 1985–2019, https://www.statistikdatabasen.scb.se/pxweb/en/ssd/START__UF__UF0506/Utbildning/(26.04.2021). Share of people(aged 0–19) with an ­equivalised disposable income below the at-risk-of-poverty threshold, which is set at 60% of the national median ­equivalised disposable income(2018) in per cent SCB 2020d: At-risk-of-poverty rate by region and age. Year 2015– 2019, https://www.statistikdatabasen.scb.se/pxweb/en/ssd/ START__HE__HE0110__HE0110I/Tab4InkDesoN/(26.04.2021). Average monthly nominal gross salary in SEK, at place of residence(2019) SCB 2020e: Average monthly salary in the primary municipal‑ ities by municipality and sex. Year 2007–2019, https://www.statistikdatabasen.scb.se/pxweb/en/ssd/START__AM__AM0106__ AM0106A/Kommun17g/(26.04.2021). Calculation: annual gross income women / annual gross income men × 100, at place of ­residence(2018) in per cent SCB 2020f: Gross pay(SEK) and number of statements of income distributed by region of work and residence and sex. Year 2005–2019, https://www.statistikdatabasen.scb.se/pxweb/en/ ssd/START__AM__AM0302__AM0302B/LSUMkommunkon/ (21.03.2021). Employees in the medical and health ­sector per 100 inhabitants over 65 years of age (2018) SCB 2020g: Gainfully employed 16+ years by region of residence (RAMS), by region, industry SNI2007 and sex. Year 2008–2018, https://www.statistikdatabasen.scb.se/pxweb/en/ssd/START__ AM__AM0207__AM0207J/NattSNI07KonK/(26.04.2021). Purchase price for permanent dwelling buildings(no site leasehold rights), average in SEK 1,000(2019) SCB 2020h: Sold one- and two-dwelling buildings by region and type of real estate. Year 1981–2019, https://www.statistikdatabasen.scb.se/pxweb/en/ssd/START__BO__BO0501__BO0501B/ FastprisSHRegionAr/(21.03.2021). Investment per capita by local government/ municipality in SEK 1,000(average of 2016 –2019) SCB 2020i: Investment in municipalities by region and activity. Year 2016–2020, https://www.statistikdatabasen.scb.se/pxweb/en/ssd/ START__OE__OE0107__OE0107B/InvestKommun/(26.04.2021). Voting rates the Riksdag election(2018) in per cent SCB 2020j: Voting rates in elections to the Riksdag, County Councils and Municipal Councils by region. Year of election 1973–2018, https://www.statistikdatabasen.scb.se/pxweb/en/ssd/START__ ME__ME0104__ME0104D/ME0104T4/(21.03.2021). Balance of all immigration and emigration between Swedish counties per 1,000 inhabitants(2014–2019) SCB 2020k: Migration by region, age and sex. Year 1997–2020, https://www.statistikdatabasen.scb.se/pxweb/en/ssd/START__ BE__BE0101__BE0101J/Flyttningar97/(26.04.2021).

FRIEDRICH-EBERT-STIFTUNG – POLITICS FOR EUROPE 20 ANNEX B: Methodological notes The spatial typology of Sweden was computed in a combined statistical procedure consisting of a prinicipal component and a cluster analysis. This procedure involves three steps. In the first step all variables were standardised by z-score transformation. Then, since many of the 11 selected disparity indicators are potentially correlated, a principal component analysis was conducted in order to reduce complexity and to avoid any potential bias caused by multicollinearity. The principal component analysis merges the initial selection of indicators to a lower number of uncorrelated“super-variables”, socalled principal components. The number of principal components chosen for the cluster analysis explains more than 90% of total variance in the data. In the final step, a hierarchical cluster analysis using the Ward method was conducted. In this procedure, the initial observations are hierarchically merged using a minimum variance criterion. The point where to stop the clustering procedure, and hence the resulting number of clusters, is chosen by the data analyst. Several solutions have been tested and discussed within the research team. The final typology of four clusters was selected based on its intuitiveness and relevance to identify spatial disparities in Sweden.

ANNEX 21 ANNEX C: Indicator value ranges Indicator Employment rate(%) Demographic dependency ratio(%) Highly qualified workforce(%) At-risk-of-poverty rate(age 0–19)(%) Monthly gross salary(SEK) Gender pay gap(annual gross income women / annual gross income men × 100)(%) Medical and health related employment(per 100 inhabitants over 65 years of age) House prices(in SEK 1,000) Investment per capita by local government/municipality (in SEK 1,000, average of 2016–2019) Voter turnout(%) Overall internal net migration balance(per 1,000 inhabitants) Year 2018 2019 2018 2018 2019 2018 Value range from... to... 66.4(Eda) to 89.1(Habo) 43.3(Sundbyberg) to 95.5(Pajala) 14.4(Filipstad) to 55.8(Lund) 5.0(Täby) to 45.0(Högsby) 28,200(Åsele) to 36,900(Sollentuna) 64.7(Danderyd) to 92.7(Sorsele) 2018 20.4(Båstad) to 66.2(Umeå) 2019 2016–2019 330(Åsele) to 11,209(Danderyd) 1,028(Laxå) to 29,472(Gällivare) 2018 2014–2019 72.8(Haparanda) to 93.9(Lomma) –126.0(Ljusnarsberg) to 119.7(Trosa)

FRIEDRICH-EBERT-STIFTUNG – POLITICS FOR EUROPE 22 Literature Andersson, L.-F.; Eriksson R. H.; Haneweihman E. 2018: Växande ­regionala obalanser, î: Ekonomisk debatt, 46(8), pp. 1–12. EU Fonder 2021: https://eufonder.se/(02.02.2021). Eurostat 2021: LOCAL ADMINISTRATIVE UNITS(LAU), https://ec.europa. eu/eurostat/web/nuts/local-administrative-units(02.02.2021). Hultkrantz, L.; Söderström, H. T.; Andersson, B. 2011: Marknad och politik, Stockholm. Lindell, Johanna; Pelling, Lisa 2021: Det svenska missnöjet, Stockholm. Nilsson, J.-E. 2012: Från aktiv lokaliseringspolitik till regional politik, ­Tillväxtanalys Working paper/PM 2012:18, Östersund, https://www.tillvaxtanalys.se/download/18.56c9e251715b3d6b3a2cc4c/1587454405576/ PM_2020_09_N%C3%A4ringslivets%20rapportering%20om%20milj%C3%B6relaterade%20risker%20i%20leverant%C3%B6rskedjan.pdf (02.02.2021). Nordmark, S.; Svensson, L. 2019: ­Temautvärdering ­Regionalfonden – en regional analys och nationell syntes, Tillväxtverket, ­Östersund, https://tillvaxtverket.se/download/18.4102188d16b93e0181d7929/1561625924283/TV%C3%84X022_Syntesrapport2019sista_190626.pdf(02.02.2021). Organisation for Economic Cooperation and Development(OECD) 2021: Statistical Database, National Accounts, https://stats.oecd.org/ (02.02.2021). OECD 2020: OECD Regions and Cities at a Glance 2020, Paris, https://www.oecd-ilibrary.org/docserver/959d5ba0-en.pdf?expires=1616603070&id=id&accname=guest&checksum=1C7DD18907C1AC8D1056B344BE82235F(02.02.2021). Pontarollo, N.; Serpieri, C. 2020: Challenges and Opportunities to ­Regional Renewal in the European Union, in: International Regional ­Science Review 44(1), pp. 142–169, https://journals.sagepub.com/doi/ pdf/10.1177/0160017620931591(02.02.2021). Statens Offentliga Utredningar(SOU) 2017: För Sveriges ­landsbygder – en sammanhållen politik för arbete, hållbar tillväxt och välfärd, ­Slutbetänkande av landsbygdskommitten, Stockholm. SOU 2020: Starkare kommuner– med kapacitet att klara välfärds‑ uppdraget, Slutbetänkande av Kommunutredningen, Stockholm. Swedish Statistics Agency(SCB) 2021a: Population, https://www. statistikdatabasen.scb.se/pxweb/sv/ssd/START__BE__BE0101__BE0101A/ FolkmangdNov/(02.02.2021). SCB 2021b: Taxation, https://www.statistikdatabasen.scb.se/pxweb/en/ ssd/(26.04.2021). SCB 2020c: Population 16–74 Years of Age by Region, Highest Level of ­Education, Age and Sex. Year 1985–2019, https://www.statistikdatabasen.scb.se/pxweb/en/ssd/START__UF__UF0506/Utbildning/ (26.04.2021). SCB 2020d: At-Risk-of-Poverty Rate by Region and Age. Year 2015– 2019, https://www.statistikdatabasen.scb.se/pxweb/en/ssd/START__HE__ HE0110__HE0110I/Tab4InkDesoN/(26.04.2021). SCB 2020e: Average Monthly Salary in the Primary Municipalities by Municipality and Sex. Year 2007–2019, https://www.statistikdatabasen.scb. se/pxweb/en/ssd/START__AM__AM0106__AM0106A/Kommun17g/ (26.04.2021). SCB 2020f: Gross Pay(SEK) and Number of Statements of Income Distri‑ buted by Region of Work and Residence and Sex. Year 2005–2019, https://www.statistikdatabasen.scb.se/pxweb/en/ssd/START__AM__ AM0302__AM0302B/LSUMkommunkon/(21.03.2021). SCB 2020g: Gainfully Employed 16+ Years by Region of Residence (RAMS), by Region, Industry SNI2007 and Sex. Year 2008–2018, https:// www.statistikdatabasen.scb.se/pxweb/en/ssd/START__AM__AM0207__ AM0207J/NattSNI07KonK/(26.04.2021). SCB 2020h: Sold One- and Two-Dwelling Buildings by Region and Type of Real Estate. Year 1981–2019, https://www.statistikdatabasen.scb.se/ pxweb/en/ssd/START__BO__BO0501__BO0501B/FastprisSHRegionAr/ (21.03.2021). SCB 2020i: Investment in Municipalities by Region and Activity. Year 2016–2020, https://www.statistikdatabasen.scb.se/pxweb/en/ssd/ START__OE__OE0107__OE0107B/InvestKommun/(26.04.2021). SCB 2020j: Voting Rates in Elections to the Riksdag, County Councils and Municipal Councils by Region. Year of Election 1973–2018, https:// www.statistikdatabasen.scb.se/pxweb/en/ssd/START__ME__ME0104__ ME0104D/ME0104T4/(21.03.2021). SCB 2020k: Migration by Region, Age and Sex. Year 1997–2020, https:// www.statistikdatabasen.scb.se/pxweb/en/ssd/START__BE__BE0101__ BE0101J/Flyttningar97/(26.04.2021). Widuto, A. 2019: Regional Inequalities in the EU, European Parliamentary Research Service, Brussels, http://www.europarl.europa.eu/RegData/ etudes/BRIE/2019/637951/EPRS_BRI(2019)637951_EN.pdf(02.02.2021). Åberg, R. 2013: Tjugohundratalets arbetsmarknad – fortsatt uppkvalificering eller jobbpolarisering?, in: Ekonomisk debatt, 41(2), pp. 6–15.

List of figures and tables 23 List of figures and tables 7 Figure 1 Local tax rate(2020) and population growth(1990–2020) 11 Figure 2 The Swedish disparity map 12 Table 1 Spatial typology of socioeconomic disparities in Sweden 14 Table 2 Bandwidth of indicator values for the spatial types 16 Table 3 Summary of spatial-based recommendations for action

Friedrich-Ebert-Stiftung www.fes.de The Friedrich-Ebert-Stiftung(FES) is the oldest political foundation in Germany with a rich tradition dating back to its foundation in 1925. Today, it remains loyal to the legacy of its namesake and campaigns for the core ideas and values of social democracy: freedom, justice and solidarity. It has close ties to social democracy and the free trade unions. FES promotes the advancement of social democracy, in particular through: – political educational work to strengthen civil society; – think tanks; – international cooperation with our international network of offices in more than 100 countries; – support for talented young people; – maintaining the collective memory of social democracy with archives, libraries and more. Arena Idé www.arenaide.se Arena Idé is a progressive, non-partisan think tank based in Stockholm, Sweden, focusing on labour market and economic policy, welfare and democracy issues. Arena Idé is part of the Arenagruppen and is funded by Swedish trade unions. Foundation for European Progressive Studies(FEPS) www.feps-europe.eu The Foundation for European Progressive Studies(FEPS) is the think tank of the social democratic political family at the EU level. Its mission is to develop innovative research, policy advice, training and debates to inspire and inform progressive politics and policies across Europe. FEPS operates as a hub for thinking to facilitate the emergence of progressive answers to the challenges that Europe faces. Today FEPS benefits from a solid network of 68 member organisations. In addition to this network of organisations that are active in the promotion of progressive values, FEPS also has an extensive network of partners, including renowned universities, scholars, policymakers and activists. ILS – Research Institute for Regional and Urban ­Development gGmbH www.ils-forschung.de/en/ The ILS is an urban research institute which explores the dynamics and diversity of urban change in international comparison. The aim of research is a better understanding of transformation processes, in order to gain insights into the sustainable development and design of urban areas of different size and scale. IMPRINT © 2021 FES Nordic Countries Dr. Philipp Fink Front cover/Design/ Typesetting: Heike Wächter Design concept: www.bergsee-blau.de The views expressed in this publication are not necessarily those of the FriedrichEbert-Stiftung and the partner organisations for this publication. Commercial use of all media published by the Friedrich-Ebert-Stiftung(FES) is not permitted without the written consent of the FES.

More Equality in Europe! The inability of democratic actors and procedures to provide rapid responses to social-economic issues has led to widespread disenchantment of political and democratic systems across Europe. As the benefits of economic growth and increasing employment have been unequally spread, therefore creating regional disparities, perceived and experienced social-economic inequalities and injustices have deepened and played into the hands of right-wing populists. But what are the answers to these challenges? How should policies in the EU member states and the EU tackle regional socio-economic disparities? With the project“Unequal Europe – Tackling Regional Disparities in Europe”, the Friedrich-Ebert-Stiftung and the Foundation for European Progressive Studies(FEPS), put forward progressive policy recommendations based on the disparity reports for several European countries for both the respective national and European level. https://fes.de/unequal-sweden