Devolution and the regional health divide: a longitudinal ecological study of 14 countries in Europe

Devolution and the regional health divide: a longitudinal ecological study of 14 countries in Europe Abstract Background Greater regional devolution can reduce economic inequalities between regions; however, the impact on health inequalities is not clear. We investigated the association between changes over time in the level of devolution in European countries and regional economic and health inequalities. Methods We used the proportion of government expenditure controlled by subnational levels of government as our measure of devolution in 14 European countries between 1995 and 2011. Fixed effects linear regression models were used to analyse trends in the level of devolution, trends in regional economic inequalities (Gini-coefficient) and trends in regional health inequalities (slope index) in life expectancy. Results Each additional percentage of government expenditure managed at subnational level reduced the Gini-coefficient of regional GDP by −0.17 points (95% CI: −0.33 to −0.01; P = 0.04). However, it increased the slope index of regional life expectancy by 23 days (95% CI: −2 to 48; P = 0.07). When trends in regional economic inequalities were controlled for, as a potential mediator—increased devolution—was significantly associated with an increase in health inequalities between regions (P = 0.01). Conclusions Increased devolution does not appear to reduce regional health inequalities—even when it reduces regional economic inequalities—and it could be associated with increased health inequalities. policy, economics, health, inequalities Introduction In Europe there has been a growing trend in recent years towards greater devolution, resulting in the delegation of power and resources from central to lower levels of government.1 One key aspect of this devolution has been fiscal devolution where the proportion of public expenditure controlled by subnational levels of government has increased. In the UK, for example, substantial powers and resources have been devolved to Scotland and Wales in recent years2 and giving city regions greater control over resources is central to the UK government’s strategy to stimulate economic growth in less economically developed regions.3 In Denmark in 2007, large reforms devolved resources and gave additional competences to local and regional authorities.4 The federalism rules introduced in Italy in 2009 increased the fiscal autonomy of the regions.5 Economic differences between regions are a major cause of health inequalities in Europe, with poorer regions within countries experiencing worse health than more wealthy regions.6 Strategies which reduce these economic inequalities could therefore have a beneficial impact on regional health inequalities. There are two opposing arguments which indicate the pathways in which fiscal devolution may influence regional economic inequalities. Firstly, it is argued that local decision makers have a greater understanding of local needs, are more accountable to local populations, and have opportunities for innovation, thus devolution can result in more efficient and effective services.7,8 Where savings are made, resources can be re-invested locally without the inefficiencies that could be introduced by the involvement of central government. Secondly, the opposite argument is fiscal devolution could increase economic inequalities by diminishing the scope of national governments to redistribute resources from wealthier regions to poorer regions.9–11 The two pathways outlined above also have implications for health inequalities. More effective and efficient local services could help reduce regional health inequalities, whilst reduced redistribution of resources could increase health inequalities (Fig. 1). For example, the Greater Manchester Health and Social Care Devolution Plan proposes to reduce inequalities through redesign of services, including integration across sectors, tailoring to local needs and upgrading prevention.12 Fig. 1 View largeDownload slide Scatter plots showing the association between the average annual change in the level of devolution in each country and (i) the average annual change in regional economic inequalities and (ii) the average annual change in regional health inequalities. Number of years included on the analysis for each country shown in key below. AT – Austria (17 years), CZ – Czech Republic (17 years), DE – Germany (12 years), DK – Denmark (12 years), FI – Finland (17 years), FR – France (17 years), ES – Spain (17 years), HU – Hungry (16 years), IT – Italy (17 years), NO – Norway (15 years), PL – Poland (2 years), PT – Portugal (17 years), SE – Sweden (17 years), UK – United Kingdom (17 years). Fig. 1 View largeDownload slide Scatter plots showing the association between the average annual change in the level of devolution in each country and (i) the average annual change in regional economic inequalities and (ii) the average annual change in regional health inequalities. Number of years included on the analysis for each country shown in key below. AT – Austria (17 years), CZ – Czech Republic (17 years), DE – Germany (12 years), DK – Denmark (12 years), FI – Finland (17 years), FR – France (17 years), ES – Spain (17 years), HU – Hungry (16 years), IT – Italy (17 years), NO – Norway (15 years), PL – Poland (2 years), PT – Portugal (17 years), SE – Sweden (17 years), UK – United Kingdom (17 years). Yet, the relative importance of these two pathways could be different for health inequalities than for economic inequalities. For example, the ability of central governments to target resources at more disadvantaged regions was an effective part of the strategy to reduce health inequalities in England.13,14 The benefits of this may outweigh those resulting from the increased efficiency and effectiveness that devolution brings. Devolution could also have an impact on health inequalities indirectly through its effect on economic inequalities. The literature to date relating fiscal devolution to regional inequalities has focused on economic differences between regions, with much of the evidence indicating greater fiscal devolution is associated with reduced economic inequalities between regions.3,10,11,15–17 This suggests, at least for regional economic inequalities, that the increased efficiency and effectiveness devolution brings outweighs the negative effect from reduced redistribution. We do not know whether fiscal devolution across Europe has led to either a narrowing or a widening of health differences between regions. The studies which have investigated the health inequalities effect of devolution in high income countries have focused on devolution in healthcare expenditure and reported equivocal results.18,19 With 10–20% of variations in population health estimated as attributable to the health sector, most differences in health are due to actions in other sectors known to be key determinants of health such as education, employment, environment and housing.20 Therefore, devolution could have positive or negative effects on regional health inequalities through its impact on services across multiple sectors. In this study, therefore, we used data from a panel of 14 European countries to investigate whether trends in fiscal devolution across European countries are associated with trends in regional health inequalities. We then explored the extent to which changes in regional economic inequalities mediate any effects of fiscal devolution on regional health inequalities. Method Setting, data sources and measures We measured the level of fiscal devolution in each country as the proportion of total government expenditure managed at a subnational level (state, region or local) obtained from the Organization for Economic Co-operation and Development (OECD) fiscal decentralization database.21 To measure regional health inequalities, we calculated the slope index of inequality (SII) of regional life expectancy for each country and each year. This measure indicates the absolute difference in life expectancy between the regions with the lowest and highest Gross Domestic Product (GDP) per capita within each country. It was calculated as the population weighted linear regression slope of regional life expectancy and regional GDP per capita, with regional GDP scaled between 0 and 1.22 Annual data on regional life expectancy and regional GDP per capita were obtained from the OECD for all Territorial Level 2 (TL2) regions within each country. GDP per capita was measured in USD per head, at constant Purchasing Power Parity (PPP) (base year 2010). The OECD TL2 regions for European countries are the similar to the Eurostat NUTS 2 level classification. We used this level as it represents the first administrative tier of subnational government, and in most European countries this is the level at which fiscal devolution has occurred.5 To measure regional economic inequality, we calculated the annual Gini-coefficient of regional GDP per capita,23 using the STATA package Inequa7.24 Using the approach applied by the World Bank,25 we scaled the Gini-coefficient from 0 to 100, with 0 representing perfect equality (meaning all regions have the same GDP) whilst 100 represents complete inequality (one region has all the GDP). To control for national-level economic trends that may confound the effect of devolution on health inequalities,26–28 we included annual data on national GDP and national unemployment rates obtained from the OECD and EUROSTAT, respectively.23,29 Regions that were autonomous protectorates were excluded (Azores and Madeira in Portugal, and Ceuta and Melilla in Spain). To maximize the generalizability of our analysis to Europe we included all European countries with greater than five regions and with available data between 1995 and 2011. The analysis was therefore based on data from 14 countries: Austria, Czech Republic, Denmark, Finland, France, Germany, Hungary, Italy, Norway, Poland, Portugal, Spain, Sweden and the UK. For nine countries, data were available for all 17 years between 1995 and 2011 (Austria, Czech Republic, Finland, France, Italy, Portugal, Spain, Sweden and UK). Data were only available for 16 years for Hungary, 15 years for Norway, 12 years for Germany and Denmark and 2 years for Poland. Details of country-years with missing data are given in web Appendix 1. In total, 210 country-years were available for analysis. Analysis To explore the data visually, we calculated the average absolute annual change in the SII in regional life expectancy, the level of fiscal devolution and the Gini-coefficient of regional GDP. We then graphed scatter-plots of the association between (i) the change in the level of devolution and the change in the Gini-coefficient of regional GDP and (ii) the change in the level of devolution and the change in the SII. To adjust for the potential confounders of these associations we estimated fixed effects linear regression models, additionally controlling for national economic trends using the annual national GDP and unemployment rate for each country. Fixed effects models essentially include a dummy variable for each country controlling for time-invariant differences between the countries that could act as confounders, so that the models assess the association between absolute trends in devolution and regional economic and health inequalities within countries.30 Firstly, we included the Gini-coefficient of regional GDP as the outcome and investigated the association between trends in fiscal devolution and the trend in regional economic inequality. Secondly, we included the SII of regional life expectancy as the outcome to investigate the association between trends in fiscal devolution and the trend in regional health inequality. Thirdly, we additionally controlled for the trend in the Gini-coefficient of regional GDP—whilst including the SII of regional life expectancy as the outcome—to explore the extent to which changes in regional economic inequalities mediated the effect of devolution on regional health inequalities. All models were weighted for population size, included time trend terms to adjust for secular trends across Europe, and used robust clustered standard errors. Statistical analyses were performed using STATA11 (full model formulae are given in web Appendix 2). Sensitivity tests We undertook several sensitivity tests to investigate the robustness of these models. To exclude any effects from the economic crisis, we limited our analyses to data prior to 2008. Because control over tax revenue—as opposed to control over expenditure—may also be an important indicator of fiscal devolution,31 we replicated our analysis using the proportion of taxation raised at the subnational level as our indicator of fiscal devolution. To take account of delayed effects of changes in devolution, we replicated models with our measure of fiscal devolution lagged by 0–3 years. As bias could result from the unbalanced nature of the panel, we repeated analysis only including countries with the full 17 years of data. Results Between 1995 and 2011, across the study countries, the average proportion of expenditure controlled by subnational governments increased slightly from 27 to 33% (see web Appendix 3). During this time those countries which tended to experience a greater increase in devolution also tended to experience a greater decline in economic differences between regions (as measured by the Gini-coefficient of regional GDP, r = 0.29, P ≤ 0.001). However, those countries which experienced a greater increase in devolution also tended to experience a greater increase in health inequalities between regions (r = 0.47, P ≤ 0.001) (Fig. 2). Fig. 2 View largeDownload slide Logic model relating fiscal devolution to regional economic inequalities and regional health inequalities. Fig. 2 View largeDownload slide Logic model relating fiscal devolution to regional economic inequalities and regional health inequalities. The multivariable regression analyses indicated broadly similar associations when controlling for economic trends in each country and country fixed effects. As Table 1 shows, each additional percentage of government expenditure managed at the subnational level reduced the Gini-coefficient of regional GDP by −0.17 points (95% CI: −0.33 to −0.01; P = 0.04). With each additional percentage of government expenditure managed at the subnational level the SII of regional life expectancy increased by 23 days (95% CI: −2 to 48), although this was just below significance at the 5% level (P = 0.07). When additionally controlling for the Gini-coefficient of regional GDP in this model, however, the association increased and was significant (P = 0.01). This suggests the association between increased devolution and higher health inequalities was mitigated to some extent through the effect of increased devolution reducing regional economic inequalities. When the mediating pathway between devolution and health inequalities via economic inequalities was held constant (by controlling for it in the analysis), we saw a larger significant association between increased devolution and increased regional health inequalities. Table 1 Association between increased devolution and change in regional economic inequalities (model 1) and regional health inequalities (model 2). Model 3 shows change in regional health inequalities associated with increased devolution, when controlling for change in regional economic inequalities   Model  Coefficient  95% CI  P-value  Change in regional Gini for each additional 1% of expenditure controlled at the subnational level.  1  −0.17  −0.33  −0.01  0.04  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level.  2  23.28  −1.67  48.23  0.07  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level. (Additionally controlling for regional Gini.)  3  31.95  7.37  56.53  0.01    Model  Coefficient  95% CI  P-value  Change in regional Gini for each additional 1% of expenditure controlled at the subnational level.  1  −0.17  −0.33  −0.01  0.04  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level.  2  23.28  −1.67  48.23  0.07  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level. (Additionally controlling for regional Gini.)  3  31.95  7.37  56.53  0.01  Note: Models based on equations C shown in online Supplementary Appendix 2 and included controls for country fixed effects, time trends, national unemployment rate and national GDP, full model results are given in online Supplementary Appendix 3. Table 1 Association between increased devolution and change in regional economic inequalities (model 1) and regional health inequalities (model 2). Model 3 shows change in regional health inequalities associated with increased devolution, when controlling for change in regional economic inequalities   Model  Coefficient  95% CI  P-value  Change in regional Gini for each additional 1% of expenditure controlled at the subnational level.  1  −0.17  −0.33  −0.01  0.04  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level.  2  23.28  −1.67  48.23  0.07  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level. (Additionally controlling for regional Gini.)  3  31.95  7.37  56.53  0.01    Model  Coefficient  95% CI  P-value  Change in regional Gini for each additional 1% of expenditure controlled at the subnational level.  1  −0.17  −0.33  −0.01  0.04  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level.  2  23.28  −1.67  48.23  0.07  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level. (Additionally controlling for regional Gini.)  3  31.95  7.37  56.53  0.01  Note: Models based on equations C shown in online Supplementary Appendix 2 and included controls for country fixed effects, time trends, national unemployment rate and national GDP, full model results are given in online Supplementary Appendix 3. Our results were broadly comparable when limiting data to prior to the economic crisis (2008), or using the proportion of taxation raised at the subnational level as our indicator of fiscal devolution, or only including countries with the full 17 years of data. The lag models indicated that the association between fiscal devolution and reduced economic inequality was greatest for a lag of 2 years. For the other models, there the association was greatest with the fiscal devolution variable lagged by 1 year (see web Appendix 4). Discussion Main findings As has been the case in previous studies10,11 we found increased devolution was associated with reduced regional economic inequalities. However, the impact on health inequalities was not straightforward. We found increases in devolution were associated with increased regional health inequalities, although this was only significant if trends in regional economic inequalities were controlled for. This suggests the direct effect of increased devolution was to increase heath inequalities, but because devolution also tended to reduce regional economic inequalities, this mitigated some of the direct negative impacts of devolution on regional health inequalities. What is already known There has been a growing trend in Europe towards greater fiscal devolution.1 Several studies have shown greater fiscal devolution is associated with decreased economic inequalities between regions within countries.3,10,11,15–17 It is thought localized decision-making may lead to more effective and efficient use of resources as it is informed by local knowledge about needs and is accountable to local populations and this may contribute to reduced regional economic inequalities.9 We do not know the health inequalities impact of devolution policy. The few studies which have investigated the health inequalities effect of devolution in high income countries have tended to investigate the effects of devolution on healthcare expenditure. An Italian study,32 found a reform which devolved a large proportion of healthcare expenditure was associated with a reduction in inequalities in self-reported health within regions, but it did not investigate the effect on inequalities between regions. Costa-Font and Gil18 found no association between devolution of healthcare expenditure in Spain and inequalities in reported health status. A Canadian study found healthcare devolution was followed by an increase in inequalities in health status.19 However, although we know public investment across multiple sectors (education, employment, environment and housing) has the potential to influence health, we do not know whether increased fiscal devolution of public expenditure in general is associated with a decrease or an increase in health differences between regions within European countries. What this study adds Our results indicate that whilst devolution may be an effective approach to reducing regional economic inequalities, this does not necessarily translate into reduced health inequalities. In fact, the opposite may be true, and devolution may increase health inequalities between regions. There are several reasons why fiscal devolution could be associated with reduced regional differences in GDP whilst being associated with increased regional differences in life expectancy. Firstly, it could be that the ability to redistribute resources from wealthier regions to poorer regions is particularly important for addressing health inequalities and this capacity is reduced in more devolved administrations, leading to under-funded health and social services in poorer regions. Other research has shown that centrally planned targeting of resources to disadvantaged areas has been associated with reduced health inequalities.13,14 Secondly, greater local control over public investment decisions may lead to local legislation and investment in projects which have primarily economic benefits and limited health benefits. Investment in infrastructure projects, attracting outside investment or public popularity may be prioritized above health; and thus, economic efficiencies may not translate into population health benefits. Local governments may also be more susceptible to lobbying from corporations and local businesses, undermining their capacity to implement public health legislation—such as controlling access to alcohol or unhealthy foods, or reducing environmental pollution. Whilst local businesses may seek to reduce regulations that have public health benefits but are seen as a barrier to economic growth.33 Finally, it could also be that the reductions in regional differences in GDP associated with increased fiscal devolution do not represent a reduction in inequalities in the life chances of the populations living in different regions. For example, in more devolved administrations where there is greater regional control over the setting of tax rates, poorer regions may try and compete with other regions to attract external investment by reducing taxes. This could lead to increased capital flows into poorer regions—reducing regional differences in GDP, however, these increases in regional GDP may not translate into widespread improvements in wages.34 Furthermore, the reduced levels of taxation in these regions then reduce income for public service provision, further widening inequalities in life chances.35 Our findings have important implications for policies that aim to devolve greater resources to subnational levels. They indicate that even if such policies reduce economic differences between regions, they could still increase health differences. Therefore, policies should not be purely focused on economic growth or just on reducing economic inequalities between regions. Mitigating actions will need to be taken to ensure heath inequalities are not increased by these policies and to maximize the potential for such policies to reduce health inequalities.36 Strengths and limitations of this study Our findings need to be understood in the context of several limitations. Firstly, because our study analysed data at the aggregate level, we cannot know whether the findings reflect associations at the individual level. For example, it could be that whilst there was an association between increased devolution and increased regional health inequalities, health inequalities could have concurrently reduced within regions as was found in the study of healthcare devolution in Italy.32 Secondly, the measure of devolution we used—the proportion of government expenditure managed at the subnational level—only reflects a component of devolution. In practice, the nature of devolution differs greatly between countries, with differences in the level of devolution between tiers of government and between regions. Models vary between federal governments with three tiers, such as Belgium, to unitary governments with one level but with some autonomous regions, such as in Portugal.37 In some countries, devolution relates primarily to decisions over expenditure, in others to tax raising powers, and in others to legislative freedoms. In this study, we have focused on one measure of fiscal devolution, and we have not been able to investigate differential impacts of alternative approaches to devolution across the study countries. However, our findings did not change when using the proportion of tax revenue raised at the subnational level (rather than the level of expenditure controlled at the subnational level) as the measure of devolution. Thirdly, due to limitations in the data available the sample size was relatively small; therefore, the analysis may have been underpowered and our estimates are relatively imprecise. Fourthly, it is possible the associations we observed were due to other confounding factors that we were not able to take into account in the analysis. However, by analysing change over time across countries, our analysis provides more robust evidence than simple cross-sectional analyses. This enabled us to control for potentially unobserved time-invariant confounders that could vary between countries. Finally, our investigation of whether regional economic inequalities mediate the effect of fiscal devolution on regional health inequalities assumes trends in regional health inequalities are not a major cause of regional economic inequalities. If this was the case, we would be conditioning on a common effect of exposure and outcome and the increased association observed in model 3 could then be the result of collider bias.38 Conclusion Across the 14 European countries included in our analysis, whilst increased fiscal devolution was associated with decreased regional economic inequalities, it was associated with increased regional health inequalities. Policies which promote increased fiscal devolution should aim to minimize any negative impacts on health inequalities and to maximize positive impacts. This could include ensuring that devolution is combined with policies for inclusive economic growth39 and a progressive system for redistribution of resources both between regions and between individuals through a comprehensive welfare system.22 Supplementary data Supplementary data are available at the Journal of Public Health online. References 1 Polverari L. Does devolution increase accountability? Empirical evidence from the implementation of European Union Cohesion Policy. Reg Stud  2015; 49: 1074– 86. Google Scholar CrossRef Search ADS   2 Local Government Association. Briefing—Greater Manchester Health and Social Care Partnership, 2015. http://www.gmhsc.org.uk/ (accessed 5 March 2015, date last accessed). 3 City Growth Commission. Unleashing Metro Growth Final Recommendations of the City Growth Commission. London; 2014. 4 Blöchliger H, Vammalle C. Reforming Fiscal Federalism and Local Government: Beyond the Zero-Sum Game . Paris: OECD Publishing, 2012. Google Scholar CrossRef Search ADS   5 Council of European Municipalities and Regions. Decentralisation at a Crossroads: Territorial Reforms in Europe in Times of Crisis . Brussels: Council of European Municipalities and Regions, 2013. 6 Marmot M. Health Inequalities in the EU: Final Report of a Consortium . Brussels: European Union, 2013. 7 Robalino DA, Picazo OF, Voetberg A. Does Fiscal Decentralization Improve Health Outcomes? Evidence from a Cross-Country Analysis . Washington: World Bank, 2001. Google Scholar CrossRef Search ADS   8 Oates WE. On the Theory and Practice of Fiscal Decentralization . Lexington: Institute for Federalism and Intergovernmental Relations, 2006. 9 Asfaw A, Frohberg K, James KS et al.   Modeling the Impact of Fiscal Decentralization on Health Outcomes: Empirical Evidence from India . Bonn. Center for Development Research, University of Bonn; 2004. 10 Canaleta CG, Arzoz PP, Garate MP. Regional economic disparities and decentralisation. Urban Stud  2004; 41: 71– 94. Google Scholar CrossRef Search ADS   11 Lessmann C. Regional inequality and decentralization: an empirical analysis. Environ Plann A  2012; 44: 1363– 88. Google Scholar CrossRef Search ADS   12 Greater Manchester Combined Authority. Taking Charge of our Health and Social Care in Greater Manchester: The Plan. Manchester, 2015. 13 Barr B, Bambra C, Whitehead M. The impact of NHS resource allocation policy on health inequalities in England 2001–11: longitudinal ecological study. Br Med J  2014; 348: g3231. Google Scholar CrossRef Search ADS   14 Barr B, Higgerson J, Whitehead M. Investigating the impact of the English health inequalities strategy: time trend analysis. Br Med J  2017; 358: j3310. Google Scholar CrossRef Search ADS   15 Beila J, Hennl A, Kaiser A. Combining federalism and decentralization: comparative case studies on regional development policies in Switzerland, Austria, Denmark, and Ireland. Comp Polit Stud  2012; 45: 447– 76. Google Scholar CrossRef Search ADS   16 Ezcurra R, Pascual P. Fiscal decentralization and regional disparities: evidence from several European Union countries. Environ Plann  2008; 40: 1185– 1201. Google Scholar CrossRef Search ADS   17 Governatori M, Yim D Fiscal Decentralisation and Fiscal Outcomes . European Commission, Directorate-General for Economic and Financial Affairs; Brussels, 2012. 18 Costa-Font J, Gil J. Exploring the Pathways of Inequality in Health, Access and Financing in Decentralised Spain . London: London School of Economics and Political Science, 2008. 19 Zhong H. The impact of decentralization of health care administration on equity in health and health care in Canada. Int J Health Care Finance Econ  2010; 10: 219– 37. Google Scholar CrossRef Search ADS PubMed  20 Health Affairs. Health Policy Brief. The Relative Contribution of Multiple Determinants to Health Outcomes . Maryland: Health Affairs and Robert Wood Johnson Foundation, 2014. 21 OECD. OECD Fiscal Decentralisation Database. [Online]. http://www.oecd.org/tax/federalism/oecdfiscaldecentralisationdatabase.htm#C_3 (23 March 2015, date last accessed). 22 Mackenbach J, Kunst E. Measuring the magnitude of socio-economic inequalities in health: an overview of available measures illustrated with two examples from Europe. Soc Sci Med  1997; 44: 757– 71. Google Scholar CrossRef Search ADS PubMed  23 OECD. Regional Statistics and Indicators. [Online]. http://www.oecd.org/cfe/regional-policy/regionalstatisticsandindicators.htm. (23 March 2015, date last accessed). 24 Van Kerm P Stata Module to Compute Measures of Inequality, Ideas . Boston College, Department of Economics; 2007. https://ideas.repec.org/c/boc/bocode/s416401.html (13 October 2016, date last accessed). 25 The World Bank. GINI Index (World Bank Estimate). [Online]. https://data.worldbank.org/indicator/SI.POV.GINI (25 June 2015, date last accessed). 26 Cribb J, Joyce R, Phillip D. Living Standards, Poverty and Inequality in the UK: 2012 . London: The Institute for Fiscal Studies, 2012. 27 Council of European Municipalities and Regions. Crisis and Decentralisation. 2014. http://www.ccre.org/en/activites/view/13 (7 August 2015, date last accessed). 28 Owyang MT, Sekhposyan T, Vermann EK. Output and unemployment how do they relate today? Reg Economist  2013; 21: 4– 9. 29 Eurostat. Labour Markets at Regional Level. [Online] http://ec.europa.eu/eurostat/statistics-explained/index.php/Labour_market_statistics_at_regional_level (12 October 2015, date last accessed). 30 Torres-Reyna O. Panel Data Analysis Fixed and Random Effects Using Stata . Princeton: Princeton University, 2007. 31 Jiménez-Rubio D. The impact of fiscal decentralization on infant mortality rates: evidence from OECD countries. Soc Sci Med  2011; 73: 1401– 7. Google Scholar CrossRef Search ADS PubMed  32 Di Novi C, Piacenza M, Robone S et al.  . How Does Fiscal Decentralization Affect Within-Regional Disparities in Well-being? Evidence From Health Inequalities in Italy . York: University of York, 2015. 33 Katikireddi SV, Smith KE, Stuckler D et al.  . Devolution of power, revolution in public health? J Public Health (Bangkok)  2017; 39: 241– 7. 34 Alves J, Peralta S, Perelman J. Efficiency and equity consequences of decentralization in health: an economic perspective. Rev Port de Saúde Pública  2013; 31: 74– 83. Google Scholar CrossRef Search ADS   35 Social Mobility Commission. State of the Nation 2017: Social Mobility in Great Britain . London: Social Mobility Commission, 2017. 36 OECD. OECD Territorial Reviews: Sweden 2010 . Paris: OECD Publishing, 2010. 37 Allain-Dupré D Decentralisation Trends in OECD Countries: A Comparative Perspective for Ukraine [Presentation]: OECD, Regional Development Policy Division; 2016. 38 Cole SR, Platt RW, Schisterman EF et al.  . Illustrating bias due to conditioning on a collider. Int J Epidemiol  2010; 39: 417– 20. Google Scholar CrossRef Search ADS PubMed  39 Saunders M, Barr B, McHale P et al.   Key Policies for Addressing the Social Determinants of Health and Health Inequities. Health Evidence Network (HEN), Synthesis Report 52. Copenhagen. WHO Regional Office for Europe. 2017. © The Author(s) 2018. Published by Oxford University Press on behalf of Faculty of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Public Health Oxford University Press

Devolution and the regional health divide: a longitudinal ecological study of 14 countries in Europe

Loading next page...
 
/lp/ou_press/devolution-and-the-regional-health-divide-a-longitudinal-ecological-KjbMGlfbEF
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of Faculty of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
1741-3842
eISSN
1741-3850
D.O.I.
10.1093/pubmed/fdy014
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Greater regional devolution can reduce economic inequalities between regions; however, the impact on health inequalities is not clear. We investigated the association between changes over time in the level of devolution in European countries and regional economic and health inequalities. Methods We used the proportion of government expenditure controlled by subnational levels of government as our measure of devolution in 14 European countries between 1995 and 2011. Fixed effects linear regression models were used to analyse trends in the level of devolution, trends in regional economic inequalities (Gini-coefficient) and trends in regional health inequalities (slope index) in life expectancy. Results Each additional percentage of government expenditure managed at subnational level reduced the Gini-coefficient of regional GDP by −0.17 points (95% CI: −0.33 to −0.01; P = 0.04). However, it increased the slope index of regional life expectancy by 23 days (95% CI: −2 to 48; P = 0.07). When trends in regional economic inequalities were controlled for, as a potential mediator—increased devolution—was significantly associated with an increase in health inequalities between regions (P = 0.01). Conclusions Increased devolution does not appear to reduce regional health inequalities—even when it reduces regional economic inequalities—and it could be associated with increased health inequalities. policy, economics, health, inequalities Introduction In Europe there has been a growing trend in recent years towards greater devolution, resulting in the delegation of power and resources from central to lower levels of government.1 One key aspect of this devolution has been fiscal devolution where the proportion of public expenditure controlled by subnational levels of government has increased. In the UK, for example, substantial powers and resources have been devolved to Scotland and Wales in recent years2 and giving city regions greater control over resources is central to the UK government’s strategy to stimulate economic growth in less economically developed regions.3 In Denmark in 2007, large reforms devolved resources and gave additional competences to local and regional authorities.4 The federalism rules introduced in Italy in 2009 increased the fiscal autonomy of the regions.5 Economic differences between regions are a major cause of health inequalities in Europe, with poorer regions within countries experiencing worse health than more wealthy regions.6 Strategies which reduce these economic inequalities could therefore have a beneficial impact on regional health inequalities. There are two opposing arguments which indicate the pathways in which fiscal devolution may influence regional economic inequalities. Firstly, it is argued that local decision makers have a greater understanding of local needs, are more accountable to local populations, and have opportunities for innovation, thus devolution can result in more efficient and effective services.7,8 Where savings are made, resources can be re-invested locally without the inefficiencies that could be introduced by the involvement of central government. Secondly, the opposite argument is fiscal devolution could increase economic inequalities by diminishing the scope of national governments to redistribute resources from wealthier regions to poorer regions.9–11 The two pathways outlined above also have implications for health inequalities. More effective and efficient local services could help reduce regional health inequalities, whilst reduced redistribution of resources could increase health inequalities (Fig. 1). For example, the Greater Manchester Health and Social Care Devolution Plan proposes to reduce inequalities through redesign of services, including integration across sectors, tailoring to local needs and upgrading prevention.12 Fig. 1 View largeDownload slide Scatter plots showing the association between the average annual change in the level of devolution in each country and (i) the average annual change in regional economic inequalities and (ii) the average annual change in regional health inequalities. Number of years included on the analysis for each country shown in key below. AT – Austria (17 years), CZ – Czech Republic (17 years), DE – Germany (12 years), DK – Denmark (12 years), FI – Finland (17 years), FR – France (17 years), ES – Spain (17 years), HU – Hungry (16 years), IT – Italy (17 years), NO – Norway (15 years), PL – Poland (2 years), PT – Portugal (17 years), SE – Sweden (17 years), UK – United Kingdom (17 years). Fig. 1 View largeDownload slide Scatter plots showing the association between the average annual change in the level of devolution in each country and (i) the average annual change in regional economic inequalities and (ii) the average annual change in regional health inequalities. Number of years included on the analysis for each country shown in key below. AT – Austria (17 years), CZ – Czech Republic (17 years), DE – Germany (12 years), DK – Denmark (12 years), FI – Finland (17 years), FR – France (17 years), ES – Spain (17 years), HU – Hungry (16 years), IT – Italy (17 years), NO – Norway (15 years), PL – Poland (2 years), PT – Portugal (17 years), SE – Sweden (17 years), UK – United Kingdom (17 years). Yet, the relative importance of these two pathways could be different for health inequalities than for economic inequalities. For example, the ability of central governments to target resources at more disadvantaged regions was an effective part of the strategy to reduce health inequalities in England.13,14 The benefits of this may outweigh those resulting from the increased efficiency and effectiveness that devolution brings. Devolution could also have an impact on health inequalities indirectly through its effect on economic inequalities. The literature to date relating fiscal devolution to regional inequalities has focused on economic differences between regions, with much of the evidence indicating greater fiscal devolution is associated with reduced economic inequalities between regions.3,10,11,15–17 This suggests, at least for regional economic inequalities, that the increased efficiency and effectiveness devolution brings outweighs the negative effect from reduced redistribution. We do not know whether fiscal devolution across Europe has led to either a narrowing or a widening of health differences between regions. The studies which have investigated the health inequalities effect of devolution in high income countries have focused on devolution in healthcare expenditure and reported equivocal results.18,19 With 10–20% of variations in population health estimated as attributable to the health sector, most differences in health are due to actions in other sectors known to be key determinants of health such as education, employment, environment and housing.20 Therefore, devolution could have positive or negative effects on regional health inequalities through its impact on services across multiple sectors. In this study, therefore, we used data from a panel of 14 European countries to investigate whether trends in fiscal devolution across European countries are associated with trends in regional health inequalities. We then explored the extent to which changes in regional economic inequalities mediate any effects of fiscal devolution on regional health inequalities. Method Setting, data sources and measures We measured the level of fiscal devolution in each country as the proportion of total government expenditure managed at a subnational level (state, region or local) obtained from the Organization for Economic Co-operation and Development (OECD) fiscal decentralization database.21 To measure regional health inequalities, we calculated the slope index of inequality (SII) of regional life expectancy for each country and each year. This measure indicates the absolute difference in life expectancy between the regions with the lowest and highest Gross Domestic Product (GDP) per capita within each country. It was calculated as the population weighted linear regression slope of regional life expectancy and regional GDP per capita, with regional GDP scaled between 0 and 1.22 Annual data on regional life expectancy and regional GDP per capita were obtained from the OECD for all Territorial Level 2 (TL2) regions within each country. GDP per capita was measured in USD per head, at constant Purchasing Power Parity (PPP) (base year 2010). The OECD TL2 regions for European countries are the similar to the Eurostat NUTS 2 level classification. We used this level as it represents the first administrative tier of subnational government, and in most European countries this is the level at which fiscal devolution has occurred.5 To measure regional economic inequality, we calculated the annual Gini-coefficient of regional GDP per capita,23 using the STATA package Inequa7.24 Using the approach applied by the World Bank,25 we scaled the Gini-coefficient from 0 to 100, with 0 representing perfect equality (meaning all regions have the same GDP) whilst 100 represents complete inequality (one region has all the GDP). To control for national-level economic trends that may confound the effect of devolution on health inequalities,26–28 we included annual data on national GDP and national unemployment rates obtained from the OECD and EUROSTAT, respectively.23,29 Regions that were autonomous protectorates were excluded (Azores and Madeira in Portugal, and Ceuta and Melilla in Spain). To maximize the generalizability of our analysis to Europe we included all European countries with greater than five regions and with available data between 1995 and 2011. The analysis was therefore based on data from 14 countries: Austria, Czech Republic, Denmark, Finland, France, Germany, Hungary, Italy, Norway, Poland, Portugal, Spain, Sweden and the UK. For nine countries, data were available for all 17 years between 1995 and 2011 (Austria, Czech Republic, Finland, France, Italy, Portugal, Spain, Sweden and UK). Data were only available for 16 years for Hungary, 15 years for Norway, 12 years for Germany and Denmark and 2 years for Poland. Details of country-years with missing data are given in web Appendix 1. In total, 210 country-years were available for analysis. Analysis To explore the data visually, we calculated the average absolute annual change in the SII in regional life expectancy, the level of fiscal devolution and the Gini-coefficient of regional GDP. We then graphed scatter-plots of the association between (i) the change in the level of devolution and the change in the Gini-coefficient of regional GDP and (ii) the change in the level of devolution and the change in the SII. To adjust for the potential confounders of these associations we estimated fixed effects linear regression models, additionally controlling for national economic trends using the annual national GDP and unemployment rate for each country. Fixed effects models essentially include a dummy variable for each country controlling for time-invariant differences between the countries that could act as confounders, so that the models assess the association between absolute trends in devolution and regional economic and health inequalities within countries.30 Firstly, we included the Gini-coefficient of regional GDP as the outcome and investigated the association between trends in fiscal devolution and the trend in regional economic inequality. Secondly, we included the SII of regional life expectancy as the outcome to investigate the association between trends in fiscal devolution and the trend in regional health inequality. Thirdly, we additionally controlled for the trend in the Gini-coefficient of regional GDP—whilst including the SII of regional life expectancy as the outcome—to explore the extent to which changes in regional economic inequalities mediated the effect of devolution on regional health inequalities. All models were weighted for population size, included time trend terms to adjust for secular trends across Europe, and used robust clustered standard errors. Statistical analyses were performed using STATA11 (full model formulae are given in web Appendix 2). Sensitivity tests We undertook several sensitivity tests to investigate the robustness of these models. To exclude any effects from the economic crisis, we limited our analyses to data prior to 2008. Because control over tax revenue—as opposed to control over expenditure—may also be an important indicator of fiscal devolution,31 we replicated our analysis using the proportion of taxation raised at the subnational level as our indicator of fiscal devolution. To take account of delayed effects of changes in devolution, we replicated models with our measure of fiscal devolution lagged by 0–3 years. As bias could result from the unbalanced nature of the panel, we repeated analysis only including countries with the full 17 years of data. Results Between 1995 and 2011, across the study countries, the average proportion of expenditure controlled by subnational governments increased slightly from 27 to 33% (see web Appendix 3). During this time those countries which tended to experience a greater increase in devolution also tended to experience a greater decline in economic differences between regions (as measured by the Gini-coefficient of regional GDP, r = 0.29, P ≤ 0.001). However, those countries which experienced a greater increase in devolution also tended to experience a greater increase in health inequalities between regions (r = 0.47, P ≤ 0.001) (Fig. 2). Fig. 2 View largeDownload slide Logic model relating fiscal devolution to regional economic inequalities and regional health inequalities. Fig. 2 View largeDownload slide Logic model relating fiscal devolution to regional economic inequalities and regional health inequalities. The multivariable regression analyses indicated broadly similar associations when controlling for economic trends in each country and country fixed effects. As Table 1 shows, each additional percentage of government expenditure managed at the subnational level reduced the Gini-coefficient of regional GDP by −0.17 points (95% CI: −0.33 to −0.01; P = 0.04). With each additional percentage of government expenditure managed at the subnational level the SII of regional life expectancy increased by 23 days (95% CI: −2 to 48), although this was just below significance at the 5% level (P = 0.07). When additionally controlling for the Gini-coefficient of regional GDP in this model, however, the association increased and was significant (P = 0.01). This suggests the association between increased devolution and higher health inequalities was mitigated to some extent through the effect of increased devolution reducing regional economic inequalities. When the mediating pathway between devolution and health inequalities via economic inequalities was held constant (by controlling for it in the analysis), we saw a larger significant association between increased devolution and increased regional health inequalities. Table 1 Association between increased devolution and change in regional economic inequalities (model 1) and regional health inequalities (model 2). Model 3 shows change in regional health inequalities associated with increased devolution, when controlling for change in regional economic inequalities   Model  Coefficient  95% CI  P-value  Change in regional Gini for each additional 1% of expenditure controlled at the subnational level.  1  −0.17  −0.33  −0.01  0.04  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level.  2  23.28  −1.67  48.23  0.07  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level. (Additionally controlling for regional Gini.)  3  31.95  7.37  56.53  0.01    Model  Coefficient  95% CI  P-value  Change in regional Gini for each additional 1% of expenditure controlled at the subnational level.  1  −0.17  −0.33  −0.01  0.04  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level.  2  23.28  −1.67  48.23  0.07  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level. (Additionally controlling for regional Gini.)  3  31.95  7.37  56.53  0.01  Note: Models based on equations C shown in online Supplementary Appendix 2 and included controls for country fixed effects, time trends, national unemployment rate and national GDP, full model results are given in online Supplementary Appendix 3. Table 1 Association between increased devolution and change in regional economic inequalities (model 1) and regional health inequalities (model 2). Model 3 shows change in regional health inequalities associated with increased devolution, when controlling for change in regional economic inequalities   Model  Coefficient  95% CI  P-value  Change in regional Gini for each additional 1% of expenditure controlled at the subnational level.  1  −0.17  −0.33  −0.01  0.04  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level.  2  23.28  −1.67  48.23  0.07  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level. (Additionally controlling for regional Gini.)  3  31.95  7.37  56.53  0.01    Model  Coefficient  95% CI  P-value  Change in regional Gini for each additional 1% of expenditure controlled at the subnational level.  1  −0.17  −0.33  −0.01  0.04  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level.  2  23.28  −1.67  48.23  0.07  Change in SII (in days of life expectancy) for each additional 1% of expenditure controlled at the subnational level. (Additionally controlling for regional Gini.)  3  31.95  7.37  56.53  0.01  Note: Models based on equations C shown in online Supplementary Appendix 2 and included controls for country fixed effects, time trends, national unemployment rate and national GDP, full model results are given in online Supplementary Appendix 3. Our results were broadly comparable when limiting data to prior to the economic crisis (2008), or using the proportion of taxation raised at the subnational level as our indicator of fiscal devolution, or only including countries with the full 17 years of data. The lag models indicated that the association between fiscal devolution and reduced economic inequality was greatest for a lag of 2 years. For the other models, there the association was greatest with the fiscal devolution variable lagged by 1 year (see web Appendix 4). Discussion Main findings As has been the case in previous studies10,11 we found increased devolution was associated with reduced regional economic inequalities. However, the impact on health inequalities was not straightforward. We found increases in devolution were associated with increased regional health inequalities, although this was only significant if trends in regional economic inequalities were controlled for. This suggests the direct effect of increased devolution was to increase heath inequalities, but because devolution also tended to reduce regional economic inequalities, this mitigated some of the direct negative impacts of devolution on regional health inequalities. What is already known There has been a growing trend in Europe towards greater fiscal devolution.1 Several studies have shown greater fiscal devolution is associated with decreased economic inequalities between regions within countries.3,10,11,15–17 It is thought localized decision-making may lead to more effective and efficient use of resources as it is informed by local knowledge about needs and is accountable to local populations and this may contribute to reduced regional economic inequalities.9 We do not know the health inequalities impact of devolution policy. The few studies which have investigated the health inequalities effect of devolution in high income countries have tended to investigate the effects of devolution on healthcare expenditure. An Italian study,32 found a reform which devolved a large proportion of healthcare expenditure was associated with a reduction in inequalities in self-reported health within regions, but it did not investigate the effect on inequalities between regions. Costa-Font and Gil18 found no association between devolution of healthcare expenditure in Spain and inequalities in reported health status. A Canadian study found healthcare devolution was followed by an increase in inequalities in health status.19 However, although we know public investment across multiple sectors (education, employment, environment and housing) has the potential to influence health, we do not know whether increased fiscal devolution of public expenditure in general is associated with a decrease or an increase in health differences between regions within European countries. What this study adds Our results indicate that whilst devolution may be an effective approach to reducing regional economic inequalities, this does not necessarily translate into reduced health inequalities. In fact, the opposite may be true, and devolution may increase health inequalities between regions. There are several reasons why fiscal devolution could be associated with reduced regional differences in GDP whilst being associated with increased regional differences in life expectancy. Firstly, it could be that the ability to redistribute resources from wealthier regions to poorer regions is particularly important for addressing health inequalities and this capacity is reduced in more devolved administrations, leading to under-funded health and social services in poorer regions. Other research has shown that centrally planned targeting of resources to disadvantaged areas has been associated with reduced health inequalities.13,14 Secondly, greater local control over public investment decisions may lead to local legislation and investment in projects which have primarily economic benefits and limited health benefits. Investment in infrastructure projects, attracting outside investment or public popularity may be prioritized above health; and thus, economic efficiencies may not translate into population health benefits. Local governments may also be more susceptible to lobbying from corporations and local businesses, undermining their capacity to implement public health legislation—such as controlling access to alcohol or unhealthy foods, or reducing environmental pollution. Whilst local businesses may seek to reduce regulations that have public health benefits but are seen as a barrier to economic growth.33 Finally, it could also be that the reductions in regional differences in GDP associated with increased fiscal devolution do not represent a reduction in inequalities in the life chances of the populations living in different regions. For example, in more devolved administrations where there is greater regional control over the setting of tax rates, poorer regions may try and compete with other regions to attract external investment by reducing taxes. This could lead to increased capital flows into poorer regions—reducing regional differences in GDP, however, these increases in regional GDP may not translate into widespread improvements in wages.34 Furthermore, the reduced levels of taxation in these regions then reduce income for public service provision, further widening inequalities in life chances.35 Our findings have important implications for policies that aim to devolve greater resources to subnational levels. They indicate that even if such policies reduce economic differences between regions, they could still increase health differences. Therefore, policies should not be purely focused on economic growth or just on reducing economic inequalities between regions. Mitigating actions will need to be taken to ensure heath inequalities are not increased by these policies and to maximize the potential for such policies to reduce health inequalities.36 Strengths and limitations of this study Our findings need to be understood in the context of several limitations. Firstly, because our study analysed data at the aggregate level, we cannot know whether the findings reflect associations at the individual level. For example, it could be that whilst there was an association between increased devolution and increased regional health inequalities, health inequalities could have concurrently reduced within regions as was found in the study of healthcare devolution in Italy.32 Secondly, the measure of devolution we used—the proportion of government expenditure managed at the subnational level—only reflects a component of devolution. In practice, the nature of devolution differs greatly between countries, with differences in the level of devolution between tiers of government and between regions. Models vary between federal governments with three tiers, such as Belgium, to unitary governments with one level but with some autonomous regions, such as in Portugal.37 In some countries, devolution relates primarily to decisions over expenditure, in others to tax raising powers, and in others to legislative freedoms. In this study, we have focused on one measure of fiscal devolution, and we have not been able to investigate differential impacts of alternative approaches to devolution across the study countries. However, our findings did not change when using the proportion of tax revenue raised at the subnational level (rather than the level of expenditure controlled at the subnational level) as the measure of devolution. Thirdly, due to limitations in the data available the sample size was relatively small; therefore, the analysis may have been underpowered and our estimates are relatively imprecise. Fourthly, it is possible the associations we observed were due to other confounding factors that we were not able to take into account in the analysis. However, by analysing change over time across countries, our analysis provides more robust evidence than simple cross-sectional analyses. This enabled us to control for potentially unobserved time-invariant confounders that could vary between countries. Finally, our investigation of whether regional economic inequalities mediate the effect of fiscal devolution on regional health inequalities assumes trends in regional health inequalities are not a major cause of regional economic inequalities. If this was the case, we would be conditioning on a common effect of exposure and outcome and the increased association observed in model 3 could then be the result of collider bias.38 Conclusion Across the 14 European countries included in our analysis, whilst increased fiscal devolution was associated with decreased regional economic inequalities, it was associated with increased regional health inequalities. Policies which promote increased fiscal devolution should aim to minimize any negative impacts on health inequalities and to maximize positive impacts. This could include ensuring that devolution is combined with policies for inclusive economic growth39 and a progressive system for redistribution of resources both between regions and between individuals through a comprehensive welfare system.22 Supplementary data Supplementary data are available at the Journal of Public Health online. References 1 Polverari L. Does devolution increase accountability? Empirical evidence from the implementation of European Union Cohesion Policy. Reg Stud  2015; 49: 1074– 86. Google Scholar CrossRef Search ADS   2 Local Government Association. Briefing—Greater Manchester Health and Social Care Partnership, 2015. http://www.gmhsc.org.uk/ (accessed 5 March 2015, date last accessed). 3 City Growth Commission. Unleashing Metro Growth Final Recommendations of the City Growth Commission. London; 2014. 4 Blöchliger H, Vammalle C. Reforming Fiscal Federalism and Local Government: Beyond the Zero-Sum Game . Paris: OECD Publishing, 2012. Google Scholar CrossRef Search ADS   5 Council of European Municipalities and Regions. Decentralisation at a Crossroads: Territorial Reforms in Europe in Times of Crisis . Brussels: Council of European Municipalities and Regions, 2013. 6 Marmot M. Health Inequalities in the EU: Final Report of a Consortium . Brussels: European Union, 2013. 7 Robalino DA, Picazo OF, Voetberg A. Does Fiscal Decentralization Improve Health Outcomes? Evidence from a Cross-Country Analysis . Washington: World Bank, 2001. Google Scholar CrossRef Search ADS   8 Oates WE. On the Theory and Practice of Fiscal Decentralization . Lexington: Institute for Federalism and Intergovernmental Relations, 2006. 9 Asfaw A, Frohberg K, James KS et al.   Modeling the Impact of Fiscal Decentralization on Health Outcomes: Empirical Evidence from India . Bonn. Center for Development Research, University of Bonn; 2004. 10 Canaleta CG, Arzoz PP, Garate MP. Regional economic disparities and decentralisation. Urban Stud  2004; 41: 71– 94. Google Scholar CrossRef Search ADS   11 Lessmann C. Regional inequality and decentralization: an empirical analysis. Environ Plann A  2012; 44: 1363– 88. Google Scholar CrossRef Search ADS   12 Greater Manchester Combined Authority. Taking Charge of our Health and Social Care in Greater Manchester: The Plan. Manchester, 2015. 13 Barr B, Bambra C, Whitehead M. The impact of NHS resource allocation policy on health inequalities in England 2001–11: longitudinal ecological study. Br Med J  2014; 348: g3231. Google Scholar CrossRef Search ADS   14 Barr B, Higgerson J, Whitehead M. Investigating the impact of the English health inequalities strategy: time trend analysis. Br Med J  2017; 358: j3310. Google Scholar CrossRef Search ADS   15 Beila J, Hennl A, Kaiser A. Combining federalism and decentralization: comparative case studies on regional development policies in Switzerland, Austria, Denmark, and Ireland. Comp Polit Stud  2012; 45: 447– 76. Google Scholar CrossRef Search ADS   16 Ezcurra R, Pascual P. Fiscal decentralization and regional disparities: evidence from several European Union countries. Environ Plann  2008; 40: 1185– 1201. Google Scholar CrossRef Search ADS   17 Governatori M, Yim D Fiscal Decentralisation and Fiscal Outcomes . European Commission, Directorate-General for Economic and Financial Affairs; Brussels, 2012. 18 Costa-Font J, Gil J. Exploring the Pathways of Inequality in Health, Access and Financing in Decentralised Spain . London: London School of Economics and Political Science, 2008. 19 Zhong H. The impact of decentralization of health care administration on equity in health and health care in Canada. Int J Health Care Finance Econ  2010; 10: 219– 37. Google Scholar CrossRef Search ADS PubMed  20 Health Affairs. Health Policy Brief. The Relative Contribution of Multiple Determinants to Health Outcomes . Maryland: Health Affairs and Robert Wood Johnson Foundation, 2014. 21 OECD. OECD Fiscal Decentralisation Database. [Online]. http://www.oecd.org/tax/federalism/oecdfiscaldecentralisationdatabase.htm#C_3 (23 March 2015, date last accessed). 22 Mackenbach J, Kunst E. Measuring the magnitude of socio-economic inequalities in health: an overview of available measures illustrated with two examples from Europe. Soc Sci Med  1997; 44: 757– 71. Google Scholar CrossRef Search ADS PubMed  23 OECD. Regional Statistics and Indicators. [Online]. http://www.oecd.org/cfe/regional-policy/regionalstatisticsandindicators.htm. (23 March 2015, date last accessed). 24 Van Kerm P Stata Module to Compute Measures of Inequality, Ideas . Boston College, Department of Economics; 2007. https://ideas.repec.org/c/boc/bocode/s416401.html (13 October 2016, date last accessed). 25 The World Bank. GINI Index (World Bank Estimate). [Online]. https://data.worldbank.org/indicator/SI.POV.GINI (25 June 2015, date last accessed). 26 Cribb J, Joyce R, Phillip D. Living Standards, Poverty and Inequality in the UK: 2012 . London: The Institute for Fiscal Studies, 2012. 27 Council of European Municipalities and Regions. Crisis and Decentralisation. 2014. http://www.ccre.org/en/activites/view/13 (7 August 2015, date last accessed). 28 Owyang MT, Sekhposyan T, Vermann EK. Output and unemployment how do they relate today? Reg Economist  2013; 21: 4– 9. 29 Eurostat. Labour Markets at Regional Level. [Online] http://ec.europa.eu/eurostat/statistics-explained/index.php/Labour_market_statistics_at_regional_level (12 October 2015, date last accessed). 30 Torres-Reyna O. Panel Data Analysis Fixed and Random Effects Using Stata . Princeton: Princeton University, 2007. 31 Jiménez-Rubio D. The impact of fiscal decentralization on infant mortality rates: evidence from OECD countries. Soc Sci Med  2011; 73: 1401– 7. Google Scholar CrossRef Search ADS PubMed  32 Di Novi C, Piacenza M, Robone S et al.  . How Does Fiscal Decentralization Affect Within-Regional Disparities in Well-being? Evidence From Health Inequalities in Italy . York: University of York, 2015. 33 Katikireddi SV, Smith KE, Stuckler D et al.  . Devolution of power, revolution in public health? J Public Health (Bangkok)  2017; 39: 241– 7. 34 Alves J, Peralta S, Perelman J. Efficiency and equity consequences of decentralization in health: an economic perspective. Rev Port de Saúde Pública  2013; 31: 74– 83. Google Scholar CrossRef Search ADS   35 Social Mobility Commission. State of the Nation 2017: Social Mobility in Great Britain . London: Social Mobility Commission, 2017. 36 OECD. OECD Territorial Reviews: Sweden 2010 . Paris: OECD Publishing, 2010. 37 Allain-Dupré D Decentralisation Trends in OECD Countries: A Comparative Perspective for Ukraine [Presentation]: OECD, Regional Development Policy Division; 2016. 38 Cole SR, Platt RW, Schisterman EF et al.  . Illustrating bias due to conditioning on a collider. Int J Epidemiol  2010; 39: 417– 20. Google Scholar CrossRef Search ADS PubMed  39 Saunders M, Barr B, McHale P et al.   Key Policies for Addressing the Social Determinants of Health and Health Inequities. Health Evidence Network (HEN), Synthesis Report 52. Copenhagen. WHO Regional Office for Europe. 2017. © The Author(s) 2018. Published by Oxford University Press on behalf of Faculty of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

Journal of Public HealthOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off