Policy Goals and Outcomes in 'Three Worlds of Welfare Capitalism'

Headey, Bruce; Muffels, Ruud Article Policy Goals and Outcomes in 'Three Worlds of Welfare Capitalism' Schmollers Jahrbuch – Zeitschrift für Wirtschaftsund Sozialwissenschaften. Journal of Applied Social Science Studies Provided in Cooperation with: Duncker & Humblot, Berlin Suggested Citation: Headey, Bruce; Muffels, Ruud (2003) : Policy Goals and Outcomes in 'Three Worlds of Welfare Capitalism', Schmollers Jahrbuch – Zeitschrift für Wirtschaftsund Sozialwissenschaften. Journal of Applied Social Science Studies, ISSN 1865-5742, Duncker & Humblot, Berlin, Vol. 123, Iss. 1, pp. 27-41, https://doi.org/10.3790/schm.123.1.27 This Version is available at: https://hdl.handle.net/10419/292036 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Policy Goals and Outcomes in ‘Three Worlds of Welfare Capitalism’ By Bruce Headey and Ruud Muffels Abstract Measures of income mobility should be valuable for assessing performance in achieving national policy goals. We propose a portfolio of measures to assess policy goals relating to 1. economic growth and rising living standards 2. equality of opportunity 3. equality of outcomes 4. income security and 5. social solidarity. The measures are used to assess recent policy performance in the US, Germany and the Netherlands. These countries are taken as cases (‘best cases’ in terms of economic performance) of Esping-Andersen’s (1990) ‘three worlds of welfare capitalism’. JEL Classification: D31 There is an almost infinite number of ways in which income mobility could be conceptualised and measured, but no agreement among social scientists about which measures are most useful for analytic or policy purposes. The most commonly used measures of mobility at present are changes in the quantile ranks of income units between time periods (e.g. changes in quintile ranks between t 1 and t 2 ), but it often not clear why these measures are selected; they have no obvious policy relevance. The purpose of this paper is to suggest measures which are valuable from a public policy standpoint (Fields/Ok 1996; Jenkins 2000; van Kerm 2001). We assume that policy goals broadly reflect public demands, and begin by discussing measures of mobility which an individual or family might find relevant as they think about their own subjective economic welfare and reflect on how it has changed during, say, the last decade. In the main part of the paper the suggested measures are used to assess policy outcomes in the US, Germany and the Netherlands in 1987–1996. These three countries may be regarded as examples – indeed, as leading economic performers or ‘best cases’– of what Esping-Andersen (1990) termed ‘the three worlds of welfare capitalism’. In this typology Germany is viewed as the prototypical conservative, corporatist type of welfare-capitalist regime, the Netherlands is a social democratic regime (albeit a borderline one in Esping-Andersen’s framework), and the US is the leading liberal regime. We shall find that Schmollers Jahrbuch 123 (2003) 1 Schmollers Jahrbuch 123 (2003), 27–42 Duncker & Humblot, Berlin OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.27 | Generated on 2023-04-04 12:31:31 28 Bruce Headey and Ruud Muffels one’s choice of mobility measures makes an enormous difference to assessment of which country is most mobile, and indeed to understanding what kinds of policy packages they are actually delivering to their citizens. Subjective economic welfare – then policy goals As an approach to inferring policy goals relating to mobility, we begin by asking what changes in income an individual or family might wish for in reviewing its subjective economic welfare over the last decade or so. For present purposes we define subjective economic welfare as satisfaction with one’s material standard of living. The individual or family might first ask: “How much has our income gone up or down in real terms in the last ten years?” This is a question relating to absolute mobility, and it is reasonable to suppose that subjective welfare is greatly affected by absolute mobility. The public policy goal which absolute mobility translates into is rising GDP per capita, where the benefits are dispersed widely rather than concentrated on particular groups. A second question that might be asked is: “How have I/we done – how has my income changed – relative to other people in this country? Am I getting ahead, am I doing as well as I and others would have expected?” This is a question about relative mobility and translates into the public policy goal of equality of opportunity, or at least widespread opportunities. In measuring relative mobility we need to specify ‘relative to what?’ A peer group or income yardstick has to be used to assess how relatively well or badly particular individuals or households are faring. Alternative yardsticks are discussed in the Methods section. A third question which might spring to mind is: “Have I had a rough ride or an easy ride? Has my income been stable and secure, or has it been risky and precarious?” This raises issues about security of living standards and there is every reason to suppose that people’s subjective welfare is substantially affected by their perceptions of income security. We can label this dimension of mobility wave mobility or income risk. The relevant public policy goal, which welfare states promote to a greater or lesser extent, is income security. Additional public policy goals – collective goals So far we have three dimensions of mobility – absolute mobility, relative mobility and income risk – which affect subjective welfare and can readily be translated into public policy goals. We now add two policy goals of a collective nature. One is income equality – one aspect of equality of outcomes – and the other is income solidarity. Income equality, for someone who believes in Schmollers Jahrbuch 123 (2003) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.27 | Generated on 2023-04-04 12:31:31 Policy Goals and Outcomes 29 it, is a collective rather than an individual goal in that it makes little sense to say, “I want my income to become more equal to other people’s”. But a person could reasonably say, ” I want our society to become more equal”. Income equality is closely related to income mobility, because the more people’s incomes fluctuate over time, the more equal they become in the medium and long term. This linkage led Shorrocks (1978) to propose that multi-year measures of income equality could serve as measures of mobility. The final goal considered here – income solidarity, an aspect of social solidarity – is presumably what people have in mind when they worry about social exclusion, when they worry that some members of society are not enjoying the fruits of economic growth, not sharing in the general prosperity. The collective goal of social solidarity may be expressed as, “I want us to stick together as a society – to share the gains and share the losses”. In this view, government policy should promote a shared economic fate, not allowing some sections of society to get ahead while others flounder, and perhaps ensuring that during a recession all sections bear the losses or sacrifices required. What expectations or hypotheses would we have about policy performance in relation to these goals in a liberal welfare-capitalist state (the US), a corporatist state (Germany) and a social democratic regime (the Netherlands)? Extrapolating from Esping-Andersen (1990) and our own previous research on these countries (Goodin et al. 1999), we would expect that the liberal US regime performed best in relation to its priority goal of achieving high levels of economic efficiency and rising living standards. The Dutch social democratic regime would be expected to perform best in regard to egalitarian goals; equality of opportunity and equality of outcomes. Corporatist Germany would be predicted to do best in pursuing goals of income security and social solidarity. Methods The three panels The three panel surveys analysed here are the American Panel Study of Income Dynamics (PSID), the German Socio-Economic Panel (GSOEP) and the Dutch Socio-Economic Panel (SEP). All have over 10,000 respondents in the period considered, and they are the only three national economic panels to have run for ten consecutive years or more. The PSID began in 1968 and has continued ever since. Low-income households were initially over-sampled, because the study was partly paid for by the Office of Economic Opportunity. One respondent answers on behalf of each household and the sample is renewed and kept more or less representative by interviewing ‘split-offs’; that is, people who leave their original household and move to a new one (e.g. children leaving home to get married). LongitudiSchmollers Jahrbuch 123 (2003) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.27 | Generated on 2023-04-04 12:31:31 30 Bruce Headey and Ruud Muffels nal weights are used to adjust for panel attrition and other sources of sample bias. The GSOEP and the SEP both began in 1984. The GSOEP initially oversampled foreigners (guest-worker households) and also added a supplementary post-unification immigrant sample in 1995. The German and Dutch panels also follow split-offs and also, of course, require use of longitudinal weights whenever multi-year analysis is undertaken. The PSID and GSOEP files have been adapted for comparability by the German Institute for Economic Research (DIW) and Cornell University (Cornell University, 2001). We have prepared an equivalent Dutch file. 1 This paper covers the years 1987–96, since this is the last decade of data available for all three countries. Only West German data are used, since East Germans were not interviewed until 1990. In all three countries analysis is restricted to households headed by men or women of prime working age (25– 59). These are the households mainly affected by the equity and efficiency goals we assess via income mobility measures. The issues facing retirement age households and younger student age households are quite different. Measures Equivalised income The aim is to measure mobility in households’ material standard of living. So, following usual practice, all income measures have been equivalised, in this case by use of the International Experts’ equivalence scale, which requires dividing incomes by the square root of household size (Buhmann et al. 1988). This is almost the same as the current OECD equivalence scale of 1.0 for the first adult, 0.5 for other adults and 0.3 for children. In parts of the paper we shall want to look at the impact of government – the tax-transfer system – on mobility. For this purpose it makes sense to equivalise market incomes too, in order to compare mobility of market or pre-government incomes with disposable or post-government incomes. The formula used for assessing the impact of government, derived from Kakwani (1986) and Ringen (1991), is: Impact of government (%) =100* (post-government mobility – pre-government mobility)/pre-government mobility . Schmollers Jahrbuch 123 (2003) 1 1The GSOEP Cross-National Equivalent File is described in Burkhauser et al. (2001) and Cornell University (2001). The information on the English version of the SEP data is available at http://wsa.magw.nl/index_uk.htm. OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.27 | Generated on 2023-04-04 12:31:31 Policy Goals and Outcomes 31 Measuring four types of income mobility Absolute mobility is simply the absolute (inflation adjusted) change in households’ disposable incomes during the decade. This is used to measure policy goals relating to (a) economic efficiency and rising living standards (b) aspects of equality of opportunity and (c) aspects of equality of outcomes. Measurement of relative mobility requires a choice of standards or yardsticks – change in income relative to what? Perhaps the two standards most obviously relevant to policy goals and subjective welfare are mobility measured as change in income percentile rank and mobility as gains/losses relative to national median income. For the first measure we compare the percentile rank of each person’s income in 1996 with his/her starting rank in 1987. Similarly, mobility relative to national median income is the percentage of median income one received in 1996 compared to 1987. The best way to measure wave mobility or income risk is not obvious. At first sight a valid measure would be the mean or median coefficient of variation of the ten annual incomes people in each country received in 1987–96. But this measure is open to the objection that it confounds upside and downside risk; the coefficient would be the same if one’s income steadily increased or steadily declined throughout the decade. Probably when policy makers or the public think about income security they really only have in mind security against downside risk. So the measure we propose is the number of times in the decade a person’s income in the current year declined by more than 10% in real terms compared to the previous year. This very straightforward measure uncovers large differences among the countries. A measure of shared directional mobility should capture the extent to which a nation’s people share the same economic fate – get richer or poorer together. This is assessed by measuring the percentage of the population whose own real incomes rose or fell by within plus or minus 10% of national per capita economic growth during the decade. We test the sensitivity of this threshold by also considering changes between plus or minus 25%, between plus or minus 25–50%, or at the extreme by plus or minus 50%. Results Goal 1 – rising living standards, economic efficiency absolute mobility First, we present background information on economic growth in the three countries. Schmollers Jahrbuch 123 (2003) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.27 | Generated on 2023-04-04 12:31:31 32 Bruce Headey and Ruud Muffels Table 1 Economic Growth Per Capita in 1987–96 (%) US Germany Netherlands Real economic growth 23.9 24.7 28.0 Population growth 9.5 8.3 7.4 Economic growth per capita 14.4 16.4 20.6 Notes: All values are in percentages. Calculated from growth indices in OECD Economic Outlook (1999). Our first expectation proves false. The liberal US welfare-capitalist regime, which gives highest priority to economic efficiency and growth, did not have the highest growth rate. The comparison is a fair one because in all three countries the decade began and ended with reasonable growth and there was a recession in the middle. The American recession was the worst; the Dutch one was barely a pause. (If the period were extended to the present the American and Dutch relative performances would be the same, but Germany would slip back). The widespread impression that the US economy has outperformed the economies of most of Western Europe in the last ten to fifteen years is not correct in per capita growth terms, although it is true in employment terms. Next we present evidence that relates more directly to rising living standards, and hence to absolute mobility of incomes. Table 2 shows mean and median increases in real equivalent incomes and the percentage of the population in each country whose incomes were higher at the end of the decade than the beginning (‘winners’). Table 2 Economic Efficiency & Rising Living Standards: Households with Heads Aged 25–59 Absolute Mobility 1987–96 United States Germany Netherlands Pregov’t income Postgov’t income Gov’t impact Pregov’t income Postgov’t income Gov’t impact Pregov’t income Postgov’t income Gov’t impact Average increase 74.6 58.3 – 21.8 45.5 31.6 – 30.5 61.6 49.0 – 20.5 Median increase 13.5 25.7 90.3 24.1 21.1 – 12.4 11.3 19.8 75.2 Winners 59.6 66.0 10.7 67.6 68.8 1.8 60.0 69.1 15.2 Notes: All values are in percentages. The government impact is the percent difference between the postand pre-government income increases. Winners are those whose income rises over the period. Schmollers Jahrbuch 123 (2003) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.27 | Generated on 2023-04-04 12:31:31 Policy Goals and Outcomes 33 The comparisons here do not show a clear pattern. American mean and median increases in disposable income were highest but there were slightly more winners in the Netherlands and Germany than in the US. The evidence on the impact of government (the third column for each country) implies that the American and Dutch tax-transfer systems – or really the tax side – favoured middle income earners more than the German, having a more positive impact on median incomes and on the percentage who emerged as ‘winners’. The main point, however, is that our expectation that living standards would have risen most strongly in liberal US is not confirmed. In practice, the performance of the three types of regime in promoting rising living standards was quite similar. Goal 2 – equality of opportunity Our first measure of equality of opportunity, changes in people’s percentile rank in the income distribution, is the measure most commonly used by sociologists. Arguably, though, it has the drawback that it is easier to move up and down the ranks in a society with a more equal income distribution than in a more unequal society, because the same dollar gain or loss will produce more movement in the more equal society (Fritzell, 1990; Gustafsson, 1994). Table 3 is intended to answer the question, ‘What was the median change in rank by 1996 of people starting in different quintiles in 1987?’ In practice, we show results only for the top and bottom quintiles, because these are the only ones for which the international comparison shows differences. (In all countries members of the 2 nd quintile moved up a bit on average, the 4 th quintile moved down a bit, and the middle quintile stayed put). Table 3 Equality of Opportunity: The Chance to Get Ahead: Households with Heads Aged 25–59 Percentile Ranks Mobility 1987–96 United States Germany Netherlands Pregov’t income median percentile change Postgov’t income median percentile change Gov’t impact Pregov’t income median percentile change Postgov’t income median percentile change Gov’t impact Pregov’t income median percentile change Postgov’t income median percentile change Gov’t impact Bottom quintile 8 8 0 14 15 7.1 13 14 7.7 Top quintile – 7 – 8 14.3 – 9 – 8 – 11.1 – 15 – 17 11.8 Notes: All values are in percentages. The government impact is the percent difference between the postand pre-government income median percentile change. Schmollers Jahrbuch 123 (2003) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.27 | Generated on 2023-04-04 12:31:31 34 Bruce Headey and Ruud Muffels Results here are not quite in line with expectations. In the social democratic regime, the Netherlands, there was the greatest mobility, but it was especially marked at the top end of the distribution rather than the bottom. Opportunity for the lowest quintile to get ahead was just as high in Germany (the difference between the two countries not being significant at the .05 level). Both European countries recorded higher mobility at the bottom end than the US. At the top end the Netherlands stands out as a country in which, exaggerating slightly, one might say ‘it is easy to get rich, but hard to stay rich’. The impact of government in all three countries is negligible. The tax-transfer system does reduce inequality in all three countries (Goodin et al. 1999) but it does little to alter people’s income ranks; perhaps a politically dangerous thing to do. Table 4 Equality of Opportunity: The Chance to Get Ahead: Households with Heads Aged 25–59 Mobility Relative to Median (Mainstream) Income 1987–96 United States Germany Netherlands Pregov’t income median percentile change Postgov’t income median percentile change Gov’t impact Pregov’t income median percentile change Postgov’t income median percentile change Gov’t impact Pregov’t income median percentile change Postgov’t income median percentile change Gov’t impact Bottom quintile 17 17 0 14 13 - 7.1 13 13 0 Top quintile – 30 – 32 6.7 – 36 – 23 – 36.1 – 48 – 37 – 22.9 Notes: All values are in percentages. The government impact is the percent difference between the postand pre-government income median percentile change. Table 4 provides a second measure of equality of opportunity – mobility relative to national median income. This measure does not have the problem of being likely to show more apparent mobility in a more equal society than an unequal one. The table shows median outcomes in 1996 for people starting in the top and bottom quintiles in 1987. Here results are plainly not in line with expectations. On this measure, the US bottom quintile had most chance to get ahead – moving up 17% relative to national median disposable income in 1996 compared to their position in 1987. The Dutch and German bottom quintiles both gained 13% relative to median income. At the top end of the distribution there is considerably more mobility, and it again appears that the Netherlands is a hard place in which to retain a high income. The impact of government is again negligible at the bottom end, and also at the top end in the US. Schmollers Jahrbuch 123 (2003) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.27 | Generated on 2023-04-04 12:31:31 Policy Goals and Outcomes 41 Ringen, S. (1991), Households, standard of living and inequality, in: The Review of Income and Wealth, 37, 1–13. Shorrocks, A. F. (1978), Income inequality and income mobility, in: Journal of Economic Theory, 46, 376–93. van Kerm, P. (2001), Essays on Income Mobility and Income Distribution Dynamics, Naumur. Schmollers Jahrbuch 123 (2003) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.27 | Generated on 2023-04-04 12:31:31