Multidimensional welfare comparisons of EU member states before, during, and after the financial crisis: a dominance approach


How did the EU states’ populations fare during the financial crisis in key dimensions such as income, health, and education? We seek to answer this question by way of welfare comparisons between countries and within countries over time, using EU-SILC data. Our study is novel in implementing a multidimensional first order dominance (FOD) comparison approach on the basis of multi-level indicators. FOD only requires that outcomes can be ranked from worse to better within each dimension and is therefore suitable for the analysis of multidimensional ordinal data. We find that the countries most often dominated are southern and eastern European member states, and the dominant countries are mostly northern and western European member states. However, for most country comparisons, there is no dominance relationship. Moreover, only a few member states have experienced a temporal dominance improvement in welfare, while no member states have experienced a temporal dominance deterioration during the financial crisis.

This is a preview of subscription content, log in to check access.


  1. 1.

    Ravallion (2012) refers to this as ‘mashup indices’.

  2. 2.

    The present paper focuses on comparisons of welfare in population distributions with ordinal multidimensional data. For comparisons of inequality across populations with ordinal data, we refer to Allison and Foster (2004), Gravel and Moyes (2012), Balestra and Ruiz (2015), Sonne-Schmidt et al. (2016), and Cowell and Flachaire (2017).

  3. 3.

    See Tabel 7 in Appendix A.1 for a grouping of countries to specific regions.

  4. 4.

    Note that there can be multiple indicators for the same welfare dimension as exemplified here, where both mean years of schooling for adults and expected years of schooling for children are used as indicators in the education dimension. In this paper, we use a single indicator for each welfare dimension included.

  5. 5.

    For example, an indicator in the health dimension is that the respondent considers her own health as fair or above, and the indicator in the education dimension is whether or not the respondent has completed primary education.

  6. 6.

    Note that FOD and orthant stochastic orderings are equivalent in the one-dimensional setting.

  7. 7.

    The first proof of the equivalence between (i) and (iii) is usually attributed to Lehmann (1955) (however, see also Levhari et al. 1975). The first formulation and proof of the equivalence between (i) and (ii) is not easy to trace back to its roots, but Kamae et al. (1977) observed that the equivalence between (i) and (ii) is a corollary of Strassen’s Theorem (Strassen 1965). Østerdal (2010) provides a constructive direct proof of this for the finite case.

  8. 8.

    In the one-dimensional case, f first order dominates g if and only if \(F(x) \le G(x)\) for all \(x \in X\), where \(F(\cdot )\) and \(G(\cdot )\) are the cumulative distribution functions corresponding to f and g, respectively. For a review of FOD in both a one-dimensional and multidimensional welfare setting using binary indicators, we refer to Siersbæk et al. (2017).

  9. 9.

    Specifically, when minimizing an objective function under a set of conditions that yields a lower bound of the objective function, rounded data inputs as well as a large number of estimated transfers that all have a true value of zero are computationally handled as very small positive numbers. However, when adding up multiple very small positive numbers, the result may be a sum that is ‘large’. The analyst in practice thus has to define a threshold for when the objective function is interpreted as zero. For example, when minimizing an objective function that is bounded to zero, an estimated minimum value of the objective function of, say, \(10^{-10}\) has to be interpreted; is it a true zero (in which case FOD is observed) or something larger than zero (in which case FOD is not observed). If the threshold is not chosen ‘correctly’, this may lead to the conclusion that a dominance exists when in fact there is no dominance (but ‘close’ to dominance).

  10. 10.

    The number of LCSs is quantified by Sampson and Whitaker (1988) using the number of levels in each indicator. Strictly speaking, Sampson and Whitaker (1988) provide the number of upper comprehensive sets, which is, however, equal to the number of LCSs. For three dimensions with binary indicators, the total number of LCSs is 20. If the number of levels of each indicator is three, the total number of LCSs is 980, and if four levels of each indicator are used, the total number of LCSs increases to 232,848.

  11. 11.

    The Matlab code for identifying all LCSs and checking FOD is available on the following web-page: (also available from the authors upon request). The empty LCS and the full set of all outcomes are omitted in the code since the corresponding sums using definition (i) are 0 and 1, respectively.

  12. 12.

    Computationally efficient algorithms capable of handling multiple indicators and levels is grounds for further research. For the bivariate case, efficient algorithms are provided in Range and Østerdal (2019).

  13. 13.

    A straightforward test of the multiple inequalities obtained in condition (i) of FOD can be applied using, for example, testing of moment inequality constraints, which is frequently used to test instrument validity in econometric instrumental variable regressions (see, e.g., Chen and Szroeter 2014; Huber and Mellace 2015). This, however, implies a null-hypothesis of dominance.

  14. 14.

    See Efron and Tibshirani (1993) for a general treatment of bootstrapping, Arndt and Tarp (2017) for another applications to FOD, and Kobus et al. (2019) for an application to orthant dominance.

  15. 15.

    Equivalized total net income uses the OECD-modified scale (first proposed by Hagenaars et al. 1996). This assigns a weight of 1 to the first adult in the household, a weight of 0.5 to each additional member of the household aged 14 and over, and a weight of 0.3 to household members aged less than 14. The household’s total net income is divided by this equivalized number of persons to get equivalized total net income (per person in the household). See OECD (2013b) for more information. The income reference period is the previous calendar or tax year for all countries except for the United Kingdom (current year) and Ireland (last 12 months), see Eurostat (2019).

  16. 16.

    The ISCED (International Standard Classification of Education) is developed by UNESCO (United Nations Educational, Scientific and Cultural Organization) to facilitate cross-country comparisons of education systems since these vary in terms of structure. We use the ISCED 1997, which ranges from 0 (pre-primary education) to 6 (second stage of tertiary education); see UNESCO (2006).

  17. 17.

    Since definition (i) in Sect. 3 uses weak inequalities, a country will always dominate itself. For simplicity, these ‘self-dominances’ are not included in Tables 2 through 4, nor in the remainder of the paper.

  18. 18.

    The maximum number of potential dominances for n countries is \((n^2-n)/2\). n is raised to the second power to obtain all country combinations. The subtraction of n in the nominator is to exclude self-dominance, whereas the 2 in the denominator is due to the fact that if country A dominates country B, B cannot dominate A. Since \(n=24\) in this paper, the maximum number of potential dominances is \((24^2-24)/2=276\).

  19. 19.

    Note that the years of analysis in Alkire and Apablaza (2016) are 2006, 2009, and 2012, where only 2009 is somewhat directly comparable.

  20. 20.

    We thank an anonymous referee for a useful discussion on this topic.

  21. 21.

    Though in a slightly different set-up, see also Hussain et al. (2016) for an example of refining dimensions to analyze the ‘depth’ of FOD.

  22. 22.

    The fifth LCS that includes all outcomes is redundant since the probability mass functions both sum to 1.

  23. 23.

    The combination of lower- and upper orthant dominance is equivalent to FOD in the two-dimensional binary case used here. This is, however, not true for more general two-dimensional distributions.

  24. 24.

    Other aggregation thresholds have been used as well. While the results naturally are sensitive to the threshold choice, the ones shown in Table 11 have been chosen to exemplify the difference between multi-level and binary indicators.

  25. 25.

    In general we will observe weakly more dominances as the number of outcome combinations decrease.


  1. Alkire S (2002) Dimensions of human development. World Dev 30(2):181–205

    Article  Google Scholar 

  2. Alkire S, Foster J (2011a) Counting and multidimensional poverty measurement. J Public Econ 95(7–8):476–487

    Article  Google Scholar 

  3. Alkire S, Foster J (2011b) Understandings and misunderstandings of multidimensional poverty measurement. J Econ Inequal 9(2):289–314

    Article  Google Scholar 

  4. Alkire S, Foster J, Seth S, Santos ME, Roche JM, Ballon P (2015) Multidimensional poverty measurement and analysis. Oxford University Press, Oxford

    Google Scholar 

  5. Alkire S, Apablaza M (2016) Multidimensional poverty in Europe 2006–2012: illustrating a methodology. OPHI Working Paper No 74, University of Oxford

  6. Alkire S, Apablaza M, Jung E (2014) Multidimensional poverty measurement for EU-SILC countries. OPHI Research in Progress 36c, University of Oxford

  7. Allison RA, Foster JE (2004) Measuring health inequality using qualitative data. J Health Econ 23(3):505–524

    Article  Google Scholar 

  8. Anderson G (1996) Nonparametric tests of stochastic dominance in income distributions. Econometrica 64(5):1183–1193

    Article  Google Scholar 

  9. Arndt C, Tarp F (2017) Measuring poverty and wellbeing in developing countries. Oxford University Press, Oxford

    Google Scholar 

  10. Arndt C, Distante R, Hussain MA, Østerdal LP, Huong PL, Ibraimo M (2012) Ordinal welfare comparisons with multiple discrete indicators: a first order dominance approach and application to child poverty. World Dev 40(11):2290–2301

    Article  Google Scholar 

  11. Arndt C, Hussain MA, Vincenzo S, Tarp F, Østerdal LP (2016) Poverty mapping based on first-order dominance with an example from Mozambique. J Int Dev 28(1):3–21

    Article  Google Scholar 

  12. Arrow JK (1971) A utilitarian approach to the concept of equality in public expenditures. Quart J Econ 85(3):409–415

    Article  Google Scholar 

  13. Atkinson AB, Bourguignon F (1987) Income distribution and differences in need. In: Feiwek GR (ed) Arrow and the Foundations of the Theory of Economic Policy, Harvester Wheatsheaf. London UK

  14. Atkinson AB (1992) Measuring poverty and differences in family composition. Economica 59(233):1–16

    Article  Google Scholar 

  15. Atkinson AB, Bourguignon F (1982) The comparison of multi-dimensioned distributions of economic status. Rev Econ Stud 49(2):183–201

    Article  Google Scholar 

  16. Balestra C, Ruiz N (2015) Scale-invariant measurement of inequality and welfare in ordinal achievements: an application to subjective well-being and education in OECD countries. Soc Choice Welf 123(2):479–500

    Google Scholar 

  17. Barrett GF, Donald SG (2003) Consistent tests for stochastic dominance. Econometrica 71(1):71–104

    Article  Google Scholar 

  18. Bossert W, Chakravarty SR, D’Ambrosio C (2013) Multidimensional poverty and material deprivation with discrete data. Rev Income Wealth 59(1):29–43

    Article  Google Scholar 

  19. Bourguignon F (1989) Family size and social utility: income distribution dominance criteria. J Econom 42(1):67–80

    Article  Google Scholar 

  20. Bourguignon F, Chakravarty SR (2003) The measurement of multidimensional poverty. J Econ Inequal 1(1):25–49

    Article  Google Scholar 

  21. Buchmann C (1996) The debt crisis, structural adjustment and women’s education: implications for status and social development. Int J Comp Sociol 37(1):5

    Article  Google Scholar 

  22. Chen LY, Szroeter J (2014) Testing multiple inequality hypotheses: a smoothed indicator approach. J Econom 178:678–693

    Article  Google Scholar 

  23. Copeland AH (1951) A “reasonable” social welfare function. University of Michigan Seminar on Applications of Mathematics to the Social Sciences

  24. Cowell FA, Flachaire E (2017) Inequality with ordinal data. Economica 84(334):290–321

    Article  Google Scholar 

  25. Crawford I (2005) A nonparametric test of stochastic dominance in multivariate distributions. Discussion Papers in Economics, DP 12/05, University of Surrey

  26. Davidson R, Duclos JY (2000) Statistical inference for stochastic dominance and for the measurement of poverty and inequality. Econometrica 68(6):1435–1464

    Article  Google Scholar 

  27. Davidson R, Duclos JY (2013) Testing for restricted stochastic dominance. Econom Rev 32(1):84–125

    Article  Google Scholar 

  28. De Beer P (2012) Earnings and income inequality in the EU during the crisis. Int Labor Rev 151(4):313–331

    Article  Google Scholar 

  29. Delhey J (2004) Quality of life in Europe. Life satisfaction in an enlarged Europe. European Foundation for the improvement of living and working conditions. Office for Official Publications of the European Commission, Luxembourg

    Google Scholar 

  30. Duclos JY, Échevin D (2011) Health and income: a robust comparison of Canada and the US. J Health Econ 30(2):293–302

    Article  Google Scholar 

  31. Duclos JY, Sahn DE, Younger SD (2006) Robust multidimensional poverty comparisons. Econ J 116(514):943–968

    Article  Google Scholar 

  32. Duclos JY, Sahn DE, Younger SD (2007) Robust multidimensional poverty comparisons with discrete indicators of well-being. In: Jenkins SP, Micklewright J (eds) Inequality and poverty re-examined. Oxford University Press, Oxford

    Google Scholar 

  33. Dyckerhoff R, Mosler K (1997) Orthant orderings of discrete random vectors. J Stat Plan Inference 62(2):193–205

    Article  Google Scholar 

  34. Efron B, Tibshirani RJ (1993) An introduction to the bootstrap. Chapman and Hall/CRC, New York

    Google Scholar 

  35. European Commission (2009) Economic crisis in Europe: causes, consequences and responses. Eur Economy 7, Accessed 10 Feb 2017

  36. European Union (2011) Interinstitutional style guide. Luxembourg: Publications Office of the European Unio. Accessed 9 May 2018

  37. Eurostat (2019) Glossary: equivalised disposable income. Accessed 10 Jan 2019

  38. Foster J, Greer J, Thorbecke E (1984) A class of decomposable poverty measures. Econometrica 52(3):761–766

    Article  Google Scholar 

  39. Gravel N, Moyes P (2012) Ethically robust comparisons of bidimensional distributions with an ordinal attribute. J Econ Theory 147(4):1384–1426

    Article  Google Scholar 

  40. Gravel N, Mukhopadhyay A (2010) Is India better off today than 15 years ago? A robust multidimensional answer. J Econ Inequal 8(2):173–195

    Article  Google Scholar 

  41. Gravel N, Moyes P, Tarroux B (2009) Robust international comparisons of distributions of disposable income and regional public goods. Economica 76(303):432–461

    Article  Google Scholar 

  42. Hagenaars AJ, De Vos K, Zaidi MA (1996) Poverty statistics in the late 1980s: research based on micro-data. Office for official publications of the European Communities. Eurostat, Luxembourg

    Google Scholar 

  43. Higham NJ (2002) Accuracy and stability of numerical algorithms. Society for Industrial and Applied Mathematics, Philadelphia

    Google Scholar 

  44. Huber M, Mellace G (2015) Testing instrument validity for LATE identification based on inequality moment constraints. Rev Econ Stat 97(2):398–411

    Article  Google Scholar 

  45. Hussain MA (2016) EU country rankings’ sensitivity to the choice of welfare indicators. Soc Indic Res 125(1):1–17

    Article  Google Scholar 

  46. Hussain MA, Jørgensen MM, Østerdal LP (2016) Refining population health comparisons: a multidimensional first order dominance approach. Soc Indic Res 129(2):1–21

    Article  Google Scholar 

  47. Jürges H (2007) True health vs response styles: exploring cross-country differences in self-reported health. Health Econ 16(2):163–178

    Article  Google Scholar 

  48. Kamae T, Krengel U, O’Brien GL (1977) Stochastic inequalities on partially ordered spaces. Ann Probab 5(6):899–912

    Article  Google Scholar 

  49. Kannai Y (1980) The ALEP definition of complementarity and least concave utility functions. J Econ Theory 22(1):115–117

    Article  Google Scholar 

  50. Kentikelenis A, Karanikolos M, Papanicolas I, Basu S, McKee M, Stuckler D (2011) Health effects of financial crisis: omens of a Greek tragedy. Lancet 378(9801):1457–1458

    Article  Google Scholar 

  51. Kobus M, Półchłopek O, Yalonetzky G (2019) Inequality and welfare in quality of life among OECD countries: non-parametric treatment of ordinal data. Soc Indic Res 143(1):201–232

    Article  Google Scholar 

  52. Kolm S (1977) Multidimensional egalitarianism. Quart J Econ 91(1):1–13

    Article  Google Scholar 

  53. Lehmann EL (1955) Ordered families of distributions. Ann Math Stat 26(3):399–419

    Article  Google Scholar 

  54. Levhari D, Paroush J, Peleg B (1975) Efficiency analysis for multivariate distributions. Rev Econ Stud 42(1):87–91

    Article  Google Scholar 

  55. Marling TG, Range TM, Sudhölter P, Østerdal LP (2018) Decomposing bivariate dominance for social welfare comparisons. Math Soc Sci 95:1–8

    Article  Google Scholar 

  56. McCaig B, Yatchew A (2007) International welfare comparisons and nonparametric testing of multivariate stochastic dominance. J App Econom 22(5):951–969

    Article  Google Scholar 

  57. McFadden D (1989) Testing for stochastic dominance. In: Fomby TB, Seo TK (eds) Studies in the economics of uncertainty. Springer-Verlag, New York, pp 113–134

    Google Scholar 

  58. Mosler KC, Scarsini MR (1991) Stochastic orders and decision under risk. Institute of Mathematical Statistics, Heyward

    Google Scholar 

  59. Muller C, Trannoy A (2011) A dominance approach to the appraisal of the distribution of well-being across countries. J Public Econ 95(3–4):239–246

    Article  Google Scholar 

  60. OECD (2013) Education indicators in focus. Chapter B: financial and human resources invested in education. What is the impact of the economic crisis on public education spending? Organisation for economic co-operation and development. Accessed 3 Feb 2017

  61. OECD (2013) OECD Framework for statistics on the distribution of household income, consumption and wealth. Chapter 8: framework for integrated analysis. Organisation for economic co-operation and development. Accessed 9 May 2018

  62. OECD/Eurostat/UNESCO Institute for Statistics (2015) ISCED 2011 operational manual: guidelines for Classifying National Education Programmes and related qualifications. OECD Publishing., Accessed 15 Dec 2019

  63. Permanyer I, Hussain MA (2018) First order dominance techniques and multidimensional poverty indices: an empirical comparison of different approaches. Soc Indic Res 137(3):867–893

    Article  Google Scholar 

  64. Range TM, Østerdal LP (2019) First order dominance: stronger characterization and a bivariate checking algorithm. Math Program 173(1–2):193–219

    Article  Google Scholar 

  65. Ravallion M (2011) On multidimensional indices of poverty. J Econ Inequal 9(2):235–248

    Article  Google Scholar 

  66. Ravallion M (2012) Mashup indices of development. World Bank Res Observer 27(1):1–32

    Article  Google Scholar 

  67. Sampson AR, Whitaker LR (1988) Positive dependence, upper sets, and multidimensional partitions. Math Oper Res 13(2):254–264

    Article  Google Scholar 

  68. Schady NR (2004) Do macroeconomic crises always slow human capital accumulation? World Bank Econ Rev 18(2):131–154

    Article  Google Scholar 

  69. Sen AK (1970) Collective choice and social welfare. Holden-Day, San Francisco

    Google Scholar 

  70. Sen AK (1976) Poverty: an ordinal approach to measurement. Econometrica 44(2):219–231

    Article  Google Scholar 

  71. Shaked M, Shanthikumar JG (2007) Stochastic orders. Springer Science and Business Media, New York

    Google Scholar 

  72. Siersbæk N, Østerdal LP, Arndt C (2017) Multidimensional first order dominance comparisons of population wellbeing. In: Arndt C, Tarp F (eds) Measuring poverty and wellbeing in developing countries. Oxford University Press, Oxford

    Google Scholar 

  73. Sonne-Schmidt C, Tarp F, Østerdal LP (2016) Ordinal bivariate inequality: concepts and application to child deprivation in Mozambique. Rev Income Wealth 62(3):559–573

    Article  Google Scholar 

  74. Strassen V (1965) The existence of probability measures with given marginals. Ann Math Stat 36(2):423–439

    Article  Google Scholar 

  75. Stuckler D, Basu S, Suhrcke M, Coutts A, McKee M (2009) The public health effect of economic crises and alternative policy responses in Europe: an empirical analysis. Lancet 374(9686):315–323

    Article  Google Scholar 

  76. Thomas D, Beegle K, Frankenberg E, Sikoki B, Strauss J, Teruel G (2004) Education in a crisis. J Dev Econ 74(1):53–85

    Article  Google Scholar 

  77. UNDP (1990) Human Development Report 1990—concept and measurement of human development. United Nations Development Programme. Accessed 29 May 2017

  78. UNDP (2011) Human Development Report 2011—sustainability and equity: a better future for all. United Nations Development Programme. Accessed 12 Feb 2016

  79. UNDP (2014) Human Development Report 2014—sustaining human progress: reducing vulnerabilities and building resilience. United Nations Development Programme. Accessed 12 Feb 2016

  80. UNESCO (2006) ISCED 1997: International standard classification of education. United Nations educational, scientific and cultural organization. Accessed 19 Jan 2016

  81. UNSD (1999) Standard country or area codes for statistical use. United Nations statistics division publications catalogue, Series M, No. 49/Rev.4. Accessed 9 May 2018

  82. Whelan CT, Nolan B, Maître B (2014) Multidimensional poverty measurement in Europe: an application of the adjusted headcount approach. J Eur Soc Policy 24(2):183–197

    Article  Google Scholar 

  83. World Bank (1990) World development report 1990: poverty. Oxford University Press, New York

    Google Scholar 

  84. Yalonetzky G (2013) Stochastic dominance with ordinal variables: conditions and a test. Econom Rev 32(1):126–163

    Article  Google Scholar 

  85. Yalonetzky G (2014) Conditions for the most robust multidimensional poverty comparisons using counting measures and ordinal variables. Soc Choice Welf 43(4):773–807

    Article  Google Scholar 

  86. Østerdal LP (2010) The mass transfer approach to multivariate discrete first order stochastic dominance: direct proof and implications. J Math Econ 46(6):1222–1228

    Article  Google Scholar 

Download references


We are grateful to Andreas Bjerre-Nielsen, Dorte Gyrd-Hansen, Casper Worm Hansen, Lene Holbæk, Giovanni Mellace, Troels Martin Range, Peter Sudhölter, and participants at the DGPE workshop (November 2015, Sønderborg, Denmark), CBS Inequality Platform 1st Workshop on Health and Inequality (April 2019, Copenhagen, Denmark), and IFABS 2019 Conference (December 2019, Medellin, Colombia) for useful inputs and suggestions. We are also grateful to the Coordinating Editor Marc Fleurbaey and two anonymous referees for very constructive comments.


Financial support from the Independent Research Fund Denmark (Grant-ID: DFF-6109-000132) is gratefully acknowledged.

Author information



Corresponding author

Correspondence to Nikolaj Siersbæk.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Financial support from the Independent Research Fund Denmark (Grant-ID: DFF-6109-000132) is gratefully acknowledged.



A.1 Country abbreviations

See Table 7.

Table 7 Abbreviations for included EU member states and regional groupings

A.2 Illustration of LCSs and description of orthant stochastic orderings

Table 8 illustrates all LCSs in the bivariate case with binary indicators. Let dimension A be the row dimension and dimension B be the column dimension. In each dimension an individual can either be in outcome 0 or 1, where 1 is best. This yields four LCSs in total: \(\hbox {LCS}_1=\{(0,0)\}\), \(\hbox {LCS}_2=\{(0,0),(1,0)\}\), \(\hbox {LCS}_3=\{(0,0),(0,1)\}\), and \(\hbox {LCS}_4=\{(0,0),(1,0),(0,1)\}\).Footnote 22 To check for FOD between two distributions f and g using definition (i), one simply has to check that the following four inequalities are satisfied:

$$\begin{aligned}&{(\hbox {i}_1)} \ g(0,0) \le f(0,0)\\&{(\hbox {i}_2)} \ g(0,0) + g(1,0) \le f(0,0) + f(1,0)\\&{(\hbox {i}_3)} \ g(0,0) + g(0,1) \le f(0,0) + f(0,1)\\&{(\hbox {i}_4)} \ g(0,0) + g(1,0) + g(0,1) \le f(0,0) + f(1,0) + f (0,1) \end{aligned}$$
Table 8 Illustrating all LCSs, bivariate and binary

FOD requires only ordinal data and does not require assumptions about the strength of preferences for each dimension, the relative desirability of changes among levels within or between dimensions, and substitutability or complementarity between dimensions (e.g., Arndt et al. 2012). On the contrary, orthant dominance concepts following, e.g., Atkinson and Bourguignon (1982) require such assumptions. Particularly, f lower orthant dominates g if

$$\begin{aligned}&{(\hbox {i}_l)} \ \sum _{z \le y} g(z) \ge \sum _{z\le y} f(z) \ \text {for all} \ y \in Y. \end{aligned}$$

Condition (\(\hbox {i}_l\)) corresponds to checking inequality \(\hbox {i}_1\), \(\hbox {i}_2\), and \(\hbox {i}_3\) (i.e., using only \(\hbox {LCS}_1\), \(\hbox {LCS}_2\), and \(\hbox {LCS}_3\)). Correspondingly, f upper orthant dominates g if

$$\begin{aligned}&{(\hbox {i}_u)} \ \sum _{z \ge y} g(z) \le \sum _{z\ge y} f(z) \ \text {for all} \ y \in Y. \end{aligned}$$

Condition (\(\hbox {i}_u\)) corresponds to checking inequality \(\hbox {i}_2\), \(\hbox {i}_3\), and \(\hbox {i}_4\) (i.e., using only \(\hbox {LCS}_2\), \(\hbox {LCS}_3\), and \(\hbox {LCS}_4\)).Footnote 23

Orthant dominance concepts are defined from classes of functions U+ (non-decreasing ALEP substitutes) or U- (non-decreasing ALEP complements) for lower and upper orthant dominance, respectively. These classes of functions require further assumptions including specified signs on the second order partial or cross-derivatives of the underlying utility function (corresponding to ALEP substitutability/complementarity/neutrality of the underlying utility functions, see Kannai 1980). For recent applications see, e.g., (Yalonetzky 2014) and (Kobus et al. 2019).

A.3 Sample sizes

See Table 9.

Table 9 Samples sizes for each country-year combination in the EU-SILC data

A.4 Data structure

See Table 10.

Table 10 Example of data structure using sample data for Germany in 2005

A.5 Illustration of temporal development in indicators

See Figs. 1, 2 and 3.

Fig. 1

Source: Calculations based on data from Eurostat

Income: Share of the population in each of the four categories. Notes: Four quartiles of equivalized annual net income, see Table 1, for Austria (AT), Germany (DE), Denmark (DK), and Hungary (HU)

Fig. 2

Source: Calculations based on data from Eurostat

Health: Share of the population in each of the five categories. Notes: Five outcomes of self-reported health, see Table 1, for Austria (AT), Germany (DE), Denmark (DK), and Hungary (HU)

Fig. 3

Source: Calculations based on data from Eurostat

Education: Share of the population in each of the three categories. Notes: Three educational outcomes based on ISCED 2011 levels ranging from 0 to 8, using (1) early childhood educational development, pre-primary education, pre-primary education, and lower secondary educaiton, (2) upper secondary and post-secondary education, and (3) short-cycle tertiary education, Bachelor’s or equivalent level, Master’s or equivalent level, and Doctoral or equivalent level. The FOD analyses use ISCED 1997, see Table 1. ISCED 2011 levels 0–2 correspond to ISCED 1997 levels 1–2 used in the FOD analyses with an additional category of early childhood educational development added to the ISCED 2011 levels, ISCED 2011 levels 3–4 correspond to ISCED 1997 levels 3–4, and ISCED 2011 levels 5-8 correspond to ISCED 1997 levels 5–6 (OECD/Eurostat/UNESCO Institute for Statistics 2015). Comparisons are for Austria (AT), Germany (DE), Denmark (DK), and Hungary (HU).

A.6 Binary FOD analyses

As mentioned, applications of FOD in a welfare context have until now used binary indicators. To show the consequences of using multi-level indicators instead of binary ones, analyses similar to those in Sect. 5 but using binary indicators have been conducted. Instead of applying the multi-level indicators outlined in Table 1, we combine the levels into binary outcomes as shown in Table 11, where the second column from the right shows the multi-level indicators applied in the previously reported results, and the rightmost column shows how the multi-level indicators are aggregated into binary indicators.Footnote 24

Table 11 Description of welfare dimensions and binary indicators

The spatial results are shown in Tables 12, 13, and 14 and the corresponding Copeland scores are shown in Table 15. The temporal results are shown in Table 16. These five tables are the analogous to Tables 2 through 6, the only difference being that the results are obtained using the binary indicators described in Table 11. Naturally, the multidimensional binary analyses yield several more dominances than the corresponding multidimensional multi-level analyses.Footnote 25 In particular, we never obtain an indeterminate result in each dimension analyzed separately since the distribution in a given dimension is fully described by a single number, e.g., those who are worse off. With binary indicators, we obtain between 138 and 149 multidimensional dominances, i.e., around three and a half times more than we do in the multi-level analyses.

The overall results are similar to those obtained in the multi-level analyses, i.e., the countries most often dominated are southern and eastern European member countries, and the dominant countries are mostly northern and western European member states. The Copeland scores using binary indicators are also similar to the ones using multi-level indicators where almost no northern or western European countries are in the bottom half of the ranking and almost no southern and eastern European countries are in the top half of the ranking. Though the use of binary indicators enables us to more easily compare European countries and to obtain a more complete ranking, it comes at the price of less refined indicators.

Table 12 Spatial first order dominances using binary indicators, 2005
Table 13 Spatial first order dominances using binary indicators, 2009
Table 14 Spatial first order dominances using binary indicators, 2013
Table 15 Copeland scores and corresponding ranking of EU member states using binary indicators
Table 16 Temporal first order dominances using binary indicators

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Hussain, M.A., Siersbæk, N. & Østerdal, L.P. Multidimensional welfare comparisons of EU member states before, during, and after the financial crisis: a dominance approach. Soc Choice Welf (2020).

Download citation