Abstract
This paper introduces a new graphical tool: the mean deviation concentration curve. Using a unified approach, we derive the associated dominance conditions that identify robust rankings of absolute socioeconomic health inequality for all indices obeying Bleichrodt and van Doorslaer’s (J Health Econ 25:945–957, 2006) principle of income-related health transfer. We also derive dominance conditions that are compatible with other transfer principles available in the literature. To make the identification of all robust orderings implementable using survey data, we discuss statistical inference for these dominance tests. To illustrate the empirical relevance of the proposed approach, we compare joint distributions of income and health-related behavior in the United States.
Similar content being viewed by others
Notes
It is important to note that the well-known concentration index is a rank-dependent index.
Exceptions are noted for cases where the variable of interest is not ratio-scale as in the work of Allison and Foster [1] for pure health inequality and the work of Makdissi and Yazbeck [28] in the context of socioeconomic health inequality. Indeed, these two papers explore a particular type of dominance for absolute socioeconomic health inequality. Specifically, because of their focus on ordinal data, Allison and Foster [1] and Makdissi and Yazbeck [28] develop a dominance-based approach embedded within an index-based approach (embedded dominance approach). Nevertheless, both papers do not develop dominance conditions that accounts for the uncertainty around the specific mathematical form of the index itself in the context of absolute inequality.
In fact, this new proposed measurement tool is based on a mathematical transformation of the tools available in the literature.
We provide dominance conditions the second-order and higher orders.
In this paper, we assume that this health measure is a ratio-scale variable.
The assumptions made on this social weight function are discussed in details in Sects. “Health achievement-based socioeconomic health inequality indices”, “Identifying robust orderings of health distributions”, “Principle of income-related health transfer: mean deviation concentration curve”, and “Pro-poor transfer sensitivity principles: Health achievement-based inequality indices”.
While this paper’s focus in on absolute inequality indices, the relation between achievement indices and inequality indices will play a key role in this paper if we want to consider higher order ethical principles.
It is important to note that a marginal transfer of health \(\delta _h\) can be viewed as a transfer in resources that have an impact on health.
A more formal definition of this set and all subsequent sets of indices are given Appendix A.
It should be noted that we can also define a subset of health achievement indices that pass the upside down test and make these symmetric indices compatible with an AKS approach. However, there is no need to restrict achievement indices to obtain robust rankings.
It is important to note that a well-defined class of indices involves imposing mathematical assumptions on the social weight function. These assumptions have an interpretation in terms of ethical principles.
Note that while Jann [16] proposes a “concentration” curve version of Moyes’s [29] absolute Lorenz curve in his Stata package, Jann [16] does does not lay the theoretical foundations nor provides ethical principles underlying this “concentration” curve version for absolute Lorenz. In addition, Jann [16] does not derive the associated dominance conditions including higher order conditions that allow researchers to test for dominance.
For information regarding the detailed proofs for all dominance Theorems please refer to Appendix B.
Note that if Theorem 1 holds, the conditions in the subsequent theorems also hold since \(\Lambda _{AW}^s\subset \Lambda _{AW}^{s-1}\subset \cdots \subset \Lambda _{AW}^2\subset \Lambda _A^2\). In case one identifies a robust ranking with Theorem 1, there is no need to check for these additional conditions.
The authors adapt Fishburn and Willig [10] generalized transfer sensitivity principles to the context of socioeconomic health inequality.
Note that one does not need to impose A.3 at order 2 because \(MDC^2(1)=0\) by definition. Having \(\nu ^{(s-2)}(1)MDC^s(1)=0\) or \(\omega ^{(s-2)}(1)MDC^s(1)=0\) plays an important role in the proofs in the appendix. Imposing A.3 is not a very strong assumption. It only eliminates social weights function that would assign negative values on some socioeconomic ranks. However, because the technical fact that \(MDC(1)=0\), we do not need this assumption at order 2 to impose as little assumptions as possible.
We denote by \(\Lambda _{A\rho }^2\subset \Lambda _A^2\) the set of all indices obeying this ethical principle.
The details are in Appendix C.
Details can be found Appendix D.
Restricting the analysis to subintervals of [0,1] does not complicate the analysis.
We compute equivalent income by dividing family income by the square root of household size.
It should be highlighted that if we were dealing with healthy behaviors a concentration of this behavior among the lower socioeconomic groups would mean a lower socioeconomic inequality for that behavior. Indeed, the interpretation of the direction of inequality depends on the nature of the health related behavior.
We would like to highlight that with the current methodology we cannot dig into the mechanisms behind these differences. This goes beyond the scope of this paper but is an interesting avenue to explore in future work on the topic.
In this empirical illustration, we chose to tests for order 2 to 4.
References
Allison, R.A., Foster, J.E.: Measuring health inequality using qualitative data. J. Health Econ. 23, 505–524 (2004)
Atkinson, A.B.: On the measurement of inequality. J. Econ. Theory 2, 244–263 (1970)
Avendano Pabon, M.: Smoking and inequalities. The Lancet 368, 1417–1418 (2006)
Bilger, M., Kruger, E.J., Finkelstein, E.A.: Measuring socioeconomic inequality in obesity, looking beyond the obesity threshold. Health Econ. 26, 1052–1056 (2017)
Bleichrodt, H., van Doorslaer, E.: A welfare economics foundation for health inequality measurement. J. Health Econ. 25, 945–957 (2006)
Davidson, R., Duclos, J.-Y.: Testing for restricted stochastic dominance. Econ. Rev. 32, 84–125 (2013)
Erreygers, G., Clarke, P., Van Ourti, T.: “Mirror, mirror, on the wall, who is this land is the fairest of all’’ - Distributional sensitivity in the measurement of socioeconomic inequality in health. J. Health Econ. 31, 257–270 (2012)
Erreygers, G., Van Ourti, T.: Measuring socioeconomic inequality in health, health care and health financing by means of rank-dependent indices: A recipe for good practice. J. Health Econ. 30, 685–694 (2011)
Faure, M., Gravel, N.: Reducing inequalities among unequals. Int. Econ. Rev. 62, 357–404 (2021)
Fishburn, P.C., Willig, R.D.: Transfer principles in income redistribution. J. Public Econ. 25, 323–328 (1984)
Guo, X., Post, T., Wong, W.K., Zhu, L.X.: Moment conditions for almost stochastic dominance. Econ. Lett. 124, 163–167 (2014)
Guo, X., Wong, W.K.: Multivariate stochastic dominance for risk averters and risk seekers. RAIRO - Oper. Res. 50, 575–586 (2016)
Guo, X., Wong, W.K., Zhu, L.X.: Almost stochastic dominance for risk averters and risk seekers. Financ. Res. Lett. 19, 15–21 (2016)
Harper, S., Lynch, J.: Commentary: using innovative inequality measures in epidemiology. Int. J. Epidemiol. 36, 926–928 (2007)
Harper, S., King, N.B., Meersman, S.C., Reichman, M.E., Breen, N., Lynch, J.: Implicit value judgments in the measurement of health inequalities. Milbank Q. 88, 4–29 (2010)
Jann, B.: Estimating Lorenz and concentration curves. Stata J. 16, 837–866 (2016)
Kelly, M., Morgan, A., Bonnefoy, J., Butt, J., Bergman, V., Mackenbach, J.P.: The social determinants of health: developing an evidence base for political action. World Health Organization Commission on Social Determinants of Health (2007)
Khaled, M., Makdissi, P., Yazbeck, M.: Income-Related Health Transfers Principles and Orderings of Joint Distributions of Income and Health. J. Health Econ. 57, 315–331 (2018)
King, N.B., Harper, S., Young, Y.E.: Use of relative and absolute effect measures in reporting health inequalities: structured review. BMJ 345, e5774 (2012)
Kjellsson, G., Gerdtham, U.-G., Petrie, D.: Lies, damned lies, and health inequality measurements. Understanding the value judgments. Epidemiology 26, 673–680 (2015)
Kolm, S.-C.: Unequal inequalities I. J. Econ. Theory 12, 416–442 (1976)
Kolm, S.-C.: Unequal inequalities II. J. Econ. Theory 13, 82–111 (1976)
Leshno, M., Levy, H.: Preferred by “all’’ and preferred by “most’’ decision makers: almost stochastic dominance. Manag. Sci. 48, 1074–1085 (2002)
Liu, L., Meyer, J.: Stochastic superiority, Working Paper 2107, Private Entreprise Research Center, Taxas A&M University (2021)
Mackenbach, J.P., Kunst, A.E.: Measuring the magnitude of socio-economic inequalities in health: an overview of available measures illustrated with two examples from Europe. Soc. Sci. Med. 44, 757–771 (1997)
Mackenbach, J.P., Stirbu, I., Roskam, A.-J.R., Schaap, M.M., Menvielle, G., Leinsalu, M., Kunst, A.E.: Socioeconomic inequalities in health in 22 European countries. N. Engl. J. Med. 358, 2468–2481 (2008)
Makdissi, P., Yazbeck, M.: Measuring socioeconomic health inequalities in presence of multiple categorical information. J. Health Econ. 34, 84–95 (2014)
Makdissi, P., Yazbeck, M.: Robust Rankings of Socioeconomic Health Inequality Using a Categorical Variable. Health Econ. 26, 1132–1145 (2017)
Moyes, P.: A new concept of Lorenz domination. Econ. Lett. 23, 203–207 (1987)
Mussard, S., Pi Alperin, M.N.: Accounting for risk factors on health outcomes: The case of Luxembourg. Eur. J. Oper. Res. 291, 1180–1197 (2021)
Mussard, S., Pi Alperin, M.N., Thireau, V.: Health inequality indices and exogenous risk factors: an illustration on Luxembourgish workers. Eur. J. Health Econ. 19, 1285–1301 (2018)
Oliver, A., Healey, A., Le Grand, J.: Addressing health inequalities. The Lancet 360, 565–567 (2002)
Regidor, E.: Measures of health inequalities: part 2. J. Epidemiol. Community Health 58, 900–903 (2004)
Regidor, E., Sánchez, E., de la Fuente, L., Luquero, F.J., de Mateo, S., Domínguez, V.: Major reduction in AIDS-mortality inequalities after HAART: the importance of absolute differences in evaluating interventions. Soc. Sci. Med. 68, 419–426 (2009)
Schechtman, E., Shelef, A., Yitzhaki, S., Zitikis, R.: Testing hypotheses about absolute concentration curves and marginal conditional stochastic dominance. Economet. Theor. 24, 1044–1062 (2008)
Shorrocks, A.F.: Ranking income distributions. Economica 50, 3–17 (1983)
Sen, A.K.: On economic inequality. Clarendon Press, Oxford (1973)
Wagstaff, A.: Inequality aversion, health inequalities and health achievement. J. Health Econ. 21, 627–641 (2002)
Wagstaff, A., Paci, P., van Doorslaer, E.: On the measurement of inequalities in health. Soc. Sci. Med. 33, 545–557 (1991)
Whang, Y.-J.: Econometric analysis of stochastic dominance. Concepts, methods, tools, and applications. Cambridge University Press, Cambridge, UK (2019)
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This research was funded by the Social Sciences and Humanities Research Council of Canada. The authors wish to thank Owen O’Donnell his constructive comments.
Appendices
A Sets of indices
Proofs are based on the following mathematical definition of the set of indices.
\(\Lambda _A^2\) is the set of all absolute socioeconomic health inequality indices, \(I_A(f_{Y,H})\) such that
-
(a)
\(\nu (p)\in \Re \),
-
(b)
\(\nu (p)\) is continuous and differentiable almost everywhere over \(\left[ 0,1\right] \),
-
(c)
\(\int _0^1\nu (p)\mathrm {d}{p}=0\),
-
(d)
\(\nu ^{(1)}\left( p\right) > 0, \ \forall p \in [0,1]\).
\(\Lambda _R^2\) is the set of all absolute socioeconomic health inequality indices, \(I_R(f_{Y,H})\) such that
-
(a)
\(\nu (p)\in \Re \),
-
(b)
\(\nu (p)\) is continuous and differentiable almost everywhere over \(\left[ 0,1\right] \),
-
(c)
\(\int _0^1\nu (p)\mathrm {d}{p}=0\),
-
(d)
\(\nu ^{(1)}\left( p\right) > 0, \ \forall p \in [0,1]\).
\(\Lambda _{AW}^s\) is the set of all absolute socioeconomic health inequality indices, \(I_A(f_{Y,H})\in \Lambda _A^2\) such that
-
(a)
\(\nu (p)\in (-\infty ,1]\),
-
(b)
\(\nu (p)\) is continuous and \(s-1\)-time differentiable almost everywhere over \(\left[ 0,1\right] \),
-
(c)
\(\nu ^{(i)}(1)=0 \ \forall i\in \{1,2,\dots ,s-1\}\),
-
(d)
\((-1)^{i+1}\nu ^{(i)}(p)\geqslant 0 \ \forall p\in [0,1] \ \forall i\in \{1,2,\dots ,s-1\}\).
\(\Lambda _{A\rho }^s\) is the set of all absolute socioeconomic health inequality indices, \(I_A(f_{Y,H})\in \Lambda _A^2\) such that
-
(a)
\(\nu (1-p)=\nu (p) \ \forall p\in [0,1]\),
-
(b)
\(\nu (p)\) is continuous and \(s-1\)-time differentiable almost everywhere over \(\left[ 0,1\right] \),
-
(c)
\(\nu ^{(i)}(0.5)=0 \ \forall i\in \{1,2,\dots ,s-1\}\),
-
(d)
\((-1)^{i+1}\nu ^{(i)}(p)\geqslant 0 \ \forall p\in [0,0.5] \ \forall i\in \{1,2,\dots ,s-1\}\).
B Proofs of dominance theorems
Proof of the result in Eqs. (5) and (6) Integrating by parts Eq. (2) yields
Since by definition \(GC(0)=0\) and \(GC(1)=\mu _h\) for all indices \(I_A(f_{Y,H})\in \Lambda ^2_A\), Eq. (16) can be rewritten as
From Eq. (17), we get
Note that \(\nu ^{(1)}(p)\) is non negative. This implies that if \(GC_1(p)\ge GC_2(p)\) for all \(p\in [0,1]\), then \(\int _0^1\nu ^{(1)}(p)\left[ GC_1(p)-GC^s_2(p)\right] \mathrm {d}{p}\ge 0\). If in addition, \(\mu _{h2}\ge \mu _{h1}\), then \(\Delta I_{A12}\ge 0\). This proves for sufficiency of the condition. \(\square \)
Proof of Theorem 1
Integrating by parts equation (??), we get
Since by definition \(MDC^2(0)=MDC^2(1)=0\), \(I_A(f_{Y,H})\in \Lambda ^2_{A}\), the first term on the right hand side of the equation is nil. Also note that \(\omega ^{(1)}(p)=\nu ^{(1)}(p)\). This yields to
Let \(\Delta I_{A12}=I_A(f_{Y,H}^2)-I_A(f_{Y,H}^1)\). From equation (20), we get
Note that \(\nu ^{(1)}(p)\) is non negative. This implies that if \(MDC_2^2(p)\ge MDC_1^2(p)\) for all \(p\in [0,1]\), then \(\Delta I_{A12}\ge 0\). This proves for sufficiency of the condition.
Having provided a sufficiency condition let us now prove for the necessity of the condition. Consider now the set of indices \(I_A(f_{Y,H})\in \Lambda ^2_{A}\) for which \(\nu (p)\) takes the following form:
where \(p_c\in [0,1]\) and \(\theta \) is chosen so that \(\int _0^1\nu (p)\mathrm {d}{p}=0\). Since \(\nu (p)\) is differentiable almost everywhere, it satisfies the conditions in the definition of \(\Lambda ^2_{A}\). Differentiating Eq. (22) yields
Imagine now that \(MDC_2^2(p)<MDC_1^2(p)\) on an interval \([p_c,p_c+\varepsilon ]\) for \(\varepsilon \) that can be arbitrarily close to 0. For any \(\nu (p)\) obeying the relation in Eq. (23), the expression in Eq. (21) is negative. Hence it cannot be that \(MDC_2^s(p)<MDC_1^s(p)\) for \(p\in [p_c,p_c+\varepsilon ]\). This proves the necessity of the condition. \(\square \)
Proof of Theorem 2
First note that for \(I_A(f_{Y,H})\in \Lambda ^s_{AW}\), equation (2) can be rewritten as
Integrating by parts Eq. (24), we get
Since by definition \(MDC^2(0)=MDC^2(1)=0\), \(I_A(f_{Y,H})\in \Lambda ^2_{A}\), the first term on the right hand side of the equation is nil. This yields to
Now assume that for \(s-1\), we have
Integrating by parts Eq. (27) yields
Since by definition \(MDC^s(0)=0\) and \(\omega ^{(s-2)}(1)=0\) (because \(\nu ^{(s-2)}(1)=0\)) for all indices \(I_A(f_{Y,H})\in \Lambda ^s_{AW}\), the first term in the braces on the right hand side of the equation is nil. This yield
Given that Eqs. (27) and (29) both conform to the relation depicted in Eq. (26), it follows that Eq. (29) holds for all \(s\in \{2,3,\dots \}\). Let \(\Delta I_{A12}=I_A(f_{Y,H}^2)-I_A(f_{Y,H}^1)\). From Eq. (29), we get
Note that \((-1)^{s-1}\omega ^{(s-1)}(p)\) is non negative. This implies that if \(MDC_2^s(p)\ge MDC_1^s(p)\) for all \(p\in [0,1]\), then \(\Delta I_{A12}\ge 0\). This proves for sufficiency of the condition.
Having provided a sufficiency condition let us now prove for the necessity of the condition. Consider now the set of indices \(I_A(f_{Y,H})\in \Lambda ^s_{AW}\) for which \(\omega ^{(s-2)}(p)\) takes the following form:
where \(p_c\in [0,1]\). Since \(\omega (p)\) is differentiable almost everywhere, it satisfies the conditions in the definition of \(\Lambda ^s_{AW}\). Differentiating Eq. (31) yields
Imagine now that \(MDC^s(p)<MDC_1^s(p)\) on an interval \([p_c,p_c+\varepsilon ]\) for \(\varepsilon \) that can be arbitrarily close to 0. For any \(\omega (p)\) obeying the relation in Eq. (32), the expression in Eq. (30) is negative. Hence it cannot be that \(MDC_2^s(p)<MDC_1^s(p)\) for \(p\in [p_c,p_c+\varepsilon ]\). This proves the necessity of the condition. \(\square \)
Proof of Theorem 3
First note that for \(I_A(f_{Y,H})\in \Lambda ^s_{A\rho }\), Eq. (2) can be rewritten as
Integrating by parts Eq. (33), we get
Since by definition \(GR^2(0)=0\) and \(\nu (0.5)=0\) for all indices \(I_A(f_{Y,H})\in \Lambda ^s_{A\rho }\), the first term on the right hand side of the equation is nil. This yields to
Now assume that for \(s-1\), we have
Integrating by parts Eq. (36) yields
Since by definition \(GR^s(0)=0\) and \(\nu ^{(s-2)}(0.5)=0\) for all indices \(I_A(f_{Y,H})\in \Lambda ^s_{A\rho }\), the first term in the braces on the right hand side of the equation is nil. This yield
Given that Eqs. (35) and (38) both conform to the relation depicted in Eq. (36), it follows that Eq. (38) holds for all \(s\in \{2,3,\dots \}\). Let \(\Delta I_{A12}=I_A(f_{Y,H}^2)-I_A(f_{Y,H}^1)\). From Eq. (38), we get
Note that \((-1)^s\nu ^{(s-1)}(p)\) is non negative. This implies that if \(GR_2^s(p)\ge GR_1^s(p)\) for all \(p\in [0,0.5]\), then \(\Delta I_{A12}\ge 0\). This proves for sufficiency of the condition.
Having provided a sufficiency condition let us now prove for the necessity of the condition. Consider now the set of indices \(I_A(f_{Y,H})\in \Lambda ^s_{A\rho }\) for which \(\nu ^{(s-2)}(p)\) takes the following form:
where \(p_c\in [0,0.5]\). Since \(\nu (p)\) is differentiable almost everywhere, it satisfies the conditions in the definition of \(\Lambda ^s_{A\rho }\). Differentiating Eq. (40) yields
Imagine now that \(GR_2^s(p)<GR_1^s(p)\) on an interval \([p_c,p_c+\varepsilon ]\) for \(\varepsilon \) that can be arbitrarily close to 0. For any \(\nu (p)\) obeying the relation in Eq. (40), the expression in Eq. (39) is negative. Hence it cannot be that \(GR_2^s(p)<GR_1^s(p)\) for \(p\in [p_c,p_c+\varepsilon ]\). This proves the necessity of the condition. \(\square \)
C Estimator for \(MDC^s(p)\)
Computation of integrals containing indicator variables involving inverse of \(\widehat{F}^1_Y\). Even though \(\widehat{F}_Y\) is a step function, the following standard result holds: \(y_i \leqslant \widehat{F}_Y^{- 1} (p)\) if and only if \(\widehat{F}_Y (y_i) \leqslant p\). In what follows, We will check the formula for the estimator by induction.
First compute
Then recursively compute for \(s > 3\)
\(\square \)
Estimator for \(MDC^s(p)\). The fact that MDC(p) can be written as
results into the simple estimator for MDC(p) from a sample \((h_i,y_i)\) for \(i=1,\dots ,N\) to be given by:
where \(\overline{h}=N^{-1}\sum _{i=1}^N h_i\). From equations (48) and (49), we can also get the simple estimator for \(MDC^s(p)\)
\(\square \)
D Bootstrap algorithm
The bootstrap algorithm is constructed as follows. Assume that we have an i.i.d. sample of size \(n_0\) from the random variable corresponding to first theoretical curve \(L_0\) and and and i.i.d. sample of size \(n_1\) from the random variable corresponding to the second theoretical curve \(L_1\). Denote those samples by \(\mathcal {S}_0\) and \(\mathcal {S}_1\) respectively. Let \(\widehat{L}_0\) and \(\widehat{L}_1\) be the nonparametric estimators of \(L_0\) and \(L_1\) respectively, constructed from those two samples and let \(\widehat{L}_{01} (p) = \widehat{L}_0 (p) - \widehat{L}_1 (p)\). Let
-
1.
Repeat for \(b = 1, \ldots , B\)
-
(a)
Draw a sample of size \(n_0\) from \(\mathcal {S}_0\). Compute the nonparametric estimator \(\widehat{L}_{0 b}\).
-
(b)
Draw a sample of size \(n_1\) from \(\mathcal {S}_1\). Compute the nonparametric estimator \(\widehat{L}_{1 b}\).
-
(c)
Compute \(\widehat{L}_{01 b} (p) = \widehat{L}_{0 b} (p) - \widehat{L}_{1 b} (p)\).
-
(d)
Compute \(\widehat{\tau }_b = \sqrt{\frac{n_0 n_1}{n_0 + n_1}} \sup _p \left( \widehat{L}_{01 b} (p)-\widehat{L}_{01} (p)\right) .\)
-
(a)
-
2.
Using the sample \(\widehat{\tau }_1, \ldots , \widehat{\tau }_B\), compute the bootstrap p-value
$$\begin{aligned} \frac{1}{B} \sum _{b = 1}^B \mathbbm {1} (\widehat{\tau }_b > \widehat{\tau }). \end{aligned}$$
Rights and permissions
About this article
Cite this article
Khaled, M.A., Makdissi, P. & Yazbeck, M. On absolute socioeconomic health inequality comparisons. Eur J Health Econ 24, 5–25 (2023). https://doi.org/10.1007/s10198-022-01448-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10198-022-01448-8
Keywords
- Mean deviation concentration curves
- Generalized health concentration curves
- Generalized health range curves
- Absolute socioeconomic health inequality
- Stochastic dominance
- Inference