Skip to main content

The heterogeneity of ethnic employment gaps


This paper investigates the heterogeneity of ethnic employment gaps using a new single-index based approach. Instead of stratifying our sample by age or education, we study ethnic employment gaps along a continuous measure of employability, the employment probability minority workers would have if their characteristics were priced as in the majority group. We apply this method to French males, comparing those whose parents are North African immigrants and those with native parents. We find that both the raw and the unexplained ethnic employment differentials are larger for low-employability workers than for high-employability ones. We show in a theoretical framework that this heterogeneity can be accounted for by homogeneous underlying mechanisms and is not evidence for, say, heterogeneous discrimination. Finally, we discuss our main empirical findings in the light of simple taste-based vs. statistical discrimination models.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9


  1. 1.

    Neal and Johnson (1996) stress the roles of verbal and mathematical skills and Black et al. (2006) those of detailed degrees and fields of specialization for highly educated workers. See also Lang and Manove (2011) for a discussion.

  2. 2.

    Notable counter-examples include Flanagan (1976), Abowd and Killingsworth (1984), Cain and Finnie (1990), Welch (1990), Bound and Freeman (1992), Stratton (1993), Darity and Mason (1998), Fairlie and Sundstrom (1999), or Couch and Fairlie (2010).

  3. 3.

    Other papers make use of the decomposition technique developed in Fortin et al. (2011): Heywood and Parent (2012) focus on performance pay jobs and show that Black/White wage differentials grow with earnings, whereas in non-performance pay jobs, those differentials go to zero.

  4. 4.

    Gobillon et al. (2015) develop a different method to account for observables in the analysis of wage gaps along the distribution of wages, which could be applied to ethnic gaps.

  5. 5.

    See Section 3.4 for a discussion on the choice of the covariates.

  6. 6.

    We tried to introduce them in alternative specifications and results were not qualitatively affected.

  7. 7.

    To maintain a sufficient number of observations per cell in Fig. 3, the education covariate was grouped into 8 positions instead of 21.

  8. 8.

    See also Fortin et al. (2011) for an extensive discussion about the interpretation of decomposition methods under the CIA.

  9. 9.

    One may also note that the information about the heterogeneity of treatment along the employability score is not redundant with the one along the propensity score. See Appendix B for a discussion of similarities and differences between the propensity and employability scores.

  10. 10.

    Xie et al. (2012) also use a nonparametric method to estimate heterogeneous treatment effects. They match control units to treated ones based on the propensity score and then estimate treatment effects as a function of the propensity score by fitting a non-parametric model.

  11. 11.

    Note that ethnic gaps in skill-signaling quality are likely to be larger for low-skill workers (lower employability), than for high-skill workers (higher employability), see Arcidiacono et al. (2010)

  12. 12.

    See also Black et al. (2009) for evidence on inter-generational transmission of IQ scores.

  13. 13.

    p(q) can have a broader sense than just a risk premium and can be seen as a general cost of insecurity. Having a less secure job can have actual consequences: more difficulty to rent a property or to get a loan.


  1. Abowd JM, Killingsworth M (1984) Do minority/white unemployment differences really exist?. J Bus Econ Stat 2:64–72

    Google Scholar 

  2. Adida CL, Laitin DD, Valfort M. -A. (2014) Muslims in France: identifying a discriminatory equilibrium. J Popul Econ 27:1039–1086

    Article  Google Scholar 

  3. Aeberhardt R, Fougère D, Pouget J, Rathelot R (2010) Wages and employment of french workers with african origin. J Popul Econ 23(3):881–905

    Article  Google Scholar 

  4. Aeberhardt R, Rathelot R (2013) Les différences liées à l’origine nationale sur le marché du travail français. Revue Française d’Economie 28(1):43–71

    Article  Google Scholar 

  5. Aigner DJ, Cain GG (1977) Statistical theories of discrimination in labor markets. Ind Labor Relat Rev 30(2):175–187

    Article  Google Scholar 

  6. Algan Y, Dustmann C, Glitz A, Manning A (2010) The economic situation of first and second-generation immigrants in France, Germany and the United Kingdom. Econ J 120(542):F4–F30

    Article  Google Scholar 

  7. Altonji J, Blank R (1999) Race and gender in the labor market. In: Ashenfelter O, Card D (eds) Handbook of labor economics, vol 3C. Elsevier, Amsterdam, pp 3143–3259

  8. Arcidiacono P, Bayer P, Hizmo A (2010) Beyond signaling and human capital: education and the revelation of ability. American Economic Journal: Applied Economics 2(4):76–104

    Google Scholar 

  9. Athey S, Imbens G (2015) Machine learning for estimating heterogeneous casual effects. Stanford Graduate School of Business Working Paper No 3350

  10. Bjerk D (2007) The differing nature of black-white wage inequality across occupational sectors. J Hum Resour 42(2)

  11. Black D, Haviland A, Sanders S, Taylor L (2006) Why do minority men earn less? A study of wage differentials among the highly educated. Rev Econ Stat 88 (1):300–313

    Article  Google Scholar 

  12. Black DA, Haviland AM, Sanders SG, Taylor LJ (2008) Gender wage disparities among the highly educated. J Hum Resour 43(3):630–659

    Google Scholar 

  13. Black SE, Devereux PJ, Salvanes KG (2009) Like father, like son? A note on the intergenerational transmission of IQ scores. Econ Lett 105(1):138–140

    Article  Google Scholar 

  14. Blinder A (1973) Wage discrimination: reduced form and structural estimates. J Hum Resour 8(4):436–455

    Article  Google Scholar 

  15. Bound J, Freeman RB (1992) What went wrong? The erosion of relative earnings and employment among young black men in the 1980s. Q J Econ 107(1):201–32

    Article  Google Scholar 

  16. Cain GG, Finnie RE (1990) The black-white difference in youth employment: evidence for demand-side factors. J Labor Econ 8(1):S364–95

    Article  Google Scholar 

  17. Chandra A (2000) Labor-market dropouts and the racial wage gap: 1940–1990. Am Econ Rev Pap Proc 90(2):333–338

    Article  Google Scholar 

  18. Charles KK, Guryan J (2011) Studying discrimination: fundamental challenges and recent progress. Annual Review of Economics 3:479–511

    Article  Google Scholar 

  19. Combes P-P, Decreuse B, Laouénan M, Trannoy A (2014) Customer discrimination and employment outcomes: theory and evidence from the french labor market, forthcoming in Journal of Labor Economics

  20. Couch KA, Fairlie R (2010) Last hired, first fired? Black-white unemployment and the business cycle. Demography 47(1):227–247

    Article  Google Scholar 

  21. Darity WAJ, Mason PL (1998) Evidence on discrimination in employment: codes of color, codes of gender. J Econ Perspect 12(2):63–90

    Article  Google Scholar 

  22. Dehejia RH (2005) Program evaluation as a decision problem. J Econ 125 (1-2):141–173

    Article  Google Scholar 

  23. Edo A, Jacquemet N, Yannelis C (2014) Language skills and homophilous hiring discrimination: evidence from gender- and racially-differentiated applications, CES working paper 2013/58, University Paris I

  24. Fairlie RW, Sundstrom W (1999) The emergence, persistence and recent widening of the racial unemployment gap. Ind Labor Relat Rev 52:252–270

    Article  Google Scholar 

  25. Fernandez R, Fogli A (2009) Culture: an empirical investigation of beliefs, work, and fertility. Am Econ J Macroecon 1(1):146–177

    Article  Google Scholar 

  26. Flanagan RJ (1976) On the stability of the racial unemployment differential. Am Econ Rev 66(2):302–08

    Google Scholar 

  27. Fortin N, Lemieux T, Firpo S (2011) Decomposition methods in economics. In: Ashenfelter O, Card D (eds) Handbook of labor economics, chap 1, vol 4. Elsevier, Amsterdam, pp 1–102

  28. Gobillon L, Meurs D, Roux S (2015) Have women really a better access to best-paid jobs in the public sector? Counterfactuals based on a job assignment model, mimeo

  29. Heywood JS, Parent D (2012) Performance pay and the white-black wage gap. J Labor Econ 30(2):249–290

    Article  Google Scholar 

  30. Hirano K, Porter J (2005) Asymptotics for statistical decision rules. Working paper, Dept of Economics, University of Wisconsin

  31. Imbens GW, Wooldridge JM (2009) Recent developments in the econometrics of program evaluation. J Econ Lit 47(1):5–86

    Article  Google Scholar 

  32. Johnson WR, Neal D (1998) Basic skills and the black-white earnings gap. In: Johnson WR, Neal D (eds) The black-white test score gap. Brooking Institution, pp 480–500

  33. Lang K, Lehmann J-YK (2011) Racial discrimination in the labor market: theory and empirics. Journal of Economic Literature, p forthcoming

  34. Lang K, Manove M (2011) Education and labor market discrimination. Am Econ Rev 101(4):1467–96

    Article  Google Scholar 

  35. Manski CF (2000) Identification problems and decisions under ambiguity: empirical analysis of treatment response and normative analysis of treatment choice. J Econ 95(2):415–442

    Article  Google Scholar 

  36. Manski CF (2002) Treatment choice under ambiguity induced by inferential problems. Journal of Statistical Planning and Inference 105(1):67–82. Imprecise Probability Models and their Applications

    Article  Google Scholar 

  37. Manski CF (2004) Statistical treatment rules for heterogeneous populations. Econometrica 72(4):1221–1246

    Article  Google Scholar 

  38. Nachman DC (1975) Risk aversion, impatience, and optimal timing decisions. J Econ Theory 11(2):196–246

    Article  Google Scholar 

  39. Neal DA, Johnson WR (1996) The role of premarket factors in black-white wage differences. J Polit Econ 104(5):869–95

    Article  Google Scholar 

  40. Oaxaca R (1973) Male-female wage differentials in urban labor markets. Int Econ Rev 14(3):693–709

    Article  Google Scholar 

  41. Pissarides CA (1974) Risk, job search, and income distribution. J Polit Econ:1255–1267

  42. Pratt JW (1964) Risk aversion in the small and in the large. Econometrica:122–136

  43. Rathelot R (2014) Ethnic differentials on the labor market in the presence of asymmetric spatial sorting: set identification and estimation. Reg Sci Urban Econ 48(C):154–167

    Article  Google Scholar 

  44. Ritter JA, Taylor LJ (2011) Racial disparity in unemployment. Rev Econ Stat 93:30–42

    Article  Google Scholar 

  45. Rosenbaum PR, Rubin DB (1983) The central role of the propensity score in observational studies for causal effects. Biometrika 70:41–55

    Article  Google Scholar 

  46. Rubin D (1974) Estimating causal effects of treatments in randomized and non-randomized studies. J Educ Psychol 66:688–701

    Article  Google Scholar 

  47. Stratton LS (1993) Racial differences in men’s unemployment. Ind Labor Relat Rev 46(3):451–463

    Article  Google Scholar 

  48. Tô M (2014) Access and returns to education and the ethnic earning gap in France, mimeo

  49. Welch F (1990) The employment of black men. J Labor Econ 8(1):S26–74

    Article  Google Scholar 

  50. Xie Y, Brand JE, Jann B (2012) Estimating heterogeneous treatment effects with observational data. Sociol Methodol 42(1):314–347

    Article  Google Scholar 

Download references


We would like to thank three anonymous reviewers as well as Klaus Zimmermann for their helpful comments as well as Pierre Cahuc, Laurent Davezies, Xavier D’Haultfœ uille, Denis Fougère, Laurent Gobillon, Pauline Givord, Nicolas Jacquemet, Kevin Lang, Guy Laroque, Thomas Le Barbanchon, Dominique Meurs, Sophie Osotimehin, Sébastien Roux, Maxime Tô, Marie-Anne Valfort, and Etienne Wasmer, and the participants to the INSEE-DEEE, the CEE and the CREST-LMi seminars, the EEA, and the EALE annual conferences for their insightful remarks. Any opinions expressed here are those of the authors and not of any institution.

Author information



Corresponding author

Correspondence to Roland Rathelot.

Ethics declarations

Conflict of interests

The authors declare that they have no conflict of interest.

Additional information

Responsible editor: Klaus F. Zimmermann


Appendix A: Proofs

A.1 Consequence of the CIA: Y(F)⊥T|p(X)

Rosenbaum and Rubin (1983) prove that:

$$Y_{i}(F) \perp T_{i} |X_{i}, \: \forall i \Rightarrow Y_{i}(F) \perp T_{i} |P(T_{i}=1|X_{i}), \; \forall i. $$

Following exactly their reasoning, it is possible to prove that, for any random variables A i , B i taking values in {0,1}:

$$A_{i} \perp B_{i} |X_{i}, \: \forall i \Rightarrow A_{i} \perp B_{i} |P(A_{i}=1|X_{i}), \; \forall i. $$

Rosenbaum and Rubin (1983) consider A i =T i and B i =Y i (F). The proof finishes by taking A i =Y i (F) and B i =T i .

A.2 Employer’s best guess:

\(\hat y_{T}(x) \doteq E [y| \tilde y, x, T] = x + \tilde \varepsilon \left (\frac {\omega ^{2}}{{\sigma _{T}^{2}}+\omega ^{2}} \right )\) This point is derived from Aigner and Cain (1977). Employers’ best guess, given x, T and \(\tilde {y}\) is:

$$\hat y_{T} (x) = E [y| \tilde y, x, T]= E [x+\varepsilon| \varepsilon+\eta, x, T]=x+ E [\varepsilon|\varepsilon+\eta, T] $$

The last equality holding because xε.

The result follows then from:

$$E [\varepsilon|\varepsilon+\eta, T]=\frac{\omega^{2}}{\omega^{2}+{\sigma_{T}^{2}}}(\varepsilon+\eta)=\frac{\omega^{2}}{\omega^{2}+{\sigma_{T}^{2}}}(\tilde y-x) $$

This equation implies that \(\hat y_{T} (x) \sim N(x, \omega ^{4}/({\sigma _{T}^{2}} + \omega ^{2}))\).

A.3 Point 1, Section 5, taste-based discrimination

To offset a utility loss δ, employers set up a cutoff \(\underline c_{D} > \underline c\) such that:

$$E( \hat y | \hat y > \underline c_{D}, x)- E( \hat y | \hat y > \underline c, x) = \delta $$

Condition \(\hat y > \gamma \), (\(\gamma =\underline c\), or \(\gamma =\underline c_{D}\)) is equivalent to:

$$\frac{\sqrt{\sigma^{2} + \omega^{2}}}{\omega^{2}} (\hat y - x ) > \frac{\sqrt{\sigma^{2} + \omega^{2}}}{\omega^{2}} (\gamma - x)\\ $$

We define \(c=(\underline c-x)\frac {\sqrt {\omega ^{2}+\sigma ^{2}}}{\omega ^{2}}\), \(c_{D}=(\underline c_{D}-x)\frac {\sqrt {\omega ^{2}+\sigma ^{2}}}{\omega ^{2}}\), \(u = (\hat y - x)\frac {\sqrt {\omega ^{2}+ \sigma ^{2}}}{\omega ^{2}} \sim \mathcal N (0,1)\) and we denote λ(.)=φ(.)/Φ(.), with φ and Φ corresponding respectively to the probability distribution function and the cumulative distribution function of a \(\mathcal {N}(0,1)\). With these notations E(u|u>γ,x)=λ(γ).

The thresholds c and c D are such that:

$$\lambda(-c_{D}) - \lambda(-c) = \delta\frac{\sqrt{\sigma^{2}+\omega^{2}}}{\omega^{2}} \doteq \tilde \delta $$

If δ does not depend on x, differentiating this equation with respect to x leads to:

$$ -c_{D}^{\prime}(x)\lambda^{\prime}(-c_{D}(x)) = -c^{\prime}(x)\lambda^{\prime}(-c(x)) $$

Given that \(\lambda ^{\prime }/\lambda \) is decreasing and that c D >c, we have that:

$$ \frac{\lambda^{\prime}(-c_{D})}{\lambda(-c_{D})} > \frac{\lambda^{\prime}(-c)}{\lambda(-c)} $$

Combining Eqs. 1 and 2, and given that \(\lambda ^{\prime }<0\), we obtain:

$$ \lambda(-c_{D})(-c_{D}^{\prime}(x)) > \lambda(-c)(-c^{\prime}(x)) $$

The ratio of hiring probabilities is equal to:

$$\frac{h_{D}}{h_{F}} = \frac{ P(u > c_{D})}{ P(u > c)} = \frac{\Phi(-c_{D})}{\Phi(-c)} $$

Differentiating the ratio of employment probabilities by x, we show that the sign of the derivative is the same as the one of:

$$\lambda(-c_{D})(-c_{D}^{\prime}(x)) - \lambda(-c)(-c^{\prime}(x)) $$

From Eq. 3, we find that the ratio h D /h F should be increasing.

A.4 Point 1, Section 5, statistical discrimination in means

In this case, \(\hat y_{D} = \hat y_{F} = x + \tilde \varepsilon \frac {\omega ^{2}}{\sigma ^{2}+\omega ^{2}}\). Condition \(\hat y > \underline c\) is equivalent to:

$$\frac{\sqrt{\sigma^{2} + \omega^{2}}}{\omega^{2}} (\hat y - x ) > \frac{\sqrt{\sigma^{2} + \omega^{2}}}{\omega^{2}} (\underline c - x)\\ $$

We define \(c=(\underline c-x)\frac {\sqrt {\omega ^{2}+\sigma ^{2}}}{\omega ^{2}}\). In this case, transformed unobservables \(u = (\hat y - x)\frac {\sqrt {\omega ^{2}+\sigma ^{2}}}{\omega ^{2}}\) are distributed in a \(\mathcal N (0,1)\) in group F and \(\mathcal N(-\mu ,1)\) in group D. We denote λ(.)=φ(.)/Φ(.), with φ and Φ corresponding to the probability distribution function and the cumulative distribution function of a \(\mathcal {N}(0,1)\).


$$\frac{h_{D}}{h_{F}} = \frac{ P(u_{D} > c(x))}{ P(u_{F} > c(x))} = \frac{\Phi(-c(x)-\mu)}{\Phi(-c(x))} $$

Differentiating the ratio of employment probabilities by x, and using that \(c^{\prime }<0\), we find that the sign of the derivative is the same as the one of:

$$\lambda(-c(x)-\mu) - \lambda(-c) $$

Because −c(x)−μ<−c(x), and \(\lambda ^{\prime }<0\), we have λ(−c(x)−μ)>λ(−c), so that the ratio h D /h F is increasing.

A.5 Point 2, Section 5

Condition \(\hat y_{T} > \underline c\) is equivalent to:

$$\frac{\sqrt{{\sigma^{2}_{T}} + \omega^{2}}}{\omega^{2}} (\hat y_{T} - x ) > \frac{\sqrt{{\sigma^{2}_{T}} + \omega^{2}}}{\omega^{2}} (\underline c - x)\\ $$

or, denoting \(u_{T} = \frac {\sqrt {{\sigma ^{2}_{T}} + \omega ^{2}}}{\omega ^{2}} (\hat y_{T} - x)\), with T=D,F, \(c(x)=\frac {\sqrt {{\sigma ^{2}_{F}} + \omega ^{2}}}{\omega ^{2}} (\underline c - x)\) and \(k = \frac {\sqrt {{\sigma ^{2}_{D}} + \omega ^{2}}}{\sqrt {{\sigma ^{2}_{F}} + \omega ^{2}}}>1\), so that h F =P(u F >c) and h D =P(u D >k c).

Because u D and \(u_{F} \sim \mathcal N (0,1)\),

$$\frac{h_{D}}{h_{F}} = \frac{ P(u_{D} > kc(x))}{ P(u_{F} > c(x))} = \frac{\Phi(-kc(x))}{\Phi(-c(x))} $$

First consider the situation when σ D and σ F do not vary with x. The derivative of \(\frac {h_{D}}{h_{F}}\) with respect to x is positive if:

$$\Phi(-kc(x)) \varphi(-c(x)) c^{\prime}(x)> k \Phi(-c(x)) \varphi(-kc(x)) c^{\prime}(x) $$

and as \(c^{\prime }(x)=-\frac {\sqrt {{\sigma _{F}^{2}}+\omega ^{2}}}{\omega ^{2}}<0\), this is equivalent to:

$$\Phi(-kc(x)) \varphi(-c(x)) < k \Phi(-c(x)) \varphi(-kc(x)) $$

Noting λ(.)=φ(.)/Φ(.), this is itself equivalent to:

$$\lambda(-c(x)) < k \lambda(-kc(x)) $$

If c>0, that is \(x< \underline c\), we have as k>1, −k c(x)<−c(x), and as λ(.) is positive and decreasing, λ(−c(x))<k λ(−k c(x)). Therefore, \(\frac {h_{D}}{h_{F}}\) is increasing in x.

If c<0, that is if \(x> \underline c\), conclusion depends on the value of k: \(\frac {h_{D}}{h_{F}}\) increases in x iff λ(−c(x))<k λ(−k c(x)) and \(\frac {h_{D}}{h_{F}}\) decreases in x if λ(−c(x))>k λ(−k c(x)). Simulations show that ∀k>1 there exists a (unique) −c 0 such that ∀−c<−c 0,λ(−c(x))<k λ(−k c(x)) and ∀−c>−c 0, λ(−c(x))>k λ(−k c(x)) (more details available upon request). So, \(\frac {h_{D}}{h_{F}}\) increases with x up to a certain threshold and then decreases. The threshold depends on the employer cut-off \(\underline c\) and on the screening error variance ratio k.

A.6 Point 3, Section 5

Consider the inflow-outflow equation with e, h, and q being functions of x:

$$e(x) = \frac{h(x)}{h(x)+q(x)}. $$

Taking the derivative with respect to x leads to:

$$e^{\prime}(x) = \frac{h^{\prime}(x)q(x) - q^{\prime}(x)h(x)}{(h(x)+q(x))^{2}} $$

with the previous notations, h(x)=P(u>c(x))=Φ(−c(x)) which is increasing in x. Therefore, it suffices for e to be increasing in x, that q be non increasing in x.

A.7 Two-stage screening model

The two-stage screening model corresponds to drawing u 1, u 2 in a bivariate normal distribution such that \(u_{1}, u_{2} \sim \mathcal {N}(0,1)\) and \(cov(u_{1}, u_{2}) = \rho _{T} = \frac {\omega ^{2}}{\omega ^{2} + {\sigma ^{2}_{T}}}\).

Writing u 2=ρ T u 1+ν, with \(V(\nu ) = \sqrt {1-{\rho _{T}^{2}}}\) leads to

$$P(u_{2}>c | u_{1}>c) = \frac{P\left( \frac{\nu}{\sqrt{1-{\rho_{T}^{2}}}} > \frac{c-\rho_{T} u_{1}}{\sqrt{1-{\rho_{T}^{2}}}}\text{~~\&~~} u_{1}>c\right) }{P(u_{1}>c)}. $$

With the previous notations, it follows that:

$$P(u_{2}>c | u_{1}>c) = \frac{\int_{c}^{\infty} \Phi\left( \frac{\rho_{T} u - c}{\sqrt{1-{\rho_{T}^{2}}}}\right)\varphi(u)}{\Phi(-c)}. $$

The denominator does not depend on ρ T , and \(\frac {\rho _{T} u - c}{\sqrt {1-{\rho _{T}^{2}}}}\) is increasing in ρ T as long as u>ρ T c (which is the case here). P(u 2>c|u 1>c) is thus increasing in ρ T , and therefore decreasing in σ T . Minority workers are more likely to be dismissed than majority ones.

Appendix B: Similarities and differences between the propensity and employability scores

The employability score shares similarities with the propensity score but it differs from it. Note first that the employability score is not a balancing score in the sense defined by Rosenbaum and Rubin (1983). In general, we do not have XT|e(X). To see this, just consider two populations T=0 and T=1, and a unique explanatory variable X with values 0 and 1, and taking value 1 with probability q if T=0 and probability 1−q if T=1 ( q≠1−q). Assume also that employment Y is such that P(Y|X,T)=1/2 independent of T and X. It follows that T X|P(Y|X)=1/2.

Even if the employability score is not a balancing score, Y(F)⊥T|e(X) entails that conditional treatment effects are identified at any value of e(X). So the employability score provides a different dimension of analysis that is not redundant with nor cannot be summarized in general by the propensity score.

Furthermore, applying the same reasoning as Rosenbaum and Rubin (1983) on e(X) instead of on the propensity score, we can define balancing scores relative to Y, instead of balancing scores relative to T. Let b Y be a balancing score relative to Y, b Y is such that XY|b Y (X). Theorem 2 of Rosenbaum and Rubin (1983) says that the propensity score p(X)=P(T=1|X) is the coarsest balancing score in the sense that if b T is a balancing score (relative to T), then p = f(b T ) for some function f. Considering now Y instead of T, it follows that e(X)=P(Y=1|X) is the coarsest balancing score relative to Y.

Theorem 3 of Rosenbaum and Rubin (1983) says that if treatment assignment is strongly ignorable given X, then it is strongly ignorable given any balancing score b T (X), which holds in particular for the propensity score p(X). Considering again Y instead of T, treatment assignment is also ignorable given any balancing score relative to Y, b Y (X), in particular given the employability e(X).

To justify even more the use of the employability, we show next that it is, with the propensity score, the only other unidimensional score that could lead to the previous results in a general way. It may happen, that in specific situations, other unidimensional scores could summarize the CIA and be good candidates for a conditional analysis, but the only ones that can work on a general basis are the propensity score and the employability. To see that, it is sufficient to find an example in which they are the only valid scores (in the above sense).

Assume that there is one single covariate X, and that Y is such that P(Y=1|X)=Λ(X), with \(\Lambda (x)=\exp (x)/(1+ \exp (x))\), and T is such that P(T=1|X)=1−Λ(X).

Imagine that there is some function g such that YT|g(X) but g is neither a balancing score relative to Y nor to T: X T|g(X) and X Y|g(X). This means that there exist x 1<x 2, such that g(x 1)=g(x 2)=γ but P(Y=1|x 1)≠P(Y=1|x 2). Given the specific form of Y and T, this also means that P(T=1|x 1)≠P(T=1|x 2).

Assume without loss of generality that \(g=\gamma \Rightarrow x \in (x_{1}, x_{2})\) and that X follows a non informative distribution. It follows that P(Y=1|T=0,g=γ)<P(Y=1|T=1,g=γ). Indeed, with T=0, it is more likely that x=x 1 than x=x 2. This contradicts the fact that YT|g(X). Therefore, in general, the only scores b that are such that YT|b(X) are balancing scores relative to Y or T.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Aeberhardt, R., Coudin, É. & Rathelot, R. The heterogeneity of ethnic employment gaps. J Popul Econ 30, 307–337 (2017).

Download citation


  • Discrimination
  • Employment differentials
  • Decomposition