Oaxaca-Blinder Decomposition

decompositionwage-gaplabor-economicslinear-modelsnonlinear-modelsgroup-differences

Definition

The Oaxaca-Blinder decomposition rewrites the difference in mean outcomes between two groups as the sum of a characteristics effect (how much of the gap is explained by group differences in covariates, evaluated at a common price vector) and a coefficients effect (how much is explained by differences in slope coefficients, i.e., the wage structure). For group-specific linear regressions Yˉg=Xˉgβ^g\bar{Y}_g = \bar{X}_g'\hat{\beta}_g, the decomposition is:

Yˉ1Yˉ2=(Xˉ1Xˉ2)β^2characteristics+Xˉ1(β^1β^2)coefficients\bar{Y}_1 - \bar{Y}_2 = \underbrace{(\bar{X}_1 - \bar{X}_2)'\hat{\beta}_2}_{\text{characteristics}} + \underbrace{\bar{X}_1'(\hat{\beta}_1 - \hat{\beta}_2)}_{\text{coefficients}}

with alternative reference-group choices (group 1's β\beta, group 2's β\beta, or a pooled estimate) giving different allocations between the two terms.

Key Ideas

How It Works

Linear Case (Oaxaca-Blinder)

Run separate ordinary least squares (OLS) regressions for each group; compute group means; apply the formula above. Aggregate the characteristics effect across covariates: k(Xˉ1kXˉ2k)β^2k\sum_k (\bar{X}_{1k} - \bar{X}_{2k})\hat{\beta}_{2k}.

Nonlinear Extensions

When Y=g(Xβ,ψ)Y = g(X'\beta, \psi) for a nonlinear link function gg (probit, logit, Tobit), the linear formula does not apply exactly.

Why It Matters

Open Questions

Related