Increased access to social services is often though to decrease inequality, but in reality the effect of these increases is empirically ambiguous in many contexts. The results in this paper – both theoretical and empirical – suggest that this is not surprising, and that the relationship between access to services and inequality may be non-monotonic in a world with discrimination. Improvements in access may increase inequality initially, with further improvements decreasing it.
In the case analyzed here, I find non-monotonicities in the relationship between access to vaccination and gender differences in vaccination rates. From a policy perspective, the non-monotonicity would suggest that if policymakers care directly about the sex ratio, interventions to increase access to health inputs should focus on saturating one area rather than introducing the inputs in a more limited way in all areas. Depending on the magnitude of the effect of vaccinations on mortality, saturating one area versus providing some vaccination to all areas could make a reasonable difference in the gender imbalance in mortality.
It is worth keeping in mind, however, that the goal of gender equality may be at odds with the goal of achieving other forms of equality. In the example above, saturating one area with vaccinations and ignoring another area generates regional inequality in health outcomes, even if it serves to improve the gender balance. Ultimately, this is a tradeoff for policy-makers to consider when making choices about the distribution of health services.
Finally, as alluded to in the introduction, the theoretical framework here may help explain a variety of patterns in inequality, not just those in India. To the extent that this is more broadly applicable, it may argue for targeted interventions (Head Start rather than universal preschool, for example) that specifically focus on individuals in disadvantaged groups.