1
$\begingroup$

I've been reading that conditional quantile regression may yield results that are not interesting or generalizable in a policy context, and that unconditional quantile regressions (UQR) are a better way to go.

However, for a better critical understanding of the two approaches, can anyone tell me what are some of the limitations of UQR (with references)?

Thanks!

$\endgroup$
1
$\begingroup$

Perhaps the biggest limitation using UQR is that it can only be used to obtain global effects. In other words, you can identify how a change in a characteristic will affect a specific individual or group of individuals. What you analyze is: what would happen if EVERYONE in the population will experience a change in characteristics? In specific, what would happen with the distribution of the outcome if characteristics distribution change? and how is this translated in terms of the quantiles?

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.