Log-linearization bias: estimating returns to education through quantile regression
The present study proposes to estimate returns to education on Brazilian workers' wages through quantile regression. OLS estimation of log-linearized Mincer (1974) equation, which is traditionally present in literature, can generate a specification bias resulting from Jensen’s inequality. Media...
| Autores: | , , , |
|---|---|
| Tipo de documento: | artigo |
| Estado: | Versão publicada |
| Data de publicação: | 2022 |
| País: | Brasil |
| Recursos: | Universidade de São Paulo (USP) |
| Repositório: | Economia Aplicada |
| Idioma: | português |
| OAI Identifier: | oai:revistas.usp.br:article/147299 |
| Acesso em linha: | https://www.revistas.usp.br/ecoa/article/view/147299 |
| Access Level: | Acceso aberto |
| Palavra-chave: | returns to education log-linearization quantile regression retornos à educação log-linearização regressões quantílicas |
| Resumo: | The present study proposes to estimate returns to education on Brazilian workers' wages through quantile regression. OLS estimation of log-linearized Mincer (1974) equation, which is traditionally present in literature, can generate a specification bias resulting from Jensen’s inequality. Median estimates, as wellas the mean of quantile coefficients, presented lower coefficients than OLS estimates, indicating a possible superestimation on education returns in the mean. Ultimately, we observed that education generates bigger wage gains for upper income quantiles. |
|---|