Effects of schooling, unemployment, and income on the birth rate in Mexico: evidence from an augmented ARDL model

ABSTRACT This research was carried out to identify the variables that influenced, in the long and short term, the birth rate in Mexico for the period 1991–2023. The tested variables were the years of education, the unemployment rate, and the Gross Domestic Product (GDP) per capita, as proxies of schooling, unemployment, and income, respectively. An econometric methodology called Augmented Autoregressive Distributed Lag (A-ARDL) model was utilized, owing to its robustness over the standard ARDL model. The model was estimated and tested with and without the presence of structural breaks, due to three events of external origins to the Mexican economy: the financial crisis at the end of 2008, the fall in oil prices in 2015, and the onset of the COVID-19 pandemic in Mexico. The results suggest that women’s schooling is the main factor influencing the birth rate, and it has short- and long-term effects on this rate. The income variable does not affect the birth rate; it is not statistically significant. Also, the unemployment rate was discarded because its incorporation did not allow cointegration among the variables under study. Moreover, the econometric analysis confirms that the three events of external origins had effects on the dynamics of births in Mexico. In this sense, it is concluded that an A-ARDL model, with structural breaks, allows us to model the behavior of the birth rate in Mexico for the period 1991–2023.