Over the past three decades, educational attainment in Mexico has grown substantially. This increase in educational attainment may affect marriage patterns through the growing supply of individuals with higher education and changing preferences over their partner's education level. We use administrative marriage and birth certificate records to quantify changes in the relative education levels for both married and unmarried couples. Our results suggest that individuals who marry outside their education category prefer to match with a partner with a similar education level. That is, college graduates prefer to match with individuals with a secondary education rather than those with a primary education. We also find that assortativeness among pairs which include college graduates has grown considerably over this time period. Our findings hold across both marriage records and birth certificate records, indicating a parallel increase in assortativeness regardless of marital status.
On average, childless women observed by the Panel Study of Income Dynamics report that they intend to have more children than they actual have. A collection of intentions that record only whether respondents intend to have another child can more accurately predict the number of children they have. Errors in the formation of intentions are not required to explain this finding. Rather, if intentions record a survey respondent's most likely predicted number of children, then the average of these intentions does not necessarily equal average actual fertility, even if intentions are formed using rational expectations.