Working Paper: NBER ID: w22927
Authors: Rania Gihleb; Kevin Lang
Abstract: Some economists have argued that assortative mating between men and women has increased over the last several decades, thereby contributing to increased family income inequality. Sociologists have argued that educational homogamy has increased. We clarify the relation between the two and, using both the Current Population Surveys and the decennial Censuses/American Community Survey, show that neither is correct. The former is based on the use of inappropriate statistical techniques. Both are sensitive to how educational categories are chosen. We also find no evidence that the correlation between spouses' potential earnings has changed dramatically.
Keywords: Educational homogamy; Assortative mating; Income inequality
JEL Codes: J1; J12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| inappropriate statistical techniques (C52) | increased assortative mating by education (previous assertions) (I24) |
| sensitivity to educational categories (I24) | conclusions about assortative mating (C78) |
| grouping all college graduates together (D29) | increased homogamy (J12) |
| separating college graduates from advanced degrees (A23) | decline in homogamy (J12) |
| correlation between husbands' and wives' education (I24) | no significant increase over decades (J11) |
| changes in correlation coefficients (C10) | not indicative of a trend in assortative mating (J79) |
| categorization of educational attainment (I21) | sensitivity of findings (C52) |