I develop an estimation framework for a model tracking population changes over time where observations are recorded by multiple sources with differing frequency and precision. I apply this framework to recover the number of unmarried men and women by age and race in each state for every year over the last three decades by combining measurements from the Decennial Census, the Current Population Survey, and the American Community Survey. Estimates of the model parameters are used to generate yearly measures of sex ratios and to estimate the impact of this marriage market feature on the marriage timing decision of women. Previous research has relied on the infrequent snapshots of marriage markets features. Using the yearly estimates of sex ratios I find a decrease in waiting time to first marriage for National Longitudinal Survey of Youth 1979 (NLY79) women when the relative number of potential spouses increases. Yearly estimates of potential partner supply are also used to identify and estimate the effect of mother's marriage on children born out of wedlock. I use the Child and Young Adult Survey (CNLSY79), a survey that collects information on children of the NLSY79 women, to link children with their mothers and the marriage markets these women experienced. Ordinary least squares (OLS) regressions show that black children whose mothers marry are more likely to obtain a high school diploma relative to children whose mothers remain unwed. The effect of marital union is also positive when sex ratios are used as instruments for the mother's marriage decision. However, the instrumental variable results are not robust to the inclusion of other explanatory variables that can affect marriage market features.