Nickerson and Solomon include fixed effects to control for the overall fertility in a county and year, allowing it to vary by the number of children a woman has. This variation is measured at the state-year level using the 2000-2004 ACS surveys, which controls for variation in overall birthrates due to many economic and demographic factors, and allows for changes in overall desired family size in that county and year. This is important for distinguishing the effect of car seat laws from the impact of other secular changes that might affect birth rates.

University of Minnesota - Twin Cities

To distinguish the effect of car seat laws from other factors that will affect birth rates, Nickerson and Solomon employ numerous controls, fixed effects and interaction terms in their specifications.

Controlling for Other Factors

Nickerson, Jordan and Solomon, David H., Car Seats as Contraception (July 31, 2020). Available at SSRN: https://ssrn.com/abstract=3665046 or http://dx.doi.org/10.2139/ssrn.3665046

Car Seats as Contraception

Overall Fertility

Nickerson and Solomon include dummy variables for every combination of children’s ages. In other words, they also estimate a general probability of giving birth for each combination of children’s ages, such that the effect of car seat laws is measured relative to other families with the same number and ages of children. This allows them to measure the effect of car seat laws relative to the overall patterns in the spacing of children’s births, while putting very few structural assumptions on the shape of such birth patterns.

Combinations of Children's Ages

- Yearly values of a woman's age
- Race/ethnicity
- Quintiles of household income
- Highest education level among household members
- Presence of a male in the household

Learn Before

Related