[fosap]

Please RFA.

"The reduction they saw was confined to households that did actually have access to a car"

"They discovered that tightening these laws had no detectable effects on the rate of first and second children".

Also, if i understand it correctly, you have about 50 quasi experiments, one for each state.

[Nacraile]

Please RFA more ;)

"the third-child deterrent appears stronger among wealthier families"

So this has both the most and least effect on families that have little reason to care about needing a bigger car (because they can easily afford it, or they can't afford any car at all, respectively). When you go fishing for a correlation, form a post-hoc hypothesis for its mechanism, and then other data fits that hypothesis poorly, it's a strong sign you've found yourself a red herring. Which is the usual result of failing to properly understand the distinction between correlation and causation.

[souprock]

With increasing wealth, per-child cost goes up. It's self-imposed of course.

Getting a larger Porsche or BMW or Mercedes is going to be harder than getting a larger used Hyundai or Kia. At the extreme, you're just out of luck; there is no Ferrari with more than 4 seats.

[ashtonkem]

Most well-to-do families that I know drive a very large SUV as their secondary car, often a Q7 or a Land Rover.