What determines where you live? Housing costs, family income, the quality of schools, length of commute all matter in our housing decisions. But there are overarching factors that are even more important: the people who compose your family and where you work. Have your first child? Get a divorce? Find a new job? You’re a strong candidate for a move.

Families that move thus often find themselves with more than new neighbors. Often, their entire family life is profoundly new, with differing circumstances. And that’s why studies that treat moves as random changes in life are flawed.

Based on one such study, the New York Times has had fun this month with catchy graphics and computer-generated stories about how living in each county in the U.S. can help (or hurt) your children’s chances for success in life. But in real life, what sends your family packing to eastern Kentucky is probably different from what triggers a move to Los Angeles — and the “what” behind a move really matters.

The original study, by Raj Chetty and Nathaniel Hendren, tried to control for changes in family income and local prices. But it didn’t fully disentangle the effect of moving to a specific county from the effect of the family or job change that precipitated the move.

Most Americans agree that neighborhoods are quite important. It’s why homes in good neighborhoods cost so much more. But does a safer, cleaner, wealthier environment help kids be more successful?

In the mid-1990s, the federal government ran an experiment to find out how much neighborhoods matter. The Moving to Opportunity program gave randomly-selected residents of rough government housing projects vouchers to help pay for rent, but only if they moved to neighborhoods with low poverty. Only half the recipients used the voucher — a clue that, absent a “forcing” event, a move does not always appeal.

Follow-up studies found that those who received vouchers were no better off, on many dimensions, than those who did not. Recipients’ health did not improve; their incomes did not improve and students’ test scores did not improve. Moving to Opportunity stands as a colossal failure: It helped people escape some of America’s worst slums, without demonstrably improving their lives. Maybe neighborhood is not so important, after all.

But in a second study, Mr. Chetty and Mr. Hendren, along with Larry Katz, found some value among the wreckage of Moving to Opportunity. For children under age 13 at the time of the experiment, those who received the voucher were more likely to attend college and to earn more as young adults.

Thus, the experiment showed that neighborhood does, over time, affect opportunity — but only for those 12 and younger.

The same study found no benefits for teenagers. Perhaps the disruption of moving had overwhelmed any benefits of being in a nicer place with better schools. Or maybe it was too late for teenagers to be positively affected by their new digs.

The result is damning. There are lots of reasons to expect that moving teens from the projects to a working-class neighborhood would help: better schools, less chance of being either the victim or the friend of criminals, more jobs nearby, more peers attending college and working, and better adult role models. If a friend were to make such a move — for their children’s sake — surely you’d applaud it as a wise investment.

So how do we interpret the evidence? By ranking neighborhood in third place (or lower) as a factor that shapes children’s chances for a successful adulthood.

It is certainly far less important than family. Children thrive in stable environments. Boys, especially, benefit from having a father at home.

Income stability is also important. Income can pay for helpful opportunities or be saved to provide a vital safety net. Children will imitate parents who model hard work, thrift, and personal advancement.

And most everything else — community life, schooling, and yes, neighborhood — is heavily influenced by those first two.

 - Salim Furth is a research fellow specializing in macroeconomics in The Heritage Foundation’s Center for Data Analysis.