A medically uninsured couple from Woodbridge speaks with medical assistant Vivian Guillen at the Greater Prince William Community Health Center. (Photo by Jahi Chikwendiu/The Washington Post)

"Amy Workman and her 12-year-old son spent the holidays stuffing their Coeur d’Alene home into a moving truck," a story in the Idaho Statesman begins. "A few days after Christmas, the mother of three and her youngest son hit the road for Oregon. They headed west in search of health insurance."

Reporter Audrey Dutton's story follows one mother's decision to move from Idaho -- a state that hasn't expanded Medicaid -- into one that has, Oregon. Workman is among the 5.2 million Americans who are stuck in a Medicaid "gap" -- earning less than the poverty line of about $12,000 per year and too poor to qualify for private insurance subsidies. Her decision to move to Oregon could easily make economic sense: Medicaid spends, on average, $3,025 per adult and $2,359 per child.

So, should we expect a widespread migration from the states who fail to expand Medicaid into those that are participating? New research suggests it's unlikely: Previous expansions of the public program, coupled with the geographic clustering of non-expansion states, suggest that cases like Workman's will be the exception rather than the rule.

"The takeaway, I think, is that as states are considering whether to expand Medicaid it's unlikely they'll see an influx of beneficiaries if they expand," says Aaron Schwartz, a doctoral candidate at Harvard who published the new paper in the journal Health Affairs. "Based on past experience, we don't see any effects on migration."

Schwartz and co-author Benjamin Sommers looked at four states that recently expanded their Medicaid programs -- Arizona, Maine, New York and Massachusetts -- and those that neighbor their borders. They looked at the in- and out-of-state migration patterns of low-income adults for five years before and after each Medicaid expansion. Generally, they found that the control states - those that border the expansion states - had higher rates of in- and out-migration, both before Medicaid expanded and afterwards.

But, looking at the role of Medicaid, they saw that "changes in in-migration and out-migration following the expansions were not statistically significant."

"The largest aspect is probably just folks don't have very much income," Schwartz says. "It's an expensive prospect to move, even if its a benefit that's been valuable."

Sommers suggests that another explanation might be geography: When you look at the map of states that expanded Medicaid, and those that didn't, they tend to turn up in clusters.

 If you live Georgia, it will require a pretty long-distance move to end up in a state that is expanding Medicaid. 

"You could absolutely imagine in D.C. and Northern Virginia or on the border between eastern Indiana and western Ohio, people making that switch," co-author Sommers, also at Harvard, says. "It would be very different if the states that were expanding were right next to ones that are not expanding."

But even if there were big migrations -- between the District and Virginia, for example, or Wisconsin and any of its three neighbors -- the authors still aren't sure you'd see a significant budgetary impact for the states that are expanding.

"We're trying to look at total state impact, what happens when you think about the budget and statewide figures," Sommers says. "We don't see any evidence of a statewide impact."