A Rollercoaster Decade for Migration

New Census numbers released last week underscore an often unnoticed consequence of the what Time magazine called the “Decade from Hell”: a topsy-turvy pattern of population movement both across the U.S. and into its borders over a 10 year period which is ending with the greatest migration slowdown since the end of World War II.

These migration shifts were affected by a series of events that include a mid-decade housing bubble, followed by the financial crises and Great Recession, in addition to the mobility implications of Katrina and the 9-11 terrorist attacks. They led to boom, and then bust experiences for much of the South and West as the decade began, and windfall gains for northern and coastal states that were major donors to the earlier Sun Belt surge. The sharp migration shifts will also have implications for the 2010 Census based reapportionment of Congress.

The new Census numbers show migration and population shifts for states, through July 2009, along with revised annual estimates for earlier years in the decade. The most dramatic migration turnaround story is the decline of Florida as the nation’s pre-eminent migration magnet. Florida led all states in domestic migration for each of the first five years of this decade, falling to 2nd in 2005-06, 15th in 2006-7, and down to 44th and 45th respectively in each of the last two years — during which it lost more migrants than it gained, for the first time in recent history. (Table 1)

Clearly a poster child for the foreclosure crises, Florida was not the only state that shed its mid decade growth image. Stunningly Nevada also showed a net domestic migration loss in 2008-9 while Arizona gained only 15,000 migrants compared with 132,000 in 2004-5.

Among the 17 Sun Belt states with the greatest migration gains in 2005-6, the last year of the housing bubble, all but four (Texas, Virginia, Washington and Colorado) registered lower migration gains in each of the next three years. Texas stands as most prominent Sun Belt survivor of the last half of the decade. With a more diversified economy and smaller run-ups in housing prices (as well as fewer foreclosures), Texas emerged as the domestic migration leader in the U.S. for each of the past four years. Yet even with that, Texas’ net gain stood at 143,000 in 2008-9, well below Florida’s gain of 265,000 when it was the leader in 2004-5. Domestic migration gains are only part of overall population gains, which also include natural increase and international migration. But the impact of these migration trends are clearly reflected in relative population shifts for states like Florida, Arizona, Nevada and Texas, with the former three showing precipitous declines in the last three years, while Texas held its own. (Figure 1)

The other side of the migration equation concerns large migration donor states like New York and California. These states were major contributors to the mid-decade Sun Belt surge. Lower housing prices and easy mortgages lured many would be homeowners and upwardly mobile families to the bubble states of the Mountain West and the Southeast to escape states with high housing cost like New York and California. While still losing domestic migrants, each of these states have shown far slower drains.

New York’s loss of 98,000 domestic migrants in 2008-9 was its smallest in this decade, down from nearly a quarter million in 2004-6. California's 2008-9 loss of 98,000 migrants is less than one third of its loss of 313,000 in 2005-6. Much of New York's mid decade migration loss was to Florida, while much of California's loss was to Arizona and Nevada, according to a separate analysis of IRS migration data. Moreover of the 9 largest Northeast and Midwest out migration states in 2005-6, all but one (Michigan) lost fewer out migrants in each of the past three years.

The new Census numbers also showed noticeable declines in immigration to the U.S. over the past three years. (Figure 2) As with domestic migrants, immigrants have also responded to the recent economic downturns by “staying put” or in some cases returning home. The 2008-9 net international migration figure of 855,000 was the lowest since 2002-3 when the events of 9-11 reduced in-flows to the U.S. International migrant gains only topped one million in 2005-6.

Although these shifts associated with the recent migration slowdown hold important economic consequences for states and regions, they also have significant political consequences as the reapportionment of Congress will be based on the results of the 2010 census. To examine the impact of the recent migration slowdown, I calculated two alternative reapportionment scenarios. The first assumes that the 2004-5 domestic migration patterns continued until Census day, 2010. The second assumes that actual migration patterns continue until the same date. The results of each are shown in Table 2.

If the mid-decade boom migration patterns would have continued for another five years, Florida and Texas would have each gained 3 Congressional seats, Arizona would have gained 2, and Georgia, Nevada and Utah would have each picked up one. On the negative side, New York and Ohio would have each lost two seats in Congress. Other losers of one seat would be Pennsylvania, New Jersey, Minnesota, Michigan, Massachusetts, Iowa and Illinois.

In contrast, the more likely scenario, by projecting current migration patterns, would change the former by awarding Florida only 1 seat rather than 3, Arizona 1 seat rather then 2, and add 4 to Texas Congressional delegation rather than 3. In this scenario as well, South Carolina and Washington would each add a Congressional seat. Also in this scenario, New York would lose only one seat rather than two, and Louisiana would lose a seat. (Note: the earlier scenario was based on pre-Katrina migration)

Overall, the migration slowdown makes Texas the dominant Sun Belt winner and Florida a loser in that it picks up only one seat while Arizona also loses a potential gain of an additional Congressional seat. And because New York only loses one seat, rather than two — it retains a congressional advantage over Florida (28 seats versus 26). In the earlier scenario, Florida would have overtaken New York (28 seats versus 27).

So the rollercoaster migration decade has raised and dashed the hopes of some states, while others have fared better than they feared. But it’s clear that just as we are in a low point economically, we are also in an abnormal lull with respect to migration. When the housing and job markets eventually recover, so too will migration recover to levels and to destinations more in keeping with our recent past. But the “boom then bust” decade of 2000-2010 will leave some scars — economically and politically that will not soon be forgotten.