The United States experiences continuous social and economic change over the years and different cycles can have a sizable impact on the movement of people from place to place within the country.

We call this movement across county lines domestic migration. It is widely recognized that people make decisions to move for a variety of reasons, including economic change and job opportunities. These changes are reflected in the population estimates developed by the U.S. Census Bureau. They show that over roughly the last decade, some major domestic migration patterns were disrupted while others remained relatively unaffected.

Every year the Census Bureau releases data on domestic migration for all states, counties and metropolitan/micropolitan statistical areas. While the data are primarily created as inputs to estimate the size of the U.S. population, the data also shed light on mobility patterns and the social and economic conditions that affect them. Here, we break the analysis down into three time points across the last decade: 2005, 2010 and 2015. The goal is to show how much change can occur in domestic migration patterns even within a few years at various levels of geography.

Population estimates show that in 2005, states (Florida and Arizona) that experienced positive net domestic migration (more people moving in than out) had much less gain during 2010 but bounced back by 2015. States that were losing people to other states early in the period (California and New York), stemmed their net migration losses as mobility slowed but then returned to previous levels by the end of the period. Finally, some states (Texas, North Dakota, the District of Columbia and Colorado) continued to experience increasing net domestic migration gains throughout the entire period.

Figure 1 compares the numeric gains in states by using 2005, 2010 and 2015 estimates. The orange dots represent 2009-2010, the period’s midpoint. Green triangles represent early period migration, while the purple squares show recent patterns. Anywhere the orange dot is far from the other two markers indicates where there was a noticeable impact on a state’s domestic migration flows. When the orange dot is between the other two symbols, the impact was relatively less.

The change over time in migration is very noticeable in states like Florida and New York. Florida went from strong gains early on to much slower growth and then back to strong gains. New York had larger net losses at the beginning and end, but the net outflow slowed down considerably over the years in between.

Because the patterns are harder to detect for states with smaller overall levels of migration, it is useful to look at growth from net domestic migration relative to the size of the state’s population (Figure 2). The District of Columbia was experiencing a high rate of net domestic outmigration early in the period but this reversed course over the years. This is also true of North Dakota, a state that has reaped the benefits of job growth in an oil boom that attracted many domestic migrants from other states.

On the other hand, states like Alaska and Massachusetts enjoyed gains around the period’s midpoint but lost people both before and after. A negative impact in states that were big early gainers — Arizona, Nevada, Idaho, Georgia and North Carolina — is evident. They all experienced a slowdown during the midpoint and a gradual return to stronger gains more recently.

The broad domestic migration patterns hold true for areas within states, but there can be exceptions in metropolitan statistical areas. For example, in contrast to Florida as a whole, the Tampa metro area consistently gained people over the 10-year period. In Arizona, both Phoenix and Tucson saw slower growth through migration, closely reflecting the state trend over time.

There are slight variations in California. While the San Francisco metro area was losing domestic migrants in 2005 and 2010, it has been gaining them since 2011. In North Dakota, the growth from migration happened primarily in the northwestern part of the state in and around counties such as Williams and McKenzie, among others.

The District of Columbia has been growing from domestic migration since 2009, while it had lost consistently since at least 2000. The Washington, D.C., metro area .lost about 15,000 people from net domestic outmigration in 2005, gained 34,000 from net domestic migration in 2010, and reverted back to net domestic outmigration in 2015 (28,000 people).

Unlike fertility and mortality patterns, which typically change gradually over a number of years, domestic migration flows can change considerably even from one year to the next, making them harder to forecast. Last decade, few would have predicted the strong growth in North Dakota or the District of Columbia in recent years. We can see that social and economic trends can influence where people choose to move and how often they choose to do so. The Census Bureau’s population estimates are a key tool to inform how researchers and policymakers understand both economic and demographic change.

Benjamin C Bolender is Assistant Division Chief in the Population Estimates and Projections Area and Samuel Garrow is Survey Statistician in the Demographic Statistical Methods Division.