Restless Americans: Migration and Population Change, 2000-2007


Americans may be less mobile than in the past, but millions since 2000 have continued to be on the move, reshaping the landscape and economy of the nation. Three maps will be briefly discussed: one of population change by county, 2000-2007, one of net internal migration by county, and one of net immigration from abroad. We will then focus on the “extremes”, unusually large levels or intensities of net internal migration and of immigration.

Overall population growth

As was true in the 1990s, big growth has concentrated overwhelmingly in selected metropolitan regions --- and within them, primarily in their suburbs and exurbs. The big growth areas are concentrated in Texas (Houston, Dallas-Fort Worth, San Antonio, Austin), greater Atlanta, North Carolina (Charlotte-Raleigh), most of Florida, the Virginia and Maryland suburbs of Washington-Baltimore, the desert Southwest (Riverside-San Bernardino, Las Vegas, Phoenix, Tucson). Substantial exurban or spillover growth was common, with the Bay Area extending into California Central Valley, in far exurban New York and Pennsylvania as well as in largely once rural counties around such places as Salt Lake City, Denver, Portland, Seattle, Minneapolis, Chicago, Kansas City, Nashville, Indianapolis and Columbus.

Many smaller metropolitan areas grew, especially in the south and west. Many counties with universities appear to have also grown, notably in the South. Many rural or small-town counties with substantial growth boasted environmental amenities and a strong ‘quality of life’ appeal.

The only big population losses were New Orleans and vicinity, but there were also vast rural small town areas with small losses, characterized by continuing out-migration, but often also by natural decrease, more deaths than births (870 counties), and covering the Great Plains, but also much of the Midwest, Appalachia and the Northeast.

Overall the population grew by 20 million, 12 million from natural increase, and 8 million from immigration. Around 80 million Americans “migrated” (moved across a county line), 28 percent of the population, resulting in net gains of over 10 million in gaining areas, and net losses of the same 10 million to losing areas.


Net immigration into the US was almost 8 million in seven years, raising the legal/known share of the foreign born to almost 12 percent of the population. There were hundreds of counties --- rural, small town and small metropolitan --- where immigrants landed to take agricultural and industrial jobs or to work in service jobs or in construction. This trend is exemplified by a set of counties in far southwestern Kansas (which also had high internal out-migration—mostly Hispanics moving to meat-packing jobs) and to environmental amenity ski-resort counties in Colorado (construction, service).

The largest immigration flows continued to flow to metropolitan areas, including many large core central counties, many of which were losing heavily among domestic migrants to their suburban and exurban fringe counties. The 21 largest losing counties lost a net of 4.7 million, but gained 3.5 million immigrants. Some 40 percent of the 8 million immigrants were destined for just 8 metropolitan cores, most notably Los Angeles-San Diego, New York City, Miami-Fort Lauderdale, San Francisco-Oakland-San Jose, Chicago, Dallas-Fort Worth and Houston.

Internal Migration

The story of gains and losses from internal migration is a little more complex. Gaining counties attracted around 45 million migrants and sent out around 35 million, for a net gain of over 10 million. One obvious feature from the maps is the “donut” phenomenon, the prevalence of large central county net out-migration, surrounded by a ring of substantial suburban and exurban net in-migration (about two-thirds of which is from the central core counties).

This pattern is particularly marked in Houston, Dallas, Miami, Minneapolis, Washington, DC., Atlanta, Denver, Portland, Kansas City, Memphis, Nashville, and Indianapolis. In the cases of Los Angeles, San Francisco, Seattle, New York and Philadelphia, the ring of growth has pushed beyond the suburban counties to adjacent areas – as to the Central Valley of California or to NE Pennsylvania and Delaware.

Some core areas did gain, mostly and Southern communities such as Phoenix, Las Vegas, San Antonio, Charlotte and Raleigh, NC, and Knoxville, TN --- southern and of more recent importance. In many of these areas the “core” county also includes many areas which might be considered suburban. In this sense, the fastest “urban growth” took place in relatively low-density, auto-dominated regions as opposed to traditional urban cores. Finally, and most obvious on the map, is the continuing high growth of central Florida across most counties.

In contrast there are places so hurt by de-industrialization that the entire (or most) of the metropolitan areas have substantial out-migration. These include places like Detroit, Cleveland, Pittsburgh, Buffalo, Rochester, and Boston. In some places, notably Pittsburgh, even suburban areas are losing population.

Rural and small towns also show their own dynamics. There is also continued net-out-migration for almost half of rural small –town counties in all parts of the country, but especially in the Great Plains and Midwest and in the Mississippi delta. But on the other hand, we can see continuing f net in-migration to environmentally attractive areas, often for retirement or recreation, notably in parts of the west, but also in the Ozarks and other areas in the south, upper Midwest and Northeast.


The constants here are (1) the restless mobility of the population, (2) the dominance of suburban growth; and (3) the continuing decline of more than half of rural small-town counties. Prominent in recent years, but uncertain in the longer run are (4) our strong dependence on immigration (40 percent of net national growth), (5) the locus of fastest growth in exurbia, (6) the decline of northeastern and Midwestern industrial regions; (7) the rapid growth or rural environmental amenity counties and (8) the specific set of fast-growing metropolitan areas.

Given the severity of economic conditions, immigration could begin to slow as a result of declining employment opportunities or political opposition. Similarly exurban expansion could slow because of the housing credit collapse. It is not impossible that older industrial areas could partially recover (ample plant and housing stock) while environmental amenity areas could be hurt by recession and the housing collapse. This could also apply to some of the fastest gaining areas 2000-2005 --- notably Florida and southern California --- that have been the hardest hit in the 2007 on housing debacle.

But I believe American society is resilient, and even with needed constraints on excessive housing finance abuses, and even if we are indeed approaching the era of “peak oil,” the geographic settlement pattern of recent decades most likely will persist. People will continue to migrate for the same reasons they have for decades --- in search of cheaper, larger houses, for jobs, warmer weather or scenic beauty. So we can expect, as the financial crisis gradually recedes, continued growth in suburban, exurban and satellite zones of metropolitan areas, and a net flow southward and to amenity areas.

Richard Morrill is Professor Emeritus of Geography and Environmental Studies, University of Washington. His research interests include: political geography (voting behavior, redistricting, local governance), population/demography/settlement/migration, urban geography and planning, urban transportation (i.e., old fashioned generalist)

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Those millions on the move

Those millions on the move are the people that keep the moving industry alive. I run a moving service company and it takes a lot of intuitiveness to make things right, strategic areas are those who lead to real profits. Also the internet is a great boost for this kind of business.
Maroon at Austin movers

Like the Maps

I like the maps.

One thing I'd like to say is that not all donut migration is created equal. For example, in Indianapolis, there is large net domestic outmigration from the core county to be sure, but the region as a whole is experiencing net domestic in-migration. The Indy MSA had net domestic in-migration last year of over 11,000 - at rate to add over 100,000 per decade. It's not the sunbelt by any mean, but it also isn't the typical rust belt story. I calculated Midwest MSA migration data here:

Oops. I made a mistake. The

Oops. I made a mistake. The over 11,000 was the total net migration figure. Domestic in-migration was 8,592 of that. Sorry.

How much of the movement was due to housing mortgage games?

The housing boom here in the Shenandoah Valley, on the edge of the Washington, D.C. MSA, occurred post 9-11, 2002-2007. There was an expansion of jobs in the metro area, but housing expansion was constrained by high costs of building in the inner counties. So the builders moved to the exurban ring.

The expanded competition for land and housing doubled local housing costs, which had always been higher than rural Virginia due to the metro influence.

Thanks to the easy mortgage financing of the time, people moved up in the local housing market during the boom, buying homes that were far above what was affordable on the median incomes.

Many new household came into the region with Northern Virginia jobs, incomes and long commutes. The weren't exurban people, but couldn't afford anything closer.

The expansion of the stock exceeded the demand from household creation. The houses got bigger, even though persons per household continued to decline. Gas was getting more expensive then, but closer in housing was much more expensive. The $300,000 Valley house was $700,000 to a Million in Fairfax County.

Now the housing market is hung out to dry - extended commutes to un-affordable housing. Incomes are not going to expand to cover the excesses, in fact they will shrink. When I began as a planner in 1973, the rule was an affordable home was 2.5 times income.

By those standards, the median house should be $115,000 - it was up to $250,000. In 1974 "The Costs of Sprawl" proved the suburban model of development was inefficient and costly. With so much of our housing stock committed to this pattern, there is little flexibility going forward for adapting to a world that requires high efficiency in transportation.

The broadband that everyone wants, something that would let there be some substitution of communications for transportation, is very costly when housing units are so far apart.

In real cities, owning an automobile is optional. One could walk, bike or take transit when it exists because there is a large enough population to support it.

Small towns and cities at their core have the grid system of roads for walking communities. The car takes up too much space - flattening the urban areas in to un-walkable suburbs. If there are sidewalks, you can walk - for exercise - they are rarely connected to any job center.

Historically cities grew up and out. Since the 1950's we've grown out. This a Civil Defense matter. Check "The Reduction of Urban Vulnerability: Revisiting 1950s American Suburbanization as Civil Defence" by Kathleen A Tobin in Cold War History, Vol. 2., No.2, January 2002.

Anything that of high value, is high maintenance. Our society would rather throw things away and buy new. They've come to dislike maintenance. The infrastructure reports for the last 30 years have been saying we are not doing enough.

To take such a long view there needs to be a commitment to place, specifically the governmental organizations that manage the build environment - the city and town municipal corporations. Counties are for law and order, a court, sheriff, deed book, etc. Counties were not designed to do urban development.

These issues don't have much to do with geography. In economic terms, moves need to pay for themselves. Looking at the housing market today, that hasn't been the case.