Research Report

Scenarios for Regional Growth from 2010 to 2030

Mapping America’s Futures, Brief 1

National population projections from the Census Bureau foresee growth of nearly 49 million people between 2010 and 2030. We explore where in the United States that growth could occur using scenarios from Urban Institute’s new “Mapping America’s Futures: Population” tool. In this brief we ask: Where will the babies be born? Where will we see higher or lower levels of mortality? Where will people of every age move to and from? Where will the immigrants arrive?

This brief is one of a series that demonstrates the potential for policy and social investigation using the “Mapping America’s Futures: Population” tool. Other briefs in the series examine projections for changes in the racial and ethnic composition of the population (“Evolving Patterns in Diversity”), in the age structure of the population (“Children and Youth in an Aging America”), and in America’s labor markets (“The Labor Force in an Aging and Growing America”). The online tool allows the viewer to see the implications of different assumptions about future fertility, mortality, and migration, all of which are explained in a separate methodology brief (“Methodology and Assumptions for the Mapping America’s Futures Project”).

Our population projections are divided geographically across 740 commuting zones of the United States. Additionally, we defined 24 regions (figure 1), built from the many regional divisions defined in earlier work linked with our own observations on how neighboring commuting zones experience similar trends in population change.

Exhibit 1

Scenario Building: How Birth, Death, and Migration Add Up

Especially at local levels, we cannot predict precisely how many children will be born, how many people will die, or how many people will immigrate to, emigrate from, or move within the United States. But the Urban Institute has extrapolated, from recent trends in birth, death, and migration, multiple scenarios for future population change. Each scenario paints a different picture of how America’s population is and could be distributed across the country.

US birthrates are much higher than those in European countries and Japan. In the early years of the 21st century, just about enough children were born every year in the United States to offset mortality. Birthrates of non-Hispanic white women were stable or rising slightly. Birthrates of non-Hispanic blacks and Hispanics followed a similar trend and were higher than non-Hispanic white rates. In addition, immigrants usually travel to the United States in their prime childbearing years, so the wave of immigrants from the 1990s caused a rise in the total number of births. Generally speaking, regions with more Hispanics and non-Hispanic blacks grow faster than less diverse regions because they are younger, reducing the impact of mortality and increasing growth from new births.

During the Great Recession, however, birthrates in all groups began a relatively steep decline (see the “Children and Youth in an Aging America” brief). Whether, and for which groups, birthrates will recuperate will have marked effects on certain regions. Since young adults account for most immigrants and most new parents, today’s national landscape of 20-somethings will have an outsize impact on the regional distribution of population in 2020 and 2030 and on the future labor force of the nation.

Mortality rates, another contributor to regional growth rates, have changed slower than birthrates or immigration, but here demographers can also envision differing scenarios for the future. White women, for example, have experienced a slowdown in their increase in life expectancy, the causes of which are unclear (Astone, Martin, and Aron, forthcoming). Will this slowdown continue? Will interregional differences in mortality even out over time or become more pronounced? Evidence on the reversal of death rates in the United States shows it to be geographically variable, so it is unlikely that mortality trends will affect population growth uniformly across the country (Kindig and Cheng 2013).

Immigration, emigration, and domestic migration have rearranged the United States in dramatic ways over past decades. For many years, immigrants arrived in “gateway” metropolitan areas like Los Angeles, Miami, and New York. Since 1990, however, immigrants also have been arriving in other metropolitan areas and rural counties, especially those in the Southeast, attracted by jobs in construction, agriculture and food processing, and the service sector. The direction of domestic migration has trended from cold to warm regions for decades; but some destinations grow fast in some decades and slow in others, and there are always exceptions to the rule. Domestic migration has also been slowing, even when controlling for shifting age and educational compositions of the nation (Kaplan and Schulhofer-Wohl 2013).1 It is unclear whether and by how much movements within the United States will continue to slow or if they will begin to rise again.

Though there is uncertainty in all three drivers of local population change, we are still able to make informed projections. Differences in fertility, mortality, and migration across the years and across the nation contribute to a sense of the plausible range of possibilities. We begin our projections, described in more detail in “Methodology and Assumptions for the Mapping America’s Futures Project” brief, with decades of data from counties, including census results, local and state birth and death rates, and the Census Bureau’s 2012 projections by race and age to the year 2030.2 From these data, we compute net migration from 2000 to 2010 as the difference between how population would have changed, accounting only for births and deaths, and observed changes between census years. We report results for 740 commuting zones (CZs), areas designated in the 1990s according to economic relationships among places. We also combine CZs into 24 regions, setting boundaries based on reviews of literature and observed differences in recent patterns of population change.

Since the migration scenarios yield the sharpest differences in the national distribution of population, we look more deeply, in this brief, into the implications for population change among the 24 regions of the low- versus the high-migration scenario. Of course, many other scenarios could be devised based on assumptions specific to people of a particular age group, race or ethnicity, or area. Building such scenarios about the future based on past trends, expert judgment, policymaker questions, and public speculations is an important activity in many cities and metropolitan areas now. With the models that helped produce the scenarios we discuss here, we can expand the process to consider future projections for the whole nation.

Foundations for Scenario-Building: Local Trends and National Projections

To develop scenarios of regional growth from 2010 to 2030, we begin with information about recent growth at the national and local levels and projections of national growth from now to 2030.

Between 2000 and 2010, the United States grew by about 27.5 million people. But that growth spread unevenly across the nation (figure 2, see also table A.1 in appendix A). The growth rate in the Central Florida, Rio Grande Valley, and Texas Triangle regions was over twice the national annual growth rate of 0.94 percent. Each of these regions grew for different reasons, with the Texas Triangle growing fast because of the strength of the oil and gas economy, Central Florida because of rising tourism and speculative real estate construction, and the Rio Grande Valley because of high birthrates and immigration from Mexico and Central America.3 The Southwest Triangle region (southern California, Las Vegas, and Phoenix) also grew fast, 1.35 percent a year, as a consequence of a natural increase in its large and diverse population; persistently robust arts, entertainment, and tourism sectors; job growth in information technology; and construction and real estate speculation. Stretching from Birmingham, Alabama, to southern Virginia, the Piedmont region also grew faster than average at about 1.44 percent a year. This growth was the result of immigrants and domestic migrants arriving for jobs in agriculture, construction, finance, logistics, and manufacturing. The Piedmont region’s larger CZs were especially attractive to non-Hispanic blacks living in the Great Lakes and Northeast Corridor, to young Hispanics who moved from gateway metropolitan areas and directly from Mexico and Central America, and to new immigrants from Asia and Africa.

Meanwhile, the Delta, Great Lakes, Gulf Coast, Northern New England, and Northeast Mountains regions grew at less than half the national rate. The Great Lakes, the most populous of these regions with 33 million residents in 2010, grew by less than 300,000 people between 2000 and 2010, less than 0.1 percent a year. Northern New England also experienced slow growth, and the Northeast Mountains (which includes Pittsburgh) actually lost about 100,000 residents from 2000 to 2010. Across many regions, economic distress or stagnation produced slow job growth and low job turnover within a workforce dominated by baby boomers. The Delta and Gulf Coast regions grew slowly, but for other reasons. Hurricanes Katrina and Rita displaced hundreds of thousands of people from southern Louisiana and Mississippi, some of whom remained in other regions within the Gulf Coast but many of whom moved permanently to Atlanta, Houston, and other cities in fast-growth regions. The Delta—anchored by economically struggling Memphis, but dominated in its land area by capital-intensive agriculture—does not offer enough job opportunities to retain its young people, and it has one of the nation’s highest mortality rates.

Exhibit 2

Contrasting Pictures of Regional Change from 2010 to 2030

Looking ahead, the US Census Bureau projects that the US population will grow by nearly 49 million between 2010 and 2030. The distribution of this growth by age will look different from the changes between 2000 and 2010 (figure 3). By 2030, we will have about 5 million more people under age 20 and 8.8 million more between the ages of 20 and 39. The number of people in their 40s and 50s is set to decline by over 600,000. Most staggering of all, however, the number of people age 60 and older will grow by a projected 35.7 million. In annual terms, this means that the 60 and older population will grow by over 2.5 percent a year for the next 20 years—over three times the national population growth rate of 0.74 percent a year.

Where in the country might this population growth concentrate? Local migration varies much more over time than either fertility or mortality—so much so that many observers do not try to project local populations at all. Yet it is important to create scenarios based on reasonable assumptions so that the nation and its regions can prepare for a variety of possible situations.

Exhibit 3

Nationally, domestic migration flows between 2010 and 2030 could significantly outpace those of the past 20 years. The millennial generation will be in its 20s and 30s—high-mobility years—over these two decades. This generation substantially outnumbers generation X, which made up the lion’s share of movers between 1990 and 2010. Interregional migration also increases after people retire, historically between the ages of 65 and 70 (perhaps a little later in the future, as retirement ages increase)—and the number of people in this age range will also be high through at least 2030. But, as mentioned, domestic migration rates have slowed even for those in their 20s and 30s, and so has immigration—both as a share of the total population and in absolute terms. (Our population projections combine domestic migration and immigration into a single number of “net migrants.”) Since migration varies more than fertility and mortality, we build our scenarios based on different assumptions about the pace of migration, holding fertility and mortality constant at their 2000–10 levels (the “Mapping America’s Futures: Population” tool lets the user vary all three components of demographic change).

All our migration scenarios begin from the assumption that the national population will grow to about 358 million between 2010 and 2030 (49 million more than in 2010), and that racial and age composition will be identical. They also presume that all migration flows will occur in the same direction between 2010 and 2030 as they did from 2000 to 2010. Our high-migration scenario projects that the slowdown in migration reverses.4 This scenario is most likely to occur if an array of economic factors restarted the economies of the South and West and constricted economic opportunity in the Great Lakes. The low-migration scenario, by contrast, projects further dampening of interregional migration, which would come about if a continued weak economy offered fewer incentives and lower levels of disposable income to fuel migration. The average-migration scenario, finally, would continue the trend toward lower overall levels of migration last seen in 1980, but it would not slow migration down as much as the low-migration scenario. All three scenarios maintain average levels of fertility and mortality developed for age-race cohorts at the state and county levels.

THE SOUTH AND WEST WILL GROW FASTER THAN THE MIDWEST AND NORTHEAST

According to the average-migration scenario the fastest national growth between 2010 and 2030 will occur in the South and West (figure 4, see also table A.2–A.4 in appendix A). The fastest growth of all would occur in the Texas Triangle region, where the population would increase from 19.5 to 25.9 million, or 1.4 percent a year. Among its CZs, Houston would top 7.6 million people by 2030, Dallas would exceed 5.6 million, and Austin, Fort Worth, and San Antonio would all reach over 2.75 million. The Piedmont region would also grow fast—over 1.3 percent annually—from 22.8 to 29.6 million between 2010 and 2030. This 6.8-million person increase gives Piedmont the biggest population growth of any region between 2010 and 2030 under this scenario. Atlanta would grow by 2.4 percent a year, even faster than it did from 2000 to 2010, rising to 7.5 million by 2030; Raleigh and Charlotte, North Carolina, would top 2.7 million; and Greensboro, North Carolina, Birmingham, Alabama, and Greenville and Columbia, South Carolina, would each exceed 1 million residents. Between 2010 and 2030, the Central Florida region also would continue its fast-paced growth under this scenario, summing to 14.4 million from its current 11 million, annual growth of about 1.2 percent a year.

The largest region in the West, the Southwest Triangle, includes Los Angeles, San Diego, Las Vegas, Phoenix, Tucson, and Yuma. Its 2010 population of 29.2 million would grow to 35.9 million—0.83 percent a year—under this average-migration scenario. The Los Angeles, Las Vegas, and Phoenix commuting zones would each add over 1 million new people, a relatively slow rate of growth for massive Los Angeles and a fast rate for Las Vegas. Expanding by 3.6 million to a combined population of 18.5 million in 2030, the Front Range and Basin and Range regions would grow by over 1.0 percent a year. California’s Central Valley region would grow by just over 1.0 percent annually under this scenario, to a total of 8.8 million people.

Exhibit 4

SLOW GROWTH IN THE GREAT LAKES AND NORTHEAST

Eleven regions, with populations that make up 56 percent of the 2010 US total, would grow more slowly than the national average rate of 0.74 percent a year between 2010 and 2030 under the average-migration scenario. The future of the Great Lakes concerns many observers. Its annual growth rate between 2010 and 2030 would be practically indistinguishable from the anemic pace of 2000 to 2010, with a net increase of only 750,000 from the current 33.4 million. While Chicago would grow by 640,000 and Milwaukee by 130,000, Cleveland, Buffalo, and Youngstown would lose nearly 100,000, 75,000, and 40,000 people, respectively. Population decline would dominate in the Northeast Mountains, including most of West Virginia, western Pennsylvania, and Upstate New York; the region’s population would drop about 90,000 between 2010 and 2030. State capitals and college towns, such as Lansing in the Great Lakes region and Morgantown and State College in the Northeast Mountains region, appear better positioned to sustain their population than manufacturing centers and smaller agricultural commuting zones.

The most populous of the 24 regions is the Northeast Corridor, which spans from Richmond to Boston. In 2010, nearly one out of every six Americans—over 48 million—lived in this region. Though growth would be slow, this average-migration scenario portends a net addition of 5.6 million new residents by 2030. The population would be 53.8 million—equal to today’s population of South Africa and larger than South Korea’s. Within the Northeast Corridor under this scenario, as in the United States as a whole, the population’s center of gravity would shift southward, as Charlottesville, Fredericksburg, Richmond, and Washington, DC, grow faster than the national average and Boston, Hudson Valley, New York, Newark, and Philadelphia grow more slowly. Providence would lose population under this scenario.

The Pacific region, which includes coastal California north of Los Angeles, also is projected to have slower growth than the national average under the average-migration scenario, increasing by only 850,000 people (0.45 percent a year). Like the Northeast Corridor, the Pacific’s high housing costs, expensive entry-level labor, and constrained land markets make it an engine of innovation but a challenging environment for population growth.

COMPARING SCENARIOS: HOW MIGRATION CHANGES THE BALANCE

Migration will probably be either slower or faster than an average-migration scenario; projections are famously inaccurate, even on average, especially in particular places and cases. Broadly speaking, and logically, slow domestic migration would stem some of the outflow of residents from the Northeast and Great Lakes toward the South and migration within some of the larger regions (figure 5). Faster migration, by contrast, would generate faster flows from “sending” to “receiving” CZs and regions. Under each scenario, we assume that migrants will have the fertility and mortality rates of people of their ages, races, and sexes at their destinations. This means that, even though we change only migration, the number of births and deaths in each CZ and region also change as a result of these migration shifts.

Exhibit 5

From the “sending” side of the migration ledger, the Great Lakes may have the largest stake in migration rates. People who would move out of the Great Lakes under the average scenario will stay there in the low-migration scenario, raising its population from 33.4 to 35.6 million residents (0.31 percent a year). Though still slower than the national growth rate, it would be nearly three times the growth rate of the average-migration scenario. Among larger Great Lakes CZs, only Youngstown would lose population under the low-migration scenario. The high-migration scenario would send even more Great Lakes residents south, resulting in a net decline of over 500,000 people; the low and high scenarios differ by over 2.6 million people by 2030. Chicago would still grow under the high scenario, but Detroit would decline by over 150,000, and the population losses in Buffalo, Cleveland, and Youngstown would be much more severe than in the average scenario. Among the other 23 regions, only the Northeast Mountains would lose population under the high-migration scenario, but it appears headed toward population loss even under the low-migration scenario.

Several other regions would also grow much more with low than with high migration. In the Southwest Triangle, for example, population would rise by 6.7 million in the low scenario instead of 4.2 million in the high scenario. The difference is accounted for entirely by Los Angeles, where the population of 18 million would grow by only about a half million under the high scenario but 3.5 million, under the low scenario. Las Vegas and Phoenix, by contrast, would grow faster under the high scenario than under the low scenario. This reflects in part the role of migration from Los Angeles to Las Vegas and Phoenix, which is much greater under the high than under the low scenario. Similar tradeoffs occur for the Northeast Corridor, where the low-migration scenario adds up to a million more people for the Corridor in 2030 than the high scenario, but it would deliver over 800,000 fewer people to Washington, DC, and 750,000 fewer to the three Virginia CZs in the corridor. Meanwhile, Boston, New York, Newark, and Philadelphia all grow more under the low- than under the high-migration scenario.

Below the surface of total population change, the enormous size of the baby boomer generation masks serious prospective challenges for these “sending” regions. For much of the United States, and especially for slow-growing regions, growth in the senior population helps offset declines in people under 20 and between 40 and 59 years old. For example, even in the low scenario, the Great Lakes population under age 60 would decline between 2010 and 2030 by 1.3 million, but this would be more than compensated by an increase in the 60-and-over population of 3.4 million. Under the high-migration scenario, the loss of people under 60 would be nearly 3.3 million, while the 60 and over population would grow by 2.8 million.

On the “receiving” side of the migration ledger, the Piedmont and Texas Triangle regions present an interesting contrast. Even with low migration, the Texas Triangle would grow by over 1.3 percent a year, only a little slower than under the high-migration scenario, because its population is so young (reducing the impact of mortality) and because Hispanics and others make up so many of its residents (increasing the number of children born). The Piedmont region, which is more sensitive to migration than the Texas Triangle, would grow less than 1.1 percent a year under the low scenario. Between 2010 and 2030, the two regions would have just about identical growth rates of 1.5 percent annually under the high-migration scenario. The difference between low and high migration for the Piedmont is over 2.7 million; it is only about 800,000for the Texas Triangle. Central Florida, with a smaller, older, and less racially diverse population than Piedmont, is even more sensitive to migration: the high-migration scenario would mean growth of over 1.6 percent annually, twice as fast as the low scenario. Even the low scenario would deliver a growth rate faster than the national average, but it would also result in a much older and less racially diverse region than would occur with greater migration levels.

Conclusion

The scenarios provide food for thought about how birth, mortality, and migration might play out differently across the nation. All three of these fundamental demographic drivers will affect a region’s future age structure, labor force composition, and diversity. Conversely, a region’s age structure, labor force composition, and diversity today will affect birth, death, and migration in the future. With the briefs in this series, we hope to encourage both learning about such likely trends as the growth of seniors and increasing diversity all over the United States and the increasing number of children in much of the United States and exploration of the magnitude of the potential impacts of the demographic drivers of regional change.

Notes

1. Kaplan and Schulhofer-Wohl (2013) study gross migration flows, which are about 10 times net flows, and show that Canada and Australia have experienced slowdowns in gross migration similar to one another’s and to changes in the United States.
2. For birth and mortality rates, we begin with the Census Bureau’s national projections by race and age to 2030. We then use local and state birth and death rates from the 1990s and 2000s as the basis for expectations about how localities will differ from the national average in the future (see the “Methodology and Assumptions for the Mapping America’s Futures Project” brief).
3. In our economic analysis for this brief, we relied on data from the Bureau of Economic Analysis’s Regional Economic Accounts. We compiled estimates for income by county annually from 2001 to 2012, with incomes assigned to major code groups in the North American Industry Classification System. Some counties with small populations had missing data for some of these major groupings; even so, we tallied the incomes for all the counties in each of our 24 regions to arrive at regional income totals for the major groups. The totals and trends are therefore illustrative rather than definitive, providing a broad-brush indicator of regional economic specialization. For each region, we compared the estimated percent of total income derived from each industry code with the national average to obtain location quotients. For example, in 2012, 1.2 percent of national income was derived from mining (including oil and gas extraction), but in the Texas Triangle, 7.1 percent of income came from mining, giving the Texas Triangle an estimated income location quotient of 5.8 for mining. This indicates a high level of dependence on oil and gas extraction for income in the Texas Triangle.
4. To be more precise, this is the sum of net migrants into regions that experienced in-migration, not the gross flow of migrants.

Appendix A. Population Tables

Exhibit 6

Exhibit 7

Exhibit 8

Exhibit 9

About the Authors

Rolf Pendall is director of the Metropolitan Housing and Communities Policy Center at the Urban Institute. In this role, he leads a team of over 40 experts on a broad array of housing, community development, and economic development topics, consistent with Urban’s nonpartisan, evidence-based approach to economic and social policy.

Steven Martin is a senior research associate in the Center on Labor, Human Services, and Population at the Urban Institute, having joined in 2013. He works on various topics in social demography; his particular area of interest has been modeling demographic events across the life course. His recent work has covered a range of demographic topics across the life course, such as nonmarital childbearing, fertility timing, childlessness, union formation and dissolution, and age at entry into sexual activity as well as topics in time use, well-being, the “digital divide” (the unequal diffusion of Internet and computer use in the United States), and the quality of data from event-history surveys.

Nan Marie Astone is a senior fellow in the Center on Labor, Human Services, and Population at the Urban Institute, which she joined in 2013 after serving 24 years on the faculty of the Johns Hopkins Bloomberg School of Public Health. She is a demographer with expertise on reproductive health, the family, adolescence, and the transition to adulthood.

Austin Nichols is an Urban Institute affiliated scholar who specializes in applied econometrics, labor economics, and public finance. His research focuses on the well-being of families and social insurance programs, including work on child poverty, disability insurance, income volatility, and economic mobility (within and across generations). He also studies education, health, and labor market interventions, and determinants of poverty and economic inequality.

Kaitlin Franks Hildner is a research associate in the Metropolitan Housing and Communities Policy Center. Her research focuses on using data and analysis to better inform decisionmakers at both the policy and the programmatic level. Her current work includes providing technical assistance to US Department of Education Promise Neighborhoods grantees on how to use and develop longitudinal data systems and assisting with the National Neighborhood Indicators Partnership.

Allison Stolte is a research associate in the Center on Labor, Human Services, and Population at the Urban Institute, where her research focuses primarily on policies and service designs relating to the workforce development of low-income individuals and young parents, youth development programs, and child food security. Stolte contributes to all stages of the qualitative and quantitative research process, including research design, data collection, management, and analysis. She has conducted and participated in over 100 data-collection interviews and focus groups across the United States and has conducted statistical analyses using quantitative program data. She also leads quality control strategies and both on-site and webinar trainings for program providers regarding proper data-collection procedures.

H. Elizabeth Peters is the director of the Center on Labor, Human Services, and Population at the Urban Institute. An economic demographer, her research focuses on family economics and family policy, specifically examining the effects of public policies such as divorce laws, child support policy, child care policy, taxes, and welfare reform on family formation and dissolution; inter- and intra-household transfers; father involvement; and family investments in children. She recently completed a program project grant funded by the National Institute of Child Health and Human Development on the transition to fatherhood and is currently coprincipal investigator on an Administration for Children and Families project that examines innovative practices involving fathers in home visiting programs.

ABOUT THE URBAN INSTITUTE

The nonprofit Urban Institute is dedicated to elevating the debate on social and economic policy. For nearly five decades, Urban scholars have conducted research and offered evidence-based solutions that improve lives and strengthen communities across a rapidly urbanizing world. Their objective research helps expand opportunities for all, reduce hardship among the most vulnerable, and strengthen the effectiveness of the public sector.

Funders do not determine research findings or influence scholars’ conclusions. Urban scholars and experts are independent and empowered to share their evidence-based views and recommendations shaped by research.

Copyright © January 2015. Urban Institute. Permission is granted for reproduction of this file, with attribution to the Urban Institute. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.