Variable Population Growth is Driving an Uneven Housing Recovery in the Nation’s Large Metropolitan Areas

by George Masnick
Senior Research Fellow

We easily accept the proposition that the housing recovery will be greatest in parts of the country where population growth and job growth are occurring more rapidly. But we often forget that the longer-term trends in population growth that drive housing demand are not only highly variable across metropolitan areas, but also tend to be persistent over time within metros. Places leading the housing recovery are the same places where the engines of population growth have had the greatest long-term sustained horsepower. These are places where the young adult population is growing most rapidly. Such places have younger age structures because they are destinations for both international and domestic migration, and their younger age structures sustain population growth from higher natural increase as well. Slow growth metros, where the demand for new housing is lower, are generally places with older age structures and a lack of net in-migration – places where these variables are not likely to change in a fundamental way in the foreseeable future. That being said, there have been changes in population growth among the nation’s large metros since the end of the Great Recession that are worth noting. Metros that have grown more slowly since before 2010 are still trying to put the effects of the economic downturn behind them. Metros that have higher population growth 2010-2014 are generally those where rising housing prices and rents have squeezed household budgets most severely.

The latest release of Census Bureau 2014 population estimates for metropolitan areas underscores the existence of large differences in population growth among the nation’s large metros. The nation’s 100 largest metropolitan areas in 2014 are home to about two thirds of Americans. The largest of these is the New York-Northern New Jersey-Long Island metro at just over 20 million, and the smallest is Durham-Chapel Hill at about 550,000. If we compare population growth that took place during the first decade of this century with what has occurred more recently, we can see both the longer-term growth differences among metros and identify places where population growth has accelerated or declined between 2010 and 2014.

Figure 1 plots annual population growth in the 97 of the 100 largest metro areas in 2010 that also made this list in 2014. Most of the 97 metros cluster together at under 25,000 annual population growth for both periods, and are growing moderately, slowly, or not at all. Only a couple of dozen metros exhibit population growth that sets them apart. For these, the higher the population growth in 2000-2010, the higher the growth in 2010-2014. For example, the Houston-Sugarland-Baytown metro area added an average of 123,000 people per year in the decade 2000-2010 and 134,000 per year from 2010 to 2014. Dallas-Fort Worth-Arlington increased 126,000 annually during the 2000s and 124,000 annually so far this decade. New York-Northern New Jersey-Long Island grew at an annual rate of 124,000 during each period.

The diagonal red line separates the scatterplot into metros that grew at a faster numerical annual rate during 2010-2014 compared to 2000-2010 (above the line) from those that grew more slowly (below the line). Among other moderately-higher growth metros, Atlanta-Sandy Springs-Marietta, Phoenix-Mesa-Scottsdale, Riverside-San Bernardino-Ontario, and Las Vegas-Paradise have grown more slowly since 2010, while Los Angeles-Long Beach-Santa Ana, the DC-VA-MD-WV metro, and Miami-Fort Lauderdale-West Palm Beach have grown more rapidly. Some metros with moderate growth during the previous decade have begun to grow more rapidly and add 30,000 or more people per year during the first half of the current decade. In addition to San Francisco-Oakland-Hayward and Boston-Cambridge-Quincy, Seattle-Tacoma-Belleview, Denver-Aurora-Lakewood, San Jose-Sunnyvale-Santa Clara and San Diego-Carlsbad are on this list of metros with significantly increased population growth. These are also places with the greatest increases in housing prices.

The Charlotte-Gastonia-Concord metropolitan area is an outlier in how much slower its population has grown in the recent period compared to 2000-2010. Such a slowdown should be surprising since Charlotte was not hit by the Great Recession and the bursting of the housing bubble as much as other metros falling well below the red line in Figure 1. In fact, its faster growth during 2000-2010 stems primarily from a large (+26%) adjustment to its baseline 2010 census count. It is the only metro in the top 100 with such a large percentage adjustment, which the Census Bureau states could be "due to legal boundary updates, other geographic program changes, and Count Question Resolution action."

Decomposing population growth into its two broad components, net migration and natural increase (excess of births over deaths), allows us to better understand these recent metropolitan population growth trends. Figure 2 shows that the greater the net migration the greater the natural increase. Since migrants are generally young adults, metros that are migrant destinations have a greater excess of births over deaths. This is especially true of metros like Houston-Sugarland-Baytown, Dallas-Fort Worth-Arlington, DC-VA-MD-WV and Atlanta-Sandy Springs-Marietta, where both domestic and international migration are strongly positive (Table 1). Places that are retirement destinations like Miami-Fort Lauderdale-West Palm Beach, Orlando-Kissimmee-Sanford, and especially Tampa-St. Petersburg-Clearwater, have much lower rates of natural increase (fewer births and more deaths) because of their older age structures.

New York-Northern New Jersey-Long Island is an outlier, with its high level of international immigration largely being offset by domestic outmigration (annually 141,000 and -124,000 respectively during 2010-2014). But the New York metro’s high share of minority population (50.2 percent according to the 2010 census) produces a large growth from natural increase because of younger age structure and above-replacement fertility for minorities. Los Angeles-Long Beach-Santa Ana’s profile is similar to New York’s in that levels of recent annual international in-migration are offset by high levels of domestic migration losses (62,000 and -49,000 respectively), and its high natural increase is fueled by its minority population (67.6 percent in 2010). The Chicago-Naperville-Elgin metro area has recently experienced more than twice the level of domestic out-migration than immigration according to Census Bureau estimates. Still, Chicago’s minority population (46 percent in 2010) produces a significant level of natural increase, which has kept the Chicago metro’s overall population growth positive.

Population age structure and percent minority are demographic characteristics that change little from year to year, and these are the characteristics that largely determine growth from natural increase. Places with high natural increase should continue to produce and excess of births over deaths. Looking forward, unless patterns of domestic or international migration change dramatically in the short run, high-growth metros should remain high and low-growth metros remain low.

Remodeling Spending Expected to Accelerate into 2016

by Abbe Will
Research Analyst

After several quarters of slackening growth, home improvement spending is projected to pick-up pace moving into next year, according to the Leading Indicator of Remodeling Activity (LIRA) released today by the Remodeling Futures Program at the Joint Center for Housing Studies of Harvard University. The LIRA projects annual spending growth for home improvements will accelerate from 2.4% last quarter to 6.8% in the second quarter of 2016.

Home improvement spending continues to benefit from the last years’ upswing in housing market conditions, including new construction, price gains, and sales. Strengthening housing market conditions are encouraging owners to invest in more discretionary home improvements, such as kitchen and bath remodeling and room additions, in addition to the necessary replacements of worn components such as roofing and siding.

Although we expect remodeling activity to strengthen through the first half of 2016, further gains could be tempered. Current slowdowns in shipments of building materials and remodeling contractor employment trends, as well as restrictive consumer lending environments, are lowering remodeler sentiment and could keep spending gains in the mid-single digit range moving forward.

--

The Leading Indicator of Remodeling Activity (LIRA) is designed to estimate national homeowner spending on improvements for the current quarter and subsequent three quarters. For more information about the LIRA, including how it is calculated, please visit the LIRA page on the Joint Center’s website. The LIRA is released by the Remodeling Futures Program at the Joint Center for Housing Studies in the third week after each quarter’s closing.

Single-Family Rentals Have Risen to Nearly a Third of Rental Housing

by Rachel Bogardus Drew

Post-Doctoral Fellow

According to the Census Bureau, the national homeownership rate dropped again in the second quarter of this year, to 63.4 percent. This level represents a nearly 50 year low, and continues the trend of declining homeownership that has been in effect since the end of the mid-2000s housing boom (see my previous blog post on this topic). The flip side of lower homeownership rates, however, is higher shares of households renting their homes. Indeed, rental housing is now more in demand than it has been for decades. While new construction of rental units has picked up in response to this demand, the majority of it has been served by conversions of existing units from owner- to renter-occupied, mostly from the single-family housing stock. As a result, since 2006 the number of single-family detached homes occupied by renters has increased by a third, from 9 million to over 12 million (Figure 1), and now accounts for 29 percent of all rental housing.

Note: Other single-family housing includes attached units and manufactured housing.
Source: Tabulations of the 2000-2013 American Community Survey.

My newest paper takes a new look at single-family detached rental housing, exploring the ways in which the stock and residents of these units differ from other rental housing, and how they have changed over the last decade. It finds that single-family rentals offer an important alternative to single-family owned and multifamily rental housing. Specifically, single-family rentals allow their residents to reap many of the advantages of single-family living, such as larger units than typically found in multifamily housing, while retaining the affordability and flexibility that makes renting an attractive option to households that do not or cannot own. Because most single-family rentals were formerly owner-occupied, however, they tend to be smaller, older, and have fewer amenities than currently owned single-family units (Figure 2).

Note: Other rentals include rented single-family attached and manufactured units.

Source: Tabulations of the 2013 American Community Survey.

While the characteristics of single-family rentals align closely with those of single-family owned units, residents of these homes more closely resemble other renter households. For example, the share of minority households among single-family detached renters (39 percent) is closer to the share among multifamily renters (48 percent) than among single-family detached owner-occupants (21 percent). The same is true of the age distribution of households; 30 percent of single-family renters are under age 35, compared to 38 percent of multifamily renters in this age group but only 10 percent of single-family owner-occupants. The pattern breaks down by family type, however, as single-family detached rental units stand out as having the highest share of families with children (Figure 3).

Note: Multifamily rentals include rented single-family attached and manufactured units.
Source: Tabulations of the 2013 American Community Survey.

While single-family rentals have characteristics that are different from single-family owned and multifamily rental units, also of interest is how these units have been changing as they have grown to become a larger segment of the rental stock. Looking at these changes over time may provide some insights into whether the recent surge in single-family detached rentals is a harbinger of housing demand going forward, or a temporary reaction to the downturn in the housing and home buying markets. Most changes observed over time in the structures themselves, for instance, reflect the evolution of single-family housing in general, which continually replaces smaller, older units leaving the stock with larger and newer units in desirable locations. Some changes in the characteristics of single-family renter households, however, do not follow the same trends as in all households. One notable example of this is the share of middle-aged households (i.e., headed by someone age 35-54), which has been declining in recent years among all housing types except single-family rentals (Figure 4). The same is true of families with children, who generally prefer the features associated with single-family housing, even if they do not or cannot own their homes.

Note: Multifamily rentals include rented single-family attached and manufactured units.

Source: Tabulations of the 2000-2013 American Community Survey.

It is unlikely, however, that these shifts represent a permanent change in the rental market. The middle-aged and family households that account for large increases in single-family rentals are traditionally those most active in the trade-up and first-time home buying market. If economic conditions change in the near future such that home purchases become affordable and attainable to more households, these new classes of single-family renters will probably be among the first to seize their chance to own their own home. In such an event, detached single-family units will decline as a share of all rentals, though likely only back to their former level of around a quarter of the stock, as these units will continue to provide alternative to multifamily rentals and single-family homeownership, and a necessary component of the national housing stock.