What are the Impacts of Fertility Rates on Housing Markets?

by George Masnick
Senior Research Fellow

Since families with children are primary drivers of household formation and housing consumption, changes in fertility rates can have significant impacts on housing markets. But tracking and understanding those changes can be challenging, as illustrated by two seemingly contradictory high-profile accounts of changing fertility patterns that appeared earlier this year.

First came the Pew Research Center, which in January 2018 issued a report titled "They're Waiting Longer, but US Women today are More Likely to Have Children Than a Decade Ago." However, less than a month later, The New York Times published the seemingly contradictory headline: "American Women are Having Fewer Children than They'd Like."

Is it possible that both headlines were accurate? Is it possible that more women are having children while the overall fertility rate also is trending downward?

Answering these questions requires paying attention to both the measures being used to describe fertility trends and the data source used to measure the trend. Such an approach shows that it is quite possible for more women to become mothers and for all women to have fewer children overall.

Explaining the Trends

The General Fertility Rate (GFR)—the number of births per 1000 women age 15-49—has been trending downward over the past decade, according to the National Center for Health Statistics (NCHS). The drop is due to sharp declines in the number of children born to mothers younger than 30, somewhat but not completely offset by increases in the birthrate for mothers older than 30. Consequently, the total fertility rate has declined (Figure 1).

Figure 1. The General Fertility Rate Has Been Declining Due to Steep Declines Among Young Mothers

Note: Since births to women ate 45-49 are so few, they were excluded from this figure, which makes the GFR line an approximation of the General Fertility Rate.

Source: JCHS tabulations of National Center for Health Statistics, National Vital Statistics System, Data Brief 287, Births in the United States, 2016.

It is noteworthy that, in 2016, for the first time ever, the fertility rate for 30-34 year olds exceeded the rate for 25-29 year olds. In contrast, the birth rate for women in their early 30s was about twice the birth rate for women in their late 30s, a trend that has no changed significantly in the past decade. Fertility rates for women in their early 40s did inch upward over the past decade, but remain at exceedingly low levels (rising from just under 1 percent to just over 1 percent).

What about the increase in motherhood highlighted in the Pew report? Motherhood is measured in that report by the share of women in each cohort having ever had a live birth by age 40-44. While that report indicates that the share of mothers is rising, there are important questions about the magnitude of its reported increase in motherhood. Specifically, the Pew report is based on an analysis of the Current Population Survey's biannual June Supplement, which asks women detailed questions concerning all children they have had over a lifetime. And these CPS data appear to show a significant recent increase in motherhood.

However, comparing the CPS data on the share of older women who have become mothers with vital statistics data from the NCHS suggests that the CPS may exaggerate the recent trend toward greater motherhood (Figure 2). Although the NCHS estimates are only available until 2010, the trends from the two data sources roughly parallel one another and show a sharp downward trend followed by a trend upward. But most importantly, the upward trend in the NCHS data began in 2000 and is more modest (a 2.1 percent increase over a 10-year period), while the CPS upward trend began in 2006 and is more dramatic (a 6.0 percent increase in 10 years). Significantly, most of the upward trend in the CPS since 2006 is accounted for by the change between 2010 and 2012, a period which accounted for more than half of the 6-point gain between 2006 and 2016.

Figure 2. Motherhood Is on the Rise, but Perhaps Not as Much as the Current Population Survey Indicates

Source: JCHS tabulations of US Census Bureau, Current Population Surveys, National Center for Health Statistics, National Vital Statistics System, Data Brief 287, Births in the United States, 2016.

The NCHS percentages of women who have become mothers are higher, partly due to the fact that the NCHS women are slightly older (all age 44), while some of the CPS women (age 40-44) were still having children in their early 40s. The NCHS trend line is also much smoother because it is based on much larger number of people in the vital statistics database.

However, some of the differences are due to measurement errors that produced the lower motherhood shares in the CPS prior to 2012. As a 2015 Census Bureau working paper on this topic noted: "The June 2012 Current Population Survey (CPS) Fertility Supplement data showed a significant decrease from 2010 in the percent of women aged 35-44 who are childless... However, due to numerous changes in data and data processing, it is reasonable to think that some of the apparent changes shown in the data may be artifacts of changes in measurement, not an indication of an actual demographic shift."

We should not be surprised that the more reliable NCHS data show that the percentage of all women age 44 who are mothers has been trending only modestly upward since about 2000. One key factor in this shift is that the percentage of women in this age group who are Hispanic also increased, rising form just over 10 percent in 2000 to just under 20 percent in 2016 (Figure 3). The growth in the share of women in their early 40s who are Hispanic is due to two trends. First, the number of Hispanic women age 40-44 has been increasing, as the younger migrants from previous decades approach middle age. Second, the total number of women age 40-44 has been declining because many members of the smaller Generation-X cohort are now entering their early 40s.

Figure 3. The Increase in Motherhood is Due in Large Part to the Growing Share of Hispanic Women

Source: JCHS tabulations of US Census Bureau, Intercensal Population Estimates and 2016 Historical Series.

This is significant because Hispanic women at the end of their reproductive ages are more likely to have become mothers than non-Hispanics of the same age. Unfortunately, the NCHS fertility data are only available for all whites (including Hispanics) and all blacks (including Hispanics). Consequently, we cannot use the data to calculate motherhood by both ethnicity and race. However, the CPS, which does ask about both race and ethnicity, shows significant racial and ethnic differences in the share of women in their early 40s who have had at least one child. The Pew report, for example, which averaged CPS data for 2012, 2014, and 2016, calculated that 90 percent of Hispanic women in their early 40s had at least one child, compared to only 83 percent of non-Hispanic white women; 85 percent of non-Hispanic black women, and 86 percent of non-Hispanic Asian women.

Potential Impacts

The NYT article emphasized that the fertility decline in the US is consistent with declines in other developed countries; that American women are bearing far fewer children that they would like to; that declines in marriage (and sexual activity among unmarried women), along with increasing use of reliable contraception, are at the root of the fertility shortfall; and that the fertility decline has been widespread throughout the country. Regardless of the reasons, this delay in childbearing could have a variety of impacts not only on individuals, families, and society, but on housing markets as well.

Fewer births to teens and women in their early 20s, for example, should mean that more women are likely to complete high school, pursue higher education, and secure higher-paying jobs. The Pew report describes how the largest increases in motherhood have been among college-educated older women, the group with historically the lowest levels of completed fertility and the highest percent childless. Discounting the fact that these motherhood gains might not be as large as the CPS data indicate, and are partly driven by increases in college-educated Hispanics, such a trend could have important implications for housing demand. College-educated mothers are likely to have higher incomes which means they are more likely to have the financial resources to become homeowners, should the choose, or to rent larger units in locations better suited for growing families.

However, fewer overall births and smaller family sizes could impact housing consumption by making renting more likely or by reducing the demand for larger housing units. Moreover, fewer births will produce a smaller future labor force that may find it hard to support the very large generation of millennials when they reach retirement. If doing so requires higher taxes on young workers, then households may have less disposable income that might otherwise be used to pay for housing.

Regardless of the impact on housing, it is clear that some subtle but significant changes are likely to continue to affect both the overall fertility rate, and the total number of children in the US. The fertility decline would be further exacerbated if, as some policy makers are proposing, the country reduces the number of immigrants allowed to enter the United States, or prioritizes immigrants likely to have fewer children.

Growing Demand and Tight Supply are Lifting Home Prices and Rents, Fueling Concerns about Housing Affordability

A decade after the onset of the Great Recession, the national housing market has, by many measures, returned to normal, according to the 2017 State of the Nation’s Housing report, being released today by live webcast from the National League of Cities. Housing demand, home prices, and construction volumes are all on the rise, and the number of distressed homeowners has fallen sharply. However, along with strengthening demand, extremely tight supplies of both for-sale and for-rent homes are pushing up housing costs and adding to ongoing concerns about affordability (map + data tables). At last count in 2015, the report notes, nearly 19 million US households paid more than half of their incomes for housing (map + data tables).

National home prices hit an important milestone in 2016, finally surpassing the pre-recession peak. Drawing on newly available metro-level data, the Harvard researchers found that nominal prices in real prices were up last year in 97 of the nation’s 100 largest metropolitan areas. At the same time, though, the longer-term gains varied widely across the country, with some markets experiencing home price appreciation of more than 50 percent since 2000, while others posted only modest gains or even declines. These differences have added to the already substantial gap between home prices in the nation’s most and least expensive housing markets (map).

“While the recovery in home prices reflects a welcome pickup in demand, it is also being driven by very tight supply,” says Chris Herbert, the Center’s managing director. Even after seven straight years of  construction growth, the US added less new housing over the last decade than in any other ten-year period going back to at least the 1970s. The rebound in single-family construction has been particularly weak. According to Herbert, “Any excess housing that may have been built during the boom years has been absorbed, and a stronger supply response is going to be needed to keep pace with demand—particularly for moderately priced homes.”

Meanwhile, the national homeownership rate appears to be leveling off. Last year’s growth in homeowners was the largest increase since 2006, and early indications are that homebuying activity continued to gain traction in 2017. “Although the homeownership rate did edge down again in 2016, the decline was the smallest in years. We may be finding the bottom,” says Daniel McCue, a senior research associate at the Center.

Affordability is, of course, key. The report finds that, on average, 45 percent of renters in the nation’s metro areas could afford the monthly payments on a median-priced home in their market area. But in several high-cost metros of the Pacific Coast, Florida, and the Northeast, that share is under 25 percent. Among other factors, the future of US homeownership depends on broadening the access to mortgage financing, which remains restricted primarily to those with pristine credit.

Despite a strong rebound in multifamily construction in recent years, the rental vacancy rate hit a 30-year low in 2016. As a result, rent increases continued to outpace inflation in most markets last year. Although rent growth did slow in a few large metros—notably San Francisco and New York—there is little evidence that additions to rental supply are outstripping demand. In contrast, with most new construction at the high end and ongoing losses at the low end (interactive chart), there is a growing mismatch between the rental stock and growing demand from low- and moderate-income households.

Income growth did, however, pick up last year, reducing the number of US households paying more than 30 percent of income for housing—the standard measure of affordability—for the fifth straight year. But coming on the heels of substantial increases during the housing boom and bust, the number of households with housing cost burdens remains much higher today than at the start of last decade. Moreover, almost all of the improvement has been on the owner side. “The problem is most acute for renters. More than 11 million renter households paid more than half their incomes for housing in 2015, leaving little room to pay for life’s other necessities,” says Herbert.

Looking at the decade ahead, the report notes that as the members of the millennial generation move into their late 20s and early 30s, the demand for both rental housing and entry-level homeownership is set to soar. The most racially and ethnically diverse generation in the nation’s history, these young households will propel demand for a broad range of housing in cities, suburbs, and beyond. The baby-boom generation will also continue to play a strong role in housing markets, driving up investment in both existing and new homes to meet their changing needs as they age. “Meeting this growing and diverse demand will require concerted efforts by the public, private, and nonprofit sectors to expand the range of housing options available,” says McCue.

---

Live Webcast Today @ Noon ET

Tune into today's live webcast from the National League of Cities in Washington, DC, featuring:

Kriston Capps, Staff Writer, CityLab (panel moderator)
Chris Herbert, Managing Director, Joint Center for Housing Studies
Robert C. Kettler, Chairman & CEO, Kettler
Terri Ludwig, President & CEO, Enterprise Community Partners
Mayor Catherine E. Pugh, City of Baltimore, Maryland

Tweet questions & join the conversation on Twitter with #harvardhousingreport

Projection: US Will Add 25 Million Households by 2035

by Dan McCue
Senior Research Associate

The United States will add 13.6 million households between 2015 and 2025 and another 11.5 million households between 2025 and 2035, according to Updated Household Projections, 2015-2035: Methodology and Results, a new Joint Center working paper. This growth represents an increase from the past decade that is in line with historic rates of growth seen in the 1990s, but still well below the levels experienced in the 1970s (Figure 1). An addendum to the paper indicates that the projected growth in households could lead to continued growth in residential construction activity.

Sources: JCHS Tabulations of US Census Bureau, Housing Vacancy Surveys, Decennial Censuses, and 2016 JCHS Household Projections  

The new household projections incorporate newer population projections from the US Census Bureau that are substantially larger than the Bureau’s 2012 population projections used in our 2013 estimates of household growth, which projected that the U.S. would add 12.4 million households between 2015 and 2025 and 10.35 million households between 2025 and 2035. Our new household projections also make methodological changes related to headship rates—the ratio of households to people—designed to reflect the fact that shifts in headship rates have significantly impacted household growth over the past decade. (As the paper discusses in more detail, rather than continuing the Joint Center's recent practice of holding current headship rates constant, the new estimates use trended headship rates.)

While both changes are significant, the increases from our 2013 household projections are due entirely to the new Census Bureau population projections. In contrast, the methodological changes produce slightly lower projected growth than our previous methodology. However, the methodological changes do affect the distribution of household growth by age, race and ethnicity. In particular, they increase household growth among the oldest age groups and also among non-Hispanic whites between 2015 and 2025. In contrast, they reduce growth among 25-44 year olds as well as from black and Hispanic households.

Despite these shifts, millennials and minority households are still projected to be the main drivers of household growth in coming decades. Indeed, millennials under age of 30 in 2015 are projected to form 23 million net new households between 2015 and 2025, while 72 percent of household growth overall is expected to be non-white households. At the same time, aging of the baby boom generation will bring the number of senior households up to unprecedented heights (Figure 2). Together these forces will reshape housing demand over the next two decades.

Source: 2016 JCHS Household Projections

In addition to the estimates on household growth, the working paper includes an addendum with baseline estimates for the amount of new residential construction that might be needed to accommodate future household growth, as well as the demand for replacement units, second homes, and other changes. Combined, these factors suggest that the baseline demand for new housing units between 2015 and 2025 will range from 16.0 to 18.2 million units. While this estimate is well above most recent rates of new unit completions and mobile home placements, it is consistent with historic averages for past 10-year periods (Figure 3). Although the analysis does not factor in estimates of over- or under- supply, the estimates do suggest underlying demand will support higher construction levels and that the growth in residential construction seen over the past five years is likely to continue.

Source: JCHS Tabulations of US Census Bureau, New Residential Construction data

Read the full working paper: Updated Household Projections, 2015-2035: Methodology and Results,

Three Scenarios for the Future of Homeownership Rates

by Jonathan Spader
Senior Research Associate

Following the rise and fall in the homeownership rate over the past two decades, considerable uncertainty exists about the homeownership rate’s future trajectory. In a new working paper, I present three plausible scenarios and examine the implications of different homeownership rate outcomes for future growth in the number of homeowner and renter households. (A supplemental working paper provides additional analysis of the factors that have contributed to the homeownership rate’s decade-long decline.)

The tenure projections for growth in the number of homeowner and renter households through 2035 build directly on household growth projections also released by the Joint Center this week. We use these household growth estimates, along with data on homeownership rates from the Census Bureau’s Annual Social and Economic Supplement to the Current Population Survey (CPS/ASEC), to construct three scenarios that reflect a range of possible homeownership outcomes.

• Scenario 1 (“Base Scenario”) – Constant homeownership rates. The base scenario applies the 2015 homeownership rates by age, race/ethnicity, and family type to the projected household counts for each year. This scenario therefore describes the likely outcomes if homeownership rates stabilize near their current levels. By holding homeownership rates constant, this scenario also reveals the implications of changes in the distribution of U.S. households by age, race/ethnicity, and family type for the future homeownership rate.
• Scenario 2 (“Low Scenario”) – Continued decline through 2020 followed by constant homeownership rates. The starting point for the low scenario is the set of 2015 homeownership rates for each age, race/ethnicity, and family type category. The low scenario then projects the 2020 rates for each category by applying the 5-year cohort trends observed from 2010-2015. The 2020 homeownership rates for each age, race/ethnicity, and family type category are then held constant to project the homeownership rates for 2025, 2030, and 2035. This scenario describes the likely homeownership outcomes if the homeownership rate’s ongoing decline continues for several more years before stabilizing.
• Scenario 3 (“High Scenario”) – Homeownership rates return to pre-boom levels. The third scenario applies constant homeownership rates determined by the maximum of the 1995 and the 2015 rate for each age, race/ethnicity, and family type category. This scenario uses the 1995 homeownership rates to define the pre-boom levels that might reflect a longer-term equilibrium. It then adjusts the rates upward to the 2015 rates for older households and other groups for whom longer-term upward trends have kept the 2015 rates above their 1995 levels. The resulting homeownership rates therefore define a high scenario in which homeownership rates increase to levels slightly above than their 1995 levels, but well below their mid-2000s peaks. While such homeownership rate increases may be more plausible over longer-term periods than in the next few years, the high scenario applies these rates to all time periods, providing estimates of homeowner growth if the rates are realized within each time horizon.

The base scenario shows that changes in the distribution of households by age, race/ethnicity, and family type will not substantially alter the homeownership rate between 2015 and 2035. Rather, the homeownership rate would increase slightly from 63.5 percent in 2015 to 63.7 percent in 2025 before falling to 63.3 percent in 2035. (Figure 1) Because the base scenario holds the rates for each age, race/ethnicity, and family type category constant at their 2015 levels, the changes (or lack thereof) reflect the cumulative effect of trends in the profile of U.S. households, such as population aging, increased racial and ethnic diversity, and delayed marriage and childbirth. The upshot is that these trends largely offset one another, affecting the overall homeownership rate only minimally. Instead, increases in the number of homeowner and renter households are driven by household growth, producing 8.9 million additional homeowner households and 4.7 million additional renter households by 2025, and 15.7 million additional homeowner households and 9.4 million additional renter households by 2035. (Figures 2 and 3)

Source: JCHS tabulations of CPS ASEC data

Source: JCHS tabulations of CPS ASEC data

Source: JCHS tabulations of CPS ASEC data

While the base scenario’s projections halt the decade-long decline in the homeownership rate, the projected homeownership rates remain below their levels between 1985 and 2015. This partial recovery reflects the possibility that slowing foreclosures and a strengthening economy will ease the downward pressure on the homeownership rate in coming years, while also allowing for the foreclosure crisis and Great Recession to carry some lasting impacts. The relative importance of these offsetting pressures will only be known with time, so the base scenario’s projections should be interpreted as a reference point for homeownership outcomes if the overall rate stabilizes around its 2015 level.

The low scenario describes the consequences of continued declines through 2020 before the homeownership rate stabilizes. Under this scenario, the projected homeownership rate falls from 63.5 percent in 2015 to 60.7 percent in 2020 before leveling off at 60.8 percent in 2025 and 60.6 percent in 2035. The homeowner growth figures show that the continuation of the 2010-2015 cohort trend implies minimal growth in the number of homeowner households, adding just 755,471 additional homeowner households through 2020. In subsequent years, the eventual stabilization of the homeownership rate at 2020 levels allows household growth to add 4.9 million homeowner households through 2025 and 11.6 million homeowner households through 2035. This sluggish growth in homeowner households is accompanied by faster increases in the number of renter households, with 8.7 million additional renter households by 2025 and 13.5 million additional renter households by 2035.

The projected declines in the homeownership rate through 2020 reflects the replication of recent cohort trends from the starting point of cohorts’ already-low 2015 homeownership rates. The projected 2020 rates therefore assume a continuation of the foreclosure-related homeownership exits, tight credit conditions, weak incomes, and other factors that contributed to the homeownership rate’s recent declines. Additionally, they assume the absence of any catch-up growth due to pent up demand among households unable to buy a home in recent years or to homeownership reentries among households that experienced a foreclosure. The low scenario therefore defines a trajectory that reflects the continuation of recent declines for several more years before the homeownership rate stabilizes.

In contrast, the high scenario projections describe homeownership outcomes under assumptions that project a reversal of recent declines and returns homeownership rates to levels slightly above the pre-boom period. The projected homeownership rates for the high scenario increase from 63.5 percent in 2015 to 64.9 percent in 2020, before leveling off at 65.0 percent in 2025 and 64.7 percent in 2035. This higher homeownership rate trajectory implies the addition of 10.6 million homeowner households and 2.9 million renter households by 2025, and 17.7 million homeowner households and 7.4 million renter households by 2035.

The higher homeownership rates produced by this scenario reflect the combination of 1995 homeownership rates with an adjustment for longer-term upward trends in the homeownership attainment of certain groups, particularly older households. While there is no clear “normal” equilibrium for the homeownership rate, this scenario adopts the 1995 rates as the most recent year that precedes the housing boom and bust. Additionally, it assumes that any groups with higher levels of homeownership attainment in 2015 compared to 1995 will sustain the higher 2015 levels into the future. This assumption implies an uptick in cohort trends that fully catches up to the level defined by the maximum of the 1995 or 2015 rate. This result may be particularly tenuous for middle-aged households, who experienced the most severe effects of foreclosures and may not reach the homeownership rates of prior cohorts. To the extent that the foreclosure crisis and Great Recession have had significant impacts for some cohorts, this scenario therefore assumes that such effects will be offset by broader changes in the economy, credit conditions, or housing markets over time.

Read the working papers:

Waiting for Homeownership: Assessing the Future of Homeownership, 2015-2035

Homeowner Households and the U.S. Homeownership Rate: Tenure Projections for 2015-2035

Updated Household Projections, 2015-2035: Methodology and Results

Increased Living with Parents among 18-34 Year Olds and the Implications for Future Housing Demand

by Daniel McCue
Senior Research Associate

The rise in the number and share of adults living with their parents is a well-documented trend that became increasingly apparent after the Great Recession.  It is also increasingly meaningful to housing markets as household growth slowed markedly in this period, largely as a result of fewer young adults forming households. And it is a trend that is ongoing. A report issued by the Pew Research Center found that for the first time in the modern era a higher share of adults age 18-34 are living with parents than living with partners or spouses.

In light of this information, one might conclude that as long as the rate of young adults living with their parents remains high, household growth will continue to be depressed. But even as the rate of adults living with parents continues to grow, the Census Bureau’s Housing Vacancy Survey also reported that household growth again increased in 2015 and has been accelerating since 2012.  If young adults―who are responsible for the majority of new household formation―are still living with parents at ever higher rates, how is it that household growth is picking up?  The answer lies in the shifting age distribution of millennials, who have now begun to exit the time of life where living with parents is most common and enter older ages where living with parents is less common.  With this shift, we can maintain today’s higher levels of living with parents among young adults and still have an acceleration of household growth.     

The 18-34 year old age group is also a very wide grouping for looking at living with parents, as the rate drops sharply across these ages.  Rates start at 50 percent among adults age 20-24 and drop down to 15 percent for adults age 30-34 (Figure 1).  This pattern basically mirrors the growth in headship rates (rates of being the head of an independent household) that rise most steeply for adults in their 20s. 

Source: JCHS tabulations of US Census Bureau, 2014 American Community Survey 1-Year Estimates.

In addition to being higher, rates of living with parents have also increased much more for the younger set of adults aged 18-34 (Figure 2).  According to tabs of the ACS, rates of living with parents in 2008-2014 grew most for 20-24 and 25-29 year olds, each up by roughly 6 percentage points.  Increases taper off with age from there, dropping to 4 percentage points for those age 30-34 and 2.5 percentage points for the age 35-39 year old age group.  Similarly, household headship rates dropped most for the younger age groups under age 30 and less for those older than age 30.

Source: JCHS tabulations of US Census Bureau, 2014 American Community Survey 1-Year Estimates.

Meanwhile, over the past decade the majority of population growth for young adults was skewed towards the younger side of this 18-34 year old group as the millennials replaced the smaller, generation-X population in the 20-24 and 25-29 year old age groups.  In addition to being the ages where rates of living with parents are highest, the sharp increases in living with parents that occurred among these age groups has meant that far fewer households were formed compared to what would have been expected given the magnitude of population growth.  Tabulations from the CPS show that declines in headship rates over 2005-2015 for the 15-19, 20-24, and 25-29 year old age groups reduced household growth by 1.7 million below what would have occurred under constant rates. 

Over the next 10 years, the aging of the millennial generation will shift the bulk of population growth from the 20-24 and 25-29 year old age groups to the 30-34, 35-39, and 40-44 year old age groups (Figure 3).  At these older age groups, changes in rates of living with parents and overall household headship have been much more moderate and remain closer to recent historical levels. 

Source: JCHS tabulations of US Census Bureau, United States Population Estimates and 2014 Population Projections.

This all suggests that future expected population growth in the 30-44 year old age groups will translate more directly into household growth over the next decade, even if living with parents continues to remain high for 20-somethings.  The pick-up in annual household growth levels since 2012 as reported by the Housing Vacancy Survey is a sign that this has begun.

Are Renter Worst Case Housing Needs Easing?

by Ellen Marya
Research Associate

Every two years, the Department of Housing and Urban Development (HUD) issues its Worst Case Housing Needs Report to Congress (WCN). This report highlights the challenges faced by low-income renter households in finding affordable, good-quality housing. In addition to data on characteristics of renter households and units, HUD’s report provides a count of renters facing worst case needs—defined as households who earn less than 50 percent of the area median income (AMI) who do not receive housing assistance from the government, who also are severely cost burdened (spending more than 50 percent on income on housing costs), and/or live in severely inadequate units. 

In its most recent WCN report released in May 2015, HUD noted a full 9 percent decline in the number of households with worst case needs, falling from 8.5 million in 2011 to 7.7 million in 2013. It was the first decline in that measure since a slight (1 percent) decrease in 2005-2007 and followed two periods of increases of about 20 percent. The change was surprising given that it coincided with a time of broadly stagnant incomes, rising rents, and a rapid increase in the number of renters. Do HUD’s numbers reflect genuine improvements in conditions for renters or are other factors at work?

A potential explanation for the decrease in worst case needs explored by HUD is a change in the income limits the agency uses to identify households earning less than 50 percent of AMI (very low-income households). Between 2011 and 2013, HUD reduced the maximum income for very low-income households by $516, decreasing the number of households in this group eligible to be counted among those with worst case needs by about 1 percent. When HUD compared the tallies of renters with worst case needs using the new and old cutoffs, however, it found that only 20,000 of the 750,000 total reduction 2011–2013 could be attributed to the new lower income limit.

Much of the decrease in worst case needs is due to a drop in households with severe cost burdens, which account for the vast majority of worse case needs. HUD reported that the total number of renter households with severe cost burdens fell from 10.4 million in 2011 to 9.7 million in 2013, a decline of over 6 percent. Counter to these findings, however, calculations from the Joint Center for Housing Studies (JCHS) using a different data source, the American Community Survey, found a negligible decline (just over 1 percent) in severely cost burdened renters, from 11.3 million in 2011 to 11.2 million in 2013.

Notes: Severe burdens are defined as housing costs of more than 50% of household income. In HUD tabulations, households with zero or negative income are excluded unless they pay Fair Market Rent or more for housing. For households paying no cash rent, utility costs are used to represent housing costs. In JCHS tabulations, households with zero or negative income are assumed to have severe burdens, while renters paying no cash rent are assumed to be without burdens.

Sources: HUD Worst Case Housing Needs: 2015 Report to Congress and JCHS tabulations of US Census Bureau, American Community Surveys.

Several unique methodological differences help contextualize why HUD and JCHS estimates vary (Figure 1). The first key distinction between the measures reported by HUD and JCHS is the source data. HUD estimates of cost burdens rely on the American Housing Survey (AHS), a biennial survey jointly administered by HUD and the Census Bureau assessing characteristics of the housing stock and its occupants. JCHS calculates cost burdens using the American Community Survey (ACS), an annual Census Bureau survey of households designed to supplement the short form decennial census. The surveys vary in size and scope. The AHS reaches 50,000-70,000 housing units in its national longitudinal sample, gathering detailed information on housing quality and cost, assisted status, and location. The ACS reaches 3.0-3.5 million households in the years since its full implementation and collects information on many demographic, economic, and employment characteristics, along with selected housing cost and unit information.

In addition to their variations in design, the AHS and ACS use distinct methods for defining occupied units that result in different counts for the most basic variables of total households (equivalent to total occupied housing units) and households by tenure. While several reports have examined these differences in more depth, essentially the ACS uses a broader definition of occupancy and makes more attempts to contact sampled households. These features of the survey tend to increase the number of occupied units reported and can count households in a seasonal residence (often rented) rather than their usual residence (possibly owned), increasing the number of renter households over the AHS (Figure 2). While not unique to the 2011-2013 period to explain the divergent trends, this difference in methodology consistently results in about 2 million more renter households in the ACS over the AHS, likely contributing in part to a higher ACS count of burdened renters

Sources: HUD Worst Case Housing Needs: 2015 Report to Congress and JCHS tabulations of US Census Bureau, American Community Surveys.

There are also important distinctions in how cost burdens are measured and what adjustments are made to the data. According to its WCN report, HUD excludes households reporting zero or negative income when calculating cost burdens, unless these households report paying the local Fair Market Rent (FMR) or more for housing. In this case, HUD assumes the negative income reported to represent a temporary situation and imputes a higher income for the household. If these households pay more than FMR and live in adequate, uncrowded housing, an income slightly higher than the local area median is assigned, again assuming a temporary period of income loss. HUD also makes adjustments for households that report paying no cash rent. For these households, HUD relies on reported utility costs to represent housing costs and identify housing cost burdens.

In contrast, JCHS assumes all households reporting zero or negative income to be severely cost burdened and all those paying no cash rent to be unburdened (in the case of a household reporting both zero or negative income and no cash rent, the household is assumed to be unburdened). The difference in adjustments may have had an especially large impact in 2011-2013 as JCHS tabulations of the AHS find the number of renter households reporting zero or negative income rose by nearly 13 percent, about four times the rate of growth of renters reporting positive income. ACS numbers do not mirror this rise, as renters reporting zero or negative income increased by 3 percent 2011-2013. Excluding zero or negative income households may better isolate households with perennially low incomes from those potentially higher-wealth households reporting temporary annual business losses. However, excluding these households from ACS analysis finds that severe cost burdens still do not drop nearly as much in 2011-2013 as HUD methods shows. Subtracting all households with zero or negative incomes from the ACS burden count shifts the totals to 10.4 million severely burdened renters in 2011 and 10.3 million in 2013, a decline of just 1.4 percent—much smaller than that reported by HUD for the period. Conversely, if all zero or negative income households in the AHS were considered burdened regardless of rent level, the decline in renters with severe cost burdens 2011–2013 would be about 4.6 percent.

In addition to varying counts of zero and negative income households, a disparity in median renter income patterns between 2011 and 2013 may also explain part of the divergent cost burden trends in that period. HUD cites an increase in median renter income of 7.2 percent in 2011-2013 in real terms as a factor driving down the number of severely burdened renters. While JCHS estimates of ACS data also find an increase in real median renter income in that time—the first increase since 2006-2007—the gain is a distinctly smaller 5.2 percent. HUD notes in its WCN report that some of the observed increase in median income may be due to newly formed higher income renter households, but does not further explore this possibility. Analysis of ACS data indeed shows that an influx of higher income renters occurred over this time. Of the net 1.7 million increase in renter households measured in the ACS 2011-2013, fully 1 million or 60 percent had incomes of $75,000 or more, over twice the median renter income (Figure 3). With this group pulling up the median figure, aggregate income gains may not have impacted lower income households sufficiently to meaningfully decrease the number of severely burdened renters.

Source: JCHS tabulations of US Census Bureau, American Community Surveys.

Indeed, analysis of the most recent 2014 ACS reveals the number of severely burdened renters is once again on the rise, climbing to 11.4 million—the highest number on record. Whether new AHS data expected in the upcoming months and the next WCN report due the following spring confirm this trend or show a further drop in severely burdened renters, the results of both surveys again underscore the acute unaffordability of housing for millions of renter households. Understanding whether affordability pressures are worsening or easing is therefore crucial to making informed decisions concerning rental assistance and other housing policy actions. Given additional data showing persistent rent growth and  tightness in the rental market, the larger sample size of the ACS, the benefit of an added year of ACS data showing rising burdens, and the unusually large recent shifts in renter incomes in the AHS, it seems likely that the enduringly high severe cost burden levels reported by the ACS are more accurate and affordability pressures for renter households continue to build.

Millennial Housing Issues in Perspective: Visualizing Cohort Trends in Population Size, Household Numbers, Ownership and Renting

George Masnick
Senior Research Fellow

Today’s 41 million young adults age 25-34 have been slow to move into independent household formation and homeownership. Exactly how slowly and why, and what the future likely holds for these individuals over the next decade, is the subject of much debate. The magnitude of delayed household formation and homeownership can perhaps be better appreciated if we directly compare this young cohort of adults with the cohorts that preceded them in the age structure. The four figures below track the different cohorts’ trends between 2003 and 2013 in population size, total households, owner households and renter households as measured by the American Housing Survey.

Population Size

Figure 1 shows population size of five different 10-year birth cohorts. The youngest cohort (born 1979-1988) remained fairly constant in size between 2003 and 2013 at about 41 million. This number is slightly larger than the next oldest cohort (born 1969-1978), but not as big as the cohort born 1959-1968, which includes younger baby boomers.

The cohorts born from 1959-1968 and 1969-1978 increased slightly in size over the 2003-2013 period due to migration from abroad, underscoring the fact that cohort size among young and middle-age adults can still grow as we go forward. Why the 1979-1988 birth cohort did not also increase in size between 2003 and 2013 (it actually declined by about 300,000) is likely due to the effects of the Great Recession having had a bigger impact on 25-34 year-old immigration. The growth in the total number of annual undocumented immigrants actually turned negative during this period, and slow job growth in construction and manufacturing also had a large impact on slowing overall immigration into the under 35 age groups.

U.S. immigration law still promotes family re-unification as one of its core principles, and this provision was less impacted by the economic downturn than employment driven immigration, and probably resulted in a more sustained immigration of 35-44 and 45-54 year olds. In addition, undocumented immigrants in these age groups were more likely to have lived here longer and have children born in the U.S., so they were less likely to have left the country during the Great Recession.

From 2003-2013, the two oldest cohorts between age 55-64 and 65-74 lost population due to mortality. The oldest (born 1939-1948) declined by 22 percent and the next oldest (born 1949-1958) lost 11 percent of its population.

Number of Households

Figure 2 shows parallel cohort trends in the number of households produced by the population in Figure 1. Three things are noteworthy. First, most of a cohort’s contribution to household growth occurs as it moves from age group 15-24 to 25-34, as is visible in the sharp upturn in households among the leftmost (youngest) cohort in Figure 2. Second, the cohort born 1969-1978 (red line) appears to have formed fewer households in 2013 at age 35-44 relative to older baby boomers at the same age than its population size might have predicted. The 1969-1978 cohort is not on track to attain the household numbers achieved by the 1959-1968 and 1949-1958 cohorts (green and purple lines). Third, the two oldest cohorts, although having lost a significant share of their populations from mortality, did not reduce their household numbers proportionally.

Lower levels of household formation in the youngest two cohorts when compared to baby boomers are somewhat expected because they contain higher shares of both foreign born and minority native born, each of which have lower rates of forming independent households. They are also cohorts that have experienced delayed marriage and fertility among the native-born non-minority population, making independent household formation for the youngest members of the cohort as a whole even less likely. But if members of these cohorts are simply postponing marriage and family formation, household formation for many is also being postponed, so future upward movement in household trajectories when cohorts are still under age 45 is likely.

The fact that household numbers after age 55 do not drop as quickly as population numbers is because married couples head most households in older age groups, and if one spouse dies, the household generally survives. In addition, divorce in middle and old age generally turns one household into two, partly offsetting deaths that occur to persons who live alone. After age 75, losses from mortality increase dramatically, so it will not be until after 2020 for the oldest baby boomers, and after 2030 for the youngest and largest baby boomer cohort that significant declines in older owner households take place.

Owner and Renter Household Trends

Decomposing the cohort trends in total household numbers into owners and renters further refines our understanding of the demographic underpinnings of recent household and housing market dynamics. The youngest cohort’s shortfall in household formation, as it moved into the 25-34 age group, was especially severe on the owner side, as shown in Figure 3.

In spite of having a noticeably larger population at age 25-34 compared to the next oldest cohort (red line), and a slightly larger number of total households at the same age, owner households were almost a million fewer. In addition, this next oldest cohort also shows levels of owner household formation well below what was achieved by the cohort born 1959-1968 (green line) when it was age 35-44 in 2003. Finally, the 1959-1968 cohort had slightly fewer owners in 2013 than the next oldest cohort (purple line) at age 45-54 despite having both 4+ million more people and 1.2 million more total households. But we must not lose sight of the fact that the older 1959-1968 and 1949-1958 cohorts aged into their 40s and 50s during a very different economic period (1993-2003) with better income growth, looser mortgage lending standards and more affordable newly built housing. The number of owner households that these older cohorts achieved at ages 25-34, 35-44, and 45-54 might not be a proper benchmark by which to judge the progress of today’s younger cohorts.

Figure 4 shows that in 2013, the number of renters in the youngest cohort at age 25-34 was significantly larger than the number of owners (11 million compared to 8 million). This compares to much greater parity between the number of owners and renters in the next oldest cohort when it was age 25-34. Although the number of owners in the youngest cohort was well below the number of renters in 2013, the increase in owners between 2003 and 2013 was still larger than the increase in renters.

Looking forward, the 1979-1988 cohort is going to add many more owners over the next 10 years, while at the same time its number of renters should decline when the cohort moves between ages 25-34 and 35-44, given historical cohort transitions. In fact, this youngest cohort should continue to add owners and lose renters over the next three decades until it reaches ages 55-64. Of course, the exact numbers of owner additions will be determined by the state of the economy, by income trends, by housing prices and mortgage interest rates, and by lending practices of banks and mortgage companies. To a certain extent, future homeowner numbers will also be determined by future demographic trends in marriage, fertility, immigration and mortality that affect this age group, but these are less likely to involve significant departures from recent historical levels and are more predictable.

By examining the cohort trends in the numbers of population, households, owners and renters in the way we have, we gain a greater appreciation of the degree to which millennials have been slow to form owner households. But we also find that the next older cohort, born 1969-1978, is also well below levels achieved by baby boomers when they were the same ages. There remains room for much upward movement in owner household formation for these two youngest cohorts. However, it is unlikely that these cohorts will ever reach the 16 million owner households achieved by each 10-year baby boomer cohort without significant reductions in the obstacles they now face in becoming and remaining homeowners. Still, we should look forward to continued gains in owner household formation for the two youngest cohorts as they move into their 40s and 50s over the next decade and beyond.

Article review- “Patriarchy, Power, and Pay: The Transformation of American Families, 1800-2015”

George Masnick
Senior Research Fellow

At my age, there is little that makes my jaw drop, especially while reading an article in one of my professional journals. However, this is exactly what happened when “Patriarchy, Power, and Pay: The Transformation of American Families, 1800-2015” by Steven Ruggles appeared in the latest issue of Demography. (Another almost identical version of this paper is available free of charge here.) What struck me as amazing is the way Ruggles provides a long-term perspective on many of the demographic and economic trends taking place today that I have studied using a much shorter time frame. And by long-term, we are talking 150-200 years!

The Joint Center for Housing Studies' early effort to describe changes in household structure and the labor force participation of American womenThe Nation's Families, 1960-1990– adopted a temporal perspective from 1960-1990. Published in 1980, we thought at the time that a three-decade perspective was all that was needed to understand the dramatic changes of that era. Wrong! The longer historical perspective sheds much more light on the origins of today’s demographic shifts, particularly in household structure, and what they might mean for housing.

Ruggles begins with the trend in the share of persons age 65+ who live in multi-generational families. We have noted the increase in this household type during the past two decades, primarily due to the increasing share of Hispanic and Asian immigrants for whom multi-generational residence is more common, and have speculated about its implications for housing consumption. But since we housing researchers rarely look at trends spanning more than 30 or 40 years, we have no sense of whether the upward trend in multi-generational living is indeed all that significant.

Ruggles’ Figure 1, reproduced below, shows how slight the recent turnaround has been relative to longer-term historical levels. The high share of the labor force based in an agricultural economy drove the very high historical incidence of older Americans living in multi-generational households. Three quarters of the labor force in 1800 worked in agriculture, and farm labor still was in the majority in 1850 when the share of 65+ living in multi-generational families was also 75 percent. Ruggles explains convincingly why an agricultural based economy tied the generations together, and why the rise of wage labor off the farm split them apart.



Ruggles’ main theme is that the decline of what he calls the “corporate family” – those working in agriculture and other (often related) family businesses – and the gradual transformation of the workforce to include first only male breadwinners, and later dual earner and female breadwinner households – had the effect of making household structures both simpler and more fluid. Once again, his long-term perspective is enlightening in looking at the recent trend in such things as delayed marriage and divorce. Age at first marriage for both men and women has been rising steadily since 1960, and he predicts that the share of never-married 40-44 year old women will almost double in the near future, rising from 15 percent in 2010 to about 28 percent in 2030. Similarly, the rate at which married women are divorcing has increased steadily since 1960, showing no sign of this trend slowing. Consequently, the share of all households without a married couple present – which held near 20 percent between 1850 and 1950 – has risen to over 50 percent in 2010, and continues its upward trajectory.

Nor is it simply the case that young adults are just trading marriage for cohabitation. To be sure, this is happening to some degree, but Ruggles notes that the share of 25-29 year olds without a co-residing partner has grown from 23 percent in 1970 to 48 percent in 2007 to 54 percent today. The fastest growing household type is single-person rather than cohabiting couples, as more and more adults of all ages who never married, are separated/divorced, and are widowed live alone.

If the household is the unit of both production and consumption, greater fragmentation and instability in household structures is troublesome. The primary household production good today is the next generation, and the U.S. appears to be following the lead of many European countries in developing fertility levels below replacement. Nothing that Ruggles presents in his paper provides comfort that the recent declining fertility rates are simply due to the lingering effects of the Great Recession and will likely reverse themselves.

One contributing factor to declining fertility may be trends in income. Households have always provided the mechanism for combining incomes. To Ruggles’ dismay, the evolving global economy is leaving more American households without secure incomes. The long slide in the relative earning power of young men over the past 40 years has been mitigated by the steady rise in employment of wives. But now that fewer and fewer households contain a married couple, and given that women’s real wages have also begun to decline, aggregate household incomes for married couples has begun to decline as well. Ruggles suggests that the largest source of decline in economic opportunity for young people, especially over the past two decades and in future decades, may be the automation of both manufacturing and services made possible by new technologies.

Housing consumption broadly should follow the downward trends in employment and income. Boosting household formation and homeownership rates, especially among the young, will require a reversal of many of the long-term demographic and economic trends that Ruggles discusses.

Ruggles’ article has sixteen figures, some only going back in time to 1940, but many spanning 150 or more years. I highly recommend you take a look. Some will surely make your jaw drop too.

Reconciling Different Household Counts from Census Bureau Surveys

by Dan McCue
Senior Research Associate

One of the major challenges faced by housing analysts and demographers is the lack of consistency among various Census Bureau surveys.  Particularly troublesome is the persistently wide range of difference reported by surveys in the number of households in the US, a key measure of housing demand.  In 2013, for instance, household counts reported by various Census Bureau surveys ranged from a low of 114.6 million in the Housing Vacancy Survey (HVS), to 116.3 million in the American Community Survey (ACS), to a high of 122.5 million in the Current Population Survey’s March ASEC (CPS/ASEC) - a span of fully 7.8 million households (Figure 1).  Annual surveys also differ widely in their measures of growth in the number of households, confounding efforts to gauge recent trends.  Indeed, household growth measures for 2013 ranged from 0.3 to 1.4 million depending on the survey, leaving data observers unsure whether growth in that year was historically weak or incredibly strong.

Source: JCHS tabulations of US Census Bureau data.

Given their importance to much of our work, the Joint Center has recently released a research note to highlight some of our current thinking on Census Bureau survey counts including: the differences in household counts among annual Census Bureau surveys and their causes; which surveys we believe to have more accurate counts than others and why; how we use different household count data for different analyses at the Joint Center; and how much of a difference using alternative headship rates based on different household counts would have on the current JCHS household projections.

The research note finds that the major source of difference among survey counts of households appears to be whether or not the surveys are person-based or stock-based.  Person-based estimates, which count the number of people who report being heads of households, consistently result in higher numbers of households than stock-based estimates, which count the number of housing units that are occupied.  We don’t exactly know why there is such variance in person-based and stock-based survey results, but the magnitude of difference between these two approaches is big and can be roughly approximated by the difference in household counts of the HVS and CPS/ASEC, because they are essentially stock-based and person-based versions of the same underlying survey sample. 

Among the annual surveys, the person-weighted CPS/ASEC has an advantage in that its household counts have come in closest to decennial census counts, which we take to be the benchmark. It is also a relatively timely survey and has the longest track record of matching census growth over the decades.  However, the CPS/ASEC does have a number of other shortcomings. One of the its major shortcomings, year-to-year volatility, can be reduced by smoothing over the data using rolling averages, but that approach also reduces the timeliness of the survey for measuring short-term trends in household growth. 

The second major annual household survey from which to get household counts, the stock-based HVS, is the most timely measure, providing quarterly results in addition to annual counts, and it also offers a series of annual household counts that use a consistent weighting vintage, which provides a more stable framework for measuring annual household growth trends than surveys that adjust their underlying survey controls year to year.  However, those vintage weighting controls do not eliminate a high amount of annual volatility.  Recent results underscore this fact, for it is highly unlikely that household counts in the HVS jumped fully 1.3 million between the third and fourth quarters of 2014 as reported in the HVS.

Finally, lack of timeliness is also a major disadvantage of using the third major annual census survey, the ACS, for analyzing annual counts of households: while other annual survey results have been out for months, the 2014 ACS is not due out until late 2015.  The ACS also has much lower household counts and appears to be essentially a stock-based survey with slightly higher household counts than HVS that is most likely a result of its more inclusive rules for determining occupancy of a unit.  Still, even with its lower base of counts, as a large and detailed survey the ACS may prove to provide a reliable measure of the growth trend in households, but so far this survey has had only a few years under a consistent weighting methodology in which to judge its reliability.

Overall, there is no satisfying conclusion to which annual survey is best for measuring annual household growth trends and none is perfect. CPS/ASEC is volatile but can be smoothed over at the expense of timeliness; HVS is timely and attractive in that its counts are pinned to stable consistent weighting vintages across years but it is still volatile, and possibly the vintage controls bias growth too low; and ACS is not timely enough to be helpful in measuring recent trends and is a survey without much history in which to judge its accuracy, though it has promise as a large and detailed survey that receives a relatively high amount of resources from the Census Bureau.  In terms of the number of households, however, there is reason to believe that higher counts of the CPS/ASEC, obtained from person-based weighting approaches, do appear to be preferable to lower counts of the HVS and ACS in offering counts closest to that of the official decennial censuses.  It is largely for these reasons we have used household counts from CPS/ASEC as a key input in Joint Center household projections, which apply current headship rates (the ratio of households to people) to population projections to produce household projections that form a baseline for estimating future housing demand.

Homeownership Rates Drop to Historic Lows; Middle Class Feels the Strain of Rising Rents

The fledgling U.S. housing recovery lost momentum last year as homeownership rates continued to fall, single-family construction remained near historic lows, and existing home sales cooled, concludes The State of the Nation’s Housing report released today by the Joint Center (live webcast today @ Noon ET). In contrast, rental markets continued to grow, fueled by another large increase in the number of renter households. However, with rents rising and incomes well below pre-recession levels, the U.S. is also seeing record numbers of cost-burdened renters (view our interactive maps), including more renter households higher up the income scale.

Perhaps the most telling indicator of the state of the nation’s housing is the drop in the homeownership rate to just 64.5 percent last year. This erases nearly all of the increase from the previous two decades. In fact, the number of homeowners fell for the eighth straight year, and the trend does not appear to be abating.

The flip side of falling homeownership rates has been exceptionally strong demand for rental housing, with the 2010s on pace to be the strongest decade for renter growth in history. While soaring demand is often attributed to the millennials’ preference to rent, households aged 45–64 in fact accounted for about twice the share of renter growth as households under the age of 35. Similarly, households in the top half of the income distribution, although generally more likely to own, contributed 43 percent of the growth in renters.

The other byproduct of this surge in rental demand is that the national vacancy rate fell to its lowest point in nearly 20 years. Given the limited supply of rental units, rents rose at a 3.2 percent rate last year—twice the pace of overall inflation. To meet this demand, construction started on more multifamily units in 2014 than in any year since 1989, and if job growth continues to pick up, we could see even more demand, as young adults increasingly move out of their parents’ homes and into their own apartments.

Even before the Great Recession, the number of cost-burdened households (those paying more than 30 percent of income for housing) was on the rise. But while the cost-burdened share of homeowners began to recede in 2010 (because some homes were lost to foreclosure, and low interest rates helped other homeowners reduce their monthly costs), the cost-burdened share of renters has held near record highs. In 2013, almost half of all renters had housing cost burdens, including more than a quarter with severe burdens (paying more than 50 percent of income for housing).

But perhaps most troubling, cost burdens are climbing the income ladder, affecting growing shares of not just low-income renters but moderate- and middle-income renters as well. The cost-burdened share of renters with incomes in the $30,000–45,000 range rose to 45 percent between 2003 and 2013, while one in five renters earning $45,000–75,000 are now cost-burdened as well. While affordability for moderate income renters is hitting some cities and regions harder than others, an acute shortage of affordable housing for lowest-income renters is being felt everywhere. Between the record level of rent burdens and the plunging homeownership rate, there is a pressing need to prioritize the nation’s housing challenges in policy debates over the coming year if the country is to make progress toward the national goal of secure, decent, and affordable housing for all.