Will young adults make or break the housing industry?

In the last months there has been no shortage of public concern over the opportunities for America's millennial generation of young adults-be it their prospects for school, job, house, or life partner in the aftermath of the Great Recession. Such attention is certainly well merited.


The Great Recession was felt acutely across the American population but perhaps more so for our youth. While the unemployment rate for those over 34 peaked at about 8 percent, the unemployment rate among those between the ages of 18 and 34 peaked at 14 percent in 2010 and remains elevated, despite substantial improvement; delinquency rates on student loans have risen several percentage points since the Great Recession and even into the recovery; and the homeownership rate among young adults has dropped from a peak of 43 percent in 2005 to 37 percent in 2013 concurrent with a large increase in shared living with their parents.

As often occurs when precipitous changes appear to mark a generation, there has been much speculation regarding the economic forces driving the differences between millennials and those in my generation or my parents’ generation.

Are good jobs harder to find for today’s young adults? Will growing student loan debt hold back millennials’ homeownership opportunities? Will they ever purchase homes or do we have on our hands a “Lost Generation” consigned to a lifetime of renting, living in small condos, or even worse for all concerned-spending the coming decades living with their parents?

To be sure, these questions do not admit any straightforward answers, and it may take years before we are able to fully disentangle the interplay between the housing and labor markets in a satisfying way. But today I would like to look at millennials more closely and bring to bear some of the existing economic data and research on the issue in an attempt to better understand the trends in millennials’ behavior, with particular attention to its implications for the housing sector.

These data imply that it may be premature to suggest that trends among today’s youth are a reflection of a profound change in their preferences or attitudes in the aftermath of the Great Recession when many of these trends were on the horizon long before the recession began and they have been compounded by the large but likely temporary effects of the recession itself.

However, regardless of the cause of the recent developments, the ultimate outcomes for this generation will also depend on our policy choices.

Millennials in the economic recovery

There is no doubt that millennials have been very unlucky, given that many had to seek out a first job in what was an extraordinarily distressed labor market in 2008. As Peter Orszag recently noted-“Born in 1988? Sorry.”

The Great Recession was the most severe since the Great Depression, putting 8.6 million people put out of work, of whom 47 percent were between the ages of 18 and 34. While definitions vary, millennials are often defined as those born starting in roughly 1980 to the 2000s; that means that millennials were between 8 to 28 years old when the recession began in 2008 and are now 14 to 34 years old.

The overall unemployment rate for workers 18 to 34 peaked at 14 percent and has since come down to 9 percent, which is about 72 percent of the way back to its pre-recession average value but still unacceptably high. This is slightly less than the recovery for the overall unemployment rate-which is 83 percent of the way back to the average of the last expansion.

A range of other labor market indicators tell a similar story-a substantial recovery, but one that is not complete and is lagging somewhat behind the recovery for other age groups. Long-term unemployment for millennials, like for the broader workforce, is double its previous average. Younger men are more likely to be unemployed than young women-much like they were before the recession. African-American youth have more than twice the unemployment rate of whites, something that was equally true before the recession as after. The unemployment rate of young college graduates remains much lower than that of all other groups, although it has declined somewhat slower than those with less education. Finally, like with the overall unemployment rate, you see the labor market recovery not just in the official unemployment rate but also in broader measures of people without jobs, although the broadest measure which includes people working part-time for economic reasons, has further to go.

The Great Recession comes on top of a longer-term trend of reduced labor force participation and increased schooling by young people. Labor force participation among the youngest workers (those 18-34 years old) has fallen steadily since its peak in the late 1980s, and is now 5 percentage points lower than it was 30 years ago. However, at the same time the share of people in this age group in school has risen so that the fraction either working or in school has been roughly constant. Research has shown that both increasing enrollment and decreasing participation rates among students have contributed to the overall lower participation rate, with the former being the primary factor in the late 1980s and 1990s, and the latter driving the decline since then.

School enrollment generally has a cyclical component as young people enroll in school in greater numbers and stay longer in recessions, both because unemployment is high and in order to enhance their skills to become more competitive in a tougher job market. This increase in educational attainment was primarily at the college level with some additional enrollment at the post-college or graduate level. Encouragingly, most of this rise in college enrollments among millennials comes from students who are on track to complete their degrees on time, not from students who are avoiding the job market by taking five or six years to complete a four-year degree.

However, accompanying this increase in schooling has been an increase in the amount of student debt and an alarming increase in student debt delinquency rates. The Federal Reserve Bank of New York estimates in the first quarter that total student loan debt is nearly $1.1 trillion, more than double what it was in 2005 and larger than any other category of debt except for mortgages. Of total student loan debt today, roughly 11 percent has been categorized as seriously delinquent, up 4.4 percentage points since the end of 2005. Numerous factors can account for the increase in borrowing: growth in the student population, a shift in the student population to a higher ratio of lower-income students who are more likely to take out loans, a rise in the cost of college, lower household incomes as a result of the Great Recession, and more limited access to other forms of credit, such as home equity loans, that parents have previously used to pay for their children’s college educations. These factors manifest themselves in the rising average and median debt levels per household.

Millennials are also getting married later than previous cohorts, but as with educational attainment, the declines in marriage propensity for millennials are largely in line with a strong pre-existing trend, with little sign of a major change in the wake of the Great Recession. At least in part, the postponement of marriage may be the result of increased educational attainment-as people with college degrees tend to marry at older ages and this relationship has held even as the share of students going to college has expanded.

Finally, of particular importance to the housing market issues is the fact that record numbers of 18 to 34 year olds are currently living at home with their parents. This means they not only do not own homes but that they are also not a source of demand for rentals either.

This elevated share of young adults living at home is especially notable among millennials without jobs, which, when combined with observations about millennials’ student debt burdens and marriage rates, have raised numerous questions about how prolonged their delay into these conventional indicators of adulthood might be, including whether and when they will become homeowners. For example, it is easy to speculate how falling homeownership rates and the recovery of multifamily construction signal a shift in preferences toward renting among millennials, or at least those that have moved out of their parents’ homes.

In what follows, I will discuss whether this particular interpretation of the data is justified or whether a more nuanced view is useful in understanding the implications of millennials’ behavior for housing markets.

A closer look at Millennials

From 2012 to 2013, residential investment grew at double digit rates, making a major and timely contribution to the overall economic recovery just as fiscal support was waning. At the same time, we have seen strong home price growth, fewer underwater borrowers, and falling distressed sales. But even with this recovery, residential construction is well below its pre-recession, pre-bubble steady state levels-and in the last two quarters the residential investment component of GDP has been negative, although largely because of declining commissions associated with existing home sales than declining construction activity.

This restrained construction activity likely is a function of several factors: tight lending standards and supply material constraints to name a few. But, particularly over the long run, construction is largely demand-driven, for which a key demographic is millennials who now comprise the largest generation in U.S. history. The single largest driver of overall demand for residential construction is the number of new households formed, or “household formation.”

In recent years, getting an accurate picture of new household formation has been difficult due to an unprecedented divergence in readings from the two different surveys that are the standard sources for the data. These estimates suggest a wide range of 600,000 to 1.3 million households per year. If the reality is somewhere in between, household formation in recent years has been weak and below average household formation in previous decades.

Household formation is the product of two factors: the overall population and the “headship rate,” the share of population that either rents or owns, in effect the share of the population not living at home. The recent weak household formation readings mostly reflect a decline in headship, including for young adults. In fact, the decline in headship among young adults alone over the last decade likely resulted in a loss of over 1 million households cumulatively. Had this decline-and the consequent rise in young adults living at home-not occurred, millennials would be boosting housing demand, and in turn residential construction.

So what caused this drop in the headship rate in millennials and is it reversible? Some have speculated that this increase in those living at home and concurrent decrease in headship is a function of measurement in surveys, whereby those in college are classified as living at home. However, even among 18 to 24 year olds not enrolled in college and 25 to 34 year olds, who are presumably out of college, the share of individuals living with their parents also increased sharply during the Great Recession. So, while these measurement concerns may be part of the story, they do not appear to account for much of large increases observed for millennials.

One factor certainly at play is the business cycle and the impaired labor markets during the Great Recession. As documented in the economics literature, the headship rate has a significant cyclical component and business cycle conditions are the primary factors determining headship decisions in the short-run Indeed, since the 1970s, headship among young adults, shown by the green line plotted against an inverted axis, has roughly been the inverse of the unemployment rate for the same demographic.

However, the recovery in headship rates has fallen well short of the recovery in the unemployment rate-a point I will return to. More detailed analysis that we have undertaken looks state-by-state at the relationship between headship and unemployment confirms that cyclical factors account for a large part of the decline in headship during the recession.5 Moreover, the Great Recession strained the economic and financial resources of the parents of millennials, who may in turn have reduced their financial support of their adult children or have encouraged them to move back home to help with expenses.

These considerations suggest that as the economy and labor markets continue to improve, headship among millennials could rise and provide a boost to household formation. However, as I discuss below, the outlook for homeownership among millennials features more uncertainty in both the near and long-term.

The outlook for homeownership among Millennials

While headship is the most important factor in understanding the dynamics of housing markets, it is also important to understand whether a millennial owns or rents. On average, construction of a multifamily unit results in about 60 percent less contribution to GDP than a single-family unit would.

In the recovery so far, the decline in multifamily construction was relatively small and has fully recovered while single family fell dramatically and remains well from recovered. Some have argued that the Great Recession resulted in a profound shift in preferences for millennials toward renting so that, under this hypothesis, even if the headship rate and household formation were to recover, millennials would be a generation of renters rather than homeowners.

At this stage in the recovery, ascertaining whether there is a permanent shift in preferences is difficult to know conclusively but several factors call this interpretation of the data into question. For one, there does not appear to be a drastic shift in the homeownership rate over successive cohorts of young adults beyond the previously identified secular trend in “delayed adulthood.”

The long-run decline in ownership is present across income levels but is largest for high earners, consistent with larger declines in headship among this group. The homeownership rate among young adults features a secular downward trend explained by changing demographics rather than a drastic change in preferences. That said, as the sharp post-recession changes show, consistent with the longer-standing cyclical patterns, there is good reason to believe that the large increase in unemployment in the wake of the recession would have a commensurately large-but temporary-impact on both headship and homeownership.

Even with the economy well along the way to recovery, mortgage credit availability for households, and first-time homebuyers in particular, remains very tight and has shown only very few signs of easing. Indeed, research by Federal Reserve economist Neil Bhutta estimates that higher credit score thresholds used by lenders in the aftermath of the Great Recession can explain about 40 percent of the drop in first-time home buying in recent years relative to the early 2000s.

In contrast, builders do not appear to be facing as many constraints to finance multifamily construction as with single-family construction. Notably, issuance of private-label commercial mortgage-backed securities has recovered to pre-bubble levels and the Federal Reserve’s Senior Loan Officer Opinion Survey has reported net easing of standards for commercial real estate loans for the last 3 years.

Finally, the recovery in multifamily construction may be a byproduct of the uneven housing market recovery. To date, construction of multifamily units has been concentrated in markets with relatively inelastic housing supply. On balance, these markets feature relatively high density and currently do not have a large supply of vacant homes. In contrast, the less dense markets with a predominantly single-family housing stock were hit hardest by the recession and still have an outsized overhang of vacant homes. Hence the recovery in multifamily construction and the low levels of single-family construction that we see in the aggregate could reflect the different stages of the housing market recovery in these two types of markets.

The implications of student debt for homeownership

Many have posed the possibility for rising levels of student loan debt and delinquencies to weigh on homeownership. The theoretical justification for why this could be the case rests on at least three factors. First, all else equal, higher levels of debt can lead borrowers to trigger underwriting thresholds that make mortgages more expensive than they otherwise would be. Second, student loan delinquencies lower credit scores, which in turn raises the cost of credit. Lastly, student loan debt and delinquencies could lead young adults to live with their parents.

However, whether the burden of student loan debt is the ultimate cause of today’s low home buying activity among young adults remains an open question empirically. On the one hand, higher student loan debt is a byproduct of the higher enrollment seen in recent years. If the higher returns to education justify the investment that millennials made, then one can view today’s relatively low rates of homeownership among millennials as temporary-schooling, rather than debt, is delaying household formation and homeownership to later in the lifecycle. Eventually, these young adults will leave their parents’ nests, rent homes, get married, buy homes, and have children. In this case, the outlook for homeownership is positive. However, many young people, including those with student loans, will face costs from entering the labor market during such a terrible recession for years to come. In addition, many students financing their education through debt may have been inadequately prepared to manage the financial complications of managing debt. For both of these groups, the burden of servicing student loan debt could make it difficult to save for a down payment and buy a home.

Moreover, the student debt story is not just about averages-it is also importantly about the much greater likelihood of missing student loan payments among those who attend low-value added schools or do not complete their education. As a result, it could be having a greater impact on the housing market for these types of individuals.

While it may yet be revealed that millennials’ preferences have changed permanently in light of the recession, I hope that some of the evidence I have presented here today can temper such speculations. Although we have seen a slow but steady evolution in areas like college attendance, marriage and labor force participation for many decades, there is no strong reason to believe that millennials are dramatically different than the generations of Americans that preceded them. Rather, it is the unlucky economic times with which they were presented with that explains much of their challenge. We have the ability to do a lot to address those challenges-and the onus is on us to do so.

Author: Jason Furman, chairman of the White House’s Council of Economic Advisers, at the Zillow Housing Forum in Washington, D.C.