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.

Some Thoughts on a Surprising Household Growth Estimate

by Dan McCue
Research Manager

Many media outlets and blogs (including our own), have reported on the results of the Housing Vacancy Survey (HVS) for the fourth quarter of 2014, which showed the US homeownership rate had dropped to its lowest point in fully 20 years. But the fourth quarter HVS contained another surprising reading—one that could be even more noteworthy than the continued fall in the homeownership rate. The HVS is one of very few sources of short-term estimates of household growth – an important gauge of housing demand.  And the surprise here was that HVS data show household growth going through the roof in the fourth quarter of 2014, with year-over-year growth in excess of 1.6 million households. This comes after household growth had long been stalled out, averaging less than 600,000 per quarter for much of the previous five years (Figure 1).

The concern and attention surrounding this number breeds from the thick cloud of uncertainty behind trends in household growth. Survey data from the Census Bureau such as the HVS, ACS, and CPS/ASEC give different and sometimes conflicting measures of household growth year-to-year, each with wide margins of error, which makes it difficult for analysts to call out trends with much confidence.  Amidst this lack of clarity is a widely held anticipation, or possibly hope, that household growth, having been ‘pent-up’ after such a long period of weakness, is primed to rebound strongly and this Q4 number from HVS might signal an inflection point.

Source: US Census Bureau, Housing Vacancy Survey Revised Estimates of the Housing Inventory: 2000 to Present, Table 8a.

One possible explanation for such an abrupt change in the rate of household growth in the fourth quarter HVS would be some change in how the survey is conducted or weighted that caused a discontinuity in the series. But the folks at HVS report that there were no structural or methodological changes to the dataset that would have been behind the sharp rise.

Without any methodological justification for the sudden jump, another factor may be some degree of sampling variation that produced an abnormally high estimate for the quarter. HVS has noted that quarter to quarter variability within the survey has increased and in many respects, this Q4 number is simply a prime example of how erratic quarterly data in the HVS can be and why we prefer not to make much of any one quarter and opt instead to look at rolling averages or other smoothed versions of this data to get a sense of recent trends.  But even averaged over the previous four quarters, Q4 still pulls the annual household growth reading for 2014 up significantly, to 789,000, representing a significant increase from the 524,000 annual growth reported for 2013, although still well below the long-run pace of 1.2 million per year that the Joint Center estimates is the baseline amount to expect given current levels of adult population growth and changes.

Alternatively, this fourth quarter increase could be a sign that the HVS is simply catching up to reality with its household counts after underestimating household growth over the past several years, In fact, prior to the 4th quarter results, the primary concern with HVS estimates was that they were overly low (see previous blogs on the topic here and here), in showing continued weakness in new household formation even as the economic recovery continued to gain steam.

Indeed, there are a variety of other market indicators that would suggest that household growth has been increasing at a modest pace in recent years and more than has been suggested in HVS quarterly releases prior to Q4. Most notably, employment growth has been ratcheting up over the last three years, from 2.3 million in 2012 to 2.4 million in 2013 and 3.1 million last year. Importantly, these gains have also been felt among young adults who are so important to household formation, with the unemployment rate of those age 25-34 dropping from 8.9 to 5.9 percent over this period. The slow rise in housing starts from 550,000 in 2009 to 1.0 million last year, at the same time that vacancy rates have declined, also suggests that household growth has been picking up steadily over this period.

In short, given the nature of survey data, we are not putting too much trust in the accuracy of this one quarter’s estimate of household growth reaching a 1.6 million annual pace, but do believe household formation has been gaining momentum, which bodes well for a stronger housing recovery in 2015. But there are also headwinds. Indeed, rising rents, declining rental affordability, and rising student debt levels remain barriers to household formation for many. Given the lack of clarity, certainly there is good reason to keep an even closer eye on this important measure over the course of the coming year. 

Homeownership Rates Continue to Decline, but Minority Owners Still Better Off than Before the Housing Boom

by Rachel Bogardus Drew
Post-Doctoral Fellow

The Census Bureau recently released its 2014 Q4 Housing Vacancy Survey (HVS) data, giving us a complete look at the boom and bust in homeownership rates over the last 20 years. The HVS’ reported homeownership rate, though far from perfect, remains the most up-to-date and cited statistic on homeownership in the U.S., and thus an important barometer of the housing status of American households. One advantage of the HVS is its long time frame, going back to 1965 at the annual and regional level, which helps to put recent trends in homeownership into a historical context. Figure 1 shows that the national homeownership rate rose steadily through the late 1960s and 1970s, from 63 to 65.6 percent, before declining slightly in the early 1980s. After a decade of stagnation, the rate rose rapidly from 1994 to 2004, from 64 to a record high of 69 percent. Since then, however, the national homeownership rate has declined almost fully back to its 1994 level.

Source: U.S. Census Bureau, Housing Vacancy Survey

The roller-coaster ride of the national homeownership rate from the last 20 years, while dramatic in its own right, only tells a small part of the homeownership story. Another advantage of the HVS data is reporting on homeownership rates for subsets of households by race/ethnicity and age, which add important texture to this story. Homeownership rates by race and ethnicity, for example, have generally followed the same up-and-down trend as the national rate, although no identified racial or ethnic group has given back all their gains. Indeed, when measured by the difference in their homeownership rates relative to 1994, only blacks are close to their former level (Figure 2). Households in the ‘other’ category (mostly comprised of Asians, Pacific Islanders, and multi-race householders), which grew their share of homeowners by 12 percentage points during the boom, remain fully 7 percentage points up today.

Notes: Hispanic includes all races. All other races includes multiracial. 

Source: U.S. Census Bureau, Housing Vacancy Survey.

The reason the national homeownership rate has declined all the way back to its 1994 level, even when no individual racial or ethnic group has done so, is because of the simultaneous shift in the race/Hispanic origin composition of households in the U.S., which has increased the share of all households that are headed by a minority. Due to the lower homeownership rates of minorities, relative to non-Hispanic Whites, this shift automatically lowers the homeownership rates of the nation, independent of any changes in minority-specific homeownership rate.

The second set of data from the HVS to shed more light on homeownership trends are the changes in rates by age of householder. Similar to the breakouts of homeownership rates by race and ethnicity, most of the age groups identified share the same rise and fall trend over the last 20 years, although householders ages 65 and older did not decline appreciably following the mid-2000s boom (Figure 3). The youngest households, those under age 35, gained the most in their homeownership rate during the boom, but have since declined to one-and-a-half percentage points below their pre-boom rate. Middle-age groups, meanwhile, have seen their homeownership rates fall nearly five percentage points below their mid-90s level, with 35 to 44 year olds experiencing a full ten percentage point swing in the last ten years. These dramatic declines are greater than the overall decline in homeownership because of the shifting age composition of households, which have skewed older over the last twenty years as the baby boomers progressed into higher ownership middle and early retirement ages, being replaced in the 25 to 44 age group by the smaller and more racially diverse Generation X cohorts with lower ownership rates. The low levels among younger cohorts, however, do suggest that the national homeownership rate is only back to its mid-90s level because of sustained homeownership by older households, and that among those in their prime working years the actual homeownership rate is indeed well below so-called ‘normal’ levels.

Source: U.S. Census Bureau, Housing Vacancy Survey.

A quick glance at these data may raise some concerns about the future of homeownership, especially for younger households and minorities who will account for large shares of households going forward. Yet caution should be exercised in extrapolating these trends forward.  For one thing, the HVS data often shows unexplainable jumps in quarterly rates and counts of households, which are somewhat muted by these annual averages. Second, recently moderating house prices and a move to relax lending restrictions, along with the reduction in FHA mortgage insurance premiums, should give home buying a boost. Third, most people, and especially young adults, remain in favor of homeownership, despite the risks exposed during the recession and foreclosure crisis. For these reasons, few analysts expect homeownership rates to fall much further. If it were to happen, however, the story would change from one of returning to a pre-boom norm of about 64 percent, to entering a new era of low homeownership rates unlike anything we have seen in the last 50 years. 

Pent-Up Demand for Additional Household Formation is Fraught with Uncertainties

by George Masnick
Fellow

In early 2011, economists at the National Association of Home Builders (NAHB) reported that the slowdown in household formation that started in 2007 with the advent of the Great Recession had produced a 2.1 million household formation shortfall by 2010. The authors concluded that the demand for new housing should accelerate dramatically once the economic recovery releases this “pent-up” demand. Another pent-up demand calculation, by Jed Kolko at Trulia, estimated 2.6 million “missing households” in 2010. After three additional years in which the economy has improved on many fronts – albeit at a slow pace – the 2013 Trulia deficit in the household count was still estimated at 2.4 million. But how solid are these estimates and how likely is it that household formation rates will return to pre-recession levels?

One difficulty in making these calculations is that actual household growth estimates since 2007 vary considerably from year to year and are inconsistent among data sets (Figure 1). There is good reason to believe that the most widely used data to track household growth, the Housing Vacancy Survey (HVS, used in the NAHB calculation), has seriously underestimated the number of US households – and as a result household growth – since a revision in methodology in 2003.  The HVS’s average annual estimate of household growth since 2007 of 550-600,000 contrasts with the American Community Survey’s (ACS) estimate of 700-800,000 new households annually and the higher Current Population Survey (CPS) growth numbers of over 1 million new households annually since 2010. Without agreement on actual levels of household growth since 2007, it is quite impossible to gauge the shortfall in growth, and therefore the probable level of pent-up demand. 

Notes: 2013 ACS not available.  2010-2013 growth for the ACS a two-year average of 2010-2011 and 2011-2012 data.

The method used to estimate the “normal” level of household growth also matters. The NAHB number was based on a simple difference between “actual” household growth estimates for the 2007-2010 period, and a straight line trending of HVS household growth prior to 2007. Over the very short run this approach may be appropriate, but would not be expected to hold up over a longer period.

Kolko’s calculations are more sophisticated. Using CPS data, he computes the change in age-specific headship rates (the share of persons in an age group that head an independent household) from the average 2000-07 pre-recession levels. This change, when multiplied by the official annual population estimates for each year, gives the deficit in number of household formations in each age group due to changes in the propensity to form households. This method corrects for the effects on household formation of simple changes in the size and age structure of the adult population, which the NAHB method does not take into account. But what Kolko’s calculation does not control for is the increasing share of minorities in the population. And since Hispanics and Asians have lower headship rates than non-Hispanic whites this oversight is not trivial (Figure 2). In fact, a certain amount of the decline in household formation is due to the changing race/Hispanic origin composition of the population and not to the recent economic downturn.

This issue is exacerbated by an undercounting of growth in Hispanics and Asians over the past decade, as revealed by the results of the 2010 Census. The underestimating of Hispanic and Asian shares of the population in the CPS during the 2000s also means that pre-2010 CPS headship values are biased upward by overcounting the white share, due to incorrect population weights in the CPS survey, making the 2000-2007 benchmark headship rates too high, and exaggerating the decline in age-specific headship pre-versus-post recession.  

Even controlling for both age and race/Hispanic origin in the different surveys, we know that household formations have slowed relative to pre-recession levels, we just do not know by how much given concerns just discussed. We also know that the slowdown is likely a consequence of the recession. But, we are uncertain about whether the reduced level of household formation has been primarily driven by economic factors, or whether it is the result of more fundamental changes in attitudes and behavior regarding independent living by today’s young adults that might be partly recession-driven, but may also have deeper roots.

Lower rates of labor force participation, lower incomes of those in the labor force, rising rents, greater student loan debt and tight mortgage lending conditions are economic factors that could partly explain low levels of independent household formation. But we do not know whether these effects are likely to be short-term or long-term as an improving economy and governmental initiatives could reverse many of these factors quite quickly. 

But trends in college and graduate school enrollment, the structure of the labor force, the timing of marriage and childbearing, and attitudes about co-residence might lead millennials to form independent households according to a different timetable than the generations that preceded them, regardless of economic conditions. Going back to school for retraining is becoming increasingly necessary for technology oriented jobs in a rapidly changing economy. Employment in start-ups, freelance work, and spells of temporarily working long hours in different jobs and on various projects, followed by periods of downtime, are increasingly common. The timing and sequence of important life-course decisions such as co-habitation, marriage, and childbearing have become more fluid. Intergenerational interdependency at various life-course stages has also changed, with parents playing a larger role in financially supporting their children as young adults, in helping to raise grandchildren, and in opening their homes for spells of co-residence when their children ask. These factors may have inertia that will make them less responsive to economic changes.  

Source: Joint Center tabulations of CPS data.  Average of 2011, 2012 and 2013 values.

And even if market forces are the primary reasons for depressed rates of household formation, geographic variations in job and income growth and housing costs and availability mean that the magnitude and pace with which pent-up household formation is released should vary in different parts of the country. For all these reasons calculations about the extent of pent-up demand for housing and speculation about its causes, when demand will be released, and what kind of housing will be required to meet future demand are fraught with uncertainties. The latest Joint Center household projections hold household formation rates constant at average 2011-2013 levels, making no allowance for the future release of pent-up demand, and should therefore be considered conservative.   

A Disappointing Report on Recent Household Growth Leads to More Questions than Answers

by Dan McCue
Research Manager

Although household growth is the major driver of housing demand, getting an accurate picture of recent trends in this measure is difficult, especially when Census surveys show conflicting trends.   On January 31, the most recent Housing Vacancy Survey (HVS) was released with Q4 numbers and some annual data for 2013.  As one of the few surveys that provide timely measures of household growth, the release was much anticipated in hopes that it would shed more light on trends of a recovery seen elsewhere in the housing market, but the results were disappointing, if not somewhat confusing. 

In its recent release, the HVS reported annual household growth of just 448,800 in 2013.  This represents a 48 percent drop in household growth relative to that from 2012 and marked the lowest annual household growth measure since 2008, in the depths of the Great Recession (Figure 1).

Source: US Census Bureau, Housing Vacancy Survey

In September we noted that the HVS was showing a disconcerting slowdown in household growth after finally having picked up in 2012.  With the annual number now in, this low measure of household growth in the HVS is puzzling, at odds with an assortment of other housing market indicators that have been painting a more positive picture for housing overall.  In particular, the drop in household growth did not mesh with several other trends:

• The much higher 1.375 million annual growth reported in the 2013 Current Population Survey Annual Social and Economic Supplement (CPS/ASEC);
• The same, steady increase in jobs in 2013 as during the previous year; and
• Increased momentum in the housing market, including a further decline in vacancy rates, an increase in new home sales, and an increase in housing construction during the year.

The divergent measure of household growth from the HVS is also troubling because while the HVS is known to have a downward bias in its estimated count of households, it has been useful in tracking short-term trends in that it provides more timely estimates than other sources and has generally been subject to less sampling error.  Indeed, while the CPS/ASEC and HVS both originate from monthly CPS surveys, the HVS annual household growth number is a 12-month rolling average of year over year growth, whereas annual growth in the CPS/ASEC is year over year growth for the single March survey, making it more volatile and less reflective of trends throughout the entire year.

But this is not the only—or most important—source of difference between HVS and CPS/ASEC household counts.  These two surveys differ more fundamentally in that CPS/ASEC arrives at its estimate of households based on weights derived from estimates of the total population and the share who are heads of household, while the HVS estimates households using weights that add up to estimates of the total housing units in the country, with the household count derived as the number of housing units that are not vacant (see Note on Table 3).  During census years, the CPS/ASEC head-counting method has generally produced totals much closer to the decennial Census –the Census Bureau’s benchmark survey of people and housing--while the HVS stock-controlled method has generally produced estimates that are lower by around 3-4 million households. This suggests the HVS household estimates are generally biased lower to begin with.

With its household estimates pinned to estimates of the housing stock, the surprisingly low HVS household growth estimate may be at least in part due to overly low estimates of growth in the total housing stock.  As shown in Figure 2, over the past few years, the total housing stock estimate used by the HVS has been growing slowly and very steadily since 2011, gaining around 350,000 units a year.  At the same time the Census Bureau’s New Residential Construction surveys show a significant upturn in the number of new housing units completed in 2012 and 2013, reaching 762,000 units.  In order for the HVS estimates of changes in the housing stock to be accurate, this would suggest a surge in demolitions that roughly offset the recent surge in new construction, which seems unlikely. 

Source: JCHS tabulations of US Census Bureau, Housing Vacancy Survey and New Residential Construction data.

There is a third census survey from which household growth can be measured, the American Community Survey (ACS), that might shed more light on the recent trend. The ACS is not as timely, however, and the results for 2013 are not due to be released until late 2014.  Still, for 2012 the ACS reported household growth levels in between the HVS and CPS counts (978,000), suggesting it might prove a moderate and viable tie-breaker between the other two surveys on the direction of the recent trend. But since the ACS household estimates are linked to the same housing stock estimates as the HVS chances are the ACS, too, will be subject to a downward bias.

Overall, the discrepancies in these surveys are troubling given the importance of household growth as an indicator of the health of the economy and the housing market.  For the time being, housing analysts are flying in the dark on this key metric.

Keeping an Eye on Trends in Household Growth

by Chris Herbert
Research Director

While home prices and construction levels are on the rise, the most recent estimates suggest that the rate of household growth slowed during the first half of 2013 from what had been promising indications of a rising trend through 2012 (Figure 1).  The rise in household growth last year had been a key source of optimism that the housing recovery was being built on the solid ground of expanding demand for additional housing.  The latest indication of a possible slowdown in household growth raises a cautionary flag that we have not fully turned a corner on the slowdown in household formation that began during the Great Recession. It also raises questions about how long a recovery in the single-family for-sale market can be sustained absent greater growth in homeowner households.

Note: Annualized growth is change in trailing 4-quarter average household estimate from  previous year; year-over-year growth is change in quarterly household estimate from previous year.  Source: JCHS tabulations of US Census Bureau, Housing Vacancy Surveys.

The source of the recent estimates is the quarterly Housing Vacancy Survey (HVS) conducted by the Census Bureau, which in addition to providing data on vacancies and homeownership rates also provides the timeliest available estimates of the number of owner and renter households.  The HVS quarterly estimates are somewhat volatile, but this volatility can be dampened by examining averages of the previous four quarters, which is how the annual HVS estimates are produced. There are also non-trivial differences in the count of households from this survey compared to the decennial census for reasons that are not well understood. Over the course of the past decade household growth estimates from the HVS were on the order of 15 percent below what the decennial censuses indicated, suggesting there is a downward bias in household growth from this survey. But while there are questions about the accuracy of the overall count of households from the HVS over time, there is still reason to believe that the survey is useful as an indicator of the latest trends in the direction of household growth.

The HVS data for the second quarter (Q2) of 2013 suggest that not only was the acceleration of household growth reported for 2012 less than originally reported, but also that growth has slowed in the first half of 2013.  Back in 2012, HVS data were showing that annual household growth in the US had rebounded significantly, to a level of 980,000 for the year after averaging less than 600,000 over the previous five years.  With the Q2 2013 data release, however, the HVS also revised its 2012 annual household growth estimate down by over 100,000 households to a less robust level of 857,000. 

On top of the downward revisions to 2012 growth, HVS data now also show that in the first quarter of 2013, the year-over-year pace of household growth plummeted back to the lowest rate since mid-2010, down to 403,000. While the Q1 number could have been a statistical blip, the Q2 number shows household growth still just 746,000 - well below last year’s annual pace.  Taking these two quarters into account, the most recent annualized household growth rate is just 751,000. Given this pace household growth will have to average nearly 1.2 million in the second half of the year just to match last year’s reported growth.

The Q2 HVS data also show that the homeownership rate failed to reverse its downward course in the second quarter of 2013.  While the Q2 rate of 65.0 percent is unchanged from Q1, it is still 0.5 percentage point below the homeownership rate from a year ago.  Of particular concern is that behind the homeownership rate decline is a contraction in the number of homeowners (Figure 2).  Based on a 4-quarter moving average estimate of households, the number of homeowners was down by 178,000 in the second quarter of 2013.  This contraction is on top of newly released revisions to historical HVS data that show fewer homeowners in 2012 than originally reported. In all, the number of homeowners in HVS is 1.5 million below its peak at the start of 2007.

Note: Household estimate is 4-quarter trailing average.

Source: JCHS tabulations of US Census Bureau, Housing Vacancy Surveys.

These trends suggest that recent gains in home prices and single family home construction have been driven more by a restricted supply of homes for sale amid demand from existing owners rather than by growth in first-time homebuyers adding to the number of owners.  With continued weakness in household growth and declines in the outright number of homeowners, the latest HVS data suggest that for-sale housing markets may face challenges to a sustained recovery if the inventory of homes for sale expands but the number of households looking for homes does not.  

Strong Demand for Rental Housing Driving Gains in Multifamily Construction

by Ellen Marya
Research Assistant

As the housing recovery gains momentum, one encouraging sign has been the strong return of multifamily construction. According to the Census Bureau’s Survey of Construction and Building Permits Survey, construction on 245,300 multifamily units was started in 2012, the most since 2008. (Figure 1) The surge in construction activity is only beginning to result in new supply on the market given the long lags from the time a project is conceived until construction is completed (only 166,000 units in multifamily buildings were completed in 2012, just slightly above the low point in 2011). But looking ahead, multifamily construction will continue to accelerate, as permits for over 310,000 multifamily units were issued in 2012, also the highest level since 2008. 

Source: US Census Bureau, New Residential Construction.

Gains in permitting have been widespread: three-quarters of the 100 largest metro areas accelerated their multifamily permitting in 2012. Nationwide, the multifamily rebound is outpacing improvement in the single-family market. Multifamily permitting was up over 51 percent between 2011 and 2012, more than twice the gain in single-family permits and the third consecutive double-digit increase. The rapid recovery in the multifamily sector has led to speculation that some markets may be in danger of overbuilding. But while recent gains are dramatic, a longer-term view of both supply and demand indicates that such concerns are likely overblown—at least for now.

Current increases in multifamily permitting are from historically low levels. From a peak of over 473,000 units in 2005, multifamily permits decreased by more than 70 percent to 142,000 in 2009, the fewest in 25 years. In the context of these drastic swings, permitting is just beginning to return to levels in line with long-term averages. Nationwide, in 2012, nearly 81,000 fewer multifamily units were permitted than the average annual level from 2000 to 2009. Permitting in 34 of the 100 largest metros did top average levels from the 2000s in 2012, including seven of the top ten highest permitting areas (Figure 2).

Source: JCHS tabulations of US Census Bureau, New Residential Construction.

This boost in supply is occurring in conjunction with rapidly growing demand for rentals. Nearly 93 percent of multifamily units completed in 2012 were rentals, the highest level in decades. According to the Housing Vacancy Survey, the number of renter households increased by over 1.1 million between 2011 and 2012, marking the eighth straight year of renter growth. Rentership was especially strong in each of the top ten highest permitting areas, where growth in renter households outpaced overall household growth between 2010 and 2011, the most recent years with metro-level data available from the American Community Survey. In total, these ten metros added 154,000 households between 2010 and 2011, but the number of renter households increased by nearly 268,000.

Additional signs indicate strong rental markets in these highest permitting areas. According to MPF Research, vacancy rates in professionally-managed apartment complexes were near or under 5 percent in eight of these markets as of the fourth quarter of 2012. Monthly rents in the ten markets were up an average of 3.6 percent in the fourth quarter of 2012 from the same quarter a year earlier, compared to 3.0 percent nationwide. As construction timelines for multifamily buildings often span several years, market conditions will continue to develop during the lag between permitting and completion of new units. However, generally tight markets and enduring renter growth suggest that the robust return of multifamily construction currently represents a response to rising demand, rather than the formation of a new bubble.

A Surge in Hispanic Household Growth? The Challenge of Interpreting Short-Term Trends in Datasets that are Occasionally Adjusted

by Dan McCue
Research Manager

Interpreting year-to-year changes in annual surveys from the Census Bureau can be a tricky business, especially around decennial censuses.  Because it is the largest and most comprehensive count of the population, after each new decennial census is released, the smaller but more frequently issued surveys available from the Census Bureau, such as the Current Population Survey (CPS), Housing Vacancy Survey (HVS) and American Housing Survey (AHS), are updated, or “re-benchmarked” based on the findings from the new decennial census.  Prior to this, these surveys were controlled to extrapolations based off of the prior decennial census. While it is inevitable that ten years of extrapolation can lead controls to drift off course, failing to recognize when and how datasets are re-benchmarked to correct for this drift can lead to misinterpretations about short-term trends.  The danger is that the re-benchmarking adjustment can be misinterpreted as an actual trend that occurred in a single month or year, rather than what it really is: a discontinuity in the data due to an adjustment made to correct the net sum of ten years of extrapolation errors that had accumulated in the dataset since the last decennial census.

Take for instance, the following data overview in a recent online article:

"The latest U.S. census figures, for June, show year-over-year Hispanic homeownership increased by 7.3 percent, from 6.2 million to 6.7 million. For black-owned households during the same time, the numbers dipped by 1.3 percent, from 6.3 million to 6.2 million. Likewise, whites' homeownership also saw a slight decrease of about 1 percent, from 58.4 million to 57.8 million." - National Journal

On its face, this data leads us to conclude that the number of Hispanic homeowners surged from June 2011 to June 2012, while at the same time the number of homeowners among both blacks and whites dropped significantly, and therefore without growth in Hispanic homeownership the overall number of homeowners in the US would have dropped significantly over the past 12 months instead of growing slightly as was reported.

However, the Census Bureau’s Housing Vacancy Survey (HVS) showed that both Hispanic and non-Hispanic homeownership rates dropped during the June 2011 to June 2012 period, a time wherein Hispanics also suffered higher than average unemployment rates. At first glance, the divergence in the two reports is puzzling. However, on the Census Bureau’s HVS website, there is a short but significant sentence under the “Changes in 2012” section of the Source and Accuracy of Estimates web page:   

“Beginning in the first quarter 2012, the population controls reflect the results of the 2010 decennial census.”  - HVS Source and Accuracy of Estimates

This note is important, because the distribution of occupied households by tenure, race, and ethnicity of households is based on these population controls.  Therefore, any changes in the number of homeowners by race and ethnicity that spans across the first quarter of 2012 is also incorporating change due to the shift in the distribution of households by age, race, and tenure that occurred with the re-benchmarking of the survey..

The adjustment to Hispanic households due to the re-benchmarking appears to be significant. Looking at the Hispanic share of households in HVS before and after Q1 of 2012, we can see that the re-benchmarking in that quarter led to a significant upwards adjustment that forms a discontinuity in this series (Figure 1).  The existence of a discontinuity is corroborated by data from the Current Population Survey, which re-benchmarked to the 2010 Census in 2011. The CPS Table Creator allows us to see the impact of the re-benchmarking directly by comparing the Hispanic share of households in 2011 under both 2000 and 2010 Census weights.  It shows that the 2010 census weights raise the Hispanic share of households a full percentage point, from 11 to 12 percent, compared to the 2000 census weights.  In short, this all suggests that results from the 2010 Census found that the 2000 Census-based population extrapolations had been underestimating Hispanic household growth in the 2000s, and therefore these household counts needed to be shifted upwards in 2012 as a correction.

Figure 1:  The Shift to 2010-Based Population Controls in Q1 of 2012 in the HVS Coincides with an Apparent Discontinuity in the Hispanic Share of Householders

Source: JCHS tabulations of US Census Bureau, Housing Vacancy Survey data.

With the change in population controls in the HVS in Q1 of 2012, the amount to which the shift in the distribution of households towards Hispanic households was underestimated incrementally over the last ten years gets corrected all at once, and gets attributed as change measured between Q4 of 2011 and Q1 of 2012.  And as we see in Figure 2, the quarterly change recorded in Q1 of 2012 has a huge influence over our view of the recent trend in household and homeownership growth by Hispanic ethnicity. 

Figure 2: Concurrent with the Switch to Census 2010-Based Population Controls, The First Quarter of 2012  Has a Large Influence on the Recent Trend in Hispanic and Non-Hispanic Household Growth 

Source: JCHS Tabulations of the 1995-2011 AHS

Without the ability to compare alternative HVS household counts for Q1 of 2012 under both 2000- and 2010-based population controls, it is difficult to determine exactly how much of the change in Hispanic and non-Hispanic households and homeowners in 2011 to 2012 was due to the re-benchmarking and how much was due to actual change measurable in the survey.  We refrain from presenting alternative scenarios here, but because the quarter is such an outlier, most assumptions to smooth or discount that quarter of data would conclude with much lower Hispanic household and homeowner growth and much higher growth among non-Hispanics over the past year.  

The Resurrection of Household Growth: The Missing Link in the Housing Recovery

by Chris Herbert
Research Director

New survey data released by the Census Bureau last week provides more strong evidence that the nascent housing recovery is being built on a solid foundation.  The latest Housing Vacancy Survey (HVS) (which provides quarterly updates on the number of households as well as vacancy and homeownership rates) indicates that as of the third quarter of this year U.S. households were increasing at a rate of just over 900,000 per year, giving a strong boost to overall housing demand.  This represents a substantial increase from 2011 when the same survey found that the country only added about 635,000 households.  While still well below the rate of 1.18 million per year that the Joint Center estimates the U.S. is likely to average over the 2010-2020 decade, the upturn in 2012 indicates that the moderate but steady pace of economic growth is finally translating into increased demand for housing, which bodes well for a sustained recovery.

Much of the attention on housing market woes has focused on the supply side. Certainly, as the housing bubble was bursting at the end of 2006 the excess supply of housing was a significant contributor.  As of that time, the U.S. had added more new homes over the previous 10 year period than at any other point going back to the early 1970s when record keeping for housing completions and mobile home placements began. But after five straight years of housing starts below 1 million units a year—a level last seen in 1945 when World War II was ending—it is hard to argue that the housing market is still suffering the after effects of overbuilding. In fact, by the end of 2011 the total amount of new housing put in place over the previous 10 years was the lowest since record keeping began.

Note: Available data go back to 1974.  Source: JCHS calculations of US Census Bureau data

Instead, the lagging recovery in the housing market has been more attributable to anemic growth in the demand for housing as indicated by weak household growth.  According to the HVS, household growth fell sharply in 2007 as the housing bubble collapsed.  (See Figure 2.)  After averaging 1.35 million from 2000 through 2006, over the next 5 years annual household growth barely exceeded 600,000.  Given the trends of the last five years, the spurt in household growth to an annual rate of 900,000 through the first three quarters of this year is notable.  If the upward trend in household growth continues, housing should see a sustained recovery in 2013.

Note: 2012 is an estimate of annual average growth based on trends through the third quarter of 2012. JCHS low projection assumes that immigration in 2010-20 is half that in the US Census Bureau’s 2008 middle-series (preferred) population projection. Sources: US Census Bureau, Housing Vacancy Survey; JCHS 2010 household growth projections