Can Homebuyer Counseling Support Sustainable Homeownership?

by Jonathan Spader
Senior Research Associate

HUD recently released a progress report —including a few early findings—from what could be a ground-breaking study of homebuyer education and counseling (HEC). While it will be several more months before the full study sample is ready for analysis, the early findings offer several insights about the value of HEC in helping potential homebuyers prepare for and sustain homeownership. Equally important, they confirm that implementation of the study is on track, successfully completing a field experiment that has the potential to produce detailed evidence about the impacts of homebuyer education and counseling. (Full disclosure: I’m currently an advisor to the study and previously served as its project director.)

The HUD study offers the first large-scale randomized-control trial of homebuyer education and counseling (although existing non-experimental studies  have shown promising estimates of HEC’s impacts). When enrollment closed earlier this year, more than 5,800 study participants were enrolled in 28 cities across the United States, with enrollments primarily occurring between January 2014 and January 2016. These participants were randomly assigned to one of three study groups: 1) in-person HEC offered through local housing counseling agencies; 2) remote HEC offered via internet and telephone, and 3) a no-services control group. For more details about this study design, see the full report. 

Looking forward, the critical tests will examine how HEC influences recipient outcomes, and whether such impacts translate into improved decisions during the home purchase process and, ultimately, into sustained homeownership. To better understand these impacts, the study’s future analyses will examine multiple measures across three domains: financial knowledge and management; home and mortgage search; and, homeownership sustainability.

For now, the early results offer a few data points that focus on the initial steps toward these outcomes, comparing outcomes 12 months after enrollment for a pooled treatment group (which combines the in-person and remote groups) versus the control group. These early analyses find several statistically-significant impacts of HEC:

• Treatment group members performed significantly better than control group members on a four-question measure of mortgage literacy.
• Treatment group members are significantly more likely to indicate that they would proactively contact their lender before missing a payment, a period when the lender has the most options for finding a solution that avoids default or foreclosure.
• Treatment group members are significantly more likely to have credit scores above 620, suggesting that HEC helped participants to correct errors or otherwise improve their credit scores in advance of home purchase.

Not all of the outcomes showed statistically-significant impacts. Treatment group members were not significantly different than control group members with respect to the fourth outcome tested—whether they regularly tracked their spending against a budget. A further caveat to these findings comes from the preliminary nature of the tests, which are based on a subset of the full study sample and examine only a handful of outcomes. Nonetheless, these estimates offer initial evidence that homebuyer education and counseling can play a valuable role in helping people prepare for homeownership. To that end, the early results report includes a couple of statements from study participants that are worth quoting directly:

“Just talking to the [housing counselor], it made me realize… what I could afford and, well, what I was preapproved for… She was adding on other expenses that I had totally forgot to add, you know, ‘cause I thought I had it all together. And I hadn’t. So I ran into a few things talking to her that made me realize that I probably need to just, you know, wait.”

 –Study participant in Chicago, IL

“[HEC] gave us the idea of whether we should go for it right now or not. It is really telling us what the timing [sic] if we are not really prepared and if we don’t have enough credit or other issues …you know, maybe it is not the right time for us. So it is really helping us to make the decision of go or no go.”

–Study participant in Dallas, TX

The full analyses of study participant outcomes at 12 months from study enrollment are is due in 2018, with analyses of longer-term impacts at 42 months from enrollment due in 2020. If the initial results are any guide, these future reports are likely to offer important conclusions about the value of HEC in supporting sustainable homeownership. 

Growth in Remodeling Spending Projected to Peak in 2017

by Abbe Will
Research Analyst

Strong gains in home renovation and repair spending are expected to continue into next year before tapering, according to our latest Leading Indicator of Remodeling Activity (LIRA) released today. The LIRA projects that annual growth in home improvement and repair expenditures will continue to increase, surpassing eight percent by the second quarter of 2017 before moderating somewhat later in the year.

Homeowner remodeling activity continues to be encouraged by rising home values and tightening for-sale inventories in many markets across the country. Yet, a recent slowdown in the expansion of single family homebuilding and existing home sales could pull remodeling growth off its peak by the second half of 2017.

Even as remodeling growth trends back down, levels of spending are expected to reach new highs by the third quarter of next year. At $327 billion annually, the homeowner improvement and repair market will surpass its previous inflation-adjusted peak from 2006.

For more information about the LIRA, including how it is calculated, visit the JCHS website.

NOTE ON LIRA MODEL: As of April 21, 2016, the LIRA has undergone a major re-benchmarking and recalculation in order to better forecast a broader segment of the national residential remodeling market. For more information on this, see our earlier blog post, and read the research note: Re-Benchmarking the Leading Indicator of Remodeling Activity.

Housing Recovery by Income in Two Metros: San Francisco and St. Louis

by Alex Hermann
Research Assistant

The increases in home prices that have occurred since the Great Recession not only vary across the nation’s metropolitan areas, they also vary within many metros as well. The San Francisco metropolitan area, where home values are now 16 percent above their pre-recession peak, and the St. Louis metropolitan area, where home values are still 10 percent below their pre-recession peak, illustrate these variations.

In both areas, median home prices in low-income ZIP Codes are less likely to exceed mid-2000 peaks than median prices in high- and moderate-income ZIPs. However, the regions vary when looking at the changes in house prices between 2000 and 2016. Over that time period, the percentage increase in median prices in the Bay Area’s low-income ZIPs was greater than the increases in high- and moderate-income ones. In contrast, the percentage increase in St. Louis’ low-income ZIP Codes was much smaller than the increase in that region’s high- and moderate-income ZIP Codes. (In this analysis, low-, moderate-, and high-income ZIP Codes have a median household income under 80 percent, between 80 and 120 percent, and above 120 percent of their state’s median income, respectively.)

Changes in home price also vary within both metros. For example, metropolitan San Francisco has had the eighth strongest post-recession recovery in home prices. As a result, median home values in San Francisco’s high-income ZIP Codes are about $1.18 million dollars while the median value in low-income ones are $586,000, more than three times the median price for the U.S. as a whole, which is $186,500.

However, home values in many of the region’s ZIP Codes are still below their pre-recession peak (Figure 1). In all, 31 of San Francisco’s 142 ZIPs, or 22 percent, have yet to regain their mid-2000 peaks, including:

• 50 percent (5 of 10) of low-income ZIPs
• 35 percent (12 of 34) of moderate-income ZIPs, and
• 14 percent (14 of 98) of high-income ZIPs.

Source: JCHS tabulations of Zillow Home Value Index data and ACS 2014 5-year data

Most ZIP Codes that have not regained their peak median home values are located on the outskirts of Metro San Francisco, particularly in northern Contra Costa County. That area is home to 10 of the 14 high-income ZIP Codes where median prices have not exceeded their pre-recession peak as well as 8 of the 12 moderate-income ones and three of the five low-income ones. Most of the remaining ZIP Codes where prices are still below pre-recession peaks are in the urban areas south of Oakland along the East Bay, which includes many low and moderate-income ZIP Codes as well as two high-income ones.

Although prices in San Francisco’s low-income ZIP Codes are less likely to regain their pre-recession peaks, the trend is different when examining price changes since 2000. Overall, home values increased in all the region’s ZIP Codes. But on a percentage basis, the values in low-income ZIP Codes increased more rapidly than those in high-income areas (Figure 2).

JCHS tabulations of Zillow Home Value Index data and ACS 2014 5-year data

The story is somewhat different in metropolitan areas that have not seen San Francisco’s rapid price appreciation, such as St. Louis, where home values in June 2016 were still 10 percent below their pre-recession peak. There, median prices exceeded their peaks in only 27 of 147 ZIP Codes, most of them located in the region’s urban core and suburban Madison County. (Figure 3). These unrecovered areas include:

• 1 of 35 (3 percent) low-income ZIPs
• 6 of 55 (11 percent) moderate-income ZIPs, and
• 20 of 57 (35 percent) high-income ZIPs.

JCHS tabulations of Zillow Home Value Index data and ACS 2014 5-year data

Moreover, unlike San Francisco, prices in low-income ZIP Codes in St. Louis have grown only modestly since 2000 and have increased much less than those in high- and moderate-income ZIP Codes. In the run-up to peak, prices in low-income ZIP Codes grew only marginally faster than prices in high-income ZIPs. Additionally, the post-recession upturn in home values in low-income ZIPs lagged the increase in high-income ZIP Codes by nearly two years (Figure 4).

JCHS tabulations of Zillow Home Value Index data and ACS 2014 5-year data

What to take away from this analysis? Overall, home values in high-income ZIP Codes have outpaced home-value gains in low-income ZIPs since the price peak of the mid-2000s. When taking a broader view, low-income ZIP Codes have performed as well as high-income ZIPs since 2000 in fast-appreciating markets like San Francisco, while in many lagging markets, like St. Louis, home value gains in high-income ZIPs have typically surpassed those in low-income ZIPs. Furthermore, though income levels are important they are not determinative. The geographic patterns also underscore the fact that trends in home values are also a function of features such as density and proximity to the central city.

These relationships, and others, will be discussed in a forthcoming Joint Center working paper on home value trends since 2000.

High-Income ZIP Codes Benefit Most from Housing Recovery

by Alexander Hermann
Research Assistant

Although home prices nationally have been on the upswing since early 2012, the increases have not only been uneven across metropolitan areas but are more likely to have occurred in the most affluent parts of each metropolitan area, according to a new Joint Center analysis of Zillow home value data.

Most notably, home values in high-income ZIP Codes that are home to their region’s more affluent residents are now about 1 percent higher than their post-2005 peak, while values in low-income ZIP Codes—which increased dramatically in the early 2000s—are still about 12 percent below their pre-recession peak. Moreover, home values in moderate-income ZIP Codes are still about six percent below their pre-recession peak (Figure 1). (In this analysis, low, moderate, and high-income ZIP Codes have a median household income less than 80 percent, between 80 and 120 percent, and above 120 percent of the state median income, respectively.)

Source: JCHS tabulations of Zillow Home Value Index data and ACS 2014 5-year data.

Moreover, home prices in low-income ZIP Codes are lagging both in recovered metropolitan areas as well as in metros yet to regain their peak price. Specifically, in recovered metros, 83 percent of high-income and only 65 percent of low-income ZIP Codes had median home values matching or exceeding their peak, a full 18-point difference. In metro areas within 15 percent of peak, but still below, 22 percent of high-income ZIP Codes have recovered relative to 9 percent of low-income ZIP Codes. In metropolitan areas furthest from peak—by one measure, those that remain hardest hit—only a sliver of low-income ZIPs (5 of 699) have recovered, compared with 37 of 899 high-income ZIP Codes (4 percent). In total, across the nation, 37 percent of high-income ZIP Codes have recovered, versus only 23 percent of low-income ZIP Codes (Figure 2).

Source: JCHS tabulations of Zillow Home Value Index data and ACS 2014 5-year data.

Extending the analysis to 2000 demonstrates why high-income ZIP Codes have been more likely to recover. Low-income ZIP Code home values increased tremendously during the housing boom, but a similarly harsh decline has made recovery more difficult, and has significantly weakened low-income ZIP Code home value gains since 2000 relative to high-income ZIPs. At peak, the median home value in low-income ZIP Codes more than doubled (increasing 101 percent) from January 2000 (Figure 3). The peak median value in high-income ZIP Codes increased only 82 percent. However, the post-recession decline wiped out a large share of the relative gains low-income ZIP Codes had made. In these ZIPs, median home values (as a percent of the January 2000 home value) dropped nearly 65 percent. In high-income ZIP Codes, the drop was 38 points. This precipitous decline, and a lagging recovery, have given high-income ZIPs a narrow edge in median home value increases overall. As of June 2016, median home values in high and low-income ZIPs were 84 and 76 percent, respectively, above their 2000 median home value.

Source: JCHS tabulations of Zillow Home Value Index data and ACS 2014 5-year data.

The overall trend varies somewhat when breaking ZIP Codes down into recovered and unrecovered metros. In recovered metros, median home value gains in high-income ZIP Codes have steadily outpaced those in low-income metros over time, sharply accelerating during the recovery (Figure 4). In unrecovered metros (which include nearly 70 percent of ZIP Codes in our sample), home values in low and high-income ZIP Codes have drawn about even in the long run (Figure 5). Figure 5 also shows that the metros worse off relative to past peaks are those where low-income ZIPs saw substantial home value gains relative to their initial home value and large declines during the recession. In these unrecovered metros, ZIP Codes in both categories have median home values about 79 percent above their 2000 values.

Source: JCHS tabulations of Zillow Home Value Index data and ACS 2014 5-year data.

Note: Percentage growth derived from nominal dollars.

Source: JCHS tabulations of Zillow Home Value Index data and ACS 2014 5-year data.

In an upcoming post, we’ll take a closer look one US metro that illustrates the uneven price recovery within its own ZIP Codes – San Francisco.

Metro Data on Rental Cost Burdens Show Uneven Improvement

by Alexander Hermann
Research Assistant

The national trend in cost burdens is reflected across most metropolitan areas of the US. Looking at the 100 largest Metropolitan Statistical Areas (MSAs) by population that have not undergone geographic boundary changes between 2005 and 2015, shows that in most metros, cost burden rates declined modestly for renters in 2015, but were still high relative to their levels in 2005. (A household is defined as cost-burdened when it spends more than 30 percent of its income on housing.)

Looking at cost burden rates among the top metros as a group, we find the number of metro areas with exceedingly high cost burden rates declined in 2015. Indeed, the number of metros where cost burdens affect at least half of all renters declined from 44 metros in 2014 to 33 in 2015, which is a significant improvement from 2010 levels, when cost burdens affected half of all renters in 65 metros. In total, between 2010 and 2015, fully 83 metros saw declines in the share of cost burdened renters.

Even with these improvements, however, the share of cost-burdened renters is still above 2005 levels in most metros. More than half of rental households were cost-burdened in 33 metros in 2015, an increase of 11 metros from 2005 (Figure 1). Moreover, renter cost burden rates in 66 metros were higher in 2015 than they were in 2005.

Source: JCHS tabulations of US Census Bureau, 1-Year American Community Survey estimates via Factfinder.

This trend is also evident for the more extreme measure of severe cost burdens (those paying more than 50 percent of income for housing). From 2014 to 2015, the number of metros with severe renter cost burden rates of 25 percent or more declined from 63 to 49 of the top 100 metros, respectively (Figure 2). This is a big improvement from 2010, when 79 metros had such high rates of severely cost-burdened renters, but still worse than in 2005, when it was just 37 metros.

Source: JCHS tabulations of US Census Bureau, 1-Year American Community Survey estimates via Factfinder.

Indeed, despite significant near-term improvement, severe cost burdens have yet to return to 2005 levels in most metros. In the 100 largest metros for which data extends back to 2005, 60 had larger shares of severely cost-burdened renters 2015 than in 2005.

Lastly, initial analysis finds that the 2015 data also show the profile of metros with the highest burden rates appears to have shifted somewhat. In 2015, metros with the highest shares of severely cost-burdened renters are generally the large metros with tight housing markets along both coasts, including New York, Miami, and Honolulu. In 2005, the profile of metros with this high share of severe cost burdens was different; though some coastal metros were included (like Miami and Stockton, CA), midwest and declining industrial metros were more prevalent among the severely cost-burdened metros (including Cleveland, Detroit, Rochester, and Memphis).

We’ll post additional analysis on this dataset in the coming weeks and months.

See the full metro Excel table for a complete set of metro-level cost burden data for 2015.