Real Estate Data for the Rest of Us

Homeowners and Renters … Together?

Homeowners and renters are much more likely to be neighbors in Florida than in the New York City area. The housing bust brought owners and renters closer together in most markets as many single‐family homes became rentals.

Jed Kolko, Chief Economist
August 27, 2014

Most neighborhoods have both renters and homeowners, and that mix depends on the housing stock. Most single‐family homes are owner‐occupied and most multi‐unit buildings are rentals. But there are plenty of exceptions. High‐rise neighborhoods often have both condo owners and apartment renters. And suburban areas with mostly owner‐occupied single‐family homes often have at least a few renters sprinkled in.

Whether renters and owners live together or apart matters for two reasons. First, when every neighborhood has both renters and owners, the choice of which neighborhood to live in isn’t limited by whether you rent or buy. But in a city where neighborhoods tend to be renter‐only or owner‐only, your choices are limited. In particular, people who can’t afford to buy a home have fewer neighborhoods to choose from when looking for a rental. Second, people care who their neighbors are. That’s especially true for homeowners. When they’re asked what’s important to them about their neighbors, they say they care most about whether the people who live near them are also homeowners, according to a September 2013 Trulia survey.

Renters and homeowners are more integrated—which is to say less segregated—in some metros than in others. But overall segregation dropped between 2000 and 2010, mostly because the housing bust turned some single‐family homes in predominately owner‐occupied neighborhoods into rentals.

Where Renters and Owners Mix – And Where They Don’t
To analyze how integrated or segregated renters and owners are in different metros, we used what’s called a dissimilarity index, which is a measure of how evenly two distinct groups are distributed across a geographic area. For each metro area, this index ranges from 0 to 1:

  • 0 indicates a completely integrated metro that has the same mix of renters and owners in every neighborhood.
  • 1 indicates a completely segregated metro where every neighborhood is either all owners or all renters.

The index equals the percentage of owners or renters in a metro that would have to move to a different neighborhood for all neighborhoods to have the same mix of owners and renters. (Throughout this post, “neighborhood” means Census tract. See note below.)

Among the 100 largest U.S. metros, the top four where renters and owners are most integrated are in

Florida. In the top three—Lakeland‐Winter Haven; North Port‐Bradenton‐Sarasota; and Palm Bay-Melbourne‐Titusville—fewer than 30% of households would have to move to equalize the mix of owners and renters across all neighborhoods. Overall, the 10 metros where renters and owners are most integrated tend to be in the Sunbelt.

Top 10 Metros Where Owners and Renters are Most Integrated
# U.S. Metro Dissimilarity index of owners vs. renters
1 LakelandWinter Haven, FL 0.28
2 North PortBradentonSarasota, FL 0.29
3 Palm BayMelbourneTitusville, FL 0.29
4 Cape CoralFort Myers, FL 0.31
5 Bakersfield, CA 0.32
6 Dayton, OH 0.32
7 Jacksonville, FL 0.32
8 Little Rock, AR 0.32
9 Charleston, SC 0.33
10 Greenville, SC 0.33
Note: among 100 largest metros. The dissimilarity index ranges from 0 to 1. A lower dissimilarity index means owners and renters are more integrated (i.e. less segregated). 

The large metros where owners and renters are most segregated are clustered in the Northeast and Texas. Eight of the 10 most segregated metros are within a three‐hour train ride of New York City. Newark, NJ is the most segregated. In fact, more than half of the households living there would have to move to a different neighborhood in order to equalize the mix of renters and owners across neighborhoods.

Top 10 Metros Where Owners and Renters are Most Segregated
# U.S. Metro Dissimilarity index of owners vs. renters
1 Newark, NJ‐PA 0.53
2 Fairfield County, CT 0.48
3 Dallas, TX 0.47
4 New York, NY‐NJ 0.47
5 New Haven, CT 0.46
6 Hartford, CT 0.46
7 BethesdaRockvilleFrederick, MD 0.46
8 Washington, DC‐VA‐MD‐WV 0.46
9 EdisonNew Brunswick, NJ 0.46
10 Austin, TX 0.45
Note: among 100 largest metros. The dissimilarity index ranges from 0 to 1. A higher dissimilarity index means owners and renters are less integrated (i.e. more segregated). 

Looking across all metros, owners and renters are more likely to be integrated in smaller, lower population‐density places. Owner‐renter integration is also higher in vacation areas. Among smaller metros—those ranked 101 to 250 in population—renters and owners are most integrated in two popular vacation areas: Hilo, HI (dissimilarity index of 0.19) and Lake Havasu CityKingman, AZ (0.21). In contrast, owners and renters are most separate in college towns because off‐campus student rental housing is often clustered in particular neighborhoods. (Dormitory residents aren’t considered households and are therefore excluded from this analysis.) The most segregated smaller metros are Ann Arbor, MI (0.53); Gainesville, FL (0.53); and College StationBryan, TX (0.50)—all of which are college towns.

The Housing Bust Brought Renters and Owners Closer Together
In 70 of the 100 largest U.S. metros, the dissimilarity index fell between 2000 and 2010, which is to say renters and owners became more integrated. Integration increased most in Las Vegas, Phoenix, and North PortBradentonSarasota, FL—metros where home prices fell 50% or more during the housing bust. In fact, between 2000 and 2010, the change in renter‐owner integration was strongly correlated with the severity of the housing bust.

The housing bust and foreclosure crisis led to an overall shift from owning to renting and a decline in the homeownership rate. This shift affected renter‐owner integration primarily because many foreclosed homes were bought up by investors and rented out. As a result, fewer neighborhoods were overwhelmingly owner‐occupied in 2010 than they were in 2000. To see this, we looked at the distribution of neighborhoods by renter share (weighted by neighborhood household count). For the U.S. overall, the share of neighborhoods that were nearly all owner‐occupied (i.e., 0‐10% renters) dropped from 9.4% in 2000 to 7.4% in 2010. At the same time though, the share of overwhelmingly rental neighborhoods (i.e., 90‐100% renters) didn’t increase. Rather, the share of more evenly mixed neighborhoods increased.

When we look at the metros hit hardest by the housing bust, it’s clear that the foreclosure crisis had the biggest impact on traditionally owner‐occupied neighborhoods. In the 15 large metros where home prices fell at least 40% from peak to trough, the share of neighborhoods that were nearly all owneroccupied (i.e., 0‐10% renters) dropped from 13.5% in 2000 to 5.9% in 2010. In other words, the likelihood of living in a neighborhood with very few renters dropped by more than half in hard‐hit metros.

Thus, the housing bust reduced the segregation of owners and renters. Integration increased the most in housing markets with severe price declines, a worse foreclosure crisis, and bigger increases in single-family rentals. As a result, neighborhoods with a renter share below 10% became rarer, especially in hard‐hit metros. That’s good news for people who have been shut out of homeownership. They can now find rentals in more neighborhoods than they could before the bust. But for homeowners who want other homeowners as neighbors, it’s gotten harder to find neighborhoods that are nearly renter‐free.

Note: the dissimilarity index is calculated for each metro based on the renter‐owner mix in all Census tracts within that metro. The data come from the 2000 and 2010 decennial Census. The 2008‐2012 American Community Survey also reports owner and renter households by Census tract, but covers five years that on average are no more current than the 2010 decennial Census. Furthermore, the ACS is a sample, while the decennial Census is a complete count of households.

Dissimilarity indexes can be used to measure the integration or segregation of any two distinct groups and are often used in research about racial segregation. For formulas and discussion of dissimilarity indexes and other measures of integration or segregation, see here and here.


Hottest Homebuilding Markets of 2014

Although construction activity remains well below normal nationally, homebuilding in Boston, New York, San Jose, Houston, and several other local markets is booming.

Jed Kolko, Chief Economist
August 18, 2014

Construction activity is a fundamental measure of local housing market health. That’s because homebuilding is both a signal of where builders are betting on future housing demand as well as a creator of local jobs. In this housing recovery, construction might even be a better measure of local market health than home price changes, which have been driven in part by investors and others buying undervalued homes.

Census building permit data reveal which markets are breaking new ground. Based on permits for the first half of 2014, we projected the level of building permits for the full year of 2014 for each metro, compared with the metro’s own historical annual average level of building permits from 1990 to 2013. Of course, the historical average level of construction ranges from a lot in places like Las Vegas and Raleigh to very little in places like San Francisco and Detroit, but we’re looking at how far each metro’s 2014 construction activity is above or below its own historical average.

Where Construction is Booming
Construction activity is highest relative to the local norm in Boston, New York, San Jose, and Houston, which are on track to build at least 50% more new homes in 2014 than their local historical average. The rest of the top 10 are in CaliforniaTexas, or Oklahoma. Of course, the normal level of construction in many of these markets – particularly Boston, New York, Los Angeles, and San Francisco – is low relative to most other metros across the country, but in 2014 they’re outperforming their own historical norm.

Top 10 Metros Where New Home Construction Is Highest Above Normal
# U.S. Metro 2014 annualized permit activity relative to metro historical norm Year-over-year asking home price change, July 2014 Multi-unit building share of 2014 permits
1 Boston, MA +73% 6.0% 71%
2 New York, NY-NJ +70% 5.6% 92%
3 San Jose, CA +69% 9.1% 75%
4 Houston, TX +61% 10.4% 34%
5 Oklahoma City, OK +42% 3.4% 23%
6 Orange County, CA +40% 6.6% 57%
7 Austin, TX +39% 12.3% 44%
8 Dallas, TX +36% 7.6% 51%
9 San Francisco, CA +36% 12.1% 84%
10 Los Angeles, CA +34% 9.3% 73%
Note: among the 80 of the 100 largest metros for which sufficient local permit data are available. Sources: Census and Trulia.

… continue reading


Where “Back to School” Means Private School

One way or another, good schools cost money. Private school tuition can easily exceed monthly housing costs, but home prices in top-rated public school districts are 32% above the local average.

Jed Kolko, Chief Economist
August 13, 2014

More than two-thirds of adults with children under 12 say that the neighborhood school district is among the most important considerations when choosing a home, according to a June 2013 Trulia survey. However, some parents factor schools into their housing choices differently. Nationally, 10% of school kids grades 1-12 attend private schools, and in some neighborhoods, the majority of kids go to private school.

In recognition of the back-to-school season, we analyzed where private school enrollment is high and low across the U.S. These geographic differences reveal why parents choose private or public schools for their kids.

For parents looking to move, knowing whether neighborhood kids go to private or public schools can help them decide where to live. First, a neighborhood’s level of private school enrollment signals whether you too might want to send your kids to private school if you lived there. Second, even if you plan to send your kids to public school, the share of neighborhood children in private schools affects whether most of the neighborhood kids will be at school with them.

Who Sends Their Kids to Private School
Let’s start with two essential facts about private schools, which explain a lot about who goes to private school:

  • Essential fact #1: Just 20% of private school students attend non-sectarian schools; the other 80% are in religiously-affiliated private schools, of which half are Catholic.
  • Essential fact #2: The cost of private schools is high, but varies widely. On average, tuition is almost $11,000, not counting discounts or scholarships. This ranges from $7,000 for Catholic schools and $9,000 for other religious schools to $22,000 for non-sectarian private schools. Tuition tops out at about $40,000 for the most expensive prep schools.

Given the high cost, kids from richer families are far more likely to go to private school than kids from poorer families. Only 6% of kids in households with incomes under $50,000 attend private schools, compared with 26% of kids in households with incomes of $200,000 or more.

… continue reading


Home Price Gains Now Driven More By Jobs Than By Rebound Effect

For the first time in 26 months, no housing market had a year-over-year price increase above 15%.

The Trulia Price Monitor and the Trulia Rent Monitor are the earliest leading indicators of how asking prices and rents are trending nationally and locally. They adjust for the changing mix of listed homes and therefore show what’s really happening to asking prices and rents. Because asking prices lead sales prices by approximately two or more months, the Monitors reveal trends before other price indexes do. With that, here’s the scoop on where prices and rents are headed.

Prices Rise 0.8% Month-over-Month in July
The month-over-month increase in asking home prices of 0.8% was in line with the average monthly gain over the past year, settling back down after a 1.2% month-over-month in June. The quarter-over-quarter increase of 2.5% remains below the level of last spring’s price spurt, when the quarter-over-quarter increase was 3% or higher in March through June 2013.

Although prices aren’t rising as fast as they did in spring 2013, price increases continue to be widespread, with 97 of 100 metros posting year-over-year price gains, and 94 posting quarter-over-quarter gains.


July 2014 Trulia Price Monitor Summary

% change in asking prices # of 100 largest metros with asking-price increases % change in asking prices, excluding foreclosures
seasonally adjusted
0.8% N/A 0.9%
seasonally adjusted
2.5% 94 2.6%
Year-over-year 7.8% 97 7.3%
*Data from previous months are revised each month, so data being reported now for previous months might differ from previously reported data.

… continue reading


Can Any Local Market Predict National Home-Price Trends?

Pay extra attention to Minneapolis-St. Paul home prices – they are the best local indicator of what will happen to national home prices one year later. Home price trends in Texas, however, are a terrible guide to what’s in store for the rest of the country.

Wouldn’t it be nice if there were a local housing market that we could use as the nation’s crystal ball? If one market regularly ran ahead of the national trends, we could pay extra attention to what’s happening there in order to know what the rest of the country should expect. During the housing bubble and bust over the last decade, there were clearly markets – like Las Vegas – that had more extreme swings in prices than others did, but being more extreme isn’t the same as being first.

To see which markets – if any – tend to get ahead of the national trend, we looked at home-price changes between 1980 and 2014 in the 100 largest U.S. metros and the U.S. overall, using the Federal Housing Finance Agency (FHFA) home-price index. Our crystal-ball score, calculated for each metro individually, is the correlation between the year-over-year home price change in that metro with the year-over-year home price change for the U.S. overall one year later. In other words, we’re measuring how closely the ups and downs in a local market’s home prices match the national ups and downs one year later. Remember that correlations range from 1 to -1: the higher the correlation, the stronger the forecast. A negative correlation means that a better year for a metro’s home prices is typically followed by a worse year for the nation’s home prices (and vice versa).

The Crystal Ball Award Goes to the Twin Cities
Among the 100 largest metros, the housing market with the highest crystal-ball score is Minneapolis-St. Paul. Other markets that are relatively good bellwethers include San Diego, Ventura County, and Sacramento in California; West Palm Beach and three other Florida metros; Washington, DC; and St. Louis. In general, these markets had a more severe housing bust last decade and faster historical price growth over the past three decades than other markets. But it’s an eclectic bunch, with St. Louis, Washington, and Minneapolis-St. Paul having had a milder bust than the markets in California and Florida.

# U.S. Metro Crystal-ball score: Correlation of local price change with following year’s national price change
1 Minneapolis-St. Paul, MN-WI 0.79
2 San Diego, CA 0.76
3 West Palm Beach, FL 0.76
4 Cape Coral-Fort Myers, FL 0.73
5 Ventura County, CA 0.73
6 Washington, DC-VA-MD-WV 0.72
7 Sacramento, CA 0.72
8 Palm Bay-Melbourne-Titusville, FL 0.71
9 North Port-Bradenton-Sarasota, FL 0.71
10 St. Louis, MO-IL 0.71

This graph shows what a 0.79 correlation actually looks like. Year-over-year home prices in Minneapolis – St. Paul tend to look like national changes but a little bit ahead:

AheadofCurve … continue reading