Skip to main content

The FRED® Blog

A housing recovery without homeowners

Historically, the cost of buying a house has been positively correlated with the percent of households that own their home. During 1996 to 2006 in the United States, both the price of houses and the homeownership rate increased. This increasing trend ended abruptly with the global financial crisis, which saw house prices plunge and drove homeownership rates to historically low levels. If homeownership became less attractive in the wake of the financial crisis, we might expect both prices and homeownership to decrease. Similarly, if the current increase in house prices were driven by people buying homes to live in, we might expect the homeownership rate to increase along with prices. However, recent evidence shows that house prices and homeownership are diverging.

The graph shows that, in the wake of the financial crisis, house prices declined by over 25 percent, from an index value of around 180 to around 135. The homeownership rate also dropped from a high of over 69 percent to just over 63 percent, its lowest level since 1980. Unlike in the past, the homeownership rate continued to fall even after house prices began to recover.

Several factors could be driving the decoupling of house prices and the homeownership rate. From the housing supply side, there is a trend toward decreased construction of starter and mid-size housing units. Developers have increased the construction of large single-family homes at the expense of other segments in the market. This limited supply, particularly for starter homes, could result in increased prices for those homes and fewer new homeowners.

There are also several factors affecting housing demand. The wave of foreclosures during the recession may have made people more wary about homeownership. Tighter credit conditions may have reduced access to mortgage credit, placing homeownership out of reach for many households. Real estate investors may be buying properties to generate rental income, simultaneously bidding up the prices of homes while also decreasing the supply of homes available to potential homeowners.

All of the above explanations likely contribute somewhat to the divergence of house prices and homeownership. However, any explanation must consider that this trend isn’t just limited to the United States. In recent years, house prices and homeownership have diverged in the United Kingdom, Canada, Germany, Spain, and the euro area.

Homeownership is part of the “American Dream” and a key tool for households to build wealth. In the years since the recession, though, fewer Americans have bought homes and increasing house prices have made homeownership less attainable. What this means for the economy in the long term is unclear.

How this graph was created: From the FRED homepage, select “Browse data by…Category.” Then select “Housing” under “Production & Business Activity.” Find and select the quarterly seasonally adjusted “Homeownership Rate for the United States” series from the results. From the “Edit Graph” menu, select “Add Line” and search for “house price index.” Select the seasonally adjusted “S&P/Case-Shiller U.S. National Home Price Index” series from the results and click “Add data series.” Finally, in the “Format” tab, select “Right” for the y-axis position of Line 2.

Suggested by Daniel Eubanks, Pedro Gete, and Carlos Garriga.

View on FRED, series used in this post: CSUSHPISA, RSAHORUSQ156S

Timing the market for tax purposes Realizing capital gains when tax rates are lower

If you’re a U.S. resident, you’ve probably already filled out your tax forms. (Unless you like procrastinating. In which case, you’ve got one month left.) FRED has plenty of data from the Internal Revenue Service that describe tax filings. Some series count the number of individuals filing and the amounts in various parts of their tax declarations. This graph shows the number of people declaring net capital gains. Clearly, the number decreased significantly when the stock market was doing poorly, but it was far from zero: The troughs were about half of the peaks. The reason is that only realized capital gains are taxed—that is, when a stock is sold. Even during the worst times, many of those selling their stocks were still realizing gains compared with the prices of the stocks when they originally bought them. And timing the sale of stocks may be to a taxpayer’s advantage if it allows him or her to benefit from lower marginal tax rates. Similarly, a prolonged stock market rally doesn’t necessarily translate into immediate additional capital gains…unless people sell their stocks during the rally. But periods of high income for taxpayers may not be the right time to sell stock if it pushes them into higher tax brackets.

How this graph was created: Search for “individual income tax filing” and click on the desired series.

Suggested by Christian Zimmermann.

View on FRED, series used in this post: NCGAGI

From floor plans to roof tiles The stages of putting up new houses for sale

FRED has a lot of data on the real estate market at the regional, national, and international levels; and this graph offers a sample. Here we examine the stage of construction for new houses for sale in the U.S. Indeed, you can buy a house while it’s still in the planning stage, while it’s being constructed, or after it’s completed. Generally, most new houses are put up for sale while construction is still going on. The business cycle can affect this timing, however, and there are a few perspectives to consider.

During a downturn, when houses are more difficult to sell, the stock of completed houses for sale can build up. Also, if prospects for selling a house worsen, a builder may delay construction until a sale is made, which would increase the stock of unstarted houses for sale. On the other hand, a builder may simply abandon the idea of new construction if there are too many finished houses already on the market, decreasing the stock of unstarted houses for sale. This last story seems to apply well to the recent past. In fact, during the previous recession, the number of unstarted houses decreased.

How this graph was created: Search for “new houses for sale by stage” and select the three series. Click on “Add to Graph.” From the “Edit Graph” section, open the “Format” panel to change the graph type to “Bar” with regular stacking.

Suggested by Christian Zimmermann.

View on FRED, series used in this post: NHFSEPCS, NHFSEPNTS, NHFSEPUCS

Of places and patents Tracking changes in the geography of U.S. innovation

The first U.S. patents were granted over 225 years ago, with three granted in 1790 according to the U.S. Patent and Trademark Office. Since then, these exclusive rights for inventions have become much more common and have expanded to cover a greater variety of inventions. In 2015, nearly 300,000 U.S. inventions received patents. The growth in patents has gone hand in hand with growth in innovation, but its geographic distribution has varied over time and continues to change.

The map above shows the number of patents assigned in U.S. counties for the month of October 2016. The counties with more patents (shown by darker colors) tend to be clustered around urban centers such as Los Angeles, Phoenix, Denver, Miami, and Chicago. This distribution could be due to the higher population density in these urban counties—that is, more patents may occur simply because there are more residents. Many of these cities are also home to universities that promote research and development in technology.

Regional patent approvals can rise and fall, and some regions can surpass others. The graph above shows monthly patent approval numbers for a specific geographical region that includes Los Angeles County and nearby Santa Clara County, colloquially known as Silicon Valley. Patents in Silicon Valley began to surpass those in Los Angeles as early as 2001. The graph below (of patents per capita) shows that population growth in Santa Clara County was not the cause of its spike in patent approval. The number of patents per thousand residents was nearly identical in both counties in the early 1980s, but now Silicon Valley’s is 15 times higher than Los Angeles’s.

What explains the rise of one county’s innovation over its neighbor’s? Silicon Valley has been a leader in technological development since the early 20th century thanks to its longstanding location for (i) U.S. Navy research, (ii) Stanford University and its graduates’ jobs in computer production, and (iii) robust venture capital investment. Since about 1995, however, Internet-based firms in Silicon Valley (Google, Apple, Amazon…) have been propelled to the forefront of the tech economy. The types of inventions these companies produce, such as software processes, have been covered by patent law only fairly recently. As these companies gained prominence and profit, the number of patents they were granted increased as well and the trends seen in these graphs came to light.

NOTE: We did notice the steep drop in Santa Clara County’s patent approvals over the past six data points. It is likely due to the time it takes to grant a patent, currently averaging 24 months. Washington, DC, has a similar drop-off in new patent assignments.

How these graphs were created: Go to GeoFRED and click “Build New Map” in the upper right. From the “Tools” menu in the upper left, select “County” for the “Region Type” and type “patent” in the “Data” section’s search bar. Select “new patent assignments by county,” and then select the desired date in the “Date” menu. For the first graph, search FRED for “patents Los Angeles” and select the relevant series. From the “Edit Graph” section, under the “Add Line” tab, search for and select “patents Santa Clara.” Change the frequency of both series to “Quarterly.” To transform the first graph into the second: For the Los Angeles series, use the “Edit Line 1” option in the “Edit Graph” section to search for “population Los Angeles” in the “Customize Data” search bar. Select that series and click “Add.” In the formula bar, write a/b and click “Apply.” Repeat this process with Line 2, searching for “population Santa Clara.”

Suggested by Maria Hyrc, Diego Mendez-Carbajo, and Katrina Stierholz.

View on FRED, series used in this post: CALOSA7POP, CASANT5POP, USPTOISSUED006037, USPTOISSUED006085

Taking the pulse of the economy Connecting the San Francisco Tech Pulse with other economic indicators

The San Francisco Tech Pulse is a measure of the overall health of the American tech sector; it’s calculated using variables such as employment and consumption in the sector and investment in technology. The graph shows the Tech Pulse as well as total U.S. employment, CPI, and GDP indexed to January 2000. These other indicators are common benchmarks of general economic health: Rising GDP, slow changes in CPI, and high employment all indicate a strong economy.

During the 2008 recession, the indicators behaved as we would expect them to during such an economic downturn: employment fell steadily, as did GDP, and CPI spiked and then fell in a spell of high inflation followed by deflation. The tech pulse also plummeted, which makes sense considering it’s the sum of the above indicators in a specific area of the economy. Yet the Pulse began to rise earlier than the general indicators. This early recovery, beginning in April 2009, could indicate that the tech sector was one of the first parts of the economy to gain strength after the recession and assisted in the overall economic recovery.

However, the overall impact of technology shouldn’t be overestimated. During the earlier recession, in 2001, the other indicators remained fairly stable compared with the Tech Pulse, which decreased substantially. This drastic fall could demonstrate the opposite of the pattern we see in 2008: that the technology sector was a major loser in that recession and it was the rest of the economy that helped maintain relative stability.

How this graph was created: Search for and select “San Francisco Tech Pulse.” From the “Edit Graph” panel and the “Add Line” tab, search for and select the other series shown here: “GDP,” “CPI,” and “Employment.” In the “Units” section, select “Index (Scale value to 100 for chosen date),” set the date as January 1, 2000, and click “Apply to all.”

Suggested by Maria Hyrc and Christian Zimmermann.

View on FRED, series used in this post: CPIAUCSL, GDPC1, PAYEMS, SFTPINDM114SFRBSF

Subscribe to the FRED newsletter

Follow us

Twitter logo Google Plus logo Facebook logo YouTube logo LinkedIn logo
Back to Top