Real Estate Will Underperform Inflation for Decades
January 16, 2007
Dan Forshee is an engineer at the Boeing Company and lives in Mukilteo, Washington
The census.gov web site clearly indicates that while general prices throughout the economy tripled between 1915 and 1965, from 30.4 to 94.5, rents (the most reliable indicator in the absence of price data) only doubled, from 49.6 to 94.9! This means that if you had bought in 1915 and sold in 1965, you would have lost 33% in purchasing power. Obviously, actual results varied with about half of actual returns even lower and half higher. Tables E135-E166 of the following link provides the source for this information:
An interesting side note is that over the period of 1915 to 1965, housing costs appreciated in price by an average of only 1.32% per year, or negative 0.8% a year after inflation.
The Census data contradict the absurd fantasy often popularized in the media and culture that house prices are a wonderful investment and always provide consistent inflation-beating returns. In fact, in the last century they often underperformed inflation for many decades at a time. There are many legitimate reasons to buy real estate: stability of schools and neighbors, freedom from capricious landlords, certain tax advantages, and mortgage payments that save money every month by being less than the rental of the same unit. If the last 100 years are any indication of the future, expectations of future price appreciation are not a legitimate reason to buy real estate. A large percentage of the current buyers in the real estate boom of the 2000s are buying based on future price appreciation. Unfortunately, this means that the real estate boom of the 2000s is largely based on a myth. As real estate prices return to levels justified by the legitimate reasons, prices will fall in inflation-adjusted terms.
Combining the census data for 1914-1970 with data sets from Freddie Mac for the years 1970-2006, the following figure is obtained:
The chart indicates that housing prices in the mid-1990s were about 25% below prices in the mid-1920s.
The old adage of our grandparents that “housing is a depreciating asset” was true for a very long time.
The 1914 peak in inflation-adjusted terms was only exceeded for the first time in 2005.
Even if you don’t believe price statistics, evidence can be gathered with the just two eyes for evidence of the real estate fall in the years following the peak in 1914-1933. In free-market capitalist societies, as land prices get higher, buildings get taller. Higher prices of real estate make it profitable to build tall buildings because the higher construction costs are offset by lower land costs. Most major cities in the United States had tall buildings built between 1914 and 1933 during the real estate boom of that time frame. After the tallest building was built, it typically took about 41 years for the real estate prices to return to levels that would justify buildings of similar height. Here is a data set of example cities:
Name of City
Tallest Building Built during previous peak
Year in which the record was broken.
Number of years to break the previous peak
1914 (Smith Tower)
Not yet broken
* The space needle is taller and was built in 1962. However, it was built as a show piece, not for economic reasons and therefore is not listed. Even if it is listed, it does not appreciably change the results.
Summary by Region:
West Coast Average: 45 years
Midwest Average: 52 years
East Coast Average: 42 years
South Average: 34 years
Foreign Country Average: 32 years
Average of the five regions: 41 years
The tall building indicator indicates that it takes about 41 years for a peak real estate market to bottom out and then rise to exceed the previous peak.
The tall building indicator UNDERSTATES the time to return to a previous peak because of technological improvement in building construction. Over many decades, construction techniques and equipment improve, lowering construction costs and thereby making it more feasible to build a tall building on a lower cost of land. Assuming the tall building indicator is off by between 50% and 200% due to technology improvements, the 41 years indicated is probably closer to 60 to 120 years for a peak real estate market to exceed its prior peak. The Census and Freddie Mac data on the page 1 backs this up. The peak real estate market of 1914 was not exceeded in inflation-adjusted terms until 2005, fully 91 years later, which is in the middle of the expected range.