A Chicken and an Egg: The relationship between unemployment and office vacancy
The proverbial riddle of “the chicken and the egg” is an apt metaphor to describe the relationship between unemployment and office vacancy in a market. Which produces which? Real estate investors are searching for any indicators of how bad and how long the recession will be, and this relationship can serve as a gauge of a given market’s direction.
Unemployment and office vacancy rates tend to correlate, though not always in obvious or precise ways. Here is a standard case that demonstrates the relationship as we most commonly understand and see it.
Nortel Networks filed for bankruptcy on January 14, 2009. The company has steadily been shedding market share, jobs, and office space since the technology bubble burst in 2000. In 2007, the company shed 2,900 jobs and vacated 1 million square feet. Nortel shed 1,200 jobs and 500,000 more square feet of office space the following year. After its recent Chapter 11 filing, the company let another 3,200 employees go and may vacate up to 2 million more square feet of space. With respect to cause and effect, it appears that the company shed jobs prior to its downsizing in office space. This scenario appears to hold true in the larger national picture, at least in this current recession.
Unemployment and office vacancy have followed a generally parallel trend over the past 20 years, encompassing three recessions, including the current one. The following chart displays the national unemployment rate obtained from the U.S. Bureau of Labor Statistics and corresponding office vacancy rates from Grubb & Ellis’ Office Market Trends.
The chart above shows the drop and subsequent rise in office vacancy preceded that of unemployment in the early 1990s through 2001. However, the most recent fall and rise, taking place in 2003 and 2007, respectively, show the opposite to be true, with the unemployment rate shifting prior to office vacancy. This suggests that neither is a necessary predictor of the other. The following scatter chart provides another perspective on the data.
These charts illustrate an unmistakable relationship between unemployment and office vacancy. The trend line in the scatter chart shows a curvilinear relationship with a correlation (R²) of 0.5768. In simpler terms, this suggests that a trend is visible, but not strong enough to predict whether unemployment affects office vacancy or vice versa.
The correlation is also visible in individual markets across the country. In markets with a large volume of office space, such as the Washington, D.C. area, the correlation between unemployment and office vacancy is quite visible. The D.C. market is dominated by government-related businesses, equating to more office space per capita than, say, Detroit, which is largely dependent on manufacturing jobs that operate out of factories and warehouses more than office buildings. As such, the correlation between unemployment and office vacancy rates in D.C. is strong, as seen below:
While there are minor fluctuations, a decline from 2004 and a subsequent rise in 2008 are visible. The variances noted in 2006 through early 2007 are attributable to major swings in office space being delivered to the market. The minimal amount of new office space delivered from 2003 through 2005 was due to the high vacancy rates in the market. Improved office occupancy spurred new construction, seen in the explosion of new inventory brought to the market in late 2006. Since that time, office vacancy rates have closely mirrored the rate of unemployment.
Every market contains individual characteristics that set it apart from others nationwide, and such is the case for several southern cities. While much of the nation has seen several consecutive quarters of increasing unemployment and office vacancy, fourth-quarter 2008 office vacancy rates in Oklahoma City, Houston, and the Dallas-Fort Worth area have remained below those of two years prior. This is largely due to increased leasing activity by the energy industry when oil and gas prices were peaking.
You can see that the peaks, valleys, and plateaus of unemployment and office vacancy in Oklahoma City do not always correspond. The Oklahoma City market is quite diverse, with an employment base spread over energy, trade, government, healthcare, education, and manufacturing. This is not to say that unemployment does not affect these industries as well; however, the market’s eggs are not all in one basket, so to speak, and Oklahoma City has not experienced the dramatic upticks in unemployment and office vacancy seen nationally.
Similar to the Oklahoma City market, the Dallas-Fort Worth[1] market has not experienced the dramatic escalation in unemployment and office vacancy rates seen nationally. The unemployment rate in the Dallas area remains similar to that of three years prior. Office vacancy rates saw a discernible decline through late 2007. Early 2008 saw a significant amount of new office inventory reach the market but was quickly leased up by energy firms moving into the area when oil and gas prices were peaking. Again, we see that the more dynamic a market becomes, the less correlation is evident.
The question remains whether businesses are leasing less space because the size of their staff is shrinking, or if they’re rather reducing office space in advance of downsizing their workforce. The current economic picture and an examination of past trends shows unemployment to be a good gauge of what to expect of vacancy rates for office markets around the nation. This being the case, the current spike in unemployment forebodes a yet-to-be seen spike in office vacancy rates. But there are instances when unemployment seems not the cause but the consequence of higher office vacancy levels. This complicates the scenario, but research and analysis of the precise dynamics of a given market will help you hatch an appropriate strategy. U.S. Commercial Appraisals can help, so please contact me (rjander@ushcapp.com or 214-608-1361) before your next venture into the office market. [1] While Dallas’ office vacancy rates were relatively high in fourth quarter 2008, they have not risen significantly since the recession began.