How Big is the Shadow, Really?

by CoyDavidson on January 21, 2011

There has been quite a bit of commentary over the last year and very recently regarding the impact shadow office space will have on the market, the leased but empty office space no longer being utilized by companies as a result of the economic downturn. Some organizations have even taken a stab at estimating how much of this space exists across the country.

Commercial real estate is a data intensive discipline and after 20 years in this industry, I have seen a lot of economic statistics as property values, sales prices, cap rates, rental rates, vacancy rates and absorption statistics are widely tracked and analyzed by commercial real companies and data service providers. I recognize that data can be manipulated and interpreted in many ways and the, “garbage in | garbage out” principle can apply, but in most cases I don’t question the accuracy of widely reported data.

My first required statistical analysis class in college I ended up having to take twice, and the second time I was really paying attention. So, since shadow vacancy is data that has not been historically tracked in commercial real estate, the first time I saw an actual shadow vacancy rate reported it raised a red flag. I had to question its accuracy without understanding the methodology used to arrive at the estimated figure. I actually asked the company who reported this data more than once what methodology they utilized only to get no response, but I guess we are just supposed to take their word for it.

So how could you arrive at such a figure?

  1. You could survey every office tenant in every major market, and well we know this didn’t happen.
  2. You could survey a cross section of office tenants around the country that would result in somewhat representative sample for formal statistical analysis.
  3. You could develop some formula based on tenancy profiles, occupancy rates and reported employment statistics.

Surveying a large cross section of tenants large, medium and small from a wide variety of markets and industries around the country to produce a relevant sample for statistical analysis would be the proper approach. Consider me highly skeptical this approach was taken either and even if it was, a very small percentage of companies actually track this data in detail. I would then have to question the quality of the data.

I can only assume that some type of formula was used to arrive at this estimate. Looking at the reported numbers the only data that I could use for comparison purposes was to take a look at my clients exposure in terms of shadow vacancy. This sample is a wide cross section in terms of company size and industry but primarily concentrated in the Houston market area and skewed by the size of one large user in a particular industry. I actually have only one client with notable shadow vacancy and while we had discussions regarding disposition of the space, they opted to inventory the space as they expected to need in a relatively short time frame. So in comparison, as a percentage my client portfolio of aggregate shadow vacancy is vastly different than the estimates reported.

Shadow Vacancy Exists and its going to Impact the Market

Yes, shadow office space vacancy does exist and there is no doubt as the economy recovers and office using employment growth accelerates that companies with shadow office space are going to have a notable impact on the velocity of net office space absorption and the timing of a shift in market fundamentals toward the Landlord side of the equation. I don’t question the concept, I am just skeptical there is enough reliable data to accurately predict the volume of shadow office space, who has it, and who has already disposed of it.  Somebody’s estimate or prediction of its impact on the market is going to turn out relatively accurate, but then again I might pick the Super Bowl winner as well if I take a stab at it. Last week I would have picked the Patriots, always subject to change!

  • Pingback: Tweets that mention How Big is the Shadow, Really? — The Tenant Advisor --

  • Coy,

    I think you underestimate the effect of shadow space, but I agree with your skepticism of the data collection and somewhat disagree with the criticism of the metric.

    Even a broad survey of the market may illicit skewed results, as companies hesitate to admit how much space they actually under-utilize.

    However, our clients in professional services, i.e. large banks and law firms high in the Fortune 100, and many other companies in the market have considerable shadow space. There are still significant subleases hitting the market, but that doesn’t tell the whole story. One bank in particular put a very small sublease on the market but marginal in terms of how much space they actually under-utilized. When the time came for a certain lease expiration, they consolidated and vacated a building for 375,000 RSF and never added positive absorption to counteract it.

    As the hiring in professional services firms continues to stall, more shadow space will become subleased space or vacant space. Sometimes, as you know, companies don’t want to put disposition space “on the books” due to accounting, PR, etc, when they don’t absolutely have to for their bottom line.

    I get the idea of the post, but metrics that highlight historically unmeasured data may not be completely inane, but they also shouldn’t be reticent to explain how they achieved the data.

    Thanks for the post,


  • CoyDavidson

    Justin, perhaps the tone of my post gave the wrong impression and I should have emphasized more in the last paragraph that I was not discounting the significance of shadow office space or the impact it will have on the office market. My point is that I am skeptical that there is solid data and significant variances exist among markets, industries, small tenants / large tenants and unless they explain their methodology for estimating the volume, the source of their data and what cities were included in their data collection and tenant profiles, then I consider their estimates nothing more than and educated guess with a notable margin for error. We know many tenants have more space than they need. Is available sublease space shadow space?

    For Example purposes:
    Let’s say there is 100 million square feet of Law firm space leased nationwide and 15 million SF is actively on the sublease market. If you were going to try to determine the shadow office vacancy among law firms without surveying the tenants. How would you go about it? Wouldn’t you have to look at broad employment figures for law firms at a given point in time and compare to another period say year over year and take into account the change in the amount of leased space, available sublease space then apply some industry standard ratio of square feet per employee to estimate the volume of shadow vacancy.

    So I doubt anyone devoted the time or resources to this analysis industry by industry and market by market. So you would have to use a broader measure looking at markets city by city taking into account leased space between two periods of time, the change in available sublease space and employment figures, apply a broader ratio of square feet per employee say 250 SF, as well as hope that your office using employment figures are representative of the tenancy.

    Also, when accounting for the change in employment / head count do you have to assume there was some shadow vacancy that existed in your baseline data and what is that number? Unused office space pre-recession was considered expansion space, today its shadow space. Has there been a shift in the percentage of space companies are willing to inventory for expansion and that is a nuance that varies by industry and tenant size.

    If a company has 100,000 SF is using 60,000 SF, has 20,000 SF on the sublease market and is holding 20,000 SF for an expected expansion, then I assume that by most people’s view that 20% of their space is shadow space, as the 20,000 SF of sublease space is reflected in traditional vacancy rates. When you look at broader employment metrics how do you proportion the job losses among sublease space and shadow vacancy?

    Given the margin for error I would be hesitant to state that shadow vacancy is 4% in one market in 7% another etc and feel good about my numbers. I am not making the specific argument that all these numbers are on the low side, but I think when you see these figures you have to make some distinction as to how much of it is marketing versus solid analysis.

    I do think shadow vacancy is a significant factor and will have an impact on market fundamentals. We know there is some volume of shadow vacancy to be absorbed before we will see any economic recovery/job growth make its presence in occupancy rates. The question is can anyone really predict with relative accuracy what the percentage of leased space is shadow vacancy and if you step out and do so…I am just asking how you came up with the number?

    Thanks for the comment…always value your opinion!

Previous post:

Next post: