Trimming Data Centers: A New Initiative Aims to Trim and Optimize Federal Computer Hives

This is the age of Big Data. Given the flood of information, it is also the age of the big data center. Most of us don’t think much about these out-of-sight, out-of-mind hives of humming computers, but they are central to the modern wired economy, and they eat up energy in proportion to their importance—an annual energy consumption equivalent to the output of 35 large coal-fired power plants.

Data centers are everywhere, some of them massive (think Google, Amazon and other giants of and in the cloud), some of them small-scale operations that service retail operations and home businesses. Many do business with the federal government, and the government operates many more—which brings us to the nut of our story.

By the federal government’s estimation, based on a policy memorandum released on August 1 by Federal CIO Tony Scott, there are entirely too many of these data centers operating entirely too inefficiently, many well under 50 percent of capacity. As long ago as 2010, the Obama administration was considering ways in which to reduce their number, both to make the federal footprint more manageable and to reduce energy consumption, to say nothing of easing the way toward comprehensive security plans made possible only by knowing just how many of the things were out there—and where.

To this end, in 2014, President Obama signed into law the Federal Information Technology Acquisition Reform Act (FITARA), which requires federal agencies to undertake a comprehensive data center inventory, develop performance metrics, and provide an annual report detailing costs and savings measures. As of August 1, a new policy measure, the Data Center Optimization Initiative (DCOI), supplants a similar directive from 2010, requiring federal agencies to “develop and report on data center strategies to consolidate inefficient infrastructure, optimize existing facilities, improve security posture, achieve cost savings, and transition to more efficient infrastructure.”

As of April 2017, by the terms of the DCOI, federal agencies will not be able to budget for the construction of new data centers or “significantly” increase the size of existing ones. Instead, agencies, whether needing more space or not, will be required to analyze data needs and consider alternatives such as pooling resources with other agencies, using cloud services, or working with a third party, the last two of which would seem to offer opportunities to data entrepreneurs in the private sector.

The General Services Administration (GSA) and Office of Management and Budget (OMB) will develop standards for interagency cooperation, while GSA will develop a 
shared services marketplace, to include outside service providers. All these efforts are meant to work toward a set of larger goals, including optimizing existing data centers, closing about half of them, and arriving at inarguably significant savings, paring the data budget by at least 25 percent of the 2016 figure by the end of FY 2018 and taking costs down by half, $2.7 billion, of the $5.4 billion expenditure in FY 2014.

All of this is a tall order, reining in agencies across the government in a relatively short time, certainly shorter than most reform initiatives. To top it off, the effort must be transparent: Beginning this year, OMB will publish progress reports detailing planned and realized data center closings, cost savings, and other metrics. Updated information on the DCOI and FITARA will be made available here at the OMB website “Management and Oversight of Federal Information Technology,” and we will report on developments as they become known.

Who Really Has Leasing Authority?

What federal agencies are authorized to lease property from the private sector? If you had asked that question of any federal official before last month, you might have drawn no more than an educated guess, if not an outright blank.

Certainly there is the General Services Administration (GSA), which acts as a kind of super-landlord and property-management authority for the federal government overall. Currently, GSA leases about 192 million square feet of property from the private sector, comprising nearly 7,000 office and warehouse spaces, a vast set of holdings. But many other federal agencies and entities have the ability and authority to rent space independently of the GSA—the US Postal Service, for instance, and the Federal Aviation Administration (FAA)—whether by virtue of their enabling legislation or by some appropriations act giving them the power to do so.

Until July, no one could say for sure just how many agencies had such leasing authorization. The reason for that is that not all federal agencies, as of now, are required to submit data to the Federal Real Property Council (FRPC), which falls under the aegis of the Office of Management and Budget (OMB). To multiply acronyms, as part of this effort, GSA has been charged with creating a comprehensive inventory of federal properties, an effort known as the Federal Real Property Profile (FRPP). An agency not governed by the FRPC—and there are many loopholes—is not required to file data with the FRPP.

All this came to the fore in a recent report prepared by the US Government Accountability Office (GAO) at the direction of the House of Representatives. In hearings before the Subcommittee on Economic Development, Public Buildings, and Emergency Management of the Committee on Transportation and Infrastructure early in July, the GAO report offers testimony that much work remains to be done to “enhance information and coordination,” as the report’s title has it, “among federal entities with leasing authority.”

What the report turns up may be surprising even to those inured to problems of coordination among federal entities. The GAO surveyed 103 of them, and 52, fully half, had independent authority to lease office and warehouse space. Nearly half of them were not bound by the FRPC and so had no reason to provide information to the FRPP, even though the collective holdings of these independent authorities were significant: 944 offices and 164 warehouses, or about 16.6 million square feet of rentable space. Among these agencies were the FAA, the National Aeronautics and Space Administration (NASA), the Coast Guard, the National Oceanic and Atmospheric Administration (NOAA), and the US Patent and Trademark Office.  These specific agencies normally lease space through the GSA…except when they do it through their own leasing authorities.

The GAO report, built on data covering the period between 2001 and 2015, makes a number of interesting observations that need to be teased out further. Among them is the fact that, though the possibilities for overcharging would seem to be rife, in general the spaces leased outside GSA purview are fairly priced and in fully 38 percent of cases cost less than GSA leases, while another 30 percent were about the same. Independently leased spaces also tend to be roomier than their GSA-leased counterparts; GSA recommends that federal agencies allow 150 square feet per employee (though this metric varies widely across their agency customers), but most of the independent spaces are at least twice as big.

“The FRPP’s incomplete data set reduces its effectiveness as an oversight and accountability mechanism for entities with independent leasing authority,” the GAO report notes. More actions are surely in the offing, but at a minimum, the report recommends that non-FRPC entities be required to provide information to FRPP and that federal agencies figure out tighter standards to govern leasing.

The GSA Market Continues to Shrink

GSA Leased Inventory Trend

We have written a lot on this blog about how the GSA-leased property market is getting smaller, a trend we expect to continue easily for the next few years. Yes, one day this trend may reverse itself but there are really only a few ways that might happen, and none of them seem likely right now (and some we would not wish for). Ultimately, the investment community will need to adjust to the fact that a half-century of consistent inventory growth began to fall back in 2012 and austerity is the “new normal”.

The GSA lease inventory reached its all-time peak size in December 2012. In that month, GSA leased 198.9 million rentable square feet (RSF) across the United States and its territories. Since then, the leased space inventory has declined by 6.7 million RSF to 192.2 million RSF in March of this year (the most recent data available). It’s really not a substantial decline–just 3.4%–but it has only just begun.

GSA is driving this reduction largely through improvements in space utilization. Our team of federal leasing specialists sees this every day in virtually every transaction we work on. In fact, there is a lot of street-level evidence that the pace of downsizing will accelerate based upon lease transactions that are in procurement or recently signed.

None of this is a surprise. The early indicators date back to 2011, well before the market peak, when fiscal conservatives took control of the House of Representatives and began exerting their control of prospectus leases to force cost reductions. Prospectus-level leases, while relatively few in number (a little more than 200 at any given time), comprise roughly 1/3 of the leased square footage in GSA’s inventory. So, Congress alone has the power to make meaningful cuts to the leased space inventory. Yet, OMB removed any doubt about the direction the market was headed when it issued formal guidance to all executive agencies to freeze the size of their civilian real property inventories and, ultimately, reduce costs.

The challenge for the feds is that their workforce has not changed much since the early 1970s. Though there is a lot of pressure from Congress to reduce the size of the federal workforce, it is unlikely that the agencies will accomplish much of a decrease in headcount. The answer instead has been to transform the workplace to embrace mobile work (i.e. work from home) and reduce the square footage allocated per person in new office designs. Often, these efforts to reduce have also resulted in consolidation into fewer locations.

On the consolidation front, Washington, DC has been especially impacted. There is almost 100 MSF of leased and owned property in the National Capital Region, providing lots of opportunities to consolidate offices. There are many instances of this happening in recently completed and in-progress transactions that are not yet reflected in the inventory reduction stats. One such example is last year’s 839,000 RSF Department of Justice lease of Constitution Square 3 and 4. By the time DOJ fully occupies these new buildings in 2018 they will have vacated four downtown buildings, yielding a net space reduction of more than 200,000 RSF.

The DOJ deal also underscores the fact that not only is GSA’s leased square footage getting smaller but also the number of leases. The count of GSA leases has also declined steadily since year-end 2012. There were 8,872 GSA leases nationally in December 2012 and by March 2016 that number was reduced to 8,279 leases–a decline of 6.7%. The decline in the number of leases is happening at twice the rate of square footage reduction, which suggests that consolidation is a big factor in shrinking the government’s footprint. As anecdotal evidence, our team is working on a downtown Washington, DC building where we have agreed to terminate three GSA leases so that the tenant agencies can be consolidated into other existing leased locations (a fourth tenant consolidated last year upon the natural expiration of its lease).

There is a silver lining to all of this. Though the leased inventory is clearly getting smaller, lease terms are likely to grow longer. This is because workplace transformations require capital, which necessitates long-term lease contracts. Having engaged heavily in short-term leasing over the past several years creating a pile-up of near-term lease expirations, we can expect GSA to more often execute long-term contracts going forward. Don’t expect a watershed change–GSA lacks the leasing capacity to tackle its entire backlog at once. Yet, in the long run we anticipate a shift towards long-term leasing and, ultimately, a stabler federal leasing environment.

Sales Market Insights – Year-End 2015

Attached is our review of last year’s federally leased property sales activity. Highlights from this report include:

• Consolidation of leases, multi-agency occupancy, and relocation of leases to GSA-owned buildings continues to take place;

• From YE 2012 to YE 2015, GSA has trimmed its lease inventory by 6%, but total square footage is only 3% lower, with firm lease terms having shortened by 11% during that time;

• GSA lease solicitation opportunities have ticked up, but solicitations for new lease-construct projects continue to be well below historical averages;

• As a result, cap rates on longer-lease term assets continue to trend slightly downward, along with cap rates on medium- and shorter-lease term assets as investors seek yield further down the lease term spectrum.

To read the full report, click here.

Spotlight: Farm Service Agency

fsaIf the Farm Service Agency (FSA) sounds as if it might be some forgotten holdover of the alphabet-soup New Deal of nine decades ago, there is good reason. Although it was formally instituted only in 1996, during the last months of Bill Clinton’s first term as president, it harkens back to several programs of the US Department of Agriculture, within which the FSA is housed.

Among those initiatives of old were the Farm Security Administration, under whose auspices we have those haunting photographs of the rural Depression and Dust Bowl years, and the Agricultural Adjustment Administration, which paid subsidies to farmers to keep crops and livestock off the market, driving prices upward to yield a profit for other farmers whose produce did go up for sale. The modern FSA borrows a bit from each of them, as well as other New Deal efforts, in order to accomplish its chief mission: to keep farmers on their farms by offering low-cost loans, grants, and other incentives. A dairy producer in North Dakota might approach FSA for a loan after a loss suffered in a fierce winter snowstorm. So might an avocado farmer in California in danger of losing income to drought–or even a young urbanite deciding to go out on her own and grow, say, boutique chickens and lavender in Massachusetts.

Indeed, the FSA is responsible for administering loan and grant programs that account for the greater portion of its $15.8 billion operating budget in 2016. That’s a substantial amount of money, and to administer it the agency works from a far-flung network of more than 2,300 field offices in every state, including some locations that might seem unlikely: there’s an FSA office in the Bronx, for instance, and another in suburban Detroit, and even one on the idyllic tropical island of Kauai, in the westernmost reaches of Hawaii. That the FSA operates there as well as out in the wide-open spaces of Oklahoma and South Dakota is an indicator of not only the prevalence of agriculture in the national economy, of course, but also the reach of the government into every corner of that enterprise.

Because of this reach, to say nothing of all that money, the FSA is under close scrutiny on all sides. Not long ago, liberal health advocates chastised the agency for giving grants to tobacco farmers in the South, their objections based on moral rather than strictly economic arguments. By the same token, conservatives were outraged when it was revealed some years back that Hollywood actors, who ran what seemed to be hobby farms, were pocketing federal subsidies. Like many agencies throughout the government, FSA has been slow to modernize its IT system to streamline access to farmers, and it shared some of the responsibility for the mismanaged MIDAS (Modernize and Innovate the Delivery of Agricultural Systems) program, shut down by USDA Secretary Tom Vilsack last year after running $140 million over budget with little to show for it.

Headquartered in Washington, DC, with a large data center in Kansas City, the FSA has suffered substantial cuts in budget has staffing over the last 20 years. In 1997, the agency had more than 17,000 employees; today that number is about 12,000, and, a recent agency report warns, by next year more than half of that workforce will be eligible to retire. “New human capital initiatives designed to seek, strengthen, and retain high quality personnel are imperative to FSA’s ongoing ability to achieve its mission,” the report concludes—which, translated, means there will be job opportunities for those people with the proper training, including, it would seem, knowledge of IT operations.

Though farmers across the country are well aware of its activities, FSA is not widely known outside the agricultural sphere. Yet its sizable budget and sizable footprint make it worth knowing about—and there are plenty of opportunities to do business with FSA in every state of the Union.

Data Center Consolidation: Getting There, But Not There Yet

This figure from GAO's report shows that roughly half of all federal data center a slated for closure and that, in fact, roughly 1/3 of federal data centers have already been closed. However, it's likely these are the low-hanging fruit.

This figure from GAO’s report shows that roughly half of all federal data center a slated for closure and that, in fact, almost 1/3 of federal data centers have already been closed. So, by this metric, agencies are making good progress.

In 2010, the Office of Management and Budget (OMB), the fiscal watchdog within the federal government, issued an order requiring that federal agencies consolidate data centers in order to save money and reduce redundancies. Twenty-four agencies, including elements of the Departments of Homeland Security, Treasury, and Defense, were tasked with identifying data centers under their purview, then to determine which could be closed and at what cost savings. Six years later, these agencies have collectively identified 10,584 data centers, of which 3,125, or about 30 percent, were reported closed as of the end of FY 2015. Another 2,078 were slated to be closed by the end of 2019, for a total of 5,203, or just shy of half the total number.

That would seem to be a significant savings, but a report issued last month by the US Government Accountability Office concludes that of the nine metric targets set by OMB, such as power usage, server utilization, and “virtualization density,” only one was met by half of the agencies, with the other eight falling short of even that halfway mark. Furthermore, the report continues, the agencies as a whole have underreported potential savings; the figure now stands at $8.2 billion, but “planned savings may be higher because 10 agencies that reported planned closures from fiscal years 2016 through 2018 have not fully developed their cost savings goals for these fiscal years.”

The GAO report indicates numerous pathways to improvement, noting, for example, that across the spectrum, only about 5 percent of federal server space is actually being used. OMB has recommended but not yet put in place a uniform metric for server utilization, which would seem to be a first step toward improving the efficient use of computer resources. Perhaps surprisingly, in response, the Department of Defense has stated that it “is moving toward commercial cloud hosting services to enable the migration of workloads to more efficient environments, thus positively affecting the virtualization and density metrics.” How hardened those cloud hosting services are—safe from hacking, that is—is, we imagine, a matter under review inside the Pentagon.

That 5 percent figure, incidentally, comes from 2009, indicating another source of frustration noted by the GAO evaluators—namely, that almost all the agencies in question have been slow to provide relevant, up-to-date data. One issue, GAO observes, is that some of the agencies in question are “decentralized,” so that cumulative data are hard to come by. Another, an objection raised particularly by the Department of Transportation, is that many data centers are housed in leased space without “dedicated metering” or uniform measures of performance and savings. One implication is that owners of leased property, if not the agencies themselves, may be required to put such measures in place.

The report recommends pointedly that all of the agencies take steps to improve their reporting and forecasting. This is especially pressing, the report adds, because of federal FITARA requirements that agencies calculate annual cost savings as part of an overall strategy to improve efficiencies and reduce costs.

The report concludes, somewhat circularly, that “until agencies take action to improve progress against OMB’s metrics, including addressing any challenges identified, they could be hindered in making progress against OMB’s optimization targets.” Yet, as noted above, many agencies have struggled to develop planned data center cost savings and avoidance targets. In other words, if the agencies don’t know what they’re supposed to be saving, they won’t be able to meet OMB’s goals. Or, put more simply: “If it can’t be measured, it can’t be managed.”

A GAO Report Calls for Heightened Building Security



Last August 21, federal workers in the heart of Lower Manhattan saw a terrible situation: A disturbed man with a gun entered a building in which passports were issued and immigration cases decided. He killed a security guard, then turned the gun on himself, leaving witnesses and investigators to wonder what had driven him to commit such a horrendous act.

For many and obvious reasons, statistics on attacks and other threats that occur within federal facilities are hard to come by. The Federal Protective Service, the division of the Department of Homeland Security that is directly responsible for facility security, calculates that it responds to 534,000 “calls for service” each year, but that is a broad category that runs the range from armed intruder to minor disturbance.

DHS may provide an administrative home for FPS, but there is another fit that does not come so easily: Those federal facilities are managed by the General Service Administration (GSA), and although GSA staff in theory oversee or at least monitor what happens within them, they have no meaningful oversight over the security that FPS provides—and vice versa.
 So it is, a recent report by the Government Accountability Office (GAO) concludes, that although GSA and FPS have formulated an overall strategy for facility security, neither agency has actually signed off on it. Indeed, after spending untold hours on that strategy, the report notes that both agencies requested that the accord be shelved until “after they address other priorities,” whatever they might be and whenever that might occur.

FPS, as the agency website notes, is divided into 11 regions nationwide and protects more than 9,000 federal facilities, with more than 200 field offices and some 1,300 field agents and officers, as well as more than 13,000 contracted security guards. The agency was once operationally governed by a memorandum of agreement specifying its responsibilities with respect to GSA, but the MOA, drafted in 2006, did not accurately describe on-the-ground realities with respect to policy and organizational changes. The two agencies entered into discussions to revise the MOA in August 2015, but because this has not been effected, GAO notes, officials did not know which procedures to follow—a situation that exposes federal facilities and employees to increased risk.

These inefficiencies are probably to be expected, given that by mandate both GSA and FPS are in charge of facility security; both agencies have distinct cultures, and it is instructive that the two do not even agree on how many facilities there are under their shared aegis. (GAO gives the number as 8,900.) But this division is recent; until 2003, the report notes, FPS was housed within GSA, transferred to DHS only after the creation of the new department. A flowchart of post-move responsibilities shows overlap and redundancy; both agencies, for example, can issue security credentials allowing visitors and contractors access to sites. But, the report notes, given existing “previously found problems with the quality of data exchanged between GSA and FPS on facilities and their locations,” these credentials are not necessarily shared or even known to the other agency.

In a pointed case study, GAO identifies a GSA facility that was built to its specifications at a cost of some $75 million, comprising about 180,000 rentable square feet intended to house law enforcement agencies. However, the building was built with energy-efficient systems in place that barred the construction to FPS specifications of armories and holding cells. Because no communication had taken place between the two agencies beforehand, despite a requirement to that effect in the 2006 MOA, GSA wound up with a new facility that could not be used for its intended purpose.

Among other GAO key recommendations are that a schedule be set for a new MOA and a joint security strategy that will clarify the division of tasks and responsibilities. The report concludes that enhanced communication between and collaboration among the agencies must be made priorities: “The lack of collaboration in communicating compatible policies and procedures,” the GAO notes, “makes it difficult for the agencies to effectively implement their security mission and can negatively affect day-to-day operations.”

FBI Field Office Sales Revisited

In November 2013, I published a brief analysis looking at the cap rate trend for FBI field office sales tracked from the end of the Global Financial Collapse. FBI field offices are a good subject for this type of analysis because they are mostly build-to-suits, generally homogenous and there are enough sales of long-term (15+ year) leased assets to construct a reliable trend.

When we looked at the trend back in 2013 the most recent sale traded at a 6.8% cap rate–down considerably from an 8.7% cap rate trade at the beginning of 2009. The lowest cap rate reported for any transaction during that 5-year period was the 6.4%. At the time “low 6” cap rates felt like a pretty tight yield but based on where the trend has gone since, maybe not. The most recent field office trade was recorded at a cap rate just less than 6%–and that building had a second-generation lease.

Draw a trend line through the data and we can see that the vector has not changed much from our analysis more than two years ago. Cap rates have continued to decline about 46 bps each year.

The Big Question, of course, is will this trend continue? On the one hand, projected increases in the federal funds rate will ultimately inflate Treasuries, which will squeeze the delta between that financial instrument and cap rates. At some point this will dull interest in federal properties. On the other hand, there is a lot of capital chasing long-term federally leased assets while the number of those assets is dwindling. Among those investors seeking safe harbor real estate investments, price competition will intensify.

I won’t tackle the Big Question directly in this article but will note that sub-5% cap rates will be achieved only for long-term leased assets where at least one of these conditions is also present: 1) The real estate itself warrants more aggressive pricing (this is why many properties in big-city downtown markets like Washington, DC routinely trade in the low 5% cap range); 2) The government’s rent is at or below market rent; 3) The rent includes bumps (which is rare); 4) The mission criticality of the facility ensures the tenant’s “stickiness”, and/or; 5) The lease contains modifications (like net of electric clauses) that condition common risk factors associated with GSA leases.

Where the Federal Leases Are

The U.S. government leases space all across the nation–including in some pretty remote locations. Yet, the bulk of federal government tenancy can be found in just 20 markets. The map above shows the 20 largest markets for U.S. Government leases as measured by Core Based Statistical Area (CBSA)*.

We track nearly 7,000 properties that are leased in whole or in part by the federal government (this includes all of the buildings leased by GSA as well as many other leased by agencies using delegated or statutory leasing authority). Together, these buildings measure 213.4 million rentable square feet. 65% of this space (138.3 million rentable square feet) is located in just 20 market areas.  Chances are that if you are an investor of federal properties, you own something here:

* CBSA is the new term for a Metropolitan Statistical Area (MSA). For more info read this.

Renewal Rents Usually Increase


There are few investors of GSA-leased properties who have not engaged in hand-wringing over renewal rents, even when buying properties years in advance of the lease expiration date. Normally when presented with the question of whether the future rent is likely to increase, my response is to simply look at market rents on the whole. If they are rising then there is a good chance your renewal rent will too. One might also observe that many GSA leases are structured with levelized (non-escalating) base rents. So, if market rents increase even modestly through the term of the GSA lease, the renewal rate is bound to be higher than the previous rate. Yet, the normal laws of logic and market physics don’t always apply to GSA leasing–primarily because it is usually a “Lowest Price Technically Acceptable” procurement process that is heavily influenced by competition, delineated search area, changing mission needs of the agency, emerging legislative mandates and so forth.

It is impossible to capture all of this nuance in a single analysis but we thought it would be instructive to simply look at the renewals and see what happened. We analyzed every “single tenant” GSA property in the United States that was at least 85% occupied with a GSA lease expiring between 2005 and 2015. We also decided not to look at anything smaller than 5,000 RSF of leased space. Finally, we did not evaluate lease extensions–only renewal leases. These filters yielded 271 properties that we could look at to determine how the rental rate changed from the last month of the expiring previous lease to the first month of the new renewal lease.

The results are shown in the graph above. Every circle represents a property plotted such that the final rent of the expired previous lease is shown on the horizontal axis and the new renewal rent is shown on the vertical axis. The 45 degree trend line illustrates where the renewal rent and previous rent are exactly equal. Any properties above that line (shown in green) experienced higher renewal rents. Properties below the trend line (shown in red) experienced lower renewal rents. The red and green is shaded to reflect the percentage increase or decrease from the previous rent.

What does this tell us?

  • Renewal rents tend to increase. Of the 271 properties we analyzed, 206 enjoyed higher rents upon renewal (rents in another 14 properties did not change at all).
  • Renewal rent increases were nearly universal where the previous rent was less than $20.00/RSF.
  • As the previous rent became larger, the renewal rents were more erratic. Logically this makes sense because larger rents receive more scrutiny–and in some instances the previous rents may have been inflated due to amortization of TI and shell improvements.
  • When renewal rents increased, they did so by an average of 29.07% (22.03% median). When they decreased the average was -12.96% (-10.36% median). Across the entire 271 property set, rents increased by an average of 19.66% (14.97% median).

On the whole, this is good news for investors–especially portfolio investors hoping to maintain or increase yield. In a future article we’ll look at the factor that has much greater impact on capitalized value: renewal term.

Anticipating one final question: We did not empirically study what happened in those instances where rent decreased during the previous lease term due to de-amortization of the tenant improvement component. Yet, anecdotally, our review of the data indicates that the subsequent renewal rent had little correlation. Renewal rents appeared to rise or fall in those instances about as they did across the entire sample.