GSA Leasing: Initial Implementation of the National Broker Services Contracts Demonstrates Need for Improvements

In October 2004, GSA hired four contractors (broker firms) to carry out a portion of its leasing program. The brokers were each expected to provide broker services on a nationwide basis. Two of the brokers were dual agency firms (firms that represent both building owners and tenants) and, as such, were required to establish a "conflict wall" to help prevent electronic and physical sharing of information between broker employees within the same firm who represent GSA, and other employees within the same firm who represent building owners. In January 2007, we reported that GSA had not assessed the adequacy of the two dual agencies' conflict walls. (GSA Leasing: Initial Implementation of the National Broker Services Contracts Demonstrates Need for Improvement). Consequently, we recommended that GSA assess the adequacy of the two dual agencies' conflict wall controls and recommend actions, if applicable, to correct any identified weaknesses. GSA responded to our recommendation by assessing the adequacy of the brokers' control. On the basis of the assessment, GSA concluded, in a May 22, 2007 memorandum, that the conflict walls maintained by each of the dual agencies were satisfactory.

Our January 2007 report "GSA Leasing: Initial Implementation of the National Broker Services Contracts Demonstrates Need for Improvements" found that GSA was not requiring tenant-only brokerage firms--firms that represent only the tenant in real estate transactions--to follow three controls that could prevent them from disclosing potential or actual conflicts of interest. These three controls require brokers to (1) execute additional agreements to safeguard proprietary information, (2) notify GSA of any conflicts of interest discovered during the performance of work, and (3) include a conflict-of-interest clause in all of their subcontracts. GSA made these controls applicable to address situations that dual agency firms--firms that represent both building owners and tenants--could possibly face, but not to tenant-only brokerage firms who could face similar situations. We made a recommendation to GSA to establish consistent dual-agency and tenant-only conflict-of-interest contract requirements, including, at a minimum, the three conflict-of-interest requirements that address situations also faced by the two tenant-only firms. In response to this recommendation, in May 2007, GSA amended the National Broker Contracts so that these three conflict of interest provisions are made applicable to all contractors. In addition, GSA revised the Conflict of Interest section of the National Broker Contracts Administration Guide to reflect conflict of interest restriction changes incorporated into the contract. These changes will help to ensure that tenant-only firms are aware of all requirements applicable to their disclosure of potential or actual conflicts of interest; and GSA will have a step by step approach to determining and processing potential conflicts of interest.

Federal agencies are required to develop security programs to provide assurance that contractors and others are protecting an agency's information and information systems in a manner commensurate with the risk level assigned to the information and systems by the agency. In January 2007, we reported that GSA had not fully assessed the risk and magnitude of harm that could result from the misuse of information and information systems used on behalf of GSA by the four national brokers. (GSA Leasing: Initial Implementation of the National Broker Services Contracts Demonstrates Need for Improvements.) To address this deficiency, we recommended that GSA assess the risk and magnitude of harm that could result from unauthorized access to, or use, disclosure, disruption, modification, or destruction of, GSA information collected or maintained by the four brokers (and their subcontractors) as well as the information systems used by the brokers on behalf of GSA. As of August 30, 2007, GSA had performed the recommended risk assessment and concluded that the risk level was "moderate."

In January 2007, we reported that GSA had not tested the information security controls associated with its national brokers program, including the controls used by its four national, as required by the Federal Information Security Management Act. (GSA Leasing: Initial Implementation of the National Broker Services Contracts Demonstrates Need for Improvements.) The Act requires periodic tests--at least annually--to ensure that controls are adequate for protecting agency information, including information maintained by contractors (and subcontractors). To address this weakness, we recommended that GSA test the effectiveness of federal information security policies, procedures, and practices related to the national broker services program, including, as applicable, broker controls for safeguarding GSA's information. GSA responded to our recommendation by developing a process to test the effectiveness of the controls used to safeguard program information and, as of March 15, 2008, had completed testing at one of the four brokers. According to GSA, "The continuous monitoring required by its process means that it is never complete but must be done repeatedly..." throughout the life of the contracts.

In January 2007, we reported that GSA had not developed processes to quantify savings resulting from its implementation of its national broker services program. (GSA Leasing: Initial Implementation of the National Broker Services Contracts Demonstrates Need for Improvements). The program uses contractors (brokers) to provide brokerage services nationwide. We reported that while GSA anticipated that using national brokers would (1) reduce rental costs to the government and (2) agency savings from reduced fees, administrative expenses, and personnel by shifting costs to the national broker services contracts,it had not developed a process for quantifying these savings. We recommended that GSA develop processes for quantifying these and other savings from (1) rent reductions attributable to the brokers' greater knowledge of the commercial real estate market and (2) agency savings as specified above. In response to this recommendation, GSA extracted historical data on its prior contracts and current national broker services contracts and performed a comparative analysis (through the end of the first quarter of fiscal year 2008) which demonstrated numerous cost savings attributable to its use of the national contracts for broker services.

Under the terms of GSA's extended contracts with its four national brokers, GSA initially was required to distribute task orders between the brokers "as equitably as possible" until a record of the brokers' performance could be established. When we completed our review, GSA expected to transition to performance-based allocations on April 1, 2007--the start of the third year of the contracts. While performance-based distributions appeared to be imminent, as we reported in January 2007, GSA had not yet defined the number and types of task orders needed to establish a record of the brokers' performance. We pointed out that without this information, GSA could not demonstrate that it had established a record of performance and was ready to move to performance-based distributions. Consequently, we recommended that GSA clarify the number and types of completed task orders needed to establish a record of the brokers' performance before transitioning to performance-based distributions. GSA responded to our recommendation by delaying its performance-based distributions and by developing and, tentatively approving, a plan for subsequently allocating task orders based on performance. However, GSA suspended implementation of the plan when testing revealed flaws that would have negatively impacted the national broker services program. According to GSA, the agency has no intention of moving to performance based task orders under the current contracts, all of which are nearing completion. Although GSA was not able to successful implement a plan for implementing our recommendation, it complied with the intent of our recommendation, which was not to proceed to performance-based distributions until it was ready to do so.

In January 2007, we reported that, although GSA collects data on the number and size of the task orders distributed to the four NBS brokers it does not collect data on the geographic area (e.g., rural or urban) covered by the task orders. (GSA Leasing: Initial Implementation of the National Broker Services Contracts Demonstrates Need for Improvements). This omission was significant because GSA's contracts with the brokers specify that each broker will be provided projects on a nationwide basis in both rural and urban areas during the initial period of contract performance, as long as their performance is acceptable. Consequently, we recommended that GSA begin collecting geographic data on its work distributions (i.e., rural versus urban) to help ensure that its distributions to the brokers are equitable. By August 30, 2007, GSA had responded to this recommendation by developing a methodology and subsequently collected and analyzed the data during fiscal year 2008 to better inform its distribution of task orders between the brokers.

The national contracts and administrative guidance had numerous inaccuracies, inconsistencies, and omissions that raised questions about how GSA could ensure consistency in its regions' evaluations of the brokers' performance. Problems included inapplicable evaluation criteria; variations in the criteria identified for use at different evaluation stages by the contracts, and inconsistencies between GSA's and National Institutes of Health's (NIH) performance-related terminology. To address these problems, we recommended that GSA, among other things, (1) revise the terminology in GSA's contracts and administrative guide, as appropriate, to conform to NIH's required evaluation factors; and (2) ensure that the various evaluation stages and processes are properly and adequately described in GSA's administrative guide. In response to this recommendation, GSA revised its administrative guide to clarify when each evaluation factor is to be used in assessing contractor performance at each stage of evaluation. The revised guidance also (1) clarifies how the National Institutes of Health's required annual evaluation fits within GSA's evaluation processes and (2) describes GSA's various evaluation stages and processes.