This is the accessible text file for GAO report number GAO-15-194 entitled 'Defense Base Act Insurance: State Department Should Evaluate Its Open Market System and Incorporate Leading Practices into Any Future Single Insurer Solicitation' which was released on January 6, 2015. This text file was formatted by the U.S. Government Accountability Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products' accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. Because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately. United States Government Accountability Office: GAO: Report to Congressional Requester: January 2015: Defense Base Act Insurance: State Department Should Evaluate Its Open Market System and Incorporate Leading Practices into Any Future Single Insurer Solicitation: GAO-15-194: GAO Highlights: Highlights of GAO-15-194, a report to Congressional Requester. Why GAO Did This Study: DBA requires U.S. government contractors to buy workers’ compensation insurance for most employees working overseas. The cost of this insurance, if allowable under federal regulations, is generally reimbursable under government contracts. From 1992 until 2012, State had a contract with a single insurer to supply all State’s contractors working overseas with DBA insurance. In July 2012, State’s single insurer program ended after State unsuccessfully sought to solicit a new DBA single insurer agreement and State transitioned to a system requiring its contractors to obtain DBA insurance on the open market. However, concerns were raised about the transition and its impact on State’s costs and on small businesses’ competitiveness. To address these objectives, GAO was asked to review State’s transition. This report assesses (1) State’s management of the transition to an open market system, (2) the change’s effect on contractors’ premium rates, and (3) the change’s effect on small businesses. GAO analyzed State documents; reviewed federal and State contracting regulations; analyzed premium rate data and federal contracting data; and interviewed officials from State, the insurance industry, and contracting firms. What GAO Found: The Department of State (State) did not follow leading acquisition practices in transitioning from a single insurer Defense Base Act (DBA) program to an open market system. Leading practices emphasize adequately documenting market research, allowing enough time to complete a solicitation, and collecting and analyzing data to select among alternatives, but State took limited measures to document the market research it performed and had little time to complete its 2012 solicitation. State included provisions in the solicitation to which insurers strongly objected, received no offers, and had to cancel the solicitation 3 days before its existing single insurer contract was to expire. As a result, State had to quickly transition to an open market system without weighing the relative costs and benefits to determine which insurance system best served its needs. Until State conducts such an evaluation, it cannot be assured that the open market system is the better alternative, and unless State incorporates leading practices into any future single insurer solicitations, it risks a similar outcome. Figure: Timeline of the Department of State’s Transition to an Open Market System for Defense Base Act Insurance: [Refer to PDF for image: timeline] September 2011: Effective July 21, 2012, State and its single insurer agree to expiration of contract. February, 2012: State requests information from insurers, brokers, and contractors. June 2012: State posts solicitation for new single insurer but extends deadline. State received no bids. July 2012: State withdraws solicitation; State transitions to open market system. August 2012: State releases notice stating that it no longer has single insurer. Source: GAO (analysis); Department of State (data). GAO-15-194. [End of figure] GAO found that State contractors’ DBA premiums increased following the transition, but the increases were in a range similar to those likely to have occurred if State had continued its single insurer program. For example, median DBA premium rates increased by $1.98 per $100 of payroll. GAO analysis also shows that the increase in DBA premium rates after the transition was in a range comparable to the increase in DBA premium rates requested by State’s single DBA insurer, which said it had lost money under the prior contract. Existing data do not show a clear effect on small businesses resulting from State’s transition to an open market system, but insurers and contractors have expressed concern that the change has had or could have an adverse effect. GAO analysis of federal procurement data from fiscal years 2009 through 2013 found a decrease in the percentage of contracts awarded to small businesses, but GAO could not link this to State’s transition. Information GAO gathered from insurance industry officials and contractors shows that there is a potential for adverse effects, for example, denial of coverage and higher effective premium rates. State’s policy is to maximize opportunities for small businesses, but it has not assessed whether its transition to an open market DBA system is affecting those opportunities. Without such an assessment, State cannot be assured that it is meeting its policy goal of maximizing opportunities for small businesses. What GAO Recommends: State should (1) determine whether an open market system best suits its needs, (2) incorporate leading practices into any future single insurer solicitation, and (3) assess the effects of its transition on small businesses. State concurred with GAO’s recommendations. View [hyperlink, http://www.gao.gov/products/GAO-15-194]. For more information, contact Michael Courts at (202) 512-8980 or courtsm@gao.gov. [End of section] Contents: Letter: Background: State Did Not Follow Leading Practices in Its Transition from a Single Insurer Program to the Open Market: DBA Premiums Rose in the Open Market, but Increases Were in a Similar Range of Those Likely Had State Continued Its Single Insurer Program: Available Data Do Not Allow for Conclusions as to Whether State's Transition Affected Small Businesses, but Our Analysis Shows a Potential for Adverse Effects: Conclusions: Recommendations for Executive Action: Agency Comments: Appendix I: Objectives, Scope, and Methodology: Appendix II: Technical Appendix for Analysis of Department of State Contractors' Defense Base Act Insurance Premium Rates: Summary Statistics: Histograms of DBA Rates: Analysis of Hypothetical Scenarios: Analysis of DBA Premiums across U.S. Government Agencies: Appendix III: Comments from the Department of State: Appendix IV: GAO Contact and Staff Acknowledgments: Related GAO Products: Tables: Table 1: Change in Effective Defense Base Act (DBA) Insurance Premium Rates from a Sample of Department of State Contractors, by Contractor Category: Table 2: Change in Effective Defense Base Act (DBA) Insurance Premium Rates from a Sample of Department of State Contractors, by Country: Table 3: Advantages and Disadvantages for Small Businesses When Obtaining Defense Base Act (DBA) Insurance through a Single Insurer Program versus an Open Market System: Table 4: Count of Population and Sample Sizes: Table 5: Summary Statistics Related to Department of State's Defense Base Act (DBA) Insurance Effective Premium Rates: Table 6: Change in Effective DBA Rates from a Sample of State Contractors, by Country: Table 7: Change in Effective Defense Base Act (DBA) Rates from a Sample of State Contractors, by Contractor Category: Table 8: Results of Tests for Differences between Open Market and Hypothetical Single Insurer Effective Rates: Table 9: Comparison of Defense Base Act (DBA) Insurance Rates across Agencies, by Contractor Category: Figures: Figure 1: Trends in the Total Amount of U.S. Government-wide Defense Base Act Benefits Paid and the Number of Claims Paid from 2002 to 2013: Figure 2: Process Map of the Single Insurer and Open Market Approaches to Providing Defense Base Act Insurance: Figure 3: Key Dates and Information Related to the Department of State's 2008 Single Insurer Contract for Defense Base Act Insurance: Figure 4: Timeline of the Department of State's Transition to an Open Market System for Defense Base Act Insurance: Figure 5: The Median Open Market Effective Rate in Our Sample Was in a Likely Range Comparable to Those Requested by State's Single Insurer: Figure 6: Sample Distribution of the Department of State's Single Insurer Program Effective Rates: Figure 7: Sample Distribution of the Department of State's Single Insurer Program Effective Rates, by Contractor Category: Figure 8: Sample Distribution of the Department of State's Open Market Effective Rates: Figure 9: Sample Distribution of the Department of State's Open Market Effective Rates, by Contractor Category: Figure 10: Confidence Intervals for Median Effective Premium Rates for Defense Base Act (DBA) Insurance: Open Market Rate Paid by Department of State Contractors Compared to Our Calculated Rates under Two Hypothetical Scenarios, by Contractor Category: Abbreviations: DBA: Defense Base Act: DOD: United States Department of Defense: DOSAR: Department of State Acquisition Regulation: FAH: Department of State Foreign Affairs Handbook: FAM: Department of State Foreign Affairs Manual: FAR: Federal Acquisition Regulation: FPDS-NG: Federal Procurement Database System - Next Generation: DOL: United States Department of Labor: OIG: Office of the Inspector General: OSDBU: United States Department of State, Office of Small and Disadvantaged Business Utilization: SIGAR: Office of the Special Inspector General for Afghanistan Reconstruction: State: United States Department of State: USACE: United States Army Corps of Engineers: USAID: United States Agency for International Development: [End of section] United States Government Accountability Office: GAO: 441 G St. N.W. Washington, DC 20548: January 6, 2015: The Honorable Bob Corker: United States Senate: Dear Senator Corker: The Defense Base Act (DBA), as amended, requires U.S. government contractors to provide employees working overseas--including, in many cases, foreign nationals--with uniform levels of workers' compensation insurance.[Footnote 1] The cost of the premiums paid for DBA coverage, if allowable under the Federal Acquisition Regulation (FAR), is generally reimbursable under government contracts.[Footnote 2] According to a September 2009 Department of Defense (DOD) report, with the increase in U.S. military and civilian engagement throughout the world, the cost to the U.S. government for DBA insurance premiums sharply increased during the past decade; total premiums paid to the four largest DBA insurance providers increased from approximately $18 million in 2002 to approximately $400 million in 2007.[Footnote 3] Although more recent and comprehensive data on the total cost for DBA premiums are not available, data on the cost of DBA claims to the insurance industry suggest that DBA premium costs have continued to rise. According to Department of Labor (DOL) data, between 2002 and 2013, the total number of claims rose from 430 to 14,705, and the total amount of DBA benefits paid to claimants working for U.S. government contractors rose from about $9.5 million to about $666 million in inflation-adjusted dollars. Figure 1 shows the trends in the total amount of DBA benefits paid in inflation-adjusted dollars and in the number of claims paid from 2002 to 2013. Figure 1: Trends in the Total Amount of U.S. Government-wide Defense Base Act Benefits Paid and the Number of Claims Paid from 2002 to 2013: [Refer to PDF for image: multiple line graph] Fiscal year: 2002; Claims paid: 430; Total benefits (2013 dollars): $9.5 million. Fiscal year: 2003; Claims paid: 688; Total benefits (2013 dollars): $14.0 million. Fiscal year: 2004; Claims paid: 1,592; Total benefits (2013 dollars): $36.0 million. Fiscal year: 2005; Claims paid: 3,080; Total benefits (2013 dollars): $69.4 million. Fiscal year: 2006; Claims paid: 5,039; Total benefits (2013 dollars): $130.3 million. Fiscal year: 2007; Claims paid: 11,887; Total benefits (2013 dollars): $186.6 million. Fiscal year: 2008; Claims paid: 8,741; Total benefits (2013 dollars): $215.0 million. Fiscal year: 2009; Claims paid: 14,863; Total benefits (2013 dollars): $258.9 million. Fiscal year: 2010; Claims paid: 11,799; Total benefits (2013 dollars): $328.6 million. Fiscal year: 2011; Claims paid: 15,137; Total benefits (2013 dollars): $431.7 million. Fiscal year: 2012; Claims paid: 14,548; Total benefits (2013 dollars): $548.3 million. Fiscal year: 2013; Claims paid: 14,705; Total benefits (2013 dollars): $665.9 million. Source: GAO analysis of Departments of Commerce and Labor data. GAO-15- 194. Note: Inflation-adjusted benefits are in 2013 dollars. [End of figure] From 1992 until 2012, the Department of State (State) had an exclusive contract[Footnote 4] with a single insurer to supply all of its contractors with DBA insurance. However, in July 2012 State's single insurer program ended after State unsuccessfully sought to select a new DBA single insurer through a competitive source selection process. State now requires its contractors to obtain DBA insurance on the open market from 1 of 38 insurers authorized by DOL.[Footnote 5] You asked us to review State's transition to an open market DBA system.[Footnote 6] This report assesses (1) how State managed the transition from a single insurer program to an open market system, (2) the extent to which this change affected DBA premium rates paid by State's contractors, and (3) the extent to which this change affected small businesses. To examine State's transition to an open market system, we reviewed federal and State Department acquisition regulations, internal controls, and leading acquisition practices as reported previously by GAO. We also reviewed reports about DBA insurance issued by the State, U.S. Agency for International Development (USAID), and DOL Offices of Inspector General (OIG) as well as the Special Inspector General for Afghanistan Reconstruction (SIGAR). In addition, we reviewed contract and correspondence files from State and other agencies, and interviewed officials from State and DOL as well as officials from insurance companies, insurance brokers, contractors, and small businesses (hereafter referred to as insurers, brokers, contractors, and small businesses). The U.S. Army Corps of Engineers (USACE) made a similar transition from a single insurer program to an open market system in October 2013, and we compared how State and USACE managed their transitions to an open market system. To determine the extent to which State's transition affected DBA premium rates paid by State contractors, we obtained data on DBA premiums and premium rates from a random sample of 71 State contractors who obtained contracts through State's Office of Acquisitions Management and who performed work throughout State's transition to the open market (fiscal years 2012 to 2013) in Afghanistan, Iraq, Pakistan, Thailand, and South Africa. We identified the set of countries by analyzing country summary reports of contract actions for State and USAID in federal procurement data for fiscal years 2010 through 2013, identifying the top 10 countries for each agency in terms of number of contract actions, and choosing the countries that appeared in the top 10 list for both agencies.[Footnote 7] Results from analyzing this sample are generalizable to the population of 164 contractors that meet these selection criteria. We also conducted interviews with insurers, brokers, contractors, and small businesses. We compared the open market premium rates paid by the contractors in our sample with rates that would have been charged by a single insurer in two hypothetical scenarios under which State might have continued its single insurer program. The hypothetical scenarios were calculated from pricing information contained in State's most recent single insurer contract, signed in July 2008, as well as official documents and correspondence between State and its single insurer during the course of that contract. Data on contractors' DBA premiums and payroll were self-reported. To examine the extent to which State's transition to an open market system affected small businesses, we analyzed federal procurement data and interviewed officials from State, insurers, brokers, and contractors, including small businesses and representatives of a national association that represents government contractors. Because we used federal procurement data to assess the effects of State's transition to an open market system on small businesses, we assessed the reliability of these data. We determined that the data were sufficiently reliable for our use. We conducted this performance audit from March 2014 to January 2015 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. See appendix I for a more complete description of our scope and methodology. Background: Congress enacted DBA in 1941 to provide workers' compensation protection to employees of government contractors working at U.S. defense bases overseas. Subsequent amendments to DBA extended coverage to other classes of employees. DBA insurance provides covered employees with uniform levels of disability and medical benefits--or in the event of death provides benefits to their eligible dependents. DOL administers DBA, ensuring that workers' compensation benefits are provided for covered employees and overseeing the claims process, among other things. Under DBA, U.S. government contractors and subcontractors are required to obtain DBA insurance for all employees, including foreign nationals, unless DOL issues a waiver.[Footnote 8] The cost of DBA insurance premiums, if allocable and reasonable, is generally reimbursable under government contracts. Under the War Hazards Compensation Act,[Footnote 9] the government also reimburses insurers for DBA benefits paid if the injury or death is caused by a "war-risk hazard,"[Footnote 10] provided that the insurer did not charge its customer a war-risk hazard premium. In addition to disability and death payments, war-risk hazard benefits include funeral and burial expenses, medical expenses, and reasonable costs necessary to process the claims. In providing for DBA coverage, agencies mainly use one of two approaches: a single insurer program or an open market system. For instance, USAID has a single insurer program. DOD has generally employed an open market system. USACE, however, had a single insurer program from 2005 to 2013, and then transitioned to an open market system. State adopted a single insurer program in response to a 1991 report by its OIG, which among other things, estimated that State could save approximately 40 percent of its DBA cost by moving to a single insurer program.[Footnote 11] Under a single insurer program such as the one implemented by State, an agency selects one insurer to provide DBA insurance through a competitively selected multiyear agreement. The resulting agreement sets premium rates for the agency's contractors. The agreement may stipulate that rates will remain fixed during the entire term of the agreement, or that the rates may be adjusted up or down by mutual consent of the agency and the single insurer. To obtain DBA insurance, the agency's contractors contact a broker specified by the single insurer. The broker obtains the contractor's statement of work, assigns the contractor to a service category (for example, security), and collects premiums based on the contractor's payroll and the premium rates for the service category. Under State's single insurer program, DBA insurance was listed as a separate line item on the agency's contract with each contractor. Under an open market system, contractors must independently obtain DBA insurance coverage from an insurer licensed by DOL to underwrite DBA insurance, and they usually do this through a broker. In this system, agencies do not play a role in setting premium rates. Each contractor's selected insurer issues a DBA insurance policy that fixes premium rates generally for 1 year; after that, the rates and corresponding premiums can move up or down based in part on the insurer's assessment of the contractor's risk. The initial premium rate can vary by contractor based on the insurer's assessment of the contractor's risk. Contractors with a history of few or no claims can see a reduction in their premiums when renewing their DBA insurance, while contractors with a history of many claims can see an increase. To reduce the likelihood of claims, contractors can sometimes participate in a risk assessment and reduction program sponsored by their DBA insurer. According to State guidance issued in August 2012, it is no longer required that DBA insurance be identified as a separate line item on the agency's contract with each contractor. Depending on the type of contract, a contracting officer may include the cost of DBA insurance separately.[Footnote 12] Figure 2 provides a process map of the two approaches. Figure 2: Process Map of the Single Insurer and Open Market Approaches to Providing Defense Base Act Insurance: [Refer to PDF for image: process illustration] Single Insurer Program: 1. Agency awards contract for overseas work. [Same as open market system] 2. Contractor submits statement of work and payroll information to designated insurance broker. [Differs from open market system] 3. Broker assigns service category and submits application to insurer. [Differs from open market system] 4. Contractor pays broker, which issues proof of insurance. [Differs from open market system] 5. Agency issues notice to contractor to begin work. [Same as open market system] 6. Contractor performs work. [Same as open market system] 7. If multi-year contract, contractor applies for renewal at rates set by single insurer program. Return to step 2. [Differs from open market system] Open Market System: 1. Contractor contacts one or more insurance brokers to obtain DBA insurance pricing. [Differs from single insurer program] 2. Broker shops for insurer on behalf of contractor. [Differs from single insurer program] 3. Broker and insurer quote DBA premium, which contractor includes in bid submitted to U.S. agency. Go to #5. [Differs from single insurer program] 4. Contractor repeats initial steps with brokers, insurers to obtain new DBA insurer. Go to #6. [Differs from single insurer program] 5. Agency awards contract for overseas work. [Same as single insurer program] 6. Insurer issues policy and proof of insurance to contractor. [Differs from single insurer program] 7. Agency issues notice to contractor to begin work. [Same as single insurer program] 8. Contractor performs work. [Same as single insurer program] 9. If multi-year contract, contractor applies for renewal at annually negotiated rates: [Differs from single insurer program] Both parties satisfied with current policy? If yes, return to #6; If no, return to #4. DBA: Defense Base Act. Source: GAO (analysis); Department of State and insurance industry officials (data). GAO-15-194. [End of figure] State Did Not Follow Leading Practices in Its Transition from a Single Insurer Program to the Open Market: In 2008, State entered into a multiyear contract with a single DBA insurer. In September 2011, State and its single insurer agreed that the DBA contract would expire in July 2012. In June 2012, State posted a solicitation seeking to select a single insurer through a competitive source selection process. After State received no offers in response to its solicitation, it withdrew the solicitation 3 days before its existing contract expired and transitioned to a system requiring contractors to obtain DBA insurance on the open market. Leading acquisition practices emphasize the importance of allowing enough time to complete a solicitation, adequately documenting market research, and collecting and analyzing data, among other things. We found that State did not follow these leading practices.[Footnote 13] Specifically, State (1) had little time to complete the process of designating a DBA single insurer, (2) did not conduct a lessons learned assessment after agreeing to terminate its existing DBA contract, (3) did not adequately document market research, and (4) did not provide sufficient information to insurers. As a result, State had to quickly transition to an open market system without evaluating the relative costs and benefits involved. According to officials of insurers, brokers, and some contractors, after transitioning, State did not communicate its change in a timely manner to insurers, brokers, and contractors, which caused confusion among contractors and left some with little time to replace their expiring DBA policies. State Transitioned to an Open Market System in July 2012: Within State's Bureau of Administration, its Office of the Procurement Executive is responsible for maintaining State's acquisition regulations and procedures, and its Office of Acquisitions Management manages most of the procurement for State domestically and overseas. The Office of Acquisitions Management was in charge of the competitive source selection processes for State's single insurer program. In 2001, State first signed a contract with its most recent single insurer to provide DBA insurance to State contractors. In 2008, State signed a similar contract with the same single insurer, this one a 5- year contract consisting of a base year and 4 option years. The premium rates for the base year, or year 1, were fixed, but the contract allowed for adjustments in the rates by mutual agreement for years 2, 3, 4, and 5, with any adjustments to be based on cumulative losses sustained by State's single insurer since the start of the contract. To ensure that contractors were appropriately assigned a DBA premium rate based primarily on risk, the 2008 contract provided that contractors would be assigned to one of four types of service rate categories (listed here from highest to lowest risk): security contractors involved in aviation-related work in defined hazardous areas such as Afghanistan and Iraq, other security contractors working in the same hazardous areas, construction contractors, and service contractors. Figure 3 shows key dates and information related to State's 2008 single insurer contract. Figure 3: Key Dates and Information Related to the Department of State's 2008 Single Insurer Contract for Defense Base Act Insurance: [Refer to PDF for image: timeline] 2008: State enters into Defense Base Act (DBA) contract with single insurer. Program year: (July 22 - July 21): Program Year 1: base year; Program Year 2: State’s single insurer raises a concern about increasing DBA losses and requests premium rate increase. 2010: State acknowledges DBA losses but refuses rate increase. Program Year 3: State’s single insurer requests lump sum payment and rate increase and warns of legal action. 2011: State and single insurer agree that the contract would expire in July 2012. Program Year 4, July 2012: Contract expired. Source: GAO (analysis); Department of State (data). GAO-15-194. [End of figure] Prior to the start of year 3 of the contract in July 2010, State's single insurer raised a concern about increasing DBA losses and the potential for further increases in losses. Among other things, State's single insurer asked for an increase in premium rates for year 3. [Footnote 14] According to documentation from State and State's single insurer, the losses far exceeded what the single insurer's actuaries had anticipated and resulted from several factors, including an increase in the number of claims from State's security contractors, many of which operated in hazardous areas. The single insurer also cited an increase in the number of injured employees that had not found suitable work upon release from their contract because of poor economic conditions or the nature of their injuries, which could include post-traumatic stress disorder.[Footnote 15] State acknowledged that DBA losses had increased, but disagreed about the extent of the losses and did not agree to allow the single insurer to increase its premium rates. State also did not agree to the single insurer's request to obtain the services of an outside actuarial expert to validate the losses sustained by State's single insurer. In May 2011, prior to the start of year 4 of the contract, State's single insurer requested a lump sum payment of approximately $27 million to cover losses from year 3 of the contract as well as an increase in premium rates for year 4 of the contract.[Footnote 16] This request was accompanied by a warning of legal action. State's single insurer concluded that it had underestimated its 2010 losses and stated that a revised methodology it used to estimate losses showed that it should have charged higher premiums beginning with year 3 (July 2010-July 2011). State's single insurer also stated that its revised loss estimate methodology, which used forecasted losses rather than actual sustained losses, was consistent with insurance industry standards.[Footnote 17] In response, State noted that the single insurer had not previously used forecasted losses as part of its methodology to calculate premium rates. State disagreed with the single insurer's revised methodology and said that the single insurer's request for a lump sum payment and premium rate increases raised numerous questions about how the insurer had originally established its premium rates, along with other concerns.[Footnote 18] State and its single insurer met to discuss the insurer's requests and exchanged correspondence but could not resolve their differences. Because State and its single insurer could not agree on a solution, in September 2011, State signed a memorandum with its single insurer agreeing that State would not exercise its right to continue the program through year 5, the final option year of the contract. Between then and July 2012, a number of key events occurred. These are noted below: * September 12, 2011: State and its single insurer agreed that the DBA contract would expire on July 21, 2012.[Footnote 19] * February 9, 2012: State posted a request for information containing a DBA-related questionnaire to insurers, brokers, and contractors on a federal procurement website. Nine responded by February 28, 2012. * June 14, 2012: Seeking to designate a single insurer through a competitive source selection process, State posted a solicitation requesting that insurers submit proposals by June 27, 2012. * June 14 through July 16, 2012: State twice extended the time frame for the solicitation, ultimately setting the deadline as July 16, 2012. Responding insurers stated that the solicitation contained objectionable provisions. On July 16, 2012, one insurer filed a bid protest with GAO.[Footnote 20] * July 18, 2012: Having received no offers in response to its solicitation, State withdrew the solicitation. * July 21, 2012: State's existing single insurer program ended. * July 22, 2012: State transitioned to an open market system. * August 9, 2012: State issued a notice stating that as of July 22, 2012, State no longer had a single insurer. The notice also stated that contractors could purchase insurance from any insurer approved by DOL to provide DBA insurance. * August 28, 2012: State formally notified its contracting officers that they were required to inform contractors of the transition to an open market system as DBA insurance policies expired. State Did Not Follow Leading Practices during Its Transition to an Open Market System: State Lacks Specific Guidance for DBA Selections, but Officials Said They Followed Acquisition Regulations to the Extent Practicable: State officials said they lacked specific guidance for acquiring a new DBA contract but followed acquisition regulations to the extent practicable. The selection of a single provider of DBA insurance is not an acquisition because it does not involve the use of appropriated funds. Therefore, these selections are not subject to the FAR. State officials told us that in the absence of guidance on how to execute a DBA selection, they looked to the FAR and the Department of State Acquisition Regulation (DOSAR) as their guidance to the extent practicable to conduct what the agency labeled a "competitive source selection process." For example, both the 2008 single provider agreement and the 2012 solicitation included multiple references to the FAR, and State used certain provisions of the FAR as criteria for determining whether to exercise options on the 2008 agreement. In the absence of specific State guidance for solicitations involving DBA selections and in light of the use of federal funds,[Footnote 21] we looked to leading practices set forth in the FAR and DOSAR, as well as in other State acquisition guidance to evaluate State's management of the DBA transition.[Footnote 22] Moreover, in accordance with federal requirements, State has internal control standards that apply to competitively sourced acquisitions.[Footnote 23] These internal control standards state that appropriate internal controls apply to all State operations and administrative functions. In addition, GAO has in past reports discussed a number of other leading acquisition practices that are applicable to competitively sourced acquisitions. State Had Little Time to Complete Its June 2012 Single Insurer Solicitation and Communicate Its Decision to Transition to an Open Market System: State had little time to complete its 2012 single insurer solicitation, and when it received no offers, State had to withdraw its solicitation only 3 days before its existing DBA contract was due to expire. We have reported on the need for agencies to establish time frames for acquisition planning, including measuring lead times for presolicitation and solicitation activities.[Footnote 24] In addition, State guidance used by the Office of Acquisitions Management states that sufficient time must be allowed to perform the many steps involved in the acquisitions process.[Footnote 25] State officials told us they began planning for the June 2012 single insurer solicitation well in advance; however, they did not issue the solicitation until about a month before their existing single insurer program was due to expire. The timeline in figure 4 shows how compressed State's single insurer solicitation process became. Because State compressed its contracting efforts, it had little time to complete the solicitation process. In 2007, State did not complete the solicitation process in time to obtain a DBA single insurer; however, in that instance, State was able to reach an agreement with its single insurer to extend the contract for 6 months--an option that it did not have in this instance since it had already agreed with its single insurer that the existing contract would expire on July 21, 2012. Figure 4: Timeline of the Department of State's Transition to an Open Market System for Defense Base Act Insurance: [Refer to PDF for image: timeline] September 12, 2011: Effective July 21, 2012, State and its single insurer agree to expiration of contract. February 9, 2012: State requests information from insurers, brokers, and contractors. June 14, 2012: State posts solicitation for new single insurer, but extends deadline twice. June 14–July 18, 2012: State receives no bids. July 18, 2012: State withdraws solicitation. July 22, 2012: State transitions to open market system. August 9, 2012: State releases notice stating that it no longer has single insurer. Source: GAO (analysis); Department of State (data). GAO-15-194. [End of figure] State officials told us that their decision to transition to an open market DBA system primarily resulted from the lack of bids on its single insurer solicitation rather than from a policy assessment and decision based on an analysis of the costs and benefits of both approaches. State officials also told us that their decision to transition to an open market system was heavily influenced by discussions with DOL, which had included a proposal in its fiscal year 2014 budget to reform DBA by creating a government-wide self-insurance program. As a result, State officials said they viewed the transition to an open market system as a temporary measure. However, DOL officials told us that they decided to reevaluate whether a self- insurance program would be cost-effective after they had submitted their fiscal year 2014 budget proposal.[Footnote 26] In addition, State had little time to communicate with stakeholders, including contractors, its decision to transition to an open market system. State transitioned to an open market system on July 22, 2012, but did not formally communicate this internally or externally until August 9, 2012--about 3 weeks after the transition had occurred. According to an official of a national association that represents government contractors, and several contractors we interviewed, after State transitioned to an open market system, contractors raised a number of concerns regarding the transition, but State provided insufficient information to them. For example, according to the association official, contractors were initially unclear about whether and how long State's single insurer would continue to honor existing policies and whether contractors that had submitted bids using an estimate of DBA insurance costs based on State's single insurer program would be allowed to revise their bids. In general, officials of insurers, brokers, and some contractors told us that contractors were surprised and confused because State did not communicate its decision to transition to an open market system in a timely manner. These officials also said that State's late communication left some contractors with little time to replace their expiring single insurer policies. In addition, while State's own contracting officers were tasked with informing contractors of State's decision to use an open market system to provide DBA insurance, most of the contractors we spoke with did not learn of the change from State. State formally notified its contracting officers of its decision to end its single insurer program and transition to an open market system through a procurement bulletin issued on August 28, 2012--over a month after the transition had actually occurred. The bulletin required contracting officers to notify contractors that State no longer had a single insurer program and that contractors had to obtain future DBA insurance from a list of insurers approved by DOL. However, all but 2 of the 10 contractors we spoke with said that State's contracting officers did not notify them of the change to an open market system. Most said they learned about the change from their insurance brokers. [Footnote 27] State Did Not Conduct a Lessons Learned Assessment or Adequately Document Market Research to Inform Its 2012 Solicitation and Transition to the Open Market: State did not conduct a lessons learned assessment to inform its 2012 DBA solicitation, as suggested by leading practices, and though it conducted market research, it took limited measures to document the analysis and conclusions. State's and GAO's internal control standards maintain that significant events and information need to be clearly documented to ensure that management directives are carried out and to inform future decision making.[Footnote 28] One way agencies may meet this standard is through conducting a lessons learned assessment and market research for future competitive source selections. State encountered serious difficulties in implementing its 2008 single insurer contract but did not prepare an assessment detailing the lessons learned from the implementation of that contract. In 2004, 2008, and 2011, we reported that a knowledge base of important lessons learned from the acquisition process can help program and acquisition staff plan future acquisitions.[Footnote 29] In addition, when conducting procurements, agency heads prescribe procedures to ensure that knowledge gained from prior acquisitions is used to further refine requirements and acquisition strategies. While State's Office of Acquisitions Management has developed suggested guidance for documenting and incorporating lessons learned from previous acquisitions into preparations for future solicitations, it did not do so until August 2013 after it issued its 2012 DBA solicitation. This guidance states that lessons learned should be noted and provided to contracting officials. It also states that lessons learned documents need not be extensive but should be developed when appropriate. [Footnote 30] State took limited measures to document the market research it conducted in the course of preparing for the June 2012 solicitation. Market research is used to collect and analyze data about the capabilities in the market that could satisfy an agency's procurement needs. According to the FAR, the extent to which market research is to be conducted and documented will vary depending on various factors, including complexity and dollar value.[Footnote 31] We recently recommended that agencies should include at a minimum four basic elements in their market research documentation: the market research methods used, the time frames when staff used them, an analysis of potential sources, and a conclusion based on that analysis.[Footnote 32] In preparing for its June 2012 solicitation, State took some steps to conduct market research. State posted a request for information in February 2012 on a U.S. government website that publicizes federal procurement opportunities valued at over $25,000 to insurers, brokers, and contractors asking, among other things, whether State should continue its single insurer program, whether insurers charged minimum premiums, and whether small businesses would be adversely affected by a decision to transition to an open market system. Nine insurers, brokers, contractors, and small businesses responded to the request for information. State prepared a summary compilation of responses, but it contained no analysis and no conclusion to inform decision making. By contrast, USACE documented its market research, which included an analysis and conclusions to support its October 2013 transition to an open market system (see sidebar). Side bar: U.S. Army Corps of Engineers' (USACE) Transition to an Open Market System for Providing Defense Base Act (DBA) Insurance; In October 2013, USACE transitioned from a single insurer program to an open market system. We found that, in contrast to State, USACE did follow some key aspects of acquisition guidance and leading practices. First, USACE allowed time for a transition to an open market system. In October 2012, it prepared a decision paper to justify its transition to the open market. It also prepared a business case assessment weighing the advantages and disadvantages of an open market system. On the basis of its analysis, USACE concluded an open market system to be more cost-effective for itself and for contractors than a single insurer program. Second, USACE adequately documented its market research. For example, it documented the data collection and analysis methods used, dates when research was conducted, analysis of vendor capabilities, and a conclusion based on that analysis. Third, USACE communicated its decision in a more timely manner than State. Officials said they developed a communication plan to guide the transition, including an August 23, 2013, letter informing contractors that after October 1, 2013, they would no longer obtain DBA insurance through USACE's single insurer. USACE officials also said they sent notices to overseas contractors notifying them of the change. Source: GAO analysis of USACE documentation. GAO-15-194. [End of side bar] In addition, State did not use the February 9, 2012, request for information to ask insurers to comment on two provisions that State later included as part of the June 2012 solicitation. These two provisions, along with the lack of certain data, were objectionable to insurers. State extended the deadline and responded to insurers' questions regarding these concerns. The two provisions are described below. * An "opt out" provision allowed contractors to obtain DBA insurance on the open market instead of through the single insurer if the contractor could demonstrate that it could purchase DBA insurance on the open market at a lower cost. Insurers expressed concern to State that this opt out provision would encourage larger contractors and those with fewer claims to opt out of the program, leaving the single insurer with a pool of contractors representing a much higher level of risk. During the solicitation process, insurers asked State to withdraw this provision. State acknowledged that insurers wanted this provision withdrawn, but during the solicitation process State did not agree to the insurers' request. State officials told us this provision was included as a result of information, obtained through the request for information and other sources, indicating that some contractors could save money by purchasing insurance on the open market, and that it would allow contractors dissatisfied with the performance of the designated insurer to seek competitive alternatives if available. * A "blanket coverage" provision required the single insurer to provide coverage to all subcontractors, regardless of whether they were explicitly identified or not in State's contract with the prime contractor. State officials said they included this provision to ease the work of contracting officers, who would otherwise have to obtain proof of DBA coverage from individual subcontractors. In contrast, insurers expressed two concerns. One expressed concern that the provision contradicted applicable DOL regulations, making it harder for DOL to process claims. Authorized DBA insurers are required to report to DOL the names of every employer to which they have issued a DBA insurance policy. According to one insurer, if a covered employee is injured and reports the injury to DOL, DOL must identify the insurer that is required to provide benefits to the employee; however, this is difficult or impossible to do in the absence of information about which contractors and subcontractors are covered under the DBA policy. Additionally, insurers also expressed concern that not explicitly identifying all covered subcontractors could expose the insurers to unanticipated liabilities. Because State did not develop a lessons learned assessment, adequately document market research, and use the February 2012 request for information to ask insurers to comment on the "opt out" and "blanket" provisions, State did not have sufficient information about the provisions in its solicitation that insurers would find objectionable. This explains in part why State's single insurer solicitation received no offers and motivated one insurer to file a bid protest with GAO citing among other things a lack of information necessary for bidders to compete intelligently and on an equal basis. In addition, State did not have analysis to enable it to determine whether an open market system best fit its needs. State Did Not Provide Sufficient Information to Enable Insurers to Fully Assess Risks in State's 2012 Solicitation: As part of its June 2012 solicitation, State provided some claims and premium data, but insurers expressed concern that the data were insufficient to enable them to reliably analyze State's DBA insurance needs and estimate future claims and losses as a foundation for proposing premium rates. When conducting procurements, agencies acquire and provide sufficient data to potential service providers in order to ensure that contract requirements are clear and potential service providers have sufficient information to compete on a level playing field. Under the terms of the 2008 contract, State had required its single insurer to provide premium data quarterly and claims and premium data in semiannual, and annual reports. For the June 2012 solicitation, State provided prospective insurers with two spreadsheets--one providing data on historical premiums and the second providing data on claims. According to insurance industry officials, these data were needed to enable prospective insurers' actuaries to propose premium rates. However, according to four different insurers, both spreadsheets were missing some key data such as payroll or claims data for certain years. In our review of the solicitation, we also observed that some premium data were missing. During the solicitation process, insurers asked State to supply the missing data, but State responded that it had provided all available data. Officials of two insurers we spoke with expressed concern that not providing these data gave an unfair competitive advantage to State's previous single insurer, since only that insurer had access to the complete claims and premium history data. DBA Premiums Rose in the Open Market, but Increases Were in a Similar Range of Those Likely Had State Continued Its Single Insurer Program: Our analysis of a sample of State contractors' DBA rates during the transition from a single insurer program to an open market system showed that the rates increased as State moved to the open market system, but the increases fell in a similar range of those that were likely to have occurred if State had continued its single insurer program. Because State does not collect and analyze data on DBA costs in the open market system, we collected and analyzed data on DBA premiums and payroll from a random sample of State contractors active in fiscal years 2010 through 2013. We found that in our sample of contractors, effective DBA premium rates--rates that incorporate minimum premiums--increased for the median contractor by $1.98 per $100 of payroll as State transitioned from the single insurer program to the open market system.[Footnote 33] Our analysis also showed that the increase in effective DBA premium rates after the transition was comparable to the increase in effective DBA premium rates requested by State's single DBA insurer, which said it had lost money under the prior contract. Effective DBA Premium Rates in a Sample of State Contractors Increased as State Moved to the Open Market System: We found that in a sample of 36 contractors in five countries, effective DBA premium rates--rates that incorporate minimum premiums-- increased for the median contractor by $1.98 per $100 of payroll as State transitioned from the single insurer program to the open market system.[Footnote 34] According to insurance industry representatives, a major reason contractors' DBA costs increased following State's transition to the open market is the prevalence of minimum premiums required by DBA insurers in the open market (see sidebar for definitions of insurance-related terms).[Footnote 35] In contrast to the single insurer program, under the open market system, State's contractors may be required to pay a minimum premium set by their insurer, which is the minimum dollar amount the insurer requires to write coverage for a contractor. DBA insurance industry representatives told us that minimum DBA premiums can range between $5,000 and $25,000 per policy. Hence, open market minimum premiums can effectively increase the price of insurance, as measured in dollars per $100 of payroll (i.e., the effective premium rate), particularly for smaller contractors with fewer contracts. Side bar: Definitions of Insurance-Related Terms: Term: Loss; Definition: The dollar amount basis of a claim for damages under the terms of an insurance policy. From the claimant's perspective, a loss is the monetary benefit the claimant is entitled to receive; Term: Premium rate; Definition: The price of insurance, typically stated in dollars per $100 of covered payroll; Term: Minimum premium; Definition: The minimum dollar amount necessary to receive coverage under an insurance policy; Term: Total premium; Definition: Equal to the larger of (1) the premium rate multiplied by payroll or (2) the minimum premium; Term: Effective premium rate; Definition: Total premium divided by $100 of payroll. Because of the minimum premium, the effective premium rate may be higher than the premium rate. Source: GAO analysis of various DBA-related documents. GAO-15-194. [End of table] The increase in effective rates varied across contractor characteristics. We compared effective rates in the single insurer program with those in the open market across three sets of contractor characteristics: business size, contractor category (i.e., the nature of work performed by the contractor), and country, finding a statistically significant increase across many of these contractor characteristics. For example, for larger contractors in our sample, effective rates increased from $6.23 to $11.15 per $100 of payroll, on average; for small businesses in our sample, effective rates increased from $4.53 to $7.71 per $100 of payroll, on average.[Footnote 36] We did not find a larger increase in effective rates for small businesses compared with rates for larger contractors; this may be explained in part by differences in the type of work performed by small versus large contractors. For example, security contractors had a larger increase in effective rates compared with rates for services and construction contractors, and, as we discuss below, most of the security contractors in our sample were larger contractors. Table 1 shows the increase in the median effective premium rates for contractors in each contractor category as State transitioned from the single insurer program to the open market. Table 2 shows the increase in the median effective premium rates by country. The difference in rates shown in tables 1 and 2 suggest that State's transition to the open market may have led to variations in DBA rates according to the level of risk associated with the contractor's category and geographic location. For example, according to State and insurance industry officials, security contractors are exposed to greater levels of DBA risk, which could explain why the median rate for security contractors increased more than for other contractors in our sample. Table 1: Change in Effective Defense Base Act (DBA) Insurance Premium Rates from a Sample of Department of State Contractors, by Contractor Category: Dollars per $100 of payroll: Category: Services; (n = 15); Median single-insurer program effective rate: (A): $4.00; Median open market effective rate: (B): $4.92; Difference: (B-A): $0.92. Category: Construction; (n = 13); Median single-insurer program effective rate: (A): $5.50; Median open market effective rate: (B): $7.59; Difference: (B-A): $2.09. Category: Security; (n = 8); Median single-insurer program effective rate: (A): $8.28; Median open market effective rate: (B): $12.18; Difference: (B-A): $3.90. Source: GAO analysis of Department of State data. GAO-15-194. Legend: n = Number of observations. Note: Sample sizes and statistics above reflect contractors who reported total premium and payroll amounts under the Department of State's single insurer program and also in the open market. Single insurer program rates are calculated from the most recent DBA policy purchased by State contractors active in fiscal years 2010 through 2012. Open market rates are calculated from the first full-year, open market DBA policy purchased by the contractor after June 2012. A one- sided test (i.e., of whether open market effective rates were higher than single insurer program effective rates) showed that the median effective rate difference in all three contractor categories was significant at the 5 percent statistical significance level. [End of table] Table 2: Change in Effective Defense Base Act (DBA) Insurance Premium Rates from a Sample of Department of State Contractors, by Country: Dollars per $100 of payroll: Country: Afghanistan; (n = 12); Median single-insurer program effective rate: (A): $5.38; Median open market effective rate: (B): $8.50; Difference: (B-A): $3.12. Country: Iraq; (n = 14); Median single-insurer program effective rate: (A): $4.18; Median open market effective rate: (B): $6.75; Difference: (B-A): $2.57. Country: Pakistan; (n = 6); Median single-insurer program effective rate: (A): $5.22; Median open market effective rate: (B): $10.87; Difference: (B-A): $5.65. Source: GAO analysis of Department of State data. GAO-15-194. Legend: n = Number of observations. Note: Sample sizes and statistics above reflect contractors who reported total premium and payroll amounts under the Department of State's single insurer program and also in the open market. Single insurer program rates are calculated from the most recent DBA policy purchased by State contractors active in fiscal years 2010 through 2012. Open market rates are calculated from the first full-year, open market DBA policy purchased by the contractor after June 2012. A one- sided test (i.e., of whether open market effective rates were higher than single insurer program effective rates) showed that the median effective rate difference in all three countries was significant at the 5 percent statistical significance level. Data for Thailand and South Africa were not reported because of the small sample sizes in those countries. [End of table] Our Sample Data Suggest That Increases in DBA Rates Were in a Similar Range of Those Likely If State Had Continued Its Single Insurer Program: An increase in DBA rates for State contractors was likely to occur regardless of whether State had continued its single insurer program because State's single insurer reported it was losing money under the 2008 DBA contract. According to e-mails and other documents provided by State, State's single insurer repeatedly communicated to State that the insurer was operating at a loss under the terms of the 2008 DBA contract. Our analysis shows that the increase in State contractors' DBA premium rates following the open market transition was comparable to the increase in DBA premium rates requested by State's single insurer. We compared open market premium rates paid by our sample of State contractors with the rates we calculated that they would have paid under two hypothetical scenarios in which State continued the single insurer program for an additional year. The hypothetical scenarios are as follows: [Footnote 37] * In the highest contract rates scenario, State and the single insurer continue the single insurer program, with the insurer charging all contractors the highest rates allowable in State's 2008 single insurer contract. The 2008 DBA contract between State and its single insurer allowed for three tiers of premium rates, varying by State contractors' total DBA loss experience. In the contract, higher losses would result in higher DBA premium rates. During the actual lifespan of the 2008 contract, State's single insurer requested that the highest tier of premium rates be charged in years 3 and 4 of the contract; however, these requests were not granted because of the disputes between State and its single insurer discussed previously. * In the renegotiated rates scenario, State and the single insurer continue the single insurer program but renegotiate premium rates. The single insurer charges security contractors a higher, renegotiated security rate. For most of the contractors in our sample, the effective rates they paid in the open market were comparable to the effective rates they would have paid under the hypothetical scenarios described above. [Footnote 38] Figure 5 shows that there was a high degree of overlap between the likely ranges for the median open market effective rate calculated from our sample data and the likely ranges for the median effective rate in the two hypothetical scenarios. For services and construction contractors--48 of the 56 contractors in our sample--our analysis did not reveal a statistically significant difference between the median effective rate paid by contractors in the open market and the median effective rates that they were likely to have paid if State had continued the single insurer program. For the 8 security contractors in our sample, our analyses showed that open market effective rates were significantly lower than effective rates under the hypothetical scenarios. Open market premium rates for security contractors in our sample could have been lower than the rates calculated for them under the hypothetical scenarios in part because insurers in the open market may choose to insure only security contractors with a good loss history; in other words, contractors in our sample may have benefited from the open market to an extent that may not hold across all security contractors. Brokers told us that the open market allowed insurers to be selective about the customers they choose to insure. Additionally, the security contractors in our sample were larger contractors that may have been able to pool the risk of security contracting alongside work with lower risk exposure to reduce their DBA premiums. Only 1 of the 8 security contractors in our sample was a small business, and 1 security contractor told us that its payroll incorporated both security and clerical employees. In appendix II we describe the methodology of the analysis we conducted.[Footnote 39] Figure 5: The Median Open Market Effective Rate in Our Sample Was in a Likely Range Comparable to Those Requested by State's Single Insurer: [Refer to PDF for image: illustration] Rates are approximations: Scenario: Open market effective rates; Rates: $6.50 (range: $4.50-$8.50). Scenario: Highest contract rates scenario; Rates: $8.50 (range: $7.50-$9.50). Scenario: Renegotiated rates scenario; Rates: $5.50 (range: $5.00-$6.50). Source: GAO (analysis); Department of State (data). GAO-15-194. Note: Figure shows 95 percent confidence intervals for median effective rates for open market and hypothetical scenarios. See appendix II for further discussion. [End of figure] Available Data Do Not Allow for Conclusions as to Whether State's Transition Affected Small Businesses, but Our Analysis Shows a Potential for Adverse Effects: Our analysis of existing federal procurement data did not show a clear effect on contractors that are small businesses resulting from State's transition to an open market system, but insurers and contractors have expressed concern that the change has had or could have an adverse effect. State has not conducted an assessment of whether its 2012 transition affected small businesses' competitiveness. Our analysis of information provided by insurers, brokers, and contractors shows that there is a potential for adverse effects. State's policy calls for it to maximize opportunities for small businesses. Without conducting an assessment, State cannot be assured that it is meeting its policy goal of maximizing opportunities for small businesses. State First Adopted a Single Insurer Program in Part to Help Small Businesses' Competitiveness, Consistent with Its Policy of Maximizing Opportunities for Small Businesses: Several audit reports, including State OIG's 1991 and 1997 reports, as well as DOD's 2009 report, discussed the advantages and disadvantages for small businesses when obtaining DBA insurance through a single insurer program versus an open market system. Table 3 describes the advantages and disadvantages based on our interviews with insurers, brokers, contractors, and small businesses as well as our analysis of State's request for information in preparation for its 2012 single insurer solicitation. Table 3: Advantages and Disadvantages for Small Businesses When Obtaining Defense Base Act (DBA) Insurance through a Single Insurer Program versus an Open Market System: Single insurer program; Advantages: Guaranteed access to DBA insurance; Fixed rate; No minimum premium; Smaller administrative burden; Disadvantages: Claims history irrelevant because rates are fixed; No opportunity to negotiate lower rate by improving claims history; No incentive to reduce risk. Open market system; Advantages: Improved claims history can reduce rates; Incentive to reduce risk; Disadvantages: Minimum premiums can be relatively high per $100 payroll; Insurers can deny coverage; Larger administrative burden; DBA costs can become a factor in competing for bids, particularly in firm-fixed-price procurements. Source: GAO analysis of interviews with insurers, brokers, and contractors. GAO-15-194. [End of table] State's policy calls for it to maximize opportunities for small businesses to participate in the acquisitions process. State officials told us that State first adopted a single insurer program for DBA in 1991 in part to make it easier for small businesses to obtain DBA insurance. According to the DOSAR, State is to provide maximum opportunities for U.S. small businesses to participate in the acquisitions process.[Footnote 40] As required by law, State has created an Office of Small and Disadvantaged Business Utilization (OSDBU) to help small businesses participate in State contracting. [Footnote 41] State's OSDBU also works to ensure agency compliance with legal requirements contained in the Small Business Act (for example, determining annual small business goals for the agency). State awarded $2.1 billion in contracts to small businesses for work performed overseas in fiscal years 2009 through 2013.[Footnote 42] Approximately one of every five newly awarded contracts for work performed overseas during this period was completed by small businesses, which provide State with a broad range of services worldwide. Federal Procurement Data Are Inconclusive as to the Effect of State's Transition to the Open Market on Small Businesses: Our analysis of federal procurement data was inconclusive regarding the effects of State's transition on small businesses' ability to obtain State contracts, and the full effects might not be known for several years. We analyzed State's federal procurement data to track the percentage and value of new awards that went to small businesses before and after the shift to the open market DBA rate.[Footnote 43] We found that for all State's federally procured work performed by U.S. contractors outside the United States in fiscal years 2009 through 2013, the number of new contracts going to small businesses (as a percentage of all new contracts) declined from 22 percent to 15 percent. For work performed during the same period in only those countries affected by State's transition to an open market system, the decrease was nearly identical.[Footnote 44] There was a 2 percentage point decrease in fiscal year 2013, the first complete fiscal year after the transition for which data are available. However, that decrease may be part of the overall decline that we identified. In addition, when we looked at contracts over $25,000 in value, we found that the percentage of contracts going to small businesses actually increased by about 1 percent in fiscal year 2013. The major challenge in conducting this analysis is that federal procurement data do not specify which awards required DBA insurance. Therefore, we could not isolate those awards and examine data on them before and after the transition to the open market. Moreover, the full effects of the 2012 transition to the open market on small businesses may not be apparent for several more years. While State Had Not Observed an Effect on Small Businesses, Insurance Industry Officials and Contractors Noted a Potential for Adverse Effects: Cognizant State officials told us that they had not observed any impact of the transition to an open market system on small businesses, nor had they received complaints regarding DBA insurance after the transition from either small or large businesses. As noted previously, State issued a request for information in February 2012 and received responses from two insurers, two brokers, and five contractors. Our analysis of the responses showed that three of the four insurers and brokers responded that transitioning to an open market could negatively affect small businesses in two ways: (1) Insurers could deny them coverage and (2) insurers could charge minimum premiums significantly higher than what the small businesses paid under the single insurer program. The one small business that responded to State also reported a potential for adverse effects, citing difficulty obtaining coverage and higher premium rates. State officials told us they reviewed the responses but did not document their review or any conclusions they reached as a result of their review. Thus far, according to State officials, State's primary way of assessing the effects of its transition has been to informally monitor feedback received from small businesses. State officials told us that they have not received any complaints or concerns from small businesses about having to obtain DBA insurance through an open market system. According to State, the absence of complaints indicated that the transition to an open market system had not adversely affected small businesses. However, State officials also said that businesses may be reluctant to discuss uncertainties about DBA costs with State personnel. Industry officials noted that if the transition had not already had a negative effect on small businesses' competitiveness, it may in the future. Insurers, brokers, and contractors told us that the existence of minimum premiums under the open market system could have an adverse effect on small businesses' ability to perform work for State. As noted previously, we found the same concerns prevalent among the responses that State received for its February 2012 request for information. Data we reviewed from a major DBA broker showed that several small businesses ended up with a lower premium rate (per $100 of payroll) in the open market; for example, one small service business saw its rate decrease by half. However, the data also showed that in selected instances, small businesses were charged substantially more for premiums after State transitioned to an open market system. For example, one small service business' DBA premium was reported to have risen from $280 to $9,000. Another small business reported that its premium rose from $883 to $7,500. In both cases, the increase was due to minimum premiums. While we cannot determine what effect State's transition had on small businesses based on our analysis of federal procurement data, our discussions with insurance industry officials showed that there is a potential for adverse effects.[Footnote 45] To learn more about those effects, we interviewed a judgmentally selected sample of eight small businesses and two large contractors that worked with small subcontractors.[Footnote 46] Four of the small businesses reported experiencing no adverse effects. One small business was able to retain its previous single insurer premium rate in the open market system. [Footnote 47] Another small business told us that State's transition had no impact on it because the increase in its premium rate was minor.[Footnote 48] However, six of the small businesses we interviewed indicated a potential for adverse effects, two of which (minimum premiums and difficulty obtaining DBA insurance) were raised by respondents to State's February 2012 request for information from insurers, brokers, and contractors. After State's transition to an open market system, 3 of the 8 small businesses were required to pay minimum premiums to their insurer that increased the cost of their DBA coverage. Additionally, both of the larger contractors stated that small businesses that they subcontract with may have difficulty obtaining DBA insurance. In nearly all cases, the cost of DBA insurance (including minimum premiums) is passed on to State. However, two small businesses expressed concern that the additional costs imposed by minimum premiums would negatively affect their competitiveness against larger contractors who can more easily absorb higher DBA costs. Two of the small businesses expressed concern that under a multiyear firm-fixed-price contract, if their DBA costs increased in the second or third year, they would be forced to absorb the added cost. One small business said that an increase in its DBA premium rate forced an early cancellation of a multiyear State contract. In our interviews, some small businesses expressed uncertainty about whether State would reimburse certain DBA costs; however, State officials stated that most DBA costs are reimbursable. For example, one contractor told us that it had to absorb the increase in costs resulting from State's transition to an open market DBA requirement. In October 2014, State officials told us that contractors' concerns about reimbursement for increases in DBA costs were unfounded because State reimburses the cost of DBA insurance. As noted above, State officials said that contractors may be reluctant to discuss uncertainties about DBA costs with State personnel. However, State officials also said that in some cases involving work on State posts overseas, contractors might not be reimbursed for DBA costs. For example, one small construction firm was required to purchase DBA insurance, which included a minimum premium, in order to complete a site assessment at a State facility overseas before bidding on the contract with State. In that instance, the contractor would not have been reimbursed for the cost of DBA insurance had it not won the contract. Conclusions: DBA-required workers' compensation insurance claims have increased dramatically in recent years as the use of overseas contractors has expanded. From July 2008 to April 2014, State expended over $212 million to reimburse contractors for the cost of DBA premiums. When considering whether to adopt a single insurer program or an open market system for DBA insurance--the two main approaches to providing this coverage--agencies have a responsibility to conduct the acquisition process in a way that adequately manages U.S. funds. State's transition from a single insurer program to an open market system was not the result of a policy decision based on an analysis of costs and benefits of both approaches but was required because its 2012 solicitation to select a single insurer failed to yield any bids. State was unsuccessful in its solicitation in part because it lacked guidance for how to conduct a competitive source selection process that did not involve the use of appropriated funds. Moreover, State did not make full use of relevant leading practices, including its own standards for internal control and federal and State Department acquisition regulations. Consequently, State was forced into the open market system without knowing if that system was better suited for the agency and its contractors. As a result, State cannot be assured that its transition to an open market system was in the best interest of the agency in a period when the number of DBA claims and the amount of DBA benefits paid to claimants is at an all-time high. Recommendations for Executive Action: The Secretary of State should direct State's Office of the Procurement Executive to take the following actions: 1. Evaluate whether a single insurer program or an open market system best serves its needs. 2. Incorporate leading practices into any future single insurer solicitations by determining whether existing guidance could be used, or by developing guidance based on leading practices in federal and State Department acquisition regulations and State internal control standards. 3. Conduct an assessment to determine how State's transition to an open market DBA system is affecting small businesses. Agency Comments: We provided a draft of this report for review and comment to the Department of State, DOD, DOL, and USAID. In its comments (included in their entirety in appendix III), State concurred with all three recommendations. DOD, DOL, and USAID did not provide any comments. We are sending copies of this report to the appropriate congressional committees, the Secretary of State, the Secretary of Defense, the Secretary of Labor, the Administrator of the U.S. Agency for International Development, and other interested parties. In addition, the report is available at no charge on the GAO website at [hyperlink, http://www.gao.gov]. If you or your staff have any questions about this report, please contact me at (202) 512-8980 or courtsm@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. Sincerely yours, Signed by: Michael J. Courts: Director, International Affairs and Trade: [End of section] Appendix I: Objectives, Scope, and Methodology: Our review focused on the Department of State's (State) transition from a single insurer program to an open market system for the provision of Defense Base Act (DBA) insurance. This report examines (1) how State managed the transition from a single insurer program to an open market system, (2) the extent to which this change affected DBA premium rates paid by State's contractors, and (3) the extent to which this change affected small businesses. To provide background and context for our analysis, we reviewed prior GAO reports that discuss DBA insurance, as well as reports produced by Offices of Inspector General of State, the Department of Labor (DOL), the United States Agency for International Development (USAID), and the Special Inspector General for Afghanistan Reconstruction. In addition, we reviewed other documents from the Department of Defense (DOD), DOL, and insurance industry experts that discuss the history of DBA insurance and emerging trends in the DBA market. We also conducted interviews with officials from insurance companies, insurance brokers, contractors, and small businesses (hereafter referred to as insurers, brokers, contractors, and small businesses). To examine how State managed its transition from a single insurer program to an open market system, we reviewed the Federal Acquisition Regulation (FAR) and the Department of State Acquisition Regulation (DOSAR). In addition, we reviewed State's Foreign Affairs Manual (FAM), Foreign Affairs Handbook (FAH), and acquisition source selection manual.[Footnote 49] GAO has issued a number of reports discussing leading acquisition practices, and we also reviewed these reports. We obtained a number of State and insurance industry documents and met with representatives of State, insurers, brokers, contractors and small businesses, as well as a national association that represents government contractors. We reviewed State's 2008 DBA single insurer contract and its 2012 solicitation. We compared the applicable regulations, guidance, and leading practices with what we learned from the 2008 contract, 2012 solicitation, and other documentation. We met with officials from DOL to discuss that agency's role in the process. In October 2013, the U.S. Army Corps of Engineers (USACE) also made a transition from a single insurer program to an open market system, and we met with USACE officials to discuss how they managed their transition. We also reviewed documentation provided by USACE officials that shows how they managed their transition to an open market system. To examine the extent to which State's transition affected the DBA premiums and premium rates paid by contractors, we collected data on DBA premiums under State's single insurer program and in the open market. To determine a sample of contractors for our analysis, we obtained a list of State contracts with a principal place of performance outside of the United States from fiscal years 2010 through 2013 from State. We also obtained data from State on the contractors' country of location, the contracting officers' determination as to whether the contractors were classified as small businesses, and the contractors' contact information. To determine the population of contractors for the analysis, we applied the following selection criteria to the total dataset: 1. included contractors active throughout State's DBA policy transition (2012-2013); 2. included contractors who filed procurements through State's Office of Acquisitions Management; 3. included contractors in Afghanistan, Iraq, Pakistan, South Africa, and Thailand; 4. excluded unnamed contractors (such as "Miscellaneous Foreign Awardees"); and: 5. included the longest-performing contract per vendor in each country. We identified the set of countries by analyzing country summary reports of contract actions for State and USAID in the Federal Procurement Database System-Next Generation (FPDS-NG) for fiscal years 2010 through 2013, identifying the top 10 countries for each agency in terms of number of contract actions and choosing the countries that appeared in the top 10 list for both agencies. After applying the above selection criteria, the total population of contractors in the remaining dataset was 164; conclusions drawn from statistical tests on the resulting sample are generalizable to the population of 164 contractors that meet these selection criteria. Out of this population, we selected a simple random sample of 111 contractors, asking for information on DBA rates, premiums, and payroll during the single insurer program and during the open market. We queried for country-specific contractor data by requesting DBA information for a contract task order. Thirty contractors did not respond to our inquiry. However, in total, we did not receive DBA rate data from 40 contractors, as 10 contractors who provided responses said that DBA did not apply to them for various reasons. Among nonrespondents, 11 were in Afghanistan, 8 were in Iraq, 5 were in Pakistan, 3 were in Thailand, and 3 were in South Africa; 11 were small businesses and 19 were not. We did not see a clear pattern in the geographical location or size of nonrespondent contractors. Table 4 provides the counts of the number of responses we received to our data collection instrument. Table 4: Count of Population and Sample Sizes: Group: Population; Size: 164. Group: Sample selected; Size: 111. Group: Respondents; Size: 81. Group: Nonrespondents; Size: 30. Group: Respondents who reported having no Defense Base Act (DBA) payroll and total premium data; Size: 14. Group: Respondents who reported DBA payroll and total premium data (single insurer program or open market); Size: 67. Group: Respondents who reported DBA payroll and total premium data in both the single insurer program and the open market; Size: 36. Group: Respondents who reported paying payroll and total premium in the open market; Size: 57. Group: Respondents who reported paying payroll and total premium only in the open market; Size: 21. Group: Respondents who reported paying payroll and total premium in the single insurer program; Size: 46. Group: Respondents who reported paying payroll and total premium only in the single insurer program; Size: 10. Source: GAO analysis of State data. GAO-15-194. [End of table] We determined that it was most appropriate to use samples of contractors that had different characteristics for the different analyses that we conducted. In our analysis of the increases in effective premium rates (defined as the total premium divided by hundreds of dollars of payroll) from the single insurer program to the open market, we restricted the sample to the 36 contractors who reported paying premiums and payroll in both the single insurer program and in the open market. This was done to control for contractor-specific characteristics in the pre-transition to post- transition (pre-post) analysis. In our static analysis (i.e., we did not look at pre-post differences) comparing open market effective rates to effective rates that would have been charged in hypothetical single insurer program scenarios, we were able to incorporate additional data from a larger sample of contractors that reported paying premiums and payroll in the open market. This increased the sample size for this analysis to 56. For more details on our methodology, see appendix II. We also conducted interviews with insurers, brokers, contractors, and small businesses. To assess the reliability of the FPDS-NG data, we interviewed knowledgeable officials and reviewed publicly available documentation regarding the collection and use of the data. We determined that the data were sufficiently reliable for the purposes of this report. Contractor data on DBA premiums, premium rates, and payroll were self-reported; however, we reviewed the data provided, performed logical tests on it, conducted some follow-up with the contractors, and determined they were sufficiently reliable for our analysis. Further technical discussion of the data and analysis for this objective is provided in appendix II. To assess the extent to which State's transition affected small businesses, we interviewed officials from State, insurers, brokers, and a contractors' association that represents government contractors; analyzed data contained in the FPDS-NG, and interviewed officials of 2 large contractors and 8 small businesses. To determine what regulations and guidance State has regarding small businesses' participation in agency contracting, we examined the FAR, DOSAR, and FAH. To determine the extent to which State received and acted upon information indicating how its transition might affect small businesses, we spoke with State officials from the Office of the Procurement Executive, Office of Acquisitions Management, and Office of Small and Disadvantaged Business Utilization (OSDBU). The director of State's OSDBU, whose duties include serving as an advocate for small business participation in State contracting, told us that State is meeting its small business goals as set by the Small Business Administration. However, we found the annual Small Business Administration reports, which include measures of federal agencies' compliance with small business contracting goals, to be of limited use because the measures are based on contract awards with a U.S. principal place of performance.[Footnote 50] In order to better understand how small businesses might be affected by State's transition, we interviewed 5 of the 6 largest DBA insurers in the United States. These 6 insurers processed almost all DBA insurance claims by U.S. contractors in 2013. In addition, we interviewed 3 of the largest brokers that connect contractors to insurers and a national association that represents government contractors.[Footnote 51] We also analyzed FPDS-NG data to determine the extent to which the proportion of small businesses hired by State changed after the transition. Because the FPDS-NG does not indicate which contracts or contractors are required to obtain DBA insurance, we imposed the following limits to obtain relevant data for analysis. We (1) considered only new awards, (2) considered only U.S. firms, (3) did not consider awards for work performed in the United States, (4) distinguished between countries where DBA applies and where DBA does not apply, and (5) distinguished between contracts valued above and below $25,000. We performed a number of other analyses, including subsets of contractors who performed contracts valued above $25,000 in countries with a DBA requirement, but the results were similarly inconclusive. There are limitations on the data we used. For example, it is not possible to identify which contracts contain a DBA requirement without examining the individual contracts themselves. There is no requirement that agencies include DBA-related data in their mandatory procurement data reporting to the FPDS-NG, and State has not established a mechanism for collecting data on DBA costs in an open market system. It is therefore not possible to identify exactly which State contractors are affected, if at all, by State's transition. Moreover, a limited amount of time has passed since the transition, so the window is small for observing effects. The transition occurred on July 21, 2012, but the most recent FPDS-NG data extend only to September 30, 2013. We determined that the FPDS-NG data were sufficiently reliable for the purposes of this report. We also interviewed a judgmental sample of 10 State contractors (8 small businesses and 2 large contractors) to discover how, if at all, State's transition affected them and to uncover details that a quantitative analysis of FPDS-NG data might not reveal. We selected 8 small businesses using the judgmental selection criteria listed below. We limited our sample to contractors that: * were required to obtain DBA coverage for work performed overseas; * signed contracts with State both before and after State's transition to an open market system;[Footnote 52] * performed work for State overseas in countries affected by the transition to an open market system;[Footnote 53] * represented a range of contract dollar values and type of work performed (i.e., construction, security, or services); and: * were certified as small businesses by the Small Business Administration. We excluded contracts that were valued less than $25,000 in some of our analyses because we judged those to be less likely to include a DBA requirement because of the nature of the work performed.[Footnote 54] Because the most recent FPDS-NG data are from fiscal year 2013, we developed a list of contractors who were newly awarded State contracts in fiscal year 2013 and then searched FPDS-NG data from fiscal years 2010 through 2013 to determine whether they matched the criteria above.[Footnote 55] From the resulting narrow pool of contractors, we judgmentally selected 8 small businesses to interview. Since DBA premium rates varied in State's single insurer program according to one of four rate categories, we selected contractors in different rate categories to approximate the representation of service types and contract sizes among all State contractors.[Footnote 56] Because we intended to use these interviews to better understand the complexity and details of how small businesses experienced the transition and to uncover any details potentially not revealed in our analysis of federal procurement data--not to draw generalizable conclusions about all of the small businesses that contracted with State--we determined that eight interviews would be sufficient. To gain an additional perspective and supplement the testimony of the 8 small businesses, we interviewed 2 large contractors that work regularly with small subcontractors. We selected these 2, who meet the first 4 criteria listed above, because their responses to our request for premium rate data indicated they had insight into how the transition may have affected small businesses. The interviews, which took place in person or by telephone conference, used a standard set of questions, containing both open-and close-ended questions that allowed contractors to share details of their experiences both before and after State's transition and allowed us to compare their testimony. In all 10 interviews, we spoke with a representative whose job duties required knowledge of DBA insurance requirements (for example, a chief operating officer, accountant, or contract manager). Because this is a nongeneralizable sample, results cannot be used to make inferences about the entire population of small businesses that contracted with State during the period covered by our review. [End of section] Appendix II: Technical Appendix for Analysis of Department of State Contractors' Defense Base Act Insurance Premium Rates: To examine our second objective, the impact of the Department of State's (State) transition to the open market system on premiums paid by State's contractors, we collected data on contractors' Defense Base Act (DBA) insurance premiums, rates, and payroll during State's single insurer program and in the open market. Data on premiums, rates, and payroll were collected from a simple random sample of contractors in Afghanistan, Iraq, Pakistan, South Africa, and Thailand. We identified the set of countries by analyzing country summary reports of contract actions for State and the U.S. Agency for International Development (USAID) in the Federal Procurement Database System-Next Generation (FPDS-NG) for fiscal years 2010 through 2013, identifying the top 10 countries for each agency in terms of number of contract actions and choosing the countries that appeared in the top 10 list for both agencies. We identified the population of contractors in each country by obtaining from State a full list of State contractors whose principal place of performance was outside the United States. We also requested data on the contractors' country of location, the contracting officers' determination of whether the business was classified as a small business, and the contractors' contact information. We restricted our analysis to contracts that were procured through State's Office of Acquisitions Management. We chose a random sample of contractors in the 5 countries identified above and inquired about the DBA premiums, payroll, and premium rates during the single insurer program and in the open market system. When the task order was from an indefinite delivery/indefinite quantity contract and the contractor submitted DBA information for the overall indefinite delivery/indefinite quantity contract, we applied the DBA data to the contractor in that country. To assess the reliability of the FPDS-NG data, we interviewed knowledgeable officials and reviewed publicly available documentation regarding the collection and use of the data. We determined that the data were sufficiently reliable for the purposes of this report. Contractor data on DBA premiums, premium rates, and payroll were self-reported; however, we reviewed the data provided, performed logical tests on them, conducted some follow-up with the contractors and determined the data were sufficiently reliable for our analysis. Summary Statistics: Side bar: Definitions of Insurance-Related Terms: Term: Loss; Definition: The dollar amount basis of a claim for damages under the terms of an insurance policy. From the claimant's perspective, a loss is the monetary benefit the claimant is entitled to receive; Term: Premium rate; Definition: The price of insurance, typically stated in dollars per $100 of covered payroll; Term: Minimum premium; Definition: The minimum dollar amount necessary to receive coverage under an insurance policy; Term: Total premium; Definition: Equal to the larger of (1) the premium rate multiplied by payroll or (2) the minimum premium; Term: Effective premium rate; Definition: Total premium divided by $100 of payroll. Due to the minimum premium, the effective premium rate may be higher than the premium rate. Source: GAO analysis of various DBA-related documents. GAO-15-194. [End of side bar] Summary statistics of the effective premium rates for contractors in our sample appear in table 5. To account for the existence of minimum DBA premiums, as well as the possibility that employees performing different types of work can be covered by the same DBA policy in the open market, we computed effective premium rates paid by each contractor using the following formula: Effective premium rate is equal to: Total DBA premium divided by Payroll in hundreds of dollars. In total, 46 contractors reported their total payrolls and total premiums in the single insurer program, and 57 contractors reported their payrolls and premiums in the open market system. 36 out of 81 contractors reported paying premiums and payroll in both the single insurer program and the open market system; 10 contractors reported paying premiums and payroll in the single insurer program only. An additional 21 contractors reported paying premiums and payroll in the open market system only. Table 5: Summary Statistics Related to Department of State's Defense Base Act (DBA) Insurance Effective Premium Rates: Statistic: Number of observations; Single insurer program effective rate: 46; Open market system effective DBA rate: 57; Single insurer program effective rate: (contractors that reported in both periods): 36; Open market system effective DBA rate: (contractors that reported in both periods): 36. Statistic: Minimum; Single insurer program effective rate: $3.60; Open market system effective DBA rate: $1.00; Single insurer program effective rate: (contractors that reported in both periods): $3.60; Open market system effective DBA rate: (contractors that reported in both periods): $1.00. Statistic: Mean; Single insurer program effective rate: $5.46; Open market system effective DBA rate: $9.10; Single insurer program effective rate: (contractors that reported in both periods): $5.57; Open market system effective DBA rate: (contractors that reported in both periods): $9.81. Statistic: Median; Single insurer program effective rate: $4.95; Open market system effective DBA rate: $6.48; Single insurer program effective rate: (contractors that reported in both periods): $5.10; Open market system effective DBA rate: (contractors that reported in both periods): $8.00. Statistic: Maximum; Single insurer program effective rate: $11.18; Open market system effective DBA rate: $49.58; Single insurer program effective rate: (contractors that reported in both periods): $11.18; Open market system effective DBA rate: (contractors that reported in both periods): $49.58. Source: GAO analysis of State data. GAO-15-194. [End of table] In studying the impact of State's transition to the open market on premiums, we compared the effective premium rates paid by State contractors in the single insurer program with the effective rates paid in the open market. We conducted sign tests for median effective rate differences between the single insurer program and the open market system. We report results and p values[Footnote 57] for each test. Tables 6 and 7 below show the results of this comparison for the contractors in our sample, broken out by country and category of labor. In the country comparison, we omit Thailand and South Africa because of the low number of observations in the sample for these countries. Table 6: Change in Effective DBA Rates from a Sample of State Contractors, by Country: Country: Afghanistan; (n = 12); Median single-insurer program effective rate: (A): $5.38; Median open market effective rate: (B): $8.50; Difference: (B-A): $3.12; Median difference statistically significant?: Yes (p < 0.01). Country: Iraq; (n = 14); Median single-insurer program effective rate: (A): $4.18; Median open market effective rate: (B): $6.75; Difference: (B-A): $2.57; Median difference statistically significant?: Yes (p = 0.03). Country: Pakistan; (n = 6); Median single-insurer program effective rate: (A): $5.22; Median open market effective rate: (B): $10.87; Difference: (B-A): $5.65; Median difference statistically significant?: Yes (p = 0.02). Source: GAO analysis of State data. GAO-15-194. Legend: n = Number of observations. p = p-value for one tailed t-test (median effective rate increase > 0). Note: Sample sizes and statistics above reflect contractors who reported total premium and payroll amounts under the Department of State's single insurer program and also in the open market. Single insurer program rates are calculated from the most recent DBA policy purchased by State contractors active in fiscal years 2010 through 2012. Open market rates are calculated from the first full-year, open market DBA policy purchased by the contractor after June 2012. A one sided test (i.e., of whether open market effective rates were higher than single insurer program effective rates) showed that the median effective rate difference in all three countries was significant at the 5 percent statistical significance level. Data for Thailand and South Africa were not reported because of the small sample size in those countries. [End of table] Table 7: Change in Effective Defense Base Act (DBA) Rates from a Sample of State Contractors, by Contractor Category: Category: Services; (n = 15); Median single-insurer program effective rate: (A): $4.00; Median open market effective rate: (B): $4.92; Difference: (B-A): $0.92; Median difference statistically significant?: Yes (p < 0.01). Category: Construction; (n = 13); Median single-insurer program effective rate: (A): $5.50; Median open market effective rate: (B): $7.59; Difference: (B-A): $2.09; Median difference statistically significant?: Yes (p = 0.01). Category: Security; (n = 8); Median single-insurer program effective rate: (A): $8.28; Median open market effective rate: (B): $12.18; Difference: (B-A): $3.90; Median difference statistically significant?: Yes (p = 0.04). Source: GAO analysis of State data. GAO-15-194. Legend: n = Number of observations; p = p-value for one-tailed t-test (median effective rate increase > 0). Note: Premium rates are reported in dollars per $100 of payroll. Sample sizes and statistics above reflect contractors who reported paying premiums during both the single insurer program and the open market. A one-sided test (i.e., of whether open market effective rates were higher than single insurer program effective rates) showed that the median effective rate difference in all three contractor categories was significant at the 5 percent statistical significance level. [End of table] Histograms of DBA Rates: In this section we present a graphical depiction of the distribution of the DBA premiums and effective premiums paid by our sample of State contractors. Histograms (i.e., bar charts that show how frequently data occur in certain intervals) are provided for the overall sample, as well as disaggregated by contractor category. The histograms in figures 6 through 9 show that the distribution of premiums in both the single insurer program and the open market are skewed toward higher values on the right. As a result, we supplemented our analysis in the second findings section of the report with statistical methods that are less sensitive to skewed data. For example, we conducted two different tests for statistical significance: (1) sign tests, which test for differences in medians and (2) Wilcoxon signed ranks tests, which test for differences in overall distributions and calculated confidence intervals for the median effective rates. Because the original sample was not drawn from a stratified random sample, some categories contain relatively fewer observations than others. Figure 6: Sample Distribution of the Department of State's Single Insurer Program Effective Rates: [Refer to PDF for image: vertical bar graph] Effective rates (in dollars): $2.00-$3.99; Frequency: 10. Effective rates (in dollars): $4.00-$5.99; Frequency: 25. Effective rates (in dollars): $6.00-$7.99; Frequency: 4. Effective rates (in dollars): $8.00-$9.99; Frequency: 5. Effective rates (in dollars): $10.00-$11.99; Frequency: 2. Source: GAO (analysis); Department of State (data). GAO-15-194. [End of figure] Figure 7: Sample Distribution of the Department of State's Single Insurer Program Effective Rates, by Contractor Category: [Refer to PDF for image: 3 vertical bar graphs] Construction: Effective rates (in dollars): $2.00-$3.99; Frequency: 0. Effective rates (in dollars): $4.00-$5.99; Frequency: 15. Effective rates (in dollars): $6.00-$7.99; Frequency: 0. Effective rates (in dollars): $8.00-$9.99; Frequency: 0. Effective rates (in dollars): $10.00-$11.99; Frequency: 0. Security: Effective rates (in dollars): $2.00-$3.99; Frequency: 0. Effective rates (in dollars): $4.00-$5.99; Frequency: 0. Effective rates (in dollars): $6.00-$7.99; Frequency: 3. Effective rates (in dollars): $8.00-$9.99; Frequency: 5. Effective rates (in dollars): $10.00-$11.99; Frequency: 1. Services: Effective rates (in dollars): $2.00-$3.99; Frequency: 10. Effective rates (in dollars): $4.00-$5.99; Frequency: 10. Effective rates (in dollars): $6.00-$7.99; Frequency: 1. Effective rates (in dollars): $8.00-$9.99; Frequency: 0. Effective rates (in dollars): $10.00-$11.99; Frequency: 1. Source: GAO (analysis); Department of State (data). GAO-15-194. [End of figure] Figure 8: Sample Distribution of the Department of State's Open Market Effective Rates: [Refer to PDF for image: vertical bar graph] Open market effective rates (in dollars): $0-$1.99; Frequency: 1. Open market effective rates (in dollars): $2.00-$3.99; Frequency: 5. Open market effective rates (in dollars): $4.00-$5.99; Frequency: 20. Open market effective rates (in dollars): $6.00-$7.99; Frequency: 6. Open market effective rates (in dollars): $8.00-$9.99; Frequency: 5. Open market effective rates (in dollars): $10.00-$11.99; Frequency: 6. Open market effective rates (in dollars): $12.00-$13.99; Frequency: 7. Open market effective rates (in dollars): $14.00-$15.99; Frequency: 4. Open market effective rates (in dollars): $24.00-$25.99; Frequency: 1. Open market effective rates (in dollars): $30.00-$31.99; Frequency: 1. Open market effective rates (in dollars): $48.00-$49.99; Frequency: 1. Source: GAO (analysis); Department of State (data). GAO-15-194. [End of figure] Figure 9: Sample Distribution of the Department of State's Open Market Effective Rates, by Contractor Category: [Refer to PDF for image: 3 vertical bar graphs] Construction Open market effective rates (in dollars): $0-$1.99; Frequency: 1. Open market effective rates (in dollars): $2.00-$3.99; Frequency: 1. Open market effective rates (in dollars): $4.00-$5.99; Frequency: 3. Open market effective rates (in dollars): $6.00-$7.99; Frequency: 5. Open market effective rates (in dollars): $8.00-$9.99; Frequency: 3. Open market effective rates (in dollars): $10.00-$11.99; Frequency: 1. Open market effective rates (in dollars): $12.00-$13.99; Frequency: 2. Open market effective rates (in dollars): $14.00-$15.99; Frequency: 1. Open market effective rates (in dollars): $48.00-$49.99; Frequency: 1. Security: Open market effective rates (in dollars): $8.00-$9.99; Frequency: 1. Open market effective rates (in dollars): $10.00-$11.99; Frequency: 3. Open market effective rates (in dollars): $12.00-$13.99; Frequency: 4. Services: Open market effective rates (in dollars): $0-$1.99; Frequency: 0. Open market effective rates (in dollars): $2.00-$3.99; Frequency: 4. Open market effective rates (in dollars): $4.00-$5.99; Frequency: 17. Open market effective rates (in dollars): $6.00-$7.99; Frequency: 1. Open market effective rates (in dollars): $8.00-$9.99; Frequency: 1. Open market effective rates (in dollars): $10.00-$11.99; Frequency: 2. Open market effective rates (in dollars): $12.00-$13.99; Frequency: 1. Open market effective rates (in dollars): $14.00-$15.99; Frequency: 3. Open market effective rates (in dollars): $24.00-$25.99; Frequency: 1. Open market effective rates (in dollars): $30.00-$31.99; Frequency: 1. Source: GAO (analysis); Department of State (data). GAO-15-194. [End of figure] Analysis of Hypothetical Scenarios: We analyzed trends in contractors' DBA premiums before and after State's transition to an open market system, and examined how premium rates varied by contractor category, country, and business size. We also compared open market DBA rates to rates that would have prevailed in two hypothetical scenarios in which State continued the single insurer program. The two hypothetical scenarios are computed as follows: 1. In the highest contract rates scenario, State and the single insurer continue the single insurer program, with the insurance carrier charging all contractors the highest rates allowable in the 2008 contract along with a lump sum amount to compensate the insurance carrier for historical excess losses. This lump sum amount, which is the same for all contractors, represents a retroactive adjustment of approximately $27 million (divided equally amongst overseas contractors), paid to the single insurer as compensation for performing the second option year of the 2008 DBA contract at the lowest tier of prices when, according to the insurer, the highest tier of rates should have been charged. 2. In the renegotiated rates scenario, State and the single insurer continue the single insurer program, but renegotiate premium rates. The single insurance provider charges security contractors the higher, renegotiated security rate. The new rate is the average of the higher rate requested by the single insurer, and a rate proposed by State as its original negotiating position. All other contractors pay the same rate they were paying in the third option year. All contractors, regardless of contractor category, pay a lump sum to compensate the insurance carrier for historical excess losses. The lump sum represents the retroactive adjustment described in the scenario above. To determine whether premium rates in the hypothetical scenarios described in the report differed from open market premium rates, we conducted statistical tests of the difference between open market and hypothetical rates, by contractor category. We conducted two different tests: (1) sign tests, which test for differences in medians, and (2) Wilcoxon signed ranks tests, which test for differences in overall distributions. In table 8, we report results and p-values[Footnote 58] for each test. With the exception of rates for security contractors, we were unable to detect a robust, statistically significant difference between open market effective DBA rates and effective rates under the hypothetical scenarios. Table 8: Results of Tests for Differences between Open Market and Hypothetical Single Insurer Effective Rates: Hypothetical scenario: Highest contract rates; Contractor group: Overall sample; (n = 56); Sign test: 2-sided p < 0.01. 1-sided (actual