United States Government Accountability Office GAO Report to Congressional Requesters September 2012 DEPARTMENT OF HOMELAND SECURITY Taking Further Action to Better Determine Causes of Morale Problems Would Assist in Targeting Action Plans GAO-12-940 September 2012 DEPARTMENT OF HOMELAND SECURITY Taking Further Action to Better Determine Causes of Morale Problems Would Assist in Targeting Action Plans Highlights of GAO-12-940, a report to congressional requesters Why GAO Did This Study What GAO Found DHS is the third largest cabinet-level Department of Homeland Security (DHS) employees reported having lower department in the federal government, average morale than the average for the rest of the federal government, but employing more than 200,000 staff in a morale varied across components and employee groups within the department. broad range of jobs. Since it began Data from the 2011 Office of Personnel Management (OPM) Federal Employee operations in 2003, DHS employees Viewpoint Survey (FEVS)—a tool that measures employees’ perceptions of have reported having low job whether and to what extent conditions characterizing successful organizations satisfaction. DHS employee concerns are present in their agencies—showed that DHS employees had 4.5 percentage about job satisfaction are one example points lower job satisfaction and 7.0 percentage points lower engagement in their of the challenges the department faces work overall. Engagement is the extent to which employees are immersed in their implementing its missions. GAO has work and spending extra effort on job performance. Moreover, within most designated the implementation and transformation of DHS as a high risk demographic groups available for comparison, DHS employees scored lower on area, including its management of average satisfaction and engagement than the average for the rest of the federal human capital, because it represents government. For example, within most pay categories DHS employees reported an enormous and complex undertaking lower satisfaction and engagement than non-DHS employees in the same pay that will require time to achieve in an groups. Levels of satisfaction and engagement varied across components, with effective and efficient manner. GAO some components reporting scores above the non-DHS averages. Several was asked to examine: (1) how DHS’s components with lower morale, such as Transportation Security Administration employee morale compared with that (TSA) and Immigration and Customs Enforcement (ICE), made up a substantial of other federal employees, and (2) the share of FEVS respondents at DHS, and accounted for a significant portion of extent to which DHS and selected the overall difference between the department and other agencies. In addition, components have determined the root components that were created with the department or shortly thereafter tended to causes of employee morale, and have lower morale than components that previously existed. Job satisfaction and developed action plans to improve engagement varied within components as well. For example, employees in TSA’s morale. To address these objectives, Federal Security Director staff reported higher satisfaction (by 13 percentage GAO analyzed survey evaluations, points) and engagement (by 14 percentage points) than TSA’s airport security focus group reports, and DHS and screeners. component action planning documents, and interviewed officials from DHS and DHS has taken steps to determine the root causes of employee morale problems four components, selected based on and implemented corrective actions, but it could strengthen its survey analyses workforce size, among other things. and metrics for action plan success. To understand morale problems, DHS and selected components took steps, such as implementing an exit survey and What GAO Recommends routinely analyzing FEVS results. Components GAO selected for review—ICE, GAO recommends that DHS examine TSA, the Coast Guard, and Customs and Border Protection—conducted varying its root cause analysis efforts and add levels of analyses regarding the root causes of morale to understand leading the following, where absent: issues that may relate to morale. DHS and the selected components planned comparisons of demographic groups, actions to improve FEVS scores based on analyses of survey results, but GAO benchmarking, and linkage of root found that these efforts could be enhanced. Specifically, 2011 DHS-wide survey cause findings to action plans; and analyses did not include evaluations of demographic group differences on establish clear and measurable metrics morale-related issues, the Coast Guard did not perform benchmarking analyses, of action plan success. DHS concurred and it was not evident from documentation the extent to which DHS and its with our recommendations. components used root cause analyses in their action planning. Without these elements, DHS risks not being able to address the underlying concerns of its varied employee population. In addition, GAO found that despite having broad performance metrics in place to track and assess DHS employee morale on an agency-wide level, DHS does not have specific metrics within the action plans View GAO-12-940. For more information, that are consistently clear and measurable. As a result, DHS’s ability to assess contact David C. Maurer at (202) 512-9627 or its efforts to address employee morale problems and determine if changes firstname.lastname@example.org. should be made to ensure progress toward achieving its goals is limited. United States Government Accountability Office Contents Letter 1 Background 5 DHS Employees Reported Lower Morale than the Rest of the Federal Government, but Morale Varied across DHS Components and Employee Groups 7 DHS Took Steps to Determine Root Causes of Morale Problems and Implemented Corrective Actions, but Could Strengthen Its Efforts 17 Conclusion 34 Recommendations for Executive Action 35 Agency Comments and Our Evaluation 35 Appendix I Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies 37 Appendix II Scope and Methodology 58 Appendix III DHS and Selected Component Steps Taken to Determine Root Causes of Morale Problems 61 Appendix IV Selected Components’ Data Sources for Evaluating Morale, Other than the Federal Employee Viewpoint Survey 68 Appendix V Comments from the Department of Homeland Security 71 Appendix VI Comments from the U.S. Office of Personnel Management 73 Appendix VII GAO Contact and Staff Acknowledgments 74 Page i GAO-12-940 DHS Employee Morale Tables Table 1: 2011 FEVS Job Satisfaction and Engagement Scores by DHS Component (Sorted by Job Satisfaction Index Score) 11 Table 2: DHS-wide and Component Action Plan Goals and Examples of Low-Scoring FEVS Topics Addressed through such Goals 25 Table 3: DHS-Wide and Selected Components’ Action Planning Steps 29 Table 4: Examples of DHS-Wide and Selected Components’ Measures of Success 31 Table 5: DHS and Non-DHS Employee Engagement Index by Demographic Group for the 2011 FEVS 41 Table 6: Model Estimates of Employee Engagement Index at DHS and Other Agencies Using the 2011 FEVS 44 Table 7: OPM Employee Morale Index by DHS Component and Offices Using the 2011 FEVS 52 Table 8: Morale at Preexisting and Recently Created Components of DHS Using the 2011 FEVS 55 Table 9: Model Estimates of the Difference in Engagement and Job Satisfaction between Employees in DHS Components and Non-DHS Agencies Based on the 2011 FEVS 56 Figures Figure 1: Percentage of Satisfied DHS Employees Compared with Governmentwide Averages, 2006, 2008, 2010, and 2011 8 Figure 2: OPM Job Satisfaction and Employee Engagement Indexes in the 2011 FEVS, for Selected Categories of DHS and Non-DHS Employees 10 Figure 3: Satisfaction and Engagement by TSA Employee Group, 2011 13 Figure 4: The Extent to Which OCHCO, TSA, CBP, ICE, and the Coast Guard Incorporated Recommended Factors in Analyzing 2011 FEVS Results 20 Figure 5: OPM’s Six Steps for Action Planning to Improve FEVS Scores 28 Figure 6: Engagement Index Scores by Supervisory Status and Tenure, for DHS and Non-DHS Employees 47 Figure 7: Satisfaction Index Scores by Supervisory Status and Tenure, for DHS and Non-DHS Employees 49 Page ii GAO-12-940 DHS Employee Morale Figure 8: DHS’s 2011 Component Comparison Based on Four HCAAF Indexes 63 Figure 9: DHS HCAAF Scores since 2006 64 Abbreviations AES Annual Employee Survey CBP U.S. Customs and Border Protection CIS U.S. Citizenship and Immigration Services DHS Department of Homeland Security EE employee engagement EEESC Employee Engagement Executive Steering Committee FEMA Federal Emergency Management Agency FEVS Federal Employee Viewpoint Survey FLETC Federal Law Enforcement Training Center FOCS Federal Organizational Climate Survey GS General Schedule HCAAF Human Capital Assessment and Accountability Framework I&A Intelligence and Analysis ICE Immigration and Customs Enforcement JS job satisfaction MGMT Management Directorate MVP Most Valuable Perspective NPPD National Protection and Programs Directorate OAS Organizational Assessment Survey OCHCO Office of the Chief Human Capital Officer OIG Office of the Inspector General OPM Office of Personnel Management OS Office of the Secretary S&T Science and Technology TSA Transportation Security Administration USSS U.S. Secret Service This is a work of the U.S. government and is not subject to copyright protection in the United States. The published product may be reproduced and distributed in its entirety without further permission from GAO. However, 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. Page iii GAO-12-940 DHS Employee Morale United States Government Accountability Office Washington, DC 20548 September 28, 2012 The Honorable Susan M. Collins Ranking Member Committee on Homeland Security and Governmental Affairs United States Senate The Honorable Michael T. McCaul Chairman The Honorable William R. Keating Ranking Member Committee on Homeland Security, Subcommittee on Oversight, Investigations, and Management House of Representatives The Department of Homeland Security (DHS) is the third largest cabinet- level department in the federal government, employing more than 200,000 staff in a broad range of jobs, including aviation and border security, emergency response, cybersecurity analysis, and chemical facility inspection. The DHS workforce is situated throughout the nation, carrying out activities in support of DHS’s mission to (1) prevent terrorism and enhance security, (2) secure and manage the nation’s borders, (3) enforce and administer immigration laws, (4) safeguard and secure cyberspace, (5) ensure resilience from disasters, and (6) provide essential support to national and economic security. Since it began operations in 2003, DHS employees have reported having low job satisfaction. In 2011, for example, DHS’s scores on the Office of Personnel Management (OPM) Federal Employee Viewpoint Survey (FEVS)—a tool that measures employees’ perceptions of whether and to what extent conditions characterizing successful organizations are present in their agency—and the Partnership for Public Service’s (the Page 1 GAO-12-940 DHS Employee Morale Partnership) rankings of the Best Places to Work in the federal government, were generally low. 1 DHS employee concerns about job satisfaction are one example of the challenges the department faces in implementing its missions. In January 2003, we designated the implementation and transformation of DHS as high risk, including its management of human capital, because it represented an enormous and complex undertaking that would require time to achieve in an effective and efficient manner, and it has remained on our high-risk list since that time. 2 Improving human capital management is a DHS priority, reflected through several DHS-wide strategy documents. In June 2012, DHS provided us with its updated Integrated Strategy for High Risk Management (Integrated Strategy), which identified activities to improve employee job satisfaction scores, among other things. In addition, DHS has issued various other strategies and plans for its human capital activities and functions, such as a human capital strategic plan for fiscal years 2009 through 2013, 3 and a workforce strategy for fiscal years 2011 through 2016, which contains the 1 OPM, the central human resources agency for the federal government, has conducted the FEVS every year since 2010. Prior to 2010, OPM conducted the survey during even numbered years, beginning in 2004. The most recent survey sample of 2011 included employees from 29 major federal agencies, as well as 54 independent federal organizations. The survey results represent a snapshot in time of the perceptions of the federal workforce. In 2012 the FEVS will be implemented as a census, rather than a sample-based survey, in an effort to gather the opinions of the entire federal workforce. According to the Partnership, the Best Places to Work ranking is based on employee responses to the following three FEVS assessment items: (1) I recommend my organization as a good place to work. (2) Considering everything, how satisfied are you with your job? (3) Considering everything, how satisfied are you with your organization? 2 In determining whether a government program is high risk, we consider whether it involves national significance, a management function that is key to performance and accountability, or whether there is an inherent or systematic problem, among other things. Our prior work has identified four high-risk areas for which DHS has primary or significant responsibilities: (1) Implementing and Transforming DHS, (2) The National Flood Insurance Program, (3) Protecting the Federal Government’s Information Systems and the Nation’s Critical Infrastructure, and (4) Establishing Effective Mechanisms for Sharing Terrorism-Related Information to Protect the Homeland. GAO, Department of Homeland Security: Progress Made in Implementation and Transformation of Management Functions, but More Work Remains, GAO-10-911T (Washington, D.C.: Sept. 30, 2010). 3 DHS, Human Capital Strategic Plan, Fiscal Years 2009-2013 (Washington, D.C.). Page 2 GAO-12-940 DHS Employee Morale department’s workforce goals, objectives, and performance measures for human capital management. 4 We have previously reported that successful organizations empower and involve their employees to gain insights about operations from a frontline perspective, increase their understanding and acceptance of organizational goals and objectives, and improve motivation and morale. 5 In light of the critical nature of DHS’s mission to protect the security and economy of our nation and the importance of attracting and retaining engaged and satisfied DHS employees to perform its work, you asked us to assess DHS’s efforts to address employee morale. 6 Thus, this report addresses the following questions: • How does DHS’s employee morale compare with that of other federal government employees? • To what extent have DHS and selected components determined the root causes of employee morale and developed action plans to improve morale? To address these questions, we analyzed survey evaluations for the 2011 FEVS, focus group reports from 2007, and DHS and selected component 2011 action planning documents and compared the documents with OPM and Partnership guidance. We also interviewed officials from DHS’s Office of the Chief Human Capital Officer (OCHCO), and human capital officials from four components—U.S. Customs and Border Protection (CBP), U.S. Immigration and Customs Enforcement (ICE), Transportation Security Administration (TSA), and U.S. Coast Guard (Coast Guard). We selected these four DHS components based on their workforce size and how their 2011 Job Satisfaction and Engagement Index scores compared 4 DHS, Workforce Strategy for Fiscal Year 2011-2016 (Washington, D.C.). 5 GAO, High-Risk Series: Strategic Human Capital Management, GAO-03-120 (Washington, D.C.: January 2003). 6 For the purposes of this report, we define employee morale as being characterized by job satisfaction and employee engagement, both of which are measured in OPM’S FEVS. The job satisfaction index, composed of seven FEVS questions such as “my work gives me a feeling of personal accomplishment,” indicates the extent to which employees are satisfied with their jobs and various aspects thereof. The Engagement Index, composed of 15 FEVS questions, indicates the extent to which employees are immersed in the content of the job and energized to spend extra effort in job performance. Page 3 GAO-12-940 DHS Employee Morale with the non-DHS average. 7 The components selected had scores both above, below, and similar to the average. In addition, we interviewed representatives of employee groups within the four selected components to gather employee perspectives on drivers of morale. Details of the selected component index scores, and their statistical significance, are reported in appendix I. To compare DHS’s employee morale with that of other federal government employees, we analyzed the 2011 FEVS results and reviewed OPM survey results issued since 2004, the first full year in which survey data are available. During the course of our analysis, we interviewed knowledgeable agency officials, reviewed relevant documentation, tested data for errors, and determined that the FEVS data are sufficiently reliable for the purposes of this report. As part of this analysis, we compared 2011 DHS and non-DHS job satisfaction and engagement score results by several categories of employees, such as supervisory status, pay grade, and age. We also compared satisfaction and engagement scores within the selected components by employee group, where possible. For example, within TSA, we compared satisfaction and engagement scores reported by Transportation Security Officers, Federal Security Director staff, headquarters staff, and Federal Air Marshals. To determine the extent to which DHS and the selected components identified the root causes of employee morale and developed action plans for improvements, we reviewed agency analysis results, interviewed agency human capital officials and representatives of employee groups, and evaluated action plans for improving morale. We also compared DHS and selected components’ morale root cause analyses and related action plans with available guidance for such efforts. We conducted this performance audit from October 2011 through September 2012 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 that the evidence obtained provides a reasonable basis for our 7 Throughout this report, non-DHS refers to all federal employee FEVS responses outside of DHS. Page 4 GAO-12-940 DHS Employee Morale findings and conclusions based on our audit objectives. Appendix II contains more detailed information about our scope and methodology. Federal employees are routinely surveyed through OPM’s administration Background of the FEVS, which is administered to collect data on federal employees’ perceptions about how effectively agencies are managing their workforces. The FEVS is a tool that measures employees’ perceptions of whether, and to what extent, conditions that characterize successful organizations are present in their agencies, according to OPM. 8 This survey was administered for the first time in 2002 and then repeated in 2004, 2006, 2008, 2010, 2011, and April through June 2012. 9 The survey provides general indicators of how well the federal government is managing its human resources management systems. It also serves as a tool for OPM to assess individual agencies and their progress on strategic management of human capital, and gives senior managers employee perspectives on agency management. Specifically, the survey includes categories of questions asking employees for their perspectives on their work experience, work unit, agency, supervisor, leadership, and satisfaction. OPM intends for agency managers to use the findings to develop policies and action plans for improving agency performance. In 2011, OPM provided a summary of FEVS findings to DHS. In that report, OPM summarized DHS’s survey results relative to governmentwide averages and provided positive and negative response levels for each survey question. Also included in the report was action planning guidance for using FEVS results to improve human capital management. To guide governmentwide efforts to support agency mission results with human capital strategies and in response to the Chief Human Capital Officers Act of 2002, 10 OPM created the Human Capital Assessment and Accountability Framework (HCAAF). Agencies are evaluated by OPM on their progress in meeting HCAAF standards in areas such as talent 8 OPM, 2011 Federal Employee Viewpoint Survey, Empowering Employees, Inspiring Change, Department of Homeland Security, Agency Management Report. (Washington, D.C.). 9 Data for the 2012 FEVS will not be available until November 2012 and thus are not relied upon for this report. 10 Pub. L. No. 107-295, § 1304, 116 Stat. 2315, 2289 (2002) (codified at 5 U.S.C. § 1103(c)). Page 5 GAO-12-940 DHS Employee Morale management, which is focused on agencies having quality people with the appropriate competencies in mission-critical activities. The FEVS job satisfaction index is one of the metrics used by OPM to assess whether agencies are effectively managing the talent management system. The FEVS provides one source of information for evaluating success on other HCAAF standards as well by measuring responses to groups of FEVS questions for four indices. The four index measures are: Leadership and Knowledge Management; Results-Oriented Performance Culture; Talent Management; and Job Satisfaction. In addition, in 2011, OPM added an index to measure employee engagement, which OPM defines as the extent to which an employee is immersed in the content of the job and energized to spend extra effort in job performance. DHS’s OCHCO is responsible for implementing policies and programs to recruit, hire, train and retain DHS’s workforce. As the department-wide unit responsible for human capital issues within DHS, OCHCO provides OPM with a DHS-wide action plan every other year, with the next plan due in January 2013. OCHCO also provides guidance and oversight to the DHS components related to morale issues. For example, OCHCO provides a survey analysis and action planning tool that the components must use in response to FEVS results to develop action plans for improving employees’ positive scores. 11 These plans are to state objectives and identify actions to be taken in response to survey results. OCHCO also has provided oversight by reviewing and providing feedback on component action plans. 11 OCHCO’s authority for requiring components to use the action planning tool is based on DHS’s Human Capital Line of Business Integration and Management Directive, issued in 2004. Page 6 GAO-12-940 DHS Employee Morale Data from the 2011 FEVS show that DHS employees have lower average DHS Employees levels of job satisfaction and engagement overall and across most Reported Lower demographic groups available for comparison, such as pay grade, when compared with the average for the rest of the federal government. Levels Morale than the Rest of satisfaction and engagement vary across components, with some of the Federal components reporting satisfaction or engagement above the average for Government, but the rest of the government. Similarly, these measures of morale vary within components as well, with some employee groups reporting higher Morale Varied across morale than other groups within the same component. DHS Components and Employee Groups DHS Employees as a As shown in figure 1, DHS employees generally reported improvements Whole Reported Lower in job satisfaction index levels since 2006 that narrowed the gap between Satisfaction and DHS and the governmentwide average. 12 However, employees continue to indicate less satisfaction than the governmentwide average. 13 Engagement According to Partnership analysis of FEVS data also indicates consistent levels of low Several Measures employee satisfaction relative to other federal agencies. Similar to its 2011 ranking, 31st of 33 federal agencies, the Partnership ranked DHS 28th of 32 in 2010, 28th of 30 in 2009, and 29th of 30 in 2007 in the Best Places to Work ranking on overall scores for employee satisfaction and commitment. 14 12 Two thousand six is the first year in which job satisfaction index data were made available and can be compared between DHS and the rest of the federal government. 13 For the purposes of our report, we list governmentwide averages in some instances, which include DHS. In other instances where we were able to make statistical adjustments, we report non-DHS averages, which exclude DHS. 14 Partnership for Public Service and the Institute for the Study of Public Policy Implementation at the American University School of Public Affairs, The Best Places to Work in the Federal Government. The Partnership for Public Service’s ranking cited here is composed of large agencies, defined as agencies with more than 2,000 full-time permanent employees. Page 7 GAO-12-940 DHS Employee Morale Figure 1: Percentage of Satisfied DHS Employees Compared with Governmentwide Averages, 2006, 2008, 2010, and 2011 Note: Because the FEVS was not administered each year, the job satisfaction index and DHS versus governmentwide averages are available only for 2006, 2008, 2010, and 2011. Our analyses of 2011 FEVS results also indicate that average DHS-wide employee satisfaction and engagement scores were consistently lower when compared with average non-DHS employee scores in the same demographic groups. As shown in figure 2, comparisons of DHS with non-DHS employees by supervisory status, pay group, and tenure indicate that satisfaction and engagement are lower across many of the DHS groups where statistically significant differences are evident. 15 For example, across pay categories DHS satisfaction and engagement were lower than the scores for the same non-DHS employee pay groups, with the exception of senior executives, senior leaders, employees with less 15 Because statistical significance is a function of two things—the size of the difference and the size of the sampled groups being compared—the biggest differences are not always the differences that are significant. Page 8 GAO-12-940 DHS Employee Morale than 1 year of tenure, and General Schedule pay grades 1-6. 16 Similarly, job satisfaction and engagement scores for DHS management and non- management employees were lower than for the same non-DHS employee groups. 16 The differences between DHS and non-DHS senior leader, General Schedule 1-6, and less than 1 year of tenure satisfaction and engagement were not statistically significant. Page 9 GAO-12-940 DHS Employee Morale Figure 2: OPM Job Satisfaction and Employee Engagement Indexes in the 2011 FEVS, for Selected Categories of DHS and Non-DHS Employees Note: Estimates of job satisfaction and employee engagement have a margin of error at the 95 percent confidence level of no more than plus or minus 5.1 percentage points, except for employees in the Senior Leader pay group. Because statistical significance is a function of two things—the size of the difference and the size of the sampled groups being compared—the biggest differences are not always the differences that are significant. Page 10 GAO-12-940 DHS Employee Morale a The Job Satisfaction index, composed of seven Federal Employee Viewpoint Survey (FEVS) questions, indicates the extent to which employees are satisfied with their jobs and various aspects thereof. b The Engagement index, composed of 15 FEVS questions, indicates the extent to which employees are immersed in the content of the job and energized to spend extra effort in job performance. c The “Other” pay group may include employees who are not part of the General Schedule system, such as those in pay-band systems. d Denotes statistically significant differences between DHS and non-DHS employees that are distinguishable from zero at the 0.05 level. Employee Satisfaction and Satisfaction and engagement variation across components. As Engagement Vary across shown in table 1, the 2011 FEVS data indicate that satisfaction and and within DHS engagement levels vary across DHS components. For example, TSA is 11.6 percentage points below the non-DHS average on the job Components satisfaction index, but other large components, such as CBP and the Coast Guard, are 1 and 1.5 percentage points more satisfied than the average for the rest of the government, respectively. Three other components—the Inspector General, the Federal Law Enforcement Training Center, and the U.S. Secret Service—also score above the non- DHS engagement averages on the job satisfaction or engagement indexes. Morale varies widely across all components, with the job satisfaction index ranging from 56.9 to 71.6 percent and the employee engagement index ranging from 53.2 to 70.6 percent. Table 1: 2011 FEVS Job Satisfaction and Engagement Scores by DHS Component (Sorted by Job Satisfaction Index Score) Job Employee Satisfaction index Engagement index Federal Law Enforcement Training Center 71.6* 66.2 Inspector General 70.8 70.5* U.S. Coast Guard 70.0 70.6* U.S. Customs and Border Protection 69.5 62.9* U.S. Secret Service 69.4 68.1 Non-DHS government employees 68.5 67.1 U.S. Citizenship and Immigration Service 66.6* 64.0* Under Secretary for Management 65.7* 65.7 DHS, no sub-agency 65.2 60.5* DHS (overall) 64.0* 60.1* Federal Emergency Management Agency 63.2* 58.0* Office of the Secretary 63.0* 63.6* National Policy and Programs Directorate 61.9* 57.7* Page 11 GAO-12-940 DHS Employee Morale Job Employee Satisfaction index Engagement index U.S. Immigration and Customs Enforcement 60.6* 58.0* Science and Technology Directorate 59.7* 55.7* Office of Intelligence and Analysis 57.6* 53.2* Transportation Security Administration 56.9* 54.4* Source: GAO analysis of 2011 FEVS data. Note: Estimates of job satisfaction and employee engagement have a 95 percent margin of error of no more than plus or minus 6.3 percentage points. Asterisks indicate estimates that are distinguishable from the non-DHS estimate at the 0.05 level. Satisfaction and engagement variation within components. Within the selected components reviewed—the Coast Guard, ICE, CBP, and TSA—satisfaction and engagement varied across workgroups as well. For example, as shown in figure 3, responses by TSA employee groups varied widely, with the screening workforce reporting significantly lower satisfaction and engagement scores. 17 17 For the purposes of this report we define employee groups as the different occupational groups that exist within DHS components, such as TSA’s Federal Air Marshals, screeners, and federal security director staff. Page 12 GAO-12-940 DHS Employee Morale Figure 3: Satisfaction and Engagement by TSA Employee Group, 2011 Note: Estimates of job satisfaction and employee engagement have a 95 percent margin of error of no more than plus or minus 4.4 percentage points. Employee group scores showed variability in the three other selected DHS components as well. 18 For example: • CBP. Border Patrol employees were 8 percentage points more satisfied and 12 percentage points more engaged than CBP field operations employees. 19 18 All differences between pairs of groups in this paragraph are distinguishable from zero at the 0.05 level. 19 A border patrol agent is involved in detection, prevention and apprehension of terrorists, undocumented aliens and smugglers of aliens at or near the land border. A CBP field operations officer is responsible for, among other things, determining the nationality and identity of each applicant for admission to the United States and for preventing the entry of ineligible aliens, including criminals, terrorists, and drug traffickers. Page 13 GAO-12-940 DHS Employee Morale • Coast Guard. Satisfaction and engagement levels exceeded the non- DHS average for some civilian employee groups, such as those under the Chief of Staff for Mission Support, but some groups were substantially higher. These groups include Districts 13 and 14, which were 18.6 and 10.3 percentage points more satisfied than non-DHS employees, respectively. 20 • ICE. Homeland security investigators and immigration enforcement employees were less satisfied and engaged than many other employee groups in DHS and the average for the rest of the government. 21 Homeland security investigators were 5.5 percentage points lower on satisfaction and 8.2 percentage points lower on engagement than the non-DHS averages. Enforcement and Removal employees were also less satisfied and engaged, with scores 12.7 percentage points below on satisfaction and 14.4 percentage points lower on engagement than the non-DHS averages. In addition to variation in satisfaction and engagement levels across employee groups, representatives of DHS employee groups we interviewed identified a range of issues that may be creating lower satisfaction rates among employees. 22 These examples highlight the variety of issues that can lead to morale problems and may be unique to particular DHS components. 23 For example: • A TSA screener union representative described TSA’s performance assessment system as a key driver of morale problems among passenger screeners. According to the union representative, if a screener fails a portion of the annual examination more than three 20 The Coast Guard is organized into two major commands that are responsible for overall mission performance: the Pacific area that includes District 13 which covers the Pacific Northwest area, and District 14 which covers the Hawaii and Guam region. 21 Homeland security investigators investigate crime, human rights violations and human smuggling, smuggling of narcotics, weapons and other types of contraband, financial crimes, cybercrime, and export enforcement issues while enforcement removal operations employees work to enforce the nation’s immigration laws by ensuring the removal of aliens who pose a threat to national security or public safety. 22 For employee groups that had union representation, we interviewed union representatives who identified employee group perspectives on employee morale. For those groups without union representation, we convened focus groups of employees to discuss employee perspectives on morale. 23 The examples provided by agency officials are the observations of individual employees and are not representative of all employee opinions. Page 14 GAO-12-940 DHS Employee Morale times, the screener will be terminated. The union representative explained that failing three times is possible, even for highly effective screeners who may not be effective under testing conditions because of anxiety about the examination, resulting in a significant burden on the screeners and, therefore, the performance assessment will result in lower morale. • ICE homeland security investigators who participated in a focus group we held cited frustrations with frequent turnover in regional leadership positions, which they stated negatively affects employee morale. Unequal resource allocations across investigative groups were also described as leading to lower morale among investigators. • A union representative for CBP’s field operations employees described staffing shortages at ports of entry, temporary assignments to the southwest border that affect work-life balance, and management resistance to employee telework arrangements, among other things, as resulting in employee morale problems. • Border Patrol union representatives cited uncertainties in overtime pay policy, living conditions at small, temporary shelters for Border Patrol agents deployed in an area, and inflexible employee work scheduling practices, among other things, as creating morale problems among Border Patrol agents. Coast Guard civilian officials who participated in a focus group we held, on the other hand, provided examples of drivers of high levels of morale within the Coast Guard. The officials described a Coast Guard culture of mission focus that has led to high morale among civilian Coast Guard employees. For example, the officials stated that a sense of making a difference in maritime security and safety through work activities such as vessel inspections, contingency planning for natural disasters, and training Coast Guard employees results in employees who are engaged and satisfied with their jobs. The officials also described a Coast Guard cultural focus on team cohesion and shared successes, among both military and civilian Coast Guard personnel, both of which are recognized by Coast Guard leadership, according to the officials. For example, the officials stated that Coast Guard leadership designates a portion of awards for team successes, rather than individual achievement, which the officials we interviewed found more satisfying than individual awards. Page 15 GAO-12-940 DHS Employee Morale A statistical analysis of 2011 FEVS and employee demographic data we conducted suggested several other explanations for differences in morale 24 between DHS and non-DHS agencies: • Several of the DHS components with lower morale, such as TSA and ICE, make up a substantial share of FEVS respondents at DHS, as shown in appendix I, table 7. Those components have more influence on the agency’s overall morale score than smaller components— many of which have higher average morale scores. Consequently, the gap between DHS and the rest of the government in employee morale is driven primarily by the scores of a few large components. • DHS is not more likely than other agencies to employ the types of staff who tend to have lower morale across all agencies, as shown in appendix I, table 6. Instead, employees in the various groups we analyzed had lower morale at DHS than the same types of employees at other agencies. This suggests that the gap may be explained by factors unique to DHS, such as management practices and the nature of the agency’s work, or by differences among employees we could not analyze. 25 • DHS employees who joined the department since its creation tend to have lower morale than employees who joined the department as part of preexisting components, as shown in appendix I, table 8. DHS employees who started working for the agency between 1 and 10 years ago are less engaged than employees with similar tenures and demographics at other agencies. In addition, several of the least engaged components, such as the Intelligence and Analysis and Science and Technology divisions, as shown in appendix I, table 7, were created with the department or subsequently, rather than being added from elsewhere in the federal government. 24 We used measures of job satisfaction and employee engagement as indicators of morale in our analysis. 25 We used a statistical method called “Oaxaca decomposition” to divide the overall difference in morale between DHS and non-DHS agencies into two parts: the part explained by employee characteristics present in the FEVS (i.e. supervisory status, employee tenure, age and location), and the part explained by how those characteristics affect morale. This let us assess whether the morale gap is explained by available employee characteristics or by how those characteristics affect morale differently at DHS because of unique characteristics of the agency such as management practices or program goals. The method cannot identify specific unique characteristics that are responsible, however. Appendix I describes this analysis in more detail. Page 16 GAO-12-940 DHS Employee Morale The variation of factors that can result in morale problems as suggested by these examples, as well as the variation in levels of satisfaction and engagement among employee groups, underscores the importance of looking beyond survey scores to understand where problems, such as low employee satisfaction, are taking place within the organization, and to identify and address the causes of these problems. Appendix I provides comparisons for additional demographic groups, including age, component tenure, and location, and summarizes our statistical analysis that examines the relationships between each of these factors and satisfaction and engagement, holding constant each of the other factors. The appendix also provides a more detailed list of satisfaction and engagement estimates for components and offices, in some cases holding constant demographic differences among employees. DHS and the selected components have taken steps to understand DHS Took Steps to morale problems, such as holding focus groups, implementing an exit Determine Root survey, and routinely analyzing FEVS results. On the basis of FEVS results, DHS and the selected components planned actions to improve Causes of Morale FEVS scores. However, we found that DHS could enhance its survey Problems and analysis and monitoring of action plan results. In addition, according to Implemented DHS’s Integrated Strategy for addressing the implementing and transforming high risk area, DHS has begun implementing activities to Corrective Actions, address morale but has not yet improved DHS’s scores on OPM’s job but Could Strengthen satisfaction index or its ranking on the Partnership’s Best Places to Work in the Federal Government. Its Efforts DHS and Selected DHS’s OCHCO has taken several steps to understand morale problems Components’ Have Taken DHS-wide. Specifically, since 2007, OCHCO: Steps to Understand • Conducted focus groups DHS-wide in 2007 to determine employee Morale Problems concerns related to morale, which identified employee concerns in areas of leadership, communication, empowerment, and resources. • Performed statistical analysis in 2008 to identify workplace factors that drove employee job satisfaction, finding that the DHS mission and supervisor support, among other things, drove employee job satisfaction. • Initiated an exit survey, first administered DHS-wide in 2011, to understand why employees chose to leave their position. The survey Page 17 GAO-12-940 DHS Employee Morale found lack of quality supervision and advancement opportunities were the top reasons for leaving. 26 • Analyzed 2011 FEVS results, among other things, showing where lower scores on HCAAF indices were concentrated among several components—Intelligence and Analysis, TSA, ICE, National Protection and Programs Directorate, and the Federal Emergency Management Agency (FEMA). • Launched an Employee Engagement Executive Steering Committee (EEESC) in January 2012 that will identify action items for improving employee engagement by September 2012, according to OCHCO officials. The selected components also evaluated FEVS results to identify morale problems and considered additional information sources. For example: • TSA convened a corporate action planning team in March 2011, as part of its response to FEVS results, which relied on data sources such as the TSA-administered exit survey, employee advisory groups, and an online employee suggestion tool, to gain perspectives on systemic challenge areas and to develop plans to address morale, according to TSA officials. TSA’s action plan for improving morale, based on these sources, was completed in July 2012. • ICE considered results of a Federal Organizational Climate Survey (FOCS), last completed in March 2012, and held focus groups to gauge the extent to which employees view ICE as having an organizational culture that promotes diversity. • CBP launched a quarterly online employee survey in 2009 to solicit opinions on one specific topic per quarter, such as use of career development resources and how the resources contributed to employees’ professional growth at CBP. • The Coast Guard relied on an Organizational Assessment Survey (OAS), last administered by OPM in 2010, to understand employee morale. The OAS solicits opinions on a range of topics, including job satisfaction, leadership, training, innovation, and use of resources. It included civilian and military Coast Guard personnel, but is not administered governmentwide so comparisons between the Coast 26 Results from DHS’ Exit Survey should be interpreted with caution. Due to the method in which it is administered, the survey’s response rate in 2011 was quite low, close to 40 percent. It is likely that a higher response rate would have produced somewhat different results. Page 18 GAO-12-940 DHS Employee Morale Guard and other federal employees are limited to organizations that may use the OAS, according to Coast Guard officials. Appendix III provides more detailed descriptions of DHS’s steps to address morale problems and selected components’ 2011 FEVS analysis methods and findings. Appendix IV provides additional information on the selected components’ data sources beyond FEVS for evaluating root causes of morale, including a summary of results and how the information was used by the components. DHS and Selected For the 2011 FEVS, DHS and the selected components completed Components Conducted varying levels of analyses to determine the root causes of low morale. Limited FEVS Analyses However, DHS and the selected components conducted limited analysis in several areas that is not consistent with OPM and Partnership guidance that lays out useful factors for evaluating root causes of morale problems through FEVS analysis, as shown in figure 4. Page 19 GAO-12-940 DHS Employee Morale Figure 4: The Extent to Which OCHCO, TSA, CBP, ICE, and the Coast Guard Incorporated Recommended Factors in Analyzing 2011 FEVS Results Note: Demographic comparisons. In its guidance to federal agencies for improving employee job satisfaction, the Partnership for Public Service advises, among other things, determining demographic group differences. According to the Partnership for Public Service, demographic group analysis shows where there may be gaps in satisfaction at an agency or subcomponent (i.e., perhaps some demographic groups report lower satisfaction than others). For the purposes of this report, demographic group is used to describe any common characteristic among employees, such as pay grade, supervisory status, or work group. Benchmarking. Benchmarking agency survey results against those of similar organizations can provide a point of reference for improvements, according to the Partnership for Public Service. OPM’s management report to DHS for the 2011 FEVS also suggests, as part of action planning, comparing agency results with governmentwide results and noting which survey questions and Human Capital Assessment and Accountability Framework indexes scored lowest relative to the governmentwide averages. Page 20 GAO-12-940 DHS Employee Morale Linking analysis results with action plans. Both OPM and the Partnership for Public Service action planning guidance list analyzing survey results as a first step to developing action plans to address employee concerns. According to OPM’s guidance, the data analysis may include reviewing FEVS results and following up on survey findings with focus groups to clarify reasons for low scores. OPM’s guidance then calls for agencies to translate issues uncovered through data analysis into a set of action plan goals. Similarly, the Partnership for Public Service advises agencies to develop action plan approaches to improve employee satisfaction, based on issues the data identify, while considering the organization’s mission, culture, available time, and resources. a CBP partially benchmarked its FEVS results because it compared results with governmentwide and DHS averages, but not with those for similar organizations for the 2011 FEVS. CBP officials stated that few agencies within the United States both use FEVS and have occupations similar to those of CBP. b OCHCO and the selected components partially linked root causes with action plans because low- scoring questions were listed on action plans as the reason why actions were chosen. However, additional root cause analysis findings, such as those listed in appendix IV, were not included in the action plan documentation. Usage of the three factors described in figure 4 varied across DHS-wide and component-level 2011 FEVS analyses we reviewed. In some instances, the factors were partially or not used. For example: • Demographic group comparisons. According to our reviews of OCHCO’s analyses, OCHCO’s DHS-wide analyses did not include evaluations of demographic group differences on morale-related issues for the 2011 FEVS. According to OCHCO officials, DHS’s Office of Civil Rights and Civil Liberties reviews survey results to identify diversity issues that may be reflected in the survey, and OCHCO officials considered these results when developing one of the current (as of August 2012) DHS action plans to create policies that identify barriers to diversity. In 2007 and 2009, years in which DHS administered the Annual Employee Survey (AES), demographic comparisons were made. For example, on the basis of 2009 AES data, DHS found no significant demographic differences other than supervisors’ positive responses to questions were generally higher than those of non-supervisors and differences among pay grade levels. Because OPM now administers the survey each year, DHS is not able to make significant demographic group comparisons because of the format of the data provided by OPM, according to OCHCO officials. However, we obtained FEVS data from OPM that allowed us to make demographic group comparisons. For example, we compared DHS and non-DHS employee satisfaction and engagement scores across available demographic groups and found that both satisfaction and engagement were generally lower for DHS employees, which is summarized in appendix I, table 5. For the DHS component analyses we reviewed, TSA and CBP conducted some demographic analysis. For example, TSA compared screeners, Federal Security Director staff, Federal Air Marshals, and Page 21 GAO-12-940 DHS Employee Morale headquarters staff on each FEVS dimension (e.g., work experiences, supervisor/leader, satisfaction, and work/life). As a result, TSA was able to identify screeners as having survey scores below those of other TSA employee groups. CBP also compared race, ethnicity, gender, and program office scores. CBP found that no significant differences were present in the positive responses to the 2011 FEVS core questions when comparing race, ethnicity and gender, and found that Border Patrol employees reported higher job satisfaction than field operations employees (74 versus 66 percent on the job satisfaction index). In contrast, the Coast Guard did not conduct analysis in addition to data that was provided by DHS OCHCO. Because OCHCO’s data did not include demographic information for the 2011 FEVS, Coast Guard did not make demographic group comparisons. 27 ICE and CBP officials stated that they did not have access to 2011 FEVS data files necessary to conduct more detailed demographic comparisons. However, as shown in appendix I, we were able to make various demographic comparisons based on a more detailed data file provided by OPM, which is similar to a file that OPM makes available to agencies and the public. 28 • Benchmarking against similar organizations. TSA benchmarked its FEVS results against results from similar organizations, by comparing results with CBP, and OCHCO’s DHS-wide analysis highlighted Partnership rankings data, showing DHS’s position relative to the positions of other federal agencies as a Best Place to Work. Similarly, ICE benchmarked its FEVS results overall and for program offices, such as homeland security investigators, against other DHS components, including the U.S. Secret Service and CBP. For the 2011 FEVS, CBP performed more limited benchmarking, by comparing FEVS results with governmentwide averages. According to CBP officials, when analyzing annual employee surveys prior to 2011, CBP benchmarked its results against agencies with high positive 27 In the 2010 OAS report to the Coast Guard, OPM made demographic comparisons among several employee groups within the Coast Guard, including work groups (such as ship versus shore-based employees), race and ethnicity, and gender, and comparisons between civilian and military ranks. However, the results of the OAS analysis are not included in the Coast Guard’s FEVS-based action planning documentation. See appendix IV for a description of how the OAS is used by the Coast Guard. 28 Our file contained the same variables as the public release file, but it identified smaller demographic groups and work units. Page 22 GAO-12-940 DHS Employee Morale FEVS scores, such as the Social Security Administration, the Federal Bureau of Investigation, the Internal Revenue Service, and the Nuclear Regulatory Commission. CBP is in the initial planning phase of a larger benchmarking project that would benchmark CBP against foreign immigration, customs, and agriculture inspection agencies, such as the Canadian Border Services Agency and the Australian Customs and Border Protection Service. If approved, this benchmarking project is expected to occur in fiscal year 2013, according to CBP officials. The Coast Guard did not perform FEVS benchmarking analysis, according to the documentation we reviewed, but did make OAS-based comparisons between the Coast Guard and other organizations that use the OAS, according to Coast Guard officials. • Linkage of root causes with action plans. For both DHS-wide and selected component action plans, FEVS questions with low scores were linked with action plan areas. For example, in the DHS-wide action plan, low scores on employee satisfaction with opportunities to get a better job in the organization were linked to action plan items for enhancing employee retention. However, the extent to which DHS and the components used root causes found through other analyses to inform their action plans, such as quarterly exit survey results or additional internal component surveys, was not evident in action plan documentation (see appendix IV for a description of these additional root cause analyses). For example, • OCHCO’s DHS-wide action plan was last updated based on 2010 FEVS data and therefore did not rely on data from the DHS 2011 exit survey, since those results were not published until January 2012. Similarly, the EEESC was launched in January 2012 and therefore its efforts are not yet documented in DHS-wide action planning documents. According to OCHCO officials, the 2010 DHS-wide action plan includes consideration of results from OCHCO’s 2008 statistical analysis identifying key drivers of job satisfaction and results from the 2007 focus groups. However, linkage to items in the DHS-wide action plan to these results is not clearly identified because a new action plan template OPM introduced in 2010 did not provide an area to identify the linkage between each action and the driver, according to OCHCO officials. In addition, DHS’s September 2009 action plan indicates consideration of the 2008 key driver analysis and 2007 focus group effort that led to a focus on leadership effectiveness initiatives. Page 23 GAO-12-940 DHS Employee Morale • According to CBP and TSA officials, data from other root cause analysis efforts are not explicitly documented in action plans developed in response to FEVS results because DHS has not included linkage of other root cause analysis efforts to actions items in the FEVS action planning templates used by the components. TSA officials also stated that other root cause efforts (see appendix IV) were used to develop TSA’s July 2012 action plan update. However, the July 2012 plan did not include linkage of root cause findings other than FEVS results, such as exit survey results, to action plan items. • ICE officials stated that results from other root cause efforts, such as its FOCS, have not yet been considered in FEVS-based action planning but that ICE plans to do so in future efforts to address morale. • The Coast Guard uses information from its OAS as part of a process separate from FEVS-based action planning for addressing morale, so OAS results are not linked to FEVS-based action plans. OCHCO and component human capital officials described several reasons for the variation in root cause analysis of FEVS results. OCHCO officials described resource constraints and leadership changes within the OCHCO position as resulting in a lack of continuity in root cause analysis efforts. For example, one OCHCO official stated that because of resource constraints, OCHCO has focused more efforts on workforce planning than on morale problem analysis since 2009. ICE human capital officials stated that ICE’s human capital services were provided via a contract with CBP until 2010, when the human capital function became an independently funded part of the ICE organization. Only since moving to its current position within ICE has the human capital office been able to devote more resources to addressing morale issues, according to the officials. CBP human capital officials stated that for assessing morale issues, CBP uses both quantitative and qualitative information. However, according to the officials, qualitative evidence is preferable over quantitative survey analysis because focus groups and open-ended surveys, such as the Most Valuable Perspective online survey, allow CBP to better understand the issues affecting employees. Because of CBP human capital officials’ preference for qualitative information, CBP has not emphasized extensive quantitative analysis of survey results, such as statistical analysis that may determine underlying causes of morale problems. Without a complete understanding of which issues are driving low employee morale, DHS risks not being able to effectively address the Page 24 GAO-12-940 DHS Employee Morale underlying concerns of its varied employee population. Emphasis on survey analysis that includes demographic group comparisons, benchmarking against similar organizations, and linkage of other analysis efforts outside of FEVS within action plan documentation could assist DHS in better addressing its employee morale problems. DHS and Its Components DHS and the selected components routinely update their action plans to Completed Action Plans address employee survey results in accordance with the Office of Management and Budget’s budget guidance; the DHS-wide plan is updated every two years, and components update their plans at least annually. 29 According to OPM’s guide for using FEVS results, action planning involves, among other things, identifying goals and actions for improving low-scoring FEVS satisfaction topics such as reviewing survey results to determine steps to be taken to improve how the agency manages its workforce. DHS-wide and component action plan goals and examples of low-scoring FEVS satisfaction topics are listed in table 2. Table 2: DHS-wide and Component Action Plan Goals and Examples of Low-Scoring FEVS Topics Addressed through such Goals Example of low-scoring FEVS satisfaction topics DHS unit Summary of action plan goals to address FEVS results addressed by action plan goal DHS-wide Enhance leadership, recruitment, employee retention, and DHS unification. Opportunity to get a better job in the organization TSA Launch a corporate action planning team to study employee issues and develop Discussions with supervisors recommendations, enhance employee performance management, and improve about performance TSA communication mechanisms. ICE Advance telework opportunities, increase communication between employees and Physical conditions that allow management, and develop an awards handbook for distribution to employees. employees to do their job well CBP Develop action plans within CBP program offices to address results, enhance Policies and practices of senior communication between management and employees, create career and leaders leadership development opportunities, replace pass/fail performance appraisal with multi-leveled performance management system, implement training improvements, and maintain an existing virtual focus group to enable upward feedback to senior leaders. Coast Guard Improve communication with employees and training options. Information received from management Source: GAO analysis of DHS-wide, TSA, Coast Guard, CBP, and ICE FEVS action plans based on FEVS results. 29 The selected component action plans we reviewed were updated as of January 2012. Subsequent updates to the plans, due to OCHCO in July 2012, were not included in our evaluation. Page 25 GAO-12-940 DHS Employee Morale As part of DHS’s efforts to address our high-risk designation of implementing and transforming DHS, DHS described a plan for improving employee morale in its Integrated Strategy for High Risk Management (Integrated Strategy). In June 2012, DHS provided us with its updated Integrated Strategy, which summarized the status of the department’s activities for addressing its implementation and transformation high-risk designation. In the Integrated Strategy, DHS identified activities to improve employee job satisfaction scores, among other things. The status of the activities included ongoing analysis of the 2011 FEVS results, launch of the EEESC to address DHS scores on the HCAAF indexes, ongoing coordination between the OCHCO and components to develop action plans in response to the 2011 FEVS results, and launch of an online employee survey in the first quarter of fiscal year 2013. Within the Integrated Strategy action plan for improving job satisfaction scores, DHS reported that three of six efforts were hindered by a lack of resources. 30 For example, resources are a constraining factor for DHS’s Office of the Chief Human Capital Officer to consult with components in developing action plans in response to 2011 FEVS results. Similarly, resources are a constraining factor to deploy online focus discussions on job satisfaction-related issues. 30 The six efforts are: (1) OPM releases 2011 survey results; DHS issues employee communications and conducts analysis; (2) CHCO leads an Enterprise-wide Executive Steering Committee to develop a deliberate way forward for addressing FEVS key indices; (3) OCHCO works with components to develop action plans responding to 2011 FEVS results; (4) On-line focus discussion survey deploys across Department; (5) CHCO provides components with feedback on action plans; and (6) Updated component and headquarter action plans are due. The three efforts that are hindered by a lack of resources are: OCHCO working with components to develop action plans, on-line focus discussions, and CHCO providing components with feedback. Page 26 GAO-12-940 DHS Employee Morale DHS and Selected According to our review of the action plans created in response to the Components Generally FEVS and interviews with agency officials, DHS and the selected Followed OPM’s Six Steps components generally incorporated the six action planning steps suggested by OPM, but the agency does not have effective metrics to for Effective Action support its efforts related to monitoring. 31 (See figure 5.) Planning but Do Not Have Effective Metrics for Monitoring Efforts 31 The most recent DHS-wide action plan update was based on 2010 FEVS results. According to OCHCO officials, the next update to the action plan will be released in January 2013. Page 27 GAO-12-940 DHS Employee Morale Figure 5: OPM’s Six Steps for Action Planning to Improve FEVS Scores We found that, in general, DHS and its components are implementing the six steps for action planning as demonstrated in table 3 below. Page 28 GAO-12-940 DHS Employee Morale Table 3: DHS-Wide and Selected Components’ Action Planning Steps Action Planning Step DHS and Component Response Examples of What Has Been Done 1. Identify issues All the components and OCHCO For example, CBP’s action plan identified the FEVS questions that evaluated the FEVS results by, at a pertained to the job satisfaction dimension along with the positive minimum, determining high and low response results for each of the survey questions. positive response levels to survey questions. 2. Set goals All of the components and OCHCO For example, TSA’s action plan included a goal of enhancing developed broad goals to improve FEVS overall communication and innovation. scores. 3. Identify staff All of the components and OCHCO For example, TSA’s action plan described a cross-functional team identified personnel responsible for their that included representatives from existing councils and from all action plan. levels and organizational units to address an action item. 4. Develop plan All of the components and OCHCO For example, ICE’s action plan for increasing communication developed plans that identified actions to included utilizing ICE broadcast announcements for more tangible be taken to accomplish their goals. and employee-related purposes. 5. Implement plan All of the components and OCHCO set For example, Coast Guard’s action plan listed the redesign of the target dates and in some cases indicated civilian website by December 20, 2012. whether an action item was completed or on-going. 6. Monitor results All of the components and OCHCO For example, OCHCO’s action plan included a goal to retain an established measures of success, engaged workforce. The three measures of success listed for this providing the status of success or goal are (1) improve DHS ranking on the Partnership for Public completion dates. Service Best Places to Work, (2) reduce attrition rates, and (3) achieve an HCAAF index positive response average of 58 percent on the 2011 FEVS. Source: GAO analysis of DHS-wide, TSA, ICE, CBP, and Coast Guard action plans and U.S. Office of Personnel Management action planning guidance. Although we found that OCHCO and the four selected components are generally taking actions to execute the sixth step—monitor and evaluate the results of implementation—they have not established effective measures of the agency’s achievement of action plan goals. For example, in the DHS-wide action plan, one of the four goals is to build an effective, mission focused, diverse and inspiring cadre of leaders. A measure of success listed for this goal is that progress will be measured against the 2011 HCAAF index. For this measure, it is not clear which HCAAF index will be assessed and it does not include a target for improvement—such as a percent increase in satisfaction—by which DHS can benchmark its results. The measures of success within the DHS-wide and selected component action plans could be improved by including additional attributes of successful metrics. Specifically, in our prior work, we identified attributes of successful metrics that allow agencies to better determine whether they are meeting their goals while holding agency staff Page 29 GAO-12-940 DHS Employee Morale accountable for improving performance. 32 Three attributes relevant to the action plans are: • linkage—determines whether there is a relationship between the performance measure and the goals; • clarity—determines whether the performance measures are clearly stated; and • measurable target—determines whether, performance measures have quantifiable, numerical targets or other measurable values, where appropriate. In general, DHS and component measures satisfied the linkage attribute but did not address the clarity and measurable targets attributes. We compared DHS and the four components measures of success to the three attributes and found that all 54 measures of success incorporated the linkage attribute, 12 of the 54 measures of success did not address the clarity attribute, and 29 of the 54 measures of success did not address the measurable target attribute. As shown in table 4 below, we found that these measures demonstrate linkage because they align with the action plan goals. However, we determined that the measures demonstrate neither clarity nor a measurable target. Specifically, the measures do not demonstrate clarity because they do not provide enough detail to clearly state the metric used to measure success. They also do not demonstrate a measureable target because they do not list quantitative goals or provide a qualitative predictor of a desired outcome, which would allow the agency to better determine the extent to which they were making progress toward achieving their goals. 32 GAO, Tax Administration: IRS Needs to Further Refine Its Tax Filing Season Performance Measures, GAO-03-143 (Washington, D.C.: Nov. 22, 2002). Of the nine attributes of successful metrics listed in this report, we determined that linkage, clarity and measurable target are relevant to this evaluation. The six attributes that we did not evaluate are: objectivity, reliability, core program activities, balance, governmentwide priorities, and limited overlap. We did not include these six attributes because they were not relevant to employee morale action planning efforts. Page 30 GAO-12-940 DHS Employee Morale Table 4: Examples of DHS-Wide and Selected Components’ Measures of Success DHS Measure of Examples of Actions to unit Goal Success Achieve Goal GAO Assessment DHS- Recruit a highly Recruitment of • Streamline the hiring Clarity wide qualified and a highly process to increase The measure lacks key information that would make it diverse qualified and applicant and manager more clear—namely, what constitutes a highly qualified workforce diverse satisfaction workforce and how should “diverse workforce” be workforce • Improve manager interpreted. involvement Measurable Target • Hold managers The measure does not list quantifiable or other accountable for hiring measurable values to help determine when the goal process involvement has been reached. For example, the measure does not through performance provide a target number of employees against which evaluations DHS can benchmark its results. • Streamline all DHS job opportunity announcements TSA Performance Programs • Provide status updates to Clarity Management launched and the employee The measure lacks key information that would make it Enhancement systems performance management more clear—namely, which programs and systems deployed program TSA is to launch. • Provide training and Measurable Target resources to ensure managers and supervisors The measure does not list quantifiable or other use fair, objective, and measurable values to help determine when the goal consistent merit-based has been reached. For example, the measure does not principles during provide a target number of programs against which performance evaluations TSA can benchmark its results. CBP Continue to Increase • Initiate Division Director Clarity Implement awareness of communiqués The measure lacks key information that would make it Communication survey results • Hold a series of town hall more clear—namely, what topics in the survey results Strategies for to employees meetings across program are targeted, which employees should be included, and Disseminating offices how “increase awareness” should be interpreted. Information from • Launch a virtual town hall Measurable Target Management to meeting to address The measure does not list quantifiable or other Employees via employee concerns measurable values to help determine when the goal Technology • Launch a series of town has been reached. For example, the measure does not hall meetings in various provide a target number of employees or geographic geographic locations to locations against which CBP can benchmark its results address issues identified nor does it provide a target measurement for “increase through the FEVS awareness.” Page 31 GAO-12-940 DHS Employee Morale DHS Measure of Examples of Actions to unit Goal Success Achieve Goal GAO Assessment ICE Increasing Survey • Increase engagement with Clarity Communication employees union representatives The measure lacks key information that would make it • Development of labor more clear—namely, what issues are to be surveyed management forums and which employees should be included. within ICE Measurable Target The measure does not list quantifiable or other measurable values to help determine when the goal has been reached. For example, the measure does not provide a target number of employee survey responses against which ICE can benchmark its results. Coast Training and Develop e- • Enhance guidance Clarity Guard Development learning provided to employees The measure lacks key information that would make it course for new describing their role in more clear—namely, what is the course content or the employees individual development specific training being provided through the e-learning plans development course. Measurable Target The measure does not list quantifiable or other measurable values to help determine when the goal has been reached. For example, the measure does not provide a target number of new employees who will receive the training against which the Coast Guard can benchmark its results. Source: GAO analysis of DHS-wide, TSA, CBP, ICE and Coast Guard action plans. Officials provided several reasons why their measures of success may fall short of the attributes for successful metrics. According to OCHCO officials, OCHCO considers accomplishment of an action item step as a success and relies on the measures of success listed in its action plan as a metric for whether the action plan items were implemented. OCHCO considers whether positive responses to survey questions noted in the action plan improve over time as the outcome measure for whether action plans are effective. However, as part of its oversight and feedback on component action plans, OCHCO does not monitor or evaluate measures of success for action planning and therefore is not in a position to determine whether the measures reflect improvement. CBP officials stated that they monitor the change in FEVS results overall as the intent of the action planning is to improve their scores on the HCAAF indexes. Coast Guard officials stated that they rely on qualitative feedback from employees on action plan items, such as improved training and website updates, to measure action plan performance. TSA officials stated they assess action plan results by tracking completion dates for action items and updating OCHCO on results at least semi-annually, and ICE officials have stated they have not yet fully developed monitoring efforts to evaluate job satisfaction action planning because the human capital office Page 32 GAO-12-940 DHS Employee Morale received funding in the summer of 2011 to implement human capital programs. We acknowledge that positive responses in survey results and positive employee feedback are good indicators that action planning is working. However, until DHS and its components begin to see positive results, it is important for them to (1) understand whether they are successfully implementing the individual steps of their action plans and (2) make any necessary changes to improve on them. By not having specific metrics within the action plans that are clear and measurable, it will be more difficult for DHS to assess its efforts to address employee morale problems, as well as determine if changes should be made to ensure progress toward achieving its goals. Furthermore, effective measures are key to DHS’s action plan as it is part of a process that informs the Office of Management and Budget and OPM of DHS efforts to address survey results. According to an OPM official responsible for federal action planning to improve morale, DHS should carefully consider, for each action step, what success means to the agency, such as increased employee engagement targets. The official said that when success is defined, it should not only be clear and measurable, but should also take into account as many of the different demographic groups evaluated as possible. DHS and the Selected DHS and the selected components have initiated efforts to determine how Components Consulted other entities approach employee morale issues. DHS officials stated they Best Practices have started to review and implement what they consider to be best practices for improving employee morale, such as the following: • DHS working group—OCHCO leads a survey engagement team that holds monthly meetings during which action planning efforts from across the different components are shared and discussed. Representatives from other federal agencies such as the National Aeronautics and Space Administration and the Federal Aviation Administration have also attended these meetings and presented their action plans for addressing survey results. • Idea Factory—a TSA web-based tool adopted by DHS that empowers employees to develop, rate, and improve innovative ideas for programs, processes, and technologies. According to a DHS assessment, the Under Secretary for Management plans to use this Page 33 GAO-12-940 DHS Employee Morale tool for internal DHS employee communication so as to promote greater job satisfaction and enhance organization effectiveness. 33 Component officials we interviewed also stated they have started to review, implement, and share what they consider to be best practices for improving morale. For example: • ICE officials stated they consult with other agencies and DHS components, such as the U.S. Marshal’s Service, when addressing morale challenges and developing policies and programs. For example, the U.S. Marshal’s Service has a critical incident response program for employees encountering a traumatic event and ICE is exploring adopting a similar program. • TSA officials stated that they reached out to Marriott Corporation, CBP, and the National Aeronautics and Space Administration to identify actions for increasing employee rewards and employee confidence in leadership. • CBP officials stated they have established several ongoing working groups that routinely meet and share human capital best practices within the agency. One of these working groups has conducted benchmarking work with high-FEVS-scoring federal agencies such as the Social Security Administration, the U.S. Secret Service, the Federal Bureau of Investigation, the Internal Revenue Service and the Nuclear Regulatory Commission. • Coast Guard officials stated they share human capital best practices that may improve job satisfaction with other DHS components such as (1) their performance appraisal system which was adopted, in part, DHS-wide; (2) their automated cash award process with FEMA; and (3) Coast Guard training to supervisors with both DHS headquarters officials and FEMA. Given the critical nature of DHS’s mission to protect the security and Conclusion economy of our nation, it is important that DHS employees are satisfied with their jobs so that DHS can retain and attract the talent required to complete its work. Employee survey data indicate that when compared to other federal employees, many DHS employees report being dissatisfied and not engaged with their jobs. It is imperative that DHS understand what is driving employee morale problems and address those problems 33 DHS, Privacy Impact Assessment for the Idea Factory, January 21, 2010. Page 34 GAO-12-940 DHS Employee Morale through targeted actions that address employees’ underlying concerns. DHS has made efforts to understand morale issues across the department, but those efforts could be improved. Specifically, given the annual employee survey data available through the FEVS, DHS and its components could improve their efforts to determine root causes of morale problems by comparing demographic groups, benchmarking against similar organizations, and linking root cause findings to action plans. Uncovering root causes of morale problems could help identify appropriate actions to take in efforts to improve morale. In addition, DHS has established performance measures for its action plans to improve morale, but incorporating attributes such as improved clarity and measurable targets could better position DHS to determine whether its action plans are effective. Without doing so, DHS will have a more difficult time determining whether it is achieving its goals. To strengthen DHS’s evaluation and planning process for addressing Recommendations for employee morale, we recommend that the Secretary of Homeland Executive Action Security direct OCHCO and component human capital officials to take the following two actions: • examine their root cause analysis efforts and, where absent, add the following: comparisons of demographic groups, benchmarking against similar organizations, and linkage of root cause findings to action plans; and • establish metrics of success within the action plans that are clear and measurable. We requested comments on a draft of this report from DHS. On Agency Comments September 25, 2012, DHS provided written comments, which are and Our Evaluation reprinted in appendix V, and provided technical comments, which we incorporated as appropriate. DHS concurred with our two recommendations and described actions planned to address them. Specifically: • DHS stated that it will ensure that department-wide and component action plans are tied to root causes and that the department will conduct benchmarking against other organizations. DHS also stated that its ability to conduct demographic analysis is limited due to the data set OPM makes available to federal agencies. However, according to OPM, DHS has access to the data necessary for conducting analysis similar to our comparison of demographic groups. Page 35 GAO-12-940 DHS Employee Morale • DHS stated it will review action plans to ensure that each action is clear and measurable. We also requested comments on a draft of this report from OPM. On September 18, 2012, OPM provided a written response, which is reprinted in appendix VI. OPM’s letter indicated that it reviewed the draft report and had no comments. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the Secretary of Homeland Security, the U.S. Office of Personnel Management, and interested congressional committees. The report also will be available at no charge on GAO’s website at http://www.gao.gov. If you or your staff have any questions about this report please contact me at (202) 512-9627 or email@example.com. 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 major contributions to this report are listed in appendix VII. David C. Maurer Director, Homeland Security and Justice Issues Page 36 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies Morale at Department of Homeland Security and Other Agencies We conducted a statistical analysis of the 2011 Federal Employee Viewpoint Survey (FEVS) to assess employee morale at the Department of Homeland Security (DHS). Our analysis addressed two specific questions. First, how does morale at DHS and its components compare with morale at other agencies, holding constant demographic differences among employees? Second, to what extent is the morale gap between DHS and other agencies explained by differences in the demographic composition of the DHS workforce versus other unique characteristics of the agency or unmeasured demographic factors? This appendix explains the value of statistical analysis for understanding the employee morale gap, describes the data and methods we used, and provides additional details about our findings, which are summarized in the body of the report. In sum • DHS employees with the same demographic profiles (measured by FEVS) were about 7 percentage points less engaged and 6 points less satisfied than non-DHS employees. • Demographic differences (measured by FEVS) between DHS and other agencies are unlikely to explain the overall morale gap. Unique features of DHS (or unmeasured demographics) are more likely to be responsible. • DHS middle managers and employees with 1 to 10 years of tenure at their components—those hired after the department’s creation—have lower morale than similar employees at other departments. • Morale varies widely across DHS components, and some have similar morale as non-DHS agencies. Individual offices can strongly influence the morale gap at the component level. • The morale gap is smaller for DHS components that existed before the department was created. Understanding the Morale The morale gap between DHS and other agencies may be due to unique Gap at DHS issues within DHS or common issues faced by all agencies in similar circumstances. Unique issues might include developing an agency-wide culture, the decisions and composition of senior leaders, and the inherent uniqueness of homeland security programs. Common characteristics might include having many law enforcement and front-line customer service occupations, and having employees dispersed among many headquarters and field offices. Determining whether unique or shared issues account for the overall morale gap is important for understanding the cause of the problem. If morale at DHS was not uniquely low, compared with morale at agencies Page 37 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies with similar demographics and programs, the agency might learn from peer agencies facing similar challenges. Alternatively, if morale was lower at DHS for reasons unique to the agency, DHS might put more emphasis on understanding its own particular challenges. Distinguishing among these possible explanations can help develop a solution that is narrowly tailored to the problem. Our analysis focused on one group of shared circumstances that might explain the morale gap: employee demographics. If DHS were more likely to employ the types of workers who tend to have lower morale across all agencies of the government, the composition of the workforce might account for the gap to a greater extent than factors specific to DHS. In other words, morale at DHS may be no worse than at other agencies among demographically equivalent employees. Our analysis focused on a limited number of demographic differences, such as location and age, but attitudinal differences about pay, benefits, supervision, training, mentoring, and other human capital issues could be assessed in a similar way. We also considered how large of a morale gap there was between employees in various DHS components and work groups and non-DHS employees. The gap at the department level can mask groups of employees with higher or lower morale. Disaggregating morale into small work groups identifies areas of DHS in which morale may be high or low, and thus provides sufficiently detailed data for focused solutions to the problem. Any analysis of morale in employee surveys is limited by the fact that associations among the variables of interest may not represent cause- and-effect relationships. Nevertheless, a limited observational analysis remains useful for evaluating human capital programs. Since federal agencies cannot easily conduct high-quality randomized controlled trials of various approaches to managing their employees, the use of observational methods is common, often in the form of quantitative survey analyses or qualitative interviews and focus groups. We have previously found that a pragmatic approach to answering necessary policy questions, using the best methods and data that are feasible, is widely supported by academic experts and practitioners in policy analysis. 1 1 See GAO-10-30, “Program Evaluation: A Variety of Rigorous Methods Can Help Identify Effective Interventions,” 20-31. Page 38 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies Moreover, statistical theory has shown that observational methods can estimate cause-and-effect relationships in certain conditions. 2 Associations between morale and demographic characteristics are useful for understanding the operation of human capital programs, when interpreted cautiously and in the context of all the available evidence. Our analysis here describes patterns across the demographic groups identified in the 2011 FEVS and determines whether the aggregate differences between DHS and other agencies persists among demographically similar employees. We make no causal interpretations of these relationships, and our approach is only one that might be valid and useful. The 2011 Federal The Office of Personnel Management (OPM) provided us with a version Employee Viewpoint of the 2011 FEVS that included more detailed demographic and Survey organizational data than the file it released to the public. Specifically, our file contained the same variables as the public file but identified more detailed groups of employees. The 2011 survey included responses from 266,376 full-time, permanent federal employees, working for agencies that, according to OPM, constituted 97 percent of the executive branch workforce. OPM sampled employees within strata formed by supervisory status and organizational subgroup (e.g., component and work group). 3 This produced generally large sample sizes even for many small work groups within components, which allowed us to analyze morale among small groups of employees with an acceptable degree of precision. We focused on two types of variables in the FEVS: (1) employee demographics and (2) OPM’s Employee Engagement and Job Satisfaction indexes. A series of questions at the end of the survey collected the demographic data, rather than preexisting administrative records. OPM reported independently developing and validating the engagement indexes using factor-analytic procedures, which are common psychometric statistical methods. The survey items that made up each 2 See, for example, Donald Rubin, “The Use of Matched Sampling and Regression Adjustment to Remove Bias in Observational Studies,” Biometrics 29 (1973): 185-203, and Paul Rosenbaum, Observational Studies, New York: Springer-Verlag,1995. 3 OPM refers to these organizational subgroups as “bureaus” and “offices.” Page 39 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies index used five-point, Likert-type scales, with “agree/disagree,” “satisfied/dissatisfied,” or “good/poor” response options. 4 We used weights provided by OPM to calculate estimates and sampling variances for all analyses. The weights were the product of the unequal sampling probabilities across strata and non-response and post- stratification adjustments. Because some strata had relatively small population sizes—one-quarter with 18 employees or fewer—we corrected for finite populations. Morale Differences One explanation for lower morale at DHS is that its employees could be Between DHS Employees members of demographic groups that typically have lower morale across and Employees at Other all agencies. If this is true, the cause of morale problems and their solutions might focus less on factors that are unique to DHS and more on Agencies approaches that apply to any agency with a similar workforce. Table 5 provides basic evidence to help assess the demographic explanation. The table presents the average OPM Engagement Index for several demographic groups in the 2011 FEVS. If engagement problems at DHS were isolated to particular subgroups of employees, the morale gap should vary widely across those subgroups. In fact, engagement at DHS is lower (or statistically indistinguishable from zero) than at other agencies in each demographic subgroup we analyzed, and the gap relative to DHS does not vary by large amounts across most subgroups. However, the gap is somewhat larger among employees who were in certain subgroups, such as those who had 4 to 10 years of experience with their components and who worked outside of headquarters. 4 For a detailed description of the questions that made up the OPM indexes, see OPM, 2011 Federal Employee Viewpoint Survey, Empowering Employees, Inspiring Change, Department of Homeland Security, Agency Management Report. (Washington, D.C.), 21 and 43. Page 40 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies Table 5: DHS and Non-DHS Employee Engagement Index by Demographic Group for the 2011 FEVS OPM Employee OPM Employee 95% OPM Employee Engagement 95% Margin Engagement margin of Engagement 95% margin Index - Non-DHS of Error (+/-) Index - DHS error (+/-) Index - difference of error (+/-) All employees 67.1 0.3 60.1 0.7 -7.0* 0.7 Supervisory status Non-management 65.7 0.3 58.2 0.8 -7.5* 0.9 Management 74.1 0.4 67.0 1.1 -7.2* 1.2 Pay group Federal Wage 61.6 1.3 57.3 3.7 -4.3* 3.9 System GS 1-6 65.6 1.0 61.4 5.1 -4.2 5.2 GS 7-12 66.8 0.4 61.3 1.1 -5.5* 1.2 GS 13-15 70.5 0.4 63.0 1.2 -7.5* 1.2 Senior Executive 82.7 1.2 79.8 3.0 -3.0 3.2 Service SL 70.5 3.7 66.2 16.5 -4.4 16.9 a Other 65.6 1.1 53.1 1.4 -12.5* 1.8 Agency tenure <1 year 76.0 1.3 75.6 4.2 -0.4 4.4 1-3 years 69.7 0.6 65.2 1.6 -4.6* 1.7 4-5 years 66.0 0.9 58.4 1.8 -7.7* 2.0 6-10 years 65.8 0.6 56.0 1.1 -9.9* 1.2 11-20 years 65.1 0.6 61.3 1.7 -3.8* 1.8 20+ years 67.2 0.5 60.3 2.3 -6.9* 2.3 Location Headquarters 68.5 0.4 64.9 1.5 -3.5* 1.5 Field 66.5 0.3 59.0 0.8 -7.5* 0.8 Age <26 74.6 2.0 66.7 4.8 -7.9* 5.2 26-29 70.6 1.4 58.9 3.2 -11.7* 3.5 30-39 67.9 0.7 60.5 1.4 -7.4* 1.5 Page 41 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies OPM Employee OPM Employee 95% OPM Employee Engagement 95% Margin Engagement margin of Engagement 95% margin Index - Non-DHS of Error (+/-) Index - DHS error (+/-) Index - difference of error (+/-) 40-49 67.2 0.5 60.6 1.2 -6.6* 1.3 50-59 66.2 0.4 58.9 1.3 -7.3* 1.3 60+ 67.8 0.7 58.1 2.2 -9.8* 2.3 Race Hispanic 66.8 1.1 63.7 1.7 -3.1* 2.0 American Indian / 60.1 1.8 52.7 7.6 -7.3 7.8 Alaska Native Asian 72.4 1.2 65.0 3.4 -7.4* 3.6 Black 68.2 0.7 60.8 2.0 -7.4* 2.1 Hawaiian / Pacific 65.4 4.1 56.8 6.6 -8.5* 7.8 Islander White 67.5 0.3 58.5 0.9 -9.0* 0.9 Two or more races 62.3 1.9 56.3 4.0 -6.0* 4.4 Source: GAO analysis of 2011 FEVS data. Note: Asterisks denote differences that are statistically distinguishable from zero at the 0.05 level. a Because respondents to the 2011 FEVS reported their own pay groups, the “other” category may have included workers in various groups other than the GS system. At DHS, this group may have included Transportation Security Administration airport security screeners. Multivariate Analysis We developed several statistical models to further assess the demographic explanation. These models held constant the demographic profiles of DHS and non-DHS employees, in order to isolate the portion of the morale gap that was specifically due to non-demographic factors. The models allowed us to compare morale at DHS and other agencies among employees who were in the same demographic groups, as measured by the FEVS. To avoid methodological complications with modeling latent variables, we created a binary measure that identified whether a respondent was engaged or satisfied on each item in the respective scales. Our measure equaled 1 if the respondent gave positive answers (4 or 5) to each item in the index and 0 if the respondent gave neutral or negative responses (1,2, or 3) to at least one item. Collapsing the scale loses some information, since morale and satisfaction are continuous, latent variables. However, a collapsed measure provides some degree of comparability between OPM’s aggregate indices and our individual-level analysis, since the OPM’s indices also collapse the scale. The differences Page 42 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies among agencies and subgroups of employees are generally similar using either our measure or OPM’s. 5 We focused on the associations between broad measures of morale and fixed demographic characteristics available in the 2011 FEVS. Fixed demographics and broad measures of satisfaction are not subject to artificially high correlations that a survey’s design can produce among attitudinal measures. The models took the following form: ( ) ( E Morale ij | DHS j , Demog ij = Λ αDHS j +Demog ij β ) (1) E (Moraleij | DHS j , Demog ij ) = Λ (αDHS j +DHS j × Demog ij β D + Demog ij β G ) (2) Moraleij indicates whether employee i at agency j was engaged or satisfied, using the binary measure we calculated from the survey items that make up the OPM indexes (see above). DHSj indicates whether the employee worked for DHS, Demogij is a vector of demographic indicators (listed in table 6), Λ is the logistic function, and α and β are vectors of coefficients that estimate how morale varied among employees in different demographic groups. We included all demographic factors measured by the FEVS that plausibly could have predicted morale and were clearly causally prior to morale. We excluded pay group, however, because of its high correlation with supervisory status. Model 2 allows DHS and non-DHS employees in the same demographic groups to have β different levels of morale, as described by β D and G .We estimated each model using cluster-robust maximum likelihood methods, with 365 agency clusters (e.g., Transportation Security Administration [TSA]). Our multivariate analysis found that DHS employees remained an average of 6.4 percentage points less engaged (+/- 3.2) (see table 6) and 5.5 points less satisfied (+/- 2.2) (not shown) on our scales than employees at other agencies who had the same age, office location, race, sex, supervisory status, and tenure. This suggests that measured demographic differences between employees at DHS and other agencies 5 We discuss options for a more sophisticated analysis of engagement and satisfaction on latent, continuous scales at the end of this appendix. Page 43 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies do not fully explain the morale gap. Instead, factors that are intrinsic to DHS, such as culture or management practices, or demographic factors not measured by FEVS, such as education or occupation, are likely to be responsible. Decomposing the Morale Gap We can further explore the roles of demographics and unique DHS characteristics by performing an Oaxaca decomposition of the results of model 2, in order to compare DHS with other agencies. 6 Oaxaca decomposition can assess whether the overall morale gap is explained by the demographic characteristics of DHS employees, or whether it is explained by lower morale among DHS employees in the same demographic groups. In other words, does DHS employ an unusually large number of workers who tend to have low morale across all agencies, or do workers with the same backgrounds have uniquely lower morale at DHS? Table 6: Model Estimates of Employee Engagement Index at DHS and Other Agencies Using the 2011 FEVS DHS Other Agencies Percentage engaged Percentage engaged (GAO measure) Standard error (GAO measure) Standard error All employees 20.7 1.5 27.1 0.5 a Gap between DHS and other -6.3 agencies Portion of gap due to the 0.1 demographic composition of DHS Portion of gap due to unique -6.4 differences between DHS and non- DHS employees with similar demographic profiles Race Hispanic 27.0 1.1 29.0 0.7 American Indian / Alaska Native 13.2 3.7 21.3 0.8 6 Oaxaca decomposition is a method of disaggregating an average difference between two groups into (1) the part due to the differences in the values of the variables that determine the outcome of interest and (2) the part due to differences in the partial relationships between the predictor variables and the outcome. See Ronald Oaxaca, “Male–Female Wage Differentials in Urban Labor Markets,” International Economic Review 14 (1973). 693-709 Page 44 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies DHS Other Agencies Percentage engaged Percentage engaged (GAO measure) Standard error (GAO measure) Standard error Asian 29.0 3.5 34.4 0.8 Black 23.7 2.0 27.8 0.5 Hawaiian/Pacific Islander 23.1 2.5 28.4 1.3 White 19.0 1.5 26.7 0.5 Two or more races 18.3 3.0 20.7 0.8 Years worked for component <1 34.6 2.5 38.1 0.9 1-3 22.8 2.2 30.5 0.6 4-5 20.0 1.8 27.3 0.7 6-10 17.6 1.6 25.7 0.5 11-20 19.3 2.2 24.7 0.5 20+ 21.2 3.0 26.6 0.6 Sex Male 20.6 1.6 27.1 0.5 Female 20.9 1.5 27.2 0.5 Supervisory status Non-supervisor 18.2 1.5 23.3 0.5 Team leader 17.8 1.7 26.3 0.6 Supervisor 23.3 1.7 31.7 0.6 Manager 31.7 2.5 40.5 0.7 Executive 47.3 2.8 54.2 1.3 Age < 26 20.0 2.9 28.6 1.1 26-29 16.7 2.6 24.9 0.9 30-39 19.3 1.6 26.3 0.6 40-49 20.8 1.9 27.4 0.5 50-59 21.0 1.7 26.7 0.5 60+ 22.8 1.6 29.3 0.6 Location Headquarters 21.4 1.6 27.8 0.6 Field 20.2 1.9 26.6 0.6 Source GAO analysis of U.S. Office of Personnel Management 2011 FEVS data. Page 45 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies Note: We created a unique measure of whether each employee in the 2011 FEVS was engaged for the purposes of statistical analysis and describing differences among groups of employees. Because our measure is defined differently than the OPM Employee Engagement Index, the measures are not comparable. Our measure counts an employee as engaged if he or she gave positive answers to each item in the OPM index (see text).Engagement statistics in this table are the in-sample percentage of employees who are predicted by model 2 to be engaged on GAO’s measure. a Gap estimates and their decomposition are estimated using Oaxaca methods applied to model 2. As shown in table 6, the model suggests that the demographic profile of DHS employees (measured by FEVS) tends to slightly increase their engagement and reduce the gap compared with employees at other agencies. The demographic characteristics we can observe in FEVS reduce the overall gaps in the proportion engaged and satisfied on our scales by 0.1 and 1.0 percentage points, respectively. 7 Instead, the morale gap is better explained by unique differences in morale between DHS and other agencies among demographically similar employees. Such intrinsic differences increase the gaps in the proportion engaged and satisfied by 6.4 and 5.5 percentage points, respectively. If the demographic profile of the DHS workforce did not change, but DHS could achieve the same levels of morale as other agencies from the same types of employees, our model predicts that DHS employees would not have lower morale than employees at other agencies. DHS employees with lower-level positions and component tenure were among those with lower morale, relative to employees in other agencies. As shown in figures 6 and 7, our measures of engagement and satisfaction generally increased with seniority and decreased with tenure, among employees at DHS and other agencies. At DHS, however, morale increased more slowly as employees gained more seniority, and it declined more quickly as they spent more time at the agency. For example, the average newly hired employee at DHS and similar employees at other agencies had statistically indistinguishable levels of engagement. By their sixth years, however, satisfaction for the DHS employee declined to an average of 18 percentage points, whereas satisfaction for the non-DHS employees declined to an average of only 26 percentage points. A similar pattern exists with respect to supervisory status (see figures 6 and 7). These patterns are particularly important for explaining the overall morale gap, because DHS had about 30 percent more supervisors and about twice as many people with 6 to 10 years of 7 We omitted results for job satisfaction from table 6 to conserve space, but they generally resemble the results for engagement. Page 46 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies component tenure (as a share of all employees), compared with people at other agencies (according to FEVS). 8 Figure 6: Engagement Index Scores by Supervisory Status and Tenure, for DHS and Non-DHS Employees 8 DHS employees may have answered the 2011 FEVS question on tenure differently, depending on whether they worked for a component that existed before the department was created. Specifically, employees who reported more than 10 years of service may have interpreted the question to include their service prior to the creation of DHS. Page 47 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies Page 48 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies Figure 7: Satisfaction Index Scores by Supervisory Status and Tenure, for DHS and Non-DHS Employees Page 49 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies Morale Differences within Low employee morale is not a uniform problem throughout DHS. As DHS Components and shown in table 7, engagement varies widely across components within Work Groups the department, with employees in some components not being significantly different from the average employee at non-DHS agencies. These components include the U.S. Coast Guard (Coast Guard), Federal Law Enforcement Training Center (FLETC), Management Directorate (MGMT), and U.S. Secret Service (USSS). Job satisfaction at these components also matches or exceeds that found at other agencies (not shown in table 7). DHS has a number of components whose employees have substantially lower morale than employees at other agencies and elsewhere in the department. The large share of DHS employees working in these components accounts for the overall morale gap between DHS and other agencies. Page 50 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies Components with lower morale include Federal Emergency Management Agency (FEMA), Immigration and Customs Enforcement (ICE), Intelligence and Analysis (IA), National Protection and Programs Directorate (NPPD), Science and Technology (ST), and the TSA. The engagement scores of these components range from 9.1 to 13.9 percentage points lower than the average score for non-DHS agencies (see table 7). As a group, these components make up 46 percent of the employees interviewed for the FEVS. Consequently, the components with substantially lower morale have a large influence on the gap relative to the rest of the government, despite the fact that morale at many smaller DHS components is no worse. Morale at some of the less engaged and satisfied components is, in turn, strongly influenced by particular employee workgroups (see table 7). For example, the average engagement at TSA is 12.8 percentage points (apart from rounding) lower than at non-DHS agencies. Within TSA, however, the collectively large groups of air marshal, law enforcement, and screening workers account for much of the overall difference. A similar pattern applies to the enforcement, removal, and homeland security investigation staffs at ICE, the field operations staff at CBP, and the Federal Protective Service. Such variation within components further suggests that the morale gap is isolated to particular areas within DHS that account for a large proportion of its workforce. At other components, morale is more uniformly lower across most offices. Average engagement at all work groups within FEMA is 5.8 to 17.7 percentage points lower than the non-DHS average, with the exception of two regional offices and the offices of the Administrator and Chief of Staff. The components of ST and IA also have more consistently low morale across work groups. Page 51 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies Table 7: OPM Employee Morale Index by DHS Component and Offices Using the 2011 FEVS EE Index OPM Employee 95% EE Index 95% minus non- 95% Engagement (EE) margin of minus DHS margin of DHS margin of Index error (+/-) average error (+/-) average error (+/-) N DHS component Non-DHS 67.1 0.3 . . 0.0 0.0 250,870 Citizenship and Immigration Services 64.0 2.0 . . -3.2* 2.0 1,303 CBP 62.9 1.3 . . -4.3* 1.3 3,057 Coast Guard 70.6 2.5 . . 3.5* 2.5 863 FEMA 58.0 2.3 . . -9.1* 2.3 972 FLETC 66.2 1.9 . . -1.0 1.9 611 ICE 58.0 2.5 . . -9.1* 2.5 1,313 Intel and Analysis 53.2 4.0 . . -13.9* 4.0 150 IG 70.5 3.3 . . 3.3 3.3 307 MGMT 65.7 2.1 . . -1.4 2.2 628 NPPD 57.7 2.1 . . -9.4* 2.2 771 Science and Technology 55.7 2.8 . . -11.5* 2.8 258 USSS 68.1 2.4 . . 1.0 2.4 959 Office of Secretary 63.6 2.5 . . -3.6* 2.6 409 TSA 54.4 1.1 . . -12.8* 1.2 3,701 DHS, no sub-agency 60.5 6.3 . . -6.6* 6.3 204 TSA Non-DHS 67.1 0.3 4.4 0.9 0.0 0.0 250,870 Other DHS 62.7 0.8 0.0 0.0 -4.4* 0.9 11,805 Headquarters staff 61.5 4.4 -1.2 4.5 -5.6* 4.4 279 Law enforcement/air marshal 52.9 4.4 -9.8 4.4 -14.2* 4.4 246 Federal security director staff 64.8 2.6 2.1 2.7 -2.3 2.6 630 Screeners 50.9 1.4 -11.8 1.6 -16.2* 1.4 2,273 ICE Non-DHS 67.1 0.3 6.7 0.7 0.0 0.0 250,870 Other DHS 60.4 0.7 0.0 0.0 -6.7* 0.7 14,193 Director 66.3 23.4 5.9 23.4 -0.9 23.4 10 Acquisitions 41.2 31.0 -19.2 31.0 -26.0 31.0 13 CFO 68.3 12.6 7.9 12.6 1.2 12.6 33 Page 52 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies EE Index OPM Employee 95% EE Index 95% minus non- 95% Engagement (EE) margin of minus DHS margin of DHS margin of Index error (+/-) average error (+/-) average error (+/-) N CIO 69.8 14.1 9.4 14.1 2.6 14.1 29 Enforcement and removal operations 52.8 3.8 -7.6 3.9 -14.4* 3.8 491 Human capital 51.2 24.3 -9.2 24.3 -16.0 24.3 11 Intelligence 59.1 16.7 -1.3 16.7 -8.1 16.7 27 International affairs 57.5 11.0 -2.9 11.0 -9.6 11.0 21 Homeland security investigation 59.0 4.1 -1.4 4.2 -8.2* 4.1 433 Principle legal advisor 69.4 7.7 9.0 7.8 2.2 7.7 62 Professional responsibility 66.6 12.4 6.2 12.4 -0.6 12.4 40 Other 77.2 20.1 16.8 20.1 10.1 20.1 13 Coast Guard Non-DHS 67.1 0.3 7.6 0.7 0.0 0.0 250,870 Other DHS 59.6 0.7 0.0 0.0 -7.6* 0.7 14,643 Vice commandant 71.9 18.1 12.3 18.1 4.7 18.1 12 Chief of staff, mission support 77.8 8.9 18.2 9.0 10.6* 8.9 66 Deputy commandant for operations 89.1 5.1 29.6 5.2 22.0* 5.1 15 Force readiness command 67.4 17.8 7.9 17.8 0.3 17.8 13 Marine safety security stewardship 68.8 14.5 9.2 14.5 1.6 14.5 21 Engineering and logistics 69.9 9.8 10.3 9.9 2.8 9.8 20 Command, control, communications, computers, and IT 79.3 18.7 19.8 18.7 12.2 18.7 13 Atlantic 72.6 10.0 13.0 10.1 5.4 10.0 44 Pacific 61.0 14.5 1.4 14.5 -6.1 14.5 19 Districts 1 76.7 8.7 17.1 8.8 9.6* 8.7 46 Districts 5 69.5 8.8 9.9 8.9 2.3 8.8 69 Districts 7 66.8 12.2 7.2 12.2 -0.3 12.2 41 Districts 8 71.8 8.7 12.2 8.7 4.7 8.7 45 Districts 9 71.8 9.9 12.2 10.0 4.7 9.9 19 Districts 11 64.3 14.2 4.7 14.2 -2.9 14.2 35 Districts 13 78.5 17.4 18.9 17.4 11.4 17.4 18 Districts 14 80.2 19.4 20.6 19.4 13.0 19.4 12 Districts 17 68.7 14.5 9.1 14.5 1.6 14.5 22 Other 69.2 4.6 9.6 4.6 2.0 4.6 270 Page 53 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies EE Index OPM Employee 95% EE Index 95% minus non- 95% Engagement (EE) margin of minus DHS margin of DHS margin of Index error (+/-) average error (+/-) average error (+/-) N CBP Non-DHS 67.1 0.3 8.5 0.8 0.0 0.0 250,870 Other DHS 58.6 0.8 0.0 0.0 -8.5* 0.8 12,449 Office of the commissioner 73.2 18.4 14.5 18.5 6.0 18.4 15 Chief Counsel 58.3 15.9 -0.3 15.9 -8.8 15.9 29 Human resources management 73.3 11.5 14.7 11.6 6.2 11.5 36 Border Patrol 69.2 2.2 10.6 2.3 2.1 2.2 861 International trade 66.1 7.4 7.5 7.4 -1.1 7.4 80 Internal affairs 73.0 13.0 14.4 13.0 5.9 13.0 36 Field operations 57.5 1.9 -1.1 2.0 -9.6* 1.9 1,326 Administration and public affairs 61.5 8.4 2.9 8.4 -5.6 8.4 69 Information and Technology 66.9 6.8 8.3 6.8 -0.3 6.8 98 Training and development 59.4 9.1 0.8 9.1 -7.7 9.1 47 Internal Affairs 59.2 19.6 0.6 19.6 -7.9 19.6 16 Air and marine 58.2 6.6 -0.5 6.7 -9.0* 6.6 103 Intelligence and operations coordination 60.9 20.3 2.3 20.4 -6.2 20.3 16 Technology, innovation, and acquisition 80.1 12.4 21.5 12.4 13.0* 12.4 23 Source: GAO analysis of U.S. Office of Personnel Management 2011 FEVS data. Note: Asterisks denote differences that are distinguishable from zero at the 0.05 level. Offices listed within the components are described as they are identified in OPM’s 2011 FEVS data files provided to GAO. Comparison of Morale between One explanation for why morale varies across components focuses on Employees in Preexisting the length of time each organization has existed. Components that Components and Components existed prior to the creation of DHS may have had more time to develop Created with DHS successful cultures and management practices than components that policymakers created with the department in 2003. As a result, the preexisting components may have better morale today than components with less mature cultures and practices. To assess this explanation, we analyzed morale among two groups of components, divided according to whether the component was established with the creation of DHS or existed previously (see table 8). We considered three components to be preexisting—FLETC, USSS, and the Coast Guard—and the rest to be newly created. Because TSA was Page 54 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies created about 2 years before DHS, we included it with components that were created with DHS. Our analysis shows that employees at the more recently created components were less engaged and satisfied on average than employees at the preexisting components and at non-DHS agencies. For the preexisting components, engagement was about 2.2 percentage points higher than at the rest of the government, and the difference in satisfaction was small (less than 1.4 percentage points). In contrast, engagement and satisfaction at the more recently created components were about 8 and 5.1 percentage points lower than at the rest of the government, respectively. Table 8: Morale at Preexisting and Recently Created Components of DHS Using the 2011 FEVS OPM OPM Job Employee Satisfaction Margin of Engagement Margin of Index error (+/-) Index error (+/-) Non-DHS 68.5 0.2 67.1 0.3 Preexisting components 69.9 1.4 69.3 1.6 Components created with DHS 63.4 0.6 59.2 0.7 Source: GAO analysis of OPM 2011 FEVS data. Note: Preexisting components include FLETC, USSS, and Coast Guard, with all others classified as being created with DHS. Multivariate Analysis We developed a statistical model to confirm whether the differences among components persist, holding constant demographic differences among their employees. In an alternative version of model 1 above, we replaced DHSj with a vector of variables indicating whether the employee worked for DHS components or at an agency other than DHS. All other parts of the model were identical. The model estimates generally confirmed the differences in engagement between non-DHS and DHS component employees in the raw data (see table 9), with two exceptions. The model estimated that, holding constant demographic differences, employees in the Management Directorate and Office of the Secretary were 6.9 and 7.7 percentage points less engaged on average than employees in non-DHS agencies. This suggests that the engagement gap for employees in these offices is more similar to the gap at other offices, holding constant the demographic differences among Page 55 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies offices measured by FEVS. The model estimated that differences in satisfaction between the components and non-DHS agencies were generally similar to such differences in engagement (see table 9). The fact that differences among components remained, even among demographically equivalent employees, suggests that either unmeasured demographic variables or intrinsic characteristics of the components are responsible for the differences in morale. Table 9: Model Estimates of the Difference in Engagement and Job Satisfaction between Employees in DHS Components and Non-DHS Agencies Based on the 2011 FEVS Difference in engagement Standard Difference in satisfaction Standard DHS component (GAO measure, %) error (GAO measure, %) error CIS -5.6 0.5 -3.5 0.5 CBP -6.1 0.5 -2.5 0.6 Coast Guard 2.6 0.5 -3.3 0.5 FEMA -9.9 0.5 -5.7 0.4 FLETC 0.5 0.6 6.8 0.5 ICE -9.8 0.6 -8.4 0.5 I&A -13.9 0.5 -10.6 0.5 IG 6.4 0.9 4.2 0.8 MGMT -6.9 0.6 -3.0 0.6 NPPD -11.3 0.5 -9.0 0.5 S&T -15.7 0.6 -9.6 0.8 USSS 3.6 0.6 0.1 0.5 Office of Secretary -7.7 0.6 -8.5 0.7 TSA -12.4 0.5 -12.2 0.4 DHS, no component 7.5 0.3 3.3 1.4 Components existing prior to DHS creation 2.5 0.8 0.5 2.0 Components created with DHS (plus TSA) -8.8 1.4 -6.8 1.9 Source: GAO analysis of 2011 FEVS data. Note: Engagement statistics are the in-sample proportion of employees who are predicted by model 1 to be engaged or satisfied on GAO’s measure. Opportunities for Our analysis discussed in this appendix has a narrow scope: assessing Additional Analysis whether demographic differences among employees explain the morale differences across DHS and non-DHS employees. Consequently, DHS or others could expand and improve upon our findings. Page 56 GAO-12-940 DHS Employee Morale Appendix I: Statistical Analysis of Employee Morale at Department of Homeland Security and Other Agencies Future work could examine whether attitudinal differences among employees at DHS and other agencies explain the overall morale gap, in addition to demographic differences. The 2011 FEVS measures employee attitudes about pay, benefits, health and safety hazards, training, supervisors, and other issues that could vary meaningfully between employees at DHS and other agencies and, therefore, explain why DHS has lower morale. One might include these factors in a decomposition similar to the one we performed in this appendix. This could further assess how factors unique to DHS and factors that are common across all agencies explain the overall morale gap. A broader attitudinal analysis likely would require the use of more sophisticated statistical methods for estimating the values of and relationships among latent variables. The broad measures of morale we analyze in this appendix, such as the OPM Employee Engagement index, are made up of responses to questions on smaller dimensions, such as leadership and supervision. To avoid simply replicating the correlations that were used to create the indexes, latent variable models could be useful to examine the relationships among these concepts and compare morale on latent scales between DHS and non-DHS agencies. This was beyond the scope of our work. Page 57 GAO-12-940 DHS Employee Morale Appendix II: Scope and Methodology Appendix II: Scope and Methodology The objectives for this report were to evaluate (1) how DHS employee morale compares with that of other federal government employees and (2) to what extent DHS and its selected components determined the root causes of employee morale and developed action plans to improve morale. To address our objectives, we evaluated both DHS-wide efforts and efforts at four selected components to address employee morale—CBP, ICE, TSA, and the Coast Guard. We selected the four DHS components based on their workforce size and how their 2011 job satisfaction and engagement index scores 1 compare with the non-DHS average. 2 The components selected had scores both above, below, and similar to the average: TSA—below average on both indexes, constituting 25 percent of the DHS workforce; ICE—below average on both indexes, accounting for 9 percent of the DHS workforce; CBP—at the non-DHS average for satisfaction and below on engagement, representing 27 percent of the DHS workforce; and the civilian portion of the Coast Guard—at the non- DHS average for satisfaction and above on engagement, composing 4 percent of the DHS workforce. 3 Together these components represent 65 percent of DHS’s workforce. To evaluate how DHS’s employee morale compares with that of other federal government employees, we analyzed employee responses to the 2011 FEVS. We determined that the 2011 FEVS data were reliable for the purposes of our report, based on interviews with OPM staff, review and analysis of technical documentation of its design and administration, and electronic testing. We used two measures created by OPM—the employee job satisfaction and engagement indexes—to describe morale 1 The job satisfaction index, composed of seven FEVS questions such as “my work gives me a feeling of personal accomplishment,” indicates the extent to which employees are satisfied with their jobs and various aspects thereof. The Engagement Index, composed of 15 FEVS questions, indicates the extent to which employees are immersed in the content of the job and energized to spend extra effort in job performance. 2 Throughout this report, non-DHS refers to all federal employee FEVS responses outside of DHS. 3 The FEVS does not survey military personnel. Therefore, for the purposes of our review, our focus at the Coast Guard was on morale-related concerns for the Coast Guard’s civilian workforce, which included 8,342 employees in 2011. The Coast Guard’s civilian workforce is responsible for supporting the Coast Guard mission through over 200 different types of professional and trade fields, such as engineering, information technology, administration, and electrical work. Page 58 GAO-12-940 DHS Employee Morale Appendix II: Scope and Methodology across the federal government and within DHS. We calculated these measures for various demographic groups, DHS components, and work groups, in order to compare morale at DHS and other agencies among employees who were demographically similar, in part using statistical models. Appendix I describes our methods and findings in more detail. In addition, we interviewed employee groups about morale to identify examples of what issues may drive high and low morale within DHS. We selected the employee groups based on the size of the employee group within each selected component, ensuring we met with employees from employee groups that composed significant proportions of FEVS respondents, such as screeners from TSA (61 percent of TSA respondents) and homeland security investigators from ICE (33 percent of ICE respondents). The comments received from these interviews are not generalizable to entire groups of component employees, but provide insights into the differing issues that can drive morale. To determine the extent to which DHS and the selected components identified the root causes of employee morale and developed action plans for improvements, we reviewed analysis results, interviewed agency human capital officials and representatives of employee groups, and evaluated action plans for improving morale. To identify criteria for determining effective root cause analysis using survey data, we reviewed both OPM and Partnership for Public Service guidance for action planning based on annual employee survey results. On the basis of these guidance documents, we identified factors that should be considered in employee survey analysis that attempts to understand morale problems, such as use of demographic group comparisons, benchmarking results against results at similar organizations, and the linking results of root cause analyses to action planning efforts. We evaluated documents summarizing DHS-wide and selected component root cause analyses of the 2011 FEVS to determine whether the factors we identified were included in the analyses. In addition, we interviewed DHS officials who conducted the analyses in order to fully understand root cause analysis efforts. To identify criteria for determining agency action plans we reviewed OPM guidance for using FEVS results and previous GAO work indicating agencies’ success in measuring performance. On the basis of these guidance documents, we identified OPM’s six steps that should be considered in developing action plans and identified three attributes that were relevant for measuring action plan performance—linkage, clarity, and measurable target. We compared the action plans with these criteria to determine whether these items were included in the action plans. In addition, we interviewed DHS and component officials to identify efforts to leverage best practices for improving morale. Page 59 GAO-12-940 DHS Employee Morale Appendix II: Scope and Methodology We conducted this performance audit from October 2011 through September 2012 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 that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Page 60 GAO-12-940 DHS Employee Morale Appendix III: DHS and Selected Component Appendix III: DHS and Selected Component Steps Taken to Determine Root Causes of Morale Problems Steps Taken to Determine Root Causes of Morale Problems DHS Efforts to Determine Since 2007 DHS’s Office of the Chief Human Capital Officer (OCHCO) Root Causes of Morale has completed several efforts to determine root causes of morale DHS- Problems wide. Focus groups. In 2007 OCHCO conducted focus groups to determine employee concerns related to employee morale. DHS’s focus group effort probed for insights into four areas—(1) leadership, (2) communication, (3) empowerment, and (4) resources—and highlighted concerns raised by focus group participants in each of those areas. For example, within the leadership area, OCHCO’s focus group analysis found that the Customs and Immigration reorganization was a topic discussed by many of the U.S. Customs and Border Protection (CBP), U.S. Immigration and Customs Enforcement (ICE), and Citizenship and Immigration Services (CIS) personnel, especially what they felt was a lack of mission understanding on the part of their managers. According to the analysis, non-supervisory participants expressed dissatisfaction with the combination of three types of inspection functions to present “one face at the border.” One Face at the Border For operations at ports of entry, in September 2003 CBP issued its plan for consolidating the inspection functions formerly performed by separate inspectors from the three legacy agencies—customs inspectors from U.S. Customs, immigration inspectors and Border Patrol from the former Immigration and Naturalization Service, and the agriculture border inspectors from the Department of Agriculture’s Animal and Plant Health Inspection Service. The plan, referred to as “One Face at the Border,” called for unifying and integrating the legacy inspectors into two new positions—a CBP officer and a CBP agricultural specialist. The new CBP officer would serve as the frontline officer responsible for carrying out the priority anti-terrorism mission as well as the traditional customs and immigration inspection functions while also identifying and referring goods in need of a more extensive agricultural inspection to the agricultural specialist. CBP anticipated that having a well-trained and well-integrated workforce that could carry out the complete range of inspection functions involving the processing of individuals and goods would allow it to utilize its inspection resources more effectively and enable it to better target potentially high-risk travelers. Together, CBP envisioned the result to be more effective inspections and enhanced security at ports of entry while also accelerating the processing of legitimate trade and travel. Focus group results were distributed to DHS components for consideration in action planning efforts, according to OCHCO officials. CBP, CIS, TSA, the Federal Emergency Management Agency (FEMA), and the Federal Law Enforcement Training Center each addressed at least one of the focus group results relating to leadership, communication, empowerment, or resources in subsequent action plans, according to OCHCO officials. Page 61 GAO-12-940 DHS Employee Morale Appendix III: DHS and Selected Component Steps Taken to Determine Root Causes of Morale Problems Statistical analysis. In 2008 OCHCO performed statistical analysis of Federal Employee Viewpoint Survey (FEVS) data, beyond examining high- and low-scoring questions, in an effort to determine what workplace factors drove employee job satisfaction. Specifically, the analysis involved isolating which sets of FEVS questions most affect employee job satisfaction. The analysis found that five work areas identified in FEVS questions drive employee job satisfaction: (1) performance and rewards, (2) supervisor support, (3) physical conditions and safety, (4) senior leadership effectiveness, and (5) the DHS mission. According to OCHCO officials, DHS components were encouraged to conduct follow-up discussions at the lowest possible organizational level based on component survey scores in each of the five work areas. However, OCHCO officials stated that they are not aware of any results of this effort because OCHCO did not track or follow-up with the components on the effect of key driver discussions that may have occurred. In addition, increased emphasis on supervisor performance management training was also implemented as a result of the analysis, according to OCHCO officials. Exit survey. In 2011, DHS began administering an exit survey to understand why employees choose to leave their DHS positions. Specifically, according to OCHCO officials, the DHS exit survey was designed to determine where departing employees were moving both inside and outside of DHS, to identify barriers related to diversity, to identify reasons that veterans may be leaving DHS, and to capture feedback from interns. The 2011 exit survey found, among other things, that 27 percent of departing employees who responded to the exit survey were staying within DHS or moving to a different position, and an additional 12 percent of respondents were retiring. Lack of quality supervision and advancement opportunities were the top reasons responding employees indicated for leaving their positions. 1 Exit survey results are shared with DHS components on a quarterly and annual basis. 2011 FEVS analysis. For the 2011 FEVS, DHS’s OCHCO evaluated the results by comparing Human Capital Assessment and Accountability Framework (HCAAF) index results by component. The analysis showed 1 Results from DHS’ exit survey should be interpreted with caution. Because of the method in which the survey is administered, its response rate in 2011 was quite low, close to 40 percent. It is likely that a higher response rate would have produced somewhat different results. Page 62 GAO-12-940 DHS Employee Morale Appendix III: DHS and Selected Component Steps Taken to Determine Root Causes of Morale Problems where the lowest index scores were concentrated. As shown in figure 8, lower scores across the indexes were concentrated among several components, including Intelligence and Analysis, Transportation and Security Administration (TSA), ICE, National Protection and Programs Directorate, and FEMA. Figure 8: DHS’s 2011 Component Comparison Based on Four HCAAF Indexes The analysis also determined how DHS’s scores on the four indexes trended over time and compared with governmentwide averages. As shown in figure 9, DHS-wide scores have generally trended upward over Page 63 GAO-12-940 DHS Employee Morale Appendix III: DHS and Selected Component Steps Taken to Determine Root Causes of Morale Problems time, but continue to lag behind governmentwide averages for each index. Figure 9: DHS HCAAF Scores since 2006 Employee Engagement Executive Steering Committee (EEESC). In January 2012 the DHS Secretary directed all component heads to take steps to improve employee engagement through launch of the EEESC. According to OCHCO officials, the EEESC was launched in response to congressional concerns about DHS employee morale and the Partnership for Public Service results showing DHS’s low placement on the list of Best Places to Work. The EEESC is charged with serving as the DHS corporate body responsible for identifying DHS-wide initiatives to improve employee engagement, oversee the efforts of each DHS component to address employee engagement, and provide periodic reports to the Under Secretary for Management, Deputy Secretary, and Secretary on DHS-wide efforts to improve employee morale and engagement. Specifically, the Secretary made the following directives to component heads: Page 64 GAO-12-940 DHS Employee Morale Appendix III: DHS and Selected Component Steps Taken to Determine Root Causes of Morale Problems • develop and assume responsibility for employee engagement improvement plans, • identify and assign specific responsibilities for improved employee engagement to component senior executive performance objectives, • identify and assign a senior accountable official to serve on the EEESC, • conduct town hall meetings with employees, • attend a Labor-Management Forum meeting, and • provide monthly reports on actions planned and progress made to the Office of the Chief Human Capital Officer. As of August 2012, each of the Secretary’s directives were completed, with the exception of assigning responsibilities for improved employee engagement to Senior Executive performance objectives, which DHS plans to implement in October 2012 as part of the next senior executive performance period. The EEESC met in February 2012, and component representatives shared their latest action plans and discussed issues of joint concern. In preparation for the 2012 FEVS, the EEESC released a memorandum from the Secretary describing the responsibilities of the EEESC, highlighting department actions, and encouraging employee participation in the FEVS, which began in April 2012. The EEESC also agreed that a corresponding message should be released from component heads outlining specific component actions taken in response to past survey results and encouraging participation in the next survey. In an April 2012 EEESC meeting, the Partnership for Public Service provided a briefing describing the Best Places to Work in the Federal Government rankings and best practices across the government for improving morale scores. The EEESC members also discussed methods for improving the response rates for the upcoming survey and engaged in an action planning exercise designed to help identify actions for department-wide deployment, according to OCHCO officials. As of August 2012, EEESC action items were in development and had not been finalized. According to OCHCO officials, the EEESC plans to decide on action items by September 2012, but a projected date for full implementation has yet to be established because the actions have not been decided upon. Components Have Also In addition to the DHS-wide efforts, the components we selected for Conducted Some Root review—ICE, TSA, the U.S. Coast Guard (Coast Guard), and CBP— Cause Analyses Using conducted varying levels of analyses regarding the root causes of morale issues to inform agency action planning efforts. The selected components FEVS Results each analyzed FEVS data to understand leading issues that may relate to Page 65 GAO-12-940 DHS Employee Morale Appendix III: DHS and Selected Component Steps Taken to Determine Root Causes of Morale Problems morale, but the results indicated where job satisfaction problem areas may exist and do not identify the causes of dissatisfaction within employee groups. A discussion of the four selected components’ 2011 FEVS analysis and results are described below. TSA. In its analysis of the 2011 FEVS, TSA focused on areas of concern across groups, such as pay and performance appraisal concerns, and also looked for insight on which employee groups within TSA may be more dissatisfied with their jobs than others by comparing employee group scores on satisfaction-related questions. TSA compared its results with CBP results, as well as against DHS and governmentwide results. When comparing CBP and TSA scores, TSA found that the greatest differences in scores were on questions related to satisfaction with pay and whether performance appraisals were a fair reflection of performance. TSA scored 40 percentage points lower on pay satisfaction and 25 percentage points lower on performance appraisal satisfaction. In comparing TSA results with DHS and governmentwide results, TSA found that TSA was below the averages for all FEVS dimensions. 2 TSA also evaluated FEVS results across employee groups by comparing dimension scores for headquarters staff, the Federal Air Marshals, Federal Security Director staff, and the screening workforce. TSA found that the screening workforce scored at or below scores for all other groups across all of the dimensions. ICE. In its analysis of the 2011 FEVS, ICE analyzed the results by identifying ICE’s FEVS questions with the top positive and negative responses. ICE found that its top strength was employees’ willingness to put in the extra effort to get a job done. ICE’s top negative result was employees’ perceptions that pay raises did not depend on how well employees perform their jobs. ICE also sorted the primary low-scoring results into action planning themes, such as leadership, empowerment, and work-life balance. ICE found, among other things, that employee views on the fairness of its performance appraisals were above DHS’s average but that views on employee preparation for potential security threats were lower. When comparing ICE’s results with average governmentwide figures, ICE found, among other things, that ICE was lower on all of the HCAAF indexes, including job satisfaction. According 2 The FEVS includes questions grouped into the following dimensions: work experiences, supervisor/team leader, agency, work unit, leadership, satisfaction, and work/life. Page 66 GAO-12-940 DHS Employee Morale Appendix III: DHS and Selected Component Steps Taken to Determine Root Causes of Morale Problems to ICE human capital officials, future root cause analysis plans for the 2012 FEVS are to benchmark FEVS scores with those of similar law enforcement agencies such as the Drug Enforcement Agency; Federal Bureau of Investigation; Federal Law Enforcement Training Center; United States Secret Service; Alcohol, Tobacco and Firearms, and the U.S. Marshals. CBP. In its analysis of the 2011 FEVS, CBP focused its analysis on trends since 2006. For example, the analysis showed that CBP increased its scores by 5 or more percentage points for 36 of the 39 core FEVS questions. CBP highlighted its greatest increases in HCAAF areas, such as results-oriented performance, which showed a 21 percent improvement over 2006 responses to the question—my performance appraisal is a fair reflection of my performance. The analysis also identified areas in greatest need of improvement, which showed progress since 2006 but continued low scores, such as questions on dealing with poor performers who cannot or will not improve (28 percent positive), promotions based on merit (28 percent positive) and differences in performance are recognized (34 percent positive). 3 Coast Guard. In its review of high and low 2011 FEVS responses, the Coast Guard identified employee responses to two questions that warranted action planning items—(1) How satisfied are you with the information you receive from management on what’s going on in your organization (53 percent positive) and (2) My training needs are assessed (51 percent positive). 4 The Coast Guard officials did not provide any additional FEVS analyses that were used to inform action planning. 3 CBP noted its three FEVS scores as low, but CBP’s scores are not substantially lower than the governmentwide scores. In the 2011 FEVS, the governmentwide average for the three questions were: dealing with poor performers who cannot or will not improve (31 percent positive), promotions based on merit (36 percent positive) and differences in performance are recognized (36 percent positive). 4 The Coast Guard’s scores on these questions are not substantially different from the governmentwide averages (51 percent positive on information satisfaction with information received and 54 percent positive on training needs). However, according to an OCHCO official who monitored the Coast Guard’s action planning in previous years, these questions were addressed in the Coast Guard’s action plan because they have an impact on other low-scoring items and were important for employee satisfaction. Page 67 GAO-12-940 DHS Employee Morale Appendix IV: Selected Components’ Data Appendix IV: Selected Components’ Data Sources for Evaluating Morale, Other than the Federal Employee Viewpoint Survey Sources for Evaluating Morale, Other than the Federal Employee Viewpoint Survey Component Data source Purpose Summary of results and how used U.S. Immigration DHS exit survey Identify why employees leave the The number of exit survey respondents and Customs agency and where they are going. from ICE was too low to identify any Enforcement results and have not been used to address morale as of June 2012, according to ICE officials. Federal Organizational Last conducted in March 2012, the The survey showed low employee Climate Survey (FOCS) and FOCS is a data-gathering tool for perceptions of ICE as an organization focus groups addressing the extent to which where people trust and care for each employees perceive their other, relative to the federal average, organizational culture as one that according to ICE officials. The results incorporates mutual respect, from the FOCS and feedback from the acceptance, teamwork, and focus groups and individual one-on-one productivity among individuals who are interview sessions are provided to ICE diverse in the dimensions of human program offices with recommended differences. Additionally, ICE conducts strategies to improve the program focus groups and individual one-on-one office’s organizational climate. interview sessions to obtain clarifying information pertaining to the FOCS results and written comments. U.S. Customs and Focus groups Conducted in 2007, focus groups were The focus groups identified employees’ Border Protection launched in response to the 2006 perceived problems in specific work annual employee survey results, which environment areas, such as leaders showed CBP below DHS and lacking supervisory or communication governmentwide averages. skills. Among other things, the issues identified by focus group participants allowed CBP to develop action plans that addressed these issues, according to CBP officials. Most Valuable Perspective Launched in 2009, this survey was In the July 2012 MVP, which solicited online survey (MVP) implemented to solicit employee employee preferences for future CBP opinions on one topic per quarter as a webcasts to employees, employees mechanism for gathering further suggested retirement planning and insights on FEVS results. The MVP financial management as their top two was implemented as a continuation of preferences. CBP’s action plan the CBP focus groups completed in planning process in response to FEVS 2007. results includes consideration of MVP results, according to CBP officials. Page 68 GAO-12-940 DHS Employee Morale Appendix IV: Selected Components’ Data Sources for Evaluating Morale, Other than the Federal Employee Viewpoint Survey Component Data source Purpose Summary of results and how used Coast Guard U.S. Office of Personnel Beginning in 2002, in order to provide OPM’s report to the Coast Guard on Management Organizational the granularity, detail, and reliability the 2010 OAS results identified seven Assessment Survey (OAS) needed to ensure the best strong organizational areas (diversity, organizational value, the Coast Guard teamwork, work environment, adopted the OAS as its primary leadership and quality, communication, personnel attitude survey, according to employee involvement and supervision) Coast Guard officials. The OAS is and three areas for improvement administered to military (active and (innovation, use of resources, and reserve) and civilian personnel rewards/recognition). biennially. Coast Guard unit commanders and headquarters program managers use the OAS to support overall Coast Guard improvement. This improvement is achieved by feeding results of the OAS to Coast Guard Unit Commanders and Program Managers who then use OAS results in conjunction with other information as part of routine unit and program leadership and management. Transportation TSA exit survey Identify why employees leave the Top reasons for leaving overall were Security agency, launched in 2005. personal reasons, career Administration advancement, management, schedule, and pay. Each quarterly report includes actions managers should take to reduce turnover. A real-time reporting system is also available for each airport and office within TSA so managers can gain access to their results and use them to reduce turnover and make improvements, according to DHS officials. Results from the exit survey were also used by TSA officials in updating TSA’s action plan, according to TSA officials. However, the July 2012 action plan did not link exit survey findings to action items. Idea Factory An online tool for gathering employee Results were not available for our suggestions for agency improvement. evaluation. Each week, approximately 4,000 TSA employees log on to rate, comment, or search, or to submit ideas of their own. The Idea Factory team reviews all submissions and uses Idea Factory challenges to implement solutions to issues. National Advisory Council Two-year detail for TSA field Results were not available for our representatives to collect and address evaluation. workplace environment issues Page 69 GAO-12-940 DHS Employee Morale Appendix IV: Selected Components’ Data Sources for Evaluating Morale, Other than the Federal Employee Viewpoint Survey Component Data source Purpose Summary of results and how used DHS Ombudsman Provides informal problem resolution Results were not available for our services with the mission of promoting evaluation. fair and equitable treatment in matters involving TSA, according to TSA officials. The Ombudsman assists customers by identifying options, making referrals, explaining policies and procedures, coaching individuals on how to constructively deal with problems, facilitating dialogue, and mediating disputes. Employee Advisory Each airport and TSA headquarters Results were not available for our Committee has an employee advisory council evaluation. made up of elected members who work on understanding and addressing a variety of workplace issues. Source: GAO analysis of interviews with agency officials and documents provided by DHS. Page 70 GAO-12-940 DHS Employee Morale Appendix V: Comments from the Department Appendix V: Comments from the Department of Homeland Security of Homeland Security Page 71 GAO-12-940 DHS Employee Morale Appendix V: Comments from the Department of Homeland Security Page 72 GAO-12-940 DHS Employee Morale Appendix VI: Comments from the U.S. Office Appendix VI: Comments from the U.S. Office of Personnel Management of Personnel Management Page 73 GAO-12-940 DHS Employee Morale Appendix VII: GAO Contact and Staff Appendix VII: GAO Contact and Staff Acknowledgments Acknowledgments David C. Maurer, (202) 512-9627 or firstname.lastname@example.org GAO Contact In addition to the contact named above, Dawn Locke (Assistant Director), Staff Sandra Burrell (Assistant Director), Lydia Araya, Ben Atwater, Tracey Acknowledgments King, Kirsten Lauber, Jean Orland, Jessica Orr, and Jeff Tessin made key contributions to this report. (441016) Page 74 GAO-12-940 DHS Employee Morale GAO’s Mission The Government Accountability Office, the audit, evaluation, and investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO’s commitment to good government is reflected in its core values of accountability, integrity, and reliability. The fastest and easiest way to obtain copies of GAO documents at no Obtaining Copies of cost is through GAO’s website (http://www.gao.gov). Each weekday GAO Reports and afternoon, GAO posts on its website newly released reports, testimony, and correspondence. To have GAO e-mail you a list of newly posted Testimony products, go to http://www.gao.gov and select “E-mail Updates.” Order by Phone The price of each GAO publication reflects GAO’s actual cost of production and distribution and depends on the number of pages in the publication and whether the publication is printed in color or black and white. Pricing and ordering information is posted on GAO’s website, http://www.gao.gov/ordering.htm. Place orders by calling (202) 512-6000, toll free (866) 801-7077, or TDD (202) 512-2537. Orders may be paid for using American Express, Discover Card, MasterCard, Visa, check, or money order. Call for additional information. Connect with GAO on Facebook, Flickr, Twitter, and YouTube. Connect with GAO Subscribe to our RSS Feeds or E-mail Updates. Listen to our Podcasts. Visit GAO on the web at www.gao.gov. Contact: To Report Fraud, Waste, and Abuse in Website: http://www.gao.gov/fraudnet/fraudnet.htm E-mail: email@example.com Federal Programs Automated answering system: (800) 424-5454 or (202) 512-7470 Katherine Siggerud, Managing Director, firstname.lastname@example.org, (202) 512- Congressional 4400, U.S. Government Accountability Office, 441 G Street NW, Room Relations 7125, Washington, DC 20548 Chuck Young, Managing Director, email@example.com, (202) 512-4800 Public Affairs U.S. Government Accountability Office, 441 G Street NW, Room 7149 Washington, DC 20548 Please Print on Recycled Paper.
Department of Homeland Security: Taking Further Action to Better Determine Causes of Morale Problems Would Assist in Targeting Action Plans
Published by the Government Accountability Office on 2012-09-28.
Below is a raw (and likely hideous) rendition of the original report. (PDF)