BAND TOGETHER
COMMUNIQUÉ 

Time Critical

Union Notices

Important Update Regarding the 2008 GAO Employee Feedback Survey

Diversity Month Activities Planned for June 17

Union Contacts

Get Active Action Center – Give the IFPTE Feedback

DISCLAIMER

 --------------------------------------------------------------------------------

Important Update Regarding the 2008 GAO Employee Feedback Survey

The GAO Interim Council was officially notified by the Office of Workforce Relations on Thursday, May 22, that GAO planned to launch the 2008 GAO Employee Feedback survey by mid-June.  At that time, we expressed a number of concerns that employees have had with the survey historically, including confidentiality, appropriateness of some of the questions, some ways the results have been used by management, comparability with other federal agency surveys, the ethics of matching survey results with individual ratings (done once a few years ago), and the House oversight subcommittee's belief that a separate, independent employee survey was necessary in order to understand employees' views on some issues that are critical to GAO employees.  We also asked HCO why we were not alerted to the planned time line until only a few weeks prior to the launch of the survey, effectively eliminating the chance for substantive input from employees.

Workforce Relations staff, speaking for GAO management, have emphasized that there is not time for any changes this year but that there may be options for such discussions next year.  They explained that the employee survey must go forward, without substantive input from employees, in order to gather information to include in our annual Performance and Accountability Report, among other reasons, although last year the survey was launched on August 1, 2007.

Some of our concerns regarding the GAO Employee Feedback survey include: 

• The lack of appropriate controls in place to assure the confidentiality of responses, since the analyses are done by GAO using record-level data and the survey asks respondents to provide extensive demographic information, including mission team, office location, band level, PDP status, gender, race/ethnicity, number of years at GAO, and age; 

• The pressure some staff have experienced, particularly field office staff, directly and indirectly, to provide “positive” survey responses due to concerns about sustaining teams’ resources and the presence of teams, as well as keeping field offices open, raising questions about the quality of the data; 

• The potential misuse of survey responses when broken out by mission teams or race/ethnicity within the field offices; 

• The lack of survey questions that address critical issues for GAO employees, including GAO’s performance management system, for the purposes of comparisons over time and with other federal agencies; 

• The lack of transparency and trust in how some of the survey results are reported and actually used in evaluating and making decisions about resource allocations.

For these and other reasons, the GAO Employees Organization believes that unless these concerns are addressed, it will be difficult to have a high level of confidence in the results of the GAO Employee Feedback Survey.

While management has advised us that they will not delay the survey to give us time to provide meaningful comment on the content, we believe there is time for changes to how the survey data is managed and analyzed.  We have provided GAO management with five specific recommendations for the 2008 GAO Employee Feedback Survey:

1. The issue of confidentiality and trust has been a constant one for a number of years.  The cover letter on the current survey states that "procedures [are] in place to prevent anyone in GAO ... from knowing how specific individuals responded."  Yet identifying information is asked in such detail that it would be quite simple for those at GAO who have access to the data to know precisely how certain individuals, especially in smaller field offices and teams, responded.  A simple solution would be to have outside contractors analyze all of the data.  We recognize this would make the survey more expensive but feel the added controls would go a long way toward assuring respondents of complete confidentiality, which is essential to ensuring full and honest employee participation. 

2. Complete transparency in the analysis and presentation of the survey results is necessary to adequately assess the results.  Issues that have arisen in the past that should be addressed include: 

• The number of non-responses for each question should be summarized in tables, including non-responses on demographic questions 

• Presentation of overall survey non-response by respondent demographics (grade/band, race, sex, etc.) 

• Since some of the survey items are used for comparison with OPM's Federal Human Capital Survey (FHCS), we should consult with OPM to make sure that, technically, we are asking the questions in the same way (e.g., forced choice or not for the items in question) and whether responses to questions are comparable given differences in the survey question contexts and question ordering. 

• Since GAO is trying to build an exemplary performance appraisal system, it may be instructive to incorporate some of the questions from OPM's FHCS that deal with employee opinions on the fairness, objectivity, and impact of this process.  This would allow for baseline comparisons over time and with other agencies. 

3. It is not clear whether analyses that have been presented in the past use self-reported demographic information directly from the survey questions or whether this information is added from administrative records. For instance, we are told that Managing Directors receive their team survey results broken down by demographics such as band level and years of service yet such survey questions generally have lower response rates.  Thus, we are led to believe that this is derived from administrative records. Complete transparency is essential here -- either tell us you are using GAO administrative records or report the number not answering the question. 

4. When survey responses are broken out by mission team within field offices, there is increased risk of individuals being identified and misuse and abuse of the results.  Survey analyses should be done to show results for mission teams (staff in HQ and field offices) and for field offices (all staff within the field office).  There should be no analyses conducted that break out mission team staff within particular field offices. 

5. Because of ongoing concerns that some managers have urged employees to provide positive responses to survey questions, we request that GAO instruct all senior and mid-level managers, including those in the field offices, not to discuss anything about the survey with employees that might influence their responses, other than to ask them complete it.  We also request that this instruction to managers emphasize that there will be no negative repercussions to employees, either individually or collectively in teams or field offices, for providing any valid responses or for electing not to complete the survey. 

We will keep you posted on new developments with the 2008 GAO Employee Feedback Survey as we learn of them, including discussions we may have with GAO.  If you have comments you would like to share about this Communique, including your experiences with GAO Employee Feedback Surveys, please send an email to us at ICquestions@gaoanalysts.org.

Diversity Month Activities Planned for June 17

On June 17th, Interim Council member Jackie Harpp (Diversity - Black/African American) will participate in a moderated panel discussion on the Ivy Group's final report on disparities in performance appraisal scores between GAO's African American staff and Caucasian staff.  The discussion will take place 1:30 p.m.- 3:00 p.m. in the GAO auditorium and will focus on the broader implications of the study's findings, including how other groups may be affected by proposed recommendations to address ratings disparities. Gene Dodaro will kick off the panel discussion with opening remarks.