(ARCHIVE) Banding: The City of Columbus Experiece
Two points need to be made at the outset: 1) Columbus was not the first jurisdiction to implement banding, and 2) what will be presented is not meant as an indication of how it necessarily should be done, but rather to share how we got to where we are today.
The City of Columbus is the 15th largest city in the country, with a population of around 750,000. The city has over 9300 employees, including part time, 3343 of which are sworn police and fire.
ground: The City Charter, adopted in 1914, defines the basic structure of City government, designating the elected positions of mayor and council, and setting up the various departments needed to ensure the delivery of city services to the citizens. The charter also creates the Civil Service Commission, gives the commission responsibility for maintaining the class plan, and sets some general parameters for the structure of the class plan.
The charter states that there will be an unclassified service comprised of the elected positions of mayor and council, the mayor’s appointees (department directors and various commission members), and a select group of positions working directly for those individuals (mayor, council, directors, commissioners). All others employed by the city are in the classified service, which is divided into two groups – noncompetitive and competitive. The charter further indicates that appointments to competitive positions must be made from an eligible list created via a testing process, and in its original form, the charter mandated that such appointments be achieved using the Rule of 3. That is, the appointing authority was required to select from among the top three eligibles on the list.
The change from Rule of 3 to banding started in the early to mid nineties, when the mayor created an Operations and Efficiency Review Team, comprised of 76 individuals from around the city. The “team” included employees of businesses such as IBM, Bank One, and Ross Labs; academicians from the Ohio State University; attorneys; accountants; some city employees; and the presidents of the city employee labor unions. The mayor’s directive to this team was, “Tell me what’s wrong with my City government.” The team split up and went out to talk with administrators and employees in all of the departments, came back together to compile a list of recommendations, presented the list (organized by department) to the mayor, and then disappeared! The mayor then presented the list to the department directors with instructions to fix the problems.
Among the items on the Civil Service list was, “either modify or eliminate the Rule of 3.” In order to address the list of civil service issues, including the Rule of 3, the commission created a strategic planning committee, comprised of employees from various units and levels within Civil Service. Committee members went out to the departments and talked with our clients about the list of problems, and one piece of feedback that came through loud and clear was that the Rule of 3 was too restrictive. So part of the commission’s strategic plan to “fix things” included a charter change to replace the Rule of 3 with a banding process.
The Executive Director of the commission drafted a charter change proposal, brought the entire commission staff together, handed out the proposal, and then split the staff into small groups with instructions to 1) discuss the proposal and find the pitfalls, and 2) come up with some recommendations for avoiding the pitfalls. The feedback provided by staff initiated a number of revisions to the proposal. The resultant charter change proposal, and much of where we are today, is the direct result of the input provided by staff at that time.
Since the charter change required voter approval, and it was clear that a broad base of support would be required for a successful vote, the Commission staff held a series of stakeholder meetings – first with the mayor, council, the directors, and the unions; then with the human resources personnel, managers, supervisors, and employees in the departments; and finally with citizen groups, business groups, and editorial groups across the city. The end result was voter approval of the proposed charter change in November of 1994.
The approved charter changes included the stipulations that there be no fewer than three bands, that appointments be made from the highest band, that when there are less than five names in a band the next lower band can be certified, and that the uniformed promotional classifications be excluded from banding and continue to use the Rule of 3.
The commission started with 10-point bands for promotional classifications and 5- point bands for all others (except, obviously, uniformed promotional). It quickly became apparent that there were insufficient numbers of candidates in the top band(s), so two policy changes were implemented: First, 10-point bands were utilized for all competitive classifications. Second, where scores were previously standardized with the mean set at 74.29 and the standard deviation set at 8.57 (cut score = 70, one half standard deviation below the mean), the mean was adjusted up to 76 and the standard deviation up to 12. The end result was a more useable 90 band.
Eligible List Administration:
The Class Plan provides a framework that allows for a decentralized human resources (HR) function across City government, i.e., the placement of HR positions throughout the city. The HR Officer is a department-level position that oversees the HR function for an entire department. HR Managers, who report to the HR Officer, are found at the division level, with HR Generalists also at the division level, reporting the managers. The HR Generalist is a key classification for successful use of banding in the city.
HR Generalist is a professional level classification that requires a bachelor’s degree and two years of professional (decision-making level) human resources experience. One of the primary responsibilities of the HR Generalist is to administer the certification list, and they end up serving as a liaison with the commission throughout the hiring process as a result. The HR Generalist usually requests the list, performs preliminary screening (review of applications/resumes), coordinates and participates in the interviews, and is responsible for completing the required documentation before returning the list.
order to assist with list administration, the commission has developed a set of guidelines for use by the departments . The guidelines require the use of disposition codes to indicate the status/resolution of each certified eligible; the use of list notations to indicate date(s) and time(s) of contact attempts, the name of the person making or attempting to make contact, and any explanation needed to support/clarify the applicable disposition code or decision; the appointing authority’s signature on the list; and an admonition that the Commission may verify any of the information provided on the returned list (the Commission does, in fact, periodically verify).
One of the issues raised on the IPMAAC listserve discussion that prompted this presentation was the question of how a jurisdiction maintains the merit system principles within a banding environment. One step in that direction is the implementation of merit system principles on the front end of the process. That is, (at the risk of preaching to the choir) conducting a job analysis to define the job and to determine the important/critical knowledge, skills, abilities, and other characteristics required to do the work; designing the test based on the information provided by the job analysis (testing for the important/critical KSAOs); and, most importantly, using subject matter experts throughout the process, from job observations through data collection, test development, item/test review, and key development. In addition to following those principles, the City of Columbus conducts a minimum qualifications review of all applications.
The test products developed by Commission staff include multiple-choice exams, written work sample exams, performance exams, oral boards, and training and experience evaluations (T&E). The Commission administers both multi-phase exams, including two or more of the aforementioned exam types, and single-phase exams. Each type of exam, except for the oral board, is used for single-phase exams.
How it Looks:
Promotional eligible lists have been excluded from this data set, so Table 1 represents strictly the results of non-uniformed open competitive and uniformed entry-level (police officer and firefighter) examinations administered during this time period. It is interesting to note that the 70 and 80 bands are of similar size and that the average number of eligibles in the 90 band is slightly over 6 (for non uniformed eligible lists).
Two ways were selected to look at the data within bands to see how the banding process has been working relative to racial differences (another issue raised in the listserve discussion). The first was to look at representation of race within the group of individuals being appointed compared to representation of racial groups within the band. That is, what percentage of appointees are black/white/other, and how does that compare to percentage of black/white/other in the band. Table 2 displays this data for each band:
One can see from this data that the representation (%) of minorities within the group of appointees is less than the representation of minorities within the band, but the opposite is true for whites, in all three bands. However, as indicated, the black/white difference is not significant in the 90 band, but it is significant in both the 80 and 70 bands (based on a chisquare test).
The final step of the selection process for non-uniformed open competitive classifications in the City involves a conditional offer of employment followed by a drug screen and criminal background check. If the applicant fails either of these screenings, the conditional offer is revoked and the applicant is coded ‘CS’ (removed by Civil Service for background) in the database. Given that the ‘CS’ applicants have successfully navigated the selection process through testing and the employment interview, and the appointing authority has expressed interest in hiring, I decided to add the ‘CS’ applicants back in – as if they had been hired – to see how they were affecting the results. Table 3 displays this data for each band:
Although adding the ‘CS’ candidates cuts the difference between ‘%-of-Band’ and ‘%-of-Appointees’ for blacks nearly in half (the 90 band difference is 1.1% without ‘CS’, 0.5% with ‘CS’, the 80 band difference is 5.8% without ‘CS’, 3% with ‘CS’, and the 70 band difference is 10% without ‘CS’, 6.2% with ‘CS’), the differences are still significant in the 80 and 70 bands.
The second way chosen to assess banding results was to look at success rates and apply the 80% rule to determine disparate treatment. That is, is the success rate of minorities (% appointed) within 80% of the success rate of whites? Table 4 displays the success rates by band:
As indicated in Table 4 above, there is disparity between minorities and whites based on the 80% rule in both the 80 and 70 bands, but not in the 90 band when comparing blacks and whites.
Again, one might wonder what effect the ‘CS’ candidates have on these numbers. Table 5 displays the success rate, by band, when the ‘CS’ candidates are added back in (received a conditional offer of employment, assumed to have been ‘successful’):
As seen in the table above, adding the ‘CS’ candidates back in has the net effect of eliminating disparity between minorities and whites in the 80 band, but has no mitigating effect in the 90 or 70 band.
The uniformed results were not discussed here primarily because of the nature of the selection process. That is, most entry-level uniformed candidates are eliminated from consideration due to information uncovered in the background process rather than the result of an application/resume review and/or one or more employment interviews. Consequently, the results of the uniformed hiring process are not as germane to the banding issues as are the non-uniformed numbers. However, it should be noted that banding of the entry-level uniformed eligible lists has provided the Safety Department much greater flexibility to get candidates through the hiring process. That is, the background unit can work with large batches of candidates (50 to 100) at a time, and needn’t be as concerned with stragglers. Whereas with the Rule of 3, if a candidate was late completing the process, everyone more than two positions lower on the list was put on hold, with banding, the appointing authority need only be concerned with the cut-off between bands. Thus, they can extend conditional offers to all but four individuals in the band, at which point they can request the next lower band. For informational purposes, the tables for the uniformed entry-level eligible lists can be found in Attachment B.
Is it working?
The downside to what has been presented here is that there is some disparity between minorities and whites, especially in the 80 and 70 bands – a situation that is somewhat, but not entirely, mitigated when the ‘CS’ candidates are considered successful in the process.
On the upside, banding does appear to be working. While the 90 band represents less than 13% of all eligibles, over 25% of all appointments are being selected from that band. Similarly, the 80 band represents just under 42% of eligibles, but nearly 50% of appointments are from that band, and while the 70 band contains over 45% of eligibles, less than 25% of new hires come from that band. Also, although the 90 band represents less than 13% of all eligibles, we hire more individuals from that band than from the 70 band, which represents slightly more than 45% of all eligibles.
We also have received feedback that indicates banding is seen as a positive with the department HR personnel and supervisors. We currently survey the supervisors and/or hiring managers of 15 randomly selected new employees each quarter to determine the level of satisfaction with the hiring process. One of the questions on the survey (see Attachment C) asks if there were a sufficient number of candidates from which to select. Over the past two to three years that we have been conducting this survey, we have received only one negative response to that question.
In conclusion, banding seems to be working well for the City of Columbus. We employ merit system principles in order to develop content valid examinations, use a variety of types of exams to assess relevant KSAOs, and generally provide the appointing authority with a sufficient number of quality candidates from which to select. The development of the class plan has resulted in the placement of professional HR personnel in key positions, which, along with the implementation of a number of controls, helps to ensure above-board administration of the eligible lists. We average just over 6 eligibles in the 90 band, hire more from the 90 band than from the 70 band, and, in general, have received positive feedback from the users. Overall, the implementation of banding has been a positive step for the City.
COMPETITIVE CERTIFICATION INSTRUCTIONS
The following guidelines are to be followed when working a certification list:
ANA Appeared Not Appointed
NRC No Response from Candidate
OTH Other Disposition
CNI Candidate Not Interested
UTC Unable To Contact
DEC Declined Offer
NEP Nepotism Policy
Explanation of Codes:
ANA: This code is used to describe the candidate who was interviewed, but was not selected for the position. A brief explanation may be provided as to why the candidate was not selected. Use of this code does not allow movement to the next band.
APP: This code is used to describe the candidate who has been selected for appointment to the position.
CNI: This code is used to describe the candidate who is not interested in the position. A brief explanation must be provided as to why the candidate was not interested (e.g. dept./div., shift, hours, pay, parking, other employment, etc.).
DEC: This code is used to describe the candidate who is offered the position, but the candidate declines the offer. A brief explanation must be provided as to why the candidate declined the position.
NEP: This code is used to describe the candidate who would violate the department’s internal nepotism policy. Use of this code does not allow movement to the next band.
NRC: This code is used to describe the candidate who has been contacted (at least more than once) and there is no response from the candidate. Indicate HR contact name, date of contact, action taken (i.e., left message w/person or voicemail), and time frame given for response back. Note: A reasonable time frame should be allowed for candidates to respond.
OTH: This code is used to describe any situations that do not fit one of the existing disposition codes. A brief explanation of the circumstances must be provided. Depending on the circumstances, the use of this code may not allow movement to the next band.
UTC: This code is used to describe the candidate who the department is unable to contact. A brief explanation must be provided as to why the candidate could not be reached (e.g., telephone disconnected, telephone number changed, no message capability, or the candidate is no longer residing at the listed residence).
VOI: This code is used when the department wants to void the certification list because they no longer plan to fill the current vacancy.
The following information must be provided for each individual on the certification list:
Contact Date: Indicate the date(s) that the candidate was contacted or an attempt was made to contact and any follow-up dates.
Contact Name: Indicate the name of the department human resources person who made (or attempted to make) contact with the candidate.
Code: Indicate the disposition of the candidate using one of the disposition codes listed above.
Explanation: This space is provided for further explanation of the codes listed above. Any notes, comments or information documented about a candidate should be explained in this space.
Supervisor Certification Survey
On a quarterly basis the Civil Service Commission reviews the level of service we provide to our users. The Certification process is just one of many areas that we review and determine how well the process is working and what ways (if any) we may need to improve. Each quarter we randomly select approximately 15 new employees and their immediate supervisor to survey them on their view of how the certification/hiring process worked for them. The purpose of the survey is to find out if we are providing you with the qualified applicants you need to fill your staffing needs. These surveys are in no way a review of the employees’ job performance; instead we are reviewing the quality of the eligible list from which the employee was hired.
1. How would you rate the competence of the employee hired from this certification?
Poor Marginal Acceptable Good Excellent
2. What factors did you consider in deciding your overall competence rating?
3. Are you happy with your selection? YES NO
4. In general, do you find that there are several qualified applicants on the eligible list or are there not enough qualified applicants on the list?
5. Are there any applicant qualities that you think are lacking from the eligible lists you typically use? If so, please list and briefly explain.
6. Are there any obstacles you face in filling positions (i.e., shift work, parttime)?
7. Is there normally an eligible list available to you when you have a vacancy?
8. Do you have any questions or concerns about the current certification process?
9. Do you normally find someone available at Civil Service when you need assistance?
10. Are there any issues you believe are not being addressed by Civil Service? Please explain.
11. Do you believe Civil Service is responsive?
12. Is there anything you would like to add?
Thank you for your time in completing this survey!