Personnel Policy Board
February 7, 2001
Bismarck, North Dakota
I. Opening Business:
A. The meeting convened at 1:00 p.m. Members present included Penny Miller, Chair; Justice Mary Muehlen Maring; Dion Ulrich; Rita Hannesson; and Judge Simonson participated by conference call. Guest: Robert Bjorklund, Personnel Consultant.
B. Minutes of the previous Board meeting, in the form disseminated prior to this meeting, were approved on a motion by Judge Simonson, seconded by Rita Hannesson.
II. New Business:
The first area, Job Descriptions, Mr. Bjorklund found them to be very concise and clearly written. However, they do not document physical requirements or risks/hazards associated with the positions. Additionally, the essential duties and functions need to coincide with the principal duties of the position. He recommended that the descriptions include a history section outlining when the position description was adopted, when changes were made in the content, when the description was last reviewed or evaluated by the Court.
Job Evaluation System:
Mr. Bjorklund explained that during his review of the current job evaluation system he discovered certain strengths within the system. He stated that the plan documents and explains how the job evaluation factors are to be evaluated by raters; the spreadsheets maintained by the Court documents the points allocated to a job on each of the 15 job evaluation subfactors, however, it is not always clear how the points were arrived at on each of the subfactors; the number of salary grades created by the system appears reasonable in that there are a sufficient number of levels to accommodate differences in responsibility levels across an organization the size of the Supreme Court and the point spread between grades is a constant 99 points and because of the significant point difference, it creates a relatively stable job hierarchy particularly in the lower half of the hierarchy. Mr. Bjorklund explained that during his review he discovered that the system contained some shortcomings. For example, the job evaluation research generally shows that three to four job evaluation factors in most systems differentiate or are responsible for the impact of the final outcomes. If true, the Courts have a system that assesses 15 job evaluation factors and many of these factors have little impact on the final outcome. Further, there is some overlap in several of the job evaluation factors (i.e. measuring similar characteristics), for example, Factor 3 and Factor 4 assess complexity as a measure of number of employees. Mr. Bjorklund said that because job evaluation is more of an "art" than science, a job evaluation system comprised of fifteen different factors lends itself to challenges of objectively and fairness. He stated that many of the factors lend themselves to rater subjectivity due to lack of clear definitions or provide too much rater judgment, which makes many of the factors unreliable. Mr. Bjorklund explained that even if some of the factors are unreliable, the subjectivity does not impact the overall reliability of the grade outcome. Lastly, Mr. Bjorklund explained that in developing a job evaluation, it is a commonly accepted HR practice to assign points either using a geometric or arithmetic approach to assess the job value. His review of the system suggests that there is no clear standard used in the development of the system currently used. He said that this causes problems when jobs are rated higher on a factor and then may actually receive fewer points (i.e. Supervision).
Reliability Testing - Job Evaluation System
Mr. Bjorklund explained that job evaluation systems are expected to produce meaningful and reliable job evaluation ratings and outcomes. He said that reliability tests the questions: will different job evaluation raters arrive at similar rating outcomes for the same job? Will different job raters evaluate job evaluation factors in a similar and consistent manner? He explained that to test these questions he selected eight job descriptions none of which were in the same current salary grade. He then had two separate job analysts independently evaluate each of the eight job descriptions using the Court's job evaluation system. He discovered that the reliability of the factors were rated as follows: five factors were rated "good", one was rated "marginal", one factor was rated "acceptable" and nine factors were rated "clearly unacceptable".
Mr. Bjorklund reviewed the Court's current compensation structure (i.e. pay plan) for general human resource issues and practices. He stated that some of the strengths of the plan include the ability of employees to move through the steps in the salary range in a consistent manner. The average employee currently falls at Step 6 in the 11-Step system; approximately 25% of the employees could max out this year or next; and approximately 22% of the employees are at Steps 1-3, which reflects a relatively normal distribution of employees within the pay program. Mr. Bjorklund said that the compensation plan provides for significant movement and more meaningful increases over time for employees in Grades 2-11; differentials between maximum salaries ranges are generally between 11-16% for Grades 2-11. He said however, that it appears that beyond Grade 11, these differentials are only about 5%.
Mr. Bjorklund outlined some of the weaknesses identified in the plan-- one being that the typical compensation design strategies and practices suggest that positions of lower responsibility level generally should not be afforded the same percentage spread from range minimum to maximum because jobs at lower levels generally do not have vastly different documented performance differences and employees should achieve expected performance levels within a more limited or shortened time frame than positions at higher-grade levels. He said that because of our structure, employees at lower levels receive a greater proportion pay increase than other positions when step increases are granted resulting in the key jobs in the organization (i.e. supervisory, management and expert level positions) receiving lower proportional increases than lower level positions in the Court. This is resulting in pay compression problems at the higher levels.
Mr. Bjorklund stated that he believed that there were sufficient numbers of meaningful weaknesses inherent in the current job evaluation system to warrant its replacement. He said he would recommend a system that offers fewer but more reliable evaluation criteria that can be applied to a wide variety of different and diverse occupations across the Court. He recommended the development of a new employee position description questionnaire and job evaluation manual to replace the current system.
Mr. Bjorklund informed the members that in December 2000 he met with Keithe Nelson and Carla Kolling and presented two different matrixes, CMS and DMB, that could be used to replace the current job evaluation system. He explained the advantages of using the CMS versus the DMB. The CMS is a point factor system like the current system - it assesses few criteria but operates essentially the same. Mr. Bjorklund explained that the DBM matrix is considered very complex and not conducive to the needs of the judiciary. Because CMS is a point plan, the current salary ranges can be utilized, jobs can be reassigned to these ranges, and then the ranges can be updated based upon market data. He informed the members that during the meeting, Mr. Nelson suggested including an additional factor to the CMS matrix which would take into consideration the value of attorneys to the system. Mr. Bjorklund stated by adding the new factor it may have some impact on the reliability of the instrument. However, he explained that if that matrix was chosen, he would sample test the instrument to ensure it was reliable. He informed the members that because of the familiarity of the Classification Matrix System and the fact that it is a system that would better suit the judiciary with the added factor for attorneys, he urged Mr. Nelson and Carla to choose the Classification Matrix System with the added factor addressing attorneys.
The Board expressed concern over the fact that they never had an opportunity to vote on the matrix selected. Justice Maring requested Carla to have the minutes reflect that the Board had not voted on the matrix that was proposed. Dion Ulrich questioned whether the development of the new matrix and its use on rating our positions was an extra charge or if it is covered under the current contract. Mr. Bjorklund explained that it is covered under the current contract. He informed the members that it was his understanding that he was to evaluate the current system, and if the system proved to be unreliable, he was to recommend a new matrix and rate the positions off the new matrix. He stated that there would be no reason to rate positions off the old system if it was determined that the matrix was unreliable. He commented that he would not be able to support the outcome of the ratings using the current matrix if challenged in Court.
Members questioned Mr. Bjorklund what would happen if the Board did not like the results of the new matrix. Mr. Bjorklund explained that the Board could always go back to what they had originally and that they were under no obligation to implement his recommendations.
Justice Maring inquired whether we could employ another consultant to use the matrix that he is proposing. Mr. Bjorklund explained that he is going to provide instruction manuals that outline how to apply the new matrix, but he cautions members about having another consultant use a system that they themselves did not develop. He said that the matrix is intended to be used by the judiciary, and we could not take the matrix that he developed and give it to another agency for their use.
Mr. Bjorklund informed the members that he is ready to move forward with the market analysis. He said he has tentatively identified a list of organizations to be surveyed and had presented those organizations to Mr. Nelson and Carla at the December meeting. He stated that he intends to use a blend of public organizations from within the state as well as those from states with similar demographics to North Dakota and private organizations from within the state. He will also be using other published salary information in his survey.
Chair Miller thanked Mr. Bjorklund for providing the members with a report on the progress of the review of the pay and classification program.
The recommendation submitted by Carla Kolling was approved, on a motion by Justice Maring, and seconded by Rita Hannesson. Dion Ulrich dissented.
The recommendation by Chair Miller to remove the five day per year limit on education travel was approved, on a motion by Justice Maring and seconded by Judge Simonson.
For informational purposes, Chair Miller requested Carla to forward the amendment to the Council of Presiding Judges prior to forwarding it to the Supreme Court for consideration.
Members agreed to remove the proposed Section G-Family Medical Leave With Pay from the draft.
The policy as amended was approved on a motion by Justice Maring and second by Rita Hanneson. The motion included a request to forward the proposed amendment to the Supreme Court for consideration.
Judge Simonson recommended adding the words "without a prescription" immediately following the words illegal drugs in the first full paragraph.
The policy was returned to Carla to incorporate the revisions. The policy will be revised and then forwarded to the members for review.
The meeting adjourned at 3:45 p.m.