RAISE ALLOCATION POLICY

PDF version of this policy

(Approved: General Faculty, December 6, 2006)

I.      Introduction

This document proposes a new raise allocation policy for the Department of Physics & Astronomy.  It consists of three parts.  Part A, called “Meritorious Contribution”, replaces what had been called the across-the-board raise.  The philosophy of this part is that faculty members contributing at or above the expected level for their rank should receive the full portion of this part of the raise, whereas faculty contributing less should receive a proportionately smaller portion.  The scheme proposed below provides a simple, coarse-grained method for determining each faculty member’s Part A raise.

The philosophy of Part B, called “Meritorious Achievement”, is that achievements and activities exceeding a nominal expected level of performance should be rewarded in the faculty member’s raise.  In our present raise policy, the merit-point system plays the same role as Part B.  Indeed, the system proposed below is very much in the same spirit as the current merit-point system, with new categories added, others removed, and point assignments substantially changed.  The goal of these changes has been to make the merit points better reflect the goals and values of our faculty.  These include (but are not limited to): (a) Improving the research environment of our department, (b) improving the overall reputation of our department, and (c) improving the quality of education we provide.  (Extraordinary service is viewed as indirectly contributing to these goals.)

It is worth emphasizing that no system will provide a perfect measure of merit.  In part this is because some meritorious accomplishments are difficult to quantify.  But perhaps of even greater importance is the realization that there exist among our faculty valid differences of opinion as to the relative worth of various meritorious accomplishments.  Our goal should not be to strive for perfection, but rather to develop a policy that is substantially more equitable than our present policy and that will be deemed fair and reasonable by a consensus of the faculty. 

Finally, Part C, called “Head’s Discretion”, is simply a continuation of the corresponding part of our current raise allocation policy, albeit as a larger percentage of the overall raise pool.  The philosophy of this part is that various circumstances related to raises but not covered by Parts A and B periodically arise.  Examples include (but are not limited to): salary compression and inversion; retention counter-offers; transient shortcomings of this policy to be corrected at its next review; and rare or unusual cases of meritorious achievement not properly covered by the merit-point system in Part B.  This part of the raise-allocation policy permits the Head to deal with such circumstances in an appropriate and efficient manner.  While the Head is given great latitude in the distribution of Part C raises, it is understood that he/she will not use this portion of the raise pool capriciously.

 II.      Part A:  Meritorious Contribution

 For this part of the proposed policy, every job-related activity will be classified as contributing to one of three categories: research, teaching, and service.  An incomplete but representative list of the types of activities falling into each category is given below.  At the time of annual evaluations, each faculty member will rate himself/herself in each of these three categories on a scale of 0-4: 

0 = performed well below the expected level for his/her rank

1 = performed somewhat below the expected level for his/her rank

2 = performed roughly at the expected level for his/her rank

3 = performed somewhat above the expected level for his/her rank

4 = performed well above the expected level for his/her rank

He/she will also provide a justification for each rating.  (This could be done, for example, by directed reference to his/her Faculty Activity Summary.)  The Head will then judge whether or not these proposed ratings are reasonable, and make adjustments (up or down) accordingly.  The faculty member’s overall rating, R, will be the sum of the ratings assigned by the Head for each of the three categories.  Thus, a “meets expectations” rating in each category yields an overall rating of R=6.  If R≥6, then that faculty member will be entitled to 100% of his/her portion of the Part A raise.  If R<6, then the percentage of his/her portion of the Part A raise to which that faculty member will be entitled is (R/6)x100%, i.e., proportionately lower.  The expectation is that most faculty members would have R≥6 and thus earn their full portion of the Part A raise.

One of the virtues of this approach is that a given faculty member can achieve his/her full portion of this raise even if he/she underperforms in one of the categories.  For example, if Abe’s research program has wound down as he approaches retirement age, he can continue to earn his full Part A raise by excelling in teaching and/or service.  Alternatively, consider Beth who is focusing her efforts primarily on establishing a top-notch, world-renowned research program.  She may teach only one course a year (and possibly not all that well), but makes up for her teaching shortfall by far exceeding expectations in research and meeting expectations in service.  Thus she would earn her full Part A raise. 

One might ask, “What about the faculty member who excels in all categories?”  Suppose, for example, Charlie gets 4’s in research and teaching and a 3 in service, so that R = 11.  Why should he get the same Part A raise as Abe or Beth, who each only had, say, R = 6?  The answer to this question is that Charlie will be rewarded for his meritorious achievements in Part B.  The goal of Part A is to assess the extent to which each faculty member has met the expectations of his/her rank.

Another virtue of this approach is that it is coarse-grained.  Faculty members do not have to keep track of every work-related task they perform in a given year.  Rather, they only need to make the case that they fall into one of five rating levels in each of the three categories.  Additionally, a faculty member is not hamstrung by a system that specifies each point-worthy activity a priori, but rather is able to get “credit” for activities that may not have been foreseen when the system was constructed. 

The following are non-exhaustive lists of the types of activities associated with each of the three categories.  The purpose of these lists is to guide the faculty members and the Head in their thinking.  No effort has been made here to assign relative merit to the various activities or distinguish relative merit within a given activity (e.g., invited talks vs. contributed talks).  Each faculty member is expected to show good, reasonable judgment.  Since the Head is the final arbiter of ratings, inflated rating suggestions will undoubtedly result in correction by the Head.

• Research:  Publishing papers; obtaining research grants; giving research talks; funding students and/or postdocs; submitting grant proposals; submitting papers; winning research prizes or awards; obtaining patents; having heavily cited papers
• Teaching:  Teaching courses; earning strongly positive student evaluations; mentoring students in research; winning teaching prizes or awards; contributing to curriculum development; publishing educational materials; writing letters of recommendation for students and/or postdocs
• Service:  Serving on departmental committees; holding a departmental administrative position; serving on college or university committees or governing bodies; engaging in public outreach; serving as an officer or council member of a professional society; participating in conference or workshop organization; editing journals or other professional publications; refereeing papers and proposals; serving on review panels, advisory boards, and the like; writing external letters of reference for faculty positions, prizes, etc.

One issue that is not addressed in this proposal is the meaning of “expected level for ones rank”.  Expectations in teaching, research, and service are each tied to rank in very different ways.  If Part A of this proposal is adopted, the faculty will need to clarify this issue either through a series of illustrative examples or by providing definitions of rank-related expectations in each of the three categories.

The following is a description of how Part A raises are computed once ratings for all faculty members are established.  Define the following symbols:

T_A = total raise pool for Part A raises (see Section VI)

S_i = pre-raise salary of ith faculty member

R_i = rating for ith faculty member, as defined above

k_i = fraction of Part A raise earned by ith faculty member = min(1, R_i/6)

N = normalization factor = T_A/∑k_iS_i

Based on these definitions, the Part A raise earned by the ith faculty member is given by:

                         ∆S_i^(A)=Nk_iS_i                                   (1)

III.      Part B: Meritorious Achievement

The overall structure of this part of the proposed policy is largely identical to our current merit point system.  A list of categories for which merit points are given and their associated merit points is created and agreed upon by the faculty.  At the time of annual evaluations, each faculty member provides the Head with a list of merit-point requests, with justification given for each item on the list.  The Head then judges whether the requests are reasonable and makes adjustments (up or down) accordingly.  In addition, the Head may award merit points that the faculty member did not think to request. 

Where this proposal differs in significant ways from the current merit point system is in the categories for which merit points are given (i.e., new categories are added and some old categories are removed) and in the numerical values for merit points for each category.  In many more cases than before, categories are assigned ranges of merit points, which reflects the reality that not all accomplishments in a given category are equally meritorious.  For some categories, the manner for determining merit points is specified precisely.  This is done when relative merit in that category is (roughly) quantifiable.  However, for other categories, the manner for determining merit points is left vague and somewhat open-ended.  The faculty and the Head will need to use good judgment and try to be as fair as possible in these cases.  Since the Head makes final merit-point assignments, all such judgment calls will be made on an equal footing.

The following is a description of how Part B raises are computed once merit points have been determined for all faculty members.   Define the following symbols:

T_B = total raise pool for Part B raises (see Section VI)

f_p = fraction of T_B assigned on a salary-percentage basis

f_d = fraction of T_B assigned on a constant-dollar basis = 1 - f_p

S_i = pre-raise salary of ith faculty member

m_i = number of merit points awarded to ith faculty member

N_p = normalization factor for salary-percentage portion =f_pT_B/∑m_iS_i

N_d = normalization factor for constant-dollar portion = f_dT_B/ ∑m_i

Based on these definitions, the Part B raise earned by the ith faculty member is given by:

                                                      ∆S_i^(A)=N_pm_iS_i+N_dm_i                                          (2)

The remainder of this section is a listing of the categories for which merit points are given and the merit points (or range of merit points) associated with each.  For organizational convenience, some categories are grouped under one heading with subcategories.  In some cases (especially new categories) discussion and rationale are provided.  No attempt has been made to separate research and teaching categories where they overlap.  For example, papers or grants for research and teaching are treated on an equal footing.

One important guiding philosophy here has been to construct a merit point system that reflects the goals described at the beginning of this document.  Some categories and their associated merit points pertain directly to one (or more) of these goals.  Others pertain more indirectly, in that they reward people who take on essential duties (e.g., extra teaching), thereby freeing other faculty members to further departmental goals.

MERIT POINTS

• Refereed Papers:  Publishing refereed papers is an important part of scholarship and should be rewarded significantly in our raise policy.  Three aspects of refereed publications are singled out for meritorious compensation:

  • Volume:  Despite frequent claims that quantity is not as important as quality in publishing, it is well known that ‘publication rate’ is an oft-cited metric of scholarly productivity.  Let n = total number of peer-reviewed journal articles in the calendar year in question.  Merit points = 3n

  • Prestige:  Publication in high-prestige journals carries significant scholarly recognition.  The faculty will decide upon a list of high-prestige journals, and this list will be updated every 2-3 years.  Let n = number of publications in high-prestige journals.  Merit points = 10n

  • Impact:  Non-author citations to scholarly publications indicate the impact of those publications to the author’s scholarly community.  Let n = # of non-author citations in the calendar year in question, including citations to any scholarly publication in the faculty member’s vita.  These data will be maintained and updated by departmental office staff, with faculty members permitted to correct errors.  Let y = # of years since Ph.D. of faculty member, rounded to the nearest year.  Merit points =   0.5n/√(y)

• Other Major Publications:  A non-exhaustive list of items in this category includes: monographs, book chapters, book editing, textbooks, other published educational materials, manuals, etc.  Citations to these publications are included in Impact above.  Merit points = 5-50 per item (Head’s discretion) 

• External Funding:  This category includes non-institutional grants and contracts from all sources: state, federal, or other.  Funding can be in support of research, instruction, facilities improvement, conferences, public outreach, etc.  Only financial funding is included here.  Other forms of competitive funding (e.g., user time at a research facility) are treated elsewhere.  Let n = sum of all external funding dollars received in the calendar year in question that are credited to faculty member.  (This includes the full amount for single-PI grants and contracts and the appropriate proportion for multi-PI grants and contracts.)  Merit points =25 √(n/$50K)    

• Internal Funding:  This category includes all competitive internal institutional grants (excluding foreign travel grants).  Funding can be in support of research, instruction, facilities improvement, conferences, public outreach, etc.  Let n = sum of all internal funding dollars received in calendar year that are credited to faculty member.  (This includes the full amount for single-PI internal grants and the appropriate proportion for multi-PI internal grants.)   Merit points = n/$5K 

• Support and Mentorship of Students and Postdocs: 

  • Postdocs:  Merit points = 5 for each full-time postdoc

  • Graduate Students:  Because of the dual importance to our department and our graduate program of mentoring and financially supporting graduate students, mentorship and financial support are each separately rewarded.

–    Mentorship:  For the ith student mentored, define n_i = fraction of calendar year mentored.  If mentorship for student is shared with other faculty, then n_i must be scaled by the appropriate fraction.  Let  n=∑n_i . Merit points = 5n

–    Financial Support:  For the ith student supported, let s_i, u_i, and f_i stand for the ninths of support for spring, summer, and fall term of the calendar year, respectively.  Define n_i = (4.5s_i + 3.0u_i + 4.5f_i)/48 = fraction of full 4-ninths annual support.  If support for student is shared with other faculty, then n_i must be scaled by the appropriate fraction.  Let n=∑n_i .  Merit points = 5n

• Undergraduates:  For the ith undergraduate mentored, define n_i = fraction of calendar year mentored.  If mentorship for student is shared with other faculty, then n_i must be scaled by the appropriate fraction.  Let n=∑n_i .  Merit points = 3n

• Competitively Awarded Use of External Research Facilities: A non-exhaustive list of items in this category includes:  Supercomputer centers, ground- or space-based telescopes, particle accelerators, probe sources (e.g., Advanced Light Source, Spallation Neutron Source), etc.  Merit points = 2-15, based on competitiveness of resource awards (Head’s discretion) 

• Invited Presentations: Invitations to give scholarly presentations are strong indicators of the importance and impact of the scholarly work, and thus should be rewarded significantly in our raise policy.  Three types of invited presentations are distinguished:

o       Major conferences or workshops:       Merit points = 8 each 

o       Minor conferences or workshops:       Merit points = 4 each 

o       Extramural colloquia or seminars:       Merit points = 2 each

• Honors, Prizes, and Awards: 

o       Extramural:     Merit points = 5-50 each  (Head’s discretion)

o       Intramural:      Merit points = 5-20 each  (Head’s discretion)

• New Patents:  Merit points = 10 each

• Major Publicity for Scholarly Achievements:  A non-exhaustive list of items in this category includes:  Press releases, articles in professional publications (e.g., in Search & Discovery), news coverage, etc.  Merit points = 2-10 each  (Head’s discretion)

• Journal Editorships:  Merit points = 10 each (editor-in-chief); 5 each (subordinate editor) 

• Major Activities Associated with Conference Organization:  A non-exhaustive list of items in this category includes:  Main organizer, program committee member, session organizer, proceedings editor, etc.  Merit points = 2-10 each (Head’s discretion) 

• Teaching Courses:  Teaching courses is an important part of the educational mission of the Department and should be rewarded significantly in our raise policy.  Two aspects of teaching courses are singled out for meritorious compensation:

  • Teaching Overload:  Merit points for teaching overloads are based on the academic year in session at the end of the calendar year in question.  The teaching overload parameter, n, for a given faculty member is determined according to the department’s teaching-load policy.  If, during the academic year in question, the faculty member also teaches a 1-2 hour seminar course, then n may be augmented by 0.5.  Negative values of n result in a merit-point penalty for not meeting teaching-load expectations.  This situation is only expected to result for additional requested releases, as described in the teaching-load balancing policy.  [Note: The Head is permitted to make suitable adjustments to account for anomalous cases arising from ‘course banking’.]  Merit points = 25n 

  • Teaching Excellence:  If student evaluations—both numerical and narrative—or other indicators are judged by the Head to indicate superior teaching, then merit points may be awarded for teaching excellence.  Merit points = 5-20  (Head’s discretion) 

• Curriculum Development: A non-exhaustive list of items in this category includes:  Major restructuring of existing courses, development of new courses, implementation of major new instructional technologies, implementation of major new course components, etc.  Merit points = 2-20 each  (Head’s discretion) 

• Major Departmental Service:  Includes service to the department that significantly exceeds standard level of service.  Merit points = 2-20 each  (Head’s discretion) 

• Major College or University Service:  Includes service to the college or university that significantly exceeds standard level of service.  Merit points = 2-20 each  (Head’s discretion) 

• Major Service to the Profession:  Includes service to the profession that significantly exceeds standard level of service.  Merit points = 2-20 each  (Head’s discretion) 

• Major Public Outreach Activities:  Merit points = 2-20 each  (Head’s discretion) 

• Miscellaneous Meritorious Research:  Merit points = 2-20 each (Head’s discretion) 

• Miscellaneous Meritorious Teaching:  Merit points = 2-20 each (Head’s discretion) 

• Miscellaneous Meritorious Service:  Merit points = 2-20 each (Head’s discretion)

Many of the merit-point categories listed above give ranges, with the specific number of merit points to be determined by the Head.  In order to give guidance to the faculty and to future policy review committees, the Head will need to provide an annual disclosure statement of how these adjustable merit points are awarded.  This could be in the form of a list of meritorious achievements in all ranged merit-point categories, with the number of merit points awarded for each.  However, care should be taken in the listing to preserve each faculty member’s privacy to the best extent possible.

IV.      Merit-Point Averaging:

In Part B (Section III), the total number of merit points for the ith faculty member—denoted by m_i in equation (2)—corresponds to the merit points earned for the calendar year in question.  An alternative would be to average merit-point totals for each faculty member over a period of, say, three years.  This would have the benefit of smoothing out fluctuations.  If raise pools were identical from year to year, then merit-point fluctuations would be less important.  However, if a given faculty member earns a large number of merit points in a lean raise-pool year and a small number in a fat raise-pool year, then that faculty member would be penalized for factors beyond his/her control.  The simplest way to implement 3-year merit-point averaging would be for the parameter m_i appearing in equation (2) to denote the sum of merit points earned by the ith faculty member in the calendar  year in question and the two preceding it.  For new hires in their first two years, the 3-year m_i value would be obtained by suitably normalizing the total number of merit points earned since being hired.  I.e., for first-year faculty members, their total number of merit points would be multiplied by 3, and for second-year faculty members, their total number of merit points over their two years would be multiplied by 1.5.

V.      Part C:  Head’s Discretion

As described in Section I, this part of the raise pool is assigned at the Head’s discretion to deal with raise issues that either are not addressed or are inadequately addressed in Parts A and B.  This may include, for example, adjustments associated with demonstrable large variations among different research subfields in various measures of productivity and impact, such as publication rates, funding levels, citation rates, speaking invitations, etc.

Let T_C stand for the total raise pool for Part C raises (see Section VI).  Then Part C raises are subject the following constraint:

                                                  ∑ ∆S_i^(C)≤ T_C                                           (3)

where ∆S_i^(C) is the Part C raise for the ith faculty member.  If the Part C raises do not exhaust T_C, then the Head adds the balance to T_B.

VI.      Raise-Pool Distribution Parameters and Computing Raises

As described in this document, raise allocations consist of three parts: Parts A, B, and C.  The total raise pool, T, is divided into these three parts according to:  T_A = g_AT, T_B = g_BT, and T_C = g_CT, where g_A + g_B + g_C = 1, so that T_A + T_B + T_C = T.  Recall that the meritorious achievement pool, T_B, is then further divided into two sub-pools: one assigned on a salary-percentage basis (f_pT_B) and the other assigned on a constant-dollar basis (f_dT_B), where f_p + f_d = 1.  The raise-pool distribution is then determined by three independent parameters: g_A, g_B, and f_d, with the other two determined by g_C = 1 - g_A - g_B and f_p = 1 - f_d.  The proposed values for these parameters, based on polling the opinions of the entire faculty last spring, are:

         g_A = 30%                               f_p = 30%

         g_B = 50%                               f_d = 70%

         g_C = 20 %

Part A and B raises for the ith faculty member are computed according to equations (1) and (2), respectively, and the Part C raise for the ith faculty member is assigned at the Head’s discretion.  The total raise for the ith faculty member is then given by: 

∆S_i=∆S_i^(A) +∆S_i^(B) +∆S_i^(C)

VII.      Policy Review

This policy must be reviewed by a departmental committee in 2-3 years after its adoption, and periodically thereafter, in order to assess its effectiveness and make recommendations for improvement.

VIII.      Examples

The following examples illustrate how faculty members with different work styles might fare under the merit-point system proposed in Section III.  In order to make comparisons on an equal footing, only seven merit-point categories are considered: external funding, graduate student and postdoctoral support and mentorship, publication volume, publication prestige, publication impact, invited presentations (all subcategories), and teaching overloads.  In these examples, teaching loads are assumed to be determined according to the companion policy proposal, including only earned overloads and releases and not additional requested overloads and releases.

(1)   Ann had $200,000 in new external funding and fully supported 3 students and 2 postdocs in the preceding year.  She published 8 papers, 2 of which were in high-profile journals, and had 60 citations to previous work.  She is 25 years past her Ph.D.  In the past year, she gave 2 invited talks at major conferences and 4 extramural colloquia.  According to the teaching-load policy, she earned a 1-course release, and therefore is not teaching an overload.

(2)   Ben had $77,000 in new external funding and supported two students for a total of 6 ninths in the preceding years.  He published 3 papers, 1 of which was in a high-profile journal, and had 28 citations to previous work.  He is 16 years past his Ph.D.  In the past year he gave 1 invited talk at a minor conference and 2 extramural colloquia.  According to the teaching-load policy, he earned a 1-course release, and therefore is not teaching an overload.

(3)   Carl’s grant was not renewed, so he had no new external funding in the previous year.  On expiring funds, he supported one student at the 3 ninths level.  He published 2 papers, neither in a high-profile journal, and had 10 citations to previous work.  He is 16 years past his Ph.D.  In the past year he gave 1 extramural colloquium.  According to the teaching-load policy, he earned no releases or overloads.

(4)   Donna no longer has grants nor applies for them.  She served as major professor to one doctoral student in the preceding year, but obviously did not support him.  She published 2 papers, neither in a high-profile journal, and had 10 citations to previous work.  She is 25 years past her Ph.D.  She gave no invited presentations in the previous year.  According to the teaching-load policy, she earned a 1-course overload. 

(5)   Evan has largely stopped conducting research, apart from the occasional thesis project with an undergraduate major, and he no longer mentors doctoral students.  He has not published at all in the preceding year, but had 12 citations to previous work.  He is 36 years past his Ph.D.  He gave no invited presentations in the previous year.  According to the teaching-load policy, he earned a 1-course overload.

The following merit points would be assigned to these five faculty members:

* If 3-year merit-point averaging had been included, Carl would have fared better, since merit points earned in previous years for his now-expired grant would be included.

Academic Programs . People . Research Groups and Centers . Activities and Events . Departmental News . Course Web Sites . Department and University Info . Athens Area Info . Links of Interest . H O M E