The college lists the following as “Hallmarks” of the compensation plan: Transparency, Predictability, Consistency, Equity, Excellence, and Accountability.
A Community Forum on Guilford College Faculty Life
I was part of a meeting on Friday with several members of college leadership, our compensation consultant Christine Riley, Jim Hood, and Natalya Shelkova (from compensation committee).
In that meeting, we discussed whether the college has been using the wrong number for setting faculty salary targets over the past three years. We know that in 2018 and 2019, and maybe in 2017, the college used the median of individual salaries of professors at our peer institutions rather than what our policy calls for, which is the median of mean faculty salaries at our peer institutions. Natalya, Jim, and I were in agreement that the median of individual salaries is both (a) not the number specified in the policy, (b) not a number that works in our target formula, and (c) a number that tends to be significantly lower than the number specified in the policy, especially for full professors. I have confirmed this gap by comparing our target salaries for this year, and the CUPA data on which they were based, with the median of peer institutional mean salaries (the number we’re supposed to be using) from AAUP. I shared that information with you around the time of the opening faculty meeting in this document.
In Friday’s meeting, Christine Riley confirmed that the intent of the policy was to use the median of institutional means salaries among our peer institutions, something that the faculty members of Compensation Committee have also confirmed.
It is clear that for each of the last three raise cycles, our salary targets at most ranks were set too low, sometimes far too low (e.g. ~$7500 for full professors in 2018). We now know that this was not merely bad data, as was cited last year, but instead it was an error in implementation. The outcome of that error has been to deny equitable raises to faculty, particularly to full professors, under the compensation policy. The money that would have gone to faculty has instead been allocated to staff, whose targets were set correctly.
I raised this issue in August 2018, immediately after the raise letters came out with very low targets for full professors. I had the magnitude of the shortfall figured out by September 2018. The college delayed in discussing this until February 2019 and then declined to take any reparative action at all. The only action it took was to change our peer group to a larger and different set of peers. Using a different peer group did not fix the error, and in July 2019, the college again set targets far too low, again without running the salary target numbers through Clerk’s Committee as indicated should be done under the policy. This review by Clerk’s has never happened in the history of the policy, despite being written into it.
In 2018, I did not know that we were asking CUPA for the wrong number. All I knew was that the targets were way too low, and I accepted that it was possible that changing the peer group might help to address what had been very inconsistent and skimpy data from CUPA for 2017 and 2018, although I still felt that the college should do something to correct for the impossibly low numbers it had used in those years.
In July 2019, Jim Hood and I reviewed the raw CUPA data in the human resources office. We discovered that the college was using not just incomplete or flawed data, but had in fact asked for entirely the wrong kind of number from CUPA. It later became clear that it had been using the wrong kind of number for at least 2018 and 2019, and perhaps also for 2017.
At the meeting on Friday, the administrators there seemed to accept that they had made these errors in implementing the salary policy for faculty. I hope they will make a public statement to that effect. After that discussion, which occupied the first half of the meeting, there was some discussion about two concerns. The first was whether the correct institutional data could be delivered by CUPA, and the second was whether using a median of means for faculty targets was fair to staff, whose targets are set based on individual medians.
The first concern, about CUPA data, is largely immaterial – if we cannot get the correct numbers from CUPA, we certainly can from AAUP, which is open-access and much more robust in reporting than is CUPA. CUPA has not served us well in any of the three raise cycles, even setting aside our asking it for the wrong numbers, because they consistently have a small minority of our peer instiutions reporting, and the numbers we’ve gotten from them have varied a great deal (unrealistically so) between years. If we can’t get CUPA to work, we can just use AAUP.
The second concern, about potential fairness questions between faculty and staff, is also immaterial, at least for the past three raise cycles. The college’s compensation committee, comprised of faculty and staff, agreed on this policy. The college published it and implemented it, but it then used incorrect numbers that significantly undercut faculty raises and compensation, three times. That is a wrong that should be righted, and I believe it to be legally actionable, although I hope we do not need to go that route.
Even if the second concern (fairness) doesn’t matter for the past three raises, it does matter for the future of the policy, because we want the faculty and staff formulas to be fair, and we want both groups of employees to reach equitable peer targets. However, I actually don’t think staff salary targets would change much if they were based on medians of institutional means rather than medians of individuals. This is because there are many, many kinds of staff positions, and most of them are unique or in very small numbers at institutions our size. Therefore, a median of individuals will be essentially the same as a median of institutional means for these positions. This is in contrast to faculty positions, where there are only four ranks we track, and they contain many faculty, who are often in those ranks for a long time, with salaries that can vary by tens of thousands of dollars at Guilford (or even by hundreds of thousands at some of our peers). I would be happy to check my prediction here if I am given the data to work with, but my sense is that there is no fairness issue in this area. Regardless, we should have followed the policy we have on the books, and we should keep following it correctly until we agree to change it. The fairness issue is, I think, a red herring at this point, given that we had a policy and didn’t follow it in a way that was already unfair.
I don’t know what happens now. It’s my strong feeling that something should be done to repair the harm caused by this error, and that the repair should be more significant and more immediate than just “we’ll do better in the future,” because all tenure-track faculty have been and continue to be negatively impacted. I would like to talk to people about what the nature of that repair should be, because while I can represent the math well, I have no current role in representing the faculty.
I am attaching five slides I prepared for the Friday meeting which lay out the argument I made. Natalya also prepared slides on the median/mean issue and discussed it very effectively. If you need help interpreting what I’ve got here, please ask questions in the comments. There’s a caption under each slide that gives some context and guidance.
I’m getting a lot of mail about the revised letters that Frank posted. Rather than answer them all separately, let me put some of that here.
The graph below shows three different models applied to three different faculty ranks.
The 2018 model (heavy dashes) used a $300 per year annual increment, as we’ve been doing since the beginning of the compensation plan. It also used salary baselines that were far too low because of spotty non-representative data we got from CUPA. Had we used robust peer data, such as from AAUP, all of those 2018 lines would be significantly higher.
The wrong 2019 model (faint dashes) was the source of the first contract letter that got sent out this year. It used updated baselines based on our bigger group of peers, and it was higher at all ranks, because the new CUPA data didn’t suck and was more representative. This wrong model used an annual increment of $300/$400/$500 for assistant/associate/full, which is why the lines are steeper than the 2018 model. This was one of the options discussed by Compensation Committee last year, but they went with a higher annual increment. That’s the error mentioned in the new contract letters.
The correct 2019 model (heavy solid lines) uses the appropriate $300/$500/$700 annual increment that Compensation Committee agreed on. That produces lines that are even steeper at higher faculty rank. However, they aren’t higher or lower than the “wrong model” lines, because the baseline for each rank got recalculated (appropriately) to reflect the average service time. The point where the wrong (light dashes) and correct (heavy solid) meet is the peer median value we use to set targets and represents the average target salary by rank. Because there was no change in peer median between the wrong and correct models, they cross at that point. The assistant lines aren’t any steeper and didn’t change between letters, because they still use the $300 annual increment.
The correction means that if you are associate or full rank, if you have less credited experience than the average amount for your rank, you likely saw a drop in target between the first and second 2019 contract letters, while if you have more experience than the average amount for your rank, you likely saw an increease in target. That’s not wrong – it’s just the model being applied with the corrected numbers.
The point of changing the increments at associate and full was to reduce compression at those ranks. For example, the old model only had about $9000 difference in target between when you got promoted to full and when you retired 3+ decades later at the end of a 45-year career. That’s a much narrower and more compressed range than most other schools have, and it meant that our new full professors had really high targets, almost as high as our late-career full professors. By increasing the slope, we spread out the range of incomes much more, to better match the shape of most careers in academia and in other fields.
Importantly, these changes in increment were not intended to move the line up. When they changed the annual increment in the corrected letter, they also had to decrease the base to ensure the mean full compensation stayed in the same place, which is what the model intended.
There is another issue with the peer data they used this year, which affected the baselines they set. I don’t think we did that correctly. I’ve raised this issue with Frank, and I’m waiting on a response. If we did it correctly, we’d likely see a fairly significant increase to some targets, and we would be where we intended to be at this point rather than artificially low (for the third year in a row).
UPDATE: The graphs below aren’t entirely accurate, and they calculate salaries for upper ranks that are too high. Please wait on an explanation from the Provost.
My previous post detailed how my salary target was not where it was supposed to be given the President’s announced salary base values. Here’s some more information.
We should be seeing a pronounced increase in salary targets this year, because last year, due to non-representative data from CUPA for our peer group, our targets were incorrectly set very low relative to our peers for full and associate rank. Jane’s announced base salary values represented this change. For many of us, that’s not what we saw in our contract letters sent out today.
Here’s what should have happened from one year to the next:
This applies two new developments to our model.
One is the shift in baselines due to two factors: (1) using a better peer group this year, and (2) that peer group’s salaries going up as they nearly always do by about 2%. Those two factors are what makes the 2019 lines higher at all ranks than the 2018 lines. Factor (1) applied to full and associate ranks, while Factor (2) applied to all ranks.
The second is the shift from a flat $300/year in target for all tenured/tenure-track faculty to a graduated $300/$500/$700 for assistant/associate/full. That’s what makes the 2019 lines for the upper ranks steeper than the 2018 lines.
My new salary target was set at $82,180 by the Dean’s office, but I think that’s wrong. Here are my calculations:
- Jane reported the base salary for full at $63,988 in her e-mail.
- I get the $5,000 for terminal degree, so that’s $68,988.
- I have worked at Guilford 22 years. They didn’t count my earlier experience when they calculated it back in 2016, which was ok with me. The full professor rate for experience is now $700/year, and 22 x $700 is $15,400. Add that to the previous and I’m at $84,388.
- The disciplinary bonus for Geology was set at $1,400. It’s possible that they have started to taper these values to zero as we approach our targets as the original plan stated. However, we did not do this in the 2018 raises, and Frank seemed steadfastly against doing that whenever we discussed it. If I’m still supposed to get the full disciplinary bonus, that’s $1,400 more, or $85,788.
I’ve written to Frank about this apparent error, which amounts to $3,608 lower target salary than I should have had. It’s possible there’s some hidden explanation I don’t understand, but as far as I know, nothing changed this year other than the base and the annual service adjustment.
Based on the numbers they gave me, it looks like the raises they applied this year amount to 11.4% of the gap between previous salary and target salary, which means the actual raise they gave is potentially about $400 lower than it should be, or 34% lower than it should be. That amount and percentage would change if they’ve done the same thing for all faculty because of the constraint of the pool size.
Of course, this doesn’t even get into the college setting the faculty salary targets way too low in previous years, which the college has yet to address or even comment on.
If you’d like to check your target, I’ve updated the salary calculator here:
I asked for Guilford’s most recent IRS Form 990. I do this for several reasons, but I got started with the 990’s after the administrator bonus fiasco at the end of the Chabotar administration in order to have a look at our reported compensation for administrators. This most recent report allowed me to track the impact of the January 2017 raises for some of our administrators. I’ve shown the raises below, compared with changes in faculty compensation as reported to the AAUP for the same years.
Data sources for administrators are 2016-17 and 2017-18 Form 990’s, which cover our fiscal year. Because of a change in our fiscal year start date, the 2016-17 report only covered 11 months, so I prorated the salaries reported in that document to 12 months before calculating the raises.
Data sources for faculty are our average salary by rank for the academic years 2016-17 and 2017-18 as reported to the AAUP by the college.
The 2016-17 numbers include part of the period covered by the large January 2017 raises. I think those raises are included in the 990 reports, because those report total dollars. I am not sure if they are included in the AAUP faculty salary reports, but I don’t think they were, because they show very little change from 2015-16 numbers. If the January 2017 raises were included in the 2016-17 data for administrators and not for faculty, then the difference between faculty and administrators is even more stark, and the faculty raise percentages should be decreased by about half to make them comparable.
For the administrators, I only included people who were working for the full year in both years and whose income was reported on the 990 in both years, with two exceptions:
- Todd Clark, whose reported compensation dropped between the two years for reasons unknown to me, and
- Kent Chabotar, whose compensation was set at $100,000 for five years following his presidency under the terms of his initial presidential contract
Barbara Lawrence is a special case, because her position changed from faculty to VP between those two years, so the high raise percentage shown probably mostly represents her new position. Her faculty compensation was high enough (>$100K) to be reported on the 990 in the earlier year, likely in part because of her involvement in the prison education project.
Analysis and Notes
With the exception of Jimmy Wilson, all of the administrators listed received higher raises (by percentage) than all faculty ranks. Jimmy’s raise percent was higher than all faculty ranks except Associate.
In dollar terms, all of the raises were higher than average faculty raises, in some cases much higher. Administrator raises for those on this list ranged from about $8,000 to $26,000, while the faculty averages changed by $2,000 to $4,300.
Jane Fernandes’ salary is set by the board and is not included in the compensation plan. The board granted her a raise during this time.
If the compensation plan was applied appropriately to all of these individuals other than Jane, for whom the plan does not apply, this outcome (higher raises for administrators than faculty) is possible if the administrators listed were all farther behind their targets than the average faculty member. I don’t know if that is true, because I don’t know the targets used for administrators, but do we know the faculty are far below their targets, so the administrators would have to be even worse off. That doesn’t seem entirely likely, given that faculty were at or near the 20th percentile among peers. However, setting targets for administrative positions is tricky, because the administrative positions we have don’t necessarily equate to similar positions at other institutions.
The raises for faculty were affected by the bad data used to set targets, as I’ve discussed earlier this year. That error was likely not made for administrators. All faculty ranks were affected by the improperly low targets in the first round of raises in January 2017, included in these data. However, the effect of these low targets is not big enough to account for the full difference in raise percentages between faculty and administrators.
Some faculty members who were farther from their targets received larger percentage raises than those who were closer. For example, my personal percent raise during this period was 8.9%, higher than the full professor average and higher than some of the administrators.
There is the potential for an apples-to-oranges issue here, because the faculty percentages are based on groups of 20-30 people, while the administrator raises are individuals. However, assuming the faculty salaries are reported correctly, that should produce no significant systematic difference in the numbers, with the possible exception that retirements, departures, or promotions of many high-salary faculty at any rank could depress the average salary for that rank to some extent, though likely not too much. When looking at the numbers, remember that the individual faculty raises are a range centered around the percentages reported, and are probably on average a little higher than the reported figure due to retirements/departures.
Back in February, I calculated an update to the Category IIB percentiles we used to publish in our Factbook. That post is here.
The AAUP has now posted data for the 2018-19 academic year. This year includes both the recent rounds of raises in January 2017 (in effect for the last half of 2016-17, although I don’t think that our AAUP salary reporting included the raises until 2017-18) and August 2018 (in effect presumably for the 2018-19 reporting).
I’ve updated each graph I made for the earlier post. Those are below, with interpretation:
Here are my interpretations of the additional year of data. Please see the earlier post for a more complete analysis.
- Of the four ranks at Guilford tracked here (heavy solid lines), all showed a modest increase in 2018.
- Nationally, the AAUP median for Category IIB schools increased more than Guilford’s raises for Associate and Assistant, which means we lost some ground against the median of our peer group at those ranks. This is not a surprise given the small size of our raise pool last year.
- Nationally, the AAUP median for Category IIB schools increased less than Guilford’s raises for Full and Instructor, which means we gained ground against the median of our peer group at those ranks.
- For Instructors, our reported modest increase in salary contrasted with the drop in the national IIB median Instructor salaries.
- For Full Professors, our reported increase in salary coming in slightly stronger than the increase in the median may have to do with the large gaps full professors had from their targets, which meant they may have gotten more of the raise pool under our formula than others who were closer – i.e., we’re still pretty far behind our peers, but we filled in a little of the gap. I suppose we could have had fewer retirements or departures than other schools, also – we lost so many folks in 2016-17 that we have fewer left to lose now, which might have elevated our numbers somewhat relative to others. This is all speculation, though.
- Fundamentally, in 2018-19, we appear to have more or less kept pace with other IIB schools in terms of dollars, but we did not make progress on closing our sizable gaps with them except at Instructor rank, which we only did because of a national decline in Instructor pay. Though not the best outcome, this is better than period from 2010-2016, when we stagnated or even lost ground in real dollars (this was even worse if you take inflation into account, which was a total of about 10% over that period).
Here is the impact on our percentiles compared to other IIB schools:
Remember that these percentiles are tracking a different thing from the dollar values above. The percentiles are only about our ranking relative to other similar schools, while the first graph is raw dollars.
Here are my interpretations of the additional year of data on percentiles. Please see the earlier post for a more complete analysis of the history.
- At all ranks except instructor, we lost ground in terms of percentiles.
- This was most pronounced at Assistant rank, which grew more strongly nationally than other ranks.
- That means that, unlike 2017-18, when we made significant upward progress in our ranking, other schools passed us, although this didn’t wipe out all of the progress we made with the January 2017 raises.
- We are now back to similar percentiles from 2013, when we were already in the midst of our very steep decline, as opposed to our heyday in 2008-09, when we were still well below where we’d set our goals at the time but (unbeknownst to us) at easily the highest level we’d experience for the next decade.
- If we reach the stated goal of our compensation plan, we will be up near the 50th percentile for IIB, which is close to both the original Compensation Plan peer group of 46 schools and to the revised peer group of ~350 schools proposed this year.
Here are the raw numbers and last year’s percentage change in table form.
|Full||$ 74,700||$ 76,200||2.0%|
|Assoc||$ 60,000||$ 60,500||0.8%|
|Asst||$ 56,200||$ 56,400||0.4%|
|Inst||$ 46,300||$ 47,100||1.7%|
|IIB Medians||2017-18||2018-19||Percent change|
|Full||$ 87,300||$ 87,800||0.6%|
|Assoc||$ 71,300||$ 72,100||1.1%|
|Asst||$ 61,200||$ 62,300||1.8%|
|Inst||$ 53,500||$ 52,500||-1.9%|
We need to have raises at least as big as last year’s not to lose ground. If we want to regain some of what we’ve lost, or even (heaven help us) reach our stated goal, we’ll need to have raises that average more like 4-5%.
Not all of that needs to come from new revenue. Every year we have some more senior, more highly compensated folks retire, and if they’re replaced, they are usually replaced with younger, lower-compensated folks, which creates room for raises for remaining faculty without adding to the overall budget.
That’s how I see it. Let me know if you have questions.
I spent some time Tuesday collecting some numbers and looking back at Guilford salaries. My goal was to recreate some of the numbers that Guilford used to report in its annual Factbook. That Factbook no longer exists, as far as I can tell, and the reporting of salaries compared to standards fizzled out in about 2013. So, it’s been a while since we had a comparison like what we used to have.
The target group we used to use for salary comparisons was Category IIB, a classification that includes institutions that grant primarily baccalaureate degrees. Guilford far exceeds the minimum degree requirements for this category and is nowhere near the postbaccalaureate degree requirements for Category IIA, even with our new Master’s program, so it’s clearly where we belong. More on these categories is available here.
I collected and processed the last seven years of the AAUP Annual Survey, which collects reports from most higher education institutions and produces an annual report. These reports are here. I selected only the Category IIB institutions in these reports. Guilford consistently reports to the AAUP like every good institution should.
I also have a record of many years of Guilford’s self-reported salaries and comparative percentiles, which I collected from the Factbooks back when they were available. I have those covering the years 2007/8 to 2013/14. These included faculty salaries at different ranks (also reported to AAUP) and a calculation of percentiles compared to the Category IIB pool. These reports also included percentiles for two categories of staff – administrative and support staff. I don’t have the data to extend these staff reports to today, and I’m not sure what data sets were used to created the reported statistics.
This first graph shows faculty salaries at four different ranks (indicated by colors), for Guilford (heavy dashed and solid lines) and AAUP Category IIB (thin dotted lines).
The dashed lines and the solid lines (representing Guilford salaries) overlap for three years, 2011-2013. As you can see, even though one was reported by Institutional Research and the other produced by me, they agree very closely, except for one blip in assistant professors in 2012. That gives me confidence that I’m doing this mostly right.
Here are insights I draw from these data sets:
- Guilford’s faculty salaries rose from 2007-2009, then stagnated or dropped from 2007-2016. At the higher ranks, average salaries in unadjusted dollars dropped, probably due to retirements coupled with very limited raises. If we scaled salaries to cost-of-living, this drop would be even more severe, as there was a mostly consistent inflation rate of 1-3% during this time.
- Probably in response to new hiring at nearer-to-market salaries, associate professors narrowed their gap with higher ranks as time went on.
- The raises in the middle of the 2016-17 year are clear at the right side of the graph. The smaller summer 2018 raises are not represented in the time range covered.
- During the period of salary stagnation, our Category IIB peers showed consistent growth in median salaries at all ranks.
- The January 2017 raises (at least as represented here), though much bigger than in recent years, did fairly little to close the widening gap between Guilford and peer institutions, as they were only a bit higher than the overall Category IIB increase.
The values for the two categories of staff are no longer reported, so they only extend to 2013. Again, there is mostly agreement between my numbers (covering 2011-2017) and the Factbook numbers (2007-2013) where they overlap, except for one year of assistant professor numbers.
Some observations from these data sets:
- As one would expect from the dollar values shown in the first graph, our percent rank dropped at all faculty ranks.
- Instructors, who were nearly at the Category IIB median in 2011, showed the greatest decline in percentile, although not the lowest overall percentile.
- Associate professors hit bottom the hardest, reaching a low-water mark of 13th percentile in 2016-17. For reference, here are the five schools on either side of us on a ranked list of IIB associate professor salaries that year:
- Saint Joseph’s College
College of Southern Nevada
Lewis-Clark State College
University of Montana-Western
Dakota Wesleyan University
Ouachita Baptist University
- Saint Joseph’s College
- Staff percentiles were nearly always higher than faculty percentiles when reported. This is likely due to several factors, such as (1) stronger market pressure for hiring new staff closer to peer salaries, (2) much higher rates of turnover, such that staff positions reset to market more frequently than faculty positions, because faculty tend to be much less mobile, especially following tenure.
- Even with these structural factors, it seems likely that institutional leadership, budgeting, prioritization, and decision-making played a role in the growing disparity between the groups. My guess is that the percentiles for staff declined with faculty, but that they stayed stronger for the staff groups than faculty groups during the lean years. We lost more staff members than faculty due to layoffs during this time, but the reported salary numbers are an average that shouldn’t change with group size alone. I can’t think of a compelling reason why the average staff salaries and their percentiles would decline much even with the layoffs, as the layoffs were not focused among highly paid indiviuals.
- Again, the raises in January 2017 provided a significant boost to faculty, but not enough to erase the previous eight years of wage stagnation and decline.
- There were raises in August 2018 also, but those raises were small, not much more than inflation, and my guess is they would lead us to lose ground to our Category IIB peers during the 2018-19 year, as they likely grew more than we did. It will be interesting to see the 2018-19 numbers to see how it plays out.
One caveat – there was a very significant number of retirements (and, sadly, deaths) among senior faculty in the 2016-17 academic year – without looking it up, I think it was 8-10 people. By itself, that would likely be big enough to affect the average numbers for Guilford full professors, as would any year with a hiring freeze for assistant professors, but without specific numbers I can’t determine the magnitude. Obviously transitions and promotions at other ranks would have an impact as well.
Ione Taylor, chair of the Academic Affairs committee of the Board of Trustees, just noted to the full board in their public meeting the need for improving faculty salaries to match our peers and competitors. She referred to this as ‘deferred people maintenance,’ naming it as a need equivalent to the deferred maintenance of our buildings and physical resources.
This represents a change in focus and in role compared to past Academic Affairs reports, and gains us a level of advocacy on the board that I have not previously seen. Ione is not alone in this. The other members of Academic Affairs have also made very encouraging statements.