Thursday, July 16, 2009

 

Salary Compression

In the discussion after the post about counteroffers a couple of days ago, several commenters raised the issue of salary compression.

For the uninitiated, 'salary compression' typically refers to new hires coming in at salaries higher than those of people who are already working there. It can happen pretty easily if internal salaries are based on pre-set, lockstep raises, but the rate of change in the outside world has been faster. Incumbent employees usually perceive salary compression as unfair, since people with less seniority are getting more money.

A few years ago my college ran into that issue with Nursing faculty. At that point, nurses could pretty much write their own tickets on the job market. (That market has since cooled considerably.) Since our salary scale was one-size-fits-all, we fell so far behind the market that for a few years, every single candidate we recruited turned us down. We eventually worked out a separate scale for Nursing faculty with the union, on the argument that if we didn't, we simply couldn't hire anyone. The union grumbled, since the idea of separate tiers cuts pretty hard against the idea of union solidarity, but acceded out of a recognition of a market-driven force majeure. Simply put, we couldn't ask the rest of the world to stop to suit our own internal taste for 'equity,' so we made the adjustment we had to make. It was either that or just drop the program, and nobody wanted that. Now, new Nursing hires make more than newly-tenured professors of anything else. Interestingly, since the Nursing market cooled, nobody on either side has proposed revisiting the salaries. Salaries are sticky, so you don't want to move off an established scale lightly; you'll never get it back.

Other than Nursing, though, we've been conscientious about sticking to a union-negotiated starting-salary scale that takes account of credentials and experience, and that offers no room for individual negotiation. Given our labor environment, a new hire coming in at x (standard) plus y (negotiated) would force us to move up our entire scale, possibly retroactively. It's just not something we can do. As a result, we really don't have salary compression in the usual sense of the term. So when I refuse to make counteroffers, I'm not trapping anybody in a salary-compressed department.

(The sense in which I'll occasionally hear allegations of salary compression is when a new hire shows up with a doctorate and years of experience elsewhere, and places above a newish incumbent person with a Master's. In that case, the issue isn't actual unfairness; it's the invisibility of the previous experience. When the griping starts, I just refer them to HR to discuss the point system. To my mind, when this happens, it's a sign of hiring well.)

The downside of a rigid salary scale, obviously, is that we lose some great people to other employers with more generous offers. That's frustrating, but in a strong union environment, it's better to eat the occasional failed search than it is to start improvising. The cost of a second-best candidate is far less than the cost of the grievances, and the negotiations, and the arbitrations, and the subsequent adjustments. (Btw, the same holds true of counteroffers. The cost of replacing a current high performer with a new hire is far less than the cost of the grievances, arbitrations, and awards that would result from deviating from the scale.) This isn't a hypothetical; my college actually went through this shortly before I got here.

That's why I reject the whole "mark-to-market" argument for counteroffers. It's less costly for me to lose the occasional star than to spend the next several years in court. Nobody -- nobody -- is irreplaceable, or worth embroiling the college in the kind of battles that counteroffers would generate. Better just to wish them well and get on with it. That's not because people are interchangeable parts, or cogs in a machine; it's simply that the very real differences between individual performers are dwarfed by the staggering cost of legal settlements.

Of course, in my preferred universe, there would be a much greater 'merit/performance' component to salaries. But that's not the world I live in, and I'm not in a position to make that happen unilaterally.

In a non-union environment, salary compression could happen much more easily. If you don't have a relatively strong centralized system for determining salaries and raises, individual deans/chairs could easily upset the apple cart in the name of getting or keeping someone they particularly want. I've even heard of universities in which elected department chairs allocate merit raises, which strikes me as insanity on a stick. (Structurally, that's basically a Tammany Hall model.) This is one of those cases in which I'm happy to work with the union, since a little discussion upfront saves untold agony and political conflict later. In fact, I'm increasingly convinced that the optimal model combines a union, a substantial performance basis for raises, and multiyear contracts instead of tenure. That's not my world, but I think it would combine reasonable amounts of security, fairness, accountability, and equity in a relatively sustainable way. Maybe someday...

Comments:
The kind of compression you note isn't the most troubling sort. The problem becomes acute when there has been a freeze in advancement on the salary scale and/or a change in the way experience is evaluated.

The result is something akin to the following:
A has a Ph.D. and 10 years experience teaching at the college.

B has an MA and 10 years experience teaching OUTSIDE the college... i.e. they're a new hire.

B makes more than A... why? Because a while back A's contract didn't include step increases, so while they put in their 10 years, they got credit for 4 or 6 on the salary scale and every one of B's years was counted in calculating salary.

The other way this happens is like this:

C is hired under a rule that limits the total amount of experience that counts toward salary... so, they have 15 years experience but only get to count 4 years per the contract.

D is hired the next year, under a new contract that has no such limits -- if they have 5 years experience, their salary will be higher than C's.

Absent a system of counter-offers, administration really should have a system to evaluate salaries on an on-going basis so that differences of this sort don't happen.
 
Chair-run merit raises aren't that bad--at least not at my school. External forces (i.e. the Dean) tells the chair how much total money for the department, and the Chair distributes the money more or less according to a point system (so many points for a teaching award, so many for an article, and so on). Some people grumble at the margin about their points, but nothing too serious. It's not like the merit pool amounts to much anyway. Maybe that's the secret, make the merit pool small enough so people can be happy they're getting merit raises without creating big enough gaps to create resentment.
 
Unfortunately, what I see at my university is the avoidance of salary compression used as an excuse to pay everyone equally badly.
 
This has become a standard genre on this blog: a post whose perspective is so deeply shaped by a career spent in strong union states that I don't recognize the logic as remotely applicable to any of the jobs I've held, all in non-union states. I wish there were an exchange program for administrators: those from Texas and Utah would see how universities work in places where workers actually have some rights, while those like Dean Dad might see how laughably naive such proposals as "multiyear contracts instead of tenure" sound outside of a union setting...
 
We've had the same challenges at our CC recruiting faculty in high demand areas (e.g., nursing). As DD suggests, the problem with bargaining "special" salary schedules for a high demand area is nobody wants to implement a salary takeback when a position is no longer high demand. Thus some faculty throughout their tenured career will earn far more than their comparable faculty colleagues, even though the justification for the salary differential no longer applies.

So far, we have maintained a consistent salary schedule for all faculty, and we have been able to fill our open positions with qualified applicants. Sure, we probably lose some well qualified candidates, but if a candidate's primary concern is a high entering salary, they wouldn't be a very good fit for our CC.
 
I like how this is handled in k-12, at least in the districts that I've dealt with:

They make a big matrix with levels of education across the top and years of experience (inside or outside of that district) down the side and pay everyone the salary that results from their cell. The entire table is adjusted each time there is a step raise (usually a little bit per every year for cost of living). Everyone can see what they can do to improve on the "education" side (around here, it seems that masters + 45 additional grad credits is where that tops out, with a small additional bonus for getting a Ph.D - I assume that the categories would be different in higher ed and I'm not sure if you have any reason to incentivise ongoing accumulation of grad credits after hire so you may have just a masters and a Ph.d column I suppose) and can see how much more money each year of experience gets them.

It seems like this solves salary compression problems if done right since everyone is on the same chart. One of the issues I'm seeing now is that, due to changing state license requirements, many teachers now have their master's degree before they start teaching. This causes them to skip about half of the "education" steps (which assume, as was the case previously, that teachers will start with a bachelor's and then work toward their master's while teaching) and means that many newer hires max out the education bonuses in their first 5 or 10 years. I worry that this will mean that they have no incentive to continue professional development later in their careers and wonder if they'll make more columns in the table as people start maxing out earlier and earlier. (I've been teaching for 3 years and am 19 credits away from maxing out the education bonuses.) The experience rows mean that you top out at about 15 years, after which you only get cost of living raises when we get a new salary table rather than experience raises.
 
This is a subject where everyone should be careful to state whether they are talking about a CC or a University....

For the record, I am at a non-union CC but our salary schedule operates very similar to the union system DD describes. It might be "board policy" rather than "union contract", but it has the same effect.

What PhilosopherP talks about up at the top is troubling, but I see it as the fault of the union and/or the faculty that negotiated that salary schedule. The administration simply implements it. If you have a freeze on pay increases to current faculty, there should be a freeze on the amount of credit someone can bring in. If you don't, it is your fault.

Similarly, I know of one instance at a university where the existing faculty voted with one voice in favor of inversion for new faculty. They said to the Dean, "just because we got screwed by the job market and inflation in the 70s is no reason to screw the new faculty. Besides, we already own houses and the new folks had to buy one." (In the end, it paid off because the performance comparisons eventually led to pay raises for the existing faculty.)

I know how university merit systems work, but I'd be interested in hearing about a merit system at a CC. What elements go into the evaluation by the Dean? Can they be as objective as grant money and the number and quality of journals you publish in?
 
I like how this is handled in k-12, at least in the districts that I've dealt with:

They make a big matrix with levels of education across the top and years of experience (inside or outside of that district) down the side and pay everyone the salary that results from their cell. The entire table is adjusted each time there is a step raise (usually a little bit per every year for cost of living). Everyone can see what they can do to improve on the "education" side (around here, it seems that masters + 45 additional grad credits is where that tops out, with a small additional bonus for getting a Ph.D - I assume that the categories would be different in higher ed and I'm not sure if you have any reason to incentivise ongoing accumulation of grad credits after hire so you may have just a masters and a Ph.d column I suppose) and can see how much more money each year of experience gets them.


This system rewards people based on inputs when you should be rewarding them based on outputs. While you can argue the above system treats people fairly, there are ways of doing that without the major flaws of this system, which provides no rewards for the employees who perform well and also offers no incentives for anyone to perform well.

Why is there so much concern with years of experience? While there is typically a measurable difference between new employees and people with a few years experience, there aren't measurable average differences between people with 5 years and 10 years experience. The common joke about someone having 1 year of experience 5 times rather than having 5 years of experience is as true in academia as it is in industry. The problem with measuring inputs instead of outputs is that you can't tell the difference between the two types.
 
Well, Anonymous at 11:49AM, you probably didn't see my comment before you wrote yours, but the question still applies:

What outputs can you objectively measure in a CC (teaching only) environment?
 
This is an excellent issue (KPIs for higher ed). Seems like if you can get people to the point where

"Educator performance should be measured by the value added to students flowing through their classrooms"

it starts in the right direction.

The tricky part is "Step 2" where you get faculty to agree on how to measure "value" in any meaningful way.
 
What outputs can you objectively measure in a CC (teaching only) environment?

There are many possible output metrics, such as counting the number of students that withdraw, fail, pass, etc. One metric I like is improvement on a subject test divided by the initial score on the test taken before the class.

No metric is perfect. Because of that fact, I don't recommend using a completely rigid objective system like the one the poster above suggested even if you are measuring the right things.

It's never possible to perfectly measure every relevant work product in a way that compares evenly with everyone else, and that's at least as true of engineers and other professionals as it is of teachers. However, most professions don't give up and measure a couple of irrelevant inputs to determine salaries.
 
Anon, I mean this in all seriousness:

Have you ever taught?

I'm not kidding. Because I have, and I can tell you that I've had classes where I've had to fail 1/3 of the students and classes where (outside of 1-2 people too busy partying to go to class) the lowest grade was a B-. Same material, same tests. Just luck of the draw. Different classes are REALLY DIFFERENT. 100 students actually isn't enough for the Law of Large Numbers to kick in, and cohort effects are enormous.

Inputs aren't well enough measured, in general, for outputs to be well-measured, and instruction itself is poorly enough understood that suggested metrics quickly become jokes.

It may be a problem worth trying to solve, but it's definitely enormously nontrivial, and it's a form of psychosis to pretend otherwise. This isn't 2003 any more; we aren't required to pretend that something is true which is obviously and patently false in order to remain in polite company.
 
I think what Anonymous was referring to was

1. Give a comprehensive final exam on the first day of class
2. Teach
3. Give a comprehensive exam on the last day of clss
4. Measure the difference

Would the faculty member be rewarded for each idividual class delta? Of course not . . . for the "strawman" reasons mentioned by PM. An X period moving average would take care of the comparabiliy issues (and is simople enough to calculate.

The real problems with this method are

1. Determining what constitutes a "comprehensive final exam" (which is a prolem with a lot of "post test only" administrations today; so isn't a new/different/unique problem); and then "enforcing" the concept of a meanignful comprehensive exam
2. Monitoring to make sure faculty members don't cheat (really hard pre-test, really easy post-test)

The KPI here (improvement coefficient) could be multiplied by the results of a *meaningful* instructor/class evaluation system (stuent opinion survey) that most campuses already have . . .

At Big State Compass Heading U, we have a faculty committee (similar to P&T) that does the faculty rack & stack for the chair. The chair adds his/her rack & stack. BOth go to the Dean. The Dean makes Merit Pay Adjustments annually.

This year the raises aren't that big . . . hoever, the net effect is we have high performing junior faculty making morethan low performing senior faculty. We have high performing senior faculty making more than low performing junior faculty.

And we pay high enough entry-level salaries to attract quality junior faculty.

Yes, there is annual squabbling over the merit pay process- but everyone is judged by the same criteria, and the process is open, and everyone knows how they are being measured.

Of course, it's a College of Business so this wasn't so hard for our faculty & administration to figure out . . .

(but no, even our enlightened solons haven't bitten off the pre-test/post-test issue; maybe because we are disgruntledly satisficed with the system we have)
 
The most important thing in the world is to eliminate the side ideas that are supposed to be slowly and accretively taught in classes (critical thinking, working in groups, research skills) and reduce every single class in the world to a single, totally agreed-upon test which every teacher must teach to.

It's not that I don't appreciate the ideas behind these kinds of initiatives, I really do. People should be held accountable, even people in positions of authority. I'm all for that huge social shift. It's just that the problems with the approach are so well-known, and so excruciatingly obvious in our experience with NCLB, that I can't understand why the conversation is still taking place at this naive a level.
 
Milo Minderbinder: "Sure, we probably lose some well qualified candidates, but if a candidate's primary concern is a high entering salary, they wouldn't be a very good fit for our CC."

The classic reasons for people to be concerned with starting salaries (aside from the fact that more money is better) are well known:

1) Most systems will increase salaries by fixed percentages; it can be difficult to impossible to ever make up a low starting salary. One is not just negotiating how much one will get this year, but over a decade or more.

2) If you're trapped by (1), and considering moving: most employers will want to know your existing salary before they make an offer. This means that even if one moves to another organization, one still might find the initial low salary dragging them down, a decade or more later.
 
In my opinion I think that salary compression is a good idea. Just take a look at some the current salary data on this profession. It's really out of whack.
 
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?