Assume that you are in charge of determining the raises for your group of 4 employees:
Further, since after all it is yourself whom we are talking about, let's assume that all four employees have exemplary performance and are the envy of every other supervisor at JPL.
Assume the following standard salary curve, based on 1996 salaries for engineers in all industries:
|Years Since BA||1996 Year Salary|
To a first approximation, inflation lifts all salaries of the salary curve up equally. Thus one can assume in the following that inflation is 0, and adjust the results for inflation at the end.
In this case of 4 exemplary employees, the salary curve then determines directly how raises should be assigned, since I've assumed that the same person is being evaluated, and that you are performing in your usual exemplary fashion at all times in your career. The average raise over the next 10 years must then be:
|Years Since BA||Yearly Average Raise|
The raise will of course be higher at the beginning of the 10 year period than the average, but all we need is the average to understand the effect.
How much money will it take to accomplish this? In the first year, one needs $1540 + $1281 + $684 + $374 = $3879, or 1.6% of the entire salary base of $39.5 + $55.7 + $68.4 + $74.7 K. Thus the "JPL average raise" must then be 1.6%.
NOTE HOW THAT 1.6% MUST BE DISTRIBUTED IN THIS CASE. The "fresh-out" + 1 year must get 3.9%, the 10-year employee must get 2.3%, the 19-year employee must get 1.0% and the 28-year employee must get 0.5%. This doesn't mean that the 19-year and 28-year employees are valued less than the others (remember they are all ranked highly - they are the same employee after all, just with different ages!) IT IS SIMPLY A REFLECTION OF THE INDUSTRY SALARY CURVE. (i.e., it's the facts of life, kid.)
Now let's add in inflation. Suppose the inflation rate is 3.0%, and the JPL average raise is 4.6%, numbers that are representative of the years before 1993. Then because 1.6% of the raise is needed to move people along the salary curve, the "inflation adjustment factor" is 4.6 - 1.6 = 3.0%, which is, not surprisingly, the rate of inflation.
What do you actually get then?
The 1-year-after-BA yourself gets 3.9 + 3.0 = 6.9% ($2726), making for a very happy fresh-out.
The 10-years-after-BA yourself gets 2.3 + 3.0 = 5.3% ($2952), making for a still pleased person.
The 19-years-after-BA yourself gets 1.0 + 3.0 = 4.0% ($2736), and is now wonndering why it is that his/he supervisor doesn't love him/her nearly as much as he used, since this is exactly an average raise in terms of the overall raise percentage allotted, and a below-average raise in terms of the distribution of the percent raises. "How can a clearly-superior employee only be getting an average raise, and a below-average percent raise?", he asks.
The 28-years-after-BA yourself gets 0.5 + 3.0 = 3.5% ($2614), and is now quite annoyed that his/her supervisor apparently hates him/her now, and doesn't value his/her work anymore. In fact, this older version of yourself is incensed that his/her raise is below both the JPL average raise, and is in fact the exact bottom percentage raise handed out. The supervisor mutters something about "well, since you are paid so much, we could do a lot of benefit for the younger people by shaving a few percentage points off your raise and giving it to them. Besides, you are making twice the salary of the person with the highest raise." The employee hears this argument, understands the rationale a bit, and may feel a little less unhappy, but is still an unhappy camper.
Note that the sum of the dollar increases represents the advertised 4.6% raise pool.
Now let's make it a little more interesting. Suppose the inflation rate is 3.0%, and the JPL average actual increase is 3.1%.
Again, because 1.6% of the raise is needed to move people along the salary curve, the "inflation adjustment factor" is 3.1 - 1.6 = 1.5%, significantly under the rate of inflation. THIS MEANS THAT IN THIS SITUATION, SALARIES IN GENERAL ARE NOT KEEPING UP WITH INFLATION. These numbers are representative of the situation for 1993-1996 in California, when there were a lot of unemployed engineers.
What do you actually get then?
The 1-year-after-BA yourself gets 3.9 + 1.5 = 5.1%, making for a pleased fresh-out, since he doesn't know that this isn't enough to move him along the salary curve properly.
The 10-years-after-BA yourself gets 2.3 + 1.5 = 3.8%, making for a person who wonders why he/she isn't as "highly valued" as he/she once was.
The 19-years-after-BA person is quite unhappy, getting 1.0 + 1.5 = 2.5%, a raise which is "insulting" and "significantly below inflation". This 19-years-after-BA yourself is wondering if his/her supervisor has taken total leave of his/her senses and who is wondering if there is a message to him/her to go get another job.
The 28-years-after-BA yourself gets 0.5 + 1.5 = 2.0%, and is now thinking about filing a lawsuit claiming age discrimination.
Yet all of these people are YOURSELF, all are ranked highly, and considered valuable employees!
At IPAC, our dividing raises into an "absolute" and a "percentage" component is just a way to accomplish giving higher raises to younger people as required by the salary curve. There are many ways to accomplish the same thing, but the facts of life remain the same.
There is a way to try to put a better face on the raises, which we do NOT use at IPAC. (This of course bears no resemblance to any actual practice at JPL or any other institution.) One can "hold back" part of the average raise and give supervisors only part of the average raise to distribute. The supervisor then can say that the average raise was only the amount the supervisor got to distribute. Since the held back portion is also distributed at raise time, just by a different person, the supervisor can end up giving everyone a raise which is claimed to be higher than the average raise that the supervisor got to distribute. This is a nice tactic, because it makes for much happier, if not any richer, employees.
For example, suppose that 0.7% of the raise is "held back". Take the first example above. Then the supervisor can claim that the "average raise" is 4.6 - 0.7 = 3.9%. The 6.9% raise fresh out is even happier, outdistancing the "average raise" by 3.0% instead of the previous 2.3%. The 5.3% 10-years-after-BA is happier, outdistancing the "average raise" by 1.4%, instead of 0.7%. The 4.0% 19-years-after-BA is not nearly as unhappy, since he/she is now at least ahead of the "average raise" by 0.1% instead of being behind by 0.6%. The 28-years-after-BA is still displeased, but not as much, since he/she is behind the "average raise" by only 0.4% instead of 1.1% and can now perhaps accept the supervisor's explanation of "the facts of life".
Note that this tactic is much less effective when the average salaries start to fall significantly behind inflation.
For actual expected raises using the actual dispersion in salaries and assuming several different distributions of years-of-experience for employees, see Detailed expected average raises versus experience.
Copyright © 1997 by Tom Chester.
Permission is freely granted to reproduce any or all of this page as long as credit is given to me at this source:
Comments and feedback: Tom Chester
Last update: 20 September 1997.