Most people are familiar with the definition of median as it pertains to divided highways. They may not be familiar with the statistical definition but they should be because it will influence their property taxes.
The corporate world uses the median compensation of other companies to justify excessive salaries and stock paid to their executives. The median is that value for which 50 per cent of numbers, when arranged in order of magnitude, lie on either side.
Some companies pay obscene amounts to “independent” consultants to obtain salaries of comparative companies and generally use the median as a benchmark. Then benchmarks are adjusted on mostly subjective criteria to set salaries of their executives. Potash Corp. paid $473,872 for such services in 2009 and $362,300 the previous year. Why do you suppose they would pay so much to compensate consultants?
We learned in the last Gazette that St. Albert obtains salary benchmarks every second year. Its comparative group is cities in Alberta with similar populations plus the Alberta government and University of Alberta. However, our city uses the 60th percentile as a benchmark and not the median (50th percentile). With only 11 comparative groups, the 60th percentile could be near or much above the median. Likely, it was the city with the seventh-highest salaries of the 11 comparators. A more legitimate calculation is the median plus a cost of living adjustment if the comparative data are a year old. By the way, federal government retirees received 0.5 per cent as a cost of living increase in 2009, while non-union staff in St. Albert received 4.5 per cent.
A problem with using the median of comparative groups is that it can readily be manipulated to ratchet up salaries. If the cities with salaries below the median increase their salaries to the median in their next comparison exercise, then the median becomes the base. If they use a percentile higher than 50 like St. Albert, the median can increase quickly. The five entities above the 50th percentile in St. Albert’s comparative group must use a high percentile or use a different comparative group in their salary models. There is plenty of scope to get the result that you want.
A second problem with this technique is that the comparative group can be cherry-picked by compensation committees or by "independent' consultants who want to fatten their own payola. I will bet, for example, that Fort McMurray is in the comparative group used by our city. We are all aware of labour costs in the tarsands. There is only one reason that corporate North America uses the median to increase salaries. It’s what might be called “the executive advantage.”
Salaries of St. Albert staff appear to be high without finding ways to ratchet them up even more.
Donald C. Thomas, St. Albert
