A front-page News headline on Labor Day read, "Who earns what? Many jobs pay thousands less in Western New York than elsewhere in nation."
While factually correct, the conclusion that "The most numerous jobs . . . pay 5 percent to 10 percent less than the national average" is oversimplified and thus slightly misleading.
The most serious and obvious problem is that the nominal data from the Bureau of Labor Statistics on which the article was based are not adjusted for differences in the cost of living.
While The News correctly noted that housing costs are lower locally than elsewhere, the article treated this exclusively as a wealth issue. In fact, when housing, food, transportation, health care and other expenses are combined, it takes less income to live in Western New York than in the nation at large.
According to the 1997-98 State and Metropolitan Area Data Book published by the U.S. Department of Commerce, the cost of living in the Buffalo-Niagara Falls metropolitan statistical area is 3.4 percent below the national average.
Thus, a Buffalo bank teller earning the local average of $21,357 per year actually has more purchasing power than his or her counterparts earning the national average of $22,007. Adjusted for these cost-of-living differences, 27 of the 42 job categories listed, covering more than half of all regional employment, pay more on average locally than nationally.
A more subtle and technical problem is that simply comparing averages ignores sampling error. For example, although there are an estimated 12,480 general office clerks in Buffalo, making it one of the most numerous local jobs, the bureau's estimate of their average pay ($19,550) is based on a sample or subset, as is the estimated national average, which makes the estimate imprecise.
I contacted the bureau and was provided with the sampling errors for each of the occupations in Buffalo. The error for the average pay of general office clerks is 1.9 percent, or about $371.45. Statistically, there is a 95 percent probability that the true average lies within plus-or-minus 1.96 standard errors of the estimated sample average.
Thus, the true average annual pay for general office clerks in Buffalo may well be as high as $20,278, yielding as much purchasing power as $20,992 earned elsewhere. This compares favorably with the reported national average of $20,250, even leaving aside the sampling error for the national estimate.
Once these adjustments are considered, it is no longer obvious that so many local jobs lag behind the national norm in terms of pay. Admittedly, many jobs do pay less here than elsewhere, but quite a few pay more.
And while there is clearly a perception of lower pay that drives workers from Buffalo to other locales, newspaper reports that overstate the gap can only fuel that perception and that exodus.
JOSEPH G. EISENHAUER