March 28, 2005

Tim Noah - Where's Your Evidence?

Via Freiheit und Wissen, founder of the Carnival of the Un-Capitalists, we find an absolutely infuriating article -- snarkily accusive, ill informed, and smug -- by Timothy Noah of Slate:

I'd thought it was a settled matter that Wal-Mart had achieved its miraculously low prices by squeezing its employees. Not so, said Scott
Settled by whom?

Just who has written a comprehensive, comparative analysis of the wages and benefits among all retail chains -- revealing that WM is the cheapest of them all? Who has demonstrated that Wal-Mart sqeezes its employees -- pays lower wages and benefits than the mom and pop stores it replaces? The UFCW? The AFL-CIO? Writers at The Nation and Alternet?

I had thought that Wal-Mart's low prices were demonstrated to be a result of a complex of factors: paying -- like everyone does -- low but competitive wages for low skilled work, large investment in information technology (which permits keeping far less inventory in its stores than competitors), opening stores in seemingly empty rural areas with only small retailers as competitors, low-cost design of facilities, cramming as much into a store as humanly possible, owning and operating its own distribution capacity, pushing suppliers to make goods cheaper, dumping suppliers who can't lower the cost of production, and a willingness (despite alleged public resistance) to import 3/4 of its goods from China -- not to mention a no-frills executive lifestyle (they fly coach and double up in standard rooms).

Mr. Noah, who did the hard work of eliminating all but "squeezing its employees" as relevant factors in the formation of Always Low Prices?

OK, well, about about those contested wage statistics?

Wal-Mart's average wage is around $10 an hour. [ - H. Lee Scott]

As Tom Geoghegan, a labor lawyer in Washington (and author of Which Side Are You On?: Trying To Be For Labor When It's Flat On Its Back) points out, the relevant number isn't the average, which would be skewed upward by the large salaries of relatively few highly-paid company executives�Scott, for example, receives, by one reckoning, 897 times the pay of the average Wal-Mart worker�but the median. In the Dec. 16 New York Review of Books, Simon Head, director of the Project on Technology and the Workplace at the Century Foundation, stated, "the average pay of a sales clerk [italics mine] at Wal-Mart was $8.50 an hour, or about $14,000 a year, $1,000 below the government's definition of the poverty level for a family of three." That the current minimum wage of $5.15 per hour leaves families even farther below the poverty line is a depressing topic for another day. [Emphasis added --ALP]

Note now that 2.9% of the hourly workforce earns the minimum wage (or less -- think waitresses), because this not so depressing data point will probably not be noted on that other day.

Anyway, I agree that the median is the relevant statistic -- if the distribution is heavily skewed. If it isn't, the mean and the median are close enough for all but nit-picking purposes. But Mr. Noah hasn't demonstrated that Wal-Mart's hourly wages of full time employees form a skewed distribution. I agree that if salaried employees wages were included, the distribution would probably be highly skewed, and the median would be a better choice. But salaried employees wages were NOT included; H. Lee Scott was talking about HOURLY wages of full time HOURLY employees.

Again, H. Lee Scott's own wages were not included in the near $10 an hour figure; neither were the wages of any other full-time salaried employee. Obsessing about the relative size of their pay to the hourly worker is unwarranted here.

(Let's play devil's advocate anyway. Assuming Mr. Scott worked 40 hours a week [he works more] for 52 weeks, his annual wages of $10million come to $4800 an hour; but if included as an hourly wage, it would bias upward the mean hourly wage of all 1.2 million U.S. employees by about $0.004 an hour. Adding in all other highly paid employees would have the same effect proportional to their salaries. Would the total mean bias sum to a dollar an hour? I'll figure it out later, but I repeat, THIS BIAS IS NOT PRESENT IN H. LEE SCOTT's figures).

Why doesn't Timothy Noah link to the Drogin report's hourly wage table? Perhaps because Wal-Mart's opponents don't use the median when discussing WM hourly wages, and perhaps because the data support H. Lee Scott's claims. It shows wages for all full time hourly WM employees in 2001. The average hourly wage was $9.26 -- in 2001 -- according to the expert used by the lawyers of the plaintiffs suing WM for sex discrimination.

Wal-Mart claims that full time hourly associates are paid $9.68 an hour (circa 2004). This makes sense, since a ($9.68-$9.26)/$9.26=4.5% increase is not unreasonable over a time period of a few years.

Also, Mr. Head rightfully claims that the average wage for a SALES CLERK is about $8.50 an hour, but of course, a sales clerk is the lowest paying full-time job at WM; why would neither he nor Tim Noah not mention this???

And Simon Head's extrapolation from hourly wages to annual salaries is incorrect. In fact, there is no need to extrapolate; actual data -- data that compensate for actual weekly hours worked (see below)-- are available. Anyway, in $2001, according to Wal-Mart opponents, sales clerks made an average of $15,500 -- not $14,000. This is absolutely dishonest on Mr. Head's part; the salary is actually $500 ABOVE the poverty threshold for a family of three. Nice try, play again.

What's next. Oh, yes, work-week:

Few people realize that about 74 percent of Wal-Mart hourly store associates work full-time, compared to 20 to 40 percent at comparable retailers.

Yes, but what exactly is a "full-time worker"? Typically, full-time is defined as 40 hours a week or more. At Wal-Mart, it's defined as 34 hours a week. So of course Wal-Mart has more "full-time" workers.

We all agree that 40 hours is a solid full-time job, and that 34 is not. Since Mr. Noah is so convinced this is a large problem, perhaps he would care to tell us how large a share of all hourly WM associates work 34 hours a week? Does he even know, or is he using the 34 hour figure in an entirely inappropriate context? (The 34 hour FT workweek is generally designed so that those working fewer hours qualify for full-time benefits). But don't take my word for the benefits problem; listen to the AFL-CIO:
In 2002, Wal-Mart raised its definition of �full-time� work from 29 to 34 hours weekly. With that change, Wal-Mart increased the number of its part-time workers to nearly 400,000, or about a third of its total workforce, and also increased barriers to workers� eligibility for job-based health care, according to a report by the minority staff of the U.S. House of Representatives Education and Workforce Committee.
In other words, the AFL-CIO used the disgusting George Miller attack paper to defend its claim that many Wal-Mart employees became part time because they worked between 29 and 34 hours... But footnote 19 of the Miller report says only that a large increase was "likely", and provides no evidence that Wal-Mart employs quasi-full-time employees. Hence, I am suggesting that there is absolutely no evidence of a large number of "full-time" Wal-Mart employees who work around 34 hours a week. Without citing any empirical evidence -- even anecdotal, or the records of a single store, or even a portion of a store (like a Tire & Lube) -- this claim is utterly worthless; repeating it is nothing but hot air. If anyone does have evidence of intentionally short work weeks, then they should show it.


Fewer hours worked, I need hardly point out, means that Wal-Mart's "full-time" employees are less likely than employees elsewhere to be able to afford premiums for any health insurance they're offered. According to Head, fewer than half of Wal-Mart's employees can afford even the company's least-expensive health plan.
Hmmm... Simon Head has no evidence of affordability, so he can't make this claim with a straight face. Wal-Mart notes:
Currently, 86 percent of Wal-Mart hourly store associates surveyed have medical insurance - 56 percent of those with coverage received health care insurance from Wal-Mart and the remainder receive health care through another source such as another employer, a family member, the military or Medicare.
Let's posit that the 14% without insurance "can't afford it", which we know is wrong, but charitable to Mr. Head's position. Mr. Head has no evidence that the 38% with non-WM coverage cannot afford the Wal-Mart plan; he has evidence that non-WM coverage is often more desirable for associates than WM's coverage, but less desirable is not the same thing as unaffordable. We can't add the 38% and the 14% because we do not know the financial situation of the 38%.

Back to Tim Noah:

In 2003, the most recent year for which I can find data, Scott sucked down $29 million (including stock-option grants). That same year, G.R. Wagoner, president and CEO of General Motors, hauled in about half that amount, $15 million. Following Scott's logic, I don't see how he can avoid knocking his own pay down to around $10 million.
H. Lee Scott does not set his own pay, so how is he supposed to knock down his own pay? (If I'm wrong, then provide evidence that Scott sets his own pay). If Tim Noah thinks that Scott is overpaid, why doesn't he contact everybody on the Wal-Mart board and tell them that they're wasting millions of dollars? He could start off like this, "Hi, I'm Tim Noah, columnist for Slate, and I know better than you how much your employees should be paid."

I'd like to hear their responses.

Posted by Kevin on March, 28 2005 at 01:49 PM