Prices, Real Wages, Wal-Mart & GE

Interesting Op-Ed in today's NYT by Pankaj Ghemawat and Ken A. Mark, called "The Price Is Right".

The authors argue that, contrary to the bad press it receives, Wal-Mart has actually greatly benefited low-income communities, especially rural ones. The argument is straightforward: yes, the authors implicitly conced, WM may (they say "supposedly") have the effect of

securing subsidies, destroying jobs in competing stores, driving employees toward public welfare systems and creating urban sprawl.

G-M cite price price reductions of 8% in rural areas and 5% in urban ones and argue that the gains that WM produces far outweigh these costs: their calculation is that

Wal-Mart saves its consumers something like $16 billion a year. And because Wal-Mart's presence forces the store's competitors to charge lower prices as well, this $16 billion figure understates the company's real impact by at least half.

They also cite a McKinsey Global Institute study, writing that

Robert Solow, a Nobel laureate in economics and an adviser on the study, noted that the most important factor in the growth of productivity was Wal-Mart. And because the study measured productivity per man hour rather than per payroll dollar, low hourly wages cannot explain the increase.

Lastly, G-M argue that even seemingly modest wage increases -- they use $2 per hour -- would "wipe out" WM's profits unless WM raised its prices, which they predict it would. G-M's conclusion is that

the debate around Wal-Mart isn't really about a Marxist conflict between capital and labor. Instead, it is a conflict pitting consumers and efficiency-oriented intermediaries like Wal-Mart against a combination of labor unions, traditional retailers and community groups.

There's a lot to talk about here.

I'm not a knee-jerk opponent of Wal-Marts, but I also think one should avoid wearing rosy glasses when thinking about these issues.

1. The most basic point that seems to get lost in these debates is that, unlike most employers, and unlike most retailers, WM is big enough to have potentially general equilibrium effects on both consumer prices and wages. For nontechnical readers, what I'm talking about is the fact that in a competitive economy, most firms are small enough that their price-setting decisions have small or zero effects on the level of prices -- when you have lots of competitors, you tend to be able to charge only "what the market will bear". WM has become big enough that it can probably move prices, up or down, to a greater extent. In other words, WM has some ability to determine what the market will bear for its competitors.
   
   Now, economists measure standard of living roughly as how much stuff (goods and/or services) households are able to afford. Few people seriously doubt that WM's presence reduces prices of the goods it sells in the markets where it operates. At the same time, WM's wages are widely believed to be low, and reduced labor costs are thought to be a key to its profitability and ability to keep prices so low (as G-M's claim about wiping out profits clearly underscores).
   
   A common -- if simplifying -- way to think about real wages is as the ratio of the hourly wage to some price index. (Think of a useful price index as an average of what all the different things you buy will cost, once we've accounted for the share of your budget for which each thing accounts.)
   
   Now, if WM lowers both the price index and the wage, then its impact on standards of living is ambiguous. It could be positive, even for low-wage workers, if goods prices are reduced more than wages are (note that you need to include the impact of WM on other firms' benefit packages, too). On the other hand, it will be negative if wages fall more than prices when WM moves in. This strikes me as a very interesting economic question to resolve -- one that someone could probably get a whole labor economics dissertation out of it if the data are available (which they may or may not be).
   
   The frustrating thing in debates I've heard about WM is that people tend to say "WM lowers prices" and then others tend to say "WM lowers wages", as if these statements are somehow mutually exclusive. There's no way to assess WM's impacts on worker-consumers without addressing both questions.
   
   If you don't believe my point, then consider this part from G-M's article:

   Wal-Mart's customers tend to be the Americans who need the most help. Our research shows that Wal-Mart operates two-and-a-half times as much selling space per inhabitant in the poorest third of states as in the richest third. And within that poorest third of states, 80 percent of Wal-Mart's square footage is in the 25 percent of ZIP codes with the greatest number of poor households. Without the much-maligned Wal-Mart, the rural poor, in particular, would pay several percentage points more for the food and other merchandise that after housing is their largest household expense.

   Yes. And they might also receive several percentage points more in wages.
   
   Serious discussions of the impact of an economic phenomenon cannot concern only the costs, nor only the benefits.



2. G-M's article says little about the social costs of losing downtowns, mom-and-pop stores, etc. I don't know how to measure these costs, and I don't even know if WM really destroys traditional outlets (though I'd be surprised if it doesn't): perhaps those stores would go under even without WM's entrance. But G-M don't even bother to discuss the possibility that "way of life" could be valuable to people. This isn't an easy issue, but it's hardly irrelevant, either.



3. I wonder about G-M's statement that "because the study measured productivity per man hour rather than per payroll dollar, low hourly wages cannot explain the increase." It really depends on whether the measure here is gross productivity or net productivity. Gross-of-labor-costs productivity equals the value of total output divided by the total number of hours worked. Net productivity equals the value of total output minus the hourly wage bill, divided by the total number of hours worked. You can increase net productivity by reducing the wage bill, holding output constant. Which means that G-M's claim about productivity per "man hour" holds only if we are using gross productivity. Each measure of productivity has a role (gross-of-labor-costs prod tells us how much stuff we can make, however it is that our society divides it up, while net-of-labor-costs productivity tells us how profitable -- and thus how globally competitive -- private businesses in the US can be).

It's too bad that G-M do such a poor job of making their argument. I would have welcomed a serious attempt to address this issue.

Update: For all you non-economists, "GE" in the title of this post refers to general equilibrium, not the manufacturer....