Column: Watch a conservative try to show that the working class has done great

The debate over the U.S. economy tends increasingly to boil down to the question of: “which economy?” There’s no question that growth overall has been decent since the depths of the recession. But for the typical rank-and-file worker, the economy still looks sluggish, even stagnant.

Progressive economist Jared Bernstein, a former advisor to Vice President Joe Biden, put his finger on this last week, observing that a “‘profits up, wages not’ recovery has lasted for years,” in which “fairly narrow groups of workers are in high demand and the rest suffer from weak bargaining power.”

Conservative economist Scott Winship of the Manhattan Institute confronted what he calls “the myth of the vanishing pay raise” head-on in Forbes. His foils were Bernstein and Steven Greenhouse of the New York Times, who also reported on the phenomenon of stagnation in the midst of apparent prosperity. Winship’s goal is to show that the average worker has done a lot better than it seems. (All three pieces appeared before the latest monthly jobs report, which showed a sharp increase in hiring in October and a drop in the unemployment rate. But monthly employment reports are notoriously volatile.)

“Their basic conclusion is just wrong,” Winship says of Bernstein and Greenhouse. But is it? Winship’s effort to prove his point tells us much not only about the true condition of working-class America, but about the complexities of trying to find a single picture that portrays how the economy is doing.

Winship’s core assertion is that inflation-adjusted “average pay has been rising since 1993 and is 23% above its January 1982 level,” which he thinks is a healthy trend. Leaving aside whether that’s healthy enough, Winship seems to have put both his thumbs on the scale in measuring the wage gains of working-class America. Let’s take them one thumb at a time.

First of all, Winship measures “average pay.” This is the “mean”--total pay divided by number of workers in the sample--rather than the “median”--the pay at the midpoint of all workers. This is such a hugely important distinction that Winship discloses it--in the very last paragraph of his article. He suggests he used average pay because Bernstein and Greenhouse did. But he also acknowledges that “as inequality has grown, growth in the average has outpaced growth in the median.”

Think of it this way: if you have 100 workers each earning $40,000, their mean and median pay is $40,000. But if a year later one is earning, say, $120,000, and the other 99 still earn $40,000, mean or “average” pay will rise to $40,800, which looks like an overall increase of 2% even though one person’s pay has tripled and everyone else’s has stagnated. Median pay, however, will still be $40,000, obviously a better gauge of the overall change. That’s important when you’re analyzing whether economic growth has diverged among economic classes. In effect, Winship seems to concede that the vanishing pay raise may not be such a “myth” when we’re talking about anyone but the top earners.

Winship’s second thumb involves his choice of an inflation measure. As he acknowledges, “nominal” or non-inflation-adjusted pay isn’t what’s important: if your paycheck and inflation have both increased 5% in a year, you haven’t made any progress. What’s important is inflation-adjusted or “real” pay, and putting a number to that is dependent on how we measure inflation.

Winship observes that Bernstein and Greenhouse rely on the consumer price index for urban consumers (the CPI-U), the best-known and most widely used inflation indicator. The CPI measures price changes in a common basket of goods and services commonly purchased by city folk.

He argues that’s a mistake, because the CPI “overstates” the rate of inflation. Instead, he advocates something known as the personal consumption expenditures price index, or PCE. Winship observes that the PCE, which is derived from purchases made by nonprofit organizations, is preferred as an inflation gauge by the Congressional Budget Office and Federal Reserve Board.

More to his liking, one suspects, is that it typically shows a lower rate of inflation than the CPI-U. (See graph below.) That’s because it weights goods and services differently, and more broadly incorporates “substitution effects”--if apples rise in price faster than oranges, some people will switch to oranges, immunizing them from inflation.

Measured against the CPI-U, hourly wages have gained a total of about 11% since 1982; measured against the PCE, they’re up 23%. If your goal is to maximize how much wages have risen, which measure would you choose?

Winship’s implicit argument is that the PCE is somehow more accurate than the CPI. That conclusion is by no means universal among economists. The accuracy of inflation indices depends wholly on what’s being measured, and these two measure different things. In healthcare, for example, the CPI counts only out-of-pocket spending by consumers; the PCE counts employer-paid premiums and government healthcare expenditures too.

The PCE might be more useful for the Fed or CBO in tracking inflation in the overall economy, but leaves much to be desired in tracking price changes experienced by the person on the street. For that, Todd Clark, an economist at the Federal Reserve Bank of Kansas City, wrote in 1999, the CPI was the better index.

Still, whichever measure is used, Winship’s own figures support the argument that wage growth has been poor, virtually since 1980. His statistics reveal only three sizable periods of inflation-adjusted growth--from 1997 through 2003, late 2005 through 2006, and 2012 through the present. A sharp increase shows up in 2008, but Winship acknowledges that this reflects a significant and presumably recession-related drop in prices that year.

For all other periods, average inflation-adjusted wages are either flat or falling. Even his vaunted 23% increase from 1982 to today, based on the PCE, works out to compound annual growth of 0.63%. You can judge for yourself whether that’s something to turn cartwheels over. Measured against the CPI-U, the compound annual gain has been 0.32%, which definitely is nothing to crow about.

One point made by both Bernstein and Greenhouse seems indisputable: however you crunch the numbers, a vast swath of working Americans feel that economic growth has left them behind. That’s been especially true during the post-recession recovery, and it corresponds to the finding of UC Berkeley economist Emmanuel Saez that fully 91% of the inflation-adjusted income growth in the U.S. in 2009-12 was captured by the top 1% of income earners.

Winship can be admired for his effort, but he’s trying to sell average Americans on the idea that their feelings of being underpaid are mostly in their imagination. It’s not surprising that so many workers aren’t buying it.

Keep up to date with the Economy Hub. Follow @hiltzikm on Twitter, see our Facebook page, or email michael.hiltzik@latimes.com.