Workers' pay hasn't kept up with the growing productivity of the American economy. It's a story that's circulated for years in the wonkier parts of the liberal internet, based in part on a series of charts produced by the Economic Policy Institute.
Most recently, Hillary Clinton has offered up her own version of the chart — framed initially as a rebuttal to Jeb Bush's notion that people should work longer hours but more broadly as part of her campaign's larger theme that a Clinton administration will bring much-needed higher pay.
But while the image is striking and depicts something real and important about the economy, it's also fairly misleading. It shows us a genuine economic trend, but not the one that readers think they see.
"That bargain has eroded. Our job is to make it strong again." pic.twitter.com/T3ARkHJRsz
— Hillary Clinton (@HillaryClinton) July 13, 2015
One problem with that chart, as deployed by the Clinton campaign, is that economic productivity simply has nothing to do with "working hard." A guy who moves furniture for a living works very hard, but he does not generate much economic value per hour worked. Highly productive workers are generally productive due to some combination of rare and valuable skills and access to useful technology. A movie star is more productive than a Broadway actor, because her performance can be seen by many more people.
But the bigger problem is that both lines are indexed to inflation — using different inflation indexes. The result is a chart that seems to suggest that further increases in productivity would be useless or unnecessary as a path to higher wages and incomes, when the real truth is the reverse.
A tale of two inflation indexes
The basic idea of inflation is easy to understand. Prices change over time, and accounts of various financial quantities should take that into account. But measuring price change is hard work. To do it, you need to define a universe of goods whose price you care about and then decide how important each item is to the overall basket.
The government does this in a few different ways.
One is the Consumer Price Index, which tries to measure the price of what a typical consumer who is not in a rural area buys. Another is the Implicit Price Deflator, which tries to measure the price of everything that is produced in the American economy.
These are different ideas.
American consumers buy airplane tickets, but they generally don't buy airplanes. Consequently, the price of a Boeing 777 isn't part of the CPI basket. The airplanes that Boeing sells to airlines — which, importantly, include many foreign airlines — also aren't part of the CPI basket. American consumers do buy computers. But a lot of computers are bought by businesses. And most successful American technology companies — Google, Apple, Microsoft, Facebook, etc. — have lots of foreign customers.
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/3910328/fredgraph.png)
It's no secret that digital technology has been getting cheaper while things like health care, college tuition, and housing in some metro areas have been getting more expensive.
Consequently, the inflation rate for consumers has been rising at a faster rate than the inflation rate for the overall US economy. When you draw a chart that uses both of these inflation indexes simultaneously, then the divergence between the two ways of looking at inflation is naturally going to drive divergence between whatever two quantities you're tracking.
The real story is more recent
When I point this out to people, a natural question they have is what this would look like if you did it right and used the same inflation index for both productivity and hourly compensation. The problem is that there's no right way to do it. You can't feed your kids a commercial jetliner or exports of business software, so saying something like, "Real wages have actually gone up a lot as long as you count a bunch of stuff that nobody buys in the price index" doesn't make much sense.
On the other hand, making business equipment and software is a very legitimate line of work. Saying, "The economy really hasn't grown much if you don't count America's most vibrant and innovative industries" is pretty blinkered.
Probably the best way to get an apples-to-apples comparison is to not adjust for inflation at all. Here is nominal GDP versus nominal compensation:
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/3911048/fredgraph-1.png)
The divergence still happens here. But it's much smaller. And rather than opening up in the early 1970s and expanding inexorably for generations, it looks like it really only opened up in the past 10 to 15 years.
To get a clearer view, you might want to look at the ration. It shows that the nominal portion of the gap is pretty clearly a direct consequence of the extent of the Great Recession:
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/3911068/fredgraph-2.png)
The best cure is not a huge structural overhaul of the American economy. It's for Janet Yellen and her colleagues at the Federal Reserve to be extremely cautious about raising interest rates. High unemployment makes it easy for employers to skimp on paying their workers, while stretches of full employment push the ratio up. To get back to pre-2000s level, we'll need more years of recovery.
The inflation thing is important
The popularity of the inflation-mixing chart among liberals is unfortunate, and has prompted public confusion. But the idea of some conservative wonks like James Sherk of the Heritage Foundation to note the issue, debunk the chart, and then move on is also misguided.
The story of the diverging price indexes is meaningful and important.
It tells us, most of all, that the enormous productivity increases inside the information technology industry have not yet led to higher productivity (or lower prices) elsewhere in the economy. The internet has not made it cheaper to get a college degree or to see a doctor. For all the talk of "disruption," computers have had a surprisingly small impact on the economy outside of the media and entertainment sectors.
Real wages really have risen much too slowly over the past 40 years. But while Clinton's version of the chart makes it look like rising productivity isn't part of the solution, looking at the divergent price indexes clarifies that it is crucial. For real wages to rise, we need the things middle-class families spend the bulk of their income on to get cheaper. That means more productivity in the big housing, health, and education sectors — not more pessimism about the potential of productivity.
