Opinion

Felix Salmon

Should the Fed be worried about wage inflation?

By Felix Salmon
May 19, 2011

Kathleen Madigan blogs — without linking to — the latest FOMC minutes, and picks up on an important point: the Fed is keeping rates low largely because it sees little risk of wage inflation. If and when wages pick up, the Fed isn’t particularly worried about inflation. “But if wages pick up,” she writes, “the Fed may step in.”

Mark Thoma is unimpressed by this: “We are much too worried about a wage-price inflation cycle breaking out and causing problems,” he says. “If the Fed is too trigger happy, it could snuff out the recovery it is hoping to bring about.”

There’s a good reason to be sanguine about rising wages, and that’s rising food and energy prices. The Fed prefers to look at core inflation when setting monetary policy, on the grounds that food and energy prices are too volatile for it to be able to control, or tend to be cyclical. But they loom large for Americans when it comes to the cost of living, and wage increases are needed to pay for higher grocery bills and more-expensive tanks of gas.

John Kemp does a good job of explaining why the Fed should look at headline rather than core inflation. But at the same time, it probably should keep an eye on wages as well. After all, wages are pretty much the one thing which wasn’t hurt by the recession:

fredgraph.png

To put this in terms even a politician can understand, the average weekly paycheck is now $787.19. That’s $59.78, or 8.2%, higher than it was at this point four years ago — and the rate of increase in wages is not slowing down. If you have a job at the next election day, you’ll likely be earning between 8% and 10% more than you were when Barack Obama was elected. This country has many problems, unemployment first and foremost among them, but stagnant wages and employees being underpaid are pretty low down the list, if they’re on it at all.

I do think that productivity gains should go more to labor and less to capital, but in the first instance that should take the form of increased hiring, rather than wage increases for the already-employed. Those of us with jobs are the fortunate ones — and yes, if we’re paid more, we’ll spend more, and that in turn is likely to show up in higher prices. The Fed is probably right to be worried about wage inflation. But that just means that it should look for ways to divert that money into new hiring, rather than raising rates to choke it off altogether.

Comments
22 comments so far | RSS Comments RSS

Whatever wage inflation exists is not caused by low interest rates. That would only happen if the low cost of capital was causing competition for labor, and that is clearly not happening, as unemployment is higher than it was four years ago.

It would be risky for the Fed to assume that inflation is caused by low interest rates – inflation is caused by demand exceeding supply, and that does not always correlate with interest rates. It’s possible that the average weekly paycheck has increased because a) the workers who didn’t lose their jobs are working more overtime, and b) more jobs were lost among low income earners, which skewed the average.

The Fed shouldn’t be worried about wage inflation; they should be worried about what the banks that they lend money to every night are doing with that money. If they did that six or seven years ago, we wouldn’t still be trying to climb out of this mess.

Posted by KenG_CA | Report as abusive
 

“After all, wages are pretty much the one thing which wasn’t hurt by the recession”

Huh? Look at your graph. Draw the trend line from before the recession. Draw the trend line from after the recession. Not the same line! The line after the recession is very far beneath the line before the recession. Therefore, wages were hurt by the recession.

Posted by q_is_too_short | Report as abusive
 

What about showing median earnings rather than average? Also, presumably the Fed is more worried about wage GROWTH than about LEVELS, no?

Posted by psummers | Report as abusive
 

I agree with the previous comment. John Kemp is making a fairly simple “How to lie with statistics” mistake. He should be looking at the mean rather than the median. I don’t entirely blame him as I couldn’t find the median income in about 10 min of searching the Fred site. There is a fed report (http://www.federalreserve.gov/pubs/bull etin/2009/pdf/scf09.pdf) which pretty clearly shows that in 2004–2007 the median wage went down while the mean went up.

If the greater productivity is going only being passed on to executive vice presidents and investment bankers I don’t care (in either the micro or the macro sense; in the macro sense I don’t care because those people will save and not spent the surplus and hence won’t boost the economy). Meanwhile, school teachers and police officers are taking 3% pay cuts in Florida (and something similar in Wisconsin). This might drive inflation in the NY botiques, but not in Sears.

Posted by ralmond | Report as abusive
 

Median earnings could have increased, also, if low income workers were laid off. Think of a department of 8 employees plus a manager, and the lowest paid worker is laid off. Not only does the average income go up, but so does the median.

What would be more interesting is total personal income, not including capital gains and dividends, divided by the total number of people working + seeking work. I’m guessing that didn’t increase 8% over those four years.

Posted by KenG_CA | Report as abusive
 

And if the fed finds a way to “divert that money into new hiring”, it should then look for a way to make us all more attractive and physically fit.

Similarly for productivity gains going to labor vs. capital. In the short run it generally goes mostly to capital; in the long run it generally goes entirely to labor; there’s precious little that regulation, let alone monetary policy, can do to change that.

Posted by dWj | Report as abusive
 

“John Kemp does a good job of explaining why the Fed should look at headline rather than core inflation.”
I disagree. Goldman economist Jan Hatzius looked at differences between headline and core inflation. Since 1987, when headline and core inflation diverged, the two numbers later converged, with the headline number doing *all* of the converging.

Posted by AHooks | Report as abusive
 

Average wage is not so average, if the gap between average and median increases: http://www.ssa.gov/oact/cola/avg_median. gif

Posted by GRRR | Report as abusive
 

But the whole point of productivity gains is eliminating jobs.

If a MacDonald’s can redo work practices or install a better frying machine there entire goal is so that they can lay off one worker per shift. It’s like a tontine. The workers each get an extra hour of work per week taking them from 34 to 35 hours and the management pockets the difference. The owners hope the good news means they can sell their shares for more. Median and mean incomes rise as capital investment eliminates the least efficient workers.

The trick is to come up with some excuse for giving people money so they can buy the newly more efficiently produced products and pumping some of that management surplus back into the economy.

Posted by spiffy76 | Report as abusive
 

“Those of us with jobs are the fortunate ones — and yes, if we’re paid more, we’ll spend more, and that in turn is likely to show up in higher prices.”

That would only be true if the increased wages compensated directly for the loss of income of the unemployed. The working are still pumping less back into the economy than when they earned less but there was higher total employment, right? And shouldn’t some be offset by the higher savings rate?

Posted by sleepyinhohokus | Report as abusive
 

Oh, quite. If Mankiw and Krugman agree on something (http://gregmankiw.blogspot.com/2011/05/ i-agree-with-paul-krugman.html), both must surely be wrong. No doubt they will appreciate your irrefutable demonstration of this: “because John Kemp says so!”

Posted by Greycap | Report as abusive
 

Wage inflation does not matter as much in the U.S. as it used to. It mattered more when the U.S. was a manufacturing oriented economy as opposed to service.

Posted by david3 | Report as abusive
 

You should factor in US inflation rate which over the last four years was 10.55%. Therefore by definition average weekly earnings are not keeping up with price increases. In fact average earnings are being pounded into the ground. Any wonder the US economy cannot dig itself out of its deficit trap it has created. The man from “Down under”. “Never just look at one side of the equation”

Posted by Rene888 | Report as abusive
 

Let’s see the same argument with median wages instead of mean wages as an ‘average.’ Go ahead, let’s hear it! And then, let’s throw in the unemployed folks — hey, they’re real people — and then take the median and do the analysis.

As for inflation, the CPI, for instance, is gimmicked. Go look at the changes to the methodology over the years for yourself.

The headline of this article should have a winky smiley on it;)

Posted by monist | Report as abusive
 

BTW, using the reference to ‘weekly’ earnings is really straining your objectivity quotient. The supposedly rising earnings your talk about are fueled by folks who are not paid daily, weekly or probably even twice=monthly. When talking about the movement of an average pulled by those at a salaried job, and those who receive bonuses, it is disingenuous to refer to ‘weekly’ wages. Real wages are what real people make, Reuters. I’m speaking to you, editor.

Posted by monist | Report as abusive
 

Oh, and let’s use real dollars, not nominal ones.

Posted by monist | Report as abusive
 

Oops, maybe I’m on at wsj.com — if so, disregard my näive comments.

Posted by monist | Report as abusive
 

I read where the Fed changed all their publicly released statistics from the base 10 number system (too volatile, all over the place) to a number system based on a quantum average of all number systems. Likewise since Einstein redefined time in general and specific terms and nobody really knows what time it is, the Fed is using a time frame based on a weighted average of all time that corrects for the wild volatility of time. Considering these minor de-volitalization adjustments, price AND wage inflation is negligible (like the risk of a catastrophic financial crisis until 5 munutes after it actually happened back in 2008) and we have absolutely nothing to worry about ..

Posted by Woltmann | Report as abusive
 

Felix – the graph you’re using overstates wage growth. Average hourly growth rates have continued to decline.
The following graph illustrates the situation: http://bit.ly/jExhw6 Wages have gone nowhere. Hence: raw material prices will largely get passed on the consumers, but no hint that this is spreading into a wider rise in prices; and with labor costs flat, that means we get a one-time jump in consumer prices, but no persistent rise in inflation.

–Pstrick Wacker

Posted by patricknewyork | Report as abusive
 

Felix, two points: the data you’ve posted is raw data, unadjusted for inflation. Real weekly earnings (adjusted for inflation) have been in a steady decline. If real wages are in decline, they’re lagging inflation, not feeding it, and are not a cause of concern for the Fed. Furthermore, strip away the top 5% of wage earners, and even the data you’ve presented – unadjusted for inflation – would show a decline. Real wages are driven by the supply/demand balance in the labor force, not by any magnanimity of employers taking pity on the financial plight of their employees. Real wages go up only when labor is in tight supply, a situation we may not ever see again.

Secondly,your comment that “… productivity gains should go more to labor and less to capital …” implies that you’re bought into economists’ hype about productivity gains driving wage increases. Nothing could be further from the truth. The BLS data for productivity and wage increases reveals that there is absolutely no relationship. The demand for labor is the only thing that drives wage increases. Consider NAICS code 3341, the BLS code for the manufacturing of computer equipment. Although productivity in that field increased 3800% from 1987 through 2005, the highest rate of increase of any labor code, compensation fell by 24%. It fell because the hours worked in that field (the demand for labor) also declined.

Posted by Pete_Murphy | Report as abusive
 

You need to show a chart of avg. hours worked per week. Avg weekly earnings are not the same as average wage

Posted by rab5566 | Report as abusive
 

That Kemp article is just crazy wrong.

“The Fed is focusing on the wrong “output gap”. It should be looking at the output gap globally and in commodity markets, rather than in the United States and the labour market and manufacturing.”

The United States Federal Reserve should target a price level determined by global commodity markets. That is just an insane idea.

Posted by AASH | Report as abusive
 

Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/
  •