Several folks have pointed to several signs that suggest inflation may be a big story in 2014.
One of the most common arguments is that labour shortages, especially in manufacturing and construction, meaning wages will have to go up to fill job vacancies.
However, Bank of America Merrill Lynch’s Ethan Harris is not buying that argument at all. From his Dec. 19 note to clients:
By some accounts, inflation is always just around the corner, even if the source of concern changes. “Inflationistas” first pointed to Fed liquidity, then to global cost pressures, and then to a weakening dollar. The most recent concern is that wages are about to pick up. After all, the unemployment rate has fallen 3 pp from its 2009 peak and the inflation-neutral unemployment rate (or NAIRU) has risen due to structural shocks to the labour market. Indeed, there are already signs of labour cost pressure. Wage inflation is coming. The Fed is behind the curve.
We disagree and argue that this is just the latest false alarm on inflation. Judging the amount of slack in the labour market requires looking at both the structural damage to the market and the amount of hidden unemployment. A reasonable weighing of the evidence suggests there is still anywhere from 1pp to 2.5pp of slack in the labour market. The pockets of wage pressure are just that: labour compensation is weak by almost any measure. Moreover, even if wages do inch higher, we are a long way from the normal 3 or 4% rate that could start to create serious pricing pressure.
Harris offers a pretty broad and thorough dismantling of the wage growth argument by showing that there’s actually a lot of slack in the labour market. We’ll highlight just a few key points.
We frequently read about the Beveridge curve, which reveals labour market efficiency by charting job openings versus the unemployment rate over time. Currently, we are watching the unemployment rate fall, yet the number of job openings remain higher than they were during earlier periods when were at similar unemployment levels (see BAML’s Chart 3 below).
This is a sign of labour market inefficiency. Some believe this means supports the idea that there aren’t enough skilled people to fill jobs.
“The shift to the right in the curve is often seen as suggesting structural problems in the labour market — companies are having trouble matching workers to jobs,” wrote Harris. “Indeed, the curve appears to have shifted to the right by about two percentage points.”
So does this elevated level of job openings relative to unemployment mean inflation is coming employers are forced to pay more?
Not necessarily. Harris says that this counterclockwise-like move in the curve is cyclical in that it is typical of past labour market cycles (see BAML’s Chart 4 above).
“This cyclical shift may occur because companies first post new positions and then actual hiring comes with a lag,” he said. “It also has become increasingly easy to post jobs on the Internet even if a company is in no hurry to fill the job. More broadly, there is evidence that companies are simply not trying as hard to fill openings.”
Harris takes issue with the argument that certain industries, like manufacturing and construction, are experiencing worker shortages whereas other industries face oversupply.
Looking at more micro data around job matching also suggests that most of the move is cyclical. These studies look for three kinds of potential job mismatch: based on industry, occupation and geography. For example, Lazear and Spletzer look at whether there is job mismatch across industries. If there was mismatch then there should be some industries with unusually high vacancies relative to unemployment compared to other industries. This would happen if some people are “trapped” in the wrong industry or occupation. Their results show that mismatch was high during the recession, but has now fallen back to pre- recession levels. They find the same results across occupations, and other studies confirm there is no unusual mismatch across geographies.
Digging deep into the underlying data, we find a very consistent pattern of spare capacity and low wage growth. The unemployment rate is higher than its 2000-07 average in all 15 industry groups and all nine occupational groups. In most groupings the excess unemployment rate is relatively close to the overall average. Even in the red-hot mining industry the unemployment rate is a few tenths higher than normal. The unemployment rate is also higher than normal in almost all of the nation, for all racial groups, for all major age groups, for both men and women, for all levels of educational attainment.
So, there doesn’t seem to be a lot of evidence that the construction and manufacturing industries are having a particularly difficult time filling jobs.
Harris also notes that if certain industries were facing labour supply constraints in a way that would affect wages, then the employment cost data would reflect these differences. They don’t.
The same story applies for labour compensation. If there was a major misallocation of workers, we would expect wide differences in compensation growth. The Employment Cost Index (ECI) slices the data into 30 components based on occupation and industry. Of these groupings, 21 have compensation growth of 1.5 to 2.5% over the last year, with 8 lower and one higher. Despite the alleged skill shortage in manufacturing and construction, compensation growth is just 1.8 and 2.1%, respectively. That is virtually identical to the overall average of 1.9%.
So, while some signs suggest labour shortages and dislocations, a closer look suggests that really isn’t the case.
What About Employers Complaining?
We often read anecdotes about how employers are struggling to find skilled workers for job openings. Indeed, the Federal Reserve’s Beige Books and various regional surveys reveal as much.
However, Harris thinks that’s just a lot of talk.
“The Fed’s Beige book has been highlighting skill shortages since 2010, even as it confirms no broad-based wage pressure,” he said.
Here’s a table breaking down the related language.
Basically, all of this talk from employers about labour shortages is just noise.
The Biggest Mystery Of All
Where have all of the construction workers gone?
After all, just a few years ago we had an enormous housing bubble, which came with tons of construction workers.
Harris thinks they’re still out there.
We are particularly sceptical about a broad-based skill shortage in construction. Housing starts have only recovered to half of their peak level. Both total construction employment and specialty trade contractors are still down about 25% from the 2006 peak. Where did all these skilled workers go? While some workers may have chosen early retirement and some immigrants have left, the unemployment rate is still almost 14% in the construction industry, compared to an historic norm of 8%. Construction skills do not require constant updating. These workers are also unlikely to have found better paying jobs elsewhere. If a skilled carpenter is forced to a take a job in the retail sector, surely he will be easy to entice back into construction.
Again, all of this is just a sampling of Harris research note.
But his bottom line is that there is definitely a lot of labour market slack in the economy. And he believes the Federal Reserve knows this and he believes this is why they’ll continue to keep monetary policy easy.
“This puts us in the same dovish camp as Janet Yellen,” said Harris.
Basically, the Fed will stay dovish until we really do have to worry about labour shortages, wage hikes, and inflation.