Jan 11th 2011, 18:53 by R.A. WASHINGTON
THE
Wall Street Journal's
Sudeep Reddy has captured a lot of attention today with an
interesting story on falling wages, the impact of a deep labour market slump. He writes:
But the decline in their fortunes points to a signature outcome of the long downturn in the labor market. Even at times of high unemployment in the past, wages have been very slow to fall; economists describe them as "sticky." To an extent rarely seen in recessions since the Great Depression, wages for a swath of the labor force this time have taken a sharp and swift fall.
The only other downturn since the Depression to see similarly large wage cuts was the 1981-82 recession. But the latest downturn is already eclipsing that one. Unemployment has stood above 9% for 20 straight months—longer than the early 1980s stretch—and is likely to remain above that level for most of 2011, putting downward pressure on wages.
Mr Reddy collects an impressive array of anecdotes to support this assertion. He finds an ex-money manager working at Starbucks, a training manager at a manufacturing firm who now works as a janitor, and so on. The stories reflect his thesis, that long spells of joblessness tend to conclude (if it all) with a re-hiring at a much lower wage. And yet Mr Reddy runs into trouble when it comes time to cite data:
Overall, U.S. wages continue to grow, but at a slow pace. Wages and salaries for civilian workers were up 1.5% before adjusting for inflation in the 12 months ended in September, according to the Labor Department's comprehensive Employment Cost Index, which compares wages in the same jobs and doesn't reflect wages of people switching careers. Over the same period, consumer prices rose 1.1%.
Here's a chart of average weekly earnings for all employees since 2007:
We observe wages levelling off and dipping just slightly during the collapse in late 2008, but by the time early 2009 rolls around earnings are rising again. Now, this trend doesn't mean that Mr Reddy is describing a false trend. When we begin pulling apart the weekly earnings data, we see that the biggest contribution to the increase is a rise in hourly earnings (up 1.8% over the past 12 months), followed by a rise in average weekly hours (up 1.5%), while total payrolls are only up 0.9%.
So what we see is a two-track labour market. Workers who never lost their jobs saw, on average, a slight decrease in hours worked and a pause in wage increases, but since that time they've been able to work more and have potentially enjoyed pay increases. But this could coincide with a group of jobless workers who have struggled to find work and who can generally only do so at a significant wage cut relative to their previous pay.
How can we explain this? I mentioned a few explanations ventured by
Rob Shimer in a
recent post:
One big issue is the problem that nominal wages aren't very flexible in a downward direction. Another issue could be that since existing firms aren't motivated to hire new and cheap workers, new firms are needed to absorb jobless workers, but new firm creation is hampered by tight credit conditions. Mr Shimer also speculated that unemployed workers could somehow be different—uniquely unskilled or improperly skilled—or they could be pinned in place by housing conditions in particularly bad job markets.
Is downward wage rigidity a problem? Mr Reddy's anecdotes indicate that many of those who've been without work for a long time are willing to take new jobs at significant pay cuts, but perhaps others are still holding out for the wages they're used to.
On the other hand, there may not be jobs available for them. Why would that be the case? Why wouldn't firms swap out older, more expensive workers for the cheaper unemployed ones available to them? One possibility is that firms are worried about the disruptive impact of such workforce turnover and have decided that it's better to keep employing existing labour at existing wages. But then we might expect new firms to start up and hire jobless workers; if the unemployed were just as productive as the employed, new businesses could operate at a significant cost advantage over competitors. But Robert Hall argues that credit conditions remain tight for new businesses, who are the big job creators.
Or it could be that jobless workers are simply much less productive than those who continue to work. Ragu Rajan indicates that this kind of structural explanation could be behind most current unemployment, and he therefore emphasises the importance of retraining. But if so many workers are now too unproductive to hire, one has to ask why firms had them on payrolls before the recession. Mr Rajan points to the unusual growth and subsequent collapse in the construction industry, but as Mr Shimer notes unemployment has basically doubled among all subgroups within the labour force. The data seem not to point toward structural factors as the primary driver of unemployment.
Perhaps the problem is a shortfall in demand, which is preventing existing firms from expanding. It could be that the real interest rate simply isn't low enough to induce firms to invest in new plants and equipment—investments that would produce corresponding jobs.
These are the factors with which economists are currently wrestling in an attempt to understand unemployment. I do think it's worth pointing out that a little bout of inflation would be helpful in resolving all of the above issues, with the possible exception of structural skills mismatch. So I continue to find perplexing criticism of the Fed's decision to resume easing.
