Structural vs. Cyclical Unemployment Revisited: Doing Nothing Is Not a Smart Option

An update on the question of structural vs. cyclical unemployment, this time with respect to policy options for each. For background, see these previous posts:  on how economists define or distinguish between structural and cyclical and a look at the situation in 2011.  Time is short and specialization is efficient, so I’ll quote Mark Thoma on this (and he’ll quote Peter Diamond and Christie Romer):

I wish I’d remembered point three when I wrote recently about the difficulty of separating cyclical and structural unemployment. I was saying, essentially, the same thing that Peter Diamnond says here (via):

Second, for the current moment, the argument about the aggregate demand side is academic, in the negative sense of the word. Current estimates I have seen of how much of the increase in unemployment from a few years ago is “structural,” rather than due to inadequate aggregate demand, still leaves enough need for aggregate demand stimulation that it is clear what direction is needed for further policies.

Third, I am skeptical of the value of attempting to separate cyclical from structural unemployment over a business cycle…. The tighter the labor market and the more valuable the filling of a vacancy, the more a firm is willing to hire a worker who is a less good match, who may need more training…. [A] worker who might be viewed as structurally unemployed, as facing serious mismatch in the current state of the economy, may be readily employable in a tight labor market. The common practice of thinking about the extent of unemployment as a sum of frictional, structural and cyclical parts misses the point…. [D]irect measures of frictional or structural unemployment… dependent on the tightness of the labor market… have limited relevance for the role of demand stimulation policies. The idea that the US economy is not adaptable and capable of dealing with the need for skills and jobs to adapt to each other is peculiar, given the long history of unemployment going up and down. When the labor market is tight and firms have trouble finding workers, they reach out to places they have not looked before and extend training in order to find workers who can fill their needs. Supporting current stimulus policies as very good for the economy is entirely compatible with taking care to avoid future inflation.

Thus, no matter how you slice it or how you define it — and even with a very generous interpretation of the structural estimates — there is still plenty of cyclical unemployment (or, perhaps more precisely, employment that will respond to an increase in demand) to worry about, and plenty for policy to do.

But suppose that, contrary to what the estimates are telling us, there is a large, dominant, structural component. Does that mean we sit on our hands and do nothing? Nope, Christy Romer makes a point I’ve made many times. Even if the problem is structural, there are still things we can do to help:

There’s this debate going on over what the source of the unemployment is: Do we not have enough aggregate demand, or is it structural? What frustrates me is the advocates of the structural theory go from saying it’s hard to turn construction workers into nurses to saying we should do nothing. If you think our problem is structural, there are things we should be doing: money for training, or helping people get out of their mortgages, or massive investment in Detroit. I don’t believe that skills are the problem here, but if that’s your point of view, there’s still a lot we can do. Saying it’s structural is not the same as saying it’s not our problem.

No matter the cause, we’ve dropped the ball on the unemployment problem (and have yet to pick it up). As I said last week, “We have enough money to pay for military action in Libya, but not for job creation?” But Bob Herbert’s last column at the NY Times says it better:

Losing Our Way, by Bob Herbert, Commentary, NY Times: So here we are pouring shiploads of cash into yet another war, this time in Libya, while simultaneously demolishing school budgets, closing libraries, laying off teachers and police officers, and generally letting the bottom fall out of the quality of life here at home.

Welcome to America in the second decade of the 21st century. An army of long-term unemployed workers is spread across the land, the human fallout from the Great Recession and long years of misguided economic policies. Optimism is in short supply. The few jobs now being created too often pay a pittance, not nearly enough to pry open the doors to a middle-class standard of living. …

The U.S. has not just misplaced its priorities. When the most powerful country ever to inhabit the earth finds it so easy to plunge into the horror of warfare but almost impossible to find adequate work for its people or to properly educate its young, it has lost its way entirely. …

Millions and millions of people still unemployed, the prospects of a slow, slow recovery of employment ahead of us (along with the permanent damage that long-term unemployment brings about), and few people in Washington seem to care.

Is Our High Unemployment Structural?

So following up on my post on the types of unemployment, when unemployment is high, how do we know if it’s due to structural or cyclical causes?  The answer is important because it tells us what kind of policy actions to take.  Do we need stimulus? (addresses cyclical), or do we need job-retraining, relocation, and education? (targets structural).  There is no clear cut way to tell how much of each type exists.

For example, suppose unemployment is 9%, much as it is today.  Let’s assume that this is a seasonally adjusted number, so we can assume that none of the 9% represents seasonal.  Let’s further assume, just for the sake of arguement, that 5% points of the 9 points represent frictional unemployment.  I should note that this is a somewhat controversial point.  I do not personally think that frictional is in fact that high nor that it we cannot get below that point. Personally I don’t see why frictional should be greater than 2-3%. My views on this are heavily influenced by Bill Mitchell and the MMT people. But since mainstream economics since Milton Friedman has defined it so, let’s accept frictional unemployment as being 5% just for the sake of this argument.  That means there’s another 4% points of the 9 points that must be either structural or cyclical. How can we tell the difference?

There’s no clear-cut way to tell as in doing some direct survey.  Instead we need to look at some indirect indicators.  Two of the more indicators we could look at are: the job seekers  to job openings ratio and unemployment by industry or state.  If our high unemployment is primarily due to structural factors then we should see a low ratio of job seekers to open jobs, and we should see wide differences in unemployment by industry or geography.  If we see low ratio of job seekers to job openings, then that means there are jobs in aggregate but apparently those seekers (unemployed) are unqualified. If we see wide differences in unemployment by industry or geography, it indicates again that jobs exist but they aren’t where the workers are.  On the other hand if the seeker-openings ratio is high, then that means there just aren’t enough jobs period – workers outnumber jobs.  If unemployment is high across all industries and locations, then again there is no unsatisfied sector. There is no structural unemployment.

So with that in  mind, let’s look at the data.  First, via StateofWorkingAmerica.org, we have the job seekers-to-job openings ratio (called the JOLTS data series):

Job Seekers to Job Openings Ratio, US, 2001-2010

 

 

 

 

 

 

 

 

 

 

 

 

 

Yep, it’s pretty high – between five and six workers for every open job.  Next, let’s consider unemployment across industries.  This via Paul Krugman:

here’s the increase in unemployment 2007-2010 by industry of previous employment:

DESCRIPTION

See the structural shift? Neither do I. As others have noted, basically unemployment doubled for every industry, every occupation, every state. Where are the sectors/occupations/regions gaining jobs? Nowhere to be found. There’s nothing structural about it.

But what about geographic mismatch? Let’s go to the state-by-state unemployment numbers from Calculated Risk:

Unemployment Rates by State, Dec 2010

 

 

 

 

 

 

 

 

 

 

 

Still no mismatch. Yes, there are three states under 5% and two more under 6%, but these states are the Dakotas, Nebraska, New Hampshire and Vermont.  Relatively speaking, nobody lives there.

So overall this means it’s not structural unemployment that is keeping unemployment high.  It’s cyclical unemployment.  That means the route to lower unemployment is more stimulus, not austerity. Better yet, a jobs guarantee program could work very effectively.

Types (Causes) of Unemployment

Economists classify unemployment into four types according to what caused the unemployment.  If we assume the goal is “full employment” (never mind how we might define or measure “full” right now – there’s mischief there), then what we’re really saying is that our goal is for the economy to create an appropriate a job for every willing and able worker. Then, if we find that the economy is producing some unemployment, we can and should ask ourselves: why hasn’t the economy produced an appropriate job for each worker?  We then essentially classify the economy’s failure to put every available worker into a job as due to one of four reasons.

  • Seasonal
  • Frictional
  • Cyclical
  • Structural

From a policy standpoint, two are troubling and two aren’t.  The two non-troubling causes or types of unemployment are seasonal and frictional.

Seasonal unemployment means the worker (and his/her skills) is unemployed because it’s the wrong time of year.  A classical example is a downhill ski instructor in July, or part-time holiday sales clerk in February. When time passes, winter for the instructor or November-December for the clerk, they’ll be employed again.  Employment data can be statistically massaged to remove the seasonality and so we typically look at unemployment data that is “seasonally adjusted” and we ignore seasonal unemployment.  Besides what kind of policy could we implement to fix it? Legislate Christmas in July? or pass a low mandating snow?

Frictional unemployment means that actually there is a job for the unemployed worker at the time we measured unemployment, it’s just that the worker and the job haven’t matched together yet.  Frictional represents people looking for jobs that are indeed out there for them.  Again, policy, at least at the macro level is not needed here. These workers will find their finds.  At any point in time in a healthy market economy there will always be some people between jobs.  That’s a good thing since it means people are getting matched to jobs where there’s a better fit. The only ways to drive frictional unemployment to zero is to either have instantaneous job searches or nobody ever moves to a better job.

That leaves cyclical unemployment and structural unemployment.  The difference is important in theory but difficult to identify in practice. Cyclical unemployment is workers who are out-of-jobs because employers cannot sell enough goods.  In other words, the economy is depressed. If it grows faster these people will get hired. Cyclical unemployment can fixed by appropriate macro-level stimulus policies.

Structural unemployment however, while of interest to policy-makers, cannot be fixed as easily by macro-level stimulus policies. Structural unemployment occurs when there is a mismatch at the individual worker-level between the skills, experience, qualifications, and location of the unemployed workers and what’s required for the open job opportunities. Examples of structural unemployment at the national level include:

    • technological obsolescence – we no longer need those skills (or as many with those skills). The classic example is horse livery workers once we switched to automobiles around 1910. A more modern example might include photographers or developers of film (the old silver-halide, analog stuff) after the world has switched to digital photography.
    • location mismatch – the unemployed workers live in one state and the open job opportunities exist somewhere else.
    • educational mismatches – the jobs being created require higher educational degrees or specific trainings, but the unemployed workers don’t have those qualifications.
    • inexperience – the firms hiring all want highly experienced workers or only workers who are currently employed.  The unemployed workers either don’t have experience or don’t have recent/current experience.

Structural unemployment can be reduced through policy actions, but they are different, more micro-level policies.  For example job retraining programs can reduce technological obsolescence. Programs to help people move and relocate will address location mismatch. Educational support and grants will address educational mismatches. Both government direct-hire jobs guarantee programs and employer willingess or incentives to do on-the-job training can address technological, educational, and inexperience issues.

 

Frictional, Structural, Cyclical Unemployment Defined

Mark Thoma explains the difference between cyclical, structural, and frictional unemployment:

As I noted in a previous post, economists define three types of unemployment: frictional, structural, and cyclical:

Frictional unemployment is defined as the unemployment that occurs because of people moving or changing occupations. Demographic change can also play a role in this type of unemployment since young or first-time workers tend to have higher-than-normal turnover rates as they settle into a long-term occupation. An important distinguishing feature of this type of unemployment, unlike the two that follow it, is that it is voluntary on the part of the worker.

Structural unemployment is defined as unemployment arising from technical change such as automation, or from changes in the composition of output due to variations in the types of products people demand. For example, a decline in the demand for typewriters would lead to structurally unemployed workers in the typewriter industry.

Cyclical unemployment is defined as workers losing their jobs due to business cycle fluctuations in output, i.e. the normal up and down movements in the economy as it cycles through booms and recessions over time.

In a recession, frictional unemployment tends to drop since people become afraid of quitting the job they have due to the poor chances of finding another one. People that already have another job lined up will still be willing to change jobs, though there will be fewer of them since new jobs are harder to find. However, they aren’t counted as part of the unemployed. Thus, the fall in frictional unemployment is mainly due to a fall in people quitting voluntarily before they have another job lined up.

But the drop in frictional unemployment is relatively small and more than offset by increases in cyclical and structural unemployment.