Bryan Caplan  

What Does Seasonal Employment Show About Nominal Rigidity?

Trend vs. Random Walk... Seasonal Employment (Not Wonki...
Casey Mulligan makes a thought-provoking point about summer employment:
[N]ational employment was two million to three million higher in July than it was six months before (employment estimates vary somewhat between the business surveys and the household survey).


But it's big news that the seasonal employment cycle continues to operate as normal. The cycle is, of course, the outcome of seasonal fluctuations in supply and demand, and Keynesian economists insist that supply and demand are not operating normally since the recession began and that the economy has been caught in a "liquidity trap."

This is a telling objection to weird Keynesian stories where the solution to nominal rigidity isn't lower wages.  But it's perfectly consistent with a normal Keynesian story of unemployment where excessive wages prevent full employment.  Consider: What's special about temporary summer jobs? 

1. Employers don't have to cut any worker's wage to cut workers' wages.  Just offer less than last year to the new batch of employees.

2. Insider-outsider issues aren't much of a problem.  Temporary workers, whatever their pay, pose little threat to the jobs of any full-time co-workers, so there's little reason to haze them.

3. The temporary workers leave their jobs before feelings of wage "unfairness" can do much damage to morale.

What we learn from seasonal employment, in short, is that if all workers were like temporary workers, there wouldn't be much of an unemployment problem.  Unfortunately, most workers aren't like temporary workers, and the unemployment problem is very real.  Employers who cut wages won't be popular, but they should be.

Comments and Sharing

COMMENTS (9 to date)
Doc Merlin writes:

College graduates have now a MUCH, MUCH, MUCH lower unemployment rate than people who aren't college grads. The fact that the vast majority of unemployment is at the bottom end is very telling and fits the story of price floors and payment mandates far better than it does any other story.

david writes:

On Casey Mulligan (and Mankiw): demand varies cyclically, too. This is why we used seasonally-adjusted data for most purposes. Tim Duy has graphs.

Kurbla writes:

"What we learn from seasonal employment, in short, is that if all workers were like temporary workers, there wouldn't be much of an unemployment problem. "

If they accept temporary jobs for less money? Sure. Every problem of every economic system can be solved if majority of people voluntarily accept worse position, and if we do not count that worse position as a new problem. It is trivial.

Hyena writes:

I want to throw in with Kurbla.

Just because we can get the market to clear at some price doesn't mean we should do so; if individual workers don't think the job is worth the money and wish to "bargain" with the market for more, shouldn't we defer to this opinion?

Jim Rose writes:


Jeff Miron started a major new literature by examining seasonality and asking why we adjust for seasonality and what is lost through these adjustments.

80% of post-war macroeconomic volatility is seasonal, as I recall, but this is assumed to have so little welfare costs that it is ignored.

There is a boom every summer and a recession every winter. Is this stabilised?

Maybe it is. The Fed was founded to end those seasonal banking panics during the harvesting and planting seasons because of spikes in the demand for currency. There were no seasonal banking panics in Canada because it had nation-wide branch banking rather than unit banking, and no federal monopoly over currency backing.

JPIrving writes:

Temporary high school and uni workers aren't supplying the same labor as the general unskilled population. It seems to me that many seasonal jobs can only reasonably be filled by relatively educated, higher ability young people.

I am thinking of positions where it is only appropriate to use young suburbanites or close substitutes: Camp counselor, caddy, maybe landscaper in some parts.

You can't bring a 45 year old, low skilled worker with metabolic syndrome to the county club no matter how low the wage. For that matter many aren't employable in even upper rage food service. At least no when there are healthy, presentable, young people available for $9/hr.

Mark writes:

I am assuming that most of these workers are low-skill workers. If this is true, doesn't the minimum wage present excessive nominal wages?

Ted writes:

Isn't it just more plausible teenagers have a lower reservation wage than the adult population? Not to mention a variety of government policies are keeping people's reservation wages from falling through a variety of mean-tested programs like mortgage and tax debt relief. Unemployment benefit extension prevent the reservation wage from falling as well. Also, if you have a mortgage and a family to raise, you are unlikely to go and get a minimum wage job anyway since it wouldn't cover your expenses. Demand is also seasonal anyway. Also, we shouldn't celebrate should wages fall. That would just be a sign that the economy sucks and expectations of growth are falling further due to a complacent Fed (and because prices tend to slow / fall before wages slow / fall you get another coordination problem which makes the situation worse). I also don't think it's outlandish to believe a perverse downward nominal wage rigidity exists so it could take a lot of time to adjust to a severe shock and return to "normal" in absence of Fed action.

Anyway, this doesn't prove anything, but Casey Mulligan has never understood how New Keynesian models work (for example, in his response to Gauti Eggertsson, Mulligan proved he doesn't even understand the internal workings of the NK model - at all.). All this proves is that real wages are reasonably rigid and heterogeneous agents have different expectations about real wages - which is something we already knew anyway.

@ Mark

Minimum wages are a real rigidity, not a nominal. They have qualitatively different effects. Real wage rigidities should generate exceptionally low vacancy rates. Of course, the evidence is then at odds with this conclusion since vacancy rates are higher than they should be given the unemployment figure. Of course, I would contend this might just be a data problem since we know the unemployment rate is biased downwards and we also know it is amplified during a recessions, where the error is approximately proportional to its severity. So, it's plausible we are severely underestimated the rate and so there really isn't excess vacancies, in which case real rigidities like the minimum wage / staggered Nash wage bargaining / and search & matching frictions are playing the biggest problem here.

Milton Recht writes:

It depends on how you look at wages. For example, see News N Economics blog post "The compensationless recovery" by Rebecca Wilder:

Real hourly compensation measures compensation for all workers, including wages, 401k contributions, stock options, tips, and self-employed business owner compensation....real hourly compensation....declined four consecutive quarters through Q2 2010, a first since 1979-1980. If the NBER dates the onset of the expansion at Q3 2009 (the first quarter of positive GDP growth in 2009), real hourly compensation will have dropped .7% through Q2 2010! That's pathetic compared to the average 2.5% gain during the first 4 quarters of expansion spanning the previous 10 recessions.
When one looks only at wages without employer benefit costs, it looks like wages are rising.

Since Keynes, more of the cost of compensation is in the form of employer-paid employee benefits, vacation, sick days, health, dental and life insurance, etc. Employers can decrease pay by lowering their share of benefit cost, which appears to be happening in the recovery phase of this recession.

Additionally, jobs today in industry, factories and services are much lower risk of death and injury than during Keynes' time. Employers today can reduce safety measures, which is a form of wage reduction, without lowering the nominal wage paid employees. Government wages statistics do not adjust for comparative job quality issues, such as safety. As a hypothetical example, suppose an oilrig owner cuts back on safety measures that increase the risk of injury and death to the workers on the rig. Isn't the increase risk of injury, the equivalent of a reduction in pay, especially if the employer would have had to pay more to hire with the increased work risk?

Short-term workers are not affected by benefit changes and job quality changes.

According to government statistics, there are many unfilled jobs. Looking at real hourly compensation, employers are cutting wages. One possible answer is that workers are rigid and not employers. The unfilled openings may show workers are unwilling to accept the lower wages and fill the available openings, especially if the pay reduction is in the area of benefit reductions.

Others argue that it is a skill mismatch, but at a lower wage, an employer will be willing to train an apt novice. Mankiw and other economists argue that it is the income cutoff levels in recent many government programs, such as for homeowners, etc. Income cutoffs for benefits are equivalent to a high marginal tax and a disincentive to work. Others say that a structural change in the economy is occurring.

There is also the possibility that workers have taken measures into their own hands to reduce employer costs of a worker. More people may be working off the books, which avoids the income cutoff effects, lowers employer costs, and increases employees hourly take home pay through income tax avoidance. I am curious if there has been growth in the underground economy during this recession.

Comments for this entry have been closed
Return to top