Two posts back, I explained why the "Great Vacation" idea doesn't pass the smell test. If U.S. unemployment had been caused by a negative shock to labor supply, we should have expected to see an increase in real wages.
Casey Mulligan, one of the leading proponents of the Great Vacation story, responded on his blog:
A number of bloggers have recently discovered real wages as a labor market indicator. They are at least 3 years late to the party.
Three years ago I blogged about the effect of labor supply on real wages.
I noted how real wages had risen since 2007, and predicted that they would begin to decline in 2010.
I have continued to update this work, eg here, and here.
The fact is that the real wage time series fits my recession narrative very well.
Well, in response to that, let's look at the numbers. Here, courtesy of FRED, is a graph of real compensation per hour in the nonfarm business sector:
A negative shock to labor supply should be associated with a spike in real compensation per hour. Looking at this graph, do you see such a spike? I do not. In fact, if I were to tell you that there had been a Great Vacation, and asked you to point out its beginning on that graph (without showing you the gray bars), you would probably say that it began in 2003, or maybe 2006 or 2009. You would not predict that a supply-driven recession began in 2008, when our real recession actually began.
Yes, it is true that real wages rebounded fairly rapidly from the trough to which they fell at the beginning of the Great Recession. And it is true that they climbed slightly higher after that, in 2009, reaching a peak about 2% higher than their 2006 peak. So Mulligan's statement that real wages rose during the Great Recession is correct.
However, note the size of the rise. There is no discernible increase in the rate of growth of real wages during the Great Recession. The wage growth to which Mulligan refers was slower by far, for example, than the growth that occurred between 2000 and 2004. If the Great Recession were caused by a massive negative labor supply shock, we would expect to see wages accelerate as employment fell. They did not. And the sharp downward spike in real wages in 2008 is especially hard to reconcile with a Great Vacation story.
I maintain my original case that the wage data shows no sign of a Great Vacation. If a Great Vacation in fact occurred, it had to have been a much more complicated sort of thing than the kind of negative labor supply shock that is taught in Econ 101.
"Three years ago I blogged about the effect of labor supply on real wages."ReplyDelete
Just for fun, try clicking on Mulligan's link and see how much time he actually spends on the effect on real wages. (control-f helps)
Interesting. I clicked through to look at one of the examples he gave of how he had updated his explanation. I'm not an economist and the math was more than I wanted to puzzle through, and it's the day after Xmas so I just skimmed, but when I checked out his illustrations I realized that the one thing he did NOT include was real wages. I have a feeling that if a trained economist read his paper they would spot some funny assumptions being quietly slid over.ReplyDelete
Off hand I can think of several other reasons real hourly compensation and productivity per manhour might go up under the circumstances. When workers are laid off, frequently it is the most recently hired who are let go first. Their wages are typically lower than the average for the firm. Health insurance is part of total compensation in many cases, and premiums have been rising much faster than inflation, even recently. Any given labor supply can always increase output per manhour in an emergency, or when forced to. That happens when they are scared for instance. Or when their employer demands it.ReplyDelete
I noticed Mulligan stated that the figures are not inconsistent with his hypothesis. When there are several possible causes for a phenomenon, inconsistency is a very weak threshold -- much weaker than it sounds rhetorically.
Finally, with what confidence can we accept these statistics which the Labor Department publishes. There are no error bars. Sources of uncertainty are numerous.
In Mulligan's defense he might say these are the only, and therefore the best, figures available. That is a weak defense.
For those who are interested, here are a couple of links to BLS sampling methods:ReplyDelete
Here is a quote from the Current Population Survey:
"Sources of errors in the survey estimates. There are two types of errors possible in an estimate based on a sample survey—sampling and nonsampling. The mathematical discipline of sampling theory provides methods for estimating standard errors when the probability of selection of each member of a population can be specified. The standard error,
a measure of sampling variability, can be used to compute confidence intervals that indicate a range of differences from true population values that can be anticipated because only a sample of the population has been surveyed. Nonsampling
errors such as response variability, response bias, and other types of bias occur in complete censuses as well as sample surveys. In some instances, nonsampling error may be more tightly controlled in a well-conducted survey, through which it is feasible to collect and process the data more skillfully.
Estimation of other types of bias is one of the most difficult aspects of survey work, and adequate measures of bias often cannot be made."
Even for pure sampling errors they use only a 90% confidence interval, which is misleading. The quality of the people actually doing the surveys is a source of nonsampling error not independently explored: the bureau emphasizes that its interviewers are "well trained" but having worked in the Dept. of Labor I don't think this assurance should be taken on faith.
The employer surveys also have intrinsic problems. For instance, while earnings data are reasonably complete due to IRS reporting, hours are not. They have to be obtained by sampling methods subject to the problems in the population surveys.
Remember, a wage contract specifies how much a worker will be paid for each hour of work but it does not specify how much work shall be done during that interval. It can vary as much a 40 percent even for the same worker in the same year doing the same job for the same hourly wage, which I know from personal experience both as an employee and an employer.
We are left with the problem of the error bars at a 95% confidence interval. I doubt the last decimal place can be justified.
Adding on to Luke Lea's comments, IIRC, these effects are normal (ISTR Krugman mentioning them).ReplyDelete
Companies keep 'core' workers more than 'non-core'; the second group is a lower-paid group (e.g., temps).
Sources of non-sample bias in bls earnings surveys (from bls internal report):ReplyDelete
• Cooperative Respondents: establishments
that have reported regularly and consistently
since enrollment in the survey sample;
• Intermittent Responders: establishments
that reported for a while, stopped, then
resumed reporting (e.g., they may have been
converted from “refusal” status to active
reporting status after an interval of time).
• Dropouts: establishments that reported for a
number of data collection cycles, then
stopped reporting before their period in the
sample ended; and
• Nonrespondents: establishments that
declined to participate in the survey when
initially contacted by the survey program.
All of the protocols focused on why respondents
agreed (or failed to agree) to participate in the survey
programs, and addressed issues such as:
• company policies about participation in
• company decision-making about
government survey participation
• concerns about confidentiality
• clarity of data requests
• relationship of frequency and timing of data
requests to nonresponse
• availability of requested data in
establishment records and respondent access
to those data
• preparation required to provide requested
• perceived relevance of collected survey data
• experiences with BLS personnel and effect
on survey participation
[Morgenstern's little book on the accuracy of statistical economic data is a must read for these sorts of issues; it will give you a heart palpatations if youd depend on accurate date to verify or falsify economic hypotheses.]
I notice that Mulligan doesn't have comments on his blog.ReplyDelete
Barry, what are you talking about? I have two comments on the Mulligan post that Noah linked to. There's no evidence he reads comments though.ReplyDelete
Odd. When I was over there before, there were no comments, and no obvious place for comments.ReplyDelete