Comments on economics, mystery fiction, drama, and art.

Saturday, March 29, 2014

Nate Silver's new venture FiveThirtyEight

I think I have a sense now for what we're going to see at FiveThirtyEight, based on what we've been seeing since its launch, and that's exemplified by two recent posts, one in the sports section, the other in the economics section.

In the first post, Neil Paine delves into player performance data, prompted by the "Wait...he's still in the league?" question.  Using records for all position players active since 1973, he takes data on  player age and three years' worth of wins-above-replacement (WAR) data to estimate a logit model of survival over 5 year periods.  The equation takes the following form:

S[Y(t+5)] = f(WAR(t), WAR(t-1),WAR(t-2), Age(t)]


He provides us with a list of 20 players who in fact have survived from 2009 into the 2014 season, but who had the lowest probabilities of survival (ranging from 3% to 26%).

In the second post, Ben Casselman looks at the declining labor force participation rate and attempts to determine how much of it is a result of the recent recession, how much of it is a consequence of the slow recovery, and how much of it is likely to be permanent.  He presents a chart showing the projections he gets for the 2008-2014 period (which shows a labor force participation rate declining, but generally above the actual LFPR.  In a follow-up post, he elaborates on his conclusion, which is, essentially, that the LFPR has been declining in a way largely related to business cycle factors, not from longer-term changes in the economy.

These are both interesting question, in one way or another, to one group of people or another.  I have professional research interests in career length in MLB, having done research on player career length, looking at whether ethnicity affected career length and (separately) at whether being a union player representative affected career length.  I've also done some work looking at the dramatic drops in teenage labor force participation and at the similarly striking increases in labor force participation of those age 65 and over.  So I found their posts interesting.

And intensely frustrating.

In both cases, we are given very little information about the data sets (more, interestingly, in the baseball piece than in the labor force participation piece, where we know nothing about the time period or the variables used in the analysis).  Neither Paine nor Casselman presents the actual statistical results, either in their posts or in a separate, linked document.  So we know nothing about the statistical properties of their results (e.g., whether anything is statistically significant, what the explanatory power of their analyses are).  We know nothing about the magnitude of the effects of their explanatory variables.  In Casselman's work, we don't know whether he estimated his regression in a way that let him determine whether (or how well) his results work in an out-of-sample period.  In Paine's work, we don't know whether the number of "misses"--players predicted to be out of the game--is larger than our expectations (subjective or statistical), nor does he tell us anything about players projected to survive, but who didn't.

In short, we're supposed to take the very partial, incomplete results presented and trust them.  In my world (academic research), this approach would never be acceptable.  Showing one's work is the essence of the matter in what we do, and it's the essence of why we expect people to accept out conclusions.

If FiveThirtyEight continues to present analysis in this way, I think I'll stop looking, and, when I do, I'll be thinking of it as "analysis" (scare quotes very much intended.

Friday, March 21, 2014

Job Quits and Wage Increases

A post today at FiveThirtyEight ("Want a Raise?  Quit Your Job") proposes, using anectdotal data, that one plausible way for workers to get pay increases is to quit their jobs.  I posted this comment there:

There's a question of causation here. Is wage growth stronger *because* people quit more often, or do people quit more often *when* and *because* wage growth is strong? Even the highlight example here (trucking) is suspect in that respect. Why does Covenant offer those retention benefits? Is it (as I suspect) *because* wage growth in trucking is fairly strong, so it makes sense to try to retain the drivers that have?

(Incidentally, there is older information on quit rates, it's just not showing up on-line. You can find it in print editions of old BLS publications like the Employment and Training Report of the President, which was discontinued some time in the 1980s. It was discontinued, as I recall, because the Reagan administration didn't want to do it.)
And got this response:

Yes, you're right about quits--I meant the JOLTS series specifically. Older data aren't directly comparable, though you can make some adjustments.
Re: causation, it doubtless runs both ways to some degree (people more likely to change jobs when good prospects are available). But even after controlling for occupation/industry, there's evidence is that changing jobs does lead to wage growth.
His argument would be more convincing if he actually, you know, presented the evidence.  And the evidence would be that more rapid wage increases lead changes in quit rates.  His response suggests that we need to disaggregate to the industry and occupation levels, and (frankly) I don't have the time or the energy to do that.  But just looking at aggregate quit rates and aggregate wage increases, and using data from the BLS web site, I get this:

(Click to enlarge.)

The rate of wage increases has been increasing since mid-2012, but the quit rate has barely budged.  Over the 2010 to 2014 period, the correlation between the rate at which wages are increasing and the quit rate is 0.35, which is statistically significant, but not large.  And given the likelihood that at least some of the relationship comes from the causal effect of faster wage increases on quits, the ability of rising quit rates to explain  faster wage increases seems to me to be likely to be really small.  Unless, that is, I see some evidence to the contrary, rather than assertions of a relationship.

Thursday, March 13, 2014

Annual Real GDP, 1929-2013

(Click to enlarge)

(Click to enlarge)

Sunday, March 09, 2014

What would make me a whole lot happier about the monthly jobs report

The (employer-based) employment change was +175,000 for February 2014, compared with January 2014.  More importantly, the 12-month period (February 2013 to February 2014) shows a gain of 2.16 million jobs.  Employment has been growing, on a 12-month basis, at about 1.5% to 1.8%, since September 2011.  This is a good, but not great, rate of employment growth.  It's "not great" especially coming out of a recession.  Coming out of the recession of the early 1980s, 12-month employment growth rates were above 2.5% beginning in September 1983 through November 1985.  And, from late 1983 to late 1984, annual growth rates in employment were above 4%.  Coming out of the "Great Recession," there are only two 12-month periods with an employment growth rate above 2%, and nothing above 2.17%.  (Even the tepid recovery from the 2001 recession saw six months with annual growth rates above 2%. And the average 12-month growth rate in employment since 1948 has been 1.72%.)


(Click to enlarge.)

So it's hard to work up much enthusiasm about the recovery based on growth in establishment employment.  However, the unemployment rate has dropped quite significantly, and in February was 6.7% (up slightly from 6.5% in January).  But what I'd really like to see, in addition to somewhat more robust employment growth, is some positive news about labor force participation.


(Click to enlarge.)

Despite a fractional increase in labor force participation in  January and February (up to 63.0%, from 62.8% in December), labor force participation has consistently declined during the recovery.  In fact, the 12-month percentage change in labor force participation has been negative since August 2008.  That's before the beginning of the recovery (June 200), according to the NBER Business Cycle Datng Committee.  That is, in more than four and a half years of recovery, the labor force participation has not once grown on a year-over-year basis.  This is, since monthly data on labor force participation have been available (1948), the first recovery ever during which labor force participation has not increased.

So what would make me a whole lot happier?  Not just growth in the labor force, but an increase in the labor force participation rate, which is, as of February, three percentage points below its level at the beginning of the recession.

(All data from the Bureau of Labor Statistics.)