Wednesday, January 29, 2014

Update on the job market and reallocation (BED)

Or, job market second moments.

Today the BLS released new BED numbers (I've covered these before), which now go through the second quarter of 2013. Said the BLS:

From March 2013 to June 2013 gross job gains from opening and expanding private sector establishments were 7.1 million, the U.S. Bureau of Labor Statistics reported today. Over this period, gross job losses from closing and contracting private sector establishments were 6.5 million, an increase of 191,000 from the previous quarter.

This release also included some delayed annual revisions. Here are some very important footnotes.*

Readers who aren't familiar with gross job numbers should pay attention to magnitudes. More than 7 million jobs were created during a single quarter; and 6.5 million jobs were destroyed. For comparison, the Great Recession was characterized by the net destruction of about 8.5 million jobs, total, between late 2007 and late 2009 (private sector). As you can see, job flows from reallocation absolutely dwarf the net job numbers that make headlines. The US economy is a reallocation machine. However, the rate of reallocation has been slowing, a fact I've mentioned about a million times so I'll spare you the links.

I like to separate these numbers into extensive margin (opening or closing business establishments) and intensive margin (expanding or contracting business establishments). First, extensive: Figure 1 reports the flows of employment associated with opening and closing establishments, and Figure 2 reports actual numbers of establishments that opened or closed (click for larger images).

Figure 1

Figure 2

Opening establishments seem to be doing a decent job of keeping employment demand ahead of jobs destroyed by establishments closures, maintaining that gap that opened up in late 2011. The new establishment count for 2013q2 looks unhelpful.

Now look at intensive margins; Figure 3 reports employment flows from expanding and contracting establishments, and Figure 4 reports establishment counts for these categories (click for larger images).

Figure 3

Figure 4

I would say that these numbers look promising, with all the usual caveats. But I would feel better if the entry/exit numbers looked as good as the expansion/contraction numbers.

I like to keep an eye on these numbers for several reasons. Any sort of policy discussion can benefit from knowing where the jobs are being created. The entry vs. expansion distinction is useful if we think there are costs to entry. At the firm level, a decision about opening or closing establishments is very different from a decision about expanding or contracting existing establishments.

Another reason to be aware of these numbers is to have an idea about the economy's ability to both create jobs and reallocate them. We lost over 8 million jobs to the recession, but we create nearly that many every single quarter. The problem is that while some firms are creating jobs, others are destroying them--even in (especially in?) a healthy economy. This reallocation is productivity enhancing if it arises from the closure or contraction of unproductive establishments and the opening or expansion of productive ones, but it's not obvious that all of it meets those criteria. It's also not obvious whether there are costly barriers to reallocation, the absence of which would result in more of it.

In political dialogue it's common to hear phrases like "nobody wants to hire anybody" or "the job creators are all sitting on the sidelines." These numbers reveal that there is actually a lot going on in job markets, most of it hidden by net job numbers.


*The BLS effectively expanded the sample definition in the first quarter of 2013, and it does not appear that they have done anything to fix the time series. This is very unfortunate as it limits the usefulness of looking at time series in ways that are difficult to fully grasp. The 2013q1 observation was the most obviously affected, as it reported all establishments that were added to the sample as establishment openings. For openings data, I have replaced the 2013q1 observation with the average of 2012q4 and 2013q2. I haven't dug into the data enough to know whether users can manually correct for this over the longer run. See BLS discussion here, on the bottom of the page ("Administrative Change Affecting..."). Please, BLS, do something about these time series.

It is also important to note that these numbers are seasonally adjusted, and any guess at net numbers based on the difference between two seasonally adjusted series is very, very rough. Non-SA numbers are available on the BLS website.

These numbers track business establishments, which are different from firms. Costco is a firm; your local Costco store is an establishment. Most firms consist of only one establishment. The BED is not ideal for tracking firms, as it has limited ability to correctly link establishments to the firm level.

Saturday, January 25, 2014

We need to think more about firm dynamics

I watched the Netflix Mitt documentary. This is not a political blog, and I don't like Romney any better than any other politician, but I was intrigued by something he said in a conversation with his family (starting around 41:45):

They don't know how hard it is. They have not been in a setting where you're trying to make it, where you've got a little business and you're trying to make it. They don't know how hard that is. . . . They assume it's always there, and business is always there. They don't know that businesses fail, that people go out of business--that they lose their life savings, they lose their job. They start over again! But they don't know how hard it is for a business to succeed, and they keep piling on more and more thinking "that's ok, these guys will all do fine." At some point, they don't do fine.

The line that caught me was "they assume it's always there." This is the representative firm model of the world, maybe even the exogenous income endowment model of the world, and yes it seems to me that a lot of our dialogue in both politics and economics operates from this assumption. But firm dynamics matter; and the size and age distribution matters, particularly for employment flows.

We don't think about firm dynamics enough.

Wednesday, January 22, 2014

Questions about the minimum wage

Following some musings from Tyler Cowen, Arindrajit Dube summarizes his empirical research on minimum wages. This literature is not my area, and I therefore have no opinion on the relative merits of competing empirical studies.

In my view the more relevant question is whether data on past minimum wage hikes--typically at the state level--tell us a lot about future hikes in the federal minimum wage. Even if we are persuaded that Dube's reading of the evidence is appropriate, the minimum wage is one of those issues where all future policies involve action outside of the data samples used for empirical studies. Note that this is not the case for all policy questions; I am not suggesting that all empirical research lacks policy relevance, and if your immediate reaction to my skepticism is to ask me if I find all econometric research to be useless then you're ignoring key characteristics of minimum wage policy. Different policies differ in the degree to which relevant empirical studies are externally valid.

To get to the point: Do we expect the effect of a minimum wage hike to be independent of (a) the prior level of the minimum wage and (b) the geographic region (and industrial composition) to which it applies? Is raising the wage from $6 to $7 in one state likely to have the same effect as raising the wage from $7 to $8 nationwide? It would be nice to see a model in which the employment and poverty effects of the minimum wage are linear everywhere; it seems that people relying on empirical studies to justify changes to the federal wage are assuming roughly that. But this is not the case, e.g., in the monopsonist employer model, where the effect on labor demand depends on the level of the wage. If the effects depend on levels (and other things, like industrial composition), then empirical studies are insufficient for telling us about the consequences of policies that have not yet occurred. Don't think of changes in the minimum wage only in terms of the difference between the new wage and the old; think about initial conditions as well. Theory matters here; Dube mentions the "credibility revolution," but I think credibility demands that we know why his results are the best lens through which to view potential policies in the future.

In short, the old cliche is useful here: If you support raising the minimum wage, do you support raising it to $50/hour? If not, then you admit that the effects are nonlinear. Now you need to provide evidence in favor of your preferred wage level that doesn't rely entirely on marginal effects from past data. There may be good reasons for the fact that economists have a hard time agreeing about this issue.

Friday, January 10, 2014

Small businesses and jobs

Here's a (partial) headline from earlier this week: "Small businesses keep fueling the economy." The notion that small businesses account for most job growth is politically ubiquitous; while it is technically true, it can be pretty misleading since the effect is driven by young firms. This was shown definitively in this paper (ungated here), but it's not catching on as much as it should.

Figure 1 breaks firms into four categories: large/young, large/old, small/young, and small/old (click for larger image).

Figure 1

As a group, small businesses that have been operating for 5 years or more typically destroy jobs on net. They are outpaced in job creation by both kinds of large firms (young and old) and by young small firms. Government programs that throw money at old, small businesses based on the notion that "small businesses" are big job creators strike me as a little misguided.

Figure 2 is similar, but it draws the line for "young" at startups (click for larger image).

Figure 2

I like this chart because it shows that startups were really hammered in the Great Recession (and even before it), a fact I've mentioned before. In an arithmetic sense, we "need" a lot of startups for net job growth. Firm entry is the setting for a huge amount of experimentation; startups exhibit an "up-or-out" dynamic, with some growing rapidly and others failing in short order (I did some simple cohort charts here and here). Still others reach a small "optimal" size, like your local dentist office, and won't grow again. We need more startups both to absorb the jobs destroyed by previous failed startups and to supplement the very slow job growth of incumbent firms generally. In any case, the small business meme is a good example of how aggregation can sometimes get you into trouble.

We need a better understanding of firm dynamics, in both our public dialogue and the study of macroeconomics.

Saturday, January 4, 2014

More from AEA

John Haltiwanger presented work with Foster and Grim on "cleansing" recessions. In past recessions, job destruction (i.e., shrinking or closing establishments) drove net employment declines, and total reallocation of jobs (job creation + destruction) rose in a productivity-enhancing way. In the Great Recession, the main driver of the net employment decline was a collapse in job creation; additionally, reallocation fell and the relationship between firm productivity and exit was relatively flat compared to the past. A big part of the job creation story was the lack of new firms and the beating taken by existing young firms relative to older ones. Relatively speaking, in this recession productive firms failed and unproductive ones survived. Normally, productive firms get a growth bump during recessions, but this time they didn't. All of this indicates that significant distortions were at work. We need stories to explain this; maybe financial issues are the key.

My main takeaway from this and related work is that (a) this recession was different from previous postwar recessions, and (b) there are enough strange things going on with firm dynamics that the story seems unlikely to be amenable to the simple AD/AS framework.

"Liquidity constraints, risk premia, and the macroeconomic effects of liquidity shocks" is a paper by Ivan Jaccard. This is an RBC model with a shock to collateral (he describes it as a liquidity shock). A draft is here, and slides are here.

One possible explanation for the severity of the Great Recession is that the effects of the financial crisis were amplified by the shortage of pledgeable assets created by the initial subprime shock.

I found the paper interesting because it highlighted a couple of big concepts in modern macro:

In the model economy that matches the equity premium, our main finding is that a small negative liquidity shock can generate a deep recession and a stock market crash. In the version of the model that is unable to generate plausible risk premia, in contrast, the effects of liquidity shocks are considerably smaller.

That pesky equity premium! But the model can generate large, persistent output effects of a small shock--without nominal rigidities.

For some time I've been keeping an eye on this paper by Guerrieri and Lorenzoni. Here we get a reasonably comprehensive model of credit crunch, including both heterogeneity and the ZLB. A common characteristic of representative agent ZLB models is some ploy to get the economy to the ZLB, such as a shock to discount rates. These don't seem implausible to me, but in models like Guerrieri & Lorenzoni you get the solid decline in interest rates that arises naturally in models with precautionary savings and financial shocks. The model gets a severe contraction following a financial shock. It's amplified by nominal rigidities and the presence of durable goods (houses are durable...).

Simon Gilchrist and coauthors have a paper on uncertainty shocks (paper, slides). This is a garden variety mean-preserving spread in the shock distribution; under imperfect financial markets, credit spreads rise. The model incorporates both investment adjustment costs and financial frictions. A credit spread arises from modeled agency problems, and this is the key channel through which uncertainty hits the economy.

Go look at their empirical measure of uncertainty; the discussant didn't buy it.

Here I'm going to plug my work with coauthors on endogenous uncertainty, where uncertainty is of the TFP/sales volatility variety. In our model, firms choose how many markets to operate (where "market" can mean product, geography, etc.).  More markets means diversification means lower volatility, and a firm's market exposure varies over the business cycle, so countercyclical volatility arises naturally. We need to be more careful about treating the second moment as exogenous.

I saw this paper presented (housing collateral and entrepreneurship), which I've already reviewed here. The discussant was David Sraer, and he noted that we have yet to determine the relative quantitative significance of the two proposed channels through which housing hit the economy: the Mian & Sufi consumption channel and the alternative collateral channel. This is one of the things I'm trying to do with my dissertation work.

Friday, January 3, 2014

Comments from the AEA meetings

Insert cliche about economists choosing January in the Northeast for weekend getaways (it's cheap!)

First, my stuff: My coauthor presented our paper on the secular decline in economic dynamism. A draft is here, but the presentation included new facts that are not in that draft. First: In the early 2000s, many researchers noted that various measures of volatility were on the rise among public firms. Davis, et al. showed that the opposite was true among private firms. Today, we showed that the trend among public firms reversed in the early 2000s, and now volatility is declining among all classes of firms.

One possible explanation for the decline in volatility of businesses is that perhaps shocks have become less volatile or dispersed. Our second new fact: Using TFP estimates from the manufacturing sector, we show that this is not the case. We also provide evidence that firm-level employment has become less sensitive to productivity shocks over time. You can imagine some of the potential explanations for this, but we think it's a step forward for this declining dynamism puzzle. There is an increasingly large wedge between shocks and employment outcomes at firms and establishments. I don't have access to the relevant charts now but will probably do a more thorough post about this later.

I attended a session called "What's natural?" Barsky, Melosi, and Justiniano presented a paper that estimates the natural rate of interest over time (using a Smets and Wouters model) and proposes a monetary rule that, roughly speaking, seeks to align the Federal Funds rate with the natural rate. In the model, the estimated natural rate is shown to be extremely volatile (and often below zero) and very procyclical. They call their proposed rule a "Generalized Wicksellian" rule. Michael Woodford was the discussant and was skeptical.

Mark Watson presented a paper about the natural rate of unemployment. He basically estimated a Phillips Curve. His estimate suggested that the NAIRU is around 6 percent currently. Robert Hall was his discussant; Hall made it clear that he doesn't believe in these natural rates. He also has concerns about Calvo pricing, noting that he has gotten in trouble for criticizing it because so many people "worship at the temple of Calvo." He ended with, "We need to get out of the Calvo straitjacket."

I admit that I am broadly sympathetic to this sentiment. Calvo pricing can be very useful, but the New Keynesian project generally often feels like curve fitting to me. We're really not getting at the constraints and decisions that drive nominal rigidity. We're using a lot of band-aids. In my Twitter conversations with Noah Smith, he eventually ended up with the position that "Saltwater" economics is basically about Calvo pricing, which I see as a pretty damning indictment of "Saltwater" economics as a source of policy advice. (Luckily I don't believe that "Saltwater" economics is even a thing, so it's all good).

John Fernald and Charles Jones had a very interesting paper about growth. They suggest that the ~2% growth we've had for the last 150 years was above the steady state--it was the result of transitional dynamics. It was a very Tyler Cowen-esque argument--low-hanging fruit, etc. Susanto Basu responded to it, noting that technology and abundance have given us increasing measurement difficulties, so growth in GDP may not be a great measure of welfare progress over time anyhow.

At a different session: Antoinette Schoar delivered two very interesting papers, one about family-owned firms and management and another about supervisor training at garment factories in the developing world. Go to her website.

Erik Brynjolfsson presented a paper he has with some Stanford economists and some Census Bureau economists (old draft here) describing an awesome survey of managers at 30,000 establishments. They find that "structured" management styles are associated with higher productivity and profits, and that this kind of management occurs, on average, at firms that (a) are large, (b) are located in the South and Midwest, (c) have educated employees, and (d) use IT. This survey is going to lead to a lot of interesting research.

I went to the Bernanke lecture, which was costly since there were sessions at the same time with papers more relevant to my own research, but I didn't want to miss the chance. He gave a speech discussing his tenure at the Fed. Ken Rogoff made some comments, as did Anil Kashyap. Kashyap was very interesting: he noted that leverage is not the only source of financial instability. Other mechanisms can cause fire sales, so we should not assume that watching leverage is enough. You can read his comments here. Kashyap also provided this video:



At the end, Bernanke got a standing ovation for his service. In my view he has been an excellent Fed chair.

I also wasn't watching where I was going coming out of a session once and almost ran over Kenneth Arrow. That would have been an inauspicious start to my career.