Observations on search theory in modern macro and in blogosphere macro

Let’s say you are 22, full of energy, don’t feel you need to marry soon, and have lots of cash in your bank account.  Many people in this position feel they can “date around” a lot, without fear of the repercussions.  They can enter into short-term relationships without much agonizing in advance, and simply break up if it doesn’t work out.

Alternatively, imagine you are 39, run down and vulnerable, wanting kids soon, and in a precarious financial situation.  You probably won’t date casually the same way.  You will treat every romantic relationship as if it is a significant investment.  You will be more careful, because the cost of mistakes is higher, and the cost of serial “running around” is higher too.  Search is tougher and you will apply higher standards to search.

In good times employers are like the 22-year-old and they will take chances with many different employees.  In 2009-2013, they often have seemed more like the 39-year-old.  They are waiting and watching, rather than trying out lots of dates.

Of course this analogy points to just one possible factor, it is hardly a comprehensive account of current unemployment, even if you ignore any possible problems in the story.

Note that the terms “involuntary unemployment” and “voluntary employment” do not make sense here, and usually it is a mistake to insist on one or the other.  There are jobs and for that matter dates in North Dakota, so how high does the cost of moving have to be to distinguish one category from the other?  Is theory going to supply an answer here?  No.  When you see arguments for either the “voluntary” or “involuntary” nature of unemployment, that is a good sign someone is trying to mix moral issues into positive issues.  It is also a move away from the concept of marginal analysis.

It is better to say that the quantity and quality of employer search is suboptimal and the outcomes are suboptimal too.  In this particular case, employers become choosier in their search, and raise their discount rates, without taking into account the social costs those decisions impose on the pool of available labor.

You also will note this explanation, and many others like it, differs from the traditional invocation of “sticky wages.” Sticky wages do apply to a large number of already employed workers (pdf), but the concept does not readily transfer to workers who are looking for a job.  There is a great mass of evidence for sticky nominal wages for workers who already hold jobs.  There is a good but not great deal of evidence for sticky and indeed possibly irrational reservation wages for the current unemployed, but this phenomenon is conceptually indistinguishable from what some people (not me) like to call “voluntary unemployment.”  If you insist on stickiness for the unemployed, you will quickly end up somewhere you probably do not wish to go.

Taking established labor market results for the employed and automatically transferring those same concepts to the unemployed, without critical scrutiny, is one of the most common mistakes in blogospheric economics.  You don’t see that mistake committed very much in standard academic macroeconomics and that is one of the advantages of modeling.

I sometimes hear it suggested that the sticky wage incumbent workers have “eaten up” all of the nominal gdp and there are only breadcrumbs left for the unemployed.  Yet most companies seem to be sitting on enough cash and enough profits to make additional hires if they want to; it cannot be asserted that the sticky wage incumbents are “soaking up” most corporate liquidity.  Plus in a full model velocity and credit will be endogenous so a cash constraint would end up doing all of the work here, not sticky wages.

I once heard Bryan Caplan mention (I am not sure how seriously) that insiders will resent outsiders hired at lower wages and thus the outsiders cannot get hired.  With the decline of unions if anything I observe the opposite — insider resentment is directed at the higher-wage hires — but in any case there is simply no mass of evidence behind this claim, not enough for it to serve as a major explanation of one of our largest economic and social problems.  One is led to this kind of move if one has to justify the claim that unemployment is “involuntary,” whereas a better question is why the quality of job market match is going down.

There is the macroeconomics of the academy and the macroeconomics of the blogosphere.  The latter does not devote nearly enough attention to search theory, combined with other market imperfections of course.

If you’d like to read one recent macro piece on search, there is Marcus Hagedorn and Iourii Manovskii, “Job Selection and Wages over the Business Cycle,” published in the latest AER but with an ungated version here.  It is not some kind of explanatory holy grail but it does illustrate the kind of questions contemporary macro is asking.


imagine you are 39, run down and vulnerable, wanting kids soon, and in a precarious financial situation. You probably won’t date casually the same way.


Who would want to go on a date with you, whether you are male or female?

Ah my friend, but you've never been to a Third World country. Precarious financial situation by US standards is a jackpot in the 3rd World. A whole family can live on a US social security check.

So the analogy is perfect--not only are wages sticky, but prices are too high in the USA and need to be adjusted downwards, to Philippine and African levels.

TC forgot an important part of the description:

imagine you are 39, run down and vulnerable, wanting kids soon, and in a precarious but cash rich financial situation.

Exactly. Thee 22 yr old can take a date out with a bunch a friends and see if they are compatible. The 39 yr old is expected to provide a formal evening out - more expensive and time consuming.

| There is the macroeconomics of the academy and the macroeconomics of the blogosphere. The latter does not devote nearly enough attention to search theory, combined with other market imperfections of course.

Great point. More than half of the macro seminars I attend involve search. On the other hand, I have never heard the word 'NGDP' in a seminar.

Are "bad times" periods of low NGDP? Or periods after a housing bubble bursts and NGDP is growing, where housing construction employment is a very small share of the total workforce?

Are bad times the absence of rainbows or periods after a rainstorm when rainbows appear and rain isn't in the five day forecast?

How should the Fed encourage more rainbows? By targeting them as policy, of course!

It's amazing just how controversial Friedman's aphorism that "inflation is always and everywhere a monetary phenomenon" has become. Do you really think there is nothing a central bank can do to increase spending?

Modern central banks can certainly increase aggregate spending, but it has no magical power to increase spending on one good in particular (in this case employment). What the NGDP worshipers ignore is that there's no reason to expect that the marginal dollar hot off the printing press will be spent in the same manner as the average dollar in circulation.

In other words if X% of spending is on employment, it's not so simple to say that if we increase money supply by epsilon then ipso facto employment will also rise by epsilon. Monetary expansion, at least in the medium term, is not neutral against all types of economic activity.

It doesn't take a genius to see where all those marginal dollars are flowing. There's a tremendous amount of slack and elasticity in the financial markets. The S&P went from 670 to 1600 in 4 short years. Labor force participation rates meanwhile barely budged from historic lows. Look at any other similar magnitude equity bull market in history and try to find one with such a correspondingly bad and stagnant labor market. Credit spreads, implied volatilities, TED spreads, commodity prices, implied correlations and carry-trade returns from 2009 to 2013 all tell the same story. Massive recovery in financial markets, agonizingly slow recovery in labor markets. Real estate markets somewhere in between.

The Fed can print all the money it wants. But very few of those marginal dollars are being spent on new jobs, particularly for the unskilled long-term unemployed. The vast majority are going to corporate treasuries, share buybacks, emerging market central banks and sovereign wealth funds, leveraged buyouts, collateral transformation, debt and mortgage re-financing, commodities hoarding, Chinese real estate, structured credit origination and probably most importantly bank de-leveraging (a significant portion triggered by Basel III). In the MV = PQ formula, all of these financial flows soak up the velocity neutralizing the effect of higher money supply.

Now of course on a long enough horizon the inflationary effect of money neutralizes, and eventually those marginal dollars will flow into labor markets. But we're easily talking about a horizon of a decade or more. Non-neutrality of money significantly blunts the power of central banks. For example if in the short-term only 4% of marginal dollars are spent on unemployed workers, then $1 of short-term stimulus costs you $25 of long-term inflation. Inflation also causes misery, as well as increasing the risk of financial panics. At some level of non-neutrality the tradeoff makes additional spending not worth it, even if there is persistent slack in labor markets.

Financial markets are quite messed up, but that's a different answer than "printing money wouldn't cause any inflation". I would suggest that money is fungible and so it's doesn't make much sense to talk about marginal dollars except in the very short run. Having said that, there are many pernicious ways the central bank has subsidized the banking sector, such as the interest on reserves policy. I'm not sure why capital returns are so high relative to labor these days, though it is an interesting question.

Central banks can and do reverse expansionary policy, so over a period of a decade any expansion now could be reversed. You might get some higher inflation in the meantime, but the costs of inflation are far lower than the costs of unemployment, IMO. Thanks for the thoughtful response.

Not the experience at all in financial services. They are repricing human capital down by laying off older workers and hiring young, enthusiastic kids at rock bottom prices for the firms.

I was lucky to survive the rounds of layoffs because the revenue stream I brought in was beyond question greater than my salary. I always made conservative choices with clients and their money, and we didn't lose them when things turned bad. That's because I gave them proper expectations.

I have no idea about other fields, but I am skeptical. All the economic data I see is that we are adding low paying jobs and losing it barely gaining high paying jobs. Skill mismatch might explain a lot, but repricing assets is another big aspect.

I was lucky to survive the rounds of layoffs because the revenue stream I brought in was beyond question greater than my salary.

I would expect efficient businesses to layoff older workers that don't bring in enough revenue to cover their salary and associated overhead. Granted, there are some short term vs long term aspects, but I would expect the company to keep/reward employees who behaved as you've described your behavior.

But why would a financial firm be justified in keeping advisers who didn't make conservative choices with clients and their money and thus keep them when things turned bad? That's both bad for the company and for the clients.

If you take the analogy seriously, then you can link this particular hobbyhorse with the great stagnation. Because it implies that all the companies in the marketplace are closer to the end of their own lifecycles. That is after all the essential reason why the 39 year old is "pickier".

Or you could say that the economic situation for these companies is more precarious and they can't afford to take chances with anything, eg marginal workers. But then a main problem with this ZMP idea is that the strong version goes directly against many peoples' direct experience of how hiring and firing decisions are actually made, and the weaker versions get close to the idea that well of course we have higher unemployment, the economy has been doing badly.

In fact the economy doing badly is a perfectly sufficient explanation for higher unemployment. I realize that is boring for some people.

Also, I thought the data indicated that middle tier jobs were disaapearing, what was being created was all lower tier and some higher tier.

We must always remember that there are two searches involved in employment, and they are NOT simple reciprocals of each other. Both searches are much more heavily constrained than is often imagined.

Employers are seeking to maintain and increase organizational income and wealth and at least in theory (by law) shareholder wealth. Employing lots of people is NOT a primary a goal - it's a means to an end. Labor is NOT a commodity to "buy up" when it's cheap - it's *always* expensive in total cost terms, and therefore ONLY sought when really required.

Job seekers wish to reliably provide personal income and wealth.

Both sides of the search are constrained ("pinned") by all sorts of things - moving to North Dakota is impractical if your spouse is fully employed in Seattle. Hiring all of the smart motiviated people looking for jobs in CA, who are pinned to CA in one way or another, is of no help running oil extraction in North Dakota.

So often, the better analogy would be "yes, there's a wonderful woman who would be happy to enter into a romantic relationship with my rather over 39yo self, but she lives in Italy and I live in Seattle and so we'll never meet."

So H1-B==mail order brides?

I also think the analogy is strained.

Businesses will always try to hire the most appropriate staff when they are believed to be needed (ie, when the business anticipates business which these new hires will serve). I do my share of hiring, and I do not look for the lowest cost hire. I take the wage level as given (labor at any given level as a commodity), and then hire the best person who is the best fit at that wage level. Many, many factors affect the notion of "fit". For example, I hired one of my current analysts because of her spunk. She's comes from some small village in China, working class parents (by Chinese standards); got a scholarship to Hong Kong; got a scholarship to Columbia's SIPA. Anyone who can figure out how to do that deserves a job.

Also, as a business, you can afford to make more mistakes if you financial situation is strong, that is, if your profits are high or your balance sheet robust.

Aren't search models the academic relative of DSGE models? Yes, they both add to the discussion but they are also like intellectual weeds choking out other approaches. I kind of like the eclectic macro on the blogs.

Now to your post, one thing I do not like about search models (and your analogy) is the lack of interconnections and adjustment costs. Maybe insider wage stickiness matters in a big way for the unemployed worker, albeit in an indirect way. Suppose unemployed guy lost his factory job but could get a job stocking shelves, but oh, look his kids' childcare costs a good bit more than his new wage offer. Stickiness of employed daycare worker wages can affect the gap between unemployed workers reservation wage and his offered wage, albeit indirectly. Now there are many margins to adjust on here, but a higher stream of wage offers would go a long way. Unitary, rational, perfect foresight models of economic behavior are fun to work with and ponder, but they have limits too.

Finally, why not give the 39-year-old employer in the analogy 'a little blue pill' in the form of wage insurance or pretty up the unemployed pool with retraining programs and public works. It's one thing to explain the current equilibrium, it's another to tell us how to get back to a better one.

'Of course this analogy points to just one possible factor, it is hardly a comprehensive account of current unemployment, even if you ignore any possible problems in the story.'

Or in shorter form - 'Why care about the analogy, even if you ignore any possible problems in it?'

Tecently, here sure feels like an all you can eat buffet at the Spaghetti Mill - with the patrons being asked to contribute to the process of seeing what sticks best.

Yes, recently, as in the last 10 years, it sure has seemed like Tyler is posting ideas that he has and seeing what people think of them.

Actually, I meant the last couple of weeks. The ebb and flow of posts is interesting to observe - it seems as if the great stagnation has become a bit stagnant, while the zero marginal worker is still just a bit too close to the edge of revealing how a certain class within our society views their fellow citizens.

The narratives just aren't flowing like they used to. And the comments are no longer as adoring, as any number of people with actual experience of what is posted point out how a post is flawed and/or incomplete and/or simplisitic.

The human mind: a (sometimes faulty) pattern-recognition machine.

What does this analogy add to our understanding? It's not as if we have a complete theory of search in dating.

This is like trying to explain P vs NP by saying it's like whether or not something is art.

Companies don't hire (or rather shoudn't) hire just because they have extra cash. In theory, companies use cash to hire people because they think that the new hire will produce enough profit. The reason there won't be enough profit is because companies don't think they can sell enough at current prices. So the only way to sell more is to sell at a lower price. So the question is how far would the wages of new hires have to drop to lower the marginal cost of producing an additional item or service.

Is the lumpiness of production the problem here? In how many companies could you reduce your marginal cost by hiring just 1 extra person at a cheaper wage?

Your story, at least as you've told it, would be mooted if we *had* say 6% inflation when the crisis hit and sticky wages for the currently employed were dominating firing decisions, right? It may be too late for the present crisis, but since that seems like a much easier thing to change with policy than search efficiency, shouldn't that be the primary goal going forward into the next and next and next crisis?

The analogy is more distracting than elucidating.

But I suppose the objective of the evocative analogy is to attract comments, not to provide clarity.

There are at least two ways that Tyler's opening analogy is off. He undercuts one argument later in is his post by noting that current firms are sitting on lots of money. There is an ongoing debate why they are not hiring, from Keynesian arguments about insufficient demand through Meltzer and talk radio saying it is fear of Obamacare to others going on about fear of future regulations or even tax increases, although last time I checked Obama was calling for a cut in the corporate profits tax rate. These companies are not terrified of hiring a "loser" they will be stuck with.

The other is that in dating with an eye to marriage and procreation, there is only one to be picked. In hiring the equivalent of Tyler's well-off 22 year old can hire lots of people, and still lay them off if need be. More like marrying a harem. The analogy reallyis poor all the way around.

BTW, one case that may be on Tyler's mind of insiders unhappy about new hires coming in at hire salaries is in academia.

Of course that should read "higher wages," not "hire wages," although the latter is not all that inaccurate in that indeed the higher wages of the new hires in much of academia are the wages at which the new people could be hired, :-),.

Why do firms now feel that they are in a precarious financial situation? Or is it more important that they want to have kids soon?

These courses provide high quality study materials with
high definition videos as practical training. Here is another website with Incredible Hulk coloring pages:.
That crossroad provides a choice, a choice that can be made out of fear or

Comments for this post are closed