All scholars strive to make important contributions to their discipline. ThomasSargent irrevocably transformed his.

In the early 1970s, inspired by the groundbreaking work of Robert Lucas,Sargent and colleagues at the University of Minnesota rebuilt macroeconomictheory from its basic assumptions and micro-level foundations to its broadestpredictions and policy prescriptions.

This “rational expectations revolution,” as it was later termed, fundamen-tally changed the theory and practice of macroeconomics. Prior models hadassumed that people respond passively to changes in fiscal and monetary policy;in rational expectations models, people behave strategically, not robotically.The new theory recognized that people look to the future, anticipate howgovernments and markets will act, and then behave accordingly in ways theybelieve will improve their lives.

Therefore, the theory showed, policymakers can’t manipulate theeconomy by systematically “tricking” people with policy surprises. Central banks,for example, can’t permanently lower unemployment by easing monetary policy,as Sargent demonstrated with Neil Wallace, because people will (rationally)anticipate higher future inflation and will (strategically) insist on higher wagesfor their labor and higher interest rates for their capital.

This perspective of a dynamic, random macroeconomy demanded deeperanalysis and more sophisticated mathematics. Sargent pioneered the developmentand application of new techniques, creating precise econometric methods to testand refine rational expectations theory.

But by no means has Sargent limited himself to rational expectations.Among his dozen books and profusion of research articles are key contributionsto learning theory (the study of the foundations and limits of rationality) andto economic history, including influential work on monetary standards andinternational episodes of inflation.

Interviewed here by now-retired Research director Art Rolnick, acolleague since the 1970s at the University of Minnesota and Minneapolis Fed,Sargent explores issues ranging from polar models of banking regulation andcrisis to causes of persistently high unemployment to a compelling defense ofmodern macro. Underlying the entire conversation is the “vocabulary of rationalexpectations,” observes Sargent. “In our dynamic and uncertain world, our beliefsabout what other people and institutions will do play big roles in shaping ourbehavior.”

SEPTEMBER 2010

The Region

27

Thomas Sargent

Photography by Peter Tenzer

MODERN MACROECONOMICSUNDER ATTACK

Rolnick: You have devoted your profes-sional life to helping construct and teachmodern macroeconomics. After thefinancial crisis that started in 2007,modern macro has been widely attackedas deficient and wrongheaded.

Sargent: Oh. By whom?

Rolnick: For example, by Paul Krugmanin the New York Times and Lord RobertSkidelsky in the Economist and else-where. You were a visiting professor atPrinceton in the spring of 2009. Alongwith Alan Blinder, Nobuhiro Kiyotakiand Chris Sims, you must have dis-cussed these criticisms with Krugman atthe Princeton macro seminar.

Sargent: Yes, I was at Princeton then andattended the macro seminar every week.Nobu, Chris, Alan and others alsoattended. There were interesting discus-sions of many aspects of the financialcrisis. But the sense was surely not thatmodern macro needed to be recon-structed. On the contrary, seminar par-ticipants were in the business of usingthe tools of modern macro, especiallyrational expectations theorizing, to shedlight on the financial crisis.

Rolnick: What was Paul Krugman’s opin-ion about those Princeton macro semi-nar presentations that advocated mod-ern macro?

Sargent: He did not attend the macroseminar at Princeton when I was there.

Rolnick: Oh.

Sargent: I know that I’m the one who issupposed to be answering questions, butperhaps you can tell me what popularcriticisms of modern macro you have inmind.

Rolnick: OK, here goes. Examples ofsuch criticisms are that modern macro-economics makes too much use of

sophisticated mathematics to modelpeople and markets; that it incorrectlyrelies on the assumption that asset mar-kets are efficient in the sense that assetprices aggregate information of all indi-viduals; that the faith in good outcomesalways emerging from competitive mar-kets is misplaced; that the assumption of“rational expectations” is wrongheadedbecause it attributes too much knowl-edge and forecasting ability to people;that the modern macro mainstay “realbusiness cycle model” is deficientbecause it ignores so many frictions andimperfections and is useless as a guideto policy for dealing with financialcrises; that modern macroeconomicshas either assumed away or short-changed the analysis of unemployment;that the recent financial crisis tookmodern macro by surprise; and thatmacroeconomics should be based lesson formal decision theory and more on

the findings of “behavioral economics.”Shouldn’t these be taken seriously?

Sargent: Sorry, Art, but aside from thefoolish and intellectually lazy remarkabout mathematics, all of the criticismsthat you have listed reflect either woefulignorance or intentional disregard forwhat much of modern macroeconomicsis about and what it has accomplished.That said, it is true that modern macro-economics uses mathematics and statis-tics to understand behavior in situationswhere there is uncertainty about howthe future will unfold from the past. Buta rule of thumb is that the more dynam-ic, uncertain and ambiguous is the eco-nomic environment that you seek tomodel, the more you are going to haveto roll up your sleeves, and learn and usesome math. That’s life.

Rolnick: Putting aside fear and igno-rance of math, please say more about theother criticisms.

Sargent: Sure. As for the efficient mar-kets hypothesis of the 1960s, pleaseremember the enormous amount ofgood work that responded to Hansenand Singleton’s ruinous 1983 JPE[Journal of Political Economy] findingthat standard rational expectations assetpricing theories fail to fit key features ofthe U.S. data.1 Far from taking the “effi-cient markets” outcomes for granted,important parts of modern macro areabout understanding a large and inter-esting suite of asset pricing puzzles,brought to us by Hansen and Singletonand their followers—puzzles aboutempirical failures of simple versions ofefficient markets theories. Here I havein mind papers on the “equity premiumpuzzle,” the “risk-free rate puzzle,” the“Backus-Smith” puzzle, and on and on.2

These papers have put interestingnew forces on the table that can helpexplain these puzzles, including missingmarkets, enforcement and informationproblems that impede trades, difficultestimation and inference problems con-fronting agents, preference specificationswith novel attitudes toward the timing

The Region

28SEPTEMBER 2010

It is true that modern macroeconomicsuses mathematics and statistics tounderstand behavior in situations wherethere is uncertainty about how the futurewill unfold from the past. But a rule ofthumb is that the more dynamic, uncertainand ambiguous is the economic environmentthat you seek to model, the more you aregoing to have to roll up your sleeves, andlearn and use some math.

and persistence of risk, and pessimismcreated by ambiguity and fears of modelmisspecification.

Rolnick: Tom, let me interrupt. Whyshould we at central banks care aboutwhether and how those rational expec-tations asset pricing theories can berepaired to fit the data?

Sargent: Well, there are several impor-tant reasons. One is that these theoriesprovide the foundation of our ways ofmodeling the main channels throughwhich monetary policy’s interest ratedecisions affect asset prices and the realeconomy. To put it technically, the “newKeynesian IS [investment-savings] curve”is an asset pricing equation, one of aform very close to those exposed asempirically deficient by Hansen andSingleton. Efforts to repair the assetpricing theory are part and parcel of theimportant project of building an econo-metric model suitable for providingquantitative guidance to monetary andfiscal policymakers.

Another important reason for caringis that monetary policymakers haveoften been urged to arrest bubbles inasset markets. Easier said than done.Before you can do that, you need aquantitatively reliable theory of assetprices that you can use to identify andmeasure bubbles.

Rolnick: Before I interrupted, you hadbegun responding to those criticisms ofmodern macro. Please continue.

Sargent: I have two responses to yourcitation of criticisms of “rational expec-tations.” First, note that rational expec-tations continues to be a workhorseassumption for policy analysis bymacroeconomists of all political persua-sions. To take one good example, in theSpring of 2009, Joseph Stiglitz andJeffrey Sachs independently wrote op-edpieces incisively criticizing the Obamaadministration’s proposed PPIP (Public-Private Investment Program) for jump-starting private sector purchases of toxicassets.3 Both Stiglitz and Sachs execut-

ed a rational expectations calculation tocompute the rewards to prospectivebuyers. Those calculations vividlyshowed that the administration’s pro-posal represented a large transfer of tax-payer funds to owners of toxic assets.That analysis threw a floodlight ontothe PPIP that some of its authors did notwelcome.

And second, economists have beenworking hard to refine rational expecta-tions theory. For instance, macroecono-mists have done creative work thatmodifies and extends rational expecta-tions in ways that allow us to under-stand bubbles and crashes in terms ofoptimism and pessimism that emergesfrom small deviations from rationalexpectations. An influential example ofsuch work is the 1978 QJE [QuarterlyJournal of Economics] paper by Harrisonand Kreps.4 You should also look at afascinating paper that builds onHarrison and Kreps, written by JoseScheinkman and Wei Xiong in the 2003JPE.5 As I mentioned earlier, for policy-makers to know whether and how theycan moderate bubbles, we need to havewell-confirmed quantitative versions ofsuch models up and running. We don’tyet, but we are working on it.

Rolnick: And the other criticisms?

Sargent: OK. The criticism of real busi-ness cycle models and their closecousins, the so-called New Keynesianmodels, is misdirected and reflects amisunderstanding of the purpose forwhich those models were devised.6These models were designed to describeaggregate economic fluctuations duringnormal times when markets can bringborrowers and lenders together inorderly ways, not during financial crisesand market breakdowns.

By the way, participants within boththe real business cycle and newKeynesian traditions have been sternand constructive critics of their ownworks and have done valuable creativework pushing forward the ability ofthese models to match important prop-erties of aggregate fluctuations. The

authors of papers in this literature usu-ally have made it clear what the modelsare designed to do and what they arenot. Again, they are not designed to betheories of financial crises.

Rolnick: What about the most seriouscriticism—that the recent financial cri-sis caught modern macroeconomics bysurprise?

Sargent: Art, it is just wrong to say thatthis financial crisis caught modernmacroeconomists by surprise. Thatstatement does a disservice to an impor-tant body of research to which responsi-ble economists ought to be directingpublic attention. Researchers have sys-tematically organized empirical evi-dence about past financial and exchangecrises in the United States and abroad.Enlightened by those data, researchershave constructed first-rate dynamic

The Region

29 SEPTEMBER 2010

It is just wrong to say that this financialcrisis caught modern macroeconomistsby surprise. ... Researchers have systemati-cally organized empirical evidence aboutpast financial and exchange crises in theUnited States and abroad. Enlightened bythose data, researchers have constructedfirst-rate dynamic models of the causesof financial crises and government policiesthat can arrest them or ignite them.

´

models of the causes of financial crisesand government policies that can arrestthem or ignite them. The evidence andsome of the models are well summa-rized and extended, for example, inFranklin Allen and Douglas Gale’s 2007book Understanding Financial Crises.7Please note that this work was availablewell before the U.S. financial crisis thatbegan in 2007.

Rolnick: I’ll come back to that in a sec-ond, but you haven’t said anything yetabout what is to be gained in terms ofunderstanding financial crises fromimporting insights of behavioral eco-nomics into macroeconomics.

Sargent: No, I haven’t.

FINANCIAL CRISES

Rolnick: OK then. Well, what usefulthings does macroeconomics have tosay about financial crises, what causesthem, how to manage them after theystart and what can be done to preventthem?

Sargent: A lot. In addition to the formalliterature summarized in the Allen andGale book, I want to mention the exam-ple of the 2004 book by Gary Stern andRon Feldman, Too Big to Fail.8 Thatbook doesn’t have an equation in it, butit wisely uses insights gleaned from theformal literature to frame warningsabout the time bomb for a financial cri-sis set by government regulations andpromises. Indeed, one of the focuses ofGary Stern’s long tenure as president ofthe Minneapolis Fed was steadily todraw attention to financial fragilityissues and what the government doeseither to arrest crises or, unfortunatelyas an unintended consequence, to incu-bate them.

Rolnick: Thanks for the nice words aboutGary, but please elaborate further onmacro scholarship and financial crises.

Sargent: I like to think about two polarmodels of bank crises and what govern-

ment lender-of-last-resort and depositinsurance do to arrest them or promotethem. Both models had origins inpapers written at the Federal ReserveBank of Minneapolis, one authored byJohn Kareken and Neil Wallace in 1978and the other by John Bryant in 1980,then extended by Diamond and Dybvigin 1983.9 I call them polar modelsbecause in the Diamond-Dybvig andBryant model, deposit insurance is pure-ly a good thing, while in the Kareken andWallace model, it is purely bad. Thesedifferences occur because of what thetwo models include and what they omit.

The Bryant and Diamond-Dybvigmodel starts with an environment in

which banks can do things that are veryworthwhile socially; namely, they pro-vide maturity transformation and liq-uidity transformation activities thatimprove the efficiency of the economy.They enable coalitions of people, name-ly, the banks’ depositors, to make long-term investments—loans, mortgagesand the like—while at the same time thebank’s depositors hold demand deposits,bank liabilities that are short term induration, because they can withdrawthem at any time. Banks thereby facili-tate risk-sharing among people withuncertain future liquidity needs. Theseare all good things.

But there is a potential problem herebecause for the long-term investmentsto come to fruition, enough patientdepositors must leave their funds in thebank to avoid premature liquidation of abank’s long-term investments. Withoutdeposit insurance, situations can arisethat induce even patient depositors towant to withdraw their funds early,causing the banks prematurely to liqui-date the long-term investments, withadverse affects on the realized returns.

What triggers a bank run is patientdepositors’ private incentive to with-draw early when they think that otherpatient investors are also choosing towithdraw early. Technically speaking,that amounts to multiple Nash equilib-ria. There are situations in which I run(i.e., withdraw from the bank early)because I expect you to run, and whenyou also run because you expect me torun. But there are other situations inwhich we both trust that the other per-son isn’t going to run and we don’t run.Which equilibrium prevails is anyone’sguess, or something resolved only by anextraneous random device for correlat-ing behavior, a device that economistssometimes call a “sunspot.”

So without deposit insurance, theeconomy is vulnerable to bank runs.The situations where depositors don’trun lead to good outcomes, but whenthere are bank runs, outcomes are bad.The good news in the Diamond-Dybvigand Bryant model, however, is that ifyou put in government-supplied deposit

The Region

30SEPTEMBER 2010

One of the focuses of Gary Stern’s longtenure as president of the MinneapolisFed was steadily to draw attention tofinancial fragility issues and what thegovernment does either to arrest crisesor, unfortunately as an unintendedconsequence, to incubate them.

Without deposit insurance, the economyis vulnerable to bank runs. ... The goodnews in the Diamond-Dybvig and Bryantmodel, however, is that if you put ingovernment-supplied deposit insurance ...people don’t initiate bank runs becausethey trust that their deposits are safelyinsured.

insurance, that knocks out the bad equi-librium. People don’t initiate bank runsbecause they trust that their deposits aresafely insured. And a great thing is thatit ends up not costing the governmentanything to offer the deposit insurance!It’s just good all the way around.

Rolnick: Do you think that an abstractmodel like this ever influences policy-makers?

Sargent: I believe that the Bryant-Diamond-Dybvig model has been veryinfluential generally, and in particularthat it was very influential in 2008among policymakers. A perhaps over-simplified but I think largely accurateway of characterizing the vision of manypolicy authorities in 2008 was that theycorrectly noticed that a Bryant-Diamond-Dybvig bank is not just some-thing that has “B A N K” written on itsstationary and front door. It’s any insti-tution that executes liquidity transfor-mation and maturity transformation,thereby offering a kind of intertemporalrisk-sharing.

So in 2008, there were all sorts ofinstitutions that were really banks in theeconomic sense of the Bryant-Diamond-Dybvig model but that didnot have access to explicit deposit insur-ance, institutions like money marketmutual funds, shadow banks, evenhedge funds that were doing exactlythose maturity-transforming and risk-transforming activities.

When monetary policy authorities,deposit insurance authorities and otherslooked out their windows in the fall of2008, they saw Bryant-Diamond-Dybvigbank runs all over the place. And thelogic of the Bryant-Diamond-Dybvigmodel persuaded them that if theycould arrest the runs by effectively con-vincing creditors that their loans—thatis, their short-term deposits—to these“banks” were insured, that could bedone at little or no eventual cost to thetaxpayers. You could nip the run in thebud and really prevent the next GreatDepression. This is a very optimisticview of those 2008 interventions

enlightened by the Bryant and Diamond-Dybvig model.

But Diamond and Dybvig them-selves were cautious about promotingsuch optimism. In the last part of their1983 JPE paper, Diamond and Dybvigrecommend that their readers takeseriously the message of a 1978 paper(written at the Minneapolis Fed, as Imentioned earlier) by Kareken andWallace. That paper includes some-thing important that Diamond andDybvig recognize that they left out:moral hazard.

Rolnick: And the Kareken-Wallace story?

Sargent: The main idea is that when agovernment is in the business of being alender of last resort or a deposit insurer,depending on how it regulates banks, itaffects the risk that banks take and theprobability that the government is actu-ally going to be required to exerciselender-of-last-resort and bail out facili-ties. Neil and Jack call it the “moral haz-ard” problem, which is the idea thatwhen you insure a bank, you alter itsincentives to undertake risks.

In the Kareken-Wallace model, depositinsurance is purely a bad thing.Kareken-Wallace envisions a differenteconomic setting than Bryant andDiamond-Dybvig. Of course, like allmodels, it’s an abstraction; it simplifiesthings in order to isolate key forces. TheKareken-Wallace setting has completemarkets. There are markets in all possi-ble risky claims. There are also somepeople who wanted to hold risk-freedeposits.

Kareken and Wallace compare twodifferent situations. In one, there is nodeposit insurance; depositors are ontheir own and know that their depositsare uninsured. If they want to hold risk-free deposits, they’d better hold them inbanks that are holding risk-free portfo-lios. Some very conservative banksemerge that can issue safe depositsbecause the bank portfolio managersthemselves hold assets that allow thesebanks to pay depositors in all possiblestates of the world.

Kareken and Wallace compare thatno-deposit-insurance situation toanother situation in which a govern-ment agency provides deposit insurancethat is either free or is priced too cheap-ly, meaning that it’s not priced with aproper risk-loading. Kareken andWallace show that in that situation,banks have an incentive to become asrisky as possible, and as large as possi-ble. Therefore, with a positive probabil-ity, banks will fail and taxpayers willhave to compensate banks’ depositors. Itis in banks’ shareholders’ interest that

The Region

31 SEPTEMBER 2010

The Kareken andWallace model’s predictionis that if a government sets up depositinsurance and doesn’t regulate bankportfolios to prevent them from taking toomuch risk, the government is setting thestage for a financial crisis. ... So, of thosetwo models, the Kareken-Wallace modelmakes you very cautious about lender-of-last-resort facilities and very sensitive tothe risk-taking activities of banks. TheDiamond-Dybvig and Bryant model makesyou very sensitive to runs and veryoptimistic about the ability of insuranceto cure them. Both models leave somethingout, and I think in the real world we’re ina situation where we have to worry aboutruns and we also have to worry aboutmoral hazard.

the banks organize themselves this way.This lets them gamble with the insurers’and depositors’ money.

The Kareken and Wallace model’sprediction is that if a government setsup deposit insurance and doesn’t regu-late bank portfolios to prevent themfrom taking too much risk, the govern-ment is setting the stage for a financialcrisis. On the basis of the Kareken-Wallace model, Jack Kareken wrote apaper in the Federal Reserve Bank ofMinneapolis Quarterly Review referringto the “cart before the horse.”10 Hepointed out that if you’re going toderegulate financial institutions, whichwe in the United States did in the late’70s and early ’80s (deregulation is thecart), you’d better reform deposit insur-ance first (that’s the horse). You’d bettermake it clear that financial institutionsthat take these risks are not allowed tohave access to lender-of-last-resortfacilities. But the U.S. government didn’tdo that.

So, of those two models, the Kareken-Wallace model makes you very cautiousabout lender-of-last-resort facilities andvery sensitive to the risk-taking activitiesof banks. The Diamond-Dybvig andBryant model makes you very sensitive toruns and very optimistic about the abili-ty of insurance to cure them. Both mod-els leave something out, and I think inthe real world we’re in a situation wherewe have to worry about runs and we alsohave to worry about moral hazard. Asyou know, an important theme ofresearch for macroeconomics in generaland at the Minneapolis Fed in particularhas been about how to strike a good bal-ance.

Rolnick: Jack and Neil concluded their1978 paper with a proposal for dealingwith this tension, and that was torequire much more capital than wasrequired at the time. Now the govern-ment actually requires even less capitalthan it did when Jack and Neil wrote. Ifyou go back prior to FDIC insurance,turn-of-the-century banks were hold-ing, by some estimates, 20 percent,maybe 30 percent, capital. Capital-equity

ratios were that high.What would you recommend? You

just observed that if deposit insuranceisn’t priced properly, that leads you inone direction. And Jack and Neil hadthis idea of making sure there’s a lotmore skin in the game, meaning muchcloser to what banks used to hold whenthere was no deposit insurance, no too-big-to-fail.

Sargent: The function of capital is exact-ly to protect against making risky loans.Another proposal is the narrow bankingproposal of Milton Friedman and [othereconomists at the University of]Chicago, which is a proposal to forcedeposit banks to hold safe portfolios.

Rolnick: Well, with large banks, too-big-to-fail concerns and deposit insurance, Iwould make the case to tier it based onsize. Jack and Neil made the point, Ibelieve, that shareholders of large bankscan diversify, but shareholders of small-er banks find it harder to diversify, sothey tend to be more risk-averse. Theirprediction would therefore have been, Ithink, that moral hazard is more likelyto manifest itself in larger banks—and Ithink that’s what we saw in the 2007-09financial crisis. How seriously wouldyou take the relevance of the historicalevidence that I cited?

Sargent: I would take it very seriously. Irecommend a very interesting paper byWarren Weber presented at theMinneapolis Fed conference in honor ofGary Stern this past April in whichWarren compared different privateinsurance arrangements for managingbanks’ risk-taking before the U.S. CivilWar.11

THE 2009 FISCAL STIMULUS

Rolnick: A January 2009 article quotesyou as saying, “The calculations that Ihave seen supporting the stimulus pack-age are back-of-the-envelope ones thatignore what we have learned in the last60 years of macroeconomic research.”12

What calculations had you seen?

Sargent: I said something like that to areporter. I had just read an Obamaadministration’s Council of EconomicAdvisers document e-mailed to me bymy friend [Stanford University econo-mist] John Taylor.13 I agreed with Johnthat the CEA calculations were surpris-ingly naive for 2009. They were notinformed by what we learned after 1945.

But I suspect that the council wasasked to do something quickly, and theydid what they thought was “goodenough for government work,” as someof us said during my days at thePentagon in 1968 and 1969. Back-of-envelope work can be a useful startingpoint or benchmark. But it does mis-chief when it is oversold.

In early 2009, President Obama’s eco-nomic advisers seem to have under-stated the substantial professionaluncertainty and disagreement about the

The Region

32SEPTEMBER 2010

In early 2009, I recall President Obamaas having said that while there was ampledisagreement among economists aboutthe appropriate monetary policy andregulatory responses to the financial crisis,there was widespread agreement in favorof a big fiscal stimulus among the vastmajority of informed economists. Hisadvisers surely knew that was not anaccurate description of the full range ofprofessional opinion.

wisdom of implementing a large fiscalstimulus. In early 2009, I recallPresident Obama as having said thatwhile there was ample disagreementamong economists about the appropri-ate monetary policy and regulatoryresponses to the financial crisis, therewas widespread agreement in favor of abig fiscal stimulus among the vastmajority of informed economists. Hisadvisers surely knew that was not anaccurate description of the full range ofprofessional opinion. President Obamashould have been told that there arerespectable reasons for doubting thatfiscal stimulus packages promote pros-perity, and that there are serious eco-nomic researchers who remain uncon-vinced.

Rolnick: Do any New Keynesian modelsprovide any support for the CEA num-bers?

Sargent: Some do; some don’t. I recom-mend looking at calculations by JohnTaylor and his pals.14 Based on thatwork, John remains very skeptical of the2009 CEA calculations. But Christiano,Eichenbaum and Rebelo have used vari-ants of a New Keynesian model togeth-er with particular assumptions aboutpaths of shocks to create quantitativeexamples of situations in which fiscalmultipliers can be as big as thoseassumed by the CEA.15

PERSISTENT UNEMPLOYMENT INEUROPE (AND NOW THE UNITEDSTATES?)

Rolnick: Let me go on to another set ofquestions that I have struggled toanswer. Is U.S. unemployment in thisrecession special? Is it different from theprevious 10 recessions? If so, do youhave any explanation for why that mightbe the case? Why it went so high andwhy it’s staying there as long as it is, rel-ative to the pattern of other recoveries?

I haven’t heard many economistsexpound on this, but clearly the labormarkets are behaving much differentlythan they did in previous recoveries,

and it’s not obvious to me why. I’m curi-ous what you might say about that.

Sargent: May I talk about this by linkingto some of my work with LarsLjungqvist on European unemploy-ment?

Rolnick: By all means.

Sargent: I have little new to say aboutthe details of the big rise in U.S. unem-ployment since 2008, although thefinancial crisis was a huge adverse shockto the labor market, so I suspect thatwe’ll be able to explain the rise. But themain thing that concerns me is thethreat of persistent high unemployment,and here the European experience of thelast three decades fills me with dread.

Let me begin by explaining whatmotivated Lars Ljungqvist and me tostudy European unemployment, why wehave been obsessed by it for 15 years. ToLars and me, Europe’s high unemploy-ment rate during the last three decadesrepresents an enormous waste of humanresources and individuals’ well-being,what we think is a tragedy in the lives ofthe people who have not been able toparticipate in the labor market.

Early explanations in the 1980s forEurope’s high unemployment were thatit was due to insufficient demand andwage rigidities. But soon those explana-tions came to be regarded as unsatis-factory because they couldn’t explain thepersistence of unemployment. Sometheories blamed Europe’s labor marketinstitutionswith their generous government-supplied unemployment insurance andstrong government-mandated job pro-tection.

But those theories were decisivelycriticized by Paul Krugman and otherswho pointed out that the Europeaninstitutions that liberally subsidizedunemployment and disability retire-ment were there also in the 1950s and’60s, periods when Europe had lowerunemployment rates than the UnitedStates. Therefore, Krugman and othersconcluded that you can’t blame thosegenerous European social safety nets for

the high unemployment rates thatEurope has experienced since 1980.

Here’s how Lars and I have attackedthe problem. We believe that despiteKrugman’s observation, Europe’s gener-ous unemployment compensation sys-tem has made an important contribu-tion to sustained high European unem-ployment, but that those adverse effectscame to life only after there occurredwhat seem to have been permanentchanges in the microeconomic environ-ment confronting individual workers.So the culprit was the interactions ofthose altered microeconomic conditionswith those generous European social safe-ty nets.

Rolnick: What changes in microeco-nomic conditions do you have in mind?

Sargent: Empirical microeconomistshave documented that, despite whatmacroeconomists called the “GreatModeration” in macroeconomic volatil-ity before 2007, individual workers haveexperienced more turbulent labor mar-ket outcomes since the late 1970s andearly ’80s. Empirical studies have docu-mented increased volatility of both thetransient and permanent components ofindividuals’ labor earnings. PeterGottschalk and Robert Moffitt, CostasMeghir and Luigi Pistaferri, and othershave documented that.16 David Autorand Larry Katz have assembled a con-vincing catalogue and critical summaryof the evidence.17 So if you look atinstances when a job separation causesan individual’s earnings to suffer a bigreduction, usually that individual mustlive with a substantial reduction for along time.

Lars and I use the shorthand“increased turbulence” to refer to thisincreased volatility and magnitude ofadverse earnings shocks at the time ofjob loss. In the context of several ration-al expectations models with human cap-ital dynamics and labor market frictionsthat impede the ability of displacedworkers to find new jobs, we have foundthat an increase in economic turbulencegenerates persistently high unemploy-

The Region

33 SEPTEMBER 2010

ment when combined with a generouswelfare system.

Furthermore, the same government-financed social safety net could actuallyproduce lower unemployment in a low-turbulence environment like the 1950sand 1960s. It could do this through stronggovernment-mandated job protection.But when the microeconomic turbulenceincreases to the high-turbulence post-1980 environment, that same safety netcan unleash persistently higher unem-ployment. An important element of ouranalysis is the view that a worker’shuman capital tends to grow when he orshe is employed, but deteriorates whenhe or she is not employed. We analyzedthese mechanisms in detail in twopapers, one in the JPE in 1998, anotherin Econometrica in 2008.18

That’s our explanation for the higherunemployment rate observed in Europefrom 1980 to 2007. Our vision is that anincrease in microeconomic turbulenceof individual earnings processesoccurred in both Europe and the UnitedStates. Displaced American workersfaced stingier unemployment compen-sation systems, stingier in both theirmore limited durations and their lowermonthly payments.

Rolnick: When did the microeconomicturbulence begin?

Sargent: The empirical evidence is thatit increased substantially sometime inthe late 1970s. It happens that itincreased just about when high and per-sistent unemployment broke out inEurope. This is what attracted us to it asa key part of the explanation for the per-sistent jump in unemployment inEurope relative to the United States.

Rolnick: So turbulence broke out inEurope, OK, but you get the impressionthat the Great Moderation—a decline ineconomic volatility—was taking placehere in the United States.

Sargent: Well, the so-called GreatModeration really refers to a decrease inmacroeconomic volatility. That’s why I

stress the difference between individualand aggregate volatility by emphasizingthe term microeconomic. The GreatModeration is indeed there in the aggre-gate data. An econometrician wouldthink about running a simple auto-regressive process for aggregate dataand then looking at the error variance.For aggregate data (until 2007), thaterror variance decreased. But for themicro or individual-level data, just theopposite happened: For individualworkers, the error variance—or lesstechnically, unpredictable volatility inearnings—increased.

Rolnick: And why did microeconomicearnings volatility increase?

Sargent: Lars and I believe that whenpeople now become unemployed,they’re taking a more or less permanenthit to their level of human capital, a larg-er one than they might have received

before 1980. We have a theory that peo-ple build up human capital while they’reworking on a job, but lose human capi-tal when they’re displaced from a job.We think that after 1980, people inWestern economies started sufferingbigger drops in their human capital atthe moment that they suffer a job dis-placement. Some of the forces leading tothis outcome come from various tech-nological changes going under theumbrella name of “globalization.”

Thomas Friedman’s 2005 book TheWorld Is Flat has many stories testifyingto such forces.19 By positing increasedturbulence in this sense at the micro-economic level, Lars and I have beenable both to come to grips with theobservations on aggregate unemploy-ment across Europe and the UnitedStates and also to explain some of themicro observations collected byGottschalk and Moffitt and others. Sothe Great Moderation seems not to havebeen occurring at the individual level.Just the opposite.

Our theory goes beyond the aggre-gate unemployment rate and focuses onindividuals. Our models have cohorts ofaging heterogeneous workers. Ourmodels imply that people in Europe,especially older workers, are sufferingfrom long-term unemployment becauseof the adverse incentives brought aboutby a generous social safety net when itinteracts with these human capitaldynamics. Unfortunately, the data bearthis out. In Europe, there has been along-term unemployment problemespecially affecting older workers.

Rolnick: In your model, what type oflabor market frictions impede peoplewho want to work from immediatelyfinding a job?

Sargent: The models that we like best forour purposes view unemployment as an“activity” distinct from “work” and“leisure.” We’ve cast the heart of our the-ory in several contexts, including, forexample, search models in the spirit ofGeorge Stigler and John McCall,20

where finding a job requires a time-

The Region

34SEPTEMBER 2010

After 1980, people in Western economiesstarted suffering bigger drops in theirhuman capital at the moment that theysuffer a job displacement. Some of theforces leading to this outcome come fromvarious technological changes going underthe umbrella name of “globalization.”Thomas Friedman’s 2005 book The

World Is Flat has many stories testifyingto such forces.

consuming activity of sorting throughoffers for jobs with various levels of payand compensating differences; and alsoDiamond-Mortensen-Pissarides21

matching models where an aggregatematching function imposes a conges-tion externality on workers’ activity ofwaiting for a match and firms’ activity ofwaiting for vacancies to be filled. Thesame forces come through across a vari-ety of structures, so we think there’s a lotof robustness to our basic story.

Rolnick: OK, so part of your story forlower U.S. unemployment in the pasthas to be that in the United States, espe-cially for the older workers, the safetynet wasn’t as generous. They had to goback and get retrained or whatever;therefore they chose to be more active inthe labor market than their Europeancousins did.

Sargent: Yes. In a 2003 paper in a vol-ume to honor Edmund S. Phelps, Larsand I exhibited simulations of ourmodel illustrating this.22 What would atypical, say, a 50-year-old worker do ifhe or she loses his or her job and thenimmediately gets hit by a human capitalloss? What differences in behaviorwould be exhibited by otherwise similarworkers, one facing European benefitsversus another facing U.S. benefits?

Our simulations exhibit a force thattraps the European worker in unem-ployment. Unemployment compensationsystems typically award you compensa-tion that’s linked to your earnings onyour last job; those past earnings reflectyour past human capital, not your cur-rent opportunities or current humancapital. That can make collecting unem-ployment compensation at rates reflect-ing your past (and now obsolete)human capital more desirable thanaccepting a job whose earnings reflect areturn on your current depreciated levelof human capital. This mechanism setsan incentive trap that induces theEuropean worker to withdraw fromactive labor market participation.

Rolnick: Earlier, you said that the

European experience with persistentlyhigh unemployment over the last threedecades fills you with dread about theprospects for the United States.

Sargent: The prospect that concerns memight sound like I’m hardhearted, butthat’s just the opposite of my feelings.What you’ve seen in the recent reces-sion—and it’s quite natural because it’sbeen so severe—is a tendency of

Congress to expand unemploymentbenefits, over and over again. What Larsand my theory tells us is that if, in theUnited States, we create a system whereunemployment and disability benefitsare permanently extended in their gen-erosity and their duration, we will inad-vertently put ourselves into the situationthat much of Europe has suffered forthree decades.

I don’t know enough about politics topredict whether that’s likely to happen.The unfortunate thing is you can see amultiple equilibrium trap here. Lowunemployment rates enabled the UnitedStates politically to sustain a modestunemployment compensation system.But the politics of the current situationcan imply that so long as unemploy-ment is high, we’re going to extend theduration and generosity of benefits.And that extension, done out of the bestof motives, is exactly what can lead tothe trap of persistently high unemploy-ment. An intriguing thing is that someEuropean countries like Sweden andDenmark are now moving exactly in theopposite direction.

EUROPE AND “UNPLEASANTARITHMETIC”

Rolnick: Let me ask another questionabout events in Europe. Some peoplebelieve there’s a serious conflict betweenfiscal and monetary policy, that it’s theresult of the Europeans having askedmonetary policy to do things it can’twithout real fiscal discipline. And as youand Neil pointed out 30 years ago—wasit that long ago?!—in “Some UnpleasantMonetarist Arithmetic,” you’d betterworry about those links. Is that the wayyou would interpret what’s going on inGreece, or Europe in general, and con-cern over Europe’s ability to maintainthe euro, that they face some unpleasantarithmetic that could undermine theeuro?

Sargent: The people who set up the euroclearly knew about the unpleasant arith-metic and they strove to set things up toprotect the euro from any adverse con-

The Region

35 SEPTEMBER 2010

Collecting unemployment compensationat rates reflecting your past (and nowobsolete) human capital [can be] moredesirable than accepting a job whoseearnings reflect a return on your currentdepreciated level of human capital. Thismechanism sets an incentive trap thatinduces the ... worker to withdraw fromactive labor market participation.

Low unemployment rates enabledthe United States politically to sustaina modest unemployment compensationsystem. But the politics of the currentsituation can imply that so long asunemployment is high, we’re going toextend the duration and generosity ofbenefits. And that extension, done outof the best of motives, is exactly whatcan lead to the trap of persistently highunemployment.

sequences of that arithmetic. Indeed, thewhole system was designed to force gov-ernments to balance their budgets in apresent value sense, adjusting appropri-ately for growth. Indeed, the MaastrichtTreaty actually put in fiscal rules thatamounted to overkill in the interests ofcreating a fail-safe system.

What I mean is that it put in placemore restrictive rules on fiscal policythan were needed to express the require-ment that a government’s budget had tobe balanced in the present-value sensewith little or no contributions comingfrom seigniorage revenues from theinflation tax. The treaty built redundan-cy into the rules by restricting bothdebt-to-GDP ratios and deficit-to-GDPratios.

Remember that under the gold stan-dard, there was no law that restrictedyour debt-GDP ratio or deficit-GDPratio. Feasibility and credit markets didthe job. If a country wanted to be on thegold standard, it had to balance its budg-et in a present-value sense. If you didn’trun a balanced budget in the present-value sense, you were going to have arun on your currency sooner or later,and probably sooner. So, what inducedone major Western country after anoth-er to run a more-or-less balanced budg-et in the 19th century and early 20thcentury before World War I was theirdecision to adhere to the gold standard.

Rolnick: What does the gold standardhave to do with the euro in 2010?

Sargent: The euro is basically an artifi-cial gold standard. The fiscal rules in theMaastricht Treaty were designed tomake explicit the present-value budgetbalance that was unspoken under thegold standard. In terms of the mone-tarist arithmetic, the rules made sense.

Rolnick: So what’s the problem now?

Sargent: Here is what went haywire. Inthe 2000s, France and Germany, the twokey countries at the center of the Union,violated the fiscal rules year after year.Of course, an intriguing thing about the

unpleasant arithmetic is that it’s aboutpresent values of government primarydeficits, and not just deficits for one, twoor three years. And remember that theoverkill Maastricht Treaty rules aresufficient but not necessary to sustainpresent-value budget balance, adjustedfor real economic growth, so maybe therewas no cause for alarm at that time.

But in hindsight, there was cause foralarm. The reason is that France andGermany lost the moral authority to saythat they were leading by example. Theylost the moral high ground to holdsmaller countries to the fiscal rulesintended to protect monetary policyfrom the need to monetize governmentdebt.

Rolnick: And so …

Sargent: So, a number of countries at theEuropean Union economic periphery—Greece, in particular—violated the rulesconvincingly enough to unleash thethreat of unpleasant arithmetic in those

countries. The telltale signs were persist-ently rising debt-GDP ratios in thosecountries. Of course, the unpleasantarithmetic allows them to go up for awhile, but if that goes on too long, even-tually you’re going to get a sovereigndebt crisis.

Rolnick: What could the EuropeanCentral Bank do then?

Sargent: Well, here is one thing that youcan imagine the ECB doing (which ithasn’t). It could take the stance, “If thegovernment of Greece wants to try toissue euro-denominated bonds, let themdo it, or try to do it. And if investorswant to hold euro-denominated bondsthat are understood to be liabilities ofthe Greek government, and not of theECB, let them do it. It’s not any of theECB’s business. If those bonds threatento go bad, if Greece just isn’t a good risk,that’s the bondholders’ problem. Let theinvestors bear that risk. And if Greecedefaults or renegotiates, that’s theinvestors’ problem, not the ECB’s prob-lem.”

Rolnick: Of course, the ECB hasn’t saidthat, or at least not yet!

Sargent: Well, one reason the ECB hasn’tsaid that yet is that after the financialcrisis of 2008, what seemed to someEuropean banks to be a promisingsource of higher-yielding instrumentswas sovereign debt in the form of euro-denominated bonds issued by countrieslike Greece. The banks located in thecenter of the euro area, France andGermany, hold Greek-denominateddebt, so a threat of default on Greek gov-ernment debt threatens the portfolios ofthose banks in other European coun-tries. Because it is the lender of lastresort, now it is the ECB’s business.

Rolnick: Tom, this reminds me of anexample of a breakdown in one of the linesbetween monetary and fiscal policy thatyou wrote about in your paper “Where toDraw Lines” that you presented at theStern conference we held in April.23

The Region

36SEPTEMBER 2010

In the 2000s, France and Germany, thetwo key countries at the center of theUnion, violated the fiscal rules year afteryear. ... They lost the moral high groundto hold smaller countries to the fiscal rulesintended to protect monetary policy fromthe need to monetize government debt.

Sargent: Yes, this is a big breakdown ina line between fiscal and monetarypolicy intended to be set by theMaastricht Treaty in order to enforcethat artificial gold standard, as I viewthe euro to be. Once the monetaryauthority starts assisting the fiscalauthorities of these countries, you’vedrifted from the original conception ofthe euro.

Rolnick: Would you argue that Jack andNeil’s analysis comes back into playhere, the one about too-big-to-fail andmoral hazard?

Sargent: Unfortunately, yes, that’s what Iwas trying to suggest.

Rolnick: Did things have to get to thispoint?

Sargent: Ultimately, that’s a questionabout politics, about which I know toolittle. But in purely economic terms,things could have gone differently.Here’s a “virtual history” of what couldhave happened:

France and Germany stay “holierthan thou” from beginning to end, andalways respect the fiscal limits imposedby the Maastricht Treaty. They therebyacquire the moral authority to lead byexample, and the central core of euro-area countries are running budgets thatwithout doubt are balanced in a present-value sense. Therefore, the euro is

strong. The banks of the core countries(France and Germany again) are wellregulated (the message of Kareken andWallace has been heard), so the banks inFrance and Germany are not holdingany dodgy bonds issued by govern-ments of dubious peripheral countriesthat have adopted the euro but that flirtwith violating the Maastricht Treatyrules.

In this virtual history, the ECB couldplay tough and let the Greek governmentdefault on its creditors by renegotiatingterms of the debt. For the euro, letting theGreek bondholders suffer would actuallybe therapeutic; it would strengthen theeuro by teaching peripheral countriesthat the ECB means business.

The Region

37 SEPTEMBER 2010

More About omas J. Sargent

Current Positions

William R. Berkley Professor of Economics and Business,New York University, since 2002

Senior Fellow, Hoover Institution, Stanford University, since 1987

Research Associate, National Bureau of Economic Research, 1970–73and since 1979

Previous Positions

Donald Lucas Professor of Economics, Stanford University, 1998–2002

David Rockefeller Professor of Economics, University of Chicago, 1991–98

Visiting Scholar, Hoover Institution, Stanford University, 1985–87

Visiting Professor of Economics, Harvard University, 1981–82

Ford Foundation Visiting Research Professor of Economics, University ofChicago, 1976–77

Adviser, Federal Reserve Bank of Minneapolis, 1971–87

Associate Professor of Economics, University of Minnesota, 1971–87;Professor from 1975

Associate Professor of Economics, University of Pennsylvania, 1970–71

First Lieutenant and Captain, U.S. Army; served as Staff Member andActing Director, Economics Division, Office of the Assistant Secretaryof Defense, 1968–69

Professional Affiliations

President, American Economic Association, 2007; President-elect, 2006;Vice President, 2000–01; Executive Committee, 1986–88

President, Econometric Society, 2005; First Vice President, 2004;Second Vice President, 2003; Council, 1995–99, 1987–92; Fellow, 1976

President, Society for Economic Dynamics and Control, 1989–92

Member, Brookings Panel on Economic Activity, 1973

Honors and Awards

Moore Distinguished Scholar, California Institute of Technology, 2000–01

Marshall Lecturer, Cambridge, England, 1996

Erwin Plein Nemmers Prize in Economics, Northwestern University,1996–97

Fellow, American Academy of Arts and Sciences, 1983

Fellow, National Academy of Sciences, 1983

Mary Elizabeth Morgan Prize for Excellence in Economics,University of Chicago, 1979

Most Distinguished Scholar, University of California, Berkeley, Class of 1964

Phi Beta Kappa, 1963

Publications

Author of a dozen books, including Robustness (with Lars Peter Hansen,2008; Recursive Macroeconomic Theory (with Lars Ljungqvist),2d ed., 2004; The Big Problem of Small Change (with François Velde),2002; The Conquest of American Inflation, 1999; and Bounded Rationalityin Macroeconomics, 1993. Author of more than 170 research papersfocusing on macroeconomic theory, time-series econometrics, learningtheory, fiscal and monetary policy, and economic history.

Education

Harvard University, Ph.D., 1968

University of California, Berkeley, B.A., 1964

Rolnick: Right. Although if that scenariohad been foreseen, Greece might nothave been able to issue that debt in thefirst place.

Sargent: Aha! The plot thickens. So thenwe confront again the issue of how sep-arate can monetary and fiscal policy be?In the spirit of your observation,remember that there were huge capitalgains on Italian debt after it becameclear that it would be allowed to join theeuro area. So, what really was the reasonfor those capital gains? Were they basedon expectations of a reformed and moredisciplined fiscal policy in Italy? Or wasit rather an expectation that by joiningthe euro, Italy had gained access tobailouts from other euro-zone countries?

Note that a related point pertains tothe 2009 stress tests in the United States.What did it truly mean when a bankpassed the stress test? Did it mean thatthe bank’s balance sheet was solid? Ordid it mean that since the Fed said thatbank had passed the stress test, the Fedwould make sure that henceforth thatbank would have access to lender-of-last-resort facilities?

It’s difficult to sort these things out.But notice that throughout our discus-sion, Art, we’ve been using the vocabu-lary of rational expectations. In ourdynamic and uncertain world, ourbeliefs about what other people andinstitutions will do play big roles inshaping our behavior.

Rolnick: Indeed. Thank you again, Tom.

—Art RolnickJune 15, 2010

Endnotes

1 Hansen, Lars Peter, and Kenneth J.Singleton. 1983. “Stochastic Consumption,Risk Aversion, and the Temporal Behaviorof Asset Returns.” Journal of Political Economy91(2), pp. 249–65.2 Mehra, Rajnish, and Edward C. Prescott.1985. “The Equity Premium: A Puzzle.”Journal of Monetary Economics 15(2), pp.145–61.

Weil, Philippe. 1989. “The Equity PremiumPuzzle and the Risk-Free Rate Puzzle.” Journalof Monetary Economics 24(3), pp. 401–21.

Backus, David K., and Gregor W. Smith. 1993.“Consumption and Real Exchange Rates inDynamic Economies with Non-TradedGoods.” Journal of International Economics35(3/4), pp. 297–316.3 Stiglitz, Joseph E. 2009. “Obama’s ErsatzCapitalism.” New York Times, April 1.

Sachs, Jeffrey. 2009. “Obama’s Bank PlanCould Rob the Taxpayer.” Financial Times,March 25.4 Harrison, J. Michael, and David M. Kreps.1978. “Speculative Investor Behavior in aStock Market with HeterogeneousExpectations.” Quarterly Journal of Economics92(2), pp. 323–36.5 Scheinkman, José A., and Wei Xiong. 2003.“Overconfidence and Speculative Bubbles.”Journal of Political Economy 111(6),pp. 1183–1219.6 “New Keynesian economics” refers to theschool of thought that has refined Keynes’original theories in response to the Lucascritique and rational expectations. Advocatesof New Keynesian theory rely on the idea thatprices and wages change slowly (referred toas “sticky” prices and wages) to explain whyunemployment persists and why monetarypolicy can influence economic activity. Buttheir models are adapted from the dynamicgeneral equilibrium models developed bynew classical economists such as Sargent andLucas. See http://www.econlib.org/library/Enc/NewKeynesianEconomics.html forelaboration.7 Allen, Franklin, and Douglas Gale. 2007.Understanding Financial Crises. Oxford andNew York: Oxford University Press.8 Stern, Gary H., and Ron J. Feldman. 2004.Too Big to Fail: The Hazards of Bank Bailouts.Washington, D.C.: Brookings InstitutionPress.9 Kareken, John H., and Neil Wallace. 1978.“Deposit Insurance and Bank Regulation:

A Partial-Equilibrium Exposition.” Journal ofBusiness 51(July), pp. 413–38. (Also athttp://minneapolisfed.org/research/sr/SR16.pdf.)

Bryant, John. 1980. “A Model of Reserves,Bank Runs, and Deposit Insurance. Journalof Banking & Finance 4(4), pp. 335–44.

Diamond, Douglas W., and Philip H. Dybvig.1983. “Bank Runs, Deposit Insurance, andLiquidity.” Journal of Political Economy 91(3),pp. 401–19. (Also at http://minneapolisfed.org/research/QR/QR2412.pdf.)10 Kareken, John H. 1983. “Deposit InsuranceReform or Deregulation Is the Cart, Not theHorse.” Federal Reserve Bank of MinneapolisQuarterly Review 7(Spring), pp. 1–9.11 Weber, Warren E. 2010. “Bank LiabilityInsurance Schemes Before 1865.” ResearchDepartment Working Paper 679, FederalReserve Bank of Minneapolis. (See also “AnAntebellum Lesson” in this issue of TheRegion.)12 Chicago Tribune. 2009.“The Stimulus Rush.”Jan. 13. (Also in Brannon, Ike, and ChrisEdwards. 2009. “The Troubling Return ofKeynesianism.” Tax & Budget Bulletin 52, CatoInstitute, http://www.cato.org/pubs/tbb/tbb_0109-52.pdf.)13 Romer, Christina, and Jared Bernstein.2009. “The Job Impact of the AmericanRecovery and Reinvestment Plan,”http://otrans.3cdn.net/ee40602f9a7d8172b8_ozm6bt5oi.pdf.14 Cogan, John F., Tobias Cwik, John B. Taylorand Volker Wieland. 2009. “New Keynesianversus Old Keynesian Government SpendingMultipliers.” Working Paper 47, StanfordUniversity, http://www.hoover.org/news/daily-report/15586; for a newer version of thispaper (Jan. 8, 2010), see http://www.stanford.edu/~johntayl/CCTW_100108.pdf.15 Christiano, Lawrence, Martin Eichenbaumand Sergio Rebelo. 2009. “When Is theGovernment Spending Multiplier Large?”Working Paper, Northwestern University,http://www.kellogg.northwestern.edu/faculty/rebelo/htm/multiplier.pdf.16 Gottschalk, Peter, and Robert Moffitt. 1994.“The Growth of Earnings Instability in theU.S. Labor Market.” Brookings Papers onEconomic Activity 2, pp. 217–72.

Meghir, Costas, and Luigi Pistaferri. 2004.“Income Variance Dynamics andHeterogeneity.” Econometrica 72(1), pp. 1–32.17 Katz, Lawrence F., and David H. Autor.1999. “Changes in the Wage Structure andEarnings Inequality,” in Handbook of LaborEconomics, vol. 5. Oxford and New York:

The Region

38SEPTEMBER 2010

Elsevier Science, North-Holland, pp.1463–1555.18 Ljungqvist, Lars, and Thomas J. Sargent.1998. “The European UnemploymentDilemma.” Journal of Political Economy106(3), pp. 514–50.

Ljungqvist, Lars, and Thomas J. Sargent. 2008.“Two Questions about EuropeanUnemployment.” Econometrica 76(1), pp. 1–29.19 Friedman, Thomas L. 2005. The World IsFlat: A Brief History of the Twenty-FirstCentury. New York: Farrar, Straus and Giroux.20 Stigler, George J. 1962. “Information in theLabor Market.” Journal of Political Economy70(2), pp. 94–105.

McCall, John J. 1970. “Economics ofInformation and Job Search.” QuarterlyJournal of Economics 84(1), pp. 113–26.21 Diamond, Peter A. 1982. “WageDetermination and Efficiency in SearchEquilibrium.” Review of Economic Studies49(2), pp. 217–27.

Mortensen, Dale T. 1982. “Property Rightsand Efficiency in Mating, Racing, and RelatedGames.” American Economic Review 72(5),pp. 968–79.

Pissarides, Christopher A. 1992. “Loss ofSkill During Unemployment and thePersistence of Employment Shocks.” QuarterlyJournal of Economics 107(4), pp. 1371–91.

Mortensen, Dale T., and Christopher A.Pissarides. 1999. “Unemployment Responsesto ‘Skill-Biased’ Technology Shocks: The Roleof Labor Market Policy.” Economic Journal109(455), pp. 242–65.22 Ljungqvist, Lars, and Thomas J. Sargent.2003. “European Unemployment: From aWorker’s Perspective,” in Knowledge,Information, and Expectations in ModernMacroeconomics: In Honor of Edmund S.Phelps. Philippe Aghion, Roman Frydman,Joseph Stiglitz and Michael Woodford, eds.Princeton, N.J.: Princeton University Press,pp. 326–50.23 Sargent, Thomas J. 2010. “Where to DrawLines: Stability versus Efficiency,”http://homepages.nyu.edu/~ts43/research/phillips_ver_9.pdf. (See also http://www.minneapolisfed.org/research/events/2010_04-23/index.cfm.)

The Region

39 SEPTEMBER 2010