regional economist - january 2010

8/9/2019 Regional Economist - January 2010

1/24

Inf tionMay Be the NextDragon To Slay

The RegionalEconomist

A Quarterly Reviewof Business and Economic Conditions

Vol. 18, No. 1 J n 2010

The Feder al reserve Bank oF sT. louis

C e n T r a l t o a m e r i C a s e C n m y Tm


2/24

3 P r n t M g

10 Fears about CommercialReal Estate Loans

By Rajeev Bhaskar, Yadav Gopalanand Kevin L. Kliesen

Community banks seem to havethe most to fear about the state of commercial real estate today. Teproblems with these loans, however,shouldnt derail the entire economy.

12 The Returns to Education

By Natalia Kolesnikova

Most studies estimate that thereturn to each year of education isabout 10 percent. But calcu latingthe nancial gain is not a cut-and-dried process. Even more di cult iscalculating the nonmonetary return.

c o n t n t

Inf tion M Be Next D gon To SlBy Kevin L. Kliesen

Although some think its too soon to worry about high in ation,there are risks of such happening in the medium term. Besidesthe obvious, a new bubble might be brewing in asset prices asinvestors search for higher returns, and the gap between actualoutput and potential output might be smaller than most think.

4

The Regional Economist is publishedqua e ly by he resea h a d Publi

ai s depa me s he Fede alrese ve Ba k . L uis. add esses

he a i al, i e a i al a d e i ale mi issues he day, pa i ula lyas hey apply s a es i he i h hFede al rese ve is i . Viewsexp essed a e e essa ily h se

he . L uis Fed he Fede alrese ve ys em.

Please di e y u mme s ubhayu Ba dy padhyay a 314-

444-7425 by e-mail a subhayu.ba dy padhyay@s ls. b. . Y u aals w i e him a he add ess bel w.

ubmissi a le e he ediives us he i h p s i u web

si e a d/ publish i i The RegionalEconomist u less he w i e s a es

he wise. We ese ve he i h edile e s la i y a d le h.

Director of Researchch is phe J. Walle

Senior Policy Adviserr be H. ras he

Deputy Director of Research

cle us c. c u hliDirector of Public Affairsr be J. he k

Editorubhayu Ba dy padhyay

Managing Editorl amb ski

Art DirectorJ i Williams

i le- py subs ip i s a e ee.t subs ibe, e-mail a l.a.musse@s ls. b. si up via www.s l uis ed. /publi a i s. Y u aals w i e The Regional Economist ,Publi ai s o f e, Fede al rese veBa k . L uis, B x 442, . L uis,Mo 63166.

The Eighth Federal Reserve District i ludes all ka sas, eas eMiss u i, s u he lli is a d dia a,wes e Ke u ky a d te essee, a d

he Mississippi. the i h h is if es a e i Li le r k, L uisville,

Memphis a d . L uis.

The RegionalEconomist

JaNuary 2010 | VoL. 18, no. 1

14 The Dismal ScienceTackles Happy Talk

By Rubn Hernndez-Murillo and Christopher J. Martinek

While many believe that money does buy happiness, researchshows that richer people arentnecessarily happier people.

16 c o M M n t Y P r o F L

Mobe l , Mo.

By Susan C. Tomson

Once well-known as Te MagicCity, this Missouri town has anot-so-magic formula for keepingthe economy humming: focusmore on retaining businesses

than on attracting new ones.

19 n t o n L o V r V W

Ho-H m reco e

By Kevin L. Kliesen

In the past, deep recessions werfollowed by strong recoveries.But the forecast for GDP growtthis year is 2.5-3 percentnotexactly strong.

21 t r c t o V r V W

Comp ing Job Losses

By Craig P. Aubuchon,Subhayu Bandyopadhyay,Rubn Hernndez-Murillo and Christopher J. Martinek

Te four big metropolitan areasin the Eighth DistrictSt. LouiLittle Rock, Louisville andMemphisare faring a bit bettethan the nation as a whole whenit comes to job losses.

22 cono MY t gL nc

23 r r c H n g

ov us on by J son Mwww.Mun o Mp gn . oM

2 The Regional Economist | J n 2010


3/24

J mes B ll d , P eside a d c oFede al rese ve Ba k . L uis

For the past year, the Federal Open MarketCommittee has maintained its target forthe federal funds rate at a level only fraction-ally greater than zero. During the same year,it moved monetary policy into the unchartedwaters of quantitative easing. Althoughde nitions di er, quantitative easing mosto en is de ned as a policy strategy of seekingto reduce long-term interest rates by buyinglarge quantities of nancial assets when theovernight rate is zero.

At the end of 2008, some analysts arguedthat the FOMC was out of ammunitionwhen overnight interest rates reached zero be-cause nominal interest rates, ordinarily, do notgo below zero. (Tere were some exceptionsduring the Great Depression, and negativenominal rates occasionally are observed in -nancial markets when penalties are included.)Tis assertion, however, ignores one impor-tant fact: Te Fed can continue to purchase

assets so long as the public is willing to acceptdeposits at the Federal Reserve banks in pay-ment. Central banks that engage in quantita-tive easing purchase only high-quality assetswith suitable collateral margins; doing other-wise would be to dabble in scal rather thanmonetary policy. Economic theory suggests,however, that central banks need to purchase very large amounts of such assets (relative tothe size of the economy) if quantitative easingpolicies are to a ect economic activity.

From the beginning of 2009 until early

December, the Federal Reserve under theauspices of its Large Scale Asset Purchaseprogram had bought approximately $300billion in reasury securities, $150 billion indebt securities of Fannie Mae and FreddieMac, and $1.1 trillion of xed-rate mortgage-backed securities (MBS) guaranteed by GinnieMae, Fannie Mae and Freddie Mac. Ad-ditional purchases of agency debt and MBSare in-process. When completed, the FederalReserves total assets will likely reach

$2.6 trillion, and the Federal Reserve willown between one- h and one-fourth of thetotal outstanding amounts of reasury andagency-guaranteed MBS. Te monetary baseperhaps will reach $2.4 trillion, of which $1.5trillion will be deposits of depository institu-tions at the Federal Reserve. wo years ago,in December 2007, the monetary base wasapproximately $830 billion, with only $10 to$15 billion held by banks as deposits at the Fed.

Te United States is not the only country that has pursued such massive expansionary policy during 2009. Te Bank of England,for comparison, initiated quantitative easingin March 2009 and has purchased more than175 billion in British reasuries; it also holds

more than one-quarter of all such securitiesoutstanding. Although used infrequently,central banks worldwide during the pasttwo decades have used major increases anddecreases in their balance sheets as a policy instrument in a variety of circumstances.

A forthcoming article by Richard Andersonand others in our Research division comparesthe experience of a number of countries, in-cluding the United States, the U.K., Sweden,Switzerland, Japan and Australia. 1 Teir

study suggests two lessons for policymakersthat contribute to the success of such policies.First, communication matters: It is importantthat the public be told why the increases areoccurring and be assured that the increasesare temporary, not permanent. Second, itis essential that the increases are reversedas soon as possible a er the conditions thatcaused the adoption of a quantitative easingpolicy fade. Doing both appears to forestallincreases in expected in ation that mightotherwise cause increases in actual in ation,derailing the anticipated expansionary impactof the asset purchases.

Although nal determination of the e ectsof quantitative easing awaits further research,it is likely that quantitative easing did assist

economic recovery during 2009. Economistshave yet to develop macroeconomic modelswith nancial sectors adequately detailed toexplore the channels through which quanti-tative easing boosts economic activity. Butquantitative easing has a riskif o settingpolicy actions are not taken in a timely fash-ion, the increased monetary base will fuel anundesirably large acceleration of credit and, inturn, undesirably large increases in in ation.An important part of the mechanism must bestability of in ation expectations. Credible

commitment to maintaining low future in a-tion provides a central-bank policymaker withthe exibility to double or triple the centralbanks balance sheet while not unhingingin ation expectations.

Q ntit ti e E sing: unch ted W te so Monet Polic

P r n t M g

l h u h used i eque ly,e al ba ks w ldwide

du i he pas w de adeshave used maj i easesa d de eases i hei bala eshee s as a p li y i s umei a va ie y i ums a es.

E N D N O E S1 Anderson, Richard; Gascon, Charles; and Liu, Yang.

Doubling Your Monetary Base and Surviving: SomeInternational Evidence. Forthcoming, Federal ReserveBank of St. LouisReview.

The Regional Economist | www.stlo is ed.o g 3


4/24

P o t - c r



5/24

By Kevin L. Kliesen

May Be the Next Dragon

o Slay By most metrics, the recent recession was the longest anddeepest since the 1930s. Some analysts believe that the Fed-eral Reserves and the federal governments aggressive actions toassist and stabilize the economy and fragile credit markets pre- vented an even worse outcome than actually occurred. Now, witheconomic and nancial conditions on the mend, many analystsare turning their attention to the legacy of these actions.

Foremost among the concerns of many is how to design a strat-egy that does not on the one hand raise interest rates prematurely,thereby prematurely nipping the economic recovery in the bud,while on the other hand does not keep rates too low for too long,thereby creating conditions that lead to a surge in in ation orin ation expectations. Whats needed is an e ective policy toprevent the unprecedented monetary stimulus from becoming adestabilizing in uence on price stability. Another key is accurately predicting in ation over the next few years.

Infation



6/24

Some analysts believe that in ation willremain low as long as the unemploymentrate stays well above its natural rate of unemployment (a measure of slack). Others,by contrast, believe that the risk of higherin ation has risen sharply because of theFeds large-scale asset purchase program andthe advent of large, and possibly protracted,budget de cits.

Recent Policy Actions

In some ways, the 2007-09 recession wasthe most severe since the 1930s. Te latestrecession lasted probably a little less thantwo years, roughly double the length of the

average post-World War II recession (10months). 1 As yet, though, the unemploy-ment rate remains below its post-World WarII peak of 10.8 percent, which was reached inNovember and December 1982.

Given the severity of the latest recession, itwas not surprising that government policy-makers were aggressive and innovative intheir response to it. Figure 1 shows two key measures of the response taken by FederalReserve policymakers during this period.

In Figure 1, the path of the FOMCs

federal funds interest rate target is plot-ted along with the monetary base. Temonetary base, which is sometimes calledhigh-powered money, can be thought of asthe raw material for creating money. 2 Sinceboth series are denominated di erently, thechart indexes the series to be 1.0 in January 2007. Te chart also includes vertical linesat August 2007, March 2008 and Septem-ber 2008, when key events occurred in the

nancial crises.

January 07 July 07 January 08 July 08 January 09 July 09

2.502.252.001.751.50

1.251.000.750.500.250.00

Monetary BaseFederal Funds Target

INDEX, JANUARY 2007 = 1 .0

FIGURE 1

The FOMCS Federal Funds Target Rate and the Monetary Base

no : vertical li e mark ke time i t e a cial cri i : t 2007, Marc 2008 a se tem er 2008.

sou : t r calc lati i fe eral e er e ata.

wo key points are worth noting. First, theFed began reducing its federal funds rate inSeptember 2007, about a month a er condi-tions began to deteriorate in the short-termmoney markets. Although the FOMC con-tinued to reduce its interest rate target beforeand shortly a er the crisis of Bear Stearnsin March 2008, the target then remained onhold from May to September, as rising oil andgasoline prices pushed up headline in ationto levels not seen since early 1991. In Septem-ber 2008, though, economic and nancialconditions deteriorated sharply, causing theFed to quickly reduce its interest rate target tonearly 0 percent (technically, a range from 0to 0.25 percent).

Te second takeaway from Figure 1 isthat the Federal Reserve did not begin toaggressively expand the monetary base untilSeptember 2008.3 Prior to then, the FederalReserve was aggressively lending to domesticand foreign banks and nancial institutions,but at the same time it was countering thisexpansion in bank reserves through o set-ting sales of reasury securities in its portfo-lio. Tis is known as sterilization because itprevents an increase in the monetary base.

Te Feds sterilization e orts ended in Sep-tember 2008, when nancial markets expe-rienced considerable disruption associatedwith the governments takeover of FannieMae and Freddie Mac, the failure of Lehman

Brothers and the near failure of AmericanInternational Group (AIG). At that point,more than any other in the crisis, economicactivity began to decline sharply and rapidly.By August 2009, the monetary base had risento a level that was slightly more than doubleits level in January 2007, while the FOMChad reduced its federal funds target rate by nearly 100 percent. Despite a doubling in thestock of high-powered money, the M2 mea-sure of the money supply increased by only 17 percent over the same period. 4

Te surge in the monetary base has notincreased the money supply to the sameextent both because the demand for loanshas been weak and because some bankshave been reluctant to extend credit. On thedemand side, loan growth has been anemicbecause the demand for credit typically weakens during a recessionespecially during a long and deep recession. On thesupply side, many banks have become morecircumspect in their lending practices in the



7/24

a ermath of the nancial boom and bust.Te latter could also re ect the concernsof bank regulators, who are charged withensuring the safety and soundness of thebanking system, and could stem from bank-ing laws that require banks to meet mini-mum capital requirements.

The Best Way To Forecast Infation?

Figure 1 shows the primary reason why many economists and nancial market par-ticipants worry about the potential for muchhigher in ation rates going forward: Temonetary stimulus will eventually lead to arebound in economic activity and an increasein the demand for bank loans and, thus, fastergrowth of the money supply. As price pres-sures begin to build during the recoveryinpart because rms nd it easier to raise pricesand they must compete for labor, capital andmaterialsin ation and the in ation expecta-tions of rms and households may begin toincrease. Tese in ation expectations may be exacerbated if markets believe that the Fedis not withdrawing the monetary stimulusin a timely fashion, thereby leading to higherfuture in ation rates.

A considerable amount of disagreementseems to exist among economists about thein ation outlook over the next few years.Some economists are quite worried about thepotential for much higher in ation, while

others are more concerned about the poten-tial risk of in ation falling to uncomfortably low levelsor even the possibility of de ation(a fall in the aggregate price level). Much of this disagreement re ects, on the one hand,the Federal Reserves aggressive response tothe deep recession, the nancial crisis andthe exceptionally large federal budget de cits,and on the other hand, the downward pres-sure on wages and prices that typically occursin the a ermath of a deep recession.

Figure 2 depicts one way to gauge this

disagreement. In Figure 2, the history of theBlue Chip forecasts of the average ConsumerPrice Index (CPI) in ation rate over thenext ve years is presented. Te chart showsthe average of the least optimistic in ationforecasts and the most optimistic in ationforecasts, as well as their di erence (disagree-ment). During periods when in ation tendsto be relatively high and variable, such as thelate-1980s and early 1990s, there tend to besome sizable di erences among forecasters

1986 1989 1992 1995 1998 2001 2004 2007

6

5

4

3

2

1

0

Disagreement Top 10 Bottom 10

October 2009

P E R C E N T

about the medium-term in ation outlook.By contrast, during periods when in ationtends to be relatively low and stable, suchas the mid-1990s to mid-2000s, forecasterstend to disagree less about the in ationoutlook. Since early 2007, though, the levelof in ation disagreement among forecastershas increased.

Ultimately, ones view of the in ationoutlook over the next few years depends onones view of how best to forecast in ationover that horizon. Economists use numer-ous methods to forecast in ation. Someeconomists believe that the growth rate of the money supply is an accurate predic-

tor of in ation. According to this view,popularized by monetarists, the in ationrate will ultimately be determined by thegrowth rate of the money supply relative tothe growth rate of real GDP. When money growth exceeds real GDP growthwhatMilton Friedman and others have commonly denoted as too much money chasing too fewgoodsthe in ation rate will increase. oother economists, the in ation process isa random walk, which simply means thattodays in ation will be tomorrows in ation.

Tus, if in ation is 1 percent in 2009, thenthe best forecast for in ation in 2010 is 1 per-cent. Tis view has been shown to producefairly accurate forecasts.5

According to an August 2009 survey,nearly two-thirds of professional forecast-ers surveyed by the Federal Reserve Bank of Philadelphia use some variant of thePhillips Curve to forecast in ation. TePhillips Curve is now o en known as theNew Keynesian model. In this view, todays

in ation rate depends on (i) the in ationrate expected over some horizon and (ii)the amount of slack in the economy. Teamount of slack is also o en measured as thedi erence between actual real GDP and anestimate of potential real GDP; this is termedthe output gap. Tis view also seems to holdsway among several members of the FederalOpen Market Committee.

As discussed by St. Louis Fed PresidentJames Bullard, the New Keynesian modelhas a few well-known problems as it relatesto forecasting in ation. 6 One problemis that the output gap is o en subject toconsiderable measurement error, as well as

being revised o en because of revisions toreal GDP and to estimates of the economysunderlying rate of productivity growth.Te latter a ects estimates of potential realGDP and, thus, the output gap. As a result,policymakers are o en confronted withconsiderable uncertainty about the sizeof the gap as they deliberate the stance of monetary policy.

Many New Keynesian economists assumethat the output gap matters more thanthe expected in ation rate for determin-

ing todays in ation. Tat assumption hasbeen questioned by some economists, whoinstead believe that the publics expectationof future in ation, in part determined by actions of the Federal Reserve, matter morethan the degree of economic slack currently in the economy. 7

Potential Infation Risks

Despite some disagreement about thein ation outlook over the next few years, the

FIGURE 2

Measuring Disagreement among Forecasters about the View o CPI Infation over the Next Five Years

no : di a reeme t i mea re a t e i ere ce et ee t e lea t timi tic reca ter (t 10 a era e) a t e mreca ter ( tt m 10 a era e).

sou : bl e i c mic icat r , ari i e

The Regional Economist | www.stlo is ed.o g


8/24

in ation risks stemming from the govern-ments policy responses to the nancial crisisand the so-called Great Recession will prob-ably not be immediately known becausethe economy is regularly hit by unforeseenshocks (such as large increases in oil prices),foreign economic developments and thelegacy of past policy actions. Still, there areseveral potential risks to the medium-termin ation outlook that can be identi ed. Of course, these risks must be balanced againstthe Federal Reserves commitment tomaintaining a low and stable in ation rate.Indeed, the Fed can help anchor in ationexpectations at a low level both through itswords and deeds.

Is the Output Gap Smaller than We Tink?

It is highly likely that this recession willinduce considerable structural change in theeconomy. Indeed, this development already appears to be in train since many economicresourceslabor and capitalthat wereemployed in the automotive, housing and

nancial industries will need to migrate toindustries that o er higher rates of return.One way to gauge the evolving structuralchange is by viewing the percentage of the labor force that is o en characterizedas the long-term unemployed (personsunemployed for 27 weeks or longer). As of November 2009, this percentage had risen

to 3.8 percent, its highest rate in the post-World War II period.

Tose who believe that the PhillipsCurve framework can adequately capturethe evolution of the in ation outlook overthe near term must adequately account forstructural changes that might have occurredin the boom and bust in asset prices. In its2009 Annual Report, the Bank for Interna-tional Settlements discussed these bubble-induced distortions to current estimatesof trend output growth and, hence, poten-

tial real GDP. Tus, it is conceivable thatestimates of potential real GDP at the startof the recession were too large and that thestructural adjustments noted above may have subsequently reduced potential realGDP from its arti cially high level.

While it is probably unlikely that the fallin actual real GDP during the recession hasbeen matched by the fall in potential realGDP, the size of the output gap might besmaller than conventional wisdom might

believe. If so, those who foresee little risk to the near-term in ation outlook becauseof a large, persistent output gap may betoo optimistic.

Reinfating Asset Prices

Te period following the 2001 recession isan example of how the economy can evolvein ways not readily expected. Recall thatduring the economic recovery following therecession, job growth remained consistently negative until September 2003nearly two years a er the recession ended. At thesame time, the core in ation rate was fallingsharply. From December 2001 to December2003, the year-to-year change in the coreCPI fell from about 2.75 percent to about 1percent. o confront the possibility of therisk of in ation becoming undesirably low,the FOMC announced at the conclusion of itsAug. 12, 2003, meeting that its low-interestrate policy would be maintained for a con-siderable period. In practice, this meant thatthe FOMC maintained its intended federalfunds target rate at 1 percent until the June30, 2004, meeting.

Although it is o en easy to criticize policy a er the fact, some economists subsequently concluded that the extended period of lowinterest rates created a credit boom thatstartedand prolongedsharp increases in

nancial assets and commodity and house

prices that put upward pressure on pricespaid by consumers and businesses. 8 Tesharp increase in oil and commodity priceswas especially acute. Following increasesthat averaged about 2.25 percent from 2001to 2003, the CPI in ation rate averaged 3percent from 2004 to 2007; the run-up inoil prices to more than $130 per barrel thencaused CPI in ation to accelerate sharply,averaging 5 percent over the rst three quar-ters of 2008.

In some respects, the Fed faces a similar

problem today: Policy is extraordinarily accommodative (see Figure 1), and theFOMC has said that economic conditionsare likely to warrant exceptionally low lev-els of the federal funds rate for an extendedperiod. Although low interest rates are akey part of the FOMCs strategy to boosteconomic growth and cement the health of the economic recovery, there might still bea danger of in ating asset prices by encour-aging investors and speculators to shi out

. . . he size he u puap mi h be smalle ha

ve i al wisd m mi hbelieve. s , h se wh

esee li le isk heea - e m i a i u l k

be ause a la e, pe sis eu pu ap may be p imis i .



9/24

E N D N O E S

1 Te National Bureau of Economic Research,which dates business cycle peaks and troughs,usually waits several months a er the apparentend of the recession to declare the date of thetrough.

2 In essence, high-powered money (the sum of bank reserves and currency in circulation) isused to create bank loans, which expand thesupply of money.

3 See Gavin for a detailed discussion of changes inthe monetary base during this period.

4 Broadly, M2 is the sum of currency, checkabledeposits, savings and small-time deposits, andretail money market funds. For a description andde nition of the monetary and nancial termsused throughout this article, see http://research.stlouisfed.org/publications/mt/notes.pdf.

5 See Atkeson and Ohanian.6 See Bullards presentation at http://research.

stlouisfed.org/econ/bullard/BullardNABE-FinalOct112009.pdf.

7 See Piger and Rasche.8 See aylor and Frankel.9 See Congressional Budget O ce.

10 Te Blue Chip Survey asked forecasters to gaugetheir risk of sharply higher in ation on a scaleof one to ve, with one being no risk and vesignaling great r isk.

11 SeeUnited States Financial Data to view updatedcharts of the asset and liability side of the Fedsbalance sheet. Tese charts can be accessed athttp://research.stlouisfed.org/publications/usfd/page7.pdf.

12 See Bernanke.

R E F E R E N C E S

Atkeson, Andrew; and Ohanian, Lee E. ArePhillips Curves Useful for Forecasting In a-tion? Federal Reserve Bank of MinneapolisQuarterly Review, Winter 2001, Vol. 25, No. 1,pp. 2-11.

Bank for International Settlements. 79thAnnual Report: 1 April 2008-31 March 2009.Basel, Switzerland, June 29, 2009.

Bernanke, Ben. Te Federal Reserves BalanceSheet: An Update. At the Federal ReserveBoards Conference on Key Developments inMonetary Policy, Washington, D.C., Oct. 8,2009.

Congressiona l Budget O ce. Te Budget andEconomic Outlook: An Update. Te Con-gress of the United States, August 2009.

Frankel, Je rey. Comment: Real Rates Key toCommodities Prices, Reuters.com, March2008. See www.reuters.com/article/reutersComService4/idUSDIS96078820080319?pageNumber=1&virtualBrandChannel=0&sp=true.

Gavin, William . More Money: Understand-ing Recent Changes in the Monetary Base.Federal Reserve Bank of St. Louis Review,March/April 2009, Vol. 91, No. 2, pp. 49-59.

Piger, Jeremy M.; and Rasche, Robert H. In a-tion: Do Expectations rump the Gap?International Journal of Central Banking ,December 2008, Vol. 5, No. 3, pp. 85-116.

aylor, John B. Getting O rack: How Gov-ernment Actions and Interventions Caused,Prolonged, and Worsened the FinancialCrisis. Stanford, Calif.: Hoover InstitutionPress, Stanford University, 2009.

of low-yield assets like reasury securitiesinto higher-yielding assets like commodity contracts or other tangible nancial assets.

Te Exploding Federal Budget De cit

From scal year 2002 to 2008, the U.S.federal budget de cit averaged about $305billion per year, or 2.5 percent of GDP. In

scal year 2009, though, the federal de cittotaled about $1.5 trillion, or roughly 11.25percent of GDP, according to estimates by theU.S. Congressional Budget O ce (CBO). Telarge increase in the de cit re ected legisla-tive policy actions such as the AmericanRecovery and Reinvestment Act of 2009 ( s-cal stimulus) and the roubled Asset Relief Program ( ARP), as well as an increase inmandatory government outlays associatedwith the deep recession. Te CBO projectsthat the federal budget de cit will total nearly $1.4 trillion in scal year 2010 and nearly $925 billion in scal year 2011.9

Gauging the de cits potential e ect onin ation depends on how it is nanced.o see this, consider the governments

budget constraint. In its simplest form,the constraint stipulates that if the de citis not nanced by higher taxes, it must be

nanced in one of two ways: (i) by issuingdebt to the public, which includes foreignholders of U.S. reasury securities; or (ii) by selling government debt to the central bank,

which is the Federal Reserve. Te latter, alsocalled monetization of the debt, increasesthe monetary base (high-powered money)and, thus, the money supply. For example,the Federal Reserve announced March 18,2009, that it would buy up to $300 billionof reasury securities (beyond its existingholdings at the time). Tese purchases,which were designed to help improveconditions in private credit markets, werenot sterilizedthat is, they were allowed toincrease total bank reserves and, thus, the

monetary base.Many economists appear to be concerned

about the in ationary implications of thehuge increase in government de cit spendingthat is unfolding. According to a survey pub-lished in the June 2009 Blue Chip EconomicIndicators, about 42 percent of forecasters seea relatively high risk that U.S. in ation willrise sharply within the next ve years becauseof the governments and the Feds response tothe nancial crisis and recession; another 34

percent see little or no risk; the remainder seeonly a moderate risk. 10

The Feds Strategy

A key di erence between the 2003-04episodewhen the Fed held its federalfunds interest rate target at 1 percent fromJune 2003 to June 2004and today is thatthe FOMC has used innovative measuresto dramatically expand the size of its bal-ance sheet.11 Because this expansion in themonetary base has the potential to greatly expand the nations money supply wheneconomic activity rebounds, policymak-ers are, thus, confronted with the potentialproblem of designing an e ective policy toreduce the size of the Feds balance sheetto prevent a rapid acceleration in money growth that may destabilize in ationexpectations. Improving economic and

nancial conditions have lessened the use of the Feds special lending facilities; so, someportion of these excess reserves will naturally contract on their own. Still, this processwill not be su cient to prevent a potentially destabilizing surge in money growth, whichmeans that Fed policymakers will have toadopt other, more aggressive strategies. Teo cials have discussed several methods of doing this, including paying interest on bank reserves, using conventional open marketoperations and selling outright some of the

securities and other assets held on the Fedsbalance sheet.12 Regardless of the methodused, an improving economy means that theFed must be prepared to raise its interest ratetarget to prevent an unwanted expansion inmoney growth by the banking sector.

Fed Chairman Ben Bernanke and othersenior Fed o cials are quite con dent thatthey have the tools and the determinationnecessary to prevent an unwelcome accel-eration in in ation or in ation expecta-tions. Unlike previous episodes, though,

the magnitude of the policy responses tothe nancial crisis and the Great Recessionsuggests that the FOMCs margin of errorseems much smaller than at any time in theFeds history.

Kevin L. Kliesen is an economist at the Federal Reserve Bank of St. Louis. For more on his work,see http://research.stlouisfed.org/econ/kliesen.Douglas C. Smith provided research assistance.



10/24

he U.S. economy appears to be on theroad to recovery following the deepest

and longest recession in the post-World WarII period. Despite this improvement, someanalysts and policymakers are increasingly concerned about deteriorating conditionsin the commercial real estate (CRE) sector.Defaults on CRE loans have contributed tothe recent upsurge in bank failures and asharp increase in nonperforming loans of banks. Te Federal Open Market Com-mittee noted at its Sept. 23, 2009, meetingthat many regional and small banks were vulnerable to the deteriorating performanceof commercial real estate loans. 1

How large is the commercial real estateexposure of banks, and what is the likeli-hood that problems in this sector will be

severe enough to derail the U.S. economicrecovery?

CRE Exposure

Te CRE sector has faced sharp contrac-tion over the past year, paralleling the bustthat unfolded in the housing sector twoyears ago. For example, the value of privatecommercial and o ce (herea er commer-cial) construction totaled a little less than$140 billion in 2008, about unchanged fromthe previous year. 2 By September 2009,

the nominal value of commercial construc-tion had declined by roughly 35 percent to$90.2 billion.

Like residential housing, commercialconstruction depends heavily on mortgage

nancingeither directly from commercialbanks and thri s, indirectly throughinvestors in commercial mortgage-backedsecurities (CMBS) or through other con-duits, such as private equity funds or lifeinsurance companies. As of June 30, 2009,

the size of the outstandingdebt associated with the commercialreal estate sector was $3.5 trillion. 3 Abouthalf of this amount ($1.7 trillion) was heldby banks and thri s. Of the rest, half washeld as collateral for CMBS, and the otherhalf was held by investors.

When analyzing commercial banksexposure to the downturn in the CREmarket, it is helpful to rst consider howthe valuation of these loans can changeover time. Like any asset, the market valueof a commercial property depends on itsexpected rate of return over time. Tisreturn depends on both macroeconomicfactors, such as the health of the economy (both nationally and locally), expected in a-tion and the market interest rate over the

life of the loan. But the return also dependson microeconomic factors, such as vacancy rates, property taxes, land use regulationsand the price of land.

As economic conditions deterioratedduring the latest recession, the CRE marketa ected commercial banks, investors andother nancial institutions in a couple of key ways. First, sales at businesses slowedsharply and then began to decline, causingsome rms to go out of business, vacancy rates to rise and property prices (and rents)

to fall. Te downturn in commercial prop-erty prices during this cycle was particu-larly severe. By one measure, commercialproperty prices have declined by nearly 41percent since their peak in October 2007. 4

As CRE mortgage defaults and delin-quency rates increased, banks naturally increased the level of their loan lossreserves, which adversely a ected theirearnings. Moreover, in this downturn,larger banks were also a ected by a second

factorthe valuations of com-mercial mortgage-backed securities that areheld on their balance sheets. As the valueof the collateral that determines the priceof the CMBS declined, commercial bankswere forced to mark down the value of theseassets on their balance sheets. 5 o compen-sate, banks were forced to raise additionalcapital or suspend dividends.

Now, as the housing market appears to bestabilizing, the quality of banks CRE loansis deteriorating. In the past year, averagenonperforming CRE loans (loans that are 90or more days past due or loans not accruinginterest) as a percentage of risk-based capitalhas grown considerably, from 4.47 percentin September 2008 to 7.4 percent in Sep-tember 2009. Within the banking industry,

community banks (banks with assets lessthan $1 billion) have 30.7 percent of theirloans in CRE compared with 12.1 percentfor the largest banks (banks with assetsgreater than $100 billion).

Although community banks are exposedto challenges in the CRE property markets,the accompanying chart indicates thatcommunity banks in the Eighth FederalReserve District have relatively lower levelsof CRE exposure than their national peersdo. By nature of their business model,

banks operate with relatively low levels of capital. As such, CRE loans o en representa multiple of capital. For Eighth Districtcommunity banks, CRE comprises roughly 167 percent of risk-based capital, as opposedto 201 percent for peer banks. In addition,nonperforming CRE makes up roughly 47 percent of Eighth District community banks nonperforming portfolio, while itmakes up nearly 56 percent of nonperform-ing loans for peer banks. Most important,

r c o n n r c o V r Y

Commercial Real Estate:A Drag for Some Banksbut Maybe Not for U.S. Economy By Rajeev Bhaskar, Yadav Gopalanand Kevin L. Kliesen



11/24

E N D N O E S

1 For minutes of the Sept. 23, 2009, FOMCmeeting, see www.federalreserve.gov/mon-etarypolicy/fomcminutes20090923.htm.

2 Private nonresidential construction totaledabout $416 billion in 2008.

3 Te commercial real estate total cited hereincludes outstanding debt on multifamily residential mortgages. In comparison, theoutstanding debt associated with the residen-tial sector totaled about $11 tril lion. Tesedata are published in the Federal ReservesFlow of Funds report (Z.9, able L.217).

4 See the Moodys/REAL Commercial Property Index at http://web.mit.edu/cre/research/credl/rca.html.

5 In general, community banks and reg ionalbanks have little or no exposure to CMBScompared with large banks exposure.

6 See FDIC.7 See American Economic Review.8 See Greenlee.

R E F E R E N C E S

American Economic Review. May 1993. Vol.83, No. 2. Tree papers on the recession of 1990-91, originally presented at an AmericanEconomic Association conference in 1992.

Greenlee , Jon D. estimony on residentia l andcommercial real estate before the Subcom-mittee on Domestic Policy, Committee onOversight and Government Reform, U.S.House of Representatives, in Atlanta, Ga., onNov. 2, 2009.

Federal Deposit Insurance Corp. History of theEightiesLessons for the Future. Vol 1: AnExamination of the Banking Crises of the 1980and Early 1990s. See especially Chapter 3.FDIC: 1997. Deposit insurance losses comefrom FDICs Historical Statistics on Banking;see www2.fdic.gov/hsob/. Deposit insuranceloss gure in September 2009 real dollars.

nonperforming CRE loans make up only 4percent of Eighth District banks risk-basedcapital, meaning that, as a group, these

banks have an adequate bu er to handleCRE-related losses.

Will CRE Derail the Economy?

Te collapse in the CRE market duringthe late 1980s and early 1990s o ers someguidance about the potential e ects on theU.S. macroeconomy stemming from thecurrent deterioration in CRE loan per-formance. During the economic boomthat followed the deep recession in theearly 1980s, many banking organizations

weakened underwriting standards on CREloans. By the late 1980s, the CRE marketwas experiencing tremendous stress, leadingto a collapse in CRE market activity. Tecollapse in the CRE market caused con-siderable turmoil in the banking industry,leading to tremendous losses and a largenumber of bank failures. From 1987 to1992, a little more than 1,900 banks andthri s failed, which cost the FDIC depositinsurance funds roughly $386 billion inreal terms. 6 And yet, while the economy

experienced a recession from July 1990 toMarch 1991, its not entirely clear that theCRE crisis was the major factor that causedthe recession. However, the CRE collapseand its e ect on construction activity andbank performance probably contributed tothe relatively weak recovery. 7

oday, similar concerns are being raisedabout the weakness in CRE. As the econ-omy transitions from recession to recovery,the number of bank failures is rising: From

January 2009 through early December, 130commercial banks and thri s failed, thelargest number since 1992. What is not yet

known at this point, though, is whetherthe likelihood of further CRE losses willthreaten to further weaken the bankingsystem, which is beginning to recover fromthe housing bust and the nancial crisis.According to one estimate, almost $500billion of CRE loans will be maturing overthe next few years, a potentially signi cantdefault risk if these loans are not able to bere nanced. 8 Despite these di culties, mostforecasters continue to see steady improve-ment in economic growth, rising employ-

ment and relatively low in ation in 2010and 2011.

CRE loans may pose a signi cant risk forcommunity banks in the year ahead. Justas in the late 1980s and early 1990s, it ispossible that todays commercial real estateproblems will produce adverse economicoutcomes. However, the impact is mostlikely to be seen at the local level than at thenational level.

Kevin Kliesen is an economist at the Federal Reserve Bank of St. Louis. For more on hiswork, see http://research.stlouisfed.org/econ/ kliesen/. Rajeev Bhaskar and Yadav Gopalanare research associates there. For more onGopalans work, see http://www.stlouisfed.org/ banking/pdf/SPA/Yadav_vitae.pdf.

250

100

150

200

50

0

P

E R C E N T

CRE as a Percentageof Risk-Based Capital

Nonperforming CRE as a Percentageof Total Nonperforming Loans

Nonperforming CRE toTotal Risk-Based Capital

Q3 2009 Da ta

U.S. Peer Community BanksEighth District Community Banks

Community banks are defined as those commercial bankswith $1 billion in total assets or less.

167.1

200.8

46.5 56.4

3.9 6.5

Eighth District Community Bank CRE Exposure versus National Peers

sou : li ate e rt iti a c me r ba k ( all e rt )



12/24

It is almost a given that people with moreeducation make more money than thosewith less education. But how much more isthat better education worth? Te answer ismore complicated than many would think.

Te di erential between the price of highly and poorly educated labor is given the suit-

ably evocative label the return to education.Te reference to the price di erential as areturn stems, of course, from an under-standing that education is a choice; individu-als can place themselves in a position to selltheir labor services at the higher price by investing in their human capital.

Te relationship between education andearnings is among the most widely studiedtopics in labor economics. One importantgoal is to uncover the causal impact of educa-tion on earnings. Just because a person with

a college degree earns more than a personwithout such a degree does not necessarily mean that college education causes the dif-ference in pay. Rather, the person who wentto college might have some characteristicsthat make him or her more productive in the

labor market, result-ing in higher earnings.It is possible, for example, thathigh-ability people are more likely to go tocollege and are more productive.

So, how can the e ect of a college educa-tion on earnings be isolated? In an ideal

world, researchers would make a copy of aperson, sending only one of the two to col-lege. A er the one graduates from college,earnings of the two would be compared.Only in this case could it be said with somecertainty that the di erence in earnings wasdue to the college education. 1 Of course, thissort of comparison is not feasible. Instead,researchers try to compare people who are assimilar as possible in everything but the levelof education they have.

Studies usually try to control for demo-

graphic factors, such as age, gender andrace, as well as work experience. Otherfactors that might a ect the return to educa-tion are family background, school quality and ability. Quantifying any of these factorsis a di cult task in itself. Researchers use,

for exam-ple, IQ or

aptitude test scoresas a measure of ability;

parental education is used as a measure of family background.

With so many factors to consider, studies

take di erent approaches and use di erentestimation techniques. Although all studies

nd that more education is associated withhigher earnings, the estimates of the returnto education vary. Most studies estimatethat the return to one year of schooling is, onaverage, between 8 and 13 percent. 2 In otherwords, each additional year of education isassociated with an 8-13 percent increase inhourly earnings. For practical applications,10 percent, on average, is a good estimate of the return. (It is worth pointing out that the

returns are somewhat higher for women thanfor men.)

Additional Complications

Complicating these estimates is the factthat any returns on investment in human

c o L L g

By Natalia Kolesnikova

dividuals du a i Be ef s o he s, t

Te Return to Education

Isnt Calculated Easily

Positive spillovers from education have beenfound in areas other than labor markets, too.One study has shown that higher maternaleducation improves infant health, as measuredby birth weight and gestational age. It alsoincreases the probability that a new mother ismarried, reduces parity, increases use of pre-natal care and reduces smoking, suggestingthat these are important pathways for theultimate effect on health. 8

Another study found a signi cant decreasein probability of criminal behavior and incar-ceration for people with more education. 9 The researchers noted that the externality of

Estimates of the private returns to educa-tion do not account for all the bene ts

that society receives from an individualsinvestment in education. Economic theorypredicts that an individuals education notonly boosts his or her own productivity butalso that of others. The presence of moreeducated workers leads to a knowledgespillover, making other workers moreproductive. Some recent studies havefound empirical evidence in support of thisprediction. 6

Productivity spillovers also have a positiveeffect on wages. For example, a percentagepoint increase in the supply of college gradu-ates raises high school dropouts wagesby 1.9 percent, high school graduateswages by 1.6 percent and college graduateswages by 0.4 percent, according to onestudy. 7 Not surprisingly, there is also a posi-tive impact of education on economic growthas people with more education were shownto be more likely to accept innovation andadopt new technologies.



13/24

E N D N O E S

1 Even in this case, it is not clear if the labormarket rewards skills a person learned incollege or simply reacts to a signal of higherabilities. In fact, some researchers argue thatthere is a sheepskin e ect in which diplomasand degrees matter more than actual numberof years of education. See Hungerford andSolon, as well as Belman and Heywood, formore.

2 Card provides an excellent overview of exist-ing studies.

3 See Black, Kolesnikova and aylor. 4 Te reported numbers represent an increase

in hourly earnings f rom obtaining collegeeducation (relative to having only a highschool diploma), rather than a return toone year of schooling as before.

5 See Oreopoulos and Salvanes. 6 In particular, see the work of Acemoglu and

Angrist and that of Moretti (2004a,b). 7 See Moretti (2004b). 8 See Currie and Moretti. 9 See Lochner and Moretti.

R E F E R E N C E S

Acemoglu, Daron; and Angri st, Joshua. HowLarge are Human-Capital Externalities?Evidence from Compulsory Schooling Laws.NBER Macroeconomics Annual , Vol. 15,pp. 9-59. New York: National Bureau of Economic Research, 20 00.

Belman, Dale; and Heywood, John. SheepskinE ects in the Returns to Education: AnExamination of Women and Minorities.Te Review of Economics and Statistics , 1991,Vol. 73, No. 4, pp. 720-24.

Black, Dan; Kolesnikova, Natalia; and aylor,Lowell. Earnings Functions When Wagesand Prices Vary by Location. Journal of LaborEconomics, 2009, Vol. 27, No. 1, pp. 21-47.

Card, David. Te Causal E ect of Educationon Earnings, in Orley Ashenfelter and DavidCard, eds., Handbook of Labor Economics,Vol. 3A, pp. 1801-63. Amsterda m: ElsevierScience, 1999.

Currie , Janet; and Moretti, Enrico. MothersEducation and the Intergenerational rans-mission of Human Capital: Evidence fromCollege Openings. Quarterly Journal of Economics, 2003, Vol. 118, No. 4, pp. 1495-532

Hungerford, Tomas; and Solon, Gary. Sheep-skin E ects in the Returns to Education.Te Review of Economics and Statistics , 1987,Vol. 69, No. 1, pp. 175-77.

Lochner, Lance; and Moretti, Enrico. Te E ectof Education on Crime: Evidence from PrisonInmates, Arrests and Self-Reports. Te American Economic Review, 2004, Vol. 94,No. 1, pp. 155-89.

Moretti, Enrico. Workers Education,Spillovers and Productivity: Evidence fromPlant-Level Production Functions. Te American Economic Review, 2004a, Vol. 94,No. 3, pp. 656-90.

_____. Esti mating the Social Return to HigherEducation: Evidence from Longitudinal andRepeated Cross-Sectional Data. Journal of Econometrics, 2004b, Vol. 121, No. 1-2,pp. 175-212.

Oreopoulos, Philip; and Salvanes, Kjell. HowLarge Are Returns to Schooling? Hint: MoneyIsnt Everyt hing. National Bureau of Eco-nomic Research Working Paper 15339, 2009.

capital must be realized in a speci c labormarketusually a local labor market. Teseeducational investments arent like invest-ments in stock, where a share of GeneralElectric is worth the same in New York as it isin St. Louis.

One study, conducted in part by thisauthor, found that the returns to college edu-cation are systematically lower in nicer, moreexpensive cities.3 It is not surprising thatwhen a city has attractive amenities peoplepay for these amenities in the form of highproperty prices. However, people with lowlevels of education, and therefore low lifetimeincome, nd these cities high property pricesto be a greater deterrent than do individualswith high levels of education. Well-educatedpeople (cardiologists, for example) mighteven accept a lower salary to work in thesecities than they would in less-attractive cities.On the other end of the scale, less-educatedpeople (janitors, for example) might have tobe paid more to work in these nice cities thanelsewhere because of the higher cost of living.Terefore, the discrepancy in pay betweenthose with more education and less educationis smaller than elsewhere. It is importantto point out that even though measuredmonetary returns to education are lower inmore-attractive cities, cardiologists are not atany disadvantage when they choose to locatethere. Tey are simply paying for an access

to amenities of a nice city by accepting lowerreturns to their education.

Te study also estimated the returns tocollege education for white men living inmajor U.S. cities. In 2000, a white man witha college degree earned as much as 85 percent

more than a similar white man with a highschool diploma in Dallas, but only 50 percentmore in Seattle (but he enjoyed all the goodthings that Seattle has to o er). 4 Te cross-city di erences in the returns to collegeeducation are even bigger if smaller cities areconsidered as well.

Nonmonetary Returns

Although it is di cult to determine themonetary return to education, it is practically impossible to quantify the numerous non-monetary returns. Studies have shown thatexperiences and skills acquired in schoolreverberate throughout life, not just throughhigher earnings. Schooling also a ects thedegree one enjoys work and the likelihoodof being unemployed. It leads individuals tomake better decisions about health, marriageand parenting. It also improves patience,making individuals more goal-orientedand less likely to engage in risky behavior.Schooling improves trust and social interac-tion, and may o er substantial consumption value to some students. 5

Despite the di culty in assessing thereturns to education, there is little doubtthat the importance of education will notdisappear from the public policy arena. As aresult, continued economic research on thesubject will hopefully guide e ective publicpolicy.

Natalia Kolesnikova is an economist at theFederal Reserve Bank of St. Louis. For more onher work, see http://research.stlouisfed.org/econ/ kolesnikova/index.html.

$

education is about 14-26 percent of theprivate return to schooling, suggestingthat a signi cant part of the socialreturn to education comes in theform of externalities from crimereduction.

Research done to evaluatethe social returns to educa-tion is extremely important fora variety of policy questions,such as assessing the ef ciencyof public investment in education.The issue remains one of thefrontiers of labor economics.

A percentage point increasein the number of

college grads raises:



14/24

At the heart of economic research is howeconomic policy a ects personal well-being. raditional economic measures of well-being, such as income per capita, assumethat individuals are well o to the extentthat they can satisfy their wants and needs.Under this assumption, income is generally regarded as a useful proxy for well-beingbecause greater income allows for greaterconsumption. 1 However, some critics pointout that income does not fully capture theconcept of well-being.

Nonmonetary factors, such as health,family and friends, also play a large role indetermining individual welfare. Starting inthe early 1970s, economists began studyingbroader notions of well-being by analyzingsurvey data that provide subjective individual assessments of happiness in place of

conventional income or consumption-basedmeasures of well-being. Subjective individualassessments of happiness have been used, forexample, to study the link between incomeand well-being and to study the welfaree ects of economic variables such as in a-tion and unemployment.

The Easterlin Paradox

Richard Easterlin was the rst moderneconomist to examine the link betweenindividual assessments of happiness and

income. His 1974 study uncovered a puzzlethat sparked further economic research onthe link between income and well-being.Using happiness surveys from 19 countries,Easterlin observed that, within countries,an individuals income level closely matchedself-reported happiness. Across countriesand over time, however, there was little to norelationship between income per capita andaverage happiness. Additionally, Easterlinfound that happiness in the United States had

remained stagnant despitelarge increases in averagereal personal income. Tis pattern, in whichwealthier individuals report greater happinessat any given time but average happiness doesnot increase with average income over time,is o en called the Easterlin paradox.

Te gure shows that this puzzling obser- vation persisted in the United States from1972 to 2008. Real income per capita almostdoubled over the period, while average hap-pinessas reported by respondents to theGeneral Social Surveychanged very little.2

One of the most accepted explanations forthis apparent puzzle is that individuals happi-ness is not determined by their absolute levelof income but by how their income compareswith the income of others. According to thisexplanation, societies fail to get happier with

economic progress because as economic con-ditions advance and average incomes rise, thereference standard that individuals use to judgetheir situation relative to others also rises.

At the time of Easterlins study, the analysisof reported happiness was limited to devel-oped countries because survey data from low-income countries was not available. Morerecently, the accumulation of reported hap-piness data across a lengthier time span andfor a broader array of countries has allowedeconomists to more closely examine the link

between income and well-being. One suchstudy, conducted by economists Betsey Ste- venson and Justin Wolfers, used data fromseveral surveysmost notably the GallupWorld Pollto investigate more countriesthan the original Easterlin study did. Steven-son and Wolfers, in contrast with Easterlin,found a positive association between incomeand average reported happiness across coun-tries. Tey wrote that the correlation is similarto the one found within countries between

personal income andindividual happiness reportsthat is,wealthy countries report higher averagelevels of happiness than poor countries. Teauthors also found that in several countrieswhere time series were available, peopletend to report being happier as countriesget richer, although the correlation is not asstrong. (Te United States, as noted in theEasterlin survey, remains a notable excep-tion.) Teir ndings suggest that relativeincome plays a smaller role and that absoluteincome plays a larger role in shaping happi-ness than previously thought.

The E ects o Infationand Unemployment

In contrast to policy research in othersocial sciences, economists traditionally

have been reluctant to use self-reports of well-being because of the subjective nature of those reports. Instead, economists prefer toinfer individual preferences from observedconsumption patternsan approach knownas the revealed-preference principle. How-ever, in some situations where revealed pref-erences are unable to fully assess the welfareimpact of policies or institutional features of an economy, self-reports of happiness may bea useful tool in evaluating economic policy.One particular policy issue on which subjec-

tive reports of happiness have been used toshed light is the trade-o between in ationand unemployment in terms of personalwell-being.

Economists Rafael Di ella, Robert Mac-Culloch and Andrew Oswald examinedreported happiness data from the UnitedStates and Europe and found that in ationand unemployment both reduce happiness,but unemployment costs more than in ationin terms of happiness. What is notable about

W L L - B n g

By Rubn Hernndez-Murillo and Christopher J. Martinek

Te Dismal Science

ackles Happiness Data



15/24

E N D N O E S

1 In standard utility theory, utility is derivedfrom the consumption of goods and services.

2 Te General Social Survey is a national survey sponsored by the National Opinion ResearchCenter at the University of Chicago. Te survey gathers socio-demographic characteristicsof respondents and polls them on a variety of social issues.

R E F E R E N C E S

Di ella, Rafael; MacCulloch, Robert J.; andOswald, Andrew J. Pre ferences over In a-tion and Unemployment: Evidence fromSurveys of Happiness. American EconomicReview, 2001, Vol. 91, No. 1, pp. 335-41.

Di ella, Rafael; and MacCulloch, Robert J.Some Uses of Happiness Data in Economics. Journal of Economic Perspectives, 2006,Vol. 20, No. 1, pp. 25-46.

Esterlin, Richard. Does Economic GrowthImprove the Human Lot? Some Empirica lEvidence, in Paul A. David and M.W. Reder,eds., Nations and Households in EconomicGrowth: Essays in Honour of Moses Abramovitz . L ondon: Academic Press,1974, pp. 98-125.

Gandelman, Nstor; a nd Hernndez-Murillo,Rubn. Te Impact of In ation and Unem-ployment on Subjective Personal and Country Evaluations. Federal Reserve Bank of St. LouisReview,2009, Vol. 91, No. 3, pp. 107-26.

Stevenson, Betsey; and Wolfers, Justin.Economic Growth and Subjective Well-Being: Reassessing the Easterlin Paradox.Brookings Paper on Economic Activity, 2008,No. 2, pp. 1-87.

their research is that the authors account

not only for the cost of unemployment onaverage well-beingwhat they call fearof unemploymentbut also for the directcost of individuals who actually becomeunemployed. According to these economistsestimates, individuals would prefer to take ona 1.66-percentage-point increase in in ationrather than a one-percentage-point increasein unemployment.

Additional Studies o Happiness Data

A recent study by economists Nstor

Gandelman and Rubn Hernndez-Murillothat used data for 75 countries from theGallup World Poll took a novel approach inanalyzing reported happiness. Te authorsused the responses to several unique survey questions in the Gallup World Poll to con-struct measures of well-being that are morecomprehensive.

In the Gallup survey, respondents wereasked to provide a personal assessment of their own happiness as well as a personalassessment of their countrys well-being as a

whole. Te survey also asked respondents toevaluate not only current individual happi-ness and country well-being, but also assess-ments of happiness and national well-being

ve years ago and their expectations veyears from now.

Gandelman and Hernndez-Murillo usedthe responses to these questions to con-struct measures of past, present and futurepersonal and country well-being. Teir study revealed two interesting details in happiness

data. First, individuals tend to evaluate their

personal well-being as being better than theircountrys. Second, individuals tend to expectthat their future well-being will improve.

Although Gandelman and Hernndez-Murillo did not nd any signi cant dif-ferences in the e ects of in ation andunemployment on reported happiness, they found that both in ation and unemploymentnegatively a ect past and present personalevaluations of individual and country well-being and also evaluations of present well-being relative to the future.

Comments

Research into the economics of happinesshas come a long way since Easterlins study and has gained increasing acceptance amongmainstream economists as a complement tostandard utility theory. Easterlins paradoxremains a controversial and unresolved issue,but the analysis of subjective well-being datacontinues to spread into various problemstraditionally studied in economics, sheddingnew light on such issues as the determinationof labor supply, the e ects of taxation anddemocracy, and the degree of risk-aversionin individual preferences and its impact onsavings behavior.

Rubn Hernndez-Murillo is an economist at the Federal Reserve Bank of St. Louis. Christo- pher J. Martinek is a research associate there.For more on Hernndez-Murillos work, seehttp://research.stlouisfed.org/econ/hernandez.

Happiness and Real Income per Capita in the United States

no : Mea a i e (le t cale) i t e a era e re l r m re e t t t e u.s. ge eral s cial s r e .e r e q e ti a k : ake all t et er, l a t i are t e e a ? w l a t at

are t t a , rett a r er a ? e e al e ere c e a 1, 2 a 3, re ecti el .

sou : ge eral s cial s r e ata a aila le at tt :// . rc. r /gss+we ite.

eal i c me er ca ita a e a t r calc lati i ata r m t e b rea c mic al ia t e e b rea .

197619741972 1978 1980 1982 1984 1986 1990 1992 1994 1996 1998 2000 2002 2004 2006 20081988

Mean Happiness

Real Income per Capita

350003

2

1

30000

25000

20000

15000

10000

H A P P I N E S S

S C A L E

C H A I N E D

2 0 0 5

$



16/24

he bad news came in Corey J. Meha ysrst week as president of the Moberly

Area Economic Development Corp. DuraAutomotive Systems announced the layo of 200 workers at its auto parts plant inMoberly, Mo. As part of its parent compa-nys bankruptcy reorganization, a bulwark of the local economy was losing half of itswork force to Mexico.

Tough the February 2007 decisionwas irrevocable and he was new to his job,Meha y was on the spot to try to makethe best of the companys situation, recallsRegina Reid, who was then the human

resources manager at the Dura plant inMoberly. He worked with the company and other local partners to help the excessemployees get skills assessment, job searchand retrainingfederal bene ts due them asautoworkers whose jobs le the country.

Tis sort of service has become Meha ysmodus operandi: to deliberately spend moretime working to retain local businessesand, if possible, help them growthan try-ing to woo new employers from out of town.

We do formal visits with all of our largeemployers in the area, usually three times

a year, sometimes more, he says. We sitdown with them and ask how theyre doing,if theres anything that we can do for them,if theyre having any issues. We talk abouttraining opportunities and work forcetrends.

Meha y recalls one painful rebu . Car-rollton Specialty Products, which madeand packaged Hallmark cards, spurned hisrepeated o ers to approach Hallmark a erthe card company decided to outsourcethe work of Carrolltons Moberly plant. It

closed in early 2009, with 200 jobs dis-patched to China.In contrast, Orscheln Products welcomed

Meha ys services. Te maker of vehiclecontrols announced in July that it wouldclose a plant in Ohio and move produc-tion to its Moberly plant. President RobertOrscheln credited Meha y for helping tosecure the incentives that made the movefeasible. Tese included $916,718 in state taxcredits, plus exemptions for state sales taxes

Article and photos by Susan C. Tomson

Mobe l , Mo., by he umbe sP Pu T N

City of Moberly .................... ..................... . 14,227

Randolph Co nty ..................... .................. 25,723

B R F RCERandolph Co nty ..................... .................. 12,927

uNEMP MENT R TE ..............................9.0 percent

PER C P T PERS N NC ME

Randolph Co nty ..................... ................ $25,230

* u.S. B rea of the Cens s, estimate J ly 1, 2008** H VER (B S), ctober 2009

*** BE /H VER, 2007

T P EMPL yErS

rscheln nd stries .................. .................... .......... 725

Moberly Correctional Center .................... ............. 520Moberly Regional Medical Center .................... ...... 410

Wal-Mart Perishable Food Distrib tion Center ...... 400

Moberly P blic Schools ................... .................... ... 365

ncl des rscheln Prod cts, rscheln Farm and Home, areal estate development and property management nitand the gro ps headq arters.

S uRCES: Self-reported except Moberly rea EconomicDevelopment Corp. for Wal-Mart.

ncl des part-timers

*

*

**

**

***

c o M M n t Y P r o F L

bu re e i mphasis Pays o Miss u i tt not M g c,



17/24

on energy and new equipment purchases,all in exchange for the companys pledge tocreate 100 jobs over three years.

Te company is part of Orscheln Indus-tries, which includes a real estate devel-opment and management unit and 150Orscheln Farm and Home stores in nineMidwestern states. All are based in Moberly and overseen by Roberts older brother, Barry.

Te brothers are third-generationMoberly businessmen, whose paternalgrandfather and his four brothers began

branching out from farming into otherenterprises in the 1920s. In their day,Moberly remained somewhat the magiccity, a sobriquet from years of explosivegrowth as a railroad hub halfway betweenSt. Louis and Kansas City. Rob Cater,president of the Bank of Cairo & Moberly,says the citys good times rolled throughthe 1950s, when railroading declined andmanufacturing began taking up the eco-nomic slack.

Manufacturing hit a bottom in the 1980s,

when DuPont, oastmaster and Wick Homes closed their Moberly plants, put-ting hundreds out of work, recalls MichaelBarner, director of the Moberly Area

echnical Center, a technical high schoolfunded by area school districts and housedat Moberly Public Schools. A good portionof the middle class was knocked out and hadto relocate, he says.

Te citys economic history from 1992 isrepresented by 22 shovels decking two walls

Top: t or c el pr ct , a nicker c ecle t arki rake ca le .

Bottom: e M erl rea mm it lle eer rk-relate certi cate a a ciate

r ram , t ic el t kee t e rki t e area t ate a em l a le. i

e t e t i t e el i a metal tec le ree r ram ker a r ject.

Corey Mehaffy, re i e t t e M erl rea c micde el me t r ., t m re time i t i e rete -ti t a e e recr itme t, a t e trate a ai

. be i im i e t e 22 el t e all t e c ere ce r m ere e rk , eac re re e tia r reaki i ce 1992.

of the development corporations conferenceroom. Each shovel represents a ground-breaking, ve of them for distributioncenters, one since closed.

Including the four that remain and onefrom before 1992, Moberly counts vedistribution centers today, for ScholasticBook Clubs, Goodyear Engineered Products,Orscheln Farm and Home stores, Mid-AmBuilding Supply and Wal-Mart, the newestand largest. By the development corpora-tions count, their combined work force of some 1,100 people makes distribution thecitys second-largest employment sector a ermanufacturing, which employs about 1,400.

Cater says the distribution centers havediversi ed, stabilized and regionalizedMoberlys economy, attracting workers from40 miles around with good pay and bene ts.

Although Meha y puts a priority onemployer retention, hes not about to missa chance to recruit a new company totown. Early in 2009, he landed a call centerfor Stark Bros Ful llment, a mail-ordermerchandiser based in Louisiana, Mo. Tecompany plans to eventually employ 85.

And next? City Manager Andrew Morrisenvisions a new wave of economic develop-ment in retailing. His evidence: a Lowesthat opened in November, promising120 jobs. No nancial inducements wererequired or even requested.

And beyond that? A startup company,Producers Choice Soy Energy, promises tomake up to 5 million gallons of biodieselfuel annually from 3 million bushels of soybeans bought from local farmers. Tecity abated 10 years of property taxes for thecompany and issued $16 million in bonds toimprove roads, tra c signals and storm-water control around its new $16.5 millionplant. Te company is to repay the bondsfrom its tax savings. It also stands to gain$330,000 in state tax credits if it adds 21 jobs

to the handful it began with.In October, Shapiro Brothers Inc. signeda deal to expand its scrap metal processingbusiness by buying roughly 10 acres fromRandolph County Sheltered Industries.Te purchase money will help sustain theworkshop for the handicapped, which willcontinue to occupy a building on the site.Shapiro, for its part, expects to add 10 jobs,roughly tripling its Moberly numbers overthe next three years.



18/24


19/24

n t o n L o V r V W

By Kevin L. Kliesen

During the third quarter of 2009, realGDP rose at a healthy 2.8 percentannual rate. Tis increase, the rst in a year,and the likelihood of continued moderategrowth in the fourth quarter signaled theend of the Great Recession, which startedin December 2007. However, a majority of the public probably has a di erent opinionbecause most people o en view the stateof the economy through the lens of labor

market conditions, which remain quite weak.Regardless, the discussion among economistsand forecasters is turning to the contours of the recovery in 2010 and to whether in ationwill remain quiescent. Many economists andkey policymakers expect real GDP to continueto grow this year, by about 3 percent, with CPIin ation to be about 1.75 percent. Still, thereis some concern that the recovery may weakenas the temporary measures that were designedto boost growth come to an end.

An Unusual Recession and Recoveryypically, deep recessions are followed by

strong economic recoveries. Following the1973-75 and 1981-82 recessions, real GDPgrowth averaged 7 percent during the rst fourquarters of the recovery. By comparison, therecoveries that followed the relatively shallow1990-91 and 2001 recessions produced only modest real GDP growth, and the unemploy-ment rate continued to rise well a er the reces-sion ended. Tus, if the past is any guide forthe future, forecasters should have projected

rapid real GDP growth and a sharp decline inthe unemployment rate for 2010. Yet, that isnot the case: Te consensus of forecasters surveyed in November 2009 for the PhiladelphiaFeds Survey of Professional Forecasters wasthat real GDP growth would average about2.5 percent in 2010 and that the unemploy-ment rate would remain above 10 percentfor most of the year.

Tis forecast is even more unusually low in light of the exceptionally robust

countercyclical policies put inplace to jump-start the econ-omy. On the scal side, thesetemporary measures included

scal stimulus packages in 2008 and2009, a tax credit for rst-time homebuyers and the so-called Cash for Clunkersprogram. On the monetary side, FederalReserve policymakers initiated severalinnovative lending programs designed to

improve conditions in nancial and mort-gage markets. Fed policymakers also statedtheir intention to maintain their federalfunds interest rate target at an exceptionally low level for an extended period.

The Recoverys Potholes

When attempting to project the paceof economic activity over the next severalquarters, forecasters o en look closely atfactors that in uence spending by households(consumption and housing investment) and

businesses ( xed investment). ogether,these components comprised a little less than85 percent of GDP in 2009. Te following arelikely to be key developments in uencing thepace of household and business expenditures,and thus the shape of the recovery, in 2010.

First, job gains are expected to averageabout 11,000 per month over the rst half of 2010 and then average about 150,000 permonth over the second half. With weak job growth likely moderating the pace of consumer spending, businesses are going to

be reluctant to boost outlays for structures,equipment and so ware.Second, households continue to boost

their saving rates and pay down the sizablelevels of debt that were taken on over thepast 20 years or so.

Finally, the U.S. economy appears to beundergoing some signi cant structuralchanges in the a ermath of the nancial cri-sis and Great Recession. As labor and capitalleave these industries (for example, autos

and housing), time is needed for these eco-nomic resources to become fully employed

again. For all of these reasons and more, therecovery is expected to be relatively tepidcompared with the snap-back that typically follows deep recessions.

Disagreement about the Infation Threat

Much disagreement seems to exist aboutthe outlook for in ation over the next twoto three years. Some economists believethat a high unemployment rate, subduedin ation trends and well-anchored in a-tion expectations will keep in ation low andstable over this periodand maybe beyond.Other economists, while perhaps not sens-ing an imminent threat this year, point tothe potentially in ationary consequences of doubling the monetary base, large protracted

scal budget de cits and further declines inthe dollar.

Despite these concerns, nancial marketindicators and surveys of households andbusinesses generally suggest little fear of deterioration in the in ation outlook over thenext few years. As long as in ation expecta-tions remain low and stable, long-term inter-est rates should also remain relatively low andstable. Long-term price stability will help theeconomy as it evolves in the face of structuralchange and in the labor market dislocationthat this change produces.

The reco e Might Bea 98-Po nd Weakling

Kevin L. Kliesen is an economist at the Federal Reserve Bank of St. Louis. For more on his worksee http://research.stlouisfed.org/econ/kliesen/.Douglas C. Smith provided research assistance.

us on by b u M ph son/www.donn os n s s. oM The Regional Economist | www.stlo is ed.o g 19


20/24

t r c t o V r V W

Eighth Dist ict F es Betteth n N tion in Job Losses

The Eighth Feder l Re erve Di trictis

composed of four zones, each of whichis centered around one of the four maincities: Little Rock, Louisville, Memphisand St. Louis.

MISSOURI

ILLINOIS

ARKANSASTENNESSEE

KENTUCKY

MISSISSIPPI

INDIANA

Memphis

Little Rock

Louisville

St. Louis

By Craig P. Aubuchon, Subhayu Bandyopadhyay,Rubn Hernndez-Murillo and Christopher J. Martinek

It is no secret that the recent recession hashit labor markets hard. From December2007 to October 2009, the U.S. economy lostmore than 7 million jobs, a decline of 5.3percent in total employment. In contrast,during the 2001 recession, the U.S. economy lost 1.5 million jobs, or 1.2 percent of totalemployment. Furthermore, the U.S. labormarket has yet, perhaps, to hit the bottom.A er the ve previous recessions, it took an average of about 25 months to return tofull employment, with the shortest returnfollowing the 1980 recession (10 months)and the longest return to full employmentfollowing the 2001 recession (46 months).On the plus side, in this latest recession, thefour largest Metropolitan Statistical Areas(MSAs) of the Eighth DistrictSt. Louis,Little Rock, Louisville and Memphisper-

formed somewhat better than the nation.Between December 2007 and October

2009, each of the four MSAs in the EighthDistrict experienced a lower decline in totalemployment than the nation as a whole. 1 Little Rock, with a 1.5 percent decline, lostthe fewest jobs as a percent of total employ-ment, followed by St. Louis (3.9 percent),Memphis (3.9 percent) and Louisville(4.3 percent). Tese cities represent just1.6 percent of total jobs lost during the cur-rent recession, or just over 100,000 jobs. 2

More surprising was the mix of job losses.While the latest recession fueled the long-term trend in the loss of manufacturing jobs, it also increased job losses in sectorsthat are typically considered recession-proof, such as information and nancialservices.3 A sector-by-sector comparisonbetween the four largest Eighth DistrictMSAs and the U.S. reveals that each MSAperformed better than the nation in percent-age of jobs lost between December 2007 and

October 2009 for the following categories:resources, mining and construction; manu-facturing; nancial services; and profes-sional and business services. Furthermore,each MSA experienced higher growth thanthe U.S. for government services during thissame time period. In contrast, each MSAalso performed worse than the U.S. in cer-tain industries. Both Louisvi lle and LittleRock experienced a greater decline than thenation in trade, transportation and utilities,while Memphis experienced a sharperdecline in information services. Finally,both Memphis and St. Louis experienced agreater decline in leisure and hospitality services than the U.S. for the December2007 to October 2009 period.

Te accompanying chart presents changesin employment for the period October 2008

to October 2009. Similar to the experiencesince the start of the recession, each MSAwas below the nation in percentage of jobslost for most categories during this timeperiod. Furthermore, each MSA beganto see stronger job performance (deter-mined by fewer jobs lost relative to thenation and other Eighth District MSAs,or by job growth) than the U.S. for thoseindustries with the highest relative shareof employment.

Little Rock Zone

Little Rock fared the best among EighthDistrict MSAs during the recent recession.It was the last to lose jobs over previousyears levels. November 2008 marked the

rst month since November 2002 that theLittle Rock metropolitan area experienced adecrease in year-over-year payroll employ-ment. Furthermore, between October 2008and October 2009, Little Rock experiencedthe smallest decrease in year-over-year

job losses of the four metro areas. Payrollemployment fell 1.6 percent in the LittleRock area from October 2008 to October2009well below the national experience.

Little Rock experienced the largestyear-over-year job declines in the trade/transportation/utilities (6.9 percent), man-ufacturing (5.7 percent), professional andbusiness services (4.4 percent) and otherservices (3.8 percent) sectors. Tese losseswere partly o set by gains in the leisure/hospitality (4.6 percent), education/health(2.2 percent), information (1.1 percent) andgovernment (1.4 percent) sectors.

In October 2009, Little Rock had thehighest concentration of employment, ascompared with the other three DistrictMSAs, in the aforementioned industries of growth, with the exception of leisure/hospi-

tality. Government services made up 20.5percent of the total employment in LittleRock, information services made up 2.6percent, and education and health servicesrepresented 14.6 percent, second only toSt. Louis, a city with 16.5 percent employ-ment in this sector.

Within the Little Rock Zone (a Feddemarcation), Fort Smith, Ark., and ex-arkana, Ark., posted year-over-year jobdeclines of 1.7 and 2.1, respectively, bothin excess of Little Rocks. In Fayetteville,

Ark., the decline was 1.5 percent.Louisville Zone

From October 2008 to October 2009,payroll employment in the Louisville areadropped 3 percent. Many of these lost jobswere in goods-producing industries. Morethan 16 percent of the Louisville work forceis employed in goods-producing industries,the highest proportion among the fourmajor metro areas in the District. In these



21/24

industries, the Louisville area shed 8 percentof the jobs over the past year, largely due toa 12 percent decline in resources/mining/construction jobs.

Louisville also experienced signi cant job loss in the trade/transportation/utilities(6.5 percent), professional/business services (1.4 percent) and the other services(3.3 percent) sectors. Of the four metro

areas, Louisville posted the largest year-over-year decline in the nance sector,losing 2.3 percent of total jobs. Louisvilleexperienced job growth in both the govern-ment and education/health sectors, consis-tent with other areas in the District.

Evansville, Ind., and Clarksville, enn.,two smaller metro areas in the LouisvilleZone, experienced slightly smaller year-over-year job losses (3.2 percent and 2.9percent, respectively) than the Louisvillearea did. Job growth declined 5.7 percentin Bowling Green, Ky.

Memphis Zone

Unlike the other three major metroareas, where employment did not start todecrease over previous years levels untilseveral months into the recession, Memphisexperienced year-over-year payroll employ-ment decreases throughout most of 2008and 2009. More recently, the year-over-year job loss was subdued in the Memphis area

compared with other MSAs. From October2008 to October 2009, nonfarm employmentdecreased 2 percent in the Memphis area.

Memphis employs a relatively large shareof its work force in the trade/transporta-tion/utilities sector. With greater than26 percent of the Memphis work force inthis industry, the 2.9 percent year-over-yeardecrease in payroll employment within

this sector from October 2008 to Octoberto 2009 contributed signi cantly to theoverall payroll employment decline. Joblosses in excess of the U.S. experienceoccurred in the information sector (6.9percent). Memphis also experienced sig-ni cant job loss in the professional/businessservices sector (4.8 percent). In contrast,the Memphis region experienced job growthin the education/health services sector (1.3percent) and the other services sector (3.7percent). Memphis was the only one of thefour major MSAs in the District to add jobsin the other services sector.

Jackson, Miss., a smaller MSA in theMemphis Zone, experienced a 3.7 percentdecrease in year-over-year employment.

St. Louis Zone

Over the past year, St. Louis experiencedthe largest year-over-year decline in nonfarmemployment among the four big cities at3.2 percent. As the Districts largest MSA,

St. Louis economy most closely resemblesthe national economy. In both the St. Louiseconomy and the national economy, 14percent of the work force is employed ingoods-producing industries, while 86 percentis employed in service-providing industries.Similar to the U.S. experience, job losses inthe St. Louis area were most heavily con-centrated in goods-producing industries,

in which more than 20,000 jobs were lostsince the previous year. While other largemetro areas in the District lost manufactur-ing jobs at roughly half of the 11.6 percentrate of decrease for the nation, St. Louis lost9.7 percent of its manufacturing jobs sinceOctober 2008.

Job losses in the St. Louis region werenot isolated to goods-producing industries.In the year ending October 2009, theSt. Louis metro area lost jobs in every industry category except education andhealth services, in which the number of jobsincreased by 1.2 percent. St. Louis lost asigni cant percent of total jobs in the otherservices (7.4 percent), business services(3.2 percent), trade/transportation/utili-ties (3.2 percent) and leisure/hospitality services (2.6 percent) sectors.

Columbia, Mo., and Spring eld, Mo.,two smaller MSAs in the St. Louis Zone,

continued on Page 22

10.00

5.00

0.00

5.00

10.00

15.00

20.00

Resources, Trade, ProfessionalTotal Mining and Transportation Financial and Business Education Leisure and Other

Nonfarm Construction Manufacturing and Utilities Information Activities Services and Health Hospitality Government Services

Little Rock 1.61 1.05 5.73 6.93 1.12 1.58 4.43 2.23 4.60 1.41 3.84Louisville 2.96 12.03 5.98 6.54 0.95 2.32 1.43 2.11 1.67 1.55 3.26Memphis 2.00 7.26 6.20 2.89 6.85 1.21 4.81 1.26 0.00 0.15 3.67St. Louis 3.24 11.31 9.70 3.16 2.24 1.74 3.24 1.24 2.63 0.44 7.35U.S. 4.04 15.20 11.57 4.37 5.06 4.83 5.52 2.07 2.02 0.41 2.76

Employment Growth

O C T O B E R 2 0 0 8 T O O C T O B E R 2 0 0 9 Y E A R - O V E R - Y E A R P E R C E N T C H A N G E

sou : b rea a r s tati tic a a t r calc lati



22/24

c o n o M Y t g L n c

Eleven more charts are available on the web version of this issue. Among the areas they cover are agriculture, commebanking, housing permits, income and jobs. Much of the data is speci c to the Eighth District. To go directly to these use this URL: www.stlouisfed.org/publications/re /2010/a/pdf/01-10data.pdf.

U . S . A G R I C U LT U R A L T R A D E FA R M I N G C A S H R E C E I P T S

04 05 06 07 08 09

75

60

45

30

15

0

NOTE: Data are aggregated over the past 12 months.

Exports

Imports

OctoberTrade Balance

B I L L I O N S

O F D O L L A R S

04 05 0906 07 08

190

170

150

130

110

90

NOTE: Data are aggregated over the past 1

regional economist - january 2010

Documents