September - Milken Institute · in stock market wealth Exhibit 4 lists the percentage drop in the...

16 The Milken Institute Review

By Peter Navarro and Aron Spencer

September

17Fourth Quarter 2001

H

AssessingtheCostsofTerrorism

11,2001

How much will the worst terrorist attack in history cost

America? The answer will surely be in the range of hun-

dreds of billions of dollars – and could well hurdle into

the trillions.

While this article does offer preliminary estimates of

the direct and indirect costs of the attack, pinning

down the numbers is just one reason – and not the

most important reason – to make the reckoning. For

the ultimate impact of the tragedy of September 11

depends on policy choices that have yet to be made.

Such choices range from the nature of the military

response, the methods used to increase domestic secu-

rity, as well as the means used to compensate the fami-

lies, industries and localities affected.

Arguably the most critical choices, though, will be

the ones that affect the economy as a whole – in partic-

ular, government fiscal and monetary policy, and

potential instability in world energy markets. For while

a proper accounting of the short run costs of the attack

by Peter Navarro and Aron Spencer

ap/w

ide w

orld

pho

tos

The Milken Institute Review

yield mind-numbingly large figures, the dam-age promulgated through inflation, unem-ployment and lower productivity could dwarfthe numbers linked to loss of property and

life. Indeed, if the President, the Federal Re-serve and Congress make imprudent policychoices, we may wind up on a slower growthtrack, with attendant hardships for tens ofmillions of Americans.

Exhibit 1 summarizes the broad categoriesof costs examined here. These range fromproperty damage, to an economist’s account-ing for the loss of human life, to the afore-mentioned long-term drag on productivityand growth.

property damageIt will be months before a full accounting ofthe property destroyed or damaged by terror-ists on September 11, 2001 is possible. Exhibit2 provides a preliminary reckoning, based ona combination of published reports and con-strained guesswork.

The big-ticket items include the loss offour civilian aircraft valued at $385 million,the destruction of major buildings in theWorld Trade Center with a replacement costof from $3 billion to $4.5 billion, damage to aportion of the Pentagon that will cost up to $1billion to fix, and cleanup of the hundreds ofthousands of tons of rubble at the crash sitesfor another $1.3 billion.

In addition, New York’s World FinancialCenter, 30 West Broadway, 22 CortlandtStreet, the Bankers Trust building, and 30West Street suffered severe, but repairable,damage. Estimates of the cost of repair tothese buildings are not yet available.

The terrorist attack took a very heavy tollon public works and infrastructure. Firetrucks, ambulances and police cars with anapproximate value of $35 million were lost inthe destruction. Streets were damaged, as werepublic utilities and subway infrastructure. Weestimate the loss in the range of $2 billion.

Finally, 120,000 workspaces were de-stroyed. TowerGroup Technology estimates

THE IMMEDIATE AFTERMATH

• Property damage

• Loss of human life and injuries

• Lost economic output

• Reduction in stock market wealth

• Psychological impacts of terrorism

LONGER TERM MICROECONOMIC EFFECTSOF A “TERRORISM TAX”

• Increased airline security

• Other security measures

SECTOR-SPECIFIC IMPACTS

• Advertising

• Airlines

• Insurance

• Hotel and tourism

GOVERNMENT BAILOUTS AND BUDGETARY IMPACTS

• The airlines

• New York City

• Reduced federal, state and local tax revenues

THE HIGHER OIL PRICES-WEAKER DOLLAR CONUNDRUM

• Effects of an oil price shock

• Costs of a weaker dollar

• Destabilization of the stock market and international monetary system

MACROECONOMIC COSTS

• A deeper recession

• A more volatile business cycle

• A lower long-term growth path and economicstagnation

18

P E T E R N AVA R R O (http://www.peternavarro.com) is an economist at the Graduate School of Management,University of California (Irvine) and the author of If It’sRaining in Brazil, Buy Starbucks: The Investor’s Guide toProfiting from News and Other Market Moving Events.A R O N S P E N C E R is a PhD candidate at UC (Irvine). Dr.Frank Harris helped significantly as did volunteer MBA students from UC (Irvine): Judy Allen, Holly Anderson,Carl Kim, Mike Kim, Ali Mogharabi, Arash Namvar, IndiraRenduchintala, Sunny Sang and Michael Slater. A specialthanks to Richard T. Carson for his assistance.

EXHIBIT 1: COST CATEGORIES FOR TERRORIST ATTACKS


t e r r o r i s m


computer hardware losses at $1.7 bil-lion, with software/network losses at$1.5 billion.

loss of human lifeThe terrorist attacks claimed upward of6,000 lives, in total. Some 266 innocentlives were lost on the four downed air-craft. An additional 125 are dead in thePentagon. The New York City toll is list-ed at more than 5,000 missing or con-firmed dead.

In the most profound sense, there isno way to translate life in dollars. If youask people how much they would pay toavoid a fatal accident today, the typicalanswer is everything they have.

Still, people don’t behave as if theirlives are infinitely valuable: they drivecars, use dangerous tools, sleep in hous-es without sprinkler systems, and en-gage in myriad other activities thatincrease the risk of premature death inreturn for pleasure or income or sav-ings. And economists infer value to life bymultiplying the financial gains associatedwith bearing marginally greater risk by theprobability of fatality associated with the risk.

Consider two crude examples. If a con-struction worker is willing to accept a 1 per-cent chance of a fatal accident over his work-ing life in return for $50,000 more lifetimepay, he is implicitly valuing his own life at $5million ($50,000 divided by .01). If a home-owner refuses to install a carbon monoxidedetector that costs $30 near her gas heater,thereby increasing the risk of dying by onechance in 10,000, it implies that she values herlife at no more than $3 million ($30 dividedby .0001).

Estimates based on a variety of risk-tak-ing/risk-avoiding activities generate remark-ably consistent numbers. Kip Viscusi of Har-

vard puts the value of human life at between$3 million and $7 million – numbers that arenow used by the federal government in decid-ing whether to mandate additional safetyap

/wid

e wor

ld p

hoto

s

MAJOR BUILDINGSBuildings destroyed in $3.0 billion to $4.5 billionWorld Trade Center complex

Pentagon damage $250 million to $1 billion

Cleanup of rubble from crash sites $1.3 billion

AIRCRAFTFour passenger jets $385 million

PUBLIC WORKS AND INFRASTRUCTUREFire trucks, ambulances, police cars $35 million

Infrastructure such as streets, sewer, $2 billionutility lines and subways

CORPORATE PROPERTYOffice equipment and software $3.2 billion

TOTAL $10 billion to $13 billion

COST

EXHIBIT 2: PROPERTY DAMAGE FROM TERRORIST ATTACKS


equipment in commercial aircraft. Not sur-prisingly, the value that people implicitlyplace on their own lives rises with income andeducation.

Assuming most of those lost in New Yorkwill fall at the higher end of the education andincome scales, and that there is considerablecollateral damage to their families in the formof emotional pain and income loss that is not

included in how people value their own lives,the 6,000-plus death toll translates into aneconomic cost in the range of $40 billion.

lost economic output in the immediate aftermathExhibit 3 (page 22) offers estimates for themajor categories of lost economic output.

In response to the crisis, the entire air



transportation network was shut down fortwo-and-a-half days at a cost of almost $1.5billion in lost airfares and cargo-shipping rev-enues. Federal Express lost $100 million injust 48 hours and presumably other courierservices such as UPS lost tens of millions aswell. Numerous small aviation enterprises,from crop dusters to traffic helicopter servi-ces, likewise sustained significant losses. In

addition, the major carriers are projected tolose another $3.4 billion to reduced demandby the end of September, for a total approach-ing $5 billion.

The hotel industry suffered numerousroom cancellations, at a loss in revenues of$700 million. Television and radio stationsrunning commercial-free coverage lost rough-ly $1 billion in foregone advertising revenuesap

/wid

e wor

ld p

hoto

s


CATEGORY LOSS

Lost airline and cargo shipping revenues, $4.7 billionSeptember 2001

Lost hotel industry revenues $700 millionthrough first weekend of crisis

Lost advertising revenues in television $1 billionand radio during commercial-freecoverage in the first days

Two-day partial work stoppage and $35 billionloss of productivity

Lost consumer spending and retail $6 billionsales

TOTAL $47 billion

in a matter of days. Newspapers suffered adouble squeeze of increasing the pages of cov-erage while dramatically decreasing ad space;magazines were also hard hit.

Golf tournaments, baseball games, collegeand professional football games, and numer-ous other sporting and entertainment eventswere cancelled or postponed. We can onlyguess at these impacts, though surely the lossexceeds $100 million.

A two-day partial work stoppage andattendant loss of productivity, as the nationadjusted to the shock, likely cost another $35billion. Consumer spending fell by a reported20 percent in smaller cities and 50 percent inlarger ones in the two days following the at-tack, for a loss of retail sales around $6 billion.

We acknowledge that these numbers mustbe taken with a healthy dose of skepticism.

Gross revenues lost by retailers, for example,are surely larger than the opportunity cost ofthe labor and capital services that would havebeen used to produce the revenues. And, inany event, there is clearly an element of dou-ble-counting in adding the market value ofthe labor lost to the retail value of the goodsand services not sold.

On the other hand, there is an element ofundercounting here, too. The market value ofgoods and services not sold represents a lowerbound on their value to consumers – a day atDisneyland may be worth $500, even thoughthe price of admission and assorted souvenirsis just $100. We reiterate: the point of thisadmittedly rough exercise in social account-ing is to put the losses in useful context.

short run reduction in stock market wealthExhibit 4 lists the percentage drop in themajor U.S. stock market indices and loss ofmarket capitalization from the close of themarkets on Monday, September 10 to Friday,September 21 – the end of the first week oftrading after the attack.

The New York Stock Exchange and theNasdaq, which list the lion’s share of large,publicly traded corporations, suffered double-digit drops. This translates into a total loss ofstockholder wealth, as measured by the re-duction in market capitalization, of a stagger-ing $1.7 trillion.

In addition, shareholders in other majormarkets around the world likewise sufferedsubstantial losses as of September 21. Theseforeign exchanges fell from a minimum of 6.3percent for the Nikkei to well over 10 percentfor all of the major European exchanges, withthe German exchange falling by almost 20percent.

Is the drop in the market value a measureof cost? In one sense, it may be the best single

EXHIBIT 3: LOST ECONOMIC OUTPUT IN THE IMMEDIATE AFTERMATH

NYSE (U.S.) -11.24% -$1.3 trillion

Nasdaq (U.S.) -16.05% -$407 billion

Amex (U.S.) -8.01% -$8.5 billion

DROP ININDEX AS OF

SEPTEMBER 21LOSS IN MARKETCAPITALIZATION

EXHIBIT 4: LOSS OF STOCK MARKET WEALTH

22

t e r r o r i s m


measure of the losses to the owners ofcapital linked to the physical damage, theexpected decline in corporate profitsassociated with macroeconomic instabil-ity and – most elusive – the greater riskpremium investors now attach to theownership of equity.

Again, though, we acknowledge theelement of double-counting: you can’tlegitimately add lost revenues to the air-line or hotel business to the decline in themarket capitalization of the two indus-tries. But measuring the decline in stockvalues is surely worth doing since it givessome sense of how much damageinvestors believe has been done to theirfuture claims on corporate income.

psychological impacts of terrorismThe worst terrorist attack in Americanhistory exacted a grave emotional cost toAmericans who were not directly affect-ed. This terror, and the attendant per-ceived loss of freedom to travel and workwithout terrorism’s risks, imposes aheavy psychological impact – precisely asit was intended.

The question, of course, is how wemight put a value on these intangiblecosts. One way would be to apply whateconomists call “contingent valuation” (CV)analysis. CV analysis uses survey methods toassess the willingness of people to pay forgoods not regularly traded in the market-place. Such “non-market goods” includecleaner air and water, a reduction in nuclearpower plant radiation risk, and broader pub-lic safety concerns. Thus, in the wake of theExxon Valdez oil spill, the State of Alaskacommissioned a survey in which Americanswere asked how much money they wouldhave been willing to pay collectively to undo

the damage. The (very controversial) answer:$3 billion.

In this particular case, the question is:How much would we pay to be able to fly (orwork in large office buildings, or drink thewater from public reservoirs) without fear?

The emotional cost of the attack on theWorld Trade Center surely dwarfs that associ-ated with thinking about all those dead birdsand oil-drenched sea otters in Alaska. But bywhat order of magnitude? Hypothetically, ifeach of the 100 million American householdsap

/wid

e wor

ld p

hoto

s


not living in poverty would give up a mere$1,000 to be able to forget that Osama BinLaden and his ilk walk the earth, the emo-tional damage must be equivalent to at least$100 billion.

microeconomic implications of the terrorist tax and defensive behavior Rachael Carson’s Silent Spring, released in1962, marked an important turning point inpublic awareness about the dangers of envi-ronmental pollution. The result was the birthnot only of a new government bureaucracy –the Environmental Protection Agency – butalso a plethora of new government regula-tions. While these regulations had the salu-tary effect of cleaning up the nation’s air andwater resources, they also imposed a burdenon the nation’s productive capacity. Indeed,we would argue that the new regulationturned out to be just one of a number of sup-ply-side shocks to the economy that createdirreconcilable conflicts between Americans’expectations of rising living standards and thereality of reduced productivity growth in the1970s – conflicts that led to “cost-push”stagflation of the era.

The September 11 terrorist attacks pro-mise to give rise to a comparable regulatoryresponse that creates a drag on productivity.The important lesson here from the 1970s isthat brute-force responses to real problemscan lead to great harm. It took the environ-mental bureaucracy many years to shift fromunnecessarily expensive “command and con-trol” systems of pollution abatement to morehighly targeted and economically efficientmethods. Will the nation again waste vastsums in the pursuit of legitimate public goals?

America faces two different kinds of ter-rorist threats to civil aviation. The first is the

“Lockerbie problem” – the destruction of air-craft by bombs. The second is the hijack ofaircraft for use as weapons of mass destruc-tion or, as in the past, as collateral for ransom.At present there are a number of policy op-tions on the table to address these problems –all of which potentially impose very highcosts. These options, and their microeconom-ic impacts, are summarized in Exhibit 5.

One option is to post sky marshals onevery plane flying in the U.S. At present, theU.S. government employs 500 sky marshals.Simply doubling the sky marshal force wouldcost roughly $100 million. At the other end ofthe spectrum, we could put two such mar-shals on every plane flying in the U.S. Thiswould require some 40,000 additional mar-shals. It would also lead to the creation of abureaucracy many times the size of the FBI.The annual cost would be roughly $6 billion.

A second proposal is to replace private air-port security with a government force, onepresumably managed by the same bureaucra-cy as the sky marshals. The implicit rationalefor this reverse privatization is that it is verydifficult to create an incentive structure forprivate security forces that leads to optimallevels of security delivered at least cost.

Government employees, the argumentgoes, would be paid far more than the rent-a-

Sky marshals Up to $6 billion per year

Government takeover of $1 billion per yearairport security

Retrofitting of aircraft with $450 millionanti-terrorist devices

New technology “fixes” $2 billion per year

Increased airport delays $8 to $32 billion per year

TOTAL (FIRST YEAR) Up to $41 billion

COST

EXHIBIT 5: THE MICROECONOMICCOSTS OF “TERRORIST TAX”

24

t e r r o r i s m


guards who now screen passengers, and itwould thus be easy to recruit more qualifiedemployees. Assuming the current force ofroughly 18,000 guards was increased to25,000 and that the average salary wasincreased from approximately $15,000 a yearto a government pay scale of $35,000 a year,

the wage bill alone would rise by $600 millionannually. With additional overhead of a gov-ernment security force, the net cost increasewould surely approach $1 billion annually.

A third policy option is to retrofit everyplane with a reinforced cockpit door and torequire all new planes to adopt the Israeliconfiguration of two reinforced doors to thecockpit. The retrofits alone would likely cost$50,000 per plane, for a total of roughly $450million.

U.S. regulators will no doubt also look atmore sophisticated new technology fixes.

These range from the use of computers totake control of flights seized by terrorists tomore sophisticated 3D baggage scanners,face-recognition systems, and biometric iden-tity devices, such as retinal and fingerprintscanners. Equipping the nation’s roughly9,000 commercial aircraft and 50 largest air-

ports with an array of these technologiescould cost as much as $2 billion.

Big numbers? Perhaps. But even the up-per-bound numbers work out to just $20 to$25 a ticket. Moreover, if these measures weresufficient to restore enough confidence in theAmerican public so that airline travel reachedits previous levels, the cost per ticket could fallbelow $20.

The real worry is that the intangible costsassociated with increased airport security will be ignored and, as a result, too little at-tention will be paid to passenger convenience,ap

/wid

e wor

ld p

hoto

s


permanently burdening the airline industrywith unnecessary delays. Assume that airlinetraffic falls by the 20 percent now projected bythe airline industry. This would reduce the

number of annual passenger trips in theUnited States to 525 million. If each travelermust arrive 90 minutes earlier, as recom-mended by the FAA, the cost in additionaltravel time would be 788 million hours.

What value should we put on that time?Surveys and behavioral-based studies suggestpeople place a value of about $10 per hour ontime spent in their cars, while studies of airtravel put the cost at two to four times that

much because air travelers havehigher average incomes. Accor-dingly, we estimate the value of the788 million hours lost to be some-where between $16 billion annually(valuing time at $20 an hour) and$32 billion annually (valuing time at$40 an hour).

Note, at this point, that we haveonly considered the costs of in-creased security for our nation’s air-ports. In fact, there will be consider-able expenditures by federal, stateand local governments as well as byprivate corporations to beef up secu-rity at national monuments, govern-ment buildings, nuclear powerplants, water treatment plants,sports stadiums, high-rise officebuildings, concert halls and anyplace where large numbers of peoplegather and are vulnerable to a ter-rorist attack. No estimates for theseefforts are presented here, but theincrease is likely to run into the bil-lions – and many times more if secu-rity procedures waste work andrecreation time.

longer term sector-specific costsThe four sectors likely to bear a largebrunt of the attack will be the adver-

tising, airline, insurance and hotel/tourismindustries. Of the four, only the insuranceindustry apparently has the resources to rideout the storm without great dislocation.

As noted earlier, advertising-based media ap/w

ide w

orld

pho

tos

26

t e r r o r i s m


lost as much as a billion dollars in revenueswithin just a few days of the attack. Morebroadly, industry analysts now project a 6.6percent decrease in ad revenues for theremainder of 2001 and a 7.4 percent dropduring 2002. Taking into account that part ofthese decreases were due to recessionary con-ditions prior to the attack, we value the netdrop at about $5 billion.

The airline industry is projecting a 20 per-cent reduction in demand beginning in thefourth-quarter 2001. This translates into aloss of just under $20 billion in the industry’scontribution to the gross domestic producton an annual basis. It is unclear whendemand may resume its growth, or whetherthe labor and capital freed by this industryshrinkage can be productively used elsewherein the foreseeable future. Much will dependon the course of events and the policiesadopted for improved airline security.

Similarly, the hotel and tourism industriesare hunkering down for a projected 20 per-cent decline in demand through fourth-quar-ter 2000 at a revenue loss of $5.5 billion. Thisis on top of the $700 million lost in the fourdays following the attack.

Longer term, demand for hotel rooms isprojected to drop in 2002 from its pre-attacklevels at a rate between an optimistic 3 per-cent decline to a more pessimistic 10 percentdecline. This translates into an additional $3billion to $11 billion in lost revenues.

In addition, the effects will ripple down-stream to the construction industry, as fewerof these facilities will be built. In the worst-case scenario, virtually all large hotel projectswill be put on hold for several years at a costin foregone construction revenues of about$12 billion. Again, it is unclear what portionof the labor and capital freed could be pro-ductively used in other industries.

Turning to the insurance industry, the ter-

rorist attacks only accelerate a trend: insuredcatastrophic losses have been rising dramati-cally over the past several decades. Exhibit 6summarizes some of those insured losses(adjusted for inflation) for catastrophes rang-ing from earthquakes and riots to hurricanesand terrorist bombings.

Note that two major hurricanes – Andrewand Hugo – resulted in losses of well over $20billion, while the previous World TradeCenter bombing generated claims of $542million and the Oklahoma City bombinggenerated claims of $127 million. HurricaneAndrew did force some insurance companiesinto bankruptcy. But that was the exception;the industry tends to pool its vast resourcesthrough reinsurance schemes.

Congress has limited the liability of theairlines (and their insurers) in the events ofSeptember 11. The airlines’ liability hascapped at the level of insurance at the time ofthe attacks – $1.5 billion per airline forAmerican and United, with a $500 milliondeductible each.

In addition, these companies would be re-sponsible only for the deaths of passengers.With the death count on board at 266, thisamounts to roughly $11 million per life lost,with the airline companies responsible for thefirst $1 billion of the payouts.

For persons killed on the ground, analystsestimate that an average of $500,000 will beawarded to the families of each of the victims.

CATASTROPHE INSURED LOSSES

1989 Hurricane Hugo $5 billion

1992 Los Angeles riots $844 million

1992 Hurricane Andrew $16.9 billion

1993 World Trade Center bombing $542 million

1994 Northridge, California earthquake $13 billion

1995 Oklahoma City bombing $127 million

EXHIBIT 6: INSURED LOSSES FROMOTHER DISASTERS


The actual amount per family will vary great-ly. We estimate that the total bill for lives loston the ground is somewhere on the order of$3.25 billion.

Someone must pay for expected changesin future risk. If premiums on business insur-ance of all types rise by an average of 5 per-cent, America’s annual insurance bill will riseby $8 billion. Of course, if claims don’t riseaccordingly, the $8 billion represents a trans-fer of income to the insurance industry – nota real cost to the economy.

government bailouts and lost tax revenuesCongress has approved a bailout of $18 bil-lion for the airlines. Of that, $5 billion is incash subsidies, $10 billion is loan guarantees,and $3 billion is in aid for increased securitycosts. Since loan guarantees represent modestcosts to the government – the differencebetween the cost of capital to Uncle Sam andthe cost of capital to the industry – we limitour analysis to the other items.

The City of New York will surely receivemassive aid from the federal government. Atthe time of this writing, there were no reliableestimates of the size of the aid package – Gov.Pataki is asking for $54 billion – but it willcertainly run into the billions.

On the other side of the ledger, federal,state and local tax revenues will all drop by asubstantial amount. For the federal and stategovernments, this will come in the form ofreduced corporate-tax receipts both fromcorporations operating in the affected sectorsand, more broadly, through the recessionaryeffects that will be discussed further below. Asan example of the many and varied ways theSeptember 11 attacks will ripple through thebudget, we are likely to see an extraordinaryrise in deductible charitable contributions.

This will, in turn, lower federal and state taxreceipts.

For local governments, the reduction inhotel-room demand will result in an appre-ciable drop in hotel-room taxes, which aver-age about 12 percent of the room rate. Suchtaxes constitute a significant fraction of localgovernment revenues, particularly in majortourist destination cities like Boston, SanFrancisco, San Diego, Seattle and Miami. Weestimate that the projected 3 percent to 10percent drop in hotel room demand in 2002will result in the loss of $400 million to $1.3billion in such local tax revenues.

Note that, as a first approximation, gov-ernment transfers and reduced tax revenuesrepresent no net economic losses; they arematched, dollar for dollar, by reduced liabili-ties to the individuals, corporations and local-ities affected by the terrorist attacks. But look-ing closely, there are bound to be real (if diffi-cult to measure) losses associated with thetransfers. For example, the value of the ser-vices that a local government cuts – saygarbage collection or after-school programs –will probably exceed the monetary gains tothe tourists who have cancelled their trips andthus don’t pay the tax.

the higher oil price – weaker dollar conundrumOil prices spiked after the terrorist attacks onfears of war in the Middle East. Within days,however, prices fell to levels below the attacklevel because of offsetting concerns aboutrecession and a weakening demand for oil.

As of this writing, it remains to be seenwhether oil prices will again spike in the wakeof military action in Afghanistan or else-where. While a military strike or sabotagemight disrupt the oil flow, the greater risk is acollapse of cooperation with Persian Gulf oilproducers that leads to production cutbacks

28

t e r r o r i s m


like those experienced in the 1970s.Such an embargo would likely be orga-

nized by oil-exporting countries, like Iraq,Iran and Libya, that are hostile to America.However, any successful oil embargo wouldprobably require the cooperationof the Middle East’s largest oil pro-ducer, Saudi Arabia. The Saudisaccount for half of the oil pro-duction from the Mideast. Morerelevant, they have the productioncapacity to offset any reduced out-put elsewhere in the region.

At the time of this writing, it isimpossible to predict how eventswill unfold. Let us simply note thatfor every dollar increase in theprice of a barrel of oil, America’s oilimport bill rises annually by rough-ly $3.4 billion. Moreover, higher oilprices ripple pervasively throughthe economy as American energyproducers raise their prices andenergy-intensive sectors – fromaluminum to plastics – face higherproduction costs. If governmentcopes as badly with the resultingdislocations as it did in the 1970s,the real cost would be many timesgreater than the transfers to energyproducers.

The terrorist attack has, at leastinitially, also weakened the ex-change value of the dollar. Whilethe root cause is difficult to deter-mine, foreigners fearing war andinstability may be pulling their money out ofU.S. financial markets. As investors redeemdollar-denominated assets, they exchangedollars for yen, francs or Saudi riyals, and thevalue of the dollar falls through the laws ofsupply and demand. Add to this the impact ofrising oil prices. More dollars are required to

buy the same amount of imported oil, andmore dollars spent abroad likewise meandownward pressure on the dollar.

A weaker dollar does have the benefit ofstimulating our export industries, offsetting

some of the macroeconomic weakeninglinked to the terrorist attacks. However, it alsois inflationary because it raises the cost ofimports. A 10 percent drop in the value of thedollar would raise our import bill by roughly$140 billion.

There is a broader problem here, however,ap/w

ide w

orld

pho

tos


than simply a rising import bill. A weakeningdollar that increases inflationary expectationswould severely constrain the ability of theFederal Reserve to use expansionary mone-tary policies to fight recession. Moreover, aweaker dollar undermines the economies ofour global trading partners as Americaimports fewer goods. The export-orientedeconomies of Asia – most notably Japan – areparticularly vulnerable. That’s why Tokyotried – unsuccessfully – to prop up the dollarin the days after the attack.

In this regard, there is a much larger con-cern raised by the intricate relationship be-tween rising oil prices and a weakening dollar.Consider this hypothetical chain reaction: amilitary response against the terrorists alien-ates oil producers; an OPEC boycott or othersupply disruption raises oil prices; soaring oilprices drive down the dollar; a falling dollarcreates domestic inflationary pressure; thefear of inflation constrains the Fed from fur-ther action; Fed immobility results in a drainof foreign dollars from the stock market; thedeparture of foreign investors drives downthe stock market; the decline of markets cre-ates a negative wealth effect that pushes thedomestic economy over the edge to depres-sion. This depression spreads globally first toAsia as a weakening dollar dramatically cutsAsian exports to the U.S. and then to Europeas the combined reduction in demand fromthe U.S. and Asia completely undermine theEuropean economy.

Not plausible, you say, because policymak-ers have learned their lessons from the firsttwo oil crises? You are probably right. Still...

macroenomic costsWill the worst terrorist attack in history berecessionary or inflationary – or both? Howwill the attack affect the volatility of the busi-

ness cycle and more importantly, long-runprospects for economic growth? In thinkingabout these questions, it is useful to examinethe four key macroeconomic variables inExhibit 7 for the very good times of 1993 to2000 and the very bad times of 1974 to 1982.

From 1993 to 2000, the U.S. economydemonstrated its full potential in a world ofstable prices, reduced demand for militarygoods and services, and the rapid diffusion ofexciting new technologies. Comparing thisperiod to 1974-82, we can see that productiv-ity growth was substantially higher – 2.14percent versus 0.88 percent. Inflation was vir-tually nonexistent – 1.85 percent versus 7.84percent. The unemployment rate wasremarkably low – less than 5 percent com-pared to more than 7 percent. By no coinci-dence, the economy grew, in real terms, twiceas rapidly as in the 1974-82 period – and halfa percent higher than the historical average of3.5 percent.

The very bad times of 1974-82 were initi-ated by a war and an over-expansionary fiscalpolicy in the 1960s, and compounded by sup-ply-side shocks in the early 1970s. Moreover,the Federal Reserve added to the dislocationby gyrating between contractionary andexpansionary measures. These factors led to agrowth rate a full percent lower than the aver-age of the previous 50 years.

The risk today is that we are witnessing thebeginning of a similar series of destabilizingactions and reactions that could undermine

1974-1982 1993-2000

Growth 1.98% 4.04%Inflation 7.84% 1.85%Unemploment 7.24% 4.96%Non-farm productivity 0.88% 2.14%

EXHIBIT 7: KEY MACROECONOMICVARIABLES IN BAD TIMES AND GOOD

30

t e r r o r i s m


the business climate and push the U.S. econo-my back into a 1970s-style era of stagflationand hard times. This need not be the case.Nonetheless, there is at least one plausiblenightmare scenario, which does not turn onthe aforementioned possibility of an oil crisis.

In stage one of that scenario, which playsout over the next several quarters, the macro-economic shocks of the attacks – lost produc-tion in the immediate aftermath of theattacks, plunging consumer confidence and asharp drop in retail sales, a dramatic reduc-tion in business investment, a negative“wealth effect” from a declining stock market,and so on – all conspire to push the U.S. econ-omy into a deep recession. During this stage,the Federal Reserve responds with an expan-sionary monetary policy, while the Federalgovernment engages in a variety of actionsthat effectively serve as a very strong dose ofexpansionary fiscal policies – the bailout ofthe airlines, aid to New York City, increaseddefense expenditures.

Collectively, this sharp dose of expansion-ary fiscal and monetary policies pulls theeconomy out of the recession at a rapid pace– a pace that sets the stage for increased infla-tionary pressures. Within a year or two, theeconomy overheats in much the same way itdid in the late 1960s during the Vietnam War.At the same time, massive expenditures on in-creased domestic security create additionalcost-push pressures in much the same waythe new environmental and workplace regu-lation did in the 1970s.

The stagflation of the 1970s was, in es-sence, purged by draconian changes in mone-tary policy in 1980-81 that triggered a briefbut very harsh recession and by a changingregulatory climate under President Reaganthat sharply increased business investment. Ifwe are once again caught in doleful 1970s-likecircumstances, there is no guarantee that the

nation could summon the political will tosnap back. And without it, the long-termcapacity of the economy to grow could bereduced sharply. Imagine, for example, thatinstead of growing at an average of 3.5 per-cent annually for the next decade (a half apercentage point less than in the 1994-2000period), the economy only managed to growat 2.0 percent – as in the 1970s. Instead of areal GDP of $14.4 trillion in 2011 (in 2001dollars), GDP would only reach $12.4 trillion– a difference of $2 trillion!

the final reckoningWe’ve attempted a preliminary assessment ofthe costs of the worst terrorist attack in histo-ry. Counting the value of lives lost as well asproperty damage and production of goodsand services forgone, the United States hasalready suffered losses exceeding $100 billion.Including the loss in stock market wealth –the market’s own estimate arising from ex-pectations of lower corporate profits andhigher discount rates for economic volatility– the price tag approaches $2 trillion.

The more important conclusion from thisarticle, however, is that the full picture has yetto be painted. The ultimate economic cost ofthe tragedy will turn on how successfully pol-icymakers cope with new challenges. On onelevel, they risk investing too little in resourcesthat could minimize the intangible (but verycostly) inconveniences associated with theneed for increased security. On quite anotherlevel, they must deal with the potential forincreased volatility in the business cycle thatcould reduce the economy’s growth potential.The stakes here are simply breathtaking – fail-ure to prevent 1970s-style stagflation couldreduce the GDP by trillions annually.

Indeed, if Washington overreacts to thiscrisis or makes imprudent policy choices, theterrorists will have won their heinous war. M

September - Milken Institute · in stock market wealth Exhibit 4 lists the percentage drop in the...

Documents

Transcript of September - Milken Institute · in stock market wealth Exhibit 4 lists the percentage drop in the...