Alternative Responses of Policy to External Supply Shocks
Total Page:16
File Type:pdf, Size:1020Kb
ROBERT J. GORDON NorthwesternUniversity Alternative Responses of Policy to External Supply Shocks DURING 1973 and 1974 reductionsin suppliesof food (throughnatural causes)and of oil (throughunnatural causes) simultaneously lowered the real incomeof U.S. nonfarmworkers and raisedthe rate of inflation.An inflation-cum-recessioninduced by lowersupplies of rawmaterials may call for a policyresponse different from the traditionaltonic of demandrestric- tion calledfor by a "garden-variety"inflation generated by excessdemand. In lightof the noveltyof the 1974situation, the sharpdivergence of policy recommendationsamong economists is not surprising.Some analyzedthe episodewithin the context of standardmacroeconomic demand analysis, treatingthe 1973-74acceleration of inflationas a delayedconsequence of the accelerationin monetarygrowth during 1972, and the 1974-75recession as a delayedconsequence of the sharpdeceleration in monetarygrowth that beganin June 1974.The policy advice of this group, consistinglargely of economistsgenerally identified as "monetarists,"was to maintaina con- stantor even slightlyreduced rate of growthof the moneysupply.' Arthur Note: This paper was supportedby National Science Foundation Grant GS-39701. It was inspired,as was a previous paper in another area, as an attempt to reconcilethe views of Milton Friedman and Arthur Okun. I am grateful to Michael Parkin and participantsin the Brookingspanel for helpful suggestions. 1. See Allan Meltzer,"A Plan for SubduingInflation" (a dialoguebetween Allan H. Meltzerand two editorialstaff membersof Fortune),Fortune, vol. 90 (September1974), pp. 112ff.In the same month, when the money supply (MI) had risen 5.8 percent over the preceding twelve months, Milton Friedman wrote: ". until a few months ago at 183 184 Brookings Papers on Economic Activity, 1:1975 Okunput forth the contrastingview that an attemptby policymakersto maintainfixed growth in nominalincome ignored the "macroeconomicex- ternalities"of commodityshortages: total real output falls by more than the declinein farmoutput, through an extrainduced loss of nonfarmout- put.2An implicationof Okun'sargument is that, whilestabilization policy cannotre-create the lost farmoutput, it can minimizeor eliminatethe in- duced loss of nonfarm output by promoting a higher growth rate of nominal income. The inflationin 1973and 1974can be regardedas a combinationof an underlying"hard-core" inflation, inherited from the 1960s and perhaps aggravatedby the rapid pace of economicexpansion between 1971 and 1973,with a set of four temporary"bubbles": (1) the 1972-74shortfall of farmsupplies to U.S. consumers,caused in the firsttwo yearsby buoyant foreigndemand and in the thirdby domesticsupply shortages; (2) the re- strictionof oil productionenforced by the cartel of the Organizationof PetroleumExporting Countries (OPEC); (3) the end of priceand wagecon- trols in 1974; and (4) the devaluationsof the dollar in 1971 and 1973. Althoughthese events may have permanentlyraised the pricelevel, such a one-shotrise generatesonly a temporaryincrease in the rate of inflation.3 This paperdeals with the issues raisedby an inflationinitiated not by excessdemand but by commodityshortages. Although its formalanalysis treatsan externalshock that takesthe form of a declinein farmoutput, its basicconclusions apply with only minorchanges to the casesof oil and de- valuation.What policies are availableto minimizethe indirecteffects on output?What are the conditionsunder which expansivepolicy actions best, these high interestrates have been accompaniedby extremelyhigh rates of mone- tary growth.... Recent rates of monetarygrowth are not too low. If anythingthey are still too high to bring inflation to an end in a reasonableperiod of time." See Milton Friedman,"Is MoneyToo Tight?"Newsweek, vol. 84 (September23, 1974),p. 82. Fried- man'sstand on monetarypolicy was takendespite his recognitionthat specialfactors had contributedto the 1974 inflation.He attributedroughly half of it to increasesin oil and food prices, to the lifting of price controls, and to precautionaryincreases against re- newed price controls. See Milton Friedman,"Inflation Prospects," Newsweek, vol. 84 (November4, 1974), p. 84. 2. ArthurOkun, "Incomes Inflation and the Policy Alternatives,"in "The Econo- mists Conferenceon Inflation, September5, 1974, Washington,D.C.; September23, 1974,New York, New York," vol. 1, "Report"(1974; processed),pp. 365-75. A formal analysisof the externalityargument is presentedbelow. 3. The list could perhapsbe expandedby two smallerbubbles-the increasesin prices in fear of reimpositionof controls, and the overshooting of commodity prices beyond the levelsjustified by shortagesdue to speculativeinventory hoarding. Robert J. Gordon 185 takento counteracta temporarydecline in farmoutput will causea perma- nent increasein the rateof inflation?What are the relativeadvantages and disadvantagesof income-taxreductions, food subsidies, and expansive monetarypolicy as policyresponses? Finally, how woulduniversal escala- tion (or "indexation")of wage contractsaffect the resultsof the analysis? The PolarCases To establishthe range of possibilities,the followingtwo sectionscom- pare the responsesof two hypotheticaleconomies, one with perfectflexi- bility of pricesand wagesand the other with absoluterigidity in the non- farm sector. These cases serve to illuminatethe more complicatedand relevantanalysis of a realisticeconomy in whichnonfarm prices and wages are neitherperfectly flexible nor absolutelyfixed. PERFECT PRICE FLEXIBILITY Theeconomy encounters no problemsin adjustingto an externalshock- say, a crop failure-if both farm and nonfarmprices and wages are per- fectlyflexible. In this case the marketfor nonfarmgoods and labor always clears,and no involuntaryunemployment can arise.A briefexamination of this case servesas a point of comparisonwith the diametricallyopposite case of fixed prices. The treatmentof all cases incorporatesseveral common assumptions. The economyis closed, with all outputof both sectorsproduced and con- sumedin the domesticeconomy. Farm output is exogenous,produced by a factorthat is not mobilebetween the two sectorsand consumedentirely in the nonfarmsector. The exogenoussupply of farm output, QFI is equated to the demand: = (1) QF AN (PN') whereA is a constant,ao is the nonfarmincome elasticity of demandfor farmproducts, a, is the absolutevalue of the priceelasticity (which through- out the paperis assumedto be less than unity),and PF andPN are, respec- tively, price indexes for farm and nonfarmoutput. A rearrangementof 186 BrookingsPapers on EconomicActivity, 1:1975 (1) relates the market-clearingrelative price, PF/PN, to the exogenous supplyof farmoutput and the level of nonfarmoutput, QN: (2) PF= [AQN } Ni QF For any given supply of farm products,an increasein nonfarmoutput raisesthe demandfor farm products,and hence the relativeprice, by an amountthat dependspositively on the incomeelasticity, ao, and negatively on the priceelasticity, a,. The relativeprice depends, in part,on the level of nonfarmoutput, except in the specialcase of a zero incomeelasticity. Nonfarmoutput is assumedto be producedwith labor and some other fixedfactor, like capital.Knowledge and technologyis assumedfixed, so that labor inputdetermines nonfarm output. Given the population,if the supplyof labor does not respondto changesin the real wage, both labor inputand nonfarmoutput are fixed. In this case, a crop failurechanges the relativeprice of farmproducts but not the level of nonfarmoutput. Since the wagerate that nonfarmfirms can affordto pay to a given numberof workersis limitedby nonfarmprices, any increasein the relativeprice of farmproducts reduces the real wage of workers,when the latteris defined in terms of a consumerprice index includingboth farm and nonfarm products. If, however,a lowerreal wage causes workers to reducetheir labor input, eitherby withdrawingfrom the labor force or by workingfewer hours per week, a crop failuremust reducenonfarm output.4 This responsein the nonfarmlabor marketthus providesa second relationshipbetween non- farmoutput and the relativeprice of farmproducts, in additionto equation (2) above, allowing the simultaneousdetermination of both variables.5 Hence,output and relativeprices in each sectorare beyondthe control of 4. A third case, not discussedhere, is a negativelysloped labor supply curve. Most cross-sectionevidence for the United States appearsto supporta verticalcurve for adult male workers,a positively sloped response of women and teenagersto an increase in their own real wage, and a negativeresponse of wives to an increasein their husbands' real wage. See the evidence cited in Robert J. Gordon, "The Welfare Cost of Higher Unemployment,"BPEA (1:1973), table 2, p. 159. 5. The exact form of the second relationshipis zkPF~~~~_ 0-(-1) /(b+e) V PN / whereD is a constant, k is the share of farm productsin consumerexpenditures, and b and e are, respectively,the elasticitiesof the nonfarmlabor demandand supply curves. RobertJ. Gordon 187 policymakers.If the choices of individualsbetween leisure and labor are sociallyaccepted, any reductionin employmentcaused by the voluntary withdrawalof laborinput in responseto a lowerreal wage is of no concern for stabilizationpolicy, sincethat reductionis purelyvoluntary. What, if anything,can stabilizationpolicy accomplishwhen nonfarm pricesare perfectly flexible? Aggregate-demand policy controlsthe level of nominalincome (that is, grossnational product in currentdollars), which