Public Disclosure Authorized
Public Disclosure Authorized
RESEARCH WORKING PAPER
*@
An Empirical Model
A
of Sunk Costs
and the Decision to Export
Mark J. Roberts
JamesRPTybout
Public Disclosure Authorized
Public Disclosure Authorized
POLIcy
The World Bank
IntermationalEconomics Department
International Trade Division
March 199S
L
;
r~
|POLICY
RESEARCH
WORKING PAPER 1436
Summary findings
Exports respond unpredictably to a change in real
exchange rates, suggestsevidence from the 1980s.
Recent theoretical work (Paul Krugman and Richard
Baldwin) explains this as a consequence of the sunk costs
associatedwith breaking into foreign markets. Sunk costs
include the cost of packaging,upgrading product quality,
establishing marketing channels, and accumulating
information on demand sources.
Roberts and Tybout use micro panel data to estimate a
dynamic discrete-choicemodel of participation in export
markets, a model derived from the Krugman-Baldwin
sunk-cost hyster --is framework.
Applying the model to data on manufacturingplants in
Colombia (1981489), they test for the presence of sunk
entry costs and quantify the importance of those costs in
explaining export patterns. The econometric results
reject the hypothesisthat suniccosts are zero.
The results, which control for both observed and
unobserved sources of plant heterogeneity, indicate that
prior export market experience has a substantial effect
on the probability of exporting, but its effect depreciates
fairlyquickly.The reentry costs of plants that have been
out of the export market for a year are substantially
lower than the costs of a first-timeexporter. After a year
out of the export market, however, the reentry costs are
not significantlydifferent from the entry costs.
Plant characteristicsare also associatedwith export
behavior: Large old plants owned by corporations are
more likely to export than other plants.
Variations in plant-levelcost and demand conditions
have much less effect on the profitability of exporting
than variations in macroeconomicconditions and sunk
costs do. It appears especiallydifficult to break into
fereign markets during periods of world recession.
This paper - a product of the InternationalTrade Division,InternationalEconomicsDepartment- is part of a largereffort in
the department to describethe micro foundations of export supply response. The study was fundedby the Bank's Research Support
Budget under the research project "Micro-Foundationsof SuccessfulExport Promotion" (RPO 679-20). Copies of this paper are
availablefreefrom the WorldBank,1818 H StreetNW,Washington,DC 20433. PleasecontactJermiferNgaine, room R2-052,
extension 37959 (37 pages). March 199S.
ThcPoly Rcscb lrorkg PaperS
dates
by
tPelicy
Rs
of
c wr
pDogriem to Cente
5 e
of se about
dae)opt bsum An *beaitnof th sern is
=to et the fid& ow quk*, ecx if thepresatiom areAmssm fbefypolisbed.
7he
papersam7ythe xammof deaSch
2osnd
sboud6be
dandckedoccodigly.7lOc0sq
_drrtt= candcndsi
re the
autbors'oumanbuld
onot be akds
o the WorldBaxk, its Fxeav BoardofDirasors orany ofits cmbr outrk&s
Produccd by the Policy Rescarch Dissemination Ccnter
An Empirical Model of Sunk Costs and the Decisionto Export
Mark J. Roberts
PennsylvaniaStateUniversity
and
JamesR. Tybout
GeorgetownUniversity
We would like to thank DavidCard, Terri Devine, AvinashDixit, Chris Flinn, Zvi Griliches,James
Hanna,JamesHeckman,ArielPakes,DavidRibar, DanWestbrookandan anonymousrefree for helpfil
discussionsor comments. Support for this paper was provided by the Research Administion
Depatmn of The WorldBank (RPO679-20).
SUMMar
The responsivenessof exportsto changesin the incentivestructure has long interestedpolicy
makers.But the empiricalliteraturehas providedlittle guidanceon whenor how exporterswill respond
to new incentivestructures. Researchin this area has produceda varietyof supplyelasticityestinates
that vary dramaticallyacross countriesand time periods, and few hintson how to reconcilethe diverse
results.
PaulKrugmanandRichardBaldwinhaverecentlyarguedthattheemirical literaturefailsbecause
there are sunkcostsassociatedwithbreakinginto foreignmarkets- includingupgradingproductquality,
packaging,andthe establishmentof marketingchannels. Hencethe currcnt-periodexportsupplyfunction
dependsuponthe numberandtype of producersthat wereexportingin previousperiods. Further,startup costsmean that transitorypolicychangesor macro shockscan leadto permanentchangesin market
structure,andthus that tradeflowsmay not be reversedwhena stimlus is removed.That is, sunkentry
or exit costsproduce "hysteresis'in trade flows. Finally,wlhenfuturemarketconditionsare uncertain,
sum,.costsmakepattrns of entry and exitdependentupon the stochasticprocessesthat governvariables
like dheexchangerate.
Taking the Baldwin/Krugmanperspective as a point of departure, this paper develops an
econometric model of a plant's decision to export. The model is fit u
mwiro
data for a large group of
anufacturingplants in Colcmbiafrom 1981-1989,and used to direcdyexaminethe determinantsof a
plant's exportdecisionfor consistencywith the theory.
The econometricresultsrejectthe hypothesisthat sunkcostsare zero. Theyalso revealthat the
re-entrycostsof plants that have been out of the exportmarketfor a year are substantiallyless than the
costs of a first-time exporter. Beyond a one year absence, however, the re-entry costs are not
significantlydifferentthan those faced by a new exporter. This is consistentwith the view that an
importantsource of sunk entry costsfor Colombianexportersis the need to accmulate informationon
demandsources, informationthat is likelyto depreciateuponexit from the market.
Whilethe resultsindicatethat sunk costs are a significantsourceof exportmarketpersistence,
both observed and unobservedplant characteristicsalso contibute to an individualplant's export
behavior. For example,plants that are large, old, and ownedby corporationsare all more likelyto
export.
A numberof policy implicationsemerge. For example,it appearsespeciallydifficultto break
into foreignmarketsduringperiodsof world recession. Also, althoughonly 25 percentof the plantsin
the Colombianpanel exportedduring the sampleperiod, the results imply that sufficientlyfavorable
macro conditionsand/or reductionsm sunk costs couldmake exportingprofitablefor a much larger
proportionof plants. Finaly, andmost generaly, the estimatesimplythat countriesundertaldngexport
promotionpolicies should distinguishmeasures aimed at exanding the export volume of exisitag
exportersfrom policiesaimedat promotingthe entry of new exporters.
I.
Introduction
Why is it that in some countries and time periods, a given trade and exchange rate regime
supports large scale production for foreign markets, while in other countries or time periods, the same
policies appear to induce a minimal export response? Put differently, why are cstimates of export
supply equations so sensitive to the time period or country under study?
In a recent series of papers Richard Baldwin, Paul Krugman, and Avinash Dixit have
proposed an answer.' Thay begin from the assumption that non-exporters must incur a sunk entry
cost in order to enter foreign markets. This mnakesthe current-period export supply function
dependent upon the number and type of producers that were exporting in previous periods. Further,
it means that transitory policy changes or macro shocks can lead to permanent changes in market
structure, and thus that trade flows may not be reversed when a stimulus is removed. That is, sunk
entry or exit costs produce 'hysteresis" in trade flows. Finally, when future market conditions are
uncertain, sunk costs make patterns of entry and exit depeodent upon the stochastic processes that
govern variables such as the exchange rate. Under plausible assumptions, greater uncertainty makes
trade flows less responsive to changes in these variables. None of these implications of sunk costs is
captured in standard empirical export supply functions, and all could contribute to the "instability" of
empirical relationships.2
To date, attepts to empirically validate the sunk-cost hysteresis framework have focussed on
'See, in particular, Dixit, 1989a and 1989b; Baldwin, 1988 and 1989; Baldwinand Krugman, 1989;Krugman, 1989).
2
In their review of empirical studies of price and income elasticitiesfor traded goods, Goldsteinand Khan (1985, pp.
1087-1092)report a very wide range of estimates for the supply elasticity of total exports from developed countries. They
conclude that 'excluding the United States, the supply-priceelsticity for the total exports of a representative industrial
countly appears to be in the range of one to four. The supply elasticity for U.S. exports is probably considerablyhigher
than that, perhaps even reaching ten to twelve.' They also discuss some evidence indicating that the response of export
supply to price changes is slower than demand-sideadjustments. They speculate dh this may refict start-up costs
associatedwith export production or greater unceUity assocated with sHling abroad.
I
asymmetries in the response of trade flows to exchange rate appreciation versus depreciation.3
A
limitation of this approach is that data on the volume of trade flows, even for very disaggregated
commodities, cannot distinguish the entry and exit of exporters from the supply response of
continuing exporters. With the exception of Campa (1993), the foreign market entry and exit
patterns, which are the focus of the theory, have not been exanined for consistency with the sunkcost hysteresis model.'
In this paper we develop an empirical test of the sunk-cost hysteresis model that directly
examines entry and exit patterns in plant-level panel data. We develop and estimate a dynamic
discrete choice model of the decision to export when sunk entry or exit costs are present. In essence,
this model predicts exporting status in the current period as a function of plant characteristics,
previous exporting status and a serially correlated disturbance. It not only permits us to formally test
for the presencc of sunk costs (using coefficients on lagged exporting status), it allows us to
summarize the effects of time, individual producer characteristics, and prior exporting experience on
the probability of participating in the export market. The data we use describe the export patterns of
I The empirical evidence derived from trade flow data has produced no clear consensus. Based on aggregate U.S. data,
Baldwin(1988) concludesftat the substantial appreciationof the U.S. dollar during the early 1980's resulted in a structural
shift in U.S. impon pricing equations. This is consisteantwith sunk-cost hysteresis. In contrast, Gagnon (1987) finds that
trade has been more responsiveto relative prices in the more uncertain post-BrettonWoods era, a result inconsistentwith
some versions of the hysteresis model. Using time-seriesdata for U.S manufacturingindustries. Feinberg (1992) fmds that
exports became more dispersedacross destinationmarkets as the dollar depreciated, suggesting that there was firm entry
into new country markets. The effect was weaker in industrieswhere distribution networks, and thus presumably sunk entry
costs, are more important. Parsley and Wei (1993) focus on bilateral U.S.-Canada and U.S.-Japan trade flows fbr very
disaggregated commodities. They find that both the past history of U.S. exchange rate changes and measures of exchangerate volatility had no significanteffect on trade flows. Both findingsar inconsistentwith the hysteresis model.
' Campa (1993) examines the number of foreign firms that made direct investnents in the 61 U.S. wholesaletrade
industriesover the 1981-87 period. He finds that exchange-rateuncertainty, which is proxied by the stndard deviation of
the monthly rate of growth of the exchange rate, is negativelycormlated with the number of firms investing in the U.S..
He also reports that an industry's sunk costs, which arc proxied by the advertising-salesratio and ratio of fixed assets to net
worth of firms in the industry, is negatively correlated with foreign-firmentry. Both findingsare consistentwith the
hysteresis model. Althoughnot in a trade context, related work by Bresnahan and Reiss (1991, 1994) shows how data on
net entry into a market can be used to make inferences about the ratio of sunk entry and exit costs to avenge profitability.
Their techniqueexploits the asymmetric response of the number of producers to population(demand) changes across
diffierentgeographic markets. However, as they acknowledge,persistance in behavior due to permanent cross-producer
differences in profitability can create the appearanceof sunk costs in their model.
2
Colombianmanufacturingplantsin four majorexportingindustriesover the period 1981-1989,a nineyear span characterizedby substantialchangesin aggregatedemandand real exchangerates.
The empiricalresultsstronglyrejectthe hypothesisthat sunk costs are zero. This impliesthat
prior export marketexperiencesignificantlyaffectsthe currentdecisionto export. Further, although
recentexperiencein foreignmarketsis extremelyimportant,its effectdepreciatesfairly quicklyover
time. A plantthat exportedin the prior year is up to 40 percentagepointsmore likelyto exportin
the current year han an otherwisecomparableplantthat has neverexported. But by the time a plant
has been out of the exportmarketfor two years its probabilityof exportingdifferslittle from that of a
plant that has never exported.
Severalpolicy implicationsemerge. First, althoughonly 25 percentof the plantsin our panel
exportedduringthe sampleperiod, our resultsimply that sufficientlyfavorablemacro conditions
ador reductionsin sunk costs couldmake exportingprofitablefor a much larger proportionof
plants. Second,our estimatesimplythat countriesundertaldngexportpromotionpoliciesshould
distinguishmeasuresaimedat expandingthe exportvolumeof exisitngexportersfrom policiesaimed
at promotingthe entry of new exporters.
In the next sectionof the paper we summarizethe theoreticalsunk-costmodel. The third
sectionprovidesan overviewof the patternsof exportparticipationamongColombianman
uri
plantsbetween1981and 1989. The fourthsectiondevelopsan econometricmodelof the export
decision,and the fifth sectionpresentsour results. We brieflysummarizeand draw conclusionsin
the sixth section. Readersuniterested in methodologicalissuesmay wish to skip sectionE and
readers uninterestedin econometricproblemsmay wish to skim section1".
I.
A Theoretical Model of Entry and Exit with Sunk Costs
As reviewedin Krugman(1989), sunk costs affectthe exportsupply functionfor several
3
reasons. First, and most obviously,once the sunk costs of enteringa markethave been met, a
producerwill remainin that marketas longas operatingcosts are covered. This impliesthat changes
in policy,exchangerates, or prices in foreignnmarkets
can permanentlyalter marketstructureand
thusobservedexportbehavior. For example,devaluationsthat induceentry into the exportmarket
may permanentlyincreasethe flowof exports, even if the currencysubsequentlyappreciates.
Second,even if currentconditionsappearfavorableto exporting,they may not induceentry into the
export narket if they are regardedas transitory. In this case, the expectedfuturestreamof operating
profitsmay not coverthe sunk costsof enteringforeign narkets. Thuslarge devaluationsmay induce
little responsefrom potentialexportersif they are perceivedas transitory. Finally,as formally
demonstratedby Dixit(1989a),the combinationof sunk costs and uncertaintyaboutfuturemarket
conditionscancreate an optionvalueto waiting. Dixit's simulationresdts suggestthat even small
amouts of uncertity can significantlymagnifythe degreeof persistencein a producer's expordng
status.
To motivateour empiricalwork, we beginby reviewingthe theoreticalmodelsthat generate
theseresults(see footnote 1 for references). For each periodt, le, the ti plant's expectedgross
profitswhen exportingdiffer from its expectedgross profitswhen not exportingby the amount
ir,(o,sJ. Herep, is a vector of market-levelforcingvariablesthat the planttakes as exogenous(e.g.,
the exchangerate), and s,, is a vectorof statevariablesspecificto the plant (e.g., capitalstocks and
geographiclocation). Once in the market,plants are assumedto freely adjustexport levelsin
responseto currentmarketconditions(Baldwin,1989). Thus the functionT#,,Sd representsthe
incrementto expectedprofitsassociatedwithexportingin year t, assumingthat the profit-maximizing
level of exportsis alwayschosen.
Theseprofitsare gross becausethey have not been adjustedfor the sunk costsof foreign
mariet entry or exit. Assumethat if the i* plant last exportedin year t-j (j 2 2) it facesa re-entry
4
cost of F , so uponresuning exportsin year t it earns r1(plsd - Fl . Similarly,If the plant had
never exportedpreviously,it facesan entrycost of
P57
and earns ir1(p,,sJ- P7 in its first year
exporting. Finally, a plant that exportedin periodt-1 earns r,(p,,sd)duringperiodt by continuingto
export and -X, if it exits. As in Dixit (1989a),thesesunk costs representthe direct monetarycostsof
entry and exit. The j superscriptgeneralizespreviousmodelsto allowsunk re-entrycosts to depend
on the length of absencefrom the market. This couldreflectthe increasingirrelevanceof the
knowledgeand experiencegainedin earlier years,or the increasingcost of updatingold export
products. The i subscriptallowssunk coststo vary acrossplants with differencesin size, location,
5
previouscxperience,and other plantcharacteristics.
To collapsetheseearningspcwsibilities
into a singleexpression,definethe indicatorvariable
Yi to take a valueof 1 if the plant is exportingin period t, and 0 otherwise. Also, let the exporting
historyof the plantthroughperiod t be givenby Yki,= (Y,, I j=O...J 3, whereJ1 is the age of the
plant. Tbenperiodt exportingprofitsare:
Rsd2i;') = r
where P, f
p
=
exportingexperience:
-
(Y,
(I
,
F°(IY)
*)
=
-
.Y
S
(Fi-F ?)I,-ff]- X,Y,,I(1-YY)
1
This last expressionsummarizesthe plant's most recent
1I whenthe plant's most recentexportng experienceoccurredj years,
earlier and 0 otherwise.
In periodt. managersare assumedto choosethe infinitesequenceof values 11, = (Yk,+,I
i
2 0) that maximizesthe expectedpresentvalueof payoffs. In period t, this maximizedpayoffis:
5
To keep ie notation ttable we have not added a tiwe bscrit to entry and exit costs. Inbe emp
section, we
will st whether hy vary over ime, as would be expected if there are changes in credit market condions or trade policis
that affect accessto frig
markets.
S
j
of
max E,
6- 1R
RJ, Q>
where6 is the one-perioddiscountrate and expectationsare conditionedon the plant-specific
informationset, 0 1, UsingBellman'sequation,planti's currentexportingstatuscan be represented
as the Y,,valuethat satisfies:
vt,(Od =
max(R
1 ,( )e.) + 5E,{V...(Q,, 1 .)')
whereE, denotesexpectedvaluesconditionedon the informationset 0,,. From the right-handside of
this expression,it followsthat the iuhplant will be in the exportmarketduring periodt if:
IAStsfi} + 5[E1(V,(fl,. 1) 1Y=1) - E,(VZ.1 (O.1 ) I4=0)]2
(1)
>
F° -
(F 0
.+Xi) Y,,- +
(Fill-F, ) Y,
j
J-2
where-fl;: + X,) is the sumof sunk entrycosts for a plant that neverexvported
and exit costsfor
current exporters,sometimesreferredto as the "hysteresisband" (Dixit, 1989a).
Equation(1) providesthe participationconditionthat will be estimatedin sectionV. It has
severalempiricalimplicationswe will pursue. First, if there are no sunk costs, the participation
conditioncollapsesto ri(p,si) > 0. Hence one can test the sunk-costhysteresisframeworkby asking
whether,given a plant's current gross profits, its exportinghistoryhelps explainits currentexporting
status. Second,if sunk costsdo matter, equation(1) impliesthat they appeardirectlyin each plant's
participationconditionas coefficientson binary variablesthat describeits exportinghistory.6 Hence
the magnitudeof sunk costsand the rate at whichpast experiencedepreciatescan be identified.
6The influenceof sunkcostsaLsocomesthroughthe expectedvalueterm on the left-handside, but sincethis is a nonlinearexpression.coefficientson die indiaor variablesare idtified.
6
Finally,this equationindicatesthat realizationson the variablesp, and 3s, influenceexportdecisions
throughtheir effect on w(p,,sdand their effect on the expectedfuturevalue of becomingan exporter
now. Th1islattereffect implies,for example,that exchangerate movementsthat managersconsider
transitorywill generallyhave less effect than equivalentmovementsthat are viewedas long-term
regimeshifts.
m.
The Pattern of Export Participation in Colombia
Beforediscussingestimationissuesand results, it is usefulto introduceour data base, review
the exportenviromnentin Colombia,and providesome aggregateevidenceon the patternof export
maiket participationduringthe 1980s.
The Data: The analysisin this paper is based on annualplant-leveldata collectedas part of
the Colombianmanufacturingcensusfor the years 1981-1989.This census,whichcovers all plants
with 10 or more employees,providesinformationon eachplant's geographiclocation,industry,age,
ownershipstructure,capitalstocks,investmentflows, expenditureon labor and materials,valueof
outputsold in the domesticmarket,and value of outputexported. We have matchedthe individual
plant observationsacrossyears to form a panel.7 The data are particularlywell-suitedto analyzing
exportmarketparticipationbecausethey allowus to observetransitionsof individualplantsinto and
outof the export marketand to controlfor some importantobservableplant characteristicsthat are
likelyto affect the export decision.
The PolicyRegimeand Export ParticpationRate: Table 1 illustratesthe basicpaterns
7Thecensusdataand matchingprocessare discussedin greak, detailin Roberts(1994).
7
over the sampleperiod for the 19 major exportingindustriesIn the Colombianmanufacturingsector.'
In generaltenns, the macro environmentin Colombiawas not conduciveto profitableexportingof
manufacturedgoods in the early 1980's. Respondingto illegal exports,foreigncapital inflows,and a
boomin the coffce market, the Colombianpeso appreciatedsteadilybetweenthe mid-1970'sand
1983. As shownin Table 1, this patternwas reversedaftr" 1983, with the currmncylosing
approximatelyone-halfof its valueby lY89. This partly reflectedcentralbank currency market
interventionsto ease competitivepressureson tradeablegoodsproducers.
The time-seriespatternof manufacturedexportslargelymirrors this movementin the
exchangerate. The real value of manufacturedexportsfrom the nineteenmajor exportingindustries
declinedslightlyfrom 109.6 (billion 1985pesos)in 1981to 95.9 in 1984, for an averageannual
growthrate of -4.45 percent. F-rom1984to 1989the pattern reversedand the quantityof exports
grew at an annualaveragerate of 21.1 percent."
Commercialpolicyshelteredimport-competing
producersthroughoutthe sampleperiod.
Substantialtariffbarriers werereducedslightlyafter 1984,but quantitativerestrictionson the imports
of productsthat competedwith domesticindustrieswere maintained. In additionto turning the terms
of trade againstexporters,these policesmadeit moredifficultto importraw materialsor capital
goodsthat may have been necessaryto increasethe qualityof manufacturedproducts.'"
9 The nineteenindustriesand their SITCcodesare: foodprocessing(311/312),textiles(321),clothing(322).leathr
products(323/324).
paper(341), printing(342). chemicals(351/352),plastic(356).glass(362). non-metalproducts(369).
ironandsteel (371).metal products(381),nuchinery(382/383),transportation
equipment(384),and miscellaneous
nianuficuring(390). Theseindustriesaccountfor over 96 percentof Colombia'smanufacturing
sector exportsand 85
percentof manufacturingoutputin each sampleyear.
' The real valueof exportsis measuredas the pesovalueof exportsdeflatedby an exportprice indexfrom the IMF
InternationalFinancialStatistics. This measurewill overstatethe dependenceof the physicalv.;:umeof exportson the
exchangerate becauseof valuationeffects.
"0 The long-termpromection
of the domesticmarketfrom importcompetitionappearsto havealo contributedto the low
productqualityand low productivitythat have madeit difficultfor Colombianexportersto compesein the intentional
market(WorldBank, [992).
8
Nonetheless,the bias towardimport-competing
activitieswas partly offset by exportsubsidies,which
increasedrelativeto the value of exportsby approximately50 percentbetween 1983and 1984, and
thereafterdeclined." The incentivesto exportcreatedby exportpolicy thereforewere counterto
those createdby exchangerate movements.
The net effect of these changeson the numberof exportingplants and the proportionof plants
that exportedis summarizedin the last two rows of Table 1. Again,the time-seriespattem largely
reflectsthe movementin the exchangerate. Through1984,there was net exit fromthe export
market, and a declinein the proportionof plantsexporting. After that year there was net entry and a
steadyincreasein the participationrate. There is, however,some evidenceof asymmetryin the
magnitudeof the response. The modest6 percentreal appreciationbetween 1981and 1983was
accompaniedby a declinein the exportparticipationrate from .129 tc .113, but a muchlarger (45
percent)depreciationbetween1984and 1989servedonly to increasethe participationrate to .135.
From this short timeseries it appearsthat it tookboth a substantialand persistentdevaluationto
induceentry into the exportmarket. This is consistentwith the conjecturethat potentialexporters
facedsubstantialstart-upcosts.
Surveyevidencefurther supponsthis hypothesis.'" First, to sell in developedcountry
markets,Colombianproducersweraoften requiredto investin productqualityupgrading. Second,
there was little exportinginfrastucturein the form of tradingcompaniesor distributionagents. These
companiestypicallyprovidetransportation,customs,and shippingservices,as well as informationon
" This reductionin exportsubsidiesreflecteda reductionin two governmentprogramsusedto promoteexports. The
firstprogramrebatescustomsdutiespaid on importedmaterialsand capitalequipmentfor plamsthat export. In 198041
percentof the valueof exportscamefrom plantsthat receivedrebates. This rose to 62 percentin 1984andfell to 53
percentin 1986. The secondprogramprovidesdirectsubsidiesto exportersbasedon the valueof their exports. The
averagerate of subsidyincreasedfrom7.6 percentof the valueof exportsin 1981to 15.0percentin 1985andtben fell to
8.7 percentin 1986.
I2 lThediscussionin this sectionis basedon WorldBank(1992),whichsummarizesand interpretsinterviewswiththe
managersof severalhundredColombianplants.
9
prices, potentialbuyers, and productstandardsor requirementsin other countries. The absenceof
thesemiddlemenprobablydiscouragedpotentialexporters,both by increasingthe informationcosts
they faced, and by increasingthe degreeof uncertaintyconcerningforeign marketconditions.
Apparently,however,the lackof a well-developedtradingservicessector did not affect all producers
equally. Exportersableto deal in large volumesor to ship to large marketswere relativelyless
constrainedby the absenceof tradingintermediariesbecausethey were able to sell directlyto final
buyers. This findingsuggeststhat sunk costsrose less than proportionatelywith exportvolume.
Export marketentry was also inhibitedby institutionalfactorsthat affectedexpectedprofits.
Notably,a surveyof Colombianfnancial institutionsrevealedthat none were willingto lend money
againstexportorders or lettersof credit from purchasers'banks. Lendersattributedthis unusually
conservativepracticeto their inabilityto judge whetherthe potentialborrowerscould seriously
competeabroad. It mainlyhurt first-timeexportersand existingone-product,one-countryexporters
attemptingto enter newcountryor productmarkets.
Finally, as emphasizedby Dixit (1989a),regimeuncertaintymay have inducedproducersto
delay entry into the exportmarket, even after substantialdevaluation. In the World Banksurvey,
producersciteduncertaintyaboutthe permanenceof the changein trade and exchangerate regimesas
incentivesto delay or foregoentry into the exportmarket. Theirmain concernwas apparentlythat
lobbyistsin favor of protectingdomesticindustrieswouldbe able to reversethe trend towardtrade
liberalization.
In summary,duringthe sampleperiod manyColombiannufacturs
viewedthe export
marketas more risky and less profitablethan the domesticmarket. The lack of a tradingservices
sector, accessto financing,and low productqualityall appearto have constrainedexportmarL:t
participationby raising the costsof entry or increasingthe uncertaintyof the profitabilityof
exporting.
10
Entry and Exit in the Export Market: The analytical model reviewed in Section II implies
that this combination of sunk costs and uncertainty should induce persistence in producers' exporting
stawus. That is, those who have already incurred the sunk start-up costs should be relatively likely to
export in the current period. Some preliminary evidence on this prediction is provided by transition
rates into and out of the export market, which are summarized in Table 2. Each row describes a
transition from the exporting status in column I to the status in column 2. The entries in the table are
the proportion of plants in each of the period t categories that choose each of the two possible
categories in year t+1.A3 The top panel applies to the 19 major manufacturing industries, and the
bottom panel applies to the 650 plants in the four major exporting industries - food, textiles, paper,
and chemicals - that will be used to estimate the econometric model in the next section.
The top row of each panel indicates that, of the plants that did not export in year t, more han
95 percent of them did not export in year t +1. For the plants initially in the export market, the
proportion of manufacturing plants that remain in the market from one year to the next varies from 83
percent to 91 percent over tine, and the proportion of plants in our four-industry subsample that
remains in the market varies from 85 percent to 95 percent. Clearly, there is substantial persistence
in the plant-level pattems of export market participation. Nonetheless, only 36 percent of the plants
in the subsamplethat exported at some time remained exporters for the whole sample period, and
among plants that did change exporting status, 60 percent did so more than once.14
Persistence in exporting status might be caused by sunk costs, as the hysteresis models
suggest. Alternatively, it might be caused by underlying plant heterogeneity: persistent cross-plant
'3 The datacorrespondto the group of plantsdtat were io operationin eachyear 1981-1989.Thereare 2369plants;in
this groupand theyrepresemapproximately
40 percentof thenumberof piano in operadonin any year. On average,18.1
percentof thesepla participatedin the exportmarketin any singleyear and theyaccountedfbr 62.3 percentof thetotal
numberof exportersand 61.6percentof thevalueof manufactured
exports.
'' The appendixsummarizesthe patems of multipleswitchesinto and out of the exportmarketand the abilityof the
empiricalmodelto explainthese patrns.
'11
differencesin the payofffrom exporting,rf(*),wouldexplainwhy someplants are alwaysin the
export marketand othersare alwaysout. Similarly,the fact that manyexportersenter or exit the
marketmultipletimes can also be interpretedseveralways: it couldmean that sunk costs are small,
or it couldreflectlingeringbenefitsfrom havingexportedrecently. In the next sectionwe developan
econometricframeworkthat can discriminateamongthesecompetingexplanations.
12
IV.
An Empirical Model of Export Market Participation
The Estinating Equation: Our empirical model of a plant's exporting decision begins with
the participation condition given by equation (1). Define
r5 =p,,sd)
+ [E,(V,,(Q,,.,)IYi,=l) - E,(Vf,(Q,,..)I Yj,=0)]
as the latent variable representing the expected increment to gross future profits for plant i if it
exports in period t. Export market participation is then summarized by the dynamic discrete choice
equation:15
_) YiI
if
ir,; - Fi° + (Fe +X)Y
+ E (F,` -FiJ)Y,rj > 0
0 otherwise
There are two ways we might proceed to estimate equation (2). First, we could develop a
structural representation of the participation condition by making specific assumptions about the form
of the profit function and the processes that generate s, and p," Alternatively, we could forego
identification of structnral parameters, and approximate ir', - F' as a reduced-form expression in
exogenous plant and market characteristics that are observable to producers in period t. The
advantage of the first approach is that, in principle, it allows identification of the parameters of the
profit function (inter alia) and provides a complete description of the dynamic process. Its main
disadvantage is that very restrictive parameterizations are required to make structural estimation
feasible. This problem is particularly acute in our model because the dependence of sunk entry costs
upon the length of time out of the export market implies a participation series that is a t-order
This equation is similar to those used to study labor market participationand employment (Heckman, 1981a, 1981b).
T5
"Eckstein and Wolpin (1959) and Rust (1993) summarize de literature on esiimating strucmral dynamic models of
discrete choice.
13
Markov process. Because of this difficulty, and because we do not need a structural model to assess
the role of sunk costs or to investigate the sensitivity of decisions to s,,and p, we pursue the reducedform approach.
To parameterize the reduced-form model, we assume that variation in ir;, - F: arises from
three different sources: time-specific effects that reflect industry or macro-level changes in export
conditions (A.i),observable differences in plant characteristics (Zi,),and noise (es,):
(3)
x,,- F
= , + ,
+ e,
The term jt, is an annual time effect reflecting temporal variations in export profitability and start-up
costs that are common to all plants. These timneeffects pick up the influence of credit narket
conditions, exchange rates, trade policy conditions, and other time-varyingfactors captured by p, in
the analytical model. The vector 4, controls for factors represented by s,, and F, in the analytical
model: exogenous plant-specific determinants of current operating profits and start-up costs. It
includes a constant, a set of industry dumnmiesdefined at the three-digit SITC level, a dumnnyvariable
to control for the ownership structure of the plant (proprietorship and partnership versus corporation),
and a set of two locational dunmiies to distingish the Bogota and Medillin/Cali regions from all
others.'" The vector Z, also includes several continuous variables lagged one period and measured
in logarithms: the ratio of foreign to domestic prices for output, the wage rate, capital stock, and
plant age." Relative prices and wages affect the atractiveness of domestic versus foreign markets.
Capital stock and age proxy for efficiency: in addition to scale effects, studies of industrial evolution
' The base group for comparison is a plant in the food industry that is not a corporation and that is located in the
Bogotaarea.
1 FPoreign
prices are constructedusing unit values of exports at the four-digit ISIC level; domestic prices were obtamied
at the same level from die Central Bank of Colombia.
14
suggest that efficient producers are more likely to survive and grow.
*
Additional restrictions on sunk entry and exit costs are needed to identify the model. Let -y;j
=F
- F'
(j =2,...,J
) and 'y/
= -y = F:
+ X, (j 2 J4-l),
implying that experience is
completely depreciated if it was acquired more than J years ago. (The problem of choosing J will be
discussed later). Further, let -Y/ =y j, implying that cross-plant variation in sunk costs is negligible
among producers who acquired their most recent exporting experience at die same time. Then
substituting (3) into (2), we obtain our basic estimating equation:
J
I
(4)
Y
if ° 5 /I, + jZ, +
E 1+Y
yajy
{
0
j-2
otherwise,
Properties of the disturbance term - will be discussed in the following subsection.
As noted earlier, the participation decision does not depend upon exporting history if sunk
costs are zero. Hence we can test the null hypothesis that sunk costs are unimportant in the export
decision by testing whether & and y 's are jointly equal to zero. If they are significant, we can use
them to make inferences about the rate at which export market experience decays. Using interaction
terms between the lagged participation variables and plant characteristics or macro variables, the
model can also be generalized to allow the sunk cost parameters -V'sto vary with changes in these
variables. Finally, we can use equation (4) to study the importance of temporal (s,
)
and cross-plant
(J4) variation in net expected profits from exporting (r;, - F,). In particular, we can impute
probabilities of entry or exit in response to a given shift in exogenous variables for plants with
different characteristics.
Econometric Issues: To isolate the inmportanceof sunk costs, it is critical that we control for
15
all other sources of persistence in exporting status. Muchiof this task is accomplishedby including
the vector of observable plant characteristics Z,, in equation (4). However, it is very likely that some
characteristics, such as managerial expertise or output quality, will remain unobserved and their
presence will induce serial correlation in the error term, e,,. If we use an estimator that ignores this
serial correlation, the model will incorrectly attribute its effect on exporting status to past participation
and thus overstate the importance of sunk costs.'9
Following Heckman (1981a, 1981b) we allow for serial correlation by assuming that e; is the
sum of a permanent. plant-specific component and a white-noise component: eit
a, +
(it.
Here ai
represents unobservable plant-level differences in managerial efficiency, foreign contacts, and other
factors that induce persistent plant-specific differences in the returns from exporting. We normalize
var(c4f)= 1, and assume that cov(aj,a,) = 0 v i*k, cov(Z,e,) = cov(aj,wjf)= 0 v i, t, and
cov(caj,,kso.j)
= 0 v j+0.
This specification implies that cov(ejt,e.,j)= var( 4.) =
a2,
so the parameter
o. is both the covariance between different time periods for a single plant, and the fraction of the
variance of e*,that arises from the permanent component in the error. Assuming that a and w are
each normally-distributedrandom variables, equation (4) can be estimated as a random-effects probit
model.20
There remains an additional problem. We observe a plant's export status in years 1 through
T, and our lag structure reaches back J periods, so equation (4) can be used to model the export
19This is the problem of -spurious state dependence"discussed in the empirical literature on labor market
participation. See. for example, Heckman (1981a).
20 We do not control for the a, by using plant-specificdummy variables because of the "incidentalparameters problem"
discussed in Neymanand Scon (1948), Chamberlin (1980), and Heckman (1981c). For a given number of time periods, the
number of a, values grows in direct proportion to the sample size, makingconsistent estimation as n-eo more difficult.
Under these conditions,a standard logit or probit estimator using plt-specific dummy variables will not yield consistent
slope coefficients. If the time dimension of the panel is small andlor the model is dynamic. the bias can be substantial. See
Hsiao(1986. pp. 159-161). Wright and Douglas (1975) and Heckman (1981c), for discussion of the magnitudeof the bias.
In particular, Heckman (198ic) finds that the bias in slope coefficients from a dynamicprobit with unobservableeffcs is
"disturbingly large" (p. 180) when T = S.
16
decisionin years J+I throughT. But Y,.,and Y,,_Jvalues correspondingto thesefirst J years camot
be treated as exogenousdeterminantsof Y4(J+1 < t 5 21)becauseeach dependson a,. Heckman
(1981c)suggestsdealingwith this "initialconditio'is"problemby using an approximaterepresentation
for Yi,when t < J and allowingthe disturbancesin the first J periodsto be corre!atedwith the
disturbancesin every other period. Specifically,supposethat expectedprofits in the exportmarket
duringperiods I throughJ can berepresentedwith the equation:
(S)
i ;, - I
= X Zi, + soj,
t =l,
. . .J
where7; is a vector of pre-sampleinformationfrom periodsprior to t on the fh plant, Ais a
conformableparametervectorand the randomvariablejp, is assumedto be normallydistributedwith
var(,o;,)= I, cov(so,,w,,)=O.2
Critically, we also let (i be correlated with the persistent plant
component of the error term f 2: cov(fp.,a,) = p. (Note that p can also be interpreted as the fraction
of variationdue to a; in the first J periods.) Then the correlationof laggeddependentvariableswith
a, in years 1 throughJ can be recognizedin the likelihoodfunctionby representingthe Y1,
process
usingequation(5) for periodst=l,...,J,
and equation(4) for periodst=J+I,...,T.
This specification
addsthe nuisanceparametervector X to the model as well as the parameterp. Althoughequation(5)
is an imperfectrepresentationof the processgeneratingthe data, simulationevidencesuggeststhat
Heckinan'sprocedureperformsreasonablywell.'
To constructthe likelihoodfunctionfor observationsin periods 1 throughT define:
21
In the empirical work we include all of the plant characteristics in Z7described above as explanatory variabls in the
initial-conditionsequation (5). Also included are two-year lagged values of the plant's wages, capital stck, and export
price.
2 An alternative solution to this problem is prevented because of the presec of time-varying exogenous variables in
the model. These make it impossibleto solve the model for the steady-state probabilitiesof the pre-sample Yerealizationsas
fncmtionsof data and estimable parameters.
17
(6)
i4
b = 4u, +
+
yY,,, +E-yii
+
,ad],
:J+1,
... T
Ja2
and
(7)
a,, = -(A Z, + t2,),
where ,
=
(o, I (1 -
t=...J
Cr))Ifl and
42=
The likelihoodfunctionfor a panel of n plants
(p(1-p))".
can now be writtenas:
(8)
L(A, ,,
UI '
.
FT, l,
{ II
'Ia
y
9
.
i
p
(2Y,-I)] II -[b" (2Y-1)]) 4(a) doe.
Here ii(-) is the standardnormalcumulativedistributionfunctionand
x(a)
is the normaldensity
functionfor the unobservedplant effects.
To summarize,the model of exportmarketparticipationconsistsof equations(4) and (5) and
is estimatedwith maximumlikelihoodusingthe likelihoodfunctionin equation(8).2 The
specificationof the error terms allowsfor serial correlationarising from plant-specificdifferencesin
the profitabilityof exportingand for the initialconditionsproblem. Bothcross-sectionaland temporal
variationin the data are used to identifythe coefficients. The formeris due mainlyto cross-plant
differencesin industry,location,businesstype, age, capitalstock, the relativeprice of foreignto
domesticoutput, and wages. The latter is due largelyto economy-widefluctuationsin macro
conditions,and to unobservedplant-specificshocks((i and c), whichcombinewith changesin 4, to
inducetemporalvariationin Y1,.Notethat, even if there were no variationin 4,, the transitory
'3 The integral
in equaton (8) is approximated
usingthe Hermiteintegrationformuladiscussedby Butlerand Moffit
(1982). The resultsreportedbelowused sevenevaluationpointsand were virualy unchangedfrom a five-pointevaluation.
18
shocksypkand w, would sufficeto identifythe coefficientson Y,. and ?,.
SpecificationTests: If the empiricalmodelprovidesa good approximationto the process
that generatesY,, it shouldfit the observedexportmarketparticipationpatternsup to a plant-specific
time-invariantcomponentplus serially-uncorrelated
noise. Further, thesedisturbancesshouldbe
orthogonalto each otherand to the vector Z. As discussedby Andrews(1988)and Rust (1992),
specificationtests can be used to test these assumptionson the disturbancetermjointlywith our
assumptionson the role 4, Yj,, and
l
Specifically,using the estimatedparametersof equations(4) and (5), we will repeatedly
combine actual series on the strictly exogenous variables (Zi,)with random draws on oil, i, and cal,to
generateYktrajectories,plantby plant. If the modelis valid, thesesimulatedparticipationpatterns
for the years J+1IthroughTwill differ fromthe actualones only becauseof the randomrealizations
on act,.p, and wi. Thus each possiblesequenceof zeros and onesshouldoccur with approximately
the same frequencyin both the simulatedand actualdata. Andrews' (1988)chi-squarestatistic
2'
providesa metricfor comparingthe two sets of frequencies.
V.
EconometricResults on Export Prticipation
In this sectionwe report parameterestimatesof equation(4). For all of the resultsreported
here, we focuson four major exportingindustres duringthe period 1981through 1989. These
industriesare foodproducts,textiles,paperproducts,and chemicals.25We limit our sampleto the
24 Further details of the test are provided in the Appendix.
5 We wish to limit the analysis to those industries in which Colombia appears able to comnpetein iternational markets.
The industries were chosen because they account for a substantialpercentage of total manufacturedexports and have a
relatively high proportion of plants participating in the export market. These four industriesaccount for 58.6 percent of the
value of manufacturedexports in 1984 and 59.1 percent in 1987. rTe percentage of plants in our sample that export
19
650 plants in theseindustriesthat were in operationin each sampleyear.?6 This sampleis not
representativeof the populationof manufacturingplants, however,it is appropriatefor examiningthe
effectsof sunk costson establishedproducersY The final data set consistsof 9 annual
observations,coveringthe years 1981-1989,for each of 650 plants;a total of 5850 observations.The
observationsfor 1981-1983are treated as the threepre-sampleyears and are *sed to controlfor the
initial-conditions
problemusing equation(5). The observationsfor 1984-1989are used to estimate
the role of sunk costs usingequation(4).
The parameterestimatesfor equation(4) are reportedin Table 3 for severalmodel
specifications.The mostgeneralmodel, reportedin the first column,includesthree lags of past
participation (Y,s,,
pi2
2,ij-3
)
as well as interaction terms involving Y,,.,and year dunuy variables.
Trheseinteractionterms allowthe sunk costsof a new exporterto vary over time with market
conditions.8 The remainingthree columnsin Table 3 reportresults for modelsthat restrictthe
averages8.45 percentper year in the food industry,19.7 in textiles.15.4 in paperproducts,and 45.3 in chemicals.
2 The maindifferencebetweenthe continuinggroupof plantswe analyzeand the plantsthat exit producionover the
periodis that the lattergroup has a lowerrate of entry intothe exportmarket,averaging1.9 percentper year.and a lower
degreeof persistenceonce in, averaging.795. The differences,however,do not alter the generalconclusionthat transition
rates, particularlyfor plantsthat do not export,are low and persistenceis high. Whenexaminingthe plantsthat entered
production
over the period.a patternof exportmarkettransitionsvery similarto the plantswe analyzeis observed,
particularlyafter the plantshavebeen in operationfor a few years. The mainimplicationof thesepatternsfor our analysis
is that focusingsolelyon the groupof continuingplantsdoes not distortthe patternsof exportmarkettransitionspresentin
the manufacturing
sector. It is this patternof exportmarkettransitionsthat is importantin estimatingthe econometric
model.
n A more generalframeworkwouldtreat eachplant as makingsimultaneous
decisionsto enter or exit productionand
enteror exit the exportmarket. In this case, eachplant couldbe viewedas choosingamongfour alternatives:do not
produce,produceonly for the domesticmarket,produceonly for the exportmarket,or producefor both. This approachis
unnecessarilycomplicatedfor modelingthe exportdecisionin Colombiabecausethere are no pro-ducers
that sell only in the
exportmarket. In addition,very few plantsenterproductionand the exportmarketat the same time. As a result,focusing
on the exportingbehaviorof parntsthat are alreadyin operation,as we do, providesa reasonablestartng point for
analyzingthe exportdeterminantsin Colombianmanufacturing.
2 We alsoestimatedmodelsthat allowedsunkcosts to vary withobservableplantcharacteristicsby including
interactionsbetweenlaggedparticipationand plantsize, business ype, and industry. Noneof theseadditionalinteractions
wereever statisticallysignificamand we do not reportthemhere. We alsogeneralizedthe patternof time variationin sunk
costs by interactingyear dummiesand the two periodlag in participation.Theadditionalcoefficientsfrom this specification
were also insignificanL
20
coefficientson past participation. The secondcolumnrestrictsthe interactionterms with Y.., to equal
zero, implyingsunk costsdo not vary over time. The '.ird columnrestrictsthe coefficents
on
k,
2
and kft-3to equalzero, implyingthat the sunk costs of entry are the samefor all non-
exportingplants regardlessof whetherthey had ever been exporters. The fourthcolumncombinesthe
previoustwo sets of restrictions. The final columnrestrictspast participationto have no effect on
currentparticipationand is consistentwith no sunk entry or exit costs.
Specificationtests are reportedat the bottomof each columnin Table 3.29These chi-square
statisticshave 5 degreesof freedom,and providean omnibustest of whetherthe model's
parameterizationand distributionalassumptionsare valid. The test does not reject the specificationin
either column 1 or 2 - both thesemodels includethree laggedvaluesof past export participation.
However,the models in columns3, 4, and 5, whichall restrictpastparticipationto have at most one
lag, are all rejectedby the specificationtest. This impliesthat Y,does not followa first-order
Markovprocess. Giventhat model I performsvery well by this metric,we make it the focus of our
discussionbelow.
Sunk Cost Parameters: Considerfirst the coefficientson Y,,,,
?2,
Yd
3.
a
Yk.,
interactedwith time dummies. Together,these parametersisolatethe importanceof sunk costs.
Usinga likelihoodratio test to comparemodel 1 and model 5, we find that they are jointlysignificant
with a x2(8)statisticof 206.06. This findingsupportsthe basispremiseof the hysteresisliteraturethat there are substantialsunk costsinvolvedin enteringor exiing the export market.
T
Thereare 2' = 64 possibletime pathsfor Y,,over the 1984-1989
period,someof whichare quiterare. Accordingly,
to improvethe power of our test, we group these accordingto initialexpordngstaus and the numberof swichesin export
staus over tine, arrivingat six typesof trajectories.The categoriesof plant-specific
Y,trajectoriesare: all ones,aUzeros,
beginas a one andswitchonce, beginas a one andmakemultipleswitches,beginas a zero and switcbonce, and beginas a
zero and makemultipleswitches. The actualand predictedfrequenciesfor thesesix trajectoriesare reportedin the
Appendix.
21
Lookingat individualcoefficients,we find that laggedexportparticipationY11, hlasa strong
positiveeffecton the probabilityof exporting,as expect,;d. Further, sunk entrycosts vary
significantlyover time, as coefficientson the interactiontermsbetweenY,.,and the time dummies
imply.? The signpattern on the coefficientsimpliesthat the sunk costsfell from 1984to 1985for
the typical plant, and rose steadilythereafter. One interpretationis that it is easier to break into an
expandingworld marketthan a shrinkingone: 1985was a year of relativelyrobust globalexpansion
and overvaluationin the UnitedStates,whilethe late 1980swere a periodof relativelyslow growth
in the UnitedStates.
As shownin equation(4), the coefficientson PI2 and P,3 measurethe sunk costs of a new
exporterminusthe sunk costs of a plant that last exportedtwo or three years earlier, respectively.
They indicatethat experiencetwo years agodoes significantlyreducesunk costs, relativeto the costs
of a plantthat has never exported. That is, the benefitsof past exportmarketparticipationdo not
depreciatefully upon exit. On the other hand, the coefficenton
t-3
is not significantlydifferent
than zero, implyingthat plants that last exportedthree years earlier face re-entrycosts roughlyas
large as those facedby a plant enteringthe marketfor the first time. Henceour choiceof a three
year lag structureappearsto captureall of the relevanthistory.
Expected Profits from Exporting: The remainingcoefficientsin Table 3 summarizethe
influenceof year effectsand plant characteristicson the expectedprofitabilityof exporting,net of
sunk entry costs (ir, - Pi,). (Recallthat this expressiongives the net return from exportingfor a
plant with no prior foreign marketexperience.)
The time dummiesindicatethere is variationover time, but only the 1987and 1989
30 A lirelihoodratio test for the joint signifrcance
of these interaction tens
a 2(5)staistic of 14.54.
22
is rejected at the .05 significance level with
coefficientsare significantlydifferent than zero.3" Net of entry costs, the expected future profits
from exporting were highest in 1986. This conforms to the relatively rapid entry rate between 1986
and 1987 (Table 2), and may have reflected producer anticipationof the coming years of currency
depreciation. The incentivesto begin exporting soon dissipate, however, and by 1989 expectednet
profits reach their in-sample low. Again, this is consistentwith the falling entry rates in Table 2.
Interestingly,the decline in net expected profits appears to be due to rising sunk costs rather
than falling expectationsof gross operating profits. Relative to the base year, sunk entry costs were
1.08 higher in 1989, while expected profits net of sunk costs were only .847 lower. In fact, if exit
costs (Xi)were zero, producers already exporting were doing better in 1989 than in any other year.
One interpretationof this pattern is that the real exchangerate, which was favorable to exporters in
1989, determines expectedoperating profits for incumbents, while the strength of the world
economy, which was ebbing in 1989, determinesthe ease with which new exportms can break in.
Net export profitability also varies systematicallywith observable plant characteristics.
Notably we find that increases in plant size (measuredby the plant's capital -
l),
increasesin age,
and corporate ownership all increase the probability of exporting. The plant size result may reflect
scale economy-basedexporting, as in Krugman (1984).32Alternatively, since efficient plants tend to
grow relative to others, capital stock may simply be serving as a proxy for productivity. The age
coefficient may also pick up cost differences among producers. If market forces select out inefficient
producers then older plants will tend to be more competitivein world markets, either because of cost
advantages that cannot be imitated by rivals or because they have had time to move down a learming
3
' The hypothesis
that the time dummiesare jointlyequalto zero is rejectedwitha likelihoodratiotesL The test
statisticis 23.48versusa x2(5)criticalvalueof 15.1 at the .01 significance
level.
2 Combinedwithour fmdingthat sunkcostsdo not risewith size,it is consistentwiththe well-known
positive
correlationbetweenplantsizeand exportparticipation.SeeCaves(1989)and Berry(1992)for a reviewof the evidence
relatingsizeand propensityto export.
23
33 Evenif theannualpay-offfromexportingwere the samefor youngand old plants,the
curve,
youngoneswouldperceiveless returnto breakinginto themarketbecausethey are less Iikelyto
survive.
Locationmattersas well, presumablybecauseof transportcosts. Bogota,the basecategory
in the model,is land-lockedin the Andesmountainrange, andplantsthere are amongthe leastlikely
to producefor foreignmarkets. CaliandMedellinare alsoinland,but less mountainousand closerto
the coast. Nonetheless,we estimatethat thesecitiesare as unlikelyto serve as a base for exportersas
Bogota. Perhapstheir locationaladvantageis offsetby their lackof Bogota'sagglomeration
economies.Finally,the port citiesof Cartagenaand Baranquillaare most likelyto hostexporting
plants.
Interestingly,neitherwage ratesnor exportpricesrelativeto domesticoutputpricesare
significantdeterminantsof exportingbehavior. This shouldnot be interpretedto meanthat prices
don't matter;time dummieshavealreadycontrolledfor generalmovementsin relativeprices, andthe
plant-specific
price variablesthereforereflectacross-plantdeviationsfromaveragetrendsthat can
resultfrom localmarketconditions,measurementerror, anddifferencesin mputor outputquality.
Althoughwe wouldexpectincreasesin the exportprice to increaseexportmarketparticipation,we
have no strongpriors on whetherchangesin the cost of laborshouldmakeexportingmore or less
attractivethanservicingthe domesticmarket.
UnobservedPlat Heterogeneityand Noise: The finalsourcesof vaniaon in exportstatus
are unobservederror components:persistantplantheterogeneity,a;, and transitorynoise,w. As
shownin Table 3, .336of the total unobservedvariationis due to persistentheterogeneity,and this
I A declinein the probabilityof failureas a plantageshas beenfoundby Roberts(1994)for Colombiaand by Tybout
(1994)for Chile. Liu and Tybout(1994)alsofindtht failingplantsin Colombiaare systematically
less productivethan
survivingplants. Bothpatternshavebeenfoundin datafrom the U.S.. (seeEvans(1987).Dunne.Roberts.andSamuelson
(1989).and Bailey.Hultenand Campbell(1992)).
24
fractionis significantlydifferentthan zero. Once this error componentis controlledfor, however,
our specificationtests indicatethat remainingunobservedvariationis seriallyuncorrelatedand
orthogonalto the set of explanatoryvariables.
In the pre-sampleyears, the fractionof variationdue to unobservedplant effectsis much
higher(.928). This is simplybecausethe laggedparticipationvariablesare not used to predict Y,,
during 1981-1983,and their effect is shiftedto the disturbance. An analogouseffect is present in
model 5, whichleaves out laggedparticipationvariablesfor all years. There, ai accountsfor .817 of
total unexplainedvariation. Model5 also demonstratesthat laggedparticipationis stronglycorrelated
with the vectorZ;,. Note that withoutY,,,
and
, th-3.
e remaiing variablesassumea larger
role in predictingexportingstatus. All of theseresultsconfirmthat exportmarketparticipation
equationswithoutdynamicsare seriouslymis-specified.
Sunk Costs, Heterogeneity,and Export Probabilies: Table4 quantifiesthe effectsof
observableplantcharacteristics,unobservedheterogeneity,and past participationon current export
marketparticipation. It is based on estimatesof the unrestrictedmodelreportedin column1 of Table
3. The threepanelsallow the observabledeterminantsof exportprofitability(age, capitalstock,
industryetc.) to vary. Plantsat the 25th, 50th, and 75th percentileof 3Z,are compared. Within
each panel plantsare distinguishedby whetherthey neverexported(Y,, = Y,,2
exportedthree years earlier (Y, =
= f-4. = 0;
= 0; 1r3 =
=
P
3 =
0), last
1), last exportedtwo years ago (Y,,
= 1), or exportedlastyear (Y., = 1; Y,,2 = Fr3 = 0). The rows ofthe
table summarizethe effect of unobservedheterogeneityby allowingthe normally-distnbuted
pmanent plant componentae to vary from -2 to 2.
Exporthistory mattersfor plantsthat have either above-averageiZa or above-averagea,
values. Thus,although only25 percent of our sampleever exported,roughly75 percent of the plants
might well remainexportersif they were givensome foreign marketexperience. (The probabilities
25
that they would do so range from 14 percentto nearly 100 percent.) Expectedprofits from exporting
for the remainingplantsare so low that if they were somehowgivenexport marketexperienceit
wouldnot be sufficientto makethem continuein the export market.
Table 4 also demonstrateshow quicklyexperiencedepreciates. The differencein export
probabilitiesbetweenotherwisecomparableplantsthat exportedlast year and those that last exported
two years ago is substantial,rangingfrom .30 to .40. However,the differencebetweenplants that
last exportedtwo years ago and plants that last exportedthree years ago is substantiallysmaller,
rangingfrom .10 to .15, and plants that last exportedthree years ago are not muchmore likelyto do
so than thosethat never did.
Finally,inferencesaboutthe effectsof changingmacro conditionscan also be drawn from
Table 4.3 For example,relativeto 1984. the less favorablemacro conditionsthat prevailedin 1989
shiftedthe probabilitiesthat non-exporterswouldbeginexportingapproximatelyto those in the row
directy above.
Overall,the resultsreported in Tables3 and 4 revealthat the export participationdecisionis
affectedby time-periodor macro conditions,observableplant cost or demandvariables,unobserved
time-invariantplant heterogeneity,and - importantlyfor the sunk-costhysteresismodels- prior
export marketexperience. The latter has a particularlysubstantialeffect on the probabilitya plant
exportsand this is consistentwith the plantfacing significantentry costs in the exportmarket.
VL
Conclusions
In an attemptto explainthe asymmetricpatternsof exportand importadjustmentresulting
from exchangerate movements,a recent group of papers developstheoreticalmodelsthat rely on
34Sine a one-unit change in a corrsponds to a .71 shift in bk (by equation 6 and the deflinion of
shifts bj by .71 corresponds to a one-row upward movement in Table 4.
26
.
anying that
sunk costs of entry to produce hysteresis in trade flows. Beginningwith the assumptionthat entry
into foreign markets requires exporters to incur some sunk costs, these models demonstratethat
temporary changes in market variables, such as exchange rates, can result in permanent changes in
market structure and exports because of the entry and exit of exporters. While the hypothesized
response in these models is clearly a micro process of entry and exit, empirical tests to date have
relied on aggregate or sectoral data on trade flows and prices.
In this paper we develop an econometric model of a plant's decision to export and use it to
test one of the key assumptions underlyingthe theoreticalmodels of sunk cost hysteresis. We utilize
micro data for a large group of manufacturingplants in Colombia from 1981-1989and directly
examine the determinantsof a plant's export decisionfor consistencywith the theory. The
implicationof the sunk cost models that we test is that past participationin the export market will
have a significant effect on the probabilityof exporting in the current period. Equivalently, the
presence of sunk entry or exit costs will lead to true state dependence in the export decision.
The econometric results, which control for both observed and unobserved sources of plant
heterogeneity, indicate that prior export participationhas a significanteffect on the probabilitya plant
exports. Equivalently, we reject the hypothesis that sunk costs are zero. The empirical results also
reveal that the re-entry costs of plants that have been out of the export market for a year are
substantiallyless than the costs of a first-time exporter. Beyond a one year absence, however, the
re-entry costs are not significantlydifferent than those ficed by a new exporter. This is consistent
with the view that an important source of sunk entry costs for Colombian exporters is the need to
accumulateinformation on demand sources, informationthat is likely to depreciate upon exit from the
market.
While the results indicatethat sunk costs are a significant source of export market persistence,
both observed and unobserved plant characteristics also contributeto an individualplant's export
27
behavior. Plantsthat are large, old, and ownedby corporationsare all more likelyto export.
Variationin unobservedsourcesof differencein profitabilitycan leadto as much as a 30 percentage
pointdifferencein the probabilityof exportingfor a plant with no prior experience.
This combinationof plant heterogeneityand sunk costs impliesthat the responseof aggregate
or sectoralexportsto changesin policyor the macro enviromnentwill likelybe idiosyncraticwith
respectto countryand time period. The magnitudeof the supplyresponsewill dependuponthe
numberand type of plantsalreadyparticipatingin the export market,the stabilityor pennanenceof
the policyregime, the magnitudeof temporaryshocks,and the sunk costs of entering a new market.
The latter, in turn, is likelyto vary with the degreeof informationproducershave aboutforeign
markets,the type of marketthey are likelyto enter, the type of productbeingexported,and the
policy regimne.Giventhe numberof idiosyncraticforces at work, it is not surprisingthat standard
empiricalexport supplyfunctionshave exhibitedmarkedinstabilityacross countriesand time.
Finally, our findingssuggestthat countriesundertakingexportpromotionpoliciesshould
distinguishmeasuresaimedat expandingthe exportvolumeof exisitingexportersfrom policiesaimed
at promotingthe entry of new exporters. The latter includeacions directedat reducingentrycosts
anduncertainty,such as providinginformationaboutpotentialmarkets,developingexporting
infrastructure,or providinga stablemacro and policy environment.If enteringthe exportmarketis a
moresignificanthurdle for firms than expandingtheir outputonce in the market, these entry
promotionpoliciesmay be more effectiveat expandingexportsthan direct subsidiesbased on the
valueof exports.
28
References
Andrews, Donald W. K. (1988) "Chi-square Diagnostic Tests for Econometric Models: Introduction
and Applications," Journal of Econometricsvol. 37, pp. 135-156.
Bailey, Martin, Charles Hulten, and David Campbell(1992) "Productivity Dynamics in
ManufacturingPlants," BrookingsPaperson EconomicActivity: Microeconomics1992.
Washington, D.C.: The Brookings Institution.
Baldwin, Richard (1988), "Hysteresis in Import Prices: The Beachhead Effect," Amerian Economic
Revew, Vol. 78, No. 4 (September), pp. 773-785.
Baldwin, Richard (1989), "Sunk-Cost Hysteresis", NBER Working Paper No. 2911.
Baldwin, Richard and Paul Krugman (1989), "Persistent Trade Effects of Large Exchange Rate
Shoclks,"The Quarterly Journal of Economics, Vol. 104, No.4 (November), pp. 635-654.
Berry, R. Albert (199 1), "Firm (or Plant) Size in the Analysis of Trade and Development," in G.
Helleiner(ed.), TradePolicy, Industrialization
and Development:New Perspectives, Oxford:
Clarendon Press.
Bresnahan, Timothy F. and Peter C. Reiss (1991), "Entry and Competition in Concentrated Markets,"
Journalof PolilicalEconomy,Vol. 99, pp.977-1009.
Bresnahan, Timothy F. and Peter C. Reiss (1994), 'Measuring the Importance of Sunk Costs,"
Stanford University, working paper.
Butler, J.S. and Robert Moffitt (1982), "A Computationally Efficient Quadrature Procedure for the
One-Factor Multinomial Probit Model", Econometrica, Vol. 50, No. 3 (May), pp. 761-764.
Campa, Jose Manuel (1993), "Entry by Foreign Firms in the United States Under Exchange Rate
Uncertainty," 24 The Revew of Economicsand Statistics, Vol. 75, No. 4 (November),
pp.614-622.
Caves, Richard (1989), "International Differences in Industrial Organization," in Richard Schmalensee
and Robert Willig (eds.), Handbook of Industrial Organization, Vol. 2. Amsterdam: NorthHolland.
Dixit, Avinash (1989a), "Entry and Exit Decisions Under Uncertainty," Journal of Political Economy,
Vol. 97, No. 3 (June), pp. 620-638.
Dixit, Avinash (1989b), "Hysteresis, Import Penetration, and Exchange Rate Pass-Through," The
QuarterlyJournal of Economics, Vol. 104, No. 2 (May), pp. 205-228.
Dunne, Timothy, Mark 3. Roberts, and Larry Samuelson (1989), "The Growth and Failure of U.S.
Manufactring Plants," ThzeQuarterlyJournal of Economics, Vol. 105, No. 4 (November),
pp. 671-698.
29
Eckstein, Zvi and Kenneth I. Wolpin (1989), "The Specificationand Estimation of Dynamic
Stochastic Discrete Choice Models: A Survey,' The Journal of Hunan Resources, Vol. 24,
No. 4 (Fall), pp. 562-598.
Evans, David (1987), "Tests of AlternativeTheories of Firm Growth," Journa of Political Econony,
Vol.95, No.4 (August), pp. 657-674.
Feinberg, Robert M. (1992), "Hysteresisand Export Targeting," International Journal of Industrial
Organization,Vol. 10, No. 4 (December), pp. 672-677.
Gagnon, Joseph (1987), "AdjustmentCosts and InternationalTrade Dynamics," Ph.D. Dissertation,
Stanford University.
Goldstein, Morris and Mohsin S. Khan (1985), 'Incomeand Price Effects in Foreign Trade," in
Ronald W. Jones and Peter B. Kenen (eds.), Handbook of Internaional Economics, Vol. 2,
Amsterdam: Nordt-Holland.
Hecknan, James (1981a), "Heterogeneityand State Dependence," in Sherwin Rosen (ed.), Studies in
Labor Marets, Chicago: University of Chicago Press.
Heckman, James (1981b), "Statistical Models for Discrete Panel Data," in Charles Mansld and Daniel
McFadden (eds.), The StructuralAnalysis of Discrete Data, Cambridge: MIT Press.
Heckman, James (1981c), "The IncidentalParameters Problem and the Problem of Initial Conditions
in Estimating A Discrete Time-Discrete Data Stochastic Process," in Charles Manski and
Daniel McFadden (eds.), The StructuralAnalysis of Discrete Data, Cambridge : MIT Press.
Hsiao, Cheng (1986), Analysis of Panel Data, Cambridge: Cambridge UniversityPress.
Krugman, Paul (1984), "hnport Protection as Export Promotion: Intemational Competitionin the
Presence of Oligopolyand Economies of Scale," in H. Kierzkowski (ed.), Monopolistic
Conmetitionand International Trade, Oxford: Oxford University Press.
Krugman, Paul (1989), Exchange-RateInstability, Cambridge: M1T Press.
Liu, Lili and James Tybout (1994) "ProductivityGrowth in Chile and Colombia: Panel-Based
Evidence on the Role of Entry, Exit and Learning," The World Bank, working paper.
Neyman, J. and E. Scott (1948), "Consistent EstimatesBased on Partially ConsistentObservations,"
Econometrica,Vol. 16, No.1, pp. 1-32.
Parsley, David C. and Shang-JinWei (1993), "Insignificantand InconsequentialHysteresis: The Case
of U.S. Bilateral Trade," The Review of Economics and Satistics, Vol. 75, No. 4
(November), pp. 606-613.
Roberts, Mark (1994), "The Structure of Production in Colombian Manu
Pennsylvania State University, working paper.
30
r
Industries," The
Rust, John (1993) "StructuralEstimationof MarkovDecisionProcesses,"forthcomingin R. Engle
and D. McFadden,eds., Handbookof Econometrics,vol. 4, Amsterdam: North Holland.
Tybout,James(1994), "Entry, Exit, Competition,and Productivityin the Chilean IndustrialSector,"
GeorgetownUniversity,worldngpaper.
WorldBank(1992), "ColombiaExportDuvelopmentProject," mnimeo.
Wright, B.D. and G. Douglas(1976), "BetterProceduresfor Sample-FreeItem Analysis,"Research
MemoNo. 20, StatisticsLab, Departmentof Education,Universityof Chicago.
31
Table 1
Colombian Manufactured Exports 1981-1989
(nineteen three-digit SrTC industries)
1981
1982
1983
1984
1985
1986
1987
1988
1989
Real Effective
Exchange Rate Index
117.8
125.7
125.1
114.6
100
74.4
66.3
64.4
62.7
Real ValueofExports
109.6
110.8
113.7
95.9
110.8
151.2
169.4
189.1
276.2
Export Subsidy Rate
.055
.055
.066
.099
.092
.047
.047
.042
.044
Number of Exporting
Plants
667
676
615
585
653
705
707
735
816
Proportion of Plants
that Export
.129
.128
.113
.107
.117
.122
.119
.124
.135
Table 2
Plant Transition Rates in the Export Market 1982-1989
Year t
Status
Year t+1
Status
1982- 1983
19831984
19841985
19851986
19861987
19871988
19881989
Average
1982-1989
(Nineteen three-digitmanufacturingindustries)
No Exports
Exports
No Exports
.974
.971
.957
.963
.973
.972
.958
.967
Exports
.026
.029
.043
.037
.026
.028
.042
.033
No Exports
.168
.135
.131
.108
.158
.086
.107
.128
Exports
.832
.865
.869
.892
.842
.914
.893
.872
(Four major exporting industries)
No Exports
Exports
No Exports
.971
.969
.972
.960
.983
.972
.985
.973
Exports
.029
.031
.028
.040
.017
.028
.015
.027
No Exports
.108
.101
.152
.124
.149
.085
.054
.110
Exports
.892
.899
.848
.876
.851
.915
.946
.890
Table 3
Dynamic probit model of export participation
(standard errors in parenthesis)
Explawtory variable
Intercept
Model I
-8.312(1 .456)
Model 2
Model 3
-8.3931(1.426)
*9.2850(1.433)
2.170*(.152)
1.707-(.245)
r,.,
1.933*(.261)
Y,.,0(1985
Dummy)
-.106 (.313)
-.170 (.315)
Y*.,1(1986
Dunumy)
-.009 (.318)
*.051(.321)
Y 1.*(1987Dunumy)
.185 (.324)
.202 (.327)
Y,..0(1988Dummy)
.261 (.328)
.296 (.334)
Y1.A(1989Dummy)
1.082*(.368)
1.112*(.371)
Model 4
Model S
-9.381 (l.458)
-16.954e(l.710)
1.8650(.124)
Y, -2
.732*(.195)
.777*(.193)
y.-3
.347 (.248)
.340 (.24)
1985Dununy
-.155 (.201)
-.215 (.161)
-.130 (.197)
-.221 (.163)
-.276 (.178)
1986Dummy
.004 (.189)
.018 (.159)
.026 (.187)
.019 (.161)
-.068 (.179)
1987Dummy
-.469*(.220)
-.398*(.169)
-.498*(.218)
-.436-(.171)
-.476(.1I89)
1988Dummy
-.248 (.202)
-.149 (.170)
-.286 (.201)
-.205 (.174)
-.3300(.193)
1989Dummy
-.947*(.253)
-.378'(.178)
-.902*(.249)
-.458*(.182)
-.520*(.198)
ln(Wage,.)
.216 (.147)
.231 (.145)
.256 (.149)
.236 (.150)
.4700(.173)
ln(ExponpriceM,)
-.062 (.062)
-.059 (.061)
-.064 (.062)
-.084 (.067)
.0140(.083)
1naK,.)
.247*(.045)
.239*(.043)
.291 (.041)
.2880(.043)
.5370(.052)
Agel.,
.337(.11S)
.330*(.114)
.379*(.115)
.427*(.116)
.7010(.169)
Corporation
.368+(.158)
.333*(.155)
.365*(.154)
.4500(.161)
.396 (.229)
TextilesInd. Dununy
1.018'(.198)
.999*(.194)
1.124-(.179)
1.221(. 196)
2.392*(.259)
PaperInd. Dummy
.251 (.183)
.254 (.180)
.250 (.189)
.237 (.190)
.940*(.273)
ChemicalsInd.
DUMMY
.5320(.179)
.503'(.176)
.875*(.187)
.628*(.181)
2.996*(.285)
Cali/Medellin
-.099 (.143)
-.099 (.139)
-.060 (.140)
-.123 (.143)
.5470(.190)
Otherregion
.378*(.139)
.371*(.136)
.409*(.136)
.456*(.141)
1.4300(.204)
Var(a)
.336*(.076)
.314*(.076)
.416*(.053)
.446*(.059)
.817*(.016)
Cov(qP.a)
.928*(.013)
.92280(.013)
.9170(.013)
.926*(.012)
.7790(.025)
In(L)
-845.80
-853.07
-854.09
-860.86
-948.83
Specification et
5.121
5.519
12.065*
9.S330*
SS.112*
Rejectnull hypothesisat the .05 significancelevel.
**
Rejectnull hypothesisat Ihe .10 signficance level.
Table 4
Predictedprobability of exporting
25thpercedti of A5Z.O
501hpercenile
of 17,.
75thpercetileof OZ.,
Plantcfrect(a)
0
YI3=0
=
0
0
0
1
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
-2
.000
.000
.000
.006
.000
.000
.001
.024
.001
.002
.006
.096
-1
.000
.000
.001
.037
.001
.002
.007
.104
.006
.015
.037
.277
0
.001
.004
.011
.141
.007
.016
.040
.291
.035
.071
.140
.547
1
.011
.026
.OS9
.358
.038
.077
.149
.564
.135
.225
.356
.797
.108
.197
.636
.144
.238
.371
.808
.348
.482
.633
.938
2
.056
Appendix: SpecificationTests
Andrews' (1988) specificationtest comparesrealized (Y,, Z,,) trajectoriesin our sample to
expectedtrajectoriesbased on the estimatedmodel. To constructhis X2statistic,we first partitionthe
possible(Ye,Z,,)trajectoriesinto a limitednumberof cells. In our case we have 2' - 64 possible
trajectoriesfor Y&,some of whichare veryunusual. To avoidcellsthat are nearlyempty we distinguish
six types of trajectories: Y,,= 0 V t; Yft= 0 initially,but switchesonce duringthe sanple period; f,
= 0 intially and switches at least twice during the sample period: Y,,= I v t; Y,,
I initially, but
switches once during the sample period; and Y. = 0 initially and switches at least twice during the sample
period. Lettingthe vector indicatorfunctionTI(Y Y
I, 1 ,+2 ,
...
YUTmaPY sequencesinto these6
cells, the observed frequencies of the different trajectories in our sample is the vector
P. )=-
,E(Y,,,
IYil*,2
* )
.
Elementsof this vector are reported in Table A1. I below.
Table Al.l: Observed versus Predicted Frequences of Y,,Trajectories
(based on Column1, Table 3)
Trajectory type
Observed
Frequences
Expected Frequencies
alwaysa non-exporter
.761
.743
beginas a non-exporter,switchonce
.046
.051
begin as a non-exporter,switchat
least twice
.045
.052
alwaysan exporter
.098
.089
begin as an exporter, switchonce
.017
.028
beginas an exporter, switchat least
twice
.03
.037
Next, to generate model-basedexpected values for each of these cells, we use estimatedparameter
values from Table 3 in conjunction with the observed 4, trajectories and random draws on a5, P,, and
w,, to repeatedly simulate Y',sequences, plant by plant. (The reported tests are based on 200 simulations
per plant.) Distributions for each of these random variables are based on the assumptions described in
section III. Averaging over all of the outcomes for the
rhplant, we get the probabilities it will fall in
each cell under the null hypothesis that our specification is correct: Q(1,Zu,Z1J.1 **Z,7 JI,y,Ca,P)
-
Finally, averaging these probabilities over all plants, we obtain the expected sample-wide frequencies of
each cell in the partition:
Q.(I) = 1E
Q(Z, Z ZU.
1
*ZT I p, y, or p)
The expected frequencies generated by the model in column 1 of Table 3 are reported in the
second column of Table A1.1. The test statistic is calculated as a quadratic form in the difference
between the two columns,
X2 1)= (P - Q.
by equation (15) in Andrews' appendix.
Q) ; where the
weighting matrix
W is given
Policy
Research Working
Ttle
Paper Serles
Contact
for paper
Author
Date
WPS1411 Income Inequality, Welfare, and
Poverty: An lflustraton Using
Ukrainian Data
Nanak Kakwani
January 1995
G. Evans
85783
WPS1412 ForeignTechnology Importsand
EconomicGrowth in Developing
Countries
XlaomingZhang
Heng-fuZou
January 1995
C. Jones
37754
WPS1413 Endogenous Distortionsin Product
and Labor Markets
MartinRama
Guido Tabellini
January 1995
S. Fallon
38009
WPS1414 The World Bankand LegalTechnical The World Bank
Legal Departmrent
Assistance: Initial Lessons
January 1995
K Mathemova
82782
WPS1415 China'sGDP in U.S. Dolas Based
on PurchasingPower Parity
Ren Ruoen
ChenKai
January 1995
E O`YRielIyCampbell
33707
WPS1416 Informal Regulationof Industrial
Pollutionin DevelopingCountries:
Evidencefrom Indonesia
Sheoli Pargal
David Wheeler
February1995
E Schaper
33457
WPS1417 UncrMtaintyand Gobal Warming:An
Option-PricingApproachto Policy
Andrea Baranzini
Marc Chesney
Jacques Morisset
February1995
C. Def
85148
WPS1418 The Impactof Labor Market
Regulations
Lyn Squire
Sethaput SuthlwartNarueput
February1995
G. Bayard
37460
WPS1419 Industry Structuro and Regulation
MarUnC. Stewart-Smith
February1995
N. James
82758
WPS1420 LegislativeFrameworksUsed to
Foster PetroleumDevelopment
WilliamT. Onorato
February1995
W. Onorato
81611
WPS1421 Distnbutionof Incomeand the Income ZeqikoBogetic.
Fared M. A. Hassan
Tax Burden in Bulgaria
February1995
F. Smith
36072
WPS1422 Efficiency in Bulgaria's Schools:
A NonparametricStudy
Zeljko Bogetic
Sajal Chattophadyay
February1995
F. Smith
36072
WPS1423 The Role of CommercialBanksin
Enterprise Restructuringin Central
and Eastern Europe
Millard Long
Izabela Rutkowska
February1995
R. Garner
37670
WPS1424 Terms-of-TradeShocks and Optimal
Investment Another Look at the
Laursen-Metzler Effect
Luis Serven
February1995
E. Khine
37471
Pollcy Research Working
Tine
Paper Series
Contact
for paper
Author
Date
WPSf425 On the IntersectoralMigration of
Agricultural Labor
Donald Larson
Yalr Mundlak
Febnrary1995
J. Jacobson
33710
WPS1426 Russian Unemployment:Its
Magnitude, Characteristics,and
RegionalDimensions
Sinon Commader
Ruslan Yemtsov
February1995
V. Reid
35195
WPS1427 CorporateGovernanceand Equity
Prices: Evidencefrom the Czech
and Slovak Republics
Stin Claessens
February 1995
F. Hatab
35835
WPS1428 Short-TermSupply Responseto a
Devaluation:A Model's Implications
for Primary Commodity-Exporting
DevelopingCountries
Bruno Boccara
Fabian Nsengiyumva
February1995
M. Pfeiffenberger
34963
WPS1429 The World Trade Organization's
BomardM. Hoekman
Agreementon Government
PetrosC. Mavroidis
Procurement:Expanding Disciplines,
DecliningMembership?
March1995
F. Hatab
38535
WPS1430 IntergovemmentalFiscal Relations
and PovertyAlleviationIn Viet Nam
RichardM. Bird
Jennie I. Litvack
M. GovindaRao
March 1995
G. Coward
80494
WPS1431 The Industrial Pollution Projection
System
HemamalaHettige
Paul Martin
ManjulaSingh
David Wheeler
March1995
A. Williams
37176
WPS1432 Using FinancialFuturesin Trading
and Risk Management
Ignacio Mas
Jesik R3a&Requejo
March1995
K. Binidey
81143
WPS1433 EnterpriseRestructuring in Eastem
Europe:How Much?How Fast7
Where? PreiiminaryEvidencefrom
Trade Data
BemardHoeknan
GerharCPohl
March1995
F. Hatab
35835
WPS1434 Govemment Credit Policy and
Industrial Performanoe(Japanese
MachineTool Producers,1963-91)
CharlesW. Caloiris
CharlesP. Himmnelberg
March 1995
P. Infante
37642
WPS1435 The Porical Economny
of Formal
Sector Pay and Employmentin
DevelopingCountries
Arup Banerji
J. EdgardoCampos
RichardH. Sabot
March 1995
D. Evans
37496
WPSI436 An EmpiricalModel of Sunk Costs
and the Decsion to Export
Mark J. Roberts
James R. Tybout
March 1995
J. Ngaine
37959