short-run consequences of trade liberalization: a computable general equilibrium model of zimbabwe

Short-Run Consequences of Trade Liberalization:A Computable General Equilibrium Modelof Zimbabwe

Rob Davies, University of Zimbabwe, Harare, Zimbabwe

Jørn Rattsø, Norwegian University of Science andTechnology, Trondheim, Norway

Ragnar Torvik, Central Bank of Norway, Oslo, Norway

The elimination of import controls represents a challenging adjustment process for anyeconomy. The mechanisms are investigated by the use of an economy-wide model ofZimbabwe. A benchmark version assumes import rationing and protection of domesticmarkets in an economy with unemployment of unskilled labor. Rather than modelingtrade liberalization as a decrease in tariffs, we view it as a regime shift, requiring a newmodel closure. Compared with previous computable general equilibrium studies of tradeliberalization, the analyses includes two expansionary channels, intermediate imports andsavings response. It is shown that a combined consumption boom, short-run contraction,and growing trade deficit are likely, due to drop of savings and demand switching toforeign goods. 1998 Society for Policy Modeling. Published by Elsevier Science Inc.

1. INTRODUCTION

Regulation of foreign trade has been a key feature of the Zim-babwean economy for 3 decades. During the unilateral declarationof independence (UDI) period between 1965 and 1980, interna-tional sanctions, and domestic policies to cope with them, induced

Address correspondence to Jørn Rattsø, Department of Economics, Norwegian Universityof Science and Technology, N-7055 Dragvoll, Norway.

Funding for this study was provided by the Ministry of Foreign Affairs, Norway.We are grateful for comments and discussions from seminars at Copenhagen University,Development Bank of South Africa, Harvard University, New School of Social Research,Oxford University, and Yale University; in particular, we thank Paul Collier, Jan Gunning,and Lance Taylor.

Received July 1994; final draft accepted November 1996.

Journal of Policy Modeling 20(3):305–333 (1998) 1998 Society for Policy Modeling 0161-8938/98/$19.00Published by Elsevier Science Inc. PII S0161-8938(97)00013-6

306 R. Davies et al.

involuntary import substituting industrialization and strong growthup to about 1973, when the liberation war intensified. A sophisti-cated import control system was built, which the new governmentcontinued to use after independence in 1980. The import rationingapparatus was part of a general regulatory machinery used by thegovernment to set the priorities of the economy. In particular, itwas seen as a necessary instrument to control the foreign exchangesituation. Observers agree that import compression was a majorfactor in explaining macroeconomic stability and stagnation duringthe 1980s (see Dailami and Walton, 1989; Davies and Rattsø, 1993,1996; Green and Khadhani, 1986). During the 1990s the govern-ment has moved toward trade liberalization within the frameworkof a structural adjustment program. (Government of Zimbabwe,1991).

Governments embark on trade liberalization programs to changethe workings of the economy to gain long-term benefits fromcompetition and comparative advantage. However, whatever long-run benefits might be anticipated, reforms are often postponedand opposed due to short-run adjustment costs. The present studyattempts to understand some of the processes involved in such atrade liberalization in an economy like Zimbabwe. The analysisis undertaken using a computable general equilibrium (CGE)model, emphasizing real-side sectoral balances and macroeco-nomic interactions, based on our perception of the stylized factsof the Zimbabwean economy.

The starting point is a model representing the import compres-sion regime at work through the 1980s, as analyzed by Davies,Rattsø, and Torvik (1994). The 1985 base-year version models theimport controls in place during the decade, using a policy rulefor allocating foreign exchange in which investment projects andintermediates to the export sector have priority. This links importcapacity to both capacity utilization in import-dependent indus-tries and to domestic savings via rationing of intermediates andnoncompetitive consumer goods. Rationed imports include im-portables and food, protecting their domestic markets from inter-national competition. Import capacity is determined by endogenousexports, and foreign financing (and thereby the trade balance) isexogenous.

Trade liberalization is conventionally modeled simply as a de-crease in tariffs, assuming the rest of the model unchanged. How-ever, this does not capture the essence of trade liberalization,

TRADE LIBERALIZATION IN ZIMBABWE 307

which is intended to change the fundamental workings and mecha-nisms ruling the economy. To overcome this shortcoming, wepropose to model trade liberalization as a change in model closure,thus viewing it as a regime shift, replacing old mechanisms withnew ones.

Based on the benchmark import rationing model, trade liberal-ization is analyzed as elimination of quantitative import controls.Two alternatives are investigated, representing different degreesof liberalization. Both alternatives assume an endogenous tradebalance, thereby breaking the link between export performanceand import capacity. The first, imported intermediates liberaliza-tion, is limited to lifting rationing of imported intermediates forthe construction and importables sectors. Their demand functionsfor intermediates are satisfied, and consequently their supplycurves shift out compared with the rationed benchmark. The elimi-nation of the import compression bottleneck in import-dependentsectors is clearly expansionary.

The second alternative, full liberalization, adds competitive andnoncompetitive finished goods to the list of liberalized commodi-ties. The protected importables sector is forced to compete withimports, although imperfect substitution between them is assumed(the Armington framework). The improved access to noncompeti-tive consumer goods is assumed to affect savings rates, in accor-dance with the conventional wisdom in Zimbabwe. When domesticmarkets are opened up, both contractionary and expansionarydemand side forces operate, since foreigners take a share of thedomestic market and savings rates go down. The output conse-quences of trade liberalization depend on the policy context. Twoadditional elements are analyzed: devaluation and aid inflow.

The benchmark model is sketched out in section 2, and themechanisms included in the study of the transition to trade liberal-ization are discussed in section 3. The analysis follows in two parts.The first shows the short-run consequences of the two liberaliza-tion attempts described above, sections 4 and 5. Trade liberalizingdevaluation, assumed to keep the foreign balance constant, isanalyzed in section 6. The second part investigates additionalelements of a liberalization package. Foreign aid and devaluationto prevent total output from falling are discussed in section 7.

2. MODEL GUIDELINES

The benchmark model formulation is the same as Davies et al.(1994), and describes the Zimbabwean economy of the 1980s. The


model brings together two types of dual models familiar fromdevelopment economics. The dependent economy model distin-guishing between traded and nontraded goods can be seen as thepoint of departure. With a view to including terms of trade effectsand differences in import rationing, traded goods production isdisaggregated to importables, exportables, and food agriculture.De Melo and Robinson (1989) and Bevan et al. (1990) argue fora disaggregation of the traded sector in developing countries.The exportables are often raw materials, agricultural crops, etc.,whereas the importables have the characteristics of manufacturinggoods and, in sub-Saharan Africa, foodstuffs. Since productionstructure and demand linkages differ, nontraded production isdisaggregation to services and construction. The dual flexprice-fixprice approach of Taylor (1983), differentiating agriculture fromthe rest of the economy, is adopted, although our formulationdiffers from Taylor in that both food agriculture and most nonagri-cultural outputs are supply driven.

Taking all the above considerations into account, the modeldisaggregation is based on five production sectors as follows:

• Sector 1: food agriculture, comprising part of the agriculturalsector (communal and commercial); flex-price adjustment;referred to in the text as “agriculture”

• Sector 2: nontradable consumer goods, comprising services,electricity, transport; demand determined output adjustment;“services”

• Sector 3: nontradable capital goods, comprising construction;supply-demand adjustment; “construction”

• Sector 4: exportables, comprising mining, large-scale agricul-ture and part of manufacturing: adjustment through exportof residual output; “exportables”

• Sector 5: importables, comprising the rest of manufacturing;supply-demand adjustment; “importables”

A separation is made between skilled and unskilled labor in formu-lating both production and distribution technologies. In the labormarket, skilled labor is in short supply, and the wage rate adjuststo clear the market. Unskilled labor is in permanent excess supply,and the use of unskilled labor is demand determined given aninstitutionally determined nominal wage rate.

The investment level is fixed exogenously, and the import shareresponds to relative prices. The formulation reflects the problemof identifying a well-defined investment function (see Mehlum


and Rattsø, 1993, Chhibber et al., 1989), and the understandingthat investment demand was restricted by political and businessuncertainty (see Dailami and Walton, 1989). The investment re-sponse to trade liberalization is left for future research.

The benchmark model is built around a social accounting matrix(SAM) for 1985. The parameters and the exogenous variables ofthe model are set to have the endogenous variables consistentwith the SAM. Full documentation is offered in Davies et al. (1994),and the equation set under alternative closures is presented inthe appendix.

3. MODELING IMPORT RATIONING ANDTRADE LIBERALIZATION

In the benchmark model version, the import rationing systemis based on the descriptions by Davies (1991) and Pakkiri andMoyo (1986). Available foreign exchange is allocated to satisfypriority needs (investment goods, intermediates for the ex-portables sector, and food). The importables sector is protected,and only a limited volume of importables is allowed to be imported.Any remaining foreign exchange is allocated as policy-determinedshares, to meet requirements for imported intermediates by con-struction and importables, and to noncompetitive consumer im-ports. This formulation of the foreign exchange constraint is cen-tral to the working of the model. It makes the import compressionof manufacturing and construction supply endogenously deter-mined, dependent on the existing capacity to import. A theoreticalframework for analyzing import compression is suggested byRattsø (1994a), with additional aspects developed by Rattsø(1994b) and Torvik (1994). Gibson (1985) was the first to linkexport performance and import dependent industries in a modelof Nicaragua.

The intermediate rationing motivates a detailed specification ofthe supply side of the import-dependent sectors. The formulationallows for substitution possibilities in a hierarchical system includ-ing imported and domestic intermediates, skilled and unskilledlabor, and real capital. The production structure is similar to Graiset al. (1986), but the working of the model is different since theyassume rent-seeking and no unemployment.

The supply function of the import-dependent sectors is shownin Figure 1. It is upward sloping because of fixed capital stock


Figure 1. Effects of liberalization.

implying increasing marginal costs. The supply curve under inter-mediate rationing is drawn as a solid line and is steeper whenrationing bites. Reduced availability of imported intermediatescan be compensated only imperfectly by domestic intermediates,and marginal costs are increasing for the intermediates aggregate.The rationing system offers the producers commodity-specific li-censes for intermediates, and they are not allowed to resell. Clearlythis system motivates rent-seeking, and costs associated with rent-seeking are another possible source of steeper marginal costs, assuggested by Grais et al. (1986). However, in Zimbabwe the licenseholders were old clients of the government, and observers agreethat costs associated with rent seeking were not important. In ourmodel, license holders do get an implicit rent as a result of therationing. Rationing means a higher price of the finished goodsin the protected domestic market than otherwise would be thecase.

This leads us to the other important aspect of import rationing,protection of domestic final goods markets. The import compres-sion almost eliminated imports of both importables and noncom-petitive consumer goods, whereas there was a very thin domestic


market for those imports allowed. This motivates our assumptionof fixed price rationing of final goods imports. The rents are handedto the government in the model.

Two effects of this protection are captured on the demand side.First, demand switches to domestic goods when foreign goodsare not available. Neary and Roberts (1980) have suggested amodification of the linear expenditure formulation under rationingapplied here. Since consumer imports are rationed, the ratio ofimported to domestically produced consumption is lower thandesired at the prevailing prices. The notional price of consumerimports is higher than the actual. Second, when demand for im-ports is not satisfied, part of the income can be set aside as sav-ings—in the expectation of less severe rationing in the future. Atheoretical justification is offered by Torvik (1993). Empiricalevidence is identified by Chhibber et al. (1989) and Morande andSchmidt-Hebbel (1991). In the model, this stylized fact is takencare of by linking private savings rates to the rationing of noncom-petitive consumer imports. The mechanisms may be as importantto explain savings of private firms.

Imports of food are part of government policy to regulate do-mestic food markets, in particular in response to drought. Wehave made food imports endogenously dependent on food short-ages, the gap between a target for food consumption and theactual consumption.

Trade liberalization is simulated by eliminating the rationingin two steps. The first, intermediates liberalization, does awaywith rationing of intermediate imports. The intermediate importdemand functions for construction and importables (sectors 3 and5) become unconstrained. The trade balance is endogenous, re-sponding to changes in exports and imports, and foreign financingof the deficit is assumed to be available.

Figure 1 captures the expansionary supply effect, and the supplycurve shifts outward to the dotted line. Given the demand curveD, output will expand from a to b. With fixed nominal wageof unskilled and overall unemployment, total employment canincrease. The notional price of imported inputs falls. This is theopposite effect of the contractionary liberalization analyzed byBuffie (1984). In his model, a higher relative price of importedinputs explains possible output contraction.

The second package, full liberalization, adds free trade in im-portables, noncompetitive consumer goods, and food. Imperfect


substitution is assumed between domestically produced and im-ported importables, and the market shares are determined byrelative prices. The demand function for noncompetitive consumergoods is unconstrained. The domestic price of food is determinedat the international market (given the exchange rate) and netimports equilibrate the market.

The elimination of protection adds demand effects to Figure 1.Two counteracting forces can be identified. On the one hand,private savings rates are reduced when foreign goods are madeavailable, and the consumption demand goes up given the incomelevel. Since the stock of wealth accumulated under rationing canbe higher than desired under liberalization, the immediate floweffect on consumption can be quite large (see Rattsø, 1994c, onthe dynamics of this aspect). If demand expansion dominates, wemay end up in an equilibrium like c in Figure 1. Both supply anddemand factors are expansionary under liberalization. On theother hand, demand switches to imports since consumers havebeen forced to hold an inefficient mix of domestic and foreigngoods. This effect tends to shift the demand curve to the left,possibly ending up in an equilibrium like d in Figure 1, with outputcontraction as a result of liberalization.

The standard welfare analysis of trade liberalization is basedon a model assuming full utilization of resources. Trade liberaliza-tion is shown to improve resource allocation, since imperfectionsdisturbing marginal efficiency conditions are eliminated. Staticmisallocation costs of trade restrictions have generally been shownto be small. The introduction of rent seeking has added to theestimates of possible gains. CGE studies addressing trade liberal-ization include Benjamin (1992), Condon et al. (1985), and Graiset al. (1986).

The more recent literature addresses the macroeconomic con-text of trade liberalization, as Collier and Gunning (1992). Evenin models of full employment, credibility problems may lead tounfavorable outcomes regarding the trade balance and investment.Here we realistically assume unemployment of unskilled labor,opening up for possible effect on total employment. Theoreticalstudies by Buffie (1984) and Ocampo (1987) have shown thattrade liberalization can be contractionary in this situation. Ocampoemphasizes demand switching, whereas Buffie assumes higher in-termediate import costs. As shown above, demand switching mayreduce output from import competing domestic industries. In mod-els with full employment, more labor resources are available for


Table 1: Effects of Liberalization

Intermediates Full Trade BalanceLiberalization Liberalization Preserving Devaluation

GDP 11.66 21.81 16.01Services 10.61 23.74 19.40Constr. 13.51 12.58 18.56Exportables 11.00 12.70 111.73Importables 16.73 23.35 10.54Priv. Cons. 14.49 16.07 111.94CPI 24.47 26.96 131.83Exports 11.38 17.71 117.93Imports 14.02 127.53 122.90

other sectors. In the unemployment situation, shifts in relativeprices give an incentive for expanding output in other sectors.The total employment effect depends on the strength of the supplyresponses in these other sectors, particularly exportables. Addingreal wage rigidity, as in Dixon et al. (1986), strengthens this effect.Lower import prices mean lower nominal wages and further in-creased competitiveness of the exportables sector.

4. INTERMEDIATE IMPORTS LIBERALIZATION

The import rationing captured in the benchmark version im-poses a supply constraint on the economy. When the firms cannotimport the intermediates they desire, they are forced to apply aninefficient mix of domestic and imported intermediates, raisingunit costs and reducing outputs compared to the nonrationedsituation. In the intermediate imports liberalization experiment,the demand functions of imported intermediates are unrationed.This allows imported intermediates to replace less suitable domes-tic ones, reducing marginal costs, and increasing output supply.An overall expansion of the economy is expected. However, thedemand for domestic intermediates per unit of output is alsoreduced, introducing a contractionary element into the process.

Table 1 shows the results of the experiment, presented as per-centage deviations from the rationed 1985 benchmark. The inter-mediate import-constrained importables and construction sectorsare the main sources of expansion. In the importables sector, thevalue added increases by 6.7%. Both supply and demand factorsstimulate the formerly compressed sectors. The reduced costs


associated with a more efficient mix of intermediates lead to arelative price fall, switching expenditures away from other goods.

The construction sector experiences a more moderate expansioninduced by the cost effect. Since most of the sector’s deliveriesgo to the exogenous investment component, stimulation fromincreased aggregate demand is limited. However, the fall in theprice of the rationed sectors reduces the price of domestic relativeto imported investment goods. This effect explains much of the3.5% increase in the sector’s value added.

As pointed out by Collier and Gunning (1993), import restric-tions often create an implicit subsidy for imported investmentgoods, since they imply an overvalued exchange rate comparedto what would give the same trade balance under free trade. Sinceinvestment goods have priority over other imports, they can beimported at a low cost. Our model identifies an additional reasonfor liberalization to reduce the price of domestic relative to im-ported investment goods. Since intermediate imports for domesticsectors producing investment goods were previously rationed, lib-eralization will reduce their price relative to imported investmentgoods from the cost side, even if there is no devaluation.

The positive supply response of the import rationed sectorsspreads out to the rest of the economy. GDP is increased by about1.7%. As a consequence of given capital stocks and supply ofskilled workers, the increase in unskilled employment is relativelyhigher—2.6%. Lower domestic costs of importables and construc-tion reduce the price of composite intermediates for the ex-portables sector, and its value added rises by about 1%. Thedemand-determined service sector 2 follows along, stimulated bythe economic expansion, but dampened by the expenditure switch-ing toward other less expensive goods.

It is interesting to note an important difference between theexperiment here and those in models with full employment. Withunemployment, increased labor use in the formerly rationed sec-tors does not automatically mean less labor for the other sectors.In the full employment models, this is in effect what is assumed;they have a contractionary element hidden in the assumptions,since employment outside the formerly rationed sectors must, byassumption, decrease. Furthermore, rationing in some sectors isassumed to increase employment in others. In our model, re-stricting intermediates to some sectors is likely to have adverseeffects on overall employment.


The favorable general expansion in the experiment brings with ita lower consumer price index (about 4.5%). The reduced domesticprice of importables outweighs price-increasing demand effects.Private consumption expands by about 4.5%, well above the grossdomestic product (GDP) effect. Increased foreign savings replacedomestic savings, as income distribution shifts toward labor. Theimplicit rents of the previous license holders go down, explainingthe shift away from profits.

The cost of intermediate goods liberalization can be read in thetrade balance statistics. Total imports rise by about 4%. Some ofthis is self-financed, since new export revenues come in. The nettrade deficit worsens by about 0.7% of GDP, a modest cost forthe policy reorientation. The deterioration in the trade balanceis moderate, partly because of the shift from imported to domesticinvestment goods, implying a reduced import bill for investmentgoods of about 0.5% of GDP.

All in all, the model results show that the short-run implicationsof liberalizing intermediate imports are limited. By extension thisalso suggests that the previous rationing cannot explain the growthstagnation during the 1980s, contrary to the understanding ofmost observers. The quantitative effects are dependent on ourassumptions about the degree of rationing taking place beforeliberalization, but the conclusion holds for a wide range of realisticparameter values. Our explanation of this limited short-run re-sponse is that by the time liberalization was introduced, firms hadadjusted to the rationed conditions. Those that were unable todo so did not survive. The economy had thus already rearrangeditself to minimize the consequences of rationing. Whereas onewould expect such a selection process to operate in any rationedeconomy, in Zimbabwe it was probably reinforced by variousexport incentive schemes, which had been introduced as early as1983. These provided a channel for firms to reduce the constrainton their access to intermediate imports for domestic production.It follows that even if the short-run response to intermediatesimport liberalization is small, the rationing over time can be animportant factor explaining stagnation.

5. FULL LIBERALIZATION

In the second liberalization experiment, protection of domesticmarkets is eliminated, and trade in all imports is opened up. Thenominal exchange rate is held constant to concentrate on the


responses to new trade arrangements. When finished goods areadded to the list of liberalized goods, four implications of previousrationing are modified. First, the protected importables sectorfaces competition from abroad. In the benchmark rationed situa-tion, the sector is treated as nontraded with exogenous importsand domestically determined prices. Now, imperfect substitutionis assumed between domestically produced and imported import-ables in a composite CES aggregate. It follows that output supplycan respond to cost and price conditions with a combination ofdomestic production and imports. Second, noncompetitive con-sumer imports come in. The demand function for noncompetitiveconsumption goods allocates consumer spending between domesticand foreign goods, and liberalization allows more of a given demandto be directed toward imported goods. Third, the new access toimported goods is assumed to reduce private savings rates, induc-ing higher consumer demand given the income level (as explainedin section 3). Fourth, trade in food is now determined by marketforces. The flexible food price implied by import controls is re-placed by one fixed by the world market. Imports and exports offood vary to clear the domestic market, rather than its price.

This more ambitious liberalization experiment has the sameexpansionary effects as the more limited intermediates liberaliza-tion: lower costs of the composite intermediate and higher domes-tic content of investment goods. In addition, reduced savings ratesmean more consumer demand out of a given income. However,countervailing contractionary effects are now in action. The open-ing up of the market for importables crowds domestic producersout of the market. The old protection was effective. The accessto noncompetitive consumer imports switches demand from do-mestically produced goods. The net effect on domestic demandis dependent on our calibration of key parameters, in particularthe desired switching to foreign consumer goods, the degree ofsubstitution between domestic and imported importables, and thedrop in private savings rates when imports are available. Theassumptions explained in the appendix section A.4 represent ourbest judgment. Accordingly, the net effect is a drop in demandtoward domestic goods. The appendix shows the range of parame-ters for which this holds true.

When the two types of expenditure switching dominate, anddemand is reduced, the domestic activity level goes down. GDPis reduced by close to 2% in the experiment reported in Table 1.


Again, the unskilled employment effect is relatively stronger thanthe GDP effect. Employment of unskilled workers falls by 3.3%.

The contraction can be traced back to the importables sector5. The previously protected importables sector is a key player inthe liberalization process. In the old regime, both intermediateinputs and finished importable goods were rationed. When theprotection is removed, the loss of market share is dominant. Thereduced costs associated with free access to imported intermedi-ates are not sufficient to defend the position at the domesticmarket. The changes in the composition of demand because ofliberalized noncompetitive consumer imports, add to the contrac-tionary forces. If Figure 1 is interpreted as the importables market,we end up in the new equilibrium d.

Liberalization generates structural adjustment away from non-tradable and importable goods toward exportables and domesticconstruction. The fall in intermediate input prices (importablesand construction) stimulates exportable supply, and its valueadded rises by almost 3%. The nontraded service sector contracts(less than 4%) due to income contraction and demand switchingbecause of higher relative price. The structural adjustment is ac-cording to the standard Washington recipe, but the expansion ofexports is too slow to compensate for the loss of domestic marketshares in the short-run. This is true even though the drop indomestic prices is quite large: the consumer price index (CPI) isreduced by about 7%. The relative price of importables falls withliberalization. Income distribution shifts away from profits, andfalling private savings rates, explain an expansion of private con-sumption by more than 6%, even when the economy contracts.

The short-run structural adjustment where contraction goeshand in hand with a consumption boom is paid for by a hugetrade deficit. Noncompetitive consumer imports and competitiveimportables goods are in high demand, and total imports rise bymore than 27%. Some of the added import flow is financed byclose to 8% exports revenue expansion, resulting from lower costsand lower demand for domestic goods. The demand effect notonly means less domestic consumption of export sector goods.Since the trade in food is also opened up, lower demand fordomestic goods implies food exports. Still, the trade deficit deterio-rates seriously, to close to 8% of GDP. The deficit is comparableto the first years of independence and is hardly sustainable. Thedeficit and the deindustrialization associated with it are a highshort-run price of liberalization, even if the long-run effects may


be favorable. As pointed out by Collier and Gunning (1992), tradeliberalization may be incompatible. When private agents observethe large effect on the trade balance, they will expect future devalu-ations. Hence, imports are temporarily cheap. An import boomstronger than the model predicts may be the result, worsening thetrade balance and making the liberalization even less sustainable.

The results above explain why trade liberalization is usuallyaccompanied by devaluation. A devaluation may counteract thesustainability problem and the politically unacceptable contrac-tionary consequences of liberalization. A trade-liberalizing deval-uation assumed to maintain the initial trade balance with fullliberalization is investigated in section 6. It is shown that thedevaluation necessary is quite large, and that the price level isdriven up to politically dangerous levels. Given the resistance tosuch a strong devaluation, other ways of avoiding the contractionare of interest. Two experiments, discussed in section 7, are con-structed to show how the level of GDP can be held unaffectedby liberalization—foreign aid and output preserving devaluation.

6. TRADE BALANCE–PRESERVING DEVALUATION

Devaluation is potentially an effective instrument for avoidingthe unfavorable effects of liberalization described above. The de-valuation necessary to keep the trade balance constant (in foreigncurrency) is investigated. In Collier’s (1991) terminology, this isa trade-liberalizing devaluation as opposed to a payments-improv-ing devaluation. Given the deterioration of the trade balanceshown in the full liberalization experiment, quite a large devalua-tion is needed: to leave the foreign balance unaffected, the modelasks for a nominal devaluation of 82%. Such dramatic changescould induce behavioral changes that would strain the ceterisparibus assumptions underlying our counterfactual experiments.Nevertheless, the main reactions discussed below are of interest.Table l sums up the impact on the main economic variables.

The trade-liberalizing devaluation expands the economy, withincreased GDP of 6% compared to preliberalization. The exportsector receives the most direct price signal and experiences thehighest growth. The combined devaluation and removal of ra-tioning imply a significant fall in the relative price of domesticto imported investment goods. When liberalization includes a


devaluation, the incentives to turn investment demand into domes-tic goods are stronger. The construction sector responds by in-creasing its value added by almost 9%, somewhat dampened be-cause of the increased costs of imported intermediates. Comparedwith the other sectors, the importables sector is almost unaffected.Increased availability of intermediates and increased demand areoffset completely by loss of market shares. The price effects ofincreased costs of imported intermediates and the general demandexpansion favor the demand determined service output.

The relative price shifts induced by the devaluation stimulateexports and discourage imports. Even with the increased activityin the economy, this effect is sufficiently strong to reduce imports(compared to liberalization with no devaluation). Even if importsof intermediates increase because of increased production, thereduction in consumer imports because of the strong relative priceshift is enough to outweigh this. In the savings-investment balance,the drop of private savings rates when imports are available iscompensated by higher private incomes, as foreign savings arestill basically zero.

On the price side it is worth noting that the 82% devaluationturns a fall in the CPI into a rise of above 30%. The consequenceis a sharp decline in real wages for unskilled workers. Since virtualprices are higher than actual under rationing, their utility levelis not necessarily reduced equivalently. But if compensation isdemanded, the required nominal devaluation is even higher, andthe economy may easily be thrown into accelerating inflation. Ourinterpretation of the trade-liberalizing devaluation effects is thatit solves some problems of the liberalization package by creatingother new ones. Balancing the two is a delicate political challenge.

7. OUTPUT PRESERVING FOREIGN AIDAND DEVALUATION

In the structural adjustment debate, the necessity of foreign aidinflow with reform is often emphasized. Pessimistic assessmentssuggest that foreign aid leads to deindustrialization. Starting outwith the output contraction under full liberalization (with no de-valuation), we ask how much aid is needed to preserve output(GDP). The experiment throws light on the role of foreign aid inthe economy. Furthermore, it is of interest to contrast the effectswith an output preserving devaluation, since changes in the sec-toral composition of the economy depend on how output is pre-vented from falling during liberalization.


Table 2: Output Preserving Aid and Devaluation

Output Preserving Aid Output Preserving Devaluation

GDP 11.85 11.85Services 13.24 12.82Construction 10.15 11.46Exportables 20.78 12.79Importables 13.20 11.20Priv. Cons. 17.46 11.17CPI 11.33 17.05Exports 23.43 12.89Imports 14.74 20.49

Foreign aid inflow with liberalization is part of the Washingtonprograms and is intended to overcome the threats to sustainabilityand credibility. This inflow is another possible source of expansion.The revenue involved adds to domestic demand during a periodwhen demand is switched to foreign goods. This demand effect ishighlighted here. (The aid is modeled as part of net factor pay-ments from abroad (NFPA), which feeds directly into private in-comes). An inflow of about 300 mill. Zimbabwe 1985-dollars willdo the job, amounting to a little more than 4% of GDP. Theconsequences of such an additional injection are shown in Table 2.

Compared with the effects of full liberalization, nontraded ser-vices and importables production are the main winners, as ex-pected when consumption demand goes up. The rise in consump-tion demand is significant, about 7 percentage points. The effecton the price level is small, however, since much of the demandis directed toward the service sector where the supply curve isquite flat.

The demand-expanding effect of foreign aid implies a worseningof the trade balance, although the current account is improved.If aid should be given so as to prevent the current account balancefrom worsening vis a vis the benchmark, considerably more aidthan implied by this experiment is necessary. The conclusion isin accordance with the analysis of aided trade liberalization inCollier and Gunning (1992, p. 936): “The enormous foreign ex-change requirement is liable to raise a question mark at the onsetof the reform as to sustainability.”

If aid inflow does not increase, devaluation of a little above 13%is necessary to prevent output from falling. Of course, the effects


on the economy go the same direction as with the trade balancepreserving devaluation, although the changes are smaller becausethe devaluation is substantially smaller. It is of interest to comparethe structure of the economy with the aid experiment. The keything to note is that whereas imports are higher and exports lowerwith aid, the opposite is the case here. Compared with increasedeconomic activity by devaluation, increased economic activitythrough demand injection has Dutch disease symptoms.

8. CONCLUDING REMARKS

The model analysis has shown possible reactions to trade liberal-ization in the Zimbabwean economy. Liberalization of importedintermediates will help ease the previous supply constraint and istherefore likely to be expansionary. The recipe is not costless,however, as the trade balance will worsen. Full liberalization elimi-nates the protection of domestic markets for finished goods.Allowing the demand functions for finished goods to operate canhave contractionary effects. Demand is switched away from do-mestically produced goods toward imports. Deindustrialization inthe domestic economy may result. Sustainability problems relatedto trade deficit can be hard to handle.

In efforts to overcome threats to sustainability and credibility,trade liberalization is typically accompanied by devaluation. De-valuation is expansionary and improves the trade balance ac-cording to our results, but must be large to reduce the trade deficitsignificantly. New sustainability problems arise, related to inflationand possible claims for compensation.

Foreign aid inflow is a way of coping with the trade deficit andadding demand expansion to reform. Foreign aid can help avoidcontraction, but the trade deficit will be even higher. Because ofthe demand effects of aid, the aid necessary to balance the currentaccount is beyond realistic numbers. Compared with output-reserving devaluation, demand injection by aid is counterproduc-tive if a shift to exports has high priority. Aid has Dutch diseaseeffects.

Our analysis has shown that trade liberalization creates short-run adjustment problems, and can explain why liberalization hasbeen controversial. Income distribution shifts identified when im-ports flow into the domestic market help us understand oppositionto reform. The issue is discussed in the overview by Rodrik (1994):“if reform is such a great idea, why are governments typically so


reluctant to undertake it?” The usual answer is that even if tradeliberalization improves resource allocation, there are winners andlosers at the micro level. The losers can clearly be identified asthe license holders, since they get cheap imports and high pricesin the domestic market. If losers outweigh winners in politicalinfluence, they can prevent trade liberalization from taking place.In Zimbabwe, this argument goes a long way in explaining theskepticism to trade liberalization in the protected industrial estab-lishment. Trade restrictions have been a way of hiding transfersthat are otherwise difficult to undertake.

APPENDIX: MODEL DOCUMENTATION

We present here only the essentials of the model; full documen-tation of the benchmark model is available in Rattsø and Torvik(1992). The equation set and the full list of symbols are shownbelow. The supply side of the model is described by equations 2to 25 and is brought together with the demand side by the marketclearing equation 1. The output of agriculture, sector l, is exoge-nously given by capacity (equation 2) and a flexible price levelarranges the market-clearing. Services, sector 2, are assumed tobe demand determined with markup pricing (equation 3).

In the other three sectors, production functions include skilledand unskilled labor, domestic and imported intermediates, andreal capital. They are specified as multilevel CES and Leontieffunctions. For each sector, output X is produced by a value addedaggregate and an intermediate goods aggregate, N, in fixed propor-tions. The value added is a CES function of capital stock, K, anda labor aggregate, L, (shown by the term in parentheses in equation4). The capital stock is historically given. Labor is a CES aggregateof unskilled Lu and skilled labor Ls. Intermediate goods, N. are aCES-aggregate of imported intermediates, II, and domestic inter-mediates, DI. Domestically produced intermediate goods arelinked to aggregate DI by fixed coefficients.

The allocation of intermediate imports is a key aspect of thesupply side adjustment. Exportables (sector 4) are assumed tohave priority, so that their demand for imported intermediateis satisfied (equation 6). However, construction (sector 3) andimportables (sector 5) are both import dependent and rationed.They are assumed to consume whatever quantity of importedintermediates they are given through the allocation system. The


rationed imported intermediates feed into the CES aggregate forintermediate goods (equation 8).

Prices in the three sectors are set equal to marginal costs (equa-tion 9), and the marginal costs are the weighted sum of the mar-ginal costs of intermediates and value added (equation 10).

The exportables sector is a price taker in the world market(equation 11). Since the sector is not rationed, the marginal costof its intermediate goods aggregate is the price of the aggregate(equation 12). The marginal cost of the value added aggregate isderived from the value-added production function in the normalway (equation (13). Export volume is determined as a residualgiven the supply function and the domestic demand for exportables.

The use of domestic intermediates in the rationed sectors isdetermined by standard cost-minimizing conditions. The part ofthe marginal cost associated with intermediates is influenced bythe rationing of imported intermediates (equation 14). Equations15 to 20 define costs and input-output coefficients, while the labormarket is described by equations 21 to 25. This combines a fixed-wage demand determined unskilled labor market and a full em-ployment wage-clearing skilled labor market. Unit labor costs aredetermined as a CES aggregate, and substitution between the twoskill types follows from varying demand for skilled workers.

The income generation and consumption demand aspects ofthe model capture distributional effects and rationing of importedconsumer goods (equations 26 to 34). The four income groupsdefined above have different propensities to save, an importantaspect of overall savings formation. Savings rates are influencedby access to noncompetitive imported consumer goods (equation31). The level of imports consumers desire is determined as aconstant fraction of net consumer expenditures (equation 32).However, rationing of consumer imports means that these desirescannot be fulfilled, leading to postponed consumption and in-creased demand for domestic consumer goods.

Equations 35 to 37 model the public sector accounting. Equa-tions 38 to 41 handle the foreign exchange rationing. The outlaysfor competitive imports, investment goods, and priority intermedi-ate imports to exportables are subtracted from the sources offoreign exchange to determine the rationed amount (assumed tobe positive) (equation 38). This is then allocated between importsof agricultural goods (equation 39), imported intermediates for


construction and importables, (equation 40) and imported non-competitive consumer goods (equation 41). The specification im-plies priority of imported food over intermediates and consump-tion goods; food imports are related to a target C1* for foodconsumption.

Investment is assumed to be a CES aggregate of sector 3, sector5, and imported investment goods. The composition of the invest-ment aggregate is made dependent on relative prices as in equa-tions 42 and 43. The price of the investment aggregate follows byequation 44. Finally, equation 45 is a consistency check of theinvestment-savings balance.

A1. Benchmark Rationing Model

Xi 5 o5

j51

aij · Xj 1 Ci 1 Gi 1 Ji 1 Ei 2 Mi 1 DSi (i 5 1–5) (1)

X1 5 XB1 (2)

P2 5(1 1 t2)(1 1 t2)

1 2 (1 1 t2)(1 1 t2)a22

[a12 · P1 1 a32 · P3 1 a42 · P4 1 a52 · P5 1 q2 · b2]

(3)

Xi 51

1 2 ni

([m1i · L2(12hi)/hii 1 m2i · K2(12hi)/hi]2hi/(12hi)) (i 5 3–5) (4)

Ni 5 ni · Xi (i 5 3–5) (5)

II4 5 ed4141e · P*04

MCN422d4

N4 (6)

DI4 5 ed4241 PDI4

MCN422d4

N4 (7)

Ni 5 [e1i · II2(12di)/dii 1 e2i · DI2(12di)/di

i ]2di/(12di) (i 5 3,5) (8)

Pi 5 MCi (i 5 3–5) (9)

MCi 5 ni · MCNi 1 (1 2 ni)MCVAi (i 5 3–5) (10)

P4 5 e · P*4 (11)

MCN4 5 [ed414(e · P*04)12d4 1 ed4

24 · P12d4DI4 ]1/(12d4) (12)

MCVAi 5qi

m1i

[m1i · L2(12hi)/hii 1 m2i · K2(12hi)/hi

i ]1/(12hi) · L1/hii (i 5 3–5) (13)

MCNi 5PDIi

e2i

[e1i · II2(12di)/dii 1 e2i · DIi

2(12di)/di]1/(12di) · DIi1/di

i (i 5 3,5) (14)


PDIj 5 o5

i51

fij · Pi (j 5 3–5) (15)

Bj 5 o5

i51

aij · Pi 1 a0j · e · P*0j 1 qj · bj (j 5 2–5) (16)

Aij 5 fij · DIj (i 5 1–5, j 5 3–5) (17)

a0j 5IIj

Xj

(j 5 3–5) (18)

aij 5Aij

Xj

(i 5 125, j 5 3–5) (19)

Li 5 bi · Xi (i 5 2–5) (20)

qi 5 [asii · wui

12si 1 bsii · w12si

s ]1/(12si) (i 5 2–5) (21)

Lui 5 asii (

wui

qi

)2siLi (i 5 2–5) (22)

Lsi 5 bsii (

ws

qi

)2siLi (i 5 2–5) (23)

Lu 5 o5

i52

Lui 1 Lug (24)

Ls 5 o5

i52

Lsi 1 Lsg (25)

Ya 5 P1 · X1 (26)

Yu 5 o5

i52

wui · Lui 1 wug · Lug (27)

Ys 5 o5

i52

ws · Lsi 1 ws · Lsg 1 NFPA (28)

Yz 5 o5

i52

Pi · Xi 2 Bi · Xi 2ti

11ti

Pi·Xi (29)

D 5 (1 2 sa)Ya 1 (1 2 su)Yu 1 (1 2 ss)Ys 1 (1 2 sz)(1 2 tz)Yz (30)

sj 5 s*j 1 gj1CIMPD 2 CIMPA

CIMPD2

2

(j 5 a, u, s, z) (31)

CIMPD 5mc

e · P*0c

(D 2 DB) (32)

DB 5 o5

i51

Pi · ui (33)


Ci 5 ui 1mi

(1 2 mc)Pi

(D 2 DB 2 e · P*0c · CIMPA) (i 51–5) (34)

GREV 5 tz · Yz 1 TIND (35)

TIND 5 o5

i51

ti

1 1 ti

Pi · Xi (36)

GEXP 5 o5

i51

Pi · Gi 1 wug · Lug 1 ws · Lsg

2 (P1 2 e · P*1 )M1 2 (P5 2 e · P*5 )M5 (37)

RAT 5 2e · P*5 ·M5 2 e · P*0j · J0 2 e · P*04 · II4

1 DEF 1 P2 · E2 1 P4 · E4 1 NFPA (38)

M1 5 M*1 1 h01 · (C*1 2 C1) (39)

IIi 5 II*i 1 h0i ·(RAT 2 e · P*

03 · II*3 2 e · P*05 · II*5 2 e · P*1 · M1)e · P*0i

(i 5 3, 5)

(40)

CIMPA 5 h0c ·(RAT 2 e · P*03 · II*3 2 e · P*05 · II*5 2 e · P*1 · M1)

e · P*0c

(41)

Ji 5 pti (

Pi

PCOMPJ

)2tJ (i 5 3, 5) (42)

J0 5 pt0(

e · P*0j

PCOMPJ

)2tJ (43)

PCOMPJ 5 [pt

0 · (e · P*0j)12t 1 pt3 · P12t

3 1 pt5 · P12t

5 ]1/(12t) (44)

P3 · J3 1 P5 · J5 1 e · P*0j · J0 1 o5

i51

Pi · DSi 5

sa · Ya 1 su · Yu 1 ss · Ys 1 sz(1 2 tz)Yz 1 GREV 2 GEXP 1 DEF (45)

Endogenous variables The model consists of 117 equations: 116of these are independent since the investment-savings balance canbe calculated from the rest of the equations. We therefore havethe following 116 endogenous variables:

Aij : intermediate deliveries i 5 1–5, j 5 3–5a0j : intermediate imports-output coefficients sectors

j 5 3–5aij : intermediate input-output coefficients sectors j 5 3–5,

i 5 1–5Bi : variable costs pr. unit of output sectors i 5 2–5bi : labor-output coefficients sectors i 5 3–5


Ci : total consumption levels, i 5 1–5CIMPD : desired noncompetitive consumer importsCIMPA : actual noncompetitive consumer importsD : consumer spendingDB : spending for floor consumptionDIi : domestic intermediate goods aggregate sectors

i 5 3–5E4 : exportsGEXP : gov. spendingGREV : gov. revenueJi : investment deliveries from sectors i 5 0,3,5IIi : intermediate imports sectors i 5 3–5Lui : unskilled workers sectors i 5 2–5Lsi : skilled workers sectors i 5 2–5Lu : total unskilled workersLi : labor aggregates sectors i 5 2–5M1 : imports of agricultural goodsMCi : marginal cost sectors i 5 3–5MCNi : marginal cost of intermediate goods aggregate sectors

i 5 3–5MCVAi : marginal cost of value added aggregate sectors

i 5 3–5Ni : intermediate goods aggregate to sectors i 5 3–5Pi : sectoral price levels i 5 1–5PDIi : price of domestic intermediate goods aggregate sectors

i 5 3–5PJ

COMP : price of investment aggregateqi : labor cost sectors i 5 2–5RAT : foreign exchange available for rationed goodssi : savings rates i 5 a, u, s, zTIND : indirect taxesws : wage rate skilled workersXi : sectoral output levels, i 5 1–5Ya : agr. incomeYs : wage income skilled workersYu : wage income unskilled workersYz : profit income

Parameters

aij : input-output coefficients i 5 1–5, j 5 1–2b2 : labor-output coefficient sector 2


C1* : target on food consumptionfij : shares in domestic input aggregate sectors j 5 3–5,

i 5 1–5hoi : rationing parameters i 5 1,3,5,CIIi : minimum level of imported intermediates sectors i 5 3,5M1*: constant in agricultural imports functionmc : propensity to consume noncompetitive importsmi : marginal propensity to consume: i 5 1–5ni : intermediate goods-output ratio sectors i 5 3–5s*i : constant in savings functions i 5 a,u,s,zti : indirect tax rates sectors i 5 1–5tz : tax rate of profit incomeai : distribution parameter of unskilled labor in CES aggregatebi : distribution parameter of skilled labor in CES aggregategi : savings respond parameter i 5 a,u,s,zG : elasticity of substitution in investment aggregatedi : elasticity of substitution in intermediate aggregate sectors

i 5 3–5e1i : distribution parameter of intermediate imports in intermedi-

ate aggregate sectors i 5 3–5e2i : distribution parameter of domestic intermediates in interme-

diate aggregate sectors i 5 3–5hi : elasticity of substitution in value added aggregate sectors

i 5 3–5ui : floor consumption levels i 5 1–5m1i : distribution parameter of labor in value added aggregate

sectors i 5 3–5m2i : distribution parameter of capital in value added aggregate

sectors i 5 3–5pi : distribution parameter in investment aggregate from sectors

i 5 0,3,5si : elasticity of substitution in labor aggregate i 5 2–5t2 : markup rate sector 2

Exogenous variables

DSi : changes in stocks i 5 1–5DEF : trade deficitE2 : exports sector 2e : exchange rateGi : gov. demand for commodities i 5 1–5J : investment aggregate


Ki : capital in sectors i 5 3–5LBs : supply of skilled laborLug, Lsg : government employmentM5 : competitive imports, sector 5 goodsNFPA : net factor payments from abroadPi* : foreign price of sectoral goods i 5 1,4,5P0i : foreign price of intermediate imports to sectors

i 5 1,4,5P0j : foreign price of investment goodswui, wug : wage rates unskilled workers i 5 2–5XB1 : output level sector 1

A2. Intermediates Liberalization Model

In the first liberalization model alternative, the rationing ofimported intermediates is removed, and intermediate demandfunctions for sectors 3 and 5 operate as for the initially unrationedexports sector 4. The rationing rule (equation 40) is eliminated,and the demand is determined as in (equation 6). The demandfor domestically produced intermediates is determined as in equa-tion 7, and equation 8 can be eliminated. The rationed marginalcost functions (equation 14) are replaced by (equation 12). Foodand noncompetitive consumer imports are still rationed, and M1

and CIMPAA are exogenous, whereas the rationing rules (equa-tions 39 and 41) are eliminated. The trade deficit is endogenous,and follows from the revised equation (38):

DEF 5 e · P*1 · M1 1 e·P*5 · M5 1 e · P*03 · II3 1 e · P*04 · II4 1 e · P*05 · II5

1 e · P*0c · CIMPA 1 e · P*0jJ0 2 P2 · E2 2 P4 · E4 2 NFPA

A3. Full Liberalization Model

In this version, we remove protection from imported import-ables and noncompetitive consumer goods. We introduce a com-posite importables good X*5 . It is a CES aggregate of domesticallyproduced and imported importables determined by

X5 5 jc1 1 P5

PCOMP5

22c

X*5

M5 5 jc2 1 e·P*5

PCOMP5

22c

X*5

PCOMP5 5 [jc

1 · P12c5 1 jc

2 (e · P*5 )12c]1/(12c)


where j1 and j2 are distribution parameters for domestically pro-duced and imported goods in the CES aggregate, c the elasticityof substitution, and PCOMP

5 the price of the aggregate (replaces P5

in all demand functions). Equation 1 for sector 5 is now a marketbalance for only the domestic components of demand, and M5 isleft out of the equation.

The liberalization of noncompetitive consumer goods impliesthat imports are determined by demand in equation 32 whereCIMPD is replaced by CIMPA). The consumer demand functions(equation 34) are reformulated to:

Ci 5 ui 1mi

Pi

(D 2 DB) (i 5 1–5)

Since the domestic market for food is liberalized, imports M1 aremade endogenous, with P1 determined by the world price and theexchange rate. The market equilibration is changed from a flex-price regime to an import adjusting regime in the standard openeconomy fashion.

A4. Calibration

The benchmark model is built around a social accounting matrix(SAM) for 1985 documented by Rattsø and Torvik (1992). Allvariables are calculated in millions of Zimbabwe dollars at 1985prices. The parameters and the exogenous variables of the modelare set so as to have the endogenous variables consistent with theSAM. Savings rates differ between economic groups (sa 5 su 50.22, ss 5 0.27, sz 5 0.5187). The consumption demand parametersfollow from assumptions about the income elasticities (0.8 foragriculture and 1.05 for services) and the supernumerary incomeratio (0.48).

Other key parameter values used in the simulations are as fol-lows:

Elasticities of substitutionin labor aggregate, s 5 0.5, i 5 2–5in intermediate goods aggregate, di 5 0.333, i 5 3–5in value added aggregate, hi 5 0.5, i 5 3,5 and 0.2, i 5 4in investment aggregate, G 5 0.5in sector 5 Armington aggregate, c 5 0.8

Constants in savings functions, s*a 5 0.22, s*u 5 0.18, s*s 5 0.23,s*z 5 0.4

Savings response parameters, ga 5 0, gu 5 gs 5 0.09, gz 5 0.2671


Rationing parameters, h01 5 1, h03 5 0 0535, h05 5 0.3512,h0C 5 0.5953

Constant in food import function, M*1 5 36Minimum levels of imported intermediates, I3 5 90, I5 5 600

The effect of liberalization depends in part on the pattern andextent of previously existing rationing of consumers and produc-ers. Although there is little evidence of this, we judge the followingassumptions to represent the Zimbabwe story. The productionfunctions determine the desired demand for imported intermedi-ates. Small elasticities of substitution capture the rigid productionstructure of the economy. The pure substitution effect of rationingintermediate imports assumes that 91% of the demand for inter-mediate imports by construction and importables is satisfied. Sen-sitivity analysis shows that if instead 95% of this demand hadbeen satisfied before liberalization, the economic expansion isabout 1 percentage point lower than in the experiment reportedin Table 1.

Noncompetitive consumer imports and imports of sector 5goods are assumed to be strictly rationed, with the actual importssatisfying only one-third of actual demand. The assumption isimportant for the demand switching and the savings response.Given the savings response parameters stated above, private sav-ings rates fall by about 5% points for labor and by about 15%points for profit income with liberalization. Even with this strongdrop in private savings, the demand switching dominates in thefull liberalization experiment in Table 1. If the rationing of non-competitive consumer imports and imports of sector 5 goods isabout one-half of actual demand in the base year, the demandswitching is reduced, and the model predicts output expansionwith full liberalization.

REFERENCESBenjamin, N.C. (1992) What Happens to Investment under Structural Adjustment: Results

from a Simulation Model. World Development 20(9):1335–1344.Bevan, D., Collier, P., and Gunning, J.W. (1990) Controlled Open Economies: A Neoclassi-

cal Approach to Structuralism. Oxford: Clarendon Press.Buffie, E. (1984) The Macroeconomics of Trade Liberalization. Journal of Development

Economics 17:121–137.Chhibber, A., Cottani, J., Firuzabani, R., and Walton, M. (1989) Inflation, Price Controls,

and Fiscal Adjustment in Zimbabwe. Working Paper 192, Country EconomicsDepartment, The World Bank, Washington, DC.


Collier, P. (1991) Exchange Rates and Exchange Controls in Developing Countries. GreekEconomic Review 12:132–147.

Collier, P., and Gunning, J.W. (1992) Aid and Exchange Rate Adjustment in AfricanTrade Liberalizations. Economic Journal 102:925–939.

Collier, P., and Gunning, J.W. (1993) The Impact of Liberalization on Private Investment.Mimeo, Centre for the Study of African Economies, University of Oxford.

Condon, T., Robinson, S., and Urata, S. (1985) Coping with a Foreign Exchange Crisis: AGeneral Equilibrium Model of Alternative Adjustment Mechanisms. In EconomicEquilibrium: Model Formulation and Solution (A. Manne, Ed.). Amsterdam: North-Holland.

Dailami, M., and Walton, M. (1989) Private Investment, Government Policy, and ForeignCapital in Zimbabwe. Working Paper 248, Country Economics Department, TheWorld Bank, Washington, DC.

Davies, R. (1991) Trade, Trade Management and Development in Zimbabwe. In Tradeand Development in Sub-Saharan Africa (J. Frimpong-Ansah, S. M. Ravi Kanbur,and P. Svedberg, Eds.). Manchester: Manchester University Press.

Davies, R., and Rattsø, J. (1993) Zimbabwe. In The Rocky Road to Reform: Adjustment,Income Distribution and Growth in the Developing World (L. Taylor, Ed.). Cam-bridge, MA: MIT Press.

Davies, R., and Rattsø, J. (1996) Growth, Distribution and Environment: MacroeconomicIssues in Zimbabwe. World Development 24(2):395–405.

Davies, R., Rattsø, J., and Torvik, R. (1994) The Macroeconomics of Zimbabwe in the1980s: A CGE-Model Analysis. Journal of African Economies 3(2):1–46.

Dixon, P.B., Parmenter, B.R., and Rimmer, R.J. (1986) ORANI Projections of the Short-Run Effects of a 50% Across-the-Board Cut in Protection Using Alternative DataBases. In General Equilibrium Trade Policy Modeling (T.N. Srinivasan and J. Whal-ley, Eds.). Cambridge, MA: MIT Press.

Gibson, B. (1985) A Structuralist Macromodel for Post-Revolutionary Nicaragua. Cam-bridge Journal of Economics 9:347–369.

Government of Zimbabwe (1991) Zimbabwe: A Framework for Economic Reform 1991–95.Grais, W., de Melo, J., and Urata, S. (1986) A General Equilibrium Estimation of the

Effects of Reductions in Tariffs and Quantitative Restrictions in Turkey in 1978.In General Equilibrium Trade Policy Modeling (T.N. Srinivasan and J. Whalley,Eds.). Cambridge, MA: MIT Press.

Green, R., and Kadhani, X. (1986) Zimbabwe: Transition to Economic Crises 1981-1983:Retrospect and Prospect. World Development 14(8): 1059–1083.

Mehlum, H., and Rattsø, J. (1993) Import Compression and Growth Constraints in Zim-babwe. Mimeo, Department of Economics, University of Trondheim, Norway.

Melo, J. de, and Robinson, S. (1989) Product Differentiation and the Treatment of ForeignTrade in Computable General Equilibrium Models of Small Economies. Journalof International Economics 27:47–67.

Morande, F., and Schmidt-Hebbel, K. (1991) Macroeconomics of Public Sector Deficits:The Case of Zimbabeew. Working Paper 688, Country Economics Department,The World Bank, Washington, DC.

Neary, J.P., and Roberts, K.W.S. (1980) The Theory of Household Behavior under Ra-tioning. European Economic Review 13:25–42.

Ocampo, J.A. (1987) The Macroeconomic Effect of Import Controls: A Keynesian Analy-sis. Journal of Development Economics 27:285–305.

Pakkiri, L., and Moyo, N.P. (1986) Foreign Exchange Politices: The Case of Zimbabwe.IDRC Workshop on Economic Structure and Macroeconomic Management, Har-are, Zimbabwe.


Rattsø, J. (1994a) Medium Run Adjustment under Import Compression: MacroeconomicAnalysis Relevant for Sub-Saharan Africa. Journal of Development Economics45:35–54.

Rattsø, J. (1994b) Devaluation and Monetary Policy with Import Compression. OpenEconomies Review 5:159–175.

Rattsø, J. (1994c) The Macroeconomic Role of Speculative Assets under Import Compres-sion and Financial Repression. Mimeo, Department of Economics, University ofTrondheim, Norway.

Rattsø, J., and Torvik, R. (1992) An Applied General Equilibrium Model for Zimbabwe—Documentation of Model, Social Accounting Matrix 1985 and Calibration. Mimeo,Department of Economics, Universities of Oslo and Trondheim, Norway.

Rodrik, D. (1994) Trade and Industrial Policy Reform in Developing Countries: A Reviewof Recent Theory and Evidence. In Handbook of Development Economics, vol III(J. Behrman and T.N. Srinivasan, Eds.). Amsterdam: North-Holland.

Taylor, L. (1983) Structuralist Macroeconomics. New York: Basic Books.Torvik, R. (1993) Optimal Saving in the Presence of Quotas: Macroeconomic Effects

of Trade Liberalization. Mimeo, Department of Economics, University of Oslo,Norway.

Torvik, R. (1994) Trade Policy under a Binding Foreign Exchange Constraint. Journal ofInternational Trade and Economic Development 3:15–31.

short-run consequences of trade liberalization: a computable general equilibrium model of zimbabwe

Documents