UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Vertical effective protection under mixed pricing: a study of the Canadian food industry Westera, R. S. 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1992_fall_westera_r_s.pdf [ 2.01MB ]
Metadata
JSON: 831-1.0086546.json
JSON-LD: 831-1.0086546-ld.json
RDF/XML (Pretty): 831-1.0086546-rdf.xml
RDF/JSON: 831-1.0086546-rdf.json
Turtle: 831-1.0086546-turtle.txt
N-Triples: 831-1.0086546-rdf-ntriples.txt
Original Record: 831-1.0086546-source.json
Full Text
831-1.0086546-fulltext.txt
Citation
831-1.0086546.ris

Full Text

VERTICAL EFFECTIVE PROTECTION UNDER MIXED PRICINGA Study of the Canadian Food IndustrybyR.S. WESTERAB.Sc. (Agricultural Economics), University of British ColumbiaA THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIES(Department of Agricultural Economics)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIASeptember 1992© R.S. Westera, 1992In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensive’copying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.Department of__________________The University of British ColumbiaVancouver, CanadaDate (S b?’2992DE-6 (2/88)11AbstractThe intent of this study is to adjust the measure of effective protection across nationalindustries by accounting for industry-specific pricing regimes and thereby more accuratelyestimate the direct and indirect resource allocation effects resulting from changes in tariffsand other protective instruments. The inquiry is cast in a partial equilibrium frameworkdesigned in part to evaluate and compare the measurement of effective protection asdependent upon the underlying pricing assumptions. Trade models traditionally model priceformation with the Ricardian assumption that suppliers will price to the limit of thedomestic price plus the tariff on their competing imports. Econometric eviHnee, howeverrsuggests that prices in certain imperfectly competitive markets are more a mixed functionof both the competing import price and the level of domestic costs in these industries.The empirical dimension of the study is an input-output construction that measureseffective protection according to these two types of pricing assumptions, and examines thechanges in protection that are implicit in the implementation of the 1988 Canada-U.S. TradeAgreement (CUSTA) and as it chiefly affects the Canadian food producing and processingsectors. The food processing sector is of particular interest because of the presence ofvarious degrees of imperfectly competitive behaviour among these industries and because,in contrast to most other manufacturing activities, this sector is an important user ofcommodities that are quite heavily protected. The Canadian economy is aggregated hereinto one consisting of 31 industries that feature 29 of these as comprising the food industry.mAllowing for mixed pricing suggests that the standard law-of-oneprice assumptionsignificantly overestimates effective protection in those processing and manufacturing sectorswhere imperfect competition is an important consideration. Not to account for mixed pricingin trade liberalization models may therefore seriously overestimate the calculation of gainsto trade for these industries. The results also suggest that the CUSTA continues to protectprimary agriculture while maintaining pressure on added value in the processing sector.Moving to free trade would likely result in some gains to the processing sector but valueadded in primary agriculture would be under pressure to contract by as much as 20 per centover the whole sector.ivTable of ContentsAbstract iiTable of Contents ivList of Tables viFigure viiAcknowledgements viiiChapter One - Overview 1Section 1.1 Introduction 1Section 1.1 Problem and objectives 5Section 1.2 Outline 9Chapter Two - Effective Protection and Industry Structure 12Section 2.1 Partial Equilibrium Analysis of Effective Protection 14Section 2.2 Methodological Considerations 182.2.1 Substitution 192.2.2 Non-tradeable intermediates and exchange rates 22Section 2.3 Imperfect competition and mixed pricing 242.3.1 Imperfect competition and product differentiation 262.3.2 Imperfect competition and Canadian trade modelling 292.3.3 Mixed pricing 34Chapter Three - An Input-Output Model of Vertical Protectionunder Mixed Pricing 37Section 3.1 The Model 37Section 3.2 Trade Policy Simulation: Solving the Model 43Chapter Four - Specification of the Model 46Section 4.1 Input-output Data 46Section 4.2 Pricing Parameters 52Section 4.3 Protection Data and Tariff Parameters 54VChapter Five - Results.57Section 5.1 CUSTA (1988): The Base Case with NoMacro Response 57Section 5.2 Free Trade with No Macro Response 61Section 5.3 CUSTA (1988): The Base Case with Depreciation 64Section 5.4 CUSTA (1988): The Base Case with Appreciation 66Section 5.5 Rankings of Effective Protection 67Chapter Six - Summary, Conclusions, and Recommendations 69Bibliography 76Appendices 76viList of Tables4.1 Shares of sector and resource costs 504.2 Tariffs, PSE’s and pricing parameters 545.1 Scenarios 1 and 2: CUSTA and Free Trade with e =1.0 615.2 Scenarios 3 and 4: CUSTA with appreciation (e =1.1)and depreciation (e = 0.9) 685.3 Rankings of ERPs and the rate of return 71vList of Figures2.1 Partial analysis of effective protection 16vifiAcknowledgementsGrateful acknowledgement is made here to Tim Hazledine, my advisor, for his patience,forebearance, and gracious encouragement over the years. The present work and the modelowes itself to a study previously undertaken by Messrs. Haziedine and Maundu and myselffor the Food Policy Task Force of Industry, Science and Technology Canada in 1990. Theerrors here, however, are gladly my own. I would also like to thank Mary Bohman for heradvice and support. Kathy Shynkaryk, Gwynne Sykes, Alison Sawyer, and Janet Kee helpedme put it all together and kept me from losing my mind. This very little book is for Kathy,who, with such great patience and good humour over all these many years, has kept us allfrom descending into a purgatory of countless exasperations and from looking as foolish aswe would certainly otherwise appear.1Chapter OneOverviewSection 1.1 IntroductionWith the dramatic expansion of world trade over the past few decades, and given thepresent and increasing importance of multilateral trade unions, more attention is beingdrawn to the methods and underlying assumptions of the models we use to evaluate changesin trade and policy. Traditional trade-theoretic modek have developed from and beenlargely concerned with competitive trade in homogeneous final goods. However, trade inoften highly differentiated goods and intermediates forms a significant proportion of worldtrade and especially among industrial countries. Indeed, although the net trade that occursbetween countries continues to reflect differences in resource endowments, it is found thatthe major component of international trade and its growth involves countries with similartechnologies and tastes producing similar products. These products, in turn, are commonlyintermediates in the further processing and provisioning of other goods and services whichin turn may or may not also be traded.Since tariffs and other trade policy instruments affect inputs as well as final goods,international trade theory and policy analysis have turned to examining the nature of jointproduction and the presence of intermediate inputs so as to properly evaluate the effects2of protection. Attendant to this, and with an interest in formulating some internationallycomparable index of protection, there have developed the concepts of effective protectionand the ‘price’ of an activity’s value-added product. Within economic theory this hascontributed to a number of important elaborations and advancements. Since the mid-1960’s,the final good model has evolved and been generalized to account for traded and non-traded intermediate inputs, and a beginning has been made in incorporating certainstructural characteristics of imperfect markets so as to account for the presence of non-competitive pricing and variable returns to scale. Some of this work has been directed tosuch theoretical questions as whether or not we may extend the validity of the HeckschlerOhlin, Stolper-Samuelson, and Rybczynski theorems, while other work has concentrated onthe importance of joint production, intermediate inputs, and imperfect markets to theempirical dimension of value-added and its measurement.The fundamental problem which the rate of effective protection is designed toaddress is to properly and meaningfully account for the manner in which a tariff structureimpinges upon the surplus available to any productive activity by variously affecting both thenominal output of these activities as well as the inputs that enter into them. Indeed, it hasbeen hoped that such an index would perform a role analogous to that which nominal tariffsplay in predicting changes in the gross output of final goods. In particular, by developing ameasure of tariffs as they impact upon all goods and intermediates, it has been theexpectation that such an index could be used to evaluate a particular policy change bypredicting the degree to which various primary resources, especially labour and capital,would be reallocated among these activities.3Empirical problems, however, have continued to cast doubt on the validity of the rateof effective protection as an indicator of ‘resource pull’ or its ability to determine the ‘true’effective prices for the resource and value-added components of a good. Considering thedegree of intraindustry trade and the importance granted to the estimation of such indicesgenerally and within trade models, surprisingly little success has been achieved in reconcilingthe variabifity of these estimates as dependent upon the underlying assumptions. In general,trade models have been notoriously sensitive to the parameterization of scale economies andproduct differentiation, or the incorporation of trading blocs, or the adoption of alternativeformulations of pricing behaviour.Although there exist historical precedents in the literature (see Corden, 1971), theconcept and measurement of an index of effective protection does not appear to have beenof more than passing theoretical interest until the post-war era. This seems in part due tothe general adequacy in the past of partial equilibrium analysis in the determination of thevarious costs and benefits that a tariff and taxation structure imposes upon a single industrywithin a much larger economy. In these cases it is convenient to either forego or estimatethose adjustments that may be required to properly account for the shifting of resources intoor out of that particular industry or sector. Hence, what is termed the adjusted nominal rateof protection on a good is often adequate to the purpose.Effective protection springs from the recognition that tariff structure reflects morethan a concern for nominally traded output and competing imports. Within a Sraffa-likeeconomy in which commodities are visualised as final goods that may also enter asintermediates in the production of other commodities, a change occuring in a network of4policy-induced price differentials may have a unique impact upon each and every industry.Policies change, and so do those who gain and lose. Even now, in Canada, in the midst offully implementing a Canadian-U.S. trade agreement and at the edge of another agreementwhich would include Mexico, we may commonly speak of ‘free trade’ and yet, clearly, labouris not free to seek its highest return in the same way that is becoming increasingly availableto other resources or goods or services. The return to labour will reflect instead how thatresource is embodied in the provisioning of these goods and services. So it is for everyactivity and for the owner of any resource. Each industry or activity producing and using anyof a myriad of commodities and employing its own unique combination of resources willsimilarly face a potentially quite different array of relative prices for those commodities andresources. Eventually, as the changes in. costs filter through the economy, we may expecttransformations of techniques and the very commodities themselves.Indeed, it is natural that the interest in the effect of changing protective tariffstructures on resource flows should have followed in the wake of these past few decades ofexpanding global trade. Instituted differences in world and domestic prices, barriers to trade,and the rents and resource flows implicit in those policies which have historically entrenchedthem are now of even greater importance. Increasingly we see how global our economyreally is, and we hear more of trading blocs and free-trade zones. Now, as any countrycontemplates moving towards free trade and the complete removal of an entire frameworkof frontier barriers, not only the incidence of various price wedges on output and onemployed inputs, but how these changes in all the various industries’ cost structures5permeate throughout the productive economy becomes vital in evaluating the true extentof these rents and resource flows.Because trade in differentiated goods and services is such an important componentof international development and exchange, and because the markets for these commoditiesare often highly structured and imperfectly competitive, trade models have begun toincorporate some of these elements with surprising and often contradictoiy results. As weshall see, econometric evidence over the past decade strongly suggests that the traditionalconcern with competitive price formation and its analysis is not an acceptable simplificationin modelling these industries.Section 1.2 Problem and objectives-Much of the theory and research involving price formation, and particularly within trademodels involving estimates of protection and effective prices, has operated between twopolar pricing hypotheses. One, under either the Ricardian ‘law of one price’ or the EastmanStykolt hypothesis, supposes that the domestic market responds by matching the price ofcompeting imports; at the other extreme, competitive general equilibrium models requirea cost-based pricing rule and markets are driven by the perceived demand facing thesupplier. In the context of assessments of economy-wide welfare gains to trade liberalization,large differences in these estimates follow from the adoption of one price model overanother. Both extremes in formulating prices are quite appropriate and successful inmodelling classically competitive markets with industries comprising many suppliers ofhomogeneous products, however, the presence of imperfect competition in the markets for6more heterogeneous intermediates and final goods is more problematical. In these, marketstructure often reflects a confluence of various organizational factors: industry concentrationgoverns the degree to which firms may coordinate or collude in pricing so as to restrictmarket entry, and the preference for variety and product differentiation may limit the degreeto which firms and industries capture economies of scale. Transport costs and geography andadvertising expenditures may dominate strategies in these markets. Any of these and otherimperfect competition effects may contribute to the extent to which a particular commodity’smarket and price will respond to changes in the prices of its competing imports or to thethreat of domestic market entry and changes in the costs of its various inputs.Econometric evidence exists to suggest that while both of these hypothetical limitsor ‘attractors’ seem fundamental to the specffication of observed price formation, neitherby itself is sufficient to explaining or modelling actual behaviour among many manufacturingindustries. Not only do industries vary in the degree to which either costs or import pricesmay hold sway, but it may be important to be more specific and determine when and underwhat conditions these relative considerations change.The problem addressed by the present work is to examine the various implicationsand consequences to the measurement of effective protection that follow from assumingeither (1) the standard Ricardian hypothesis that all industries price to the full extent of thetariff-laden import price (the law of one price), or (2) the mixed-pricing hypothesis thatmodels industries as variously formulating price as a structurally determined response tosome weighted average of both the price of competing imports and the level of domesticcosts.7The problem is expected to be of particular relevance with respect to many industriesin the manufacturing sector within which imperfect competition is often considered to beof prevailing and general importance. The presence of mixed pricing is expected to showthat, for many of these industries, the traditional method of assuming the Ricardianhypothesis will tend to overstate the measurement of value-added and, as a consequence,also overstate the measurement of rates of effective (as well as nominal) protection.For many countries, the manufacturing sector as a whole is often favoured indevelopment and is not only nominally protected, but, more to the point, often receivespositive effective protection: the net protection that they receive for their output is higherthan that generally levied on the shares of primary resources and processing intermediatesthat constitute their purchased inputs. In Canada, however, and in many other developedcountries, this is not true of the primary and processed food industries.Many primary agricultural products, particularly those with supply controls, such asdairy and poultry, receive positive protection while at the same time comprising a majorproportion of the inputs and costs of the processing sector. These downstream industries,and by implication, the domestic resources that they embody and use with greater relativeintensity, are therefore less protected than the nominal tariffs on their output would suggestand, in some cases, these industries may even receive negative effective protection.What is of particular interest here and motivates the focus of the present studycentering on these two food sectors follows not only from this anomalous tariff structure butknowing that this situation persists to some degree under the recently implemented CanadaU.S. Trade Agreement (CUSTA). Although tariffs have been largely eliminated between8these two countries, non-tariff barriers remain on a number of commodities that areimportant as inputs into various processing industries. These industries, then, will not onlyhave lost protection on their final goods, but will continue to absorb the higher costs ofthose protected commodities that are often completely essential to their production. Thesebecome a critical issue when evaluating trade liberalization in that these higher costs are notmerely passed on to the consumer as a means of subsidizing primary agriculture: some also-hidden proportion of these costs are borne by the economy and its resources as allocativeinefficiency losses as well as by the processor through the lowered competitiveness,production, and sales of its higher-priced output.Our formal objective, then, is to examine the implications of allowing for either tariff-limit pricing (the law-of-one-price) or mixed pricing in the vertical measurement(incorporating direct and indirect intermediate use) of effective protection, and our aim isto do so within a framework that considers the resource-allocative implications of the 1988implementation of the Canada-U.S. Trade Agreement and the particular consequences thatthis change of policy has on Canadian food industries. We will examine how the adoptionof one price model over the other affects not only the measurement of the impact of theCUSTA on these individual industries, but whether or not it may affect the rankings of theirrates of return and effective protection. The protection that results under the CUSTA iscompared with an alternative scenario in which the remaining non-tariff barriers under theagreement are removed and the economy moves to free trade. The base case CUSTAevaluation is also examined under depreciation and appreciation of the home currency.9Section 1.3 OutlineIn the following chapters, we begin by presenting the traditional effective protection modelwithin a partial equilibrium framework which assumes that all suppliers price up to theworld price plus the full tariff for their respective commodity. This model basically servesto introduce the main ideas and the examine the consequences of tariff structure onresource allocation in an economy of industries producing tradeable goods each through theemployment of a value-adding activity and inputs consisting initially of other traded goodsand intermediates. In turn, the model is discussed in the context of the critical implicationsof substitution, non-traded inputs, and currency revaluation.While a growing body of literature and empirical studies over the past two decadeshave brought recognition to the importance within trade theory of increasing returns toscale, imperfect coordination among firms, and other departures from perfect competition,more recent inquiries into and success in modelling monopolistic competition with productdifferentiation have illuminated the importance of incorporating these various considerationsin modelling price formation. The remainder of second chapter examines the industrialorganization critique of the Ricardian hypothesis and the implications that market structureand mixed-pricing may have on the traditional effective protection model.Chapter three goes on to introduce a mathematical model that assumes andincorporates mixed pricing in calculating the rate of effective protection. It is formulatedwithin a simple input-output fixed-coefficient technology framework in which the excesssupply elasticities for the rest of the world as a supplier of any tradeable commodity areinfinite at the tariffed border price of the competing import. We define a particular10measurement of effective protection and decompose this as the weighted sum of threedistinct partial effects.Chapter four discusses the empirical calibration of the model which will then beapplied to initially to a base case measurement of effective protection implicit in theCanadian industrial economy moving from the pre-1988 tariff structure to that subsequentlyimplemented under the Canada-U.S. Trade Agreement (CUSTA). The economy isrepresented as consisting of 31 industries of which 29 comprise the primary agricultural andfood processing industries while the remaining two are aggregates of other manufactures andservices. The particular aggregations of all commodities and the disaggregation of theprimary agricultural sector are detailed here, as are the various tariffs and producer subsidyequivalents that enter into determining each commodity’s price. Finally, to incorporatemixed-pricing, each industry is assigned a share parameter, as specifed by econometric andother information, that expresses price as a composite formed from some weighted averageof import prices and domestic costs.The fifth chapter summarizes the results and examines the sensitivity of the modelto the two pricing hypotheses. The base case scenario that represents the CUSTA iscompared with several other scenarios variously involving free trade or the CUSTA underexchange rate changes. The differences between the two price models and between theCUSTA and free trade scenarios are also discussed with respect to the rankings of theseindustries by their effective protection rates and their rates of return to capital.11The final chapter is directed toward briefly outlining the conclusions and limitationsof the model, and to suggest where improvements could be made and where furtherresearch appears needed.12Chapter TwoEffective Protection and Industry StructureThe primary problem which seeks to be addressed within the theory of effectiveprotection is to develop an index such that, by accounting for tariff structure and theeffective prices of all the resources that enter into any activity, and through a ranking ofthese activities based upon this index it becomes possible to predict the reallocation ofdomestic resources as a consequence to changes in that tariff structure. Within such anindex, researchers have sought to encapsulate for all goods and intermediates what thenominal protection rate does for final goods by themselves: it has been the hope that therate of effective protection could provide some indication of a measured positiverelationship between an increase in the value added by a particular activity and thatactivity’s subsequent expansion of output. As we shall see, the crux of this expectation liesin what such a measure says about changes in gross outputs and, hence, changes in income.The rate of effective protection, however, concerns itself with the employment ofintermediates and changes in net output. Gross output is the variable that is of interest whenevaluating gains to trade, whereas the rate of effective protection is directed more tosummarizing tariff structure and the response of industries to trade liberalization as itimpacts on their primary resources and value-added.13While it appears that interest in the rate of effective protection needed to await thecomparatively recent growth of world trade and a correspondingly recent interest in theeconomy-wide effects of trade liberalization, resource allocation and general equilibrium,there also lay the more pragmatic problems of model formulation and dealing with theinherent and considerable information requirement. All these matters veiy much relied uponthe development of computers and the incorporation by Leontief and others of input-outputstructure and analysis into national account-keeping and international trade theory.Geography and history also have a role in this story in that we find that the smallcountry assumption in itself is vitally important to the concept of effective protection. Theproblem was either insignificant or simply did not arise in larger and more open economieslike those of the United States andthe United Kingdom. The theory’s formal and systematicelucidation therefore awaited the interest shown in this subject in the 1960’s by researchersin such countries as Canada, Australia, Israel, and Sweden. While various authors hadhitherto examined the problem of effective rates of protection (ERP’s) and similar indexmeasures, the early theoretical work and formalization of the idea was chiefly undertakenby Johnson (1965) and Corden (1966), while Balassa (1966) and Basevi (1966) produced thethe first notable large-scale empirical studies. Much of the earlier theoretical work centredaround the use of stability conditions to provide comparative static results for dimensionallysmall systems. Larger systems and the relaxation of various assumptions, however, have leftthe theory less amenable to analysis. The stability conditions provide little structure and thepossibility of substitution among inputs has left the value of ERP’s a contentious issue,14especially within the context of general equilibrium and successfully predicting actualresource flows.Anderson and Naya (1969), Tan (1970), Jones (1971), Ramaswanii and Srinivasan(1971), Bruno (1973), Khang (1973), Ray (1973), and Bhagwati and Srinivasan (1973) allprovided early contributions to clarifying the theoretical problems of the index. Many ofthese studies are formulated under general equilibrium and remain definitive (see alsoWoodland (1977) and Woodland (1982)). Numerous empirical studies continue to beproduced in part engendered by the emphasis given to the index in GATE and other tradenegotiations and the concern within these contexts with compensation and internationalcomparability of a wide range of protective trade practices. Despite its theoretical andpractical problems, the index is of descriptive value and it remains a useful means ofsummarizing the information on the protective structure resulting from various tariffs leviedon both inputs and outputs.Section 2.1 Partial Equilibrium Analysis of Effective ProtectionTo orientate ourselves and to introduce the various terms and definitions, we present thecomponents of the rate of effective protection within a partial equilibrium framework.Essential to all of this is the allowance made here for purchased intermediates into theoutput of some commodity. These inputs in and of themselves may be also produced andtraded, and, motivating this discussion, may be variously subject to tariffs and taxes. Theproblem traditionally consists of determining the extent to which those tariffs placed uponany commodity’s trade will affect the effective cost and consequent allocation of the inputs15that enter into the production of that commodity, and, further, how the production of thatcommodity is in turn affected by the tariff and taxation structure imposed upon the marketsfor its inputs.The traded final good is considered here to be produced by an industrial activitywhich combines, to begin with and for simplicity, a single purchased and traded input andthe value-added of that activity. This value-added product is seen as the employment ofsome technique embodying some bundle of labour, capital, and other primary resources. Thestandard assumptions are to include that the value-share of the physical input coefficientsare fixed and remain identical throughout the industry producing that commodity andregardless of any ensuing change to the tariff structure. We are therefore ignoring thepossibility of any substitution occuring between traded inputs and primary factors and non-traded inputs. We also assume that the excess supply elasticities for the rest of the world asa source of any competing importable input or commodity output are infinite. The countryis small in all markets. Finally, we will make the Ricardian assumption here that alldomestic production is priced up to the c.i.f. (transportation-inclusive) world price plus thetariff: this is the law of one price.If units are chosen so as to allow for a unit of input to be the physical requirementin the production of a unit of output, we can map their demand and supply in the samespace as in Figure 2.1. Here, the foreign supply curves s and s° for the input and the value-added product are shown as being horizontal at their c.i.f. prices, p’ and p° respectively,while the domestic supply shedules for each are pictured as s’ and s°. At these prices, thevalue-added product per unit of output is (p°-p1) and Corden (1971) terms this as the16Figure2l Partial-analysis of effecti protectioneffective price of that product while p° is its nominal price.With a nominal tariff t° imposed upon the output alone, the effective price becomes(p° + t0pi) and the rate of effective protection is here t°/(p° + t0pi), or the proportionalincrease in the effective price due to a nominal tariff. With a nominal tariff imposed uponthe input alone, the effective price for the output becomes (po..pi...ti) and the ERP is nowcalculated as -t’/(p°-p’-t’). With both tariffs in place, the effective price of the value-addedbecomes (p0+ t0pit1), and the sign of the effective rate of protection, now measured as (t°ti)/(p0+topiti), is seen to depend upon the relative magnitude of the two per-unit tariffrates.Representing the domestic demand schedule for output as d°, the tariff-free ci!.world price p° results in the total quantity demanded of qd0 Under world prices and withpricc(job t)(p’+t)‘‘IS.q’0 tI’i q°1 ClO, q°3 4° ( q4 Ullits17free trade in the input market, domestic supply of this input is q’0, with a kink at A at whichthe remaining demand is being supplied by imports at price p’. As a result, a kink alsoappears at B in the schedule of domestic supply of the value-added product, and this isrepresented by the curve s°0 at quantities exceeding q’0. At given world prices, domesticproduction and supply of the output commodity is hence q°0 with imports measured as (qd00q.As will be shown, the supply of the value-added product is dependent upon theincidence of the tariffs on both the input and the commodity output. If a tariff of t1 is placedupon the input alone, the supplies of both the input and the value-added becomes kinkedat C and D respectively, resulting in a squeeze on domestic supply now reduced to (qd1q0)•If, on the other hand, a tariff t° is levied on the value-added of the output alone, domesticsupply increases to q°2. With tariffs on both the input and value-added, domestic supplyequilibrates to quantity q°3. In this case, since the nominal tariff exceeds that placed on theinput, the effective rate of protection is positive, and the tariff structure as a whole resultsin an overall increase in domestic supply of the output from (p°;q°0), but it is easy to seethat, with a relatively higher per-unit levy on the input (and hence a negative ERP), outputmay fall as well.In algebraic terms, and for the present stifi assuming one input, we can derive aformula for the effective rate of protection for the 1th industry or activity producingcommodity j and utilizing purchased input i. Then the value-added product per unit of j inindustry j in the absence of a tariff (the base case here), can be written as:Vj=p (1—a)18where p1 is the nominal price of unit off, and is the share of commodity i in the cost ofcommodity f in the base case.Defining t1 and t1 as the nominal tariffs placed respectively on commodity] as a finalgood and on commodity i as an intermediate good, then the value-added product per unitoff in industry j under the new policy imposing the tariff is measured asV1j =p[(1÷t)—a1(1+t)]Given these, g1, the rate of effective protection on the final good produced by industry j,measures the variation in value-added as a proportion of the base case level:= (v’—v)/v = (ti—a1)/(1—aIn the n-good case, this can be represented asg = (t—a1t) / (1—E a1)Since most countries move from one protective structure to another, it should benoted here that input-output tables are generally constructed with tariff-inclusiveobservations and so the elements of the matrix typically consist of a rather than a.!!, wherea1 = a11 [(1+ t1) / (1+ ti))Section 2.2 Methodological ConsiderationsThree critical assumptions have been made to facilitate the formulation and analysis ofeffective protection rates: (1) fixed input values as technical coefficients, (2) infiniteelasticities of supply for all importables, and (3) perfect substitutability between domestic19output and imports. This section briefly examines the consequences of relaxing theseassumptions and other related issues that might impart any empirical bias. In particular, thecalculation of ERPs is complicated and modified by several methodological problemsconcerning the treatment of substitution between inputs, the existence of non-tradeablegoods and the effects of changes in the rate of currency exchange.22.1 SubstitutionCorden (1966) suggested that the fixed coefficient model introduces a positive bias tocalculated effective protection rates, and a number of authors have exanilned this problemto evaluate the degree and implications of this empirical overstatement. Essentially, theqiestion turns upon the degree of substitution or the curvature of the iso-surface of inputs:in addition to the benefit of the instituted price ratios under protection, each activity isfurther free to maximize with respect to input use. A rise in the relative cost of any inputwill therefore result in a using industry to employ less of that input, if this is possible.The use of post-protection (or ‘post-substitution’) coefficients therefore leads to anover-estimation of the ERP since no account is made of the gains due to substitution.Similarly, the use of free-trade pre-substitution coefficients would result in an underestimation of the rate. Once substitution is allowed, two definitions of the ERP are possible(see Bhagwati and Srinivasan (1973), and Woodland (1982)). Corden (1971) identifies thisas an index problem and, as shown by Anderson and Naya (1969), it is the former definitionwhich is prone to bias whereas the latter involves an insurmountable data requirement: formost economies, the pre-tariff structure is simply not historically observed and not20empirically derivable. In dealing with this problem, Balassa (1965) notably employed thosecoefficients of small open economies like Belgium and The Netherlands to represent orweight the free trade technology of other industrial countries. Non-biased rates of effectiveprotection could be then calculated through weighted averages of borrowed free-trade andown-country post-tariff coefficients. In his (1971) judgement, intercountry productionfunctions and engineering coefficients tend to be reasonably uniform and comparable, andempirical studies gave evidence of very little substitutibility between primary factors andintermediates.Bhagwati and Srinivasan (1973) however argue that many countries import capitalgoods that substitute for domestic primary resources (notably, labour). Ramaswami andSrinivasan (1971) show by example that substitution effects will often be significant and that,in general, it is not possible to devise a measure of effective protection that will unerringlypredict the impact of tariff structure on gross outputs. They suggest that it is theendowments of primary factors and their substitution for other inputs that will likelygenerate the most severe empirical problems and ERPs are therefore often poor indicatorsof actual resource movements. Leith (1968) also gives an example in which the assumptionsmade concerning substitution and the presence of non-traded inputs significantly alter themagnitudes and rankings of ERPs. Corden (1971), in reply to Ramaswami and Srinivasan(1971), notes that the bias depends primarily upon a large degree of substitution and thatthe tariffs on inputs be also sufficiently high or subject to large changes. Nevertheless,Ramaswami and Srinivasan (1971) essentially laid to rest any hope of a ‘true’ measure ofeffective protection.21Jones (1971) and Khang (1973) examine the problem more rigorously and determinethe specific conditions under which perverse changes in output may be possible. In their 2-industry/3-factor models, a normal response is assured if the cross partial derivatives for anypair of factors in the production function are non-negative (the normal non-inferiorityproperty in which all inputs are substitutes). It is this restriction which, when relaxed, resultsin the possibility of resources flowing in the wrong direction in Tan (1970) and Ramaswamiand Srinivasan (1971). Jones and Khang essentially support Corden’s view in emphasizingthe importance of the composition of tariff changes: in the final analysis, perverse outcomesdepend not only upon disproportionately large changes in the prices of intermediates, butalso appropriate bias in substitution and sufficiently large differences in factor proportionsbetween industries. Bhagwati and Srinivasan (1973) go on to show that the RamaswamiSrinivasan analysis and those by Jones and Khang diverge depending upon whether or notany intermediates enter symmetrically into the production of the final goods.Bhagwati and Srinivasan (1973) is an attempt to examine and salvage what is usefulin the theory and they propose to restrict the problem by assuming that final goods do notenter into the production of other goods. As a result, in their general equilibrium modelincorporating many tradeable goods, substitutable primary factors and importedintermediates, gross outputs always equal net outputs.Bruno’s (1973) examination of the problem within an N-industry general equilibriummodel confirms the crucial nature of the normality property among the inputs and moregenerally emphasizes the requirement of functional separabifity of primary and tradeableinputs in gross output production functions. Given this and providing that no substitution22occurs between the groups of non-traded and tradeable goods and inputs, ERPs shouldcorrectly predict resource flows.Ohyama and Suzuki (1980) show that a relative increase in the price of a tradeablegood with respect to the prices of all other tradeables and importable intermediates willgenerate an increase in the net output of that good. Addressing the problem posed byRamaswami and Srinivasan, they provide the most general proposition that, if non-tradedintermediates are not substitutable, and if neither tradeable nor imported intermediates aresubstitutable for primary factors, then a relative increase in the modified Corden index ofeffective protection for a tradeable good will correctly reflect an increase in the gross outputof that good and a decrease in the gross output of some other tradeable good.2.2.2 Non-tradeable intermediates and exchange ratesThe basic issue with the presence of non-tradeables concerns their use as inputs into othergoods of which some may be tradeable, and how these complications affect their role in thetransmission of prices through the domestic economy. Within the theory of effectiveprotection and as a practical matter, the question arises as to whether or not services andsome regulated commodities should be included as primary factors or as tradeable goodswith various tariff structures assumed. Two approaches have been commonly considered.The Corden (1966 and 1971) method essentially combines the non-traded commodities withthe primary factors by assuming that any such intermediate may be decomposed entirely intodomestic added value plus any traded inputs. It is this decomposition which is a non-trivialundertaking. The Balassa (1965) method, in contrast, treats non-tradeables as though they23were traded goods but with a zero tariff applied to the final good. The Scott method (seeRay (1983) and Woodland (1982)) is similar to that of Balassa but the prices for nontradeables are allowed to change.The Balassa and Scott measures have the advantage of simplicity in calculation, butas noted by Ray (1983), the effect of the aggregation is to compromise the correct rankingof activities and the significance of the resource allocation effects. As protection is removed,tradeable goods with positive ERPs bear the total of any reduction in value-added while theprices of non-tradeables are assumed to be sticky downwards. The Corden method, however,does correctly rank its activities and has limited resource allocation significance providedthat all non-traded inputs are strictly intermediates and do not appear as final goods inconsumption.--The essence of the problem of exchange rates follows from the above and lies in thepresence of both non-tradeable goods and primary factors and their role in the transmissionof exogenous world prices throughout the domestic economy. This occurs through therelative use and degree of substitution for factors and tradeable goods and the impact thismay have on internal relative prices. Beginning with internal and external balance (fullemployment and equilibrium in payments), and supposing that the tariff structure impartsgenerally positive ERPs to a nation’s industries, the removal of that structure is likely toresult in an overall external deficit necessitating a devaluation of the currency to restorebalance in payments. In addition to this, domestic consumption and expenditure must shiftfrom both exportables and non-tradeables to that of importable goods to effect internalbalance. Resources, on the other hand, will shift in the other direction. Whether or not the24relative prices of non-tradeable goods rise or fall depends upon their substitutability inproduction and consumption for imports and exports.The problem becomes even more complex as one considers the effect of eitherbilateral or multilateral trade agreements and the exchange adjustments made by allparticipants. To incorporate all these effects so as to generate a kind of comprehensive andcombined index of protection is likely to present an insurmountable empirical difficulty.Even for a single country, the general equilibrium adjustments likely to ensue from tradeliberalization would be difficult enough to sign let alone predict with any accuracy. We maysay, however, that, although perverse outcomes are possible, the simultaneous removal oftariffs among trading partners is likely to depend upon relative size and to involve some- degree of mutual cancellation of exchange effects. In addition, there -will be a centraltendency for the relatively more protected country to undertake a devaluation as its marketsbecome subject to relatively more foreign penetration.Section 2.3 Imperfect competition and mixed pricingThus far, the competitive nature of the model is reflected in the underpinning assumptionthat domestic and world prices for all freely traded commodities differ by precisely thetariffs levied upon them. This is an assurance that comes from presuming market arbitrageand that there exist no barriers to the entry of other traders who will take the existence ofany price differential as a profit opportunity, buying in the cheaper market and selling whereit is dearer. Classically, this trading activity persists until profits existing between these twomarkets are driven to zero. As a practical and empirical matter however, there are often25cases, in which the structure of the market, through collusion, say, or economies of scale,or product differentiation, or some natural barrier, creates or allows for impediments to theentry of such traders, and so a difference in price may persist between two markets. Thepresence of imperfect competition and market power allows a monopolist or a collusiveindustry to gain this difference as pure rents while somehow excluding entry at some priceabove cost.The traditional theory of international trade rests upon the Ricardian and theHeckscher-Ohlin-Samuelson models, both of which assume perfectly competitive behaviourand the law of one price. In the Ricardian model, trade occurs on the basis of relative pricesreflecting technological differences, while the other identifies flows of resources as being dueto relative differences in factor endowments between countries. While these models havebeen and will continue to be eminently useful, they nevertheless remain inadequate inexplaining a number of interesting empirical observations. Leontieffs Paradox, for example,concerning itself with the apparent contradiction of the H-O-S model in the composition oflabour and capital in U.S. imports and exports, has engendered a host of similar studies.Many confirm this type of anomaly for other industrial countries and in other years.Although net exports and net imports continue to reflect relative endowments, a significantand increasing volume of trade occurs between countries with similar endowments andtechnologies. In particular, there exists substantial two-way intraindustry trade incommodities of similar factor intensity. There are also striking welfare anomalies. It hasbeen commonly observed that the experience of the EEC, for one example, has been oneof increased trade and income for all participants but with relatively minor reallocations of26resources between sectors. The Auto-Pact agreement between the U.S. and Canada servesas another instance of this.One of the central interests of Canadian trade modelling has concerned itself withthe trade-liberalized gains of increased scale that are debated to exist in that country’smanufacturing sector. Empirical evidence of its degree and and incidence, however, isscanty, and increasing returns and variations among industries in its degree are difficult tomeaningfully incorporate into these often linear models. Between such broadly contiguoustrading entities as Canada and the U.S., it may be that the very existence of their structuraland demand similarities makes imperfectly competitive pricing behaviour, intraindustrytrade, and national and firm levels of product diffentiation more important considerations.2.3.1 Imperfect competition and product differentiationThe conventional theory of comparative advantage concerns itself with relative endowmentsand broad classifications of industries with activities of differing ratios of factor utilization.Indeed, factors and their returns are the main focus. Since the late 1970’s however, andfollowing Spence (1976) and Dixit and Stiglitz (1977), there have developed a number ofimperfect competition models that develop the contra-Chamberlinian idea that economiescharacterized by monopolistic competition may feature too little diversity and propose thatthere exist utility gains to variety in consumption.These theoretical models (see, for example, Krugman (1979, 1980, and 1981),Brander (1980), Lancaster (1980), and Brander and Krugman (1981)) have produced notableresults and have primarily been directed toward explaining intraindustry trade. Krugman’s27(1981) two-factor model shows how both factors can gain through trade even if bothcountries have similar endowments. In particular, the more homogeneous the products, themore similar must the two countries be to allow both factors to gain. Country size alsomatters, as in Krugman (1980). Allowing two countries with symmetric demand andtechnologies to trade but allowing their relative size to vary results in each country beingan exporter of that good for which it has the greater relative domestic demand. Wages willgenerally not be equal and will tend to be higher in the country with the larger domesticmarket. Two-good models such as Krugman (1980) and Helpman (1981) in which one goodis produced under increasing returns to scale, show that, all other things being equal, thelarger country has a lower relative price and an advantage in the production and export ofthat good which is prothced under economies of scale.The existence of economies of scale in production and the utility of variety throughproduct differentiation can give rise to trading partners specializing so as to capture thebenefits of larger scale and lower unit costs. This will occur despite similar endowments andtechnologies, and this similarity in demand between the two countries will result in thisexchange being intraindustrial. What prevents each country from producing an entire rangeof products is the presence of fixed costs and it is the existence of some degree ofunexhausted economies of scale that leads to specialization and intraindustry trade indifferentiated products.Market segmentation is also seen to be an important contributing factor in explainingintraindustry trade effects and the existence of price discrimination between markets.Brander (1981) suggests that, even with undifferentiated products, cross-hauling of identical28commodities may take place within a Cournot setting. Here, trade arises through amotivation to dump or price discriminate between segmented markets: each firm perceivesits domestic and foreign markets as being separate and each formulates their outputdecisions accordingly. Although, as in Brander and Krugman (1983), reciprocal dumping ofsimilar products generates pure waste in the form of avoidable transport costs, inequffibrium, each country’s firms perceive their marginal revenue to be higher in theirexport markets and are able to overcome the relatively higher marginal cost involved inshipping abroad. Each firm operates with a smaller markup above cost in its export marketrelative to that in its domestic market.Another form of product differentiation model allows for monopolistic competitionand assumes that inereasing returns are to some degree internal and endogenous.t the firm. - -In these, importance is directly attached to the numbers of firms, their perceived elasticitiesof demand, and the extent to which market power is exploited as evidenced in their pricingbehaviour. While it remains probable that some preference for a nation’s own good is oftena significant factor in demand, firm-level product differentiation allows for distinctions inmarket power on the basis of the structural characteristics of both domestic and foreignindustries. This type of model depends more on the potential for economies of scale at thefirm level whereas the more traditional analysis of increasing returns occuring at theindustrial or national level fails to account for the role that diversity has in constraining therealization of efficient scale.Product differentiation is invoked, then, to provide for the existence of intraindustrytrade and monopolistic competition. The presence of imperfect competition allows for varing29degrees of market power and coordination among the firms of an industry in which thereis scope for scale economies as a consequence of structural fixed costs. In turn, the structureof any industry and the degree of market power exercised by its firms is reflected in theirvarying capacity to price above marginal costs to an extent bounded by the full level of tariffprotection. Under product differentiation, it is the threat of entry that prevents an industryfrom exploiting the full value of the tariff that is assumed under the standard hypothesis ofmonopolistic or focal-point pricing.2.3.2 Imperfect competition and Canadian trade modellingIf the structure of any industry creates a market which is not perfectly competitive, and ifregimes of either domestic or foreign tariffs lend themselves to the formation of smaller andmore isolated markets, it has been argued that the presence of imperfect competition andthe scope for capturing internal and external scale economies may greatly alter theevaluation of gains to trade liberalization. This is of particular concern to small openeconomies and, in Canada, it has been suggested that the failure to achieve these allegedscale economies is evidenced in the excess capacity problems faced by our manufacturingsector. Empirically, this has been confirmed in observations that the minimum-size efficientfirm in many of these industries is apparently large relative to the domestic market for theirproduction (see Fuss and Gupta (1981), Baldwin and Gorecki (1986), and Robidoux andLester (1988)).The argument that follows from this is that the domestic tariff prohibits or impedesforeign competition and thereby fosters the proliferation of protected higher-cost firms that30remain orientated to the smaller market. Similarly, the incidence of foreign tariffs also actsto prevent domestic firms from gaining access to the larger world markets. Eastman andStykolt (1967) are among those who initiated the debate and argued further that theincidence of tariff protection in itself may tend to facilitate the coordination of oligopolisticactivities.The Eastman-Stykolt (E-S) hypothesis can be characterized by an industry of collusiveoligopolists who perceive the demand for their output as being inelastic to some degree thatallows them to optimize with respect to declining marginal revenues, restrict quantities andthereby gain monopolistic rents. Foreign supply, on the other hand, is assumed to beperfectly elastic at the world price pius any domestic tariff. Domestic entry is also elastic atsome price that equals costs and so the entrenched and presumably collusive oligopoly isostensibly free to price its output between these two extremes at which entry may take place.Pricing just below the tariffed world price may just exclude foreign imports and forms anupper bound to this formed price, while a constant returns industry must price at cost to barnew domestic entrants. Under the E-S scenario, firms in these industries dissipate theserents by crowding the market and operating on average cost curves that are higher thanwould be observed if these firms were operating on an efficient scale of production.Eastman and Stykolt (1967), Wonnacott and Wonnacott (1967), and others haveargued that bilateral trade liberalization would encourage some domestic firms to competein foreign markets as well as rationalize their operations through longer runs and fewerplants, by abandoning certain product lines and specializing horizontally. The unusually largegains to Canadian-U.S. trade liberalization predicted by Harris (1984) and Cox and Harris31(1985) derive from the degree to which they see the manufacturing sector being able tocapture these scale economies. Critics of this perspective, however, would argue that if thiswere true then these firms would already be in a position to rationalize their production andinstitute more competitive pricing so as to capture larger shares of both the domestic andforeign markets.Cox and Harris, in their 1985 study of 29 Canadian industries for a benchmark year(1976), allowed for scale economies (derived from estimates for the late 1960s undertakenby Fuss and Gupta (1979)) and ad hoc across-the-board mixed pricing in twenty of these.Under the unilateral free trade scenario, they determined a welfare gain of some 4 per centwith a 10 per cent rise in the real wage, mostly attributable to rationalization of themanufacturing sector and expanding intraindustry trade. The study predicted a 20 per centgain in labour productivity with a 20 per cent decline in average fixed costs as 16 of theirmanufacturing industries expanded their output by an average of some 40 per cent. Theirmultilateral free trade estimates were even more hopeful with a welfare gain of some 8.6per cent and a 25 per cent rise in the domestic wage. Overall, manufacturing appeared togenerate the greatest benefit and employment, while industries with a high labourcomponent were the biggest losers. In testing their model, it was seen that varying thedegree of mixed pricing resulted in a very dramatic range of welfare gains. As the prices ofmanufacturing goods were modelled to exceed costs of production by 20 to 80 per cent ofthe tariff, gains increased from 2.0 to 7.8 per cent of GNP under unilateral free trade and4.0 to 16.3 per cent under multilateral free trade.32Other general equilibrium models of Canadian trade have predicted a rather widerange of generally smaller gains to unilateral and multilateral trade liberalization. While Coxand Harris do not explicitly model trade with the U.S. and the rest of the world, Hamiltonand Whalley (1985), Wigle (1988), and Brown and Stem (1988) have variously incorporatedregions of dominant trading entities so as to investigate multilateral trade issues. Thesemodels seem to suggest that the global context is an important factor in mitigating any largegains from scale economies and rationalization.Hamilton and Whalley (1985) reports results from an eight-bloc trade model in whichconstant returns are assumed for six industry aggregates. In this model and under a broadspectrum of possible scenarios, the traditional effects of trade creation and diversion appearto bc of minor siificance examining the consequees of complete trade liberalizationbetween Canada and the U.S., they estimate welfare gains to each partner to be on theorder of 0.6 and -0.04 per cent of GNP, respectively.Wigle (1988) presents a model similar to that of Hamilton-Whalley but incorporatingsome of the features found in the Harris-Cox model with a view to reconciling the broaddifferences between their two results. Here, the Hamilton-Whalley eight-bloc world market• is retained, but the six aggregated industries are divided into those assumed to possess eitherconstant or increasing returns to scale, and pricing is allowed to be of either the EastmanStykolt form (where there are fewer firms and collusion is more likely, as in motor vehiclemanufacturing) or else monopolistically competitive. This study estimates a Canadianwelfare loss of -0.1 per cent of GNP, while the U.S. gains 0.1 per cent.33The component of the Harris-Cox model instrumental in analyzing rationalizationgains are based upon the minimum efficient scale estimates reported by Fuss and Gupta(1981) in their examination of data from the late 1960s. Robidoux and Lester (1988) presenta more contemporary analysis and conclude that a majority (68 per cent) of manufacturingindustries in Canada feature some degree of increasing returns to scale but that a sizeablenumber (25 per cent) exhibit constant returns. They estimate that the cost penalty associatedwith sub-optimal scale is on the order of 3.7 per cent of the total costs of the manufacturingsector or some 2.7 per cent of GDP (in 1979 at factor cost). This placed the likely extentof sub-optimal scale to be at a lower level than that assumed by Cox and Harris (1985).Either the gains from realizing efficient scale are less than hitherto supposed or they havedeclined over the interim as scale economies were at least partially exhausted thoughproduct differentiation and market development or technical innovation.Brown and Stern (1988) undertakes a revision of the Harris-Cox model to reflect1988 tariff data and incorporates the Robidoux-Lester minimum efficient scale estimates.This study determines much smaller welfare gains to both rationalization and specialization,and, by incorporating the Hamilton-Whalley trading blocs, they estimate quite modest gainsto liberalization. Here, Canada gains 1.1 per cent of 1976 GNP while U.S. welfare rises byless than 0.1 per cent.2.3.3 Mixed pricingStudies by Haziedine (1980 and 1988) and Karikari (1988), examining Canadian and UnitedStates pricing data for the manufacturing sector, suggest that domestic pricing can appear34to be more strongly a function of both domestic costs and competing import prices. Theirresults lend support to the idea that the standard Ricardian price assumption is notappropriate for many industries. Not to allow for mixed pricing in these industries wouldtherefore mis-specify a trade model by overestimating all the direct and indirect changes inprices and value-added as well as any dependent implications this might have in thedetermining of changes in output or demand.In Hazledine (1980), an econometric model is proposed to determine the pricingresponse of various manufacturing industries to tariff protection as explained by costs andcertain structural characteristics (such as seller concentration and advertizing expenditures).An empirical test of the model on a data set of Canadian and U.S. prices in 33 industriesprovides evidence that pricing may occur between the two extremes of domestic costs. andthe fully tariff-burdened price of competing imports. As expected, monopolistic pricingbehaviour and the ability to capture any rents afforded by protection depends a great dealon concentration and the scope within any industry for its firms to collude in determininga price or some other strategy that will also bar entry.Hazledine (1985) addresses an econometric problem through the modelling of certainmanufacturing markets in a dominant-cartel-plus-fringe framework in which the domesticfirms are characterized as quantity-restricting oligopolies and any imports are supplied bythe fringe. Instead of adopting a standard pricing model based on costs, the weightings foreach industry were calculated as variables dependent upon differences in market structure.In particular, it was determined that high seller concentration was required for theseindustries to adequately take advantage of any tariff protection afforded them and that35relative costs were of greater importance to the less concentrated and more competitiveindustries.Karikari (1988), in comparing observed differences in Canadian competitivenessbetween the first half of the 1970’s and that of the second half, found pricing to significantlyvary over the two periods as combinations of the two attractors of domestic costs and importprices. During the late 1970’s, for example, when lower unit labour costs and home currencydepreciation increased apparent competitiveness in the manufacturing sector, domestic firmsfaced little penetration of their markets by competing imports and behaved more nearly asthough they were monopolists. Accordingly, their pricing behaviour was better explained asresponding to changes in their variable costs of production and upon their perceived demand-- and the elasticity of import supply. In the earlier period, however, a higher-valued dollar andtighter export markets led to price formation as being more strongly a function of importprices and tariff levels. What is more particularly interesting here is that even in this period,the results do not support the extreme law-of-one-price assumption that underlies either thetraditional partial equilibrium or Eastman-Stykolt monopoly models. Manufacturingindustries are seen then as generally not benefitting to the full extent of the tariff, and, byimplication, to proceed uncritically under such an assumption would serve to mis-specify asmuch as half the tradeable output of many small open economies.Harris (1984) and Cox and Harris (1985), in specifying their mixed-pricing modelsassigned a range of weights under various scenarios to entire blocks of the manufacturingsector to determine the sensitivity of their general equilibrium formulations to otherassumptions concerning price formation. Imperfect competition and barriers to entry had36been incorporated into these models by allowing for domestic prices to be some structuredcomposite of production costs and the tariff-inclusive border price of competing imports, andtheir results were seen to be very sensitive to the degree to which prices reflected these twoextremes. It is evident, however, that the Harris-Cox method of arbitrarily assigning weightsto the manufacturing sector as a whole can be quite feasibly improved upon if these weightscan be directly related to the empirical measurement of certain structural variables toprovide econometric industry-specific estimates of their values.37Chapter ThreeAn Input-Output Model of Vertical Protection under Mixed PricingThe present chapter introduces an intermediate-good model in which a Corden-typeindex of effective protection may be measured not only directly, as cost-share-weighted sumsof the changes in nominal tariffs on all commodities in their capacity as both final goods andinputs, but simultaneously or vertically incorporating the indirect economy-wide incidenceof new effective price& We depart from the usual Ricardian assumption which forces theprice of domestic output to equal the frontier price of competing imports. Mixed pricingallows for imperfect competition in the domestic markets for these goods and leads to theinnovation here of specifying pricing behavior to reflect any empirical differences that mayexist between industries in the relative importance of their competing import prices anddomestic costs. There then follows a discussion of how to calibrate and solve the model soas to simulate various changes in trade policy.Section 3.1 The ModelWe begin by considering a small open economy in which we can distinguish the consumptionand production of N types of commodities and an endowment of M non-producible primaryresources or factors. We picture these N industries as complexes of production activities38each of which uniquely produce units of its respective commodity through the employmentof various commodities and primary resources as inputs. Just as steel may enter into theproduction of more steel (in the next production period), it is natural to allow any of theseindustries to also employ its own commodity as an input (as it is made available from theprevious cycle of production).The technology is fixed and described by N production functions each characterizedby a vector of net outputs of commodities y= and a vector of primary resourceinputs xI=(xIl,...,xM) for each industry i. We require that each industry also employ capitaleither directly as its M + 1th primary resource input, or indirectly as embedded within itsother commodity inputs. As a. mathematical formality, we choose N such that there is nojoint production and only one technique is optimal or available in the production of anycommodity.The domestic economy is in equilibrium as consumption and production effectivelycompete for these goods. All commodities and resource factors are freely traded and theirpositive prices may be conformably mapped as vectors p= (P1’•••’PN) and w= (wl,...,wM). Theprice of capital is its rate of return, r = (rl,...,rN), which may be allowed to differ between itsalternative uses among these industries (through inherent differences in market power andrisk).With our interest in deriving the rate of effective protection as an index ofproportional differences, our confining these industries to constant returns to scale allowsus to normalize and define their activities on a per unit basis. We may represent anyparticular commodity’s unit factor cost as a weighted sum of the unit prices of all those39various resources and commodities that enter into the production of one dollar’s worth ofthat commodity. These weights correspond to the technical coefficients of the LeontiefSraffa input-output model, and we could derive these from the national account matrix ofthe economy that shows the values of the various commodities and resources that flow intothe production of any good. We can construct an NxN matrix A with elements a eachrepresenting the share of the th domestic commodity input used in the production of thecommodity output.Once we introduce trade, imported commodities also may enter into the manufactureof domestic goods, and an NxN matrix B, similar and conformable to A, can be formed withelements each representing the value share of the imported commodity employed inthern domestic production of commodity L The tariff-inclusivedomestic market prices of thesecompeting imports can be represented by the N-dimensional vector q=(q1,...,ci). Eachelement of this vector is derived as:q1=ept (3.1)where pS= (p1,•,pN) is the vector of world market prices of each commodity i andt = (tl,...,tN) conformably measures one plus the domestic ad valorem tariff rate levied oneach good. The exchange rate, e, is the domestic price of foreign currency, a scalar whichwe may vary parametrically in this partial eqffibrium model.Knowing the value of the primary resources used in each industry, we can also forma normalized NxM matrix Z consisting of elements Zik each identifying the value share ofthe ktI resource used in the production of a unit of commodity i.40Having examined in Section 2.3 econometric evidence supporting the hypothesis ofmixed pricing, and having allowed for differences in market power across our variousindustries, we now elaborate on the allowance for this power to be reflected in the pricingof domestic output. First we suppose that, for each industiy, domestic price formation canbe characterized by the relative importance of factor costs and the tariff-laden frontier priceof the competing import. Defining vector d = (dl,...,dN) as a mark-up parameter, we write,for each industry I,+ (l—s)d1c, where 0s1, d1 (3.2)in which a share parameter s = (sl,...,sN) captures in a simple way the degree to which eachindustry responds within these two bounds. In the case of s= 1, we have the case of anindustry with no market power producing a perfectly homogeneous commodity. Arbitragewill assure that the domestic market price will match that of the tariff-laden competingimport. Alternatively, at the polar extreme with any s1 = 0, price is determined solely by theinternal structure of that industry and is set as a mark-up over all non-capital costs.The technology of this economy may now be represented as N unit variable factorcost equations c = (cl,...,cN) =Apd÷ B’q+ Z’w in 2N+M variables, and for any industry I, thefactor cost of a unit of its output appears as:1=f apjni + ID bq + D ZWk (3.3)41We now turn from the domestic pricing of output and note that, with trade, manycommodities may be diverted to the export market where industry structure as well asreceived prices (and, hence, income) may vary from the domestic market. We can representthis vector of export prices in a form similar to equation (3.2). Defining and econometricallyspecifying appropriate export price mark-up and share parameters dX and s the supply priceof industry i’s exports can be written as:p=s’q’ + (l—sr)d’c1, where Osxl, dxl (3.4)where q’ for any commodity i is measured as:x * *q1 =ep/t1similar to equation (3.1) but this time deflated by t’=(tl,...,tN), one plus the foreign advalorem tariff level. Again, if s1x= 1, that particular activity can be thought of as oneproducing a homogeneous commodity. This kind of industry, with no market power, mustconsequently meet the foreign market price as well as fully absorb the impact of any foreigntariff levied against its exports. If, at the other extreme,s1=O, then the export price is beingdetermined by the structure of that industry and set purely as a mark-up over factor costs.With any industry’s output divided between domestic and export markets, its unitprice is therefore a weighted average as in:p1=up + (l—u1)p, where Ou1 (3.6)where u = (uI,...,uN) is the base case equffibrium share of that industry’s output sold in thedomestic market.42As a practical matter, we can now model any of several trade policy scenarios byaccordingly re-specifying tariffs, exchange rates and other parameters so as to solve for anew level of commodity prices p’= (p’l,...,p’N). Within the model, these new output priceswould incorporate the impact of any particular change in tariff structure as a complex ofperturbations of each commodity’s factor costs as well as the degree to which each industrywill exercise its relative power in setting its own price.For each industry, then, and similar to (3.3), these new levels of domestic and frontier priceslead to factor costs being now calculated as:c’1= ap’” + + ZW (3.7)Given this already fixed technical relationship, we go on to determine the new levels ofdomestic and export market prices. As in (3.2) and using the same share parameters, thenew price received in the domestic market for each industry’s final good is formulated soas to allow for imperfect pricing:p’=sj + (1—s)dc’, where Osl, d1 (3.8).and, substituting equations 3.7 and inverting the technical matrix, these can besimultaneously solved asP’l= [l-sd1Eaj’[sp’1 + (1—s)d(E b, + E zwk)]43Those post-policy prices that are set in the market for exports, and as in equations (3.4), arenow calculated as+ (l—s)d’c1, where OsE1, dl (3.9).and these can also be simultaneously solved as+(1_sXj)dXj(bq+ EZIkWk)]As in (3.3), the market share-weighted unit prices of output, now incorporating all of theeconomy-wide cost effects of the policy changes, are then determined asp”=up” + (l—up, where 0uL1 (3.10).Section 3.2 Trade Policy Simulation: Solving the ModelCalibration of the model begins with the normalization of all prices in the base case bychoosing units of output such that:p=p*=pd=px=w=l (3.11).In general, this means that the output in these different markets will not be measured in thesame units, but, in any case, we are only concerned here with proportional changes. As anaccounting identity for each industry, output price is also measured as44+ (3.12)which splits price into two components: total factor and non-profit costs, c, and the residualrate of return to the capital embodied in each unit, r1. We can measure r1 as the ratio of thesurplus to the value of the output, and, given (3.11), base case cost levels are calibrated asc1=1—r (3.13).Choosing the exchange rate, e, and knowing prices p and p, the vectors of initial domesticand foreign tariffs t and t8 allow us to solve equations (3.1) and (3.5) for the base-casefrontier prices q and q’.---Econometrically specifying the mixed-pricing share parameters s and sX, equations(3.2) and (3.4) can be rearranged so as to solve for the domestic and export mark-upparameters d and dx. For each industry, these ared1=(1—sq)/((l s)c (3.14)andd.x= (1_sXq)/((1—sj)c1) (3.15).In calculating vertical protection, we begin by defining value-added product per unitof ouput as the surplus of its price over its material costs: those commodity inputs that are45not primary resources. Specifically, these pre- and post-policy vectors of material costs, cmand em’, are calculated directly ascm=f ap1’+ E bq, (3.16)andc’—_+(3.17).In the base case, and for each commodity, this value-added is written as= p1 — c’ = 1 — (3.18)where we make use of (3.11), while, under the new policy structure with tariffs t’ and t”,= p—c” (3.19).The rate of effective protection then expresses the variation in value-added as a proportionof the base case component:= (v1—v)‘(3.20).Note here that adopting v1 as the denominator, which remains constant over the variousscenarios, rather thanv1’, merely allows us to compare more easily those differences betweenscenarios that are internal to this discussion. We can use (3.10) and (3.11) to decompose thevariation in value-added into three discrete parts:46= u.(l—p) + (1—u)(1—p’ + (c’—cm) (3.21)in which the first two terms represent the changes in value in the domestic and exportmarkets for output, respectively weighted by their shares of that output, while the thirdmeasures the aggregate of changes occuring in each industry’s unit material costs.47Chapter FourSpecification of the ModelThe present model depends upon three categories of information and data togenerate estimates of vertical effective protection. Input-output data describe the resourcesand commodities that flow between industries and markets and provide us with the basis ofcoefficients that fix the technology of our economy. Data oil tariffs and other mrn-tariff-barrier equivalents set the frontier price levels. Cross-section industry-specific econometricinformation allows us to describe their price-setting behaviour through a unit shareparameter. In this chapter, we will outline our methodology and describe some of the dataas well as list the various sources for our information.Section 4.1 Input-output DataStatistics Canada provides a partitioned account and demand matrix of industry input-outputinformation which is periodically revised for some levels of aggregation and then completelyrevised at the most disaggregated level in particular years. In 1985, the complete table atthe lowest and most detailed level of aggregation (Level W) is a rectangular matrix relating603 rows of commodities, margins, transfers and other payments to 271 industries and48various final demand categories. Each row traces the 1985 Canadian dollar value of thatparticular commodity as it purchased either as an input to 1985 production within eachindustry or as it is allocated to private consumption or some other final demand category(these include various government purchases, as well as material and output inventories,exports and so on). The matrix can be characterized as being very roughly diagonal withmore homogeneous resource-based or primary commodities/industries in the upper left andincreasingly complex manufactured goods and services as one moves toward the lower right.As can be imagined, the density of the matrix is rather low: most industries do not use mostof the commodities, and groups of commodities tend to pass into groups of industries. Forthose industries comprising the category of Primary Agriculture for example, the matrix listsST direct commodity inputs from that sector itself (grains for seed and eggs for hatcherie& -and so forth) as well as various goods, services and transfers from elsewhere in the economy.This leaves some 516 types of direct commodity inputs valued at zero.In turn, most of the output of this sector is allocated to food processing and finaldemand categories with only relatively small amounts passing into some other form ofmanufacturing or service activity. As our main interest here is in effective protection as itchiefly affects the food processing sector and as it affects primary agriculture as animportant supplier of its inputs, much of the downstream commodity and industry detail canbe suppressed and aggregated. Our task here is to focus on the network of primaryagriculture and food processing activities, choosing a level of detail so as to produce a bothuseful and manageable square matrix of technical coefficients.49In the 1985 W-Level input-output account (‘use’) matrix, only two industries aredistinguished as comprising all of primary agriculture: livestock and field crops. In thecommodity dimension however, Statistics Canada details the output of these two activitiesinto 23 types of commodities. For the purposes of exposition (vis a vis rates of effectiveprotection within a mixed-price framework) the importance of this primary producing sectorlies doubly in the complex nature of the sometimes very high levels of protection affordedto much of their output and in the fact that this output serves as an essential component ofthe inputs passing into a relatively unprotected food processing sector. It is desirabletherefore to detail as much as possible these commodity inputs and their variable levels ofprotection.The disaggregation of these two livestock and field crop activities into a more usefularray of activities would require some key to distributing all of these inputs consistentlyacross those industries we so choose. Since Agriculture Canada and Statistics Canadatypically provide census and industry data for this sector which preserves a twelve-industrylevel of disaggregation, and since this would be a convenient and consistent source ofdistribution weights for all of the output and 84% of all inputs to this sector (see appendixB), we elected to initially confme ourselves to these as well. Three of these ‘farm types’included various combination and specialty farms that were of little interest as inputs intoother industries. These were aggregated into another miscellaneous activity. This left a totalof those ten industries in primary agriculture that are detailed in Appendix A, and theiroutput of 23 commodities was conformably aggregated into these ten. Table 4.1 presents acomplete list of these and the other 21 industries finally chosen for this model andTable4.1SharesofSectorandResourceCostsShARESOFCOSTSACCOIJNTEDFORBY:VALUEIN(MILLIONSPrimaryFoodOtherServicesLabourRateofNetTaxesOutputExportsImportsAgricProcessManufacReturnSubsidies1CattleandCalves0.3500.1350.0750.1090.0860.2350.0113302265532Daizy0.2990.1430.0420.0670.0820.427-0.0602964003Pigs0.4900.2520.0270.0660.0440.160-0.039176416104PoultryandEggs0.4840.2550.0170.0400.0640.1350.006137730565Wheat0.0210.0000.2960.1920.0290.530-0.0683764300216SmallGrains0.0680.0000.2980.1870.0290.473-0.05526972691077Ojlseeds0.0410.0000.3370.2000.0310.430-0.03914827011468Fruit,Vegetables0.0640.0000.2890.1760.1760.355-0.060138917511329Tobacco0.0310.0000.2330.2780.2780.445-0.19135814010Misc.Agriculture0.1220.0470.1860.1640.1640.360-0.0323242189159211RedMeat Proc.0.5120.2310.0250.0890.1090.0310.0048686108752012PoultryProc.0.5300.0890.0410.1090.1620.0640.006147755513FishProc.0.0000.1110.4080.1300.2190.1260.0064401944514Fruit,Veg.Proc.0.1170.1420.2080.1760.1690.1790.0092%330084015DairyProc.0.4650.1240.0660.1400.1190.122-0.035575921311516FlourMilling0.3450.0720.0510.1990.1430.1810.0081228305017FeedMilling0.2070.3430.0870.1790.0970.089-0.00127181487818Veg.OilMilling0.6790.0000.0140.2090.0440.0430.01156315121819BiscuitIndustry0.0000.1860.1030.1640.2730.2600.014484885120Bread,Bak.Prod.0.0010.2590.0800.1450.3410.1610.0131933205321SugarRefinery0.0000.0090.0320.6040.1650.1730.0174461008022SugarConf.0.0710.1730.1060.2120.2370.1880.01399810228123Tea,CoffeeBev.0.3710.0620.0480.1910.11702010.0117463723124Misc.FoodProd.0.0330.2070.1290.2500.1770.1%0.009291818746625SoftDrinkProd.0.0000.2250.2380.1480.1890.1870.013176596826DistilleryProd.0.0260.0740.1730.2520.2170.2000.05877338633027BreweryProd.0.0070.0760.1620.2150.2690.2490.02319811917828WineIndustry0.0830.0950.2090.1980.1740.2090.032246226529TobaccoProd.0.1170.1770.1250.1590.1870.2230.0131471924330OtherManufac.0.0010.0010.39402110.2180.1530.0223129399224210163831Services0.0020.0130.1090.2650.2920.2820.0363840173422518321MEANSPrimaryAgric.0.1970.0810.1770.1370.0810.369-0.041FoodProcessing0.2740.1770.0960.1610.1590.1310.003TotalAg.Sector0.2450.1410.1260.1520.1300219-0.014ALLINDUSTRIES0.0210.0180.2280.2340.2490.2240.026U.’C51provides a summary of some of their characteristics. Appendix A is an outline of theseindustries and an itemization of their subsumed commodities. Here, one may see that‘Miscellaneous Agriculture’ also includes commodities numbered 588 to 594 which appearin the account matrix as otheiwise unallocated imports and exports. These particular itemspresent a problem that is not really possible to deal with in a satisfactory way in this model.These commodities consist of a variety of horticultural items such as raw cotton and rubberbut also, and more unfortunately, sugar, cocoa, coffees, teas, and tropical fruits, all of whichare important inputs into a number of our food processing activities. We would like todistinguish their flow into food processing but these goods are ‘otherwise unallocated’precisely because no domestic industry exists which produces them. Basing their allocationon cost structure, it would have been better, perhaps, to have included them in a field cropactivity.The W-Level input-output account matrix goes on to distinguish 19 industries in thefood processing sector which correspond to a mix of the 1970 and 1980 Standard IndustrialClassifications and includes tobacco processing and various beverage industries. These 19are as they also appear in Table 4.1. Retaining this particular configuration seemed bothadequate and for the most part natural, but, with a greater interest in this particular sector,one might otherwise elect to further distinguish and disaggregate beef and pork, or fluid andindustrial milks, or fruits and vegetables, and so on. The W-Level matrix details the outputof this sector into 72 commodity groupings and, again, Appendix A describes theiraggregation and allocation to their respective industries.52As an ideal, we would procede through all of the remaining industries that constitutethe Canadian productive economy, continuing to aggregate commodities to match theirindustries. However, the information requirement to econometrically determine all theindustry-specific pricing parameters is very large and beyond the scope of the present work.We can choose to confine ourselves here to the two quite well-defined agriculture and foodsectors which by themselves are aptly paradigmatic with respect to effective protection andpricing behaviour. Nevertheless, both the manufacturing and service sectors are importantin the shares of input costs of the other two sectors. More particularly, it is the sectoraldifferences in this input use that is of interest here. As shown in Table 4.1, Primaryagriculture employs twice the value of manufactures that enter into food processing, and the-latter uses- 20% more services. Downstream, processed food appears to only account for1.3% of the costs accounted for by the service sector but this value is nevertheless muchlarger than that accounted for by primary commodity inputs. In addition, this is 1.3% of alarge portion of the economy and, overall, the service sector is a major and expandingmarket for processed foods.Because the service sector is large in the demand for processed output and is notprotected, and because manufacturing is a relatively important supplier to primaryagriculture and eventually loses protection under the trade agreement, we chose to retainthem as separate industries. Since both are very large aggregates here, there is no scope inthis model to capture all the minute variations in costs and prices that would doubtlesslyalso be occuring among all the activities that comprise manufacturing and services. We willtherefore need to exercise due caution in interpreting any particular results for these two.53Section 4.2 Pricing ParametersIt is in allowing for mixed pricing behaviour that we chiefly depart from the standardmeasurement of protection with its assumption of the law of one price, and, in Chapter 2,we examined evidence that suggested that other considerations become important whengoods are not homogeneous. More particularly, we see that pricing among heterogeneousproducts generally reflects some degree of both import prices and domestic costs, and thatthe degree to which an industry of such a product can take advantage of a tariff levied onits competing imports is apparently proportional to the degree of seller firm concentrationin that industry. Accordingly, our algebraic model (equations 3.2 and 3.4) introduced shareparameters s and sx and econometric information will assist us in assigning values for thesein each industry.-The proper econometric estimation of domestic and export pricing equations for theCanadian food processing industries as they existed in 1988 was beyond the scope of thepresent work and so we depend upon an appropriate study of this sector undertaken withcross-section data for these industries as they existed in this country in 1982. This researchwas done by Hazledine and Maundu (1990), and the values for the share parameter s asthey appear in Table 4.2 are due to a particular weighted least squares regression performedon a sample of 25 food processing industries. The variables employed included the ratio ofCanadian and U.S. input prices, Canadian and U.S. tariff rates and 4-firm sellerconcentration ratios, Canadian shares of output exported as well as the incidence of nontariff barriers and certain regional commodities protected by transport costs. These latterwere set at s = 0.1 since the regression results supported the suggestion that for thoseTable4.2Tariffs,PSEsandPricingParameters.TariffRatesTCanTUSPSEratesTCUSTAFreeTradeT*T’T**PricingParametersSSx30OTHERMANUFACTURES31SERVICES0.0610.0470.0000.0001.0611.0471.0001.0001.0001.0001.0001.0001.001.001.001.000.990.800.990.991CATTLEANDCALVES0.0050.0051.0101.0101.0001.0001.001.000.990.992DAIRY0.1900.1751.4001.2001.4001.2001.001.000.990.993PIGS0.0000.0001.1001.0501.0501.0001.001.000.990.994POULTRYANDEGGS0.0400.0401.3201.0401.2801.0001.001.000.990.995WHEAT0.0200.0401.0201.1001.0001.0601.001.000.990.996SMALLGRAINS0.0100.0101.0101.0101.0001.0001.001.000.990.997OILSEEDS0.0000.0101.0001.1001.0001.0901.001.000.990.998FRUIT,VEGETABLES0.0300.0301.1301.0301.0001.0001.001.000.990.999TOBACCO0.0600.0601.0601.0601.1001.0001.001.000.990.9910MISCAGRICULTURE0.0100.0101.0101.0101.0001.0001.001.000.990.9911REDMEATPROCESSING0.0190.0101.0191.1371.0001.1261.001.000.500.7512POULTRYPROCESSING0,1080.0671.2011.0731.1081.0061.001.000.500.7513FISHPROCESSING0.0190.0161.0191.0161.0001.0001.001.000.400.5014FRUIT,VEGETABLEPROCESSING0.0950.0811.0951.0871.0001.0061.001.000.700.7515DAIRYPROCESSING0.0790.0401.2721.4501.1931.4101.001.000.300.7516FLOUR,MEAL,CEREALMILLING0.0280.0151.0481.0221.0201.0071.001.000.500.7517FEEDMILLING0.0300.0051.0301.0051.0001.0001.001.000.100.5018VEGETABLEOILMILLING0.1500.1201.1501.1201.0001.0001.001.000.700.7519BISCUITINDUSTRY0.0500.0051.0501.0051.0001.0001.001.000.800.7520BREAD,BAKERYPRODUCTS0.0400.0101.0401.0101.0001.0001.001.000.100.5021CANE,BEETSUGARREFINERY0.0900.0301.0501.4601.0001.4301.001.000.990.7522SUGARCONFECTIONARY0.0490.0651.0491.1051.0001.0401.001.000.800.7523TEA,COFFEEBEVERAGES0.0050.0001.0051.0001.0001.0001.001.000.700.7524MISCFOODPRODUCTS0.0800.0301.0801.0301.0001.0001.001.000.700.7525SOFTDRINKPRODUCTS0.1750.0401.1751.0401.0001.0001.001.000.100.5026DISTILLERYPRODUCTS0.0500.0401.0501.0401.0001.0001.001.000.800.5027BREWERYPRODUCTS0.0300.0151.4801.0151.4501.0001.001.000.990.5028WINEINDUSTRY0.0350.0691.5751.0691.0001.0001.001.000.800.5029TOBACCOPRODUCTS0.1650.2071.1651.2071.0001.0001.001.000.990.50(ii55industries with a regional market bounded naturally by high transport costs, tariffs did notprove to be of additional benefit in providing appreciable protection.Within the primary agriculture sector, it is usual to assume that the export marketfor their tradeable and homogeneous output is large and highly competitive. For theseindustries it is appropriate to maintain the classical model and, as price-takers, s iS set toone for each. For computational reasons however (equations 3.14 and 3.15 would involvedivision by zero), it should be noted here that s. in these and all other cases in which thestandard hypothesis holds is actually set numerically at 0.99.No econometric specification existed for the export pricing equations for theseindustries and so various types of information concerning the structure of the Canadian foodprocessing sector and the nature of their products and markets were employed so as toassign an appropriate range of values to the parameter s’. These are as they appear in Table4.2.For the ten agricultural industries, as before, assuming homogeneity of output andcompetitive price behaviour ensures that SX will equal one. The food processing industries,however, were divided into two groups. Those industries producing more highlydifferentiated national brands (breweries, wineries, distilleries, and tobacco) and thoseprotected by transport barriers (feeds, bakeries, soft drinks, and breweries again) wereassigned a value of s1x = 0.5. Because the fish processing industry is large in the shares of itsexport markets, it was also assigned a value of 0.5. For the other group comprising all theremaining industries, we assume that they have less power in their respective export marketsthan they do at home and set s=0.75.56Section 4.3 Protection Data and Tariff ParametersThe first two columns of Table 4.2 provide an adjusted schedule of the Canadian and U.S.ad valorem tariff rates as they stood in 1988 on the eve of the initial implementation of theCanada-U.S. Free Trade Agreement. Since these commodities are aggregates and manysubsumed items were already being traded freely these rates appear lower than the statutoryrates. For the first ten industries that comprise primary agriculture, these numbers representthe total duties paid as a share of the total value of each commodity group’s imports. Theseare aggregates of conformable 6-digit commodity data given in Statistics Canada catalogue65-203 (1988). Because of the emphasis in this study given to the food processing sector, wetend here to interest ourselves more with agriculture as an input into processing, and, incalculating these rates, we have excluded duties paid on produce that is consumed fresh andenters directly into the service sector and consumption.Tariffs levied by the U.S. were calculated by first comparing the statutory levels inboth countries as they appear in the appending two-volume Tariff Schedule (1987) publishedwith the Trade Agreement. The apparent ratio of U.S. to Canadian levies were then appliedto the previously calculated vector of Canadian ad valorem rates. Some of these rates alsoreflect supplementary information published in Table 1 of Hazledine and Vaughan (1986)and (for food processing) Table C2 of Lester and Morehen (1988).The base-case elements of the vectors t and t used in the model are those thatappear as columns 3 and 4 in Table 4.2. These are defined as one plus the proportional advalorem rate of protection that were in place until the implementation of the Trade57Agreement in 1988 and inclusive of the producer subsidy equivalents (PSE’s) of anymeasurable non-tariff barriers.The relevant PSE’s used for the agricultural sector were drawn from a U.S.D.A.(1988) study which presents a comparative inventory of such measures for a number ofprimary products in both Canada and the U.S.. These estimates include such measures asstabilization payments and quantity restrictions and other subsidies and barriers, some ofwhich will remain intact but most of which are scheduled to be gradually dismantledbetween 1991 and 1998. Compensating for the likelihood that the some level of these PSEswill remain installed in the attempt to negotiate a level playing field in this sector, thesewere calculated as net differences of the respective PSEs reported between the twocountries. The particular estimate fo thern poultry industry is drawn from Mosini anMeilke (1989).The estimates for t and t’ in the food processing industry are based on those ofLester and Morehen’s for that sector. Our numbers reflect the need in our model to furtherdisaggregate the various grain milling industries and to separate distilleries from breweriesand include non-tariff barriers on beer. The values for other manufactures come from theDepartment of Finance (1987), published as an assessment of the CUSTA.Table 4.2 presents two sets of vectors t’ and t” that correspond to the two protectionscenarios that we consider here. The Canada-U.S. Trade Agreement provided for an initialimmediate reduction in the tariffs applied to most goods traded between the two countriesbut which left a schedule that allowed for certain non-tariff barriers to remain in place onparticular commodities. Columns 5 and 6 present the base case scenario’s initial CUSTA58levels of remaining protection while columns 7 and 8 are identified as those vectors as theyappear in the free trade scenario with all trade barriers abolished. Our model will beassuming that the U.S. is large and the dominant price-setter in all trade flows between thetwo economies. For most commodities for which there were substantial trade flows prior tothe Agreement, the United States is both the largest importer of Canadian output and hasthe largest shares of Canadian imports.59Chapter FiveResultsThe following sections outline and table the results of four scenarios which are eachin turn evaluated according to the pricing hypothesis assumed. Initially we examine the basecase, measuring the change in value-added for each of the model’s 31 industries under the1988 implementation of the Canadian-U.S. Trade Agreement. The second section then looksat. the consequences of dismantling theremaining non-tariff barriers and moving to a fullfree trade situation. Sections 5.3 and 5.4 turn to briefly examine the effects of exchange rateappreciation and depreciation on the two CUSTA outcomes. Finally, in Section 5.5, we seehow the two pricing assumptions differ in ranking the processing industries according totheir levels of effective protection and the changes in their various returns to capital underthe two tariff scenarios.Section 5.1 CUSTA (1988): The Base Case with No Macro ResponseThe base case scenario examines the potential differences in the measurements of theimpacts deriving from the Trade Agreement under the two pricing hypotheses while initiallypresupposing that no large macroeconomic change occurs to effect an imbalance in tradepayments and a consequent change in the rate of currency exchange. The model’s results60for this scenario under either mixed pricing or the law of one price appear in Table 5.1. Foreach pricing model, G is the adjusted rate of effective protection as formulated in equation(3.20) and is defined as the change in value-added calculated as a per centage of the basecase component. Variables DPE, EPE, and MCE represent the decomposition of thechanges in value-added into the three components formulated in equation (3.21). DPE isthe change in domestic output price weighted by the share of domestic use of total output,while EPE is the change in export price weighted by the remaining share of exports fromtotal output. The changes in unit material costs (total costs net of all payments to primaryresources, (CmC’m)) is represented under MCE.Firstly, we see that the adoption of the Canada-U.S. Trade Agreement results in anoverall reduction in effective protection of the food processing and manufacturing sectors.Output-weighted means over both the agriculture and processing sectors indicate anestimated reduction in value-added ranging between 4.3 and 7.6% depending upon whetherwe assume mixed pricing or the standard hypothesis. Almost all of this reduction is likelyto be borne by the processing sector in which this range is 6.7% under mixed pricing to asmuch as 12.1% under the law of one price assumption. In the main, while nearly allindustries tend to benefit from 1.5 to 2% lower input costs, it is the general decline in theirown domestic output prices that leads to the pressure on the added value in food processingand manufacturing.The primary agriculture and service sectors seem to fare a little better. Since primaryagriculture continues to receive much of its non-tariff protection under the CUSTA, thissector nearly averages zero in the predicted changes in value-added over all of itsTable5.1Scenarios1and2:CUSTAandFreeTrade(e=1.0)SCENARIO1CUSTASCENARIO2FREETRADEMixedPricingLawof OnePriceMixedPricingLawof OnePriceGDPEEPEMCE0OPEEPEMEGDPEEPEMCE0OPEEPEMCE1CattleandCaives0.011-0.0090.0010.0120.013-0.0090.0010.0130.012-0.0090.0010.0120.013-0.0090.0010.0132Dairy0.0370.0070.0000.0100.0390.0070.0000.1)11-0.844-0.3890.0000.010-0.842-0.3890.0000.0113Pigs-0.149-0.0440.0040.015-0.144-0.0440.0040.016-0.417-0.0890.0040.016-0.415-0.0890.0040.0164PoultryendEggs-0.090-0.0340.0010.015-0.086-0.0340.0010.015-1.402-0.3040.0010.019-1.398-0.3040.0010.0195Wheat0.074-0.0040.0240.0160.077-0.0040.0240.0180.165-0.0040.0690.0160.168-0.0040.0690.0186SmaiiGrains0.019-0.0090.0010.0160.023-0.0090.0010.0180.019-0.0090.0010.0160.023-0.0090.0010.0187Oliseeds-0.046-0.0000.0020.0180.053-0.0000.0020.0200.132-0.0000.0380.0180.136-0.0000.0370.0208Fruit.vegetables-0.194-0.1180.0020.014-0.192-0.1180.0020.014-0.194-0.1180.0020.014-0.192-0.1180.0020.0149Tobacco0.13200400.0020.0200.1360.0400.0020.022-0.070-0.0560.0020.020-0.066.0.0550.0020.02210Misc.Agriculture0.009-0.0100.0000.0130.010-0.0100.0000.0140.009-0.0100.0000.0140.011-0.0100.0000.01411RedMeat Proc.0.014-0.0160.0000.0190.022-0.0160.0010.0190.124-0.0210.0100.0300.184-0.0160.0140.02912PoultryProc.-0.139-0.0540.0000.025-0.259-0.0870.0000.080.018-0.174-0.0000.1820.083-0.1630.0000.18113FishProc.0.004-0.0150.0000.0260.027-0.0180.0000.0270.004-0.0150.0000.0260.027-0.0180.0000.02714Fruit.Veg.Proc.-0.088-0.0680.0040.036-0.111-0.0840.0060.039-0.085-0.0690.0040.037-0.108-0.0840.0060.03915DairyProc.-0.102-0.0230.0010.005-0.307-0.0740.0010.010-0.035-0.2170.0040.2090.075.0.2020.0080.20716FlourMiiiing-0.024-0.0200.0000.013-0.037-0.0260.0000.014-0.052-0.0310.0000.014.0.093-0.0460.0000.01517FeedMiiiing-0.009-0.027-0.0010.0270.032-0.0270.0000.033-0.008.0.028-0.0010.0290.037-0.0270.0000.03418veg.OilMilling-0.700-0.0850.0150.001-1.005-0.1200.0200.001-0.700.0.0850.0150.001-1.005-0.1200.0200.00119BiscuitIndustry-0.041-0.035-0.0000.013-0.045-0.0410.0010.015-0.039-0.036-0.0000.015-0.041-0.0410.0010.01720Bread.Bak.Prod.-0.009-0.018-0.0000.014-0.042-0.0390.0000.017-0.010.0.023-0.0000.019-0.032-0.0380.0000.02221SugarRefinery-0.102-0.0400.0020.002-0.100-0.0400.0030.Oip20.019-0.0400.0450.0020.060-0.0400.0590.00222Sugar Conf..0.047-0.0380.0030.015-0.051-0.0440.0040.018-0.032-0.0390.0050.021-0.034.0.0440.0070.02123Tea,Coffee8ev.0.003-0.005-0.0000.0070.007-0.005-0.0000.0Q70.003-0.005-0.0000.0070.007-0.005-0.0000.00724Misc.FoodProd.-0.096-0.0580.0010.022-0.118-0.0730.0020.027-0.085-0.0600.0010.028-0.105-0.0730.0020.03125SoftDrinkProd.-0.049-0.0400.0000.024-0.303-0.1640.0000.046-0.049-0.0400.0000.024-0.303-0.1640.0000.04626DistilleryProd.-0.021-0.0270.0050.013-0.009-0.0320.0130.014-0.021-0.0270.0050.013-0.009-0.0320.0130.01427BreweryProd.-0.011-0.020-0.0000.013-0.003-0.0190.0010.015-0.761-0.424-0.0000.014-0.752-0.4230.0010.01528WIneindustry-1.039-0.4510.0000.048-1.298-0.5500.0000.051-1.039-0.4510.0000.048-1.298-0.5500.0000.05129TobaccoProd.-0.272-0.1480.0040.029-0.256-0.1480.0100.00-0.242-0.1460.0030.041-0.22541460.0100.04130OtherManufac.-0.032-0.0400.0070.022-0.033-0.0460.0100.023-0.032-0.0400.0070.022-0.033-0.0460.0100.02331Services0.010-0.000-0.0000.0060.012-0.000-0.0000.0070.011-0.000-0.0000.0070.012-0.000-0.0000.008MEANSPrimaryAgric.-0.002-0.0160.0050.0140.001-0.01600050.015.0.203-0.0900.0150.015-0.200-0.0900.0150.016FoodProcessing-0.067-0.0400.0010.020-0.121-0.0600.0020.024-0.061-0.0980.0040.062-0.046-0.1040.0070.064Total Ag.Sector4043-0.0300.0030.0174076-0.0430.0030.019-0.115-0.0940.0080.043-0.105-0.0980.0100.044ALLINDUSTRIES-0.011.0.0190.0030.013-0.014-0.0220.0040.015-0.016-0.0240.0040.016-0.016-0.0270.0050.017‘-I62activities. Through the loss of some nominal protection on their output, the only industriesthat decline here are pigs, poultry, and fruits and vegetables. The rest of this sector makesmodest gains that match or exceed the 1% gain in services. Wheat, oilseeds, and tobaccoproduction appear to be the strongest gainers but for varying reasons. Wheat and oilseedsgain in the export market but for oilseeds, it is due to the remaining foreign tariff on itsoutput. Tobacco actually receives more protection under the CUSTA scenario than priorto that agreement. As can be seen in Table 4.1, these industries also have rates of returnranging from 35 to 53%.In fact, since the agricultural industries were here supposed as already pricing up tothe full extent of the tariff, there are only very small differences between the two hypothesesfor this sector and these differences only derive from how the two pricing mo4lelsincorporate the use of primary factors and higher intermediates. Only changes in materialcosts would have resulted in major changes in this first scenario, and Table 4.1 shows thatmaterial costs account for a relatively smaller proportion of farm costs than they do foreither food processing or manufacturing, and more than 20% of these inputs come from itsown sector. The consequences of the tariff agreement and the implications of the pricinghypotheses assumed are much more important for the downstream sectors.Looking more at the individual processing industries we see that there are nosignificant gains here under the CUSTA policy framework. As one might expect in movingfrom a more protected tariff structure, sixteen of the nineteen industries experience pressureto lose value-added under the agreement, and wineries, vegetable oils, poultry, dairy, andfruit and vegetables are among those most severely squeezed. For some, it is because63sizeable non-tariff barriers remain on some of their primary inputs. For others, it is thereduction in their output prices and the relatively small size of their value-added thatbecome highly critical factors. Vegetable oil milling is an extreme case. Prior to the tradeagreement, material costs accounted for 90.2% of the output price and, out of the remainingsurplus, the rate of return accounted for a mere 4.3%. Material costs in this industry arevirtually unaffected by the CUSTA schedule however, while the price of its output drops by8.5%. As a result, even though its export market price improves, the already very smallproportion of value-added in this industry is predicted to contract by as much as 70%.Wineries, on the other hand, lose a great deal of output protection under the CUSTA butappear worse off in this model than they should because the reduction in the price ongrapes (that they receive under both the CUSTA and with free trade>is a detail that is lostin the aggregation of grapes into miscellaneous agriculture.By assuming the law of one price for all industries, pressure on value-added in theprocessing sector as a whole is generally 50 to 75% greater than that found for the base casemixed pricing model. Material costs are some 20% lower while output prices in the domesticmarket are nearly 50% lower than obtained under mixed prices. The standard modelpredicts much larger losses for the still-protected poultry and dairy processors as well as forfruit and vegetable processing, bread, and soft drinks, industries for whom regionaldifferences and high transport costs appear to be more constraining barriers to entry thanthe initial pre-1988 tariff structure. Feed mills also produce a regional product but thisindustry curiously gains in this scenario with lower than average declines in output pricecombined with somewhat larger than average input cost savings. Notably, 34% of this64industry’s inputs come from within the processing sector itself (from flour and cereal millingand from oilseed milling) from industries that lose protection hider the agreement.Section 5.2 Free Trade with No Macro ResponseMoving to free trade and eliminating those barriers that presently remain under the CUSTAhas a dramatic effect on the model’s results as illustrated for each pricing hypothesis underScenario 2 in the remaining two columnal sections of Table 5.1. Here, a 9% overall declinein output prices now leads to a squeeze on value-added in primary agriculture by as muchas 20% over the whole sector. This is chiefly due to the removal of non-tariff barriersremaining on four protected primary food industries: dairy, pigs, poultry and tobacco. All- of these activities would likely suffer massive reductions in their own output prices while -continuing to realize only very modest gains in the form of lower costs. These costs remainnear the base case levels that occur under the agreement as in Scenario 1 and these are onlyabout 1.5% lower than those existing prior to the Agreement. In fact, all the scenarios makeit quite clear that the agricultural sector has generally the least to gain with respect to lowercosts. Of these primary industries that show any improvements under free trade, only wheatand oilseeds, again, are of any great size. These industries continue to face foreign tariffsunder the CUSTA, but here, under free trade, higher export prices and negligible drops indomestic prices for output combine to contribute to quite large increases in the value-addedof these two: about 13.2% for oilseeds and as much as 16.5% for wheat.For the processing sector, the move to free trade is more of a mixed and qualifiedimprovement over the CUSTA framework. Although output prices for these industries also65show an overall decline of nearly 10% across the sector, 6.2% lower input costs contributemuch to ameliorating an otherwise bleak result. Value-added is reduced here by only 6.1%:somewhat less pressure than occurs under the initial CUSTA implementation. As theremaining trade barriers are dismantled, two processing industries capture very large costreductions: dairy and poultry each improve in this respect by 16 and 20% and, even thoughthey also lose non-tariff protection in this scenario, both activities lose less value-addedunder free trade than under the present CUSTA structure. Indeed, the sector as a wholeappears to gain about 10% in effective protection in moving from the CUSTA frameworkto free trade. Nevertheless, of all 19 food processing industries, only five would appear toreceive positive effective protection here, and, of these, only red meat processing appearsto have much scope for expansion. Value-added in that activity is predicted to increase bymore than 12%. Breweries retain a high degree of protection under the CUSTA and so joinwineries and oil mills as one of those industries likely to experience the greatest degrees ofreorganization under free trade. These three activities share losses in value-added of anorder comparable to those occuring in primary agriculture in the dairy and poultryindustries.The free trade scenario under the standard hypothesis differs only in certain respectsfrom the mixed pricing case. The two models both treat primary agriculture as beingperfectly competitive and the assumptions only influence pricing in their input markets. Asa consequence, their results for the upstream farming sector are virtually the same.The service sector, however, is also formulated in both models as being competitiveand is nominally unprotected in the pre-policy calibration, but, in not allowing for66imperfectly competitive pricing among the various input suppliers to that industry, thestandard model would have overestimated the degree of effective protection and, hence, thepotential of gains to this sector in either scenario. From these two tables and, as a firstestimate of this measurement bias, it would appear that the law of one price assumptioncould conceivably overestimate effective protection and factor returns in this ‘competitive’downstream activity by about 10% in the case of free trade, and perhaps as much as 20%in the base case CUSTA analysis. The constant-returns formulation of this model happensto be blind to the relative sizes and gross outputs of these various activities, but,nevertheless, for a large enough section of the economy, such a degree of potential bias isof considerable importance.The standardmodei shows the food processing sector as gaining much more effectiveprotection under free trade than under the CUSTA policy structure. Perhaps the mostnotable differences here are larger protection effects for red meat processing, smaller lossesfor poultry, and dairy processing, and larger losses for the more regionally differentiatedcommodities including fruit and vegetable processing, flour, feeds, and soft drinks.Overall, by assuming the law of one price, the main difference between the CUSTAoutcome and free trade is the degree to which either of these agriculture-based sectors gainsat the relative expense of the other. An economy-wide output-weighted decision based onthe standard hypothesis and assuming no rationalization and substitution gains wouldprobably just favour CUSTA (G=-0.014) over the free trade arrangement of Scenario 2(G = -0.016), but, under broader optimization assumptions, this pricing hypothesis could veryconceivably be indifferent between the two. For one thing, it would have predicted a much67greater reduction in the food processing sector (and therefore among all the imperfectlycompetitive manufacturing activities) under the CUSTA framework than the mixed pricemodel. By allowing for mixed pricing, an output-weighted evaluation of the two protectivestructures is less equivocal: the more sophisticated model shows value-added over both foodsectors to decline by an average of 11.5% under free trade in contrast to a 4.3% loss in thebase case. Over the industrial economy as a whole, effective protection and potential lossesin added value would be reduced from -1.1% with the CUSTA to -1.6% under free trade.Section 5.3 CUSTA (1988): The Base Case with DepreciationThe depreciation of a currency is often thought of as being inferior to an outcome in whichappreciation takes place although, as ever, this is primarily a matter of balance ininternational payments and changes in the exchange rate reflect and in turn modify thoseresource flows occuring between trading markets. If the value of rising imports outstrips thatof the exports required to finance this flow, downward pressure on the exchange rate andrevaluation are movements to correct any disequilibria in the international and domesticmarkets for both traded and non-traded resources.Table5.2Scenarios3and4:CUSTAwithexchangeappreciation(e=1.1)anddepreciation(e=0.9)SCENARIO3CUSTA,e-1.1SCENARIO4CUSTA,e-0.9MixedPricingLawofOnePriceMixedPricingLawofOnePriceGDPEEPEMCEGDPEEPEMEGOPEEPEMCEGDPEEPEMCE1CattieandCaives0.1180.0820.009-0.0520.1140.0820.009-0.053-0.096-0.100-0.0070.076-0.088-0.100-0.0070.0782DaIry0.1510.1100.000-0.0420.1460.1090.000-0.044-0.077-0.0970.0000.062-0.068-0.0950.0000.0653PIgs-0.0430.0430.014-0.064-0.0590.0430.014-0.066-0.254-0.131-0.0050.094-0.230-0.130-0.0050.0984PoultryandEggs0.0320.0640.003-0.0610.0160.0630.003-0.063-0.212-0.132-0.0010.090-0.189-0.131-0.0010.0935Wheat0.1850.0160.107-0.0320.1850.0160.107-0.032-0.037-0.024-0.0580.064-0.031-0.024-0.0590.0676SmallGrains0.1240.0810.011-0.0360.1240.0810.011-0.036-0.086-0.098-0.0090.069-0.079-0.098-0.0090.0727Oiiseeds0.1570.0570.046-0.0360.1570.0570.046-0.036-0.060-0.057-0.0410.073-0.052-0.057-0.0410.0768Fruit,Vegetables-0.108-0.0350.009-0.032-0.108-0.0350.009.0.032-0.279-0.201.0.0050.059-0.275.0.201-0.0050.0619Tobacco0.2500.1410.006-0.0290.2500.1410.006-0.0290.014-0.061-0.0020.0700.022-0.061-0.0020.07310Misc.AgrIculture0.1110.0860.004-0.0360.1110.0860.004-0.036-0.094-0.105-0.0040.062-0.090-0.105-0.0040.06411RedMeat Proc.0.0820.0650.011-0.0640.1230.0700.012-0.065-0.054-0.097-0.0110.101-0.079-0.103-0.0110.10312PoultryProc.-0.0860.0330.000-0.050-0.1780.0070.001-0.048-0.191-0.142-0.0000.101-0.340-0.181-0.0000.10413FishProc.0.0620.0650.002-0.0360.1290.0780.002-0.037-0.055-0.095.0.0020.088-0.076.0.115-0.0020.09114Fruit.Veg.Proc.-0.0090.0120.011.0.024-0.0200.0020.014-0.023-0.167-0.149-0.0040.097-0.203-0.16940030.10115DaIryProc.-0.0410.0650.004-0.075-0.2330.0170.004-0.070-0.162411240030.085-0.3824165-0.0030.09016FlourMilling0.0460.0660.002-0.0520.0610.0690.003-0.052-0.096-0.10740020.078-0.136-0.122-0.0020.07917FeedMilling0.0410.0540.004-0.0490.1350.0650.006-0.046-0.059-0.107-0.0050.103-0.072-0.120-0.0050.11218Veg.OilMilling.0.683-0.0140.036-0.089-1.000-0.0500.042-0.089-0.718-0.156-0.0050.091-1.011-0.189-0.0010.09119BIscuitIndustry0.0440.0390.015-0.0290.0520.0400.017-O.094127-0110-0.0150.056.0.142412140160.05920Bread,Bak.Prod.0.0160.0400.001-0.0310.0550.0580.001-0.O30-0.035-0.075-0.0010.059-0.140-0.135-0.0010.06421SugarRefinery.0.0150.0380.019-0.062-0.0110.0380.020-0.062-0.189-0.118-0.0150.0664188-0.118-0.0150.06622Sugar Conf.0.0370.0460.011-0.0390.0450.0450.012-0.038-0.132-0.121-0.0040.069-0.147-0.133-0.0040.07323Tea.CoffeeBev.0.0900.0860.004-0.0590.1080.0910.004.0.059-0.084-0.096-0.0040.0734093-0.100-0.0040.07424Mlsc.FoodProd.-0.0180.0250.006-0.036-0.0270.0160.007-0.0ç33-0.174-0.141-0.0040.081-0.209-0.162-0.0040.08725SoftDrinkProd.-0.0120.0270.000-0.0294226-0.0770.001-0.012-0.087-0.107-0.0000.077-0.380-0.251-0.0000.10326DIstilleryProd.0.0570.0310.034-0.0380.0900.0310.049-0.08-0.099.0.086-0.0240.064-0.1094095-0.0230.06627BreweryProd.0.1180.0860.007-0.0300.1270.0870.011-0.09-0.140-0.126-0.0070.057-0.132-0.12440080.06028Wineindustry-1.024-0.3900.000-0.008-1.291-0.4910.001-0.006-1.0544513-0.0000.104-1.305-0.609-0.0000.10829TobaccoProd.-0.193-0.0650.009-0.026-0.176-0.0650.017-0.026-0.351-0.231-0.0020.084-0.33642310.0030.08630OtherMantifac.0.0560.0300.02840360.0650.0280.033-0.036-0.119-0.110-0.0140.079-0.131-0.120.0.0130.08331Services0.1120.0910.008-0.0310.1120.0910.008-0.031-0.091-0.091-0.0080.044-0.089-0.091-0.0080.045MEANSPrimaryAgric.0.1060.0630.026-0.0420.1030.0630.026-0.042-0.109-0.095-0.0160.070-0.100-0.094-0.0160.073FoodProcessing0.0150.0560.008-0.055-0.0140.0430.010-0.053.0.1494135-0.0060.095-0.227-0.163-0.0060.100Total Ag.Sector0.0390.0500.015-0.0470.0190.0420.016-0.046-0.126-0.110-0.0100.080-0.171-0.128-0.0090.084ALLINDUSTRIES0.0830.0630.017-0.0350.0850.0610.019-0.034-0.105-0.101-0.0110.061-0.113-0.106.0.0110.0640)69The present model is one of partial equilibrium and, as explained earlier, nomacroeconomic mechanism exists here to interactively determine the rate of exchange withinthe model itself. As a consequence, we assume a 10% devaluation in this scenario togenerate an estimate of its effect on the various industries.As seen in Table 5.2, a 10% depreciation carries the mixed-pricing model to asolution in which the economy as a whole apparently gains an output-weighted 9.4% inadded value over the base case CUSTA Scenario in Table 5.1. The total agriculture andprocessing sectors recapture about 8% of this depreciation effect but most of this is due tothe primary industries for which enhanced trade is of greater importance and benefit andnearly matches the effective protection gains that appear here for the service sector.Overall, material costs rise by 4 to 5% over the pre-CUSTA levels but 6.3% higherdomestic output prices and 2.6% higher export prices tend to compensate and apparentlycontinue to favour the primary sector over food processing. With the CUSTA structure inplace, food processing continues to benefit least from increased export prices. In summary,most food processing industries do somewhat better with a depreciated exchange but theirgains are moderate in and of themselves and, in any case, outstripped by the rest of theeconomy.The result under the standard hypothesis for this scenario departs from the mixedprice outcome in underestimating movements in domestic prices and predicting a reductionin food processing. Again, and in general, the classical model predicts rather larger absolutevalues of gains and losses in added value although it manages to generate a similar outputweighted total for the economy as a whole.70Section 5.4 CUSTA (1988): The Base Case with AppreciationThe most favourable scenario promulgated by free trade advocates is that productivity gainscaptured by the economy will result in cheaper output with a rise in exports and aconcommitant appreciation of the currency. Within the model’s partial equilibrium structureand with no mechanism to incorporate such changes in productivity, the results that followfrom a currency appreciation are understandably severe but convey some sense of themagnitude of productivity improvements that would be required in order for the economyto begin to show gains in this direction.Table 5.2 also outlines the results for both pricing hypotheses with the modelincorporating a 10% appreciation under the CUSTA structure. Under either hypothesis, thereductioii in added value- in primary agriculture- nearly matches that for the economy--as awhole. Effective protection appears to contract in this sector by almost 11% due to a dropin both domestic and export output prices and despite considerably lower costs. Again,however, the two models diverge dramatically with regard to the processing industries. Themixed pricing assumption results in estimating a 15% decline in that sector, while thestandard model, predicting much lower domestic prices, shows a reduction of nearly 23%.In either case, added value in virtually every industry declines.Table 5.3 Rankings of ERPs and the Rate of ReturnSCENARIO 1 - COSTA FREE TRADE SCENARIO 1 - COSTA FREE TRADEMixed Law of Mixed Mixed Law of MixedPricing One Price Pricing Pricing One Price Pricing0 0 G dPR dPqR dPJRRed Meat Proc. 0.014 1 0.022 3 0.124 1 0.029 1 0.047 2 0.541 1Fish Proc. 0.004 2 0.027 2 0.004 4 -0.028 4 -0.038 5 -0.028 6Tea. Coffee 3ev. 0.003 3 0.007 4 0.003 5 -0.000 2 0.004 3 -0.000 4Feed Milling -0.009 4 0.032 1 -0.008 6 -0.019 3 0.055 1 -0.019 5Bread. Bak. Prod. -0.009 5 -0.042 8 -0.010 7 -0.029 5 -0.138 9 -0.031 7Brewery Products -0.011 6 -0.003 5 -0.761 18 -0.030 6 -0.009 4 -1.652 18Distillery Prod. -0.021 7 -0.009 6 -0.021 8 -0.069 8 -0.044 6 -0.069 8Aour.Cereai Mill. -0.024 8 -0.037 7 -0.052 13 -0.046 7 -0.072 7 -0.097 11Biscuit industry -0.041 9 -0.045 9 -0.039 11 -0.039 9 -0.102 8 -0.088 10Sugar Confect -0.047 10 -0.051 10 -0.032 9 -0.133 11 -0.145 10 -0.098 12Soft Drink Prod.-0.049 11 -0.303 16 0.049 12 -0.102 10 -0.639 16 -0.102 13Fruit. Veg. Proc.-0.088 12 -0.111 12 -0.085 14 -0.211 14 -0.270 13 -0.205 15Misc. Food Prod. -0.096 13 -0.118 13 -0.085 15 -0.202 13 -0.251 12 -0.182 14Dairy Proc. -0.102 14 -0.307 17 -0.035 10 -0.183 12 .0.550 15 -0.071 9Sugar Refinery -0.102 15 -0.100 11 0.019 2 -0.211 15 -0.206 11 0.037 2Poultry Proc. -0.139 16 -0.259 15 0.018 3 -0.558 17 -1.034 17 0.003 3Tobacco Prod. -0.272 17 -0.256 14 -0.242 16 -0.522 16 -0.490 14 -0.464 16Veg. Oil Mill. -0.700 18 -1.005 18 -0.700 17 -1.598 18 -2.291 18 -1.598 17Wine industry -1.039 19 -1.298 19 -1.039 19 -2.052 19 -2.507 19 -2.052 19Section 5.5 Rankings of Effective ProtectionWe can summarize some of the foregoing results by ranking the food processing industriesaccording to the degree to which their respective value-added products are affected by thesechanges in tariffs and protective structure. In Table 5.3, these nineteen activities are firstarranged in descending order by their ERPs as estimated under the mixed-price CUSTAmodel of Scenario 1. One can see from the next column that, at least within this particularsector, the standard law of one price assumption generates a ranking that is fairly consistentto that of the more sophisticated pricing model. The most notable exception here is the softdrink industry which retains a large foreign tariff under the CUSTA and moves down fromthe eleventh position to the sixteenth. The feed, bakery, and daiiy product industries alsoshift about somewhat, again emphasizing the degree to which the law of one price is7172misapplied to activities that are differentiated regionally and are more naturally protectedby transport costs.The complete removal of remaining trade barriers that occurs with free trade inscenario 2 results in some very dramatic changes in the ranking of these industries and theseare shown in the third column of Table 5.3. Sugar refineries and poultry processing improvein a rather spectacular fashion. At the other extreme, we find that breweries move fromsixth to the eighteenth. Flour miling and dairy processing are also much affected.Finally, we compare the ranking differences between the two pricing hypotheses asthey affect the proportional change in the rate of return. These are calculated as (r1-r’)/r,and they appear as the last three columns of Table 5.3. The ordering of these activities bychanges in r follows fairly closely that of the rates of effective protection. The only exceptionis with the wine industry which gains protection under the CUSTA but nevertheless suffersa 13% reduction in the return available to capital. In fact, with the sole exception of redmeat processing, the rate of return in all these processing industries declines and by aproportionately greater amount than their changes in total value-added. With the free tradestructure of Scenario 2, the results with mixed pricing again appear quite bleak for most ofthese activities: except for meat processing and vegetable oil milling, most industriesexperience even larger declines in profitability than are expected at the initial CUSTAimplementation.73Chapter SixSummary, Conclusions, and RecommendationsThe present work estimates the impact of the Canadian pre-1988 tariff structure bycalculating the net effect on prices through the loss of that protection occuring under (1) theinitial implementation of the Canadian-U.S. Tariff Agreement and (2) freer trade with thecomplete removal of all remaining non-tariff trade barriers. By means of an input-outputmodel, the net effects of any direct and indirect changes in commodity prices and thecalculation of effective protection allowed for the economy-wide use of these commoditiesas intermediate inputs into production. Through the incorporation of industry-specificparameters derived from an econometrically estimated pricing model, this study attemptsto examine the extent to which imperfect competition and mixed pricing behaviour due tomarket structure modifies the measurement of effective protection and the consequentimplications that the traditional assumption of tariff-limit pricing may have generally in theassessment of these rates and other related measurements of trade gains and resourceallocation effects.Again, it must be emphasized that, while higher rates of effective protection in particularindustries are generally positively correlated to increased investment and expansion in thoseindustries, actual income gains are dependent upon gross output and factor returns. Further,74for any given composition of output, factor shares may vary non-linearly with factor pricesand these distributive shares depend on this composition of output. Although the wineindustry, for example, is shown here to be rather strongly negatively affected by the CUSTApolicy structure, some efficient producers of the higher-quality wines remain profitable andhave expanded output and income since the agreement. Rates of effective protection are animperfect but useful index that goes far to summarize tariff structure when that structureis unevenly applied across intermediates and final goods. Unfortunately, its accuracy iscompromised by large changes in these various levels of protection and this effect in itselfis a common phenomenon in trade liberalization. Nevertheless, ERPs remain a good, easily-calculated indicator of relative average profitability among these industries and the pressurethat a tiered protective structure may bear upon these and the economy as a whole toreallocate its primary factors.The main results can be summarized as follows:1. The Canadian food processing sector may lose an output-weighted average of nearly6.7% of its value-added as a consequence of the initial CUSTA implementation, with thisbeing the net result of a 4.0% reduction in output prices and a 2.0% reduction in costs.Sixteen of these nineteen industries sustain negative effective protection under the 1988agreement owing largely to the imposition of significant levels of non-tariff barriers onagricultural inputs into their production as well as on some of their exported output.75Those industries with the most to lose include wineries, vegetable oil mills, and thetobacco, poultry, dairy and sugar processing industries. Those expected to improve do so toa very small degree. Red meat processing gains an estimated 1.4% in effective protectionlargely as a result of remaining foreign (U.S.) non-tariff barriers.2. The average net loss of protection (with respect to the base case pre-1988 tariffstructure) expected to occur in the processing sector with a move to free trade is about6.1%, a small improvement over the CUSTA scenario due mainly to the reduction of tariffson their primary agricultural inputs. Again, the red meat processing industry is virtually theonly industry expected to capture any sizeable gains and most other gains and losses arerather small. Wineries, breweries, vcgetable oil mills, and the tobacco processing inilustrydo much more poorly than average.3. The primary agricultural production sector continues to receive considerably moreprotection as a whole for much of its output than does that of food processing and is notexpected to experience a significant overall reduction in its surplus under the CUSTA.Wheat, oilseeds, dairy and raw tobacco are all likely to improve under the agreement whilefruit and vegetables, pigs and poultry register large predicted reductions in value-added.4. Moving to freer trade potentially results in a overall 20% reduction of value-addedin the primary food sector, which fact in itself indicates the high degree of negative effectiveprotection that presently bears upon the processing and downstream users. While wheat can76be expected to net significant improvements through enhanced export prices (there remainsa 6% foreign (U.S.) PSE on wheat under the CUSTA), cattle and small grains do onlyslightly better under free trade than under the agreement and the remaining primaryindustries are all likely to lose considerable shares of their value-adding product and surplus.This is particularly apparent among poultry, dairy, and pigs which, in this scenario, all losethe non-tariff protection that they currently enjoy under the present agreement.5. The standard Ricardian law-of-one-price assumption can be expected to significantlyoverstate gains and losses in value-added throughout that portion of the processing andmanufacturing economy that can be characterized as having scope for some degree ofimperfectly competitive pricing. Under thistraditionally competitiveassumption, reductionsin value-added in the food processing sector are about double those estimated by the moresophisticated mixed-pricing model for the base case CUSTA scenario. In the free tradescenario, the differences between the two models are less pronounced when taken as awhole but particular estimates under the standard assumption still tend to be of muchgreater magnitude. Proportional changes in export prices for this sector under the law of oneprice are double those found under mixed pricing. Nevertheless, the ranking of industriesaccording to the two protection rates is not radically altered through the adoption of onepricing assumption rather than the other. Industries dominated by regional characteristicsand protected naturally by high transport costs are among those most mis-specified by thestandard model. These include dairy and red meat processing, feed mills, bakeries, softdrinks, and breweries.776. The model is quite sensitive with respect to changes in the exchange rate, andestimates of effective protection and the rankings of many industries are strongly affected.The model essentially responds to the exchange rate as a change in scale between the pricesof tradeable intermediates and non-tradeable factors. With depreciation, for example,tradeables become relatively more expensive than the primary factors. Input substitution,however, and the degree to which different sectors and industries change relative to eachother become increasingly important and critical sources of bias as the model moves awayfrom the base case equilibrium. Expansion of the food processing sector, for example, wouldprobably depend more greatly on the expansion of the service sector while primaryagriculture would need to look more towards its commodity export markets.A1O% depreciation resuitsin a nearly 8% increase in value-added over the two foodsectors, while a similar decrease occurs under a 10% appreciation. These gains and losseshowever, are distributed unequally across the two: in examining the consequences of this inthe CUSTA mixed-price model, food processing gains considerably less than the othersectors with exchange rate depreciation and loses more than the others with appreciation.Under currency appreciation, prices of tradeables decline with respect to the prices ofdomestic resources and non-tradeable goods, and virtually every industry shows high levelsof negative effective protection due in part to the relatively higher prices for labour andcapital. To the degree that the manufacturing and service sectors realize the productivityimprovements that are anticipated for them, the resultant appreciation is likely to make iteven more relatively difficult for the two food sectors: import penetration and domesticcompetition would likely intensify.787. Because of the tariffs and barriers that remained in place with the implementationof the CUSTA, some processing industries were unfairly disadvantaged. While attemptingto moderate the 20% reduction in primary agriculture’s effective protection that wouldotherwise occur under free trade, the processing sector currently bears about 10% morepressure on its value-adding activity due to the structure implicit in the CUSTA. Vegetableoil mills, and poultry, dairy and fruit and vegetable processing are all industries that areadversely affected by these remaining barriers in the agreement.Related to the above is the question of the rents implicit in the CUSTA particularlywith respect to a number of the primary agricultural industries that especially continue tobenefit under the agreement: these include pigs, poultry, dairy, and fresh fruits andvegetables.-8. When the law of one price was assumed for all industries, measurement of effectiveprotection in the service sector was biased upwards by some 10% in the symmetricallystructured free trade scenario and by about 20% in the more complexly structured base caseCUSTA schedule. The mixed-price model estimated the degree of effective protection tothis sector as being about 1% under the actual 1988 agreement and 1.1% under free trade.On the basis of this result, therefore, we suggest that even in the case of perfectlycompetitive industries, it would appear that the standard pricing model may seriouslyoverestimate their effective protection and factor returns if they are sufficiently intensivedownstream users of differentiated intermediates produced by imperfectly competitiveindustries. As a principle, we may say that the importance of properly specifying actual79pricing behaviour becomes more critical as one considers more complex and assymetricalprotective and market structures and patterns of intermediate use. This is particularly truefor the processing and manufacturing sectors where imperfect competition is an importantand more prevailing consideration.9. The model could be improved in a number of respects beyond that of morecompletely disaggregating and specifying the manufacturing and service sectors. Within thetwo food sectors, some industries would be particularly useful to disaggregate further: amongthe primary agricultural commodities entering as inputs into processing, these include fluidand industrial milks, fresh fruits and vegetables, field crops, and the otherwise unallocatedagricultural imports; among the processing industries that would benefit±om more detail,perhaps the most important are red meats, dairy processing, feed and cereal mills, and fruitand vegetables. In addition, both the pricing model specification and the technology matrixcould be determined and updated to more accurately reflect Canadian conditions in 1988.80BibliographyAnderson, J. and S. Naya (1969), Substitution and two concepts of effective rate ofprotection, American Economic Review, vol.59, no.4, pp.607-12Balassa, B. (1965), Tariff protection in industrial countries: an evaluation, Journal ofPoliticalEconomy, vol.73, no.5, pp.573-594.Balassa, B. (1971a), Effective protection in developing countries, in Trade, Balance ofPayments, and Growth, eds. Bhagwati, et al, Amsterdam: North-Holland.Balassa, B. (1971b), The Structure of Protection in Developing Countries, Baltimore: JohnHopkins Press.Baldwin, J.R. and P.K. Gorecki (1986), The Role of Scale in Canada-US. ProductivityDifferences in the Manufacturing Sector (1970-1979), volume six of the MacDonaldCommission Report Research, Toronto: University of Toronto Press.Bhagwati, J. and T.N. Srinivasan (1973), The general equilibrium theory of effectiveprotection and resource allocation, Journal of International Economics, vol.3, no.3,pp.259-282.Brander, James A. (1981), Intra-industry trade in identical commodities, Journal ofInternational Economics, vol.11, pp.1-4.Brander, James A. and P. Krugman (1983), A reciprocal dumping model of internationaltrade, Journal of International Economics, vol.15, pp.313-321.Brown, Drusilla K. and Robert M. Stern (1988), Computable General Equilibrium Estimatesof the Gains from US. -Canadian Trade Liberalization, seminar discussion paper no.220, Ann Arbor: University of Michigan, Dept. of Economics.Bruno, M. (1973), Protection and tariff change under general equffibrium, Journal ofInternational Economics, vol.3, no.3, pp.205-2Corden, W.M. (1966), The structure of a tariff system and the effective protection rate,Journal of Political Economy, vol.74, no.3, pp.221-237.Corden, W.M. (1969), Effective protection rates in the general equilibrium model: ageometric note. Oxford Economic Papers, vol.21, no.2, pp.135-41Corden, W.M. (1971), The substitution problem in the theory of effective protection, Journalof International Economics, vol.1, no.1, pp.37-5.81Corden, W.M. (1971), The Theoiy of Protection, Oxford: Oxford University Press.Cox, David and Richard Harris (1985), Trade liberalization and industrial organization:some estimates for Canada, Journal of Political Economy, vol.93, no.1, pp.115-145.Department of Finance (1988), The Canada-US. Free Trade Agreement: an EconomicAssessment, a publivation prepared by the Fiscal and Economic Analysis Branch, copyno. 01-08-88.Dixit, A. and J.E. Stiglitz (1977), Monopolistic competition and optimum product diversity,American Economic Review, vol.67, no.3, pp.297-308Eastman, H.C. and S. Stykolt (1967), The Tariff and Competition in Canada, Toronto:Macmillan.Fuss, M.A. and V.K. Gupta (1981), A cost function approach to the estimation of minimumefficient scale, returns to scale, and sub-optimal capacity, European Economic Review,vol.15, no.1, pp.123-35Hamilton, Bob and John Whalley (1985), Geographically discriminatory trade arrangements,Review of Economics and Statistics, vol.67,, no.3, pp.446-455.Harris, Richard (1984), Applied general equilibrium analysis of small open economies andimperfect competition, American Economic Review, vol.74, no.5, pp.1016-32Harris, Richard with David Cox (1983), Trade, Industrial Policy, and CanadianManufacturing, Toronto: Ontario Research Council.Haziedine, T.J. (1980), Testing two models of pricing and protection with Canada/UnitedStates data, Journal of Industrial Economics, vol.29, no.2, pp.145-54Haziedine, T.J. (1985), Oligopolistic pricing with import competition and tariffs, a draft paperpresented to the Canadian Economics Association in April, 1985.Haziedine, T.J. (1988), Pricing and Protection: Eastman-Stykolt Reassessed, a paper presentedto the Canadian Economics Association Annual Meeting in Windsor, Canada in June3-5.Haziedine, T.J. and 0. Vaughan (1986), Canada/US. Free Trade: the Competitiveness andProspects of the B.C. Food Processing Industry, a report for the Policy Analysis andCo-ordination Unit; B.C. Ministry of Agriculture and Fisheries, (December).82Helpman, Elhanan (1981), International trade in the presence of product differentiation,economies of scale and monopolistic competition: a Chamberlin-Heckscher-Ohlinapproach, Journal of International Ecomnomics, vol.11, pp.305-340.Jones, R.W. (1971), Effective protection and substitution, Journal ofInternationalEconomics,vol.1, no.1, pp.59-8 1.Karikari, John A. (1988), International competitiveness and industry pricing in Canadianmanufacturing, Canadian Journal of Economics, vol.21, no.2, pp.410-26Khang, Chulsoon (1973), Factor substitution in the theory of effective protection: a generalequilibrium analysis, Journal of International Economics, vol.3, no.3, pp.227-44Krugman, Paul R. (1979), Increasing returns, monopolistic competition, and internationaltrade, Journal of International Economics, vol.9, no.3, pp.469-479.Krugman, Paul R. (1979), Scale economies, product differentiation, and the pattern of trade,American Economic Review, vol.70, no.5, pp.950-959.Krugman, Paul R. (1981), Intraindustry specialization and the gains from trade, Journal ofPolitical Economy, vol.89, no.5, pp.959-973.Lancaster, Kelvin J. (1980), A theory of intra-industry trade under perfect monopolisticcompetition, Journal of International Economics, vol.10, no.2, pp.151-175.Leith, J.C. (1968), Substitution and supply elasticities in calculating the effective protectionrate, Quarterly Journal of Economics, vol.82, no.4, pp.588-601Maundu Maingi (1990), The Performance of the Canadian Food, Beverages and TobaccoProcessing Industries: an extension of the profit-cost maigjn model to a pricing model,an M.Sc. thesis prepared for the Department of Agricultural Economics, Universityof British Columbia.Moschini and Meilke (1990), Tariffication with Imperfect Competitionn the Case of theCanadian Chicken Market, minieo.Ohyama, M. and K. Suzuki (1980), Interindustry flows, non-traded intermediate goods andthe theory of effective protection: a general equilibrium analysis, Journal ofInternational Economics, vol.10, no.3, pp.567-78Ramaswami, V.K. and T.N. Srinivasan (1971), Tariff structure and resource allocation in thepresence of factor substitution, in Trade, Balance of Payments and Growth, ed. 3.Bhagwati et al., Amsterdam: North Holland.83Robidoux, Benoit and John Lester (1988), Econometric Estimates of Scale Economies inCanadian Manufacturing, working paper no. 88-4, Ottawa: Department of Finance.Spence, Michael E. (1976), Product selection, fixed costs, and monopolistic competition,Review of Economic Studies, vol.43, pp.217-36.Tan, A.H.H. (1970), Differential tariffs, negative value-added and the theory of effectiveprotection, American Economic Review, vol.60, no.1, pp.107-16Thomassin, Paul and Allan Andison (1987), Agriculture Canada’s Input-Output Model PartI: Disaggregation of the Agriculture Sector, a Policy Branch working paper 6/87.Uekawa, Y. (1979), The theory of effective protection resource allocation and the StolperSamuelson theorem, Journal of International Economics, vol.9, no.1, pp.151-7Uekawa, Y. (1980), A non-traded input, substitution and effective protection, Journal ofInternational Economics, vol.10, no.4, pp.579-92U.S. Department of Agriculture (1988), Estimates of Producer and Consumer SubsidyEquivalents, ERS Staff Report AGEC880127; Economic Research Service.Wigle, Randall (1988), General equilibrium evaluation of Canada-U.S. trade liberalizationin a global context, Canadian Journal of Economics, vol.21, no.3, pp.539-64Woodland, A.D. (1977), Joint outputs, intermediate inputs and international trade theory,International Economic Review, vol.18, no.3, pp.517-33Woodland, A.D. (1982), International Trade and Resource Allocation, Amsterdam: NorthHolland.84APPENDIX A.I OUTLINE OF SUBSUMED COMMODITIESINDUSTRY/COMMODITY SIC 1980 N—LEVEL NUMBER AND COMMODITY GROUPI CATTLE AND CALVES 0112 00100 CATTLE AND CALVES2 DAIRY 0111 00900 HILK,WHOLE,FLUID,UNPROCESSED3 PIGS 0113 00300 HOGS4 POULTRY AND EGGS 0114 00400 POULTRY01000 EGGS IN THE SHELLS WHEAT 0131 00700 UHEAT,UNMILLED6 SMALl GRAINS 0132 00800 BARLEV,DATS,RYE,CORN,SRA]N,NES7 DILSEEDS 0133 01800 OIL SEEDS,NUTS AND KERNELS8 FRUIT, VEGETABLES 0151 01300 FRIJITS,FRESH, EX.TROPICA1.0152 01400 VEBETABLES,FRESH9 TOBACCO 0137 02000 TOBACCO,RAW10 MISC AGRICULTURE 0115 00200 SHEEP AND LAMBS0120 00500 OTHER LIVE ANIMALS01100 HONEY AND BEESWAX0130 NES 01200 NUTS,EDIBLE,NOT SHELLED01500 HAY,FORAGE,AND STRAW01600 SEEDS El. OIL AND SEED GRADES0160 01700 NURSERY STOCK RELATED IAT.01900 HOPS INC. IUPUI.IN02100 MINK SKINS,RANCH UNDRESSED02200 WOOL IN GREASE0210-0230 03300 SERY. INCIDENTAL TO AGR.&FOREST58800 COTTON RAW & SEMI—PROCESSED58900 NATURAL RUBBER & ALLIED GUNS59000 SUGAR , RAW59100 COCOA BEANS,UNROASTED59200 GREEN COFFEE59300 TROPICAL FRUIT-11 RED MEAT PROCESSING 1011 05200 BEEF,VEAL,MUTT&PORK,FRESH&FROZ05300 HORSE MEAT FRESH,CHILIED,FROZE05400 MEAT,CURED05500 NEAT PREP. COOKED NOT CANNED05600 MEAT PREP. CANNED05700 ANIMAL OILS & FATS & LARD05900 SAUSAGE CASINGS, NATURAL&SYNTH.06000 PRIMARY TANKAGE06100 FEEDS OF ANIMAL ORIGIN NES06200 HIDES AND SKINS,RAW,NES06300 ANIMAL MAT.FOR DRUGS & PERFUME- 06400 CUSTOM WORK MEAT & FOOD12 POULTRY PROCESSING 1012 06500 POULTRY,FRESH,CHILLED,FROZEN06600 POULTRY, CANNED13 FISH PROCESSING 1020 07500 FISH PRODUCTS14 FRUIT, VEGETABLE PROCESSING 1030 07600 FRUIT,BERRIES,DRIED,CRYSTALIZE07700 FRUITS & PREPARATIONS CANNED07800 VESET.FROZEN,DRIED & PRESERVED07900 VESETABLES&PREPARATIONS CANNED08000 SOUPS CANNED08100 INFANT&JUNIOR FOODS,CANMED08200 PICKLES,RELISHES,OTHER SAUCES08300 VINEGAR08400 OTHER FOOD PREPARATIONS10900 SOUPS, DRIED&SOUP NIIES&BASES85APPENDIX 0.2 OUTLINE OF SUBSUIED CONNODITIESINDUSTRY!COH$ODITY SIC 1980 -LEYEL NUNBER AND CONNODITY GROUP15 DAIRY PROCESSINS 1040 06700 I1ILX,NOLE,FLUID,PROCESSED06800 CREM,FRESH06900 BUTTER07000 CHEESE,CHEODAR PROCESSED07100 MILK EVAPORATED07200 ICE CREAM07300 OTHER DAIRY PRODUCTS16 FLOUR, MEAL, CEREAL MILLING 1051,1052 09000 UHEAT FLOUR09100 HEAUROUR OF OTHER CEREALSVE09200 BREAKFAST CEREAL PRODUCTS10800 PREPARED CAKE 0 SIMILAR MIXES17 FEED MILLING 1053 08500 PRIMARY OR CONCENTRATED FEEDS08600 FEED FOR COMMERCIAl. LIVESTOCK08700 FEEDS, GRAIN ORIGIN, N.E.S.08800 FEEDS OF VEGETABLE ORIGIN NES08900 PET FEEDS18 VEGETABLE OIL MILLING 1060 10300 OILSEED,KEAL & CAKE10400 YES. OILS & FATS, CRUDE19 BISCUIT INDUSTRY 1071 09300 BISCUITS20 BREAD, BAKERY PRODUCTS 1072 09400 BREAD & ROLLS09500 OTHER BAKERY PRODUCTS21 CANE, BEET SUSA REF INERI 1081 10000 BEET PUT.?10100 SUGAR10200 MOLASSES,SUSAR REFINERY PROD.22 SUSAR CONFECTIONARY 1082,1083 09600 COCOA & CHOCOLATE03800 CHOCOLATE CONFECTIONERY09900 OTHER CONFECTIONERY23 TEA, COFFEE BEVERAGES 1091 11000 COFFEE,ROASTED,GROUND,PREPARED11100 TEA24 MISC FOOD PRODUCTS 1092-1099 05800 NAR6ERINE,SHORTENING&LIKE PROD07400 MUSTARD MAYONNAISE09700 VNUTS,KERNELS & SEEDS PREPARED10600 MALT,MALT FLGUR&UHEAT STARCH10700 MAPLE SUGAR&SYRUP11200 POTATO CNIPS&SIMILAR PRODUCTS11300 MISC.F000NES25 SOFT DRINK PRODUCTS 1110 11400 SOFTDRINK CONCENTRATES&SYRUPS11500 CARBONATED BEV.,SOFT DRINKS26 DISTILLERY PRODUCTS 1120 11600 AI.COHOLIC BEVERA6ES DISTILLED27 BRENERY PRODUCTS 1130 11900 ALE BEER,STOUT & PORTER28 NINE INDUSTRY 1140 12000 NINES29 TOBACCO PRODUCTS 1210-1220 12100 TOBACCO PROCESSED,UNNANTJFACT.12200 CIGARETTES12300 TOBACCO NFS EX.CIGARETTES86APPENDIX A.3 OUTLINE OF SUBSUMED COMMODITIESINDUSTRY/COMMODITY U—LEVEl. MJNBER AND COMMODITY GROUP30 OTHER MANUFACTURES 02400 1.085 AND BOLTS 31300 COLLAPSIBLE TUBES,METAL02500 POLES,PIT PROPS FENCE—POSTS ET 31400 TRACTORS, FARM & GARDEN TYPE02600 PULPWOOD 31500 OTHER AGRICULTURAL MACHINERY02700 OTHER CRUDE 111)00 MATERIAlS 31600 MECHANICAL POWER TRANS.EQUIP.02800 CUSTOM FORESTRY 31700 PUMPS,COMPRESSORS&BLOWERS ETC.02900 FISH LANDINGS 31800 CONVEYORS,EsCAL,ELEVHo1ST MAC03000 HUNTING & TRAPPING PRODUCTS 31900 IND.TRUCKS,TRACTORS, TRAILERS E03200 6010 & ALLOYS IN PRIMARY FORM 32000 FANS,AIR CIRCULATORS&AIR UNITS03300 RADIO-ACTIVE ORESCONCENTRATES 32100 PKG.MACH,LUB.ED&OTH.MISC.MACH.• 03400 IRON ORES & CONCENTRATES 32200 INDUSTRIAL FURNACES,KILNS&OVEN03500 BAUXITE + ALUMINA 32300 MACH.IND.SPECIFIED&SPECIAI. PUR03600 METAL ORES + CONCENTRATES N.E. 32400 POWER DRIVEN HAND TOOLS03700 COAL 32500 METAL END PRODUCTS, NES03800 CRUDE MINERAL OILS 32600 REFRIB&AIR CON.ED,EX.HOUSEHOLD03900 NATURAL GAS 32700 SCALES & BALANCES04100 SULPHUR,CRUDE & REFINED 32800 VENDING MACHINES04200 ASBESTOS, UNMF6. ,CRUDE& FIBROUS 32900 OFFICE MACHINES AND EDUIPMENT04300 GYPSUM 33000 AIRCRAFT, ALL TYPES04400 SALT 33100 AIRCRAFT ENGINES04500 PEATMOSS 33200 SPECIALIZED AIRCRAFT EDUIPHENT04600 CLAI&OTNER CRUDE REFRACTORY NA 33300 IIODIFICATIONS,CONVERSJONS,SERV04700 NATURAL ABRASIVES&INDUST.DIAMO 33400 PASSENGER AUTOMOBILES & CHASSI04800 CRUDE MINERAL NES 33500 TRUCKS, CHASSIS, TRACTORS, COM04900 SAND AND GRAVEL 33600 BUSES AND CHASSIS05000 STONE,CRUDE 33700 MILITARY MOTOR YEN, MOTORCYCLE05100 SERVICES INCIDENTAL TO MINING 33800 MOBILE HOMES10500 NITROGEX FUNCTION CONPC1UND ML 33300 0TH. TRA1LERS?SENT-TRALLERS, CON-11700 ALCOHOL, NATURAL, ETHYL 34000 BODIES AND CABS FOR TRUCKS11800 BREWERS’ &DISTILLERS’ GRAINS 34100 MOTOR VEHICLE ENGINES AND PART12400 FOOTWEAR,RUBBER AND PLASTIC 34200 AUXILIARY ELECTRIC EDUIPMENT12500 TIRES & TUBES,PASSENGER CARS 34300 MOTOR YEN. ACCESS, PARTS&ASSEM12600 TIRES & TUBES,TRUCKS & BUSES 34400 AUTOMOTIVE HARDWARE, EX.SPRINS12700 TIRES I TIJBES,N.E.S. 34500 LOCOHOTIVES,CARS&TENDERS,RLY.S12800 TIRES, RETREADING 34600 SELF-PROPEL CARS12900 RECLAIMED RUBBER 34700 PARTS&ACCESS.FOR RLY,ROLL.STOC13000 RUBBER BELTS & COATED FABRICS 34800 SHIPS&BOATS,MILITARY&COM$ERCIA13100 RUBBER SHEETING SHOE STOCK ETC 34900 SUB-ASSEN8LIES,PARTS,ETC.SHIPS13200 HOSE & TUBING,MAINLY RUBBER 35000 SHIP REPAIRS13300 RUBBER WASTE I SCRAP 35100 SNOWHOBILESVcMISC.NOK-MOTOR YEN13400 RUBBER END PRODUCTS NES 35200 PLEASURE & SPORTING CRAFT13500 PLASTIC PIPE FITTINGS & SHEET 35300 SMALL ELEC.APPLIANCES, DOMESTIC13600 PLASTIC CONTAINERS&BOTTLE CAPS 35400 SPACE HEATER,HEATING STOVES El13700 PREFAB. BLDGS&STRUCTURES MES 35500 REFRI6,FREEZERS&COMB. DOMESTIC13800 PLASTIC HOSE,PAILSEND PROD.NE 35600 GAS RAN6ES&ELEC.STOYES,00MESTI13900 LEATHER 35700 T.V.,RADIO,RECORD PLAYERS14000 FOOTWEAR EX.RUBBER & PLASTIC 35800 TEL&TELEB.LINE APPARATUS&EQUIP14100 LEATHER GLOVES&MITTENS El SPOR 35900 RADIO&TV BROADCASTING&TRANS EQ14200 LEATHER BELTIN6,SHOE STOCK 36000 RADAR EQUIP. & RELATED DEVICES14300 LUGGAGE 36100 ELEC. TUBES&SENI-CDNDUCTORS ETC14400 LEATHER HANDBAGS,UALLETS ETC. 36200 ELECTRONIC EQUIPMENT COMPONENT14500 YARN, COTTON 36300 INTERIOR SISMAL,ALARN&CLOCK SY14600 YARNS MII&BLENDED&COTTON WASTE 36400 POLE LIME HARDWARE14700 FABRICS, BROAD WOVEN OF COTTON 36500 WELDING MACHINERY & EQUIPMENT14800 TIRE CORD & TIRE FABRICS 36600 EN6INES,MARINE,ELECTRIC TURBIN14900 METS & NETTING 36700 TRANSFORMERS&CONVERTERS EX.T&T15000 BLANKETS,BEDSHEETS,TOWELS&CLDT 36800 ELEC. EDUIP. INDUSTRIAL, NES87APPENDIX A.4 OUTLINE OF SUBSUMED COMMODITIESINDUSTRY/CONMODITY U-LEVEL NUMBER AND COMMODITY GROUP30 OTHER MANUFACTURES 15100 YARN OF WOOL AND HAIR 36900 BATTERIES15200 FA8RICS,BROADWOVEN,WO0L,HA!RM 37000 WIRE AND CABLE, INSULATED15300 PAPERNAKERS’ FELTS 37100 ALUM. WIRE&CASLE,NOT INSULATED15400 MAN MADE FIBRES 37200 ENCLOSED SAFETY SWITCHES ETC.15500 POLYAMIDE RESINS (NYLON) 37300 ELEC.LI6HT BULBS&TIJBES, ETC15600 YARNS, SILK, FIBRE6LASS 37400 ELECTRIC LIGHTING FIXTURES ETC15700 TIRE YARNS 37500 CEMENT15800 FABRIC, WOVEN, TEXTILE FIBRES 37600 LINE15900 FABRICS,BROAD UOVEN,MIXBLENDS 37100 CONCRETE BASIC PRODUCTS16000 RA6GA WASTE, COTTONTEXTILE NAT. 37800 SAND LINE BRICKS AND BLOCKS16100 UOOUFINE ANIMAL HAIR,SPINNING 37900 READY—NIX CONCRETE16200 THREAD,OF COTTON FIBRES 38000 BRICKS AND TILES, CLAY16300 THREAD, OF NAN-MADE FIBRES 38100 INSULATORSELEC.FITTIN6S,PORCE16400 YARN&THREAD, OTHER YES. FIBRES 38200 PLUNB.EQ, VITREOUS CHINA,& ETC16500 BALER AND BINDER TWINE 38300 REFRACTORIES16600 OTHER CORDAGE, TWINE & ROPE 38400 NATURAL STONE BASIC PROD,STRUC16700 NARROW FABRICS 3800 STONE,CLAY&CONCRETE END PROD.N16800 LACE FABRICS,BUBBINET & NET 38600 PLASTERS&OTH.GYPSU$ BASIC PROD16900 FELT, CARPET CUSHION 38700 MIN.WOOL&THERNAL INSUL.HAT.NES17000 CARPETINGIFABRIC RUGS,HATS, ETC 38800 ASBESTOS PRODUCTS17100 TEXTILE DYEING & FINISHING SER 38900 NON—METALLIC NIN.BASIC PROD.NE17200 AWNINGS, OF CLOTH & PLASTIC 39000 GLASS, PLATE, SHEET, WOOL17300 TENTS,HANNOCKS,SLEEP BAGS&SAIL 39100 GLASS CONTAINERS17400 TARPAULINS & OTHER COVERS 39200 GLASS TABLEWRE&HOUSEWRE,END&NE17500 TEXTILE CONTAINERS 39300 ABRASIVE BASIC PRODUCTS17600 VEGETABLE TEXTILE FIBRES NES 39400 AVIATION GASOLINE17700 MISC.TE1TILE FALMAT.LNC. RAGS 3900 HO-TOR GASOLINE17800 HOUSEHOLD TEXTILES, NES 39600 FUEl. OIL17900 LACES AND TEXTILE PROD. N.E.S. 39700 LUBRICATING OILS AND GREASES18000 HOSIERY 39800 BENZENE, TOLUENE AND IYLENE18100 FABRICS,KNITTEDNETTED,ELASTIC 39900 BUTANE,PROPANEIOTH.LIQ.PET.GAS18200 FABRICS, KNITTED, NES 40000 NAPHTHA18300 KNITTED WEAR 40100 ASPHALT AND COAL OILS, N.E.S.18400 CLOTHING 40200 PETROCHEMICAL FEED STOCK18500 APPAREL ACCESSORIES&OTHER MISC 40300 FERTILIZERS18600 FURS, DRESSED 40400 PLASTIC RESINSMAT.,NOT SHAPED18700 FUR PLATES, MATS AND LININGS 40500 FILN&SHEET, CELLULOSIC PLASTIC18800 FUR APPAREL 40600 ETHANOLAMINES18900 CUSTOM TAILORING 40700 ETHYLENE GLYCOL, MONO19000 PULPWOOD CHIPS 40800 PHARMACEUTICALS19100 LUMBER & TIMBER 40900 PAINTS & RELATED PRODUCTS19200 RAILWAY TIES 41000 YES. OILS,OTH.THAN CORN OIL,RE19300 WOOD WASTE 41100 GLYCERIN, REFINED19400 CUSTOM WOOD WORKING & MILLUORK 41200 DENTIFRICES, ALL KINDS19500 VENEER AND PLYWOOD 41300 SOAPS, OETERSENTS,CLEANING PROD19600 MILLUORK (WOODWORK) 41400 INDUSTRIAL CHEMICAL PREP. N.E.19700 WOOD FABRICATED NAT.FCR STRUCT 41500 TOILET PREPARATIONS & COSMETIC19800 PREFAB. BLDSS,WOOD 41600 CHLORINE19900 CONTAINERS,CLOSURES&WOOD PALLE 41700 OXYGEN20000 CASXETS,cOFFINS&OTNER MORT.SOO 41800 PHOSPHORUS20100 MISC. WOOD 41900 CHEMICAL ELEMENTS, NES20200 BARRELS & KEGS OF WOOD 42000 SULPHURIC ACID20300 WOOD END PRODUCTS,NES 42100 CARBON DIOXIDE (GAS AND DRY IC20400 HOUSEHOLD FURN.INCL.CANP&LAWN 42200 INORGANIC ACIOSIOXYGEN20500 OFFICE FURN&YISIBLE RECORD EDU 42300 AMMONIA, ANHYDROUS AND AQUA20600 SPECIAL PURPOSE FURNITURE 42400 CAUSTIC SODA (SOD.HYDROXIDE)DR88APPENDIX A.5 OUTLINE OF SUBSUMED COMMODITIESINDLISTRYICOMNODITY U—LEVEL JN8ER AND COMMODITY GROUP30 OTHER MANUFACTURES 20700 MISC. FURNITURE AND FIXTURES 42500 CALCIUM CHLORIDE20800 PORTABLE LAMPS BESIDENTIAL TYP 42600 SODIUM CHLORATE20900 PULP 42700 ALUMINUM SULPHATE21000 NEWSPRINT PAPER 42800 SODIUM PHOSPHATES21100 OTHER PAPER FOR PRINTING 42900 SODIUM CARBONATE (SODA ASH)21200 FINE PAPER 43000 SODIUM CYANIDE21300 TISSUE & SANITARY PAPER 43100 SODIUM SILICATE21400 WRAPPING PAPER 43200 METALLIC SALTS&PEROXYSALTS,WES21500 PAPER BOARD 43300 PNOT06RAPHIC&INORGANIC CHELN.21600 BLDG.PAPER 43400 ETHYLENE21700 TOWELS, NAPKINS & TOILET PAPER 43500 BUTYLENES21800 VANILLIN 43600 BUTADIENE21900 MISC. INO.PAPER MAT;BY PROO&WAS 43700 ACETYLENE22000 TILES, VINYL—ASBESTOS 43800 STYRENE MONOMER22100 PAPER CARTONS,BAGS,CANS&BOTTLE 43900 CARBON TETRACHLORIDE22200 CONVERTED PAPER,SUM, WAX OR PRI 44000 VINYLCHLORIDE MONOMER22300 CONVERTED ALUMINUM FOIL 44100 TRICHLOROETHYLENE22400 FACIAL TISSUES, ISANITARY NAPKI 44200 PERCHLOROETHYLENE22500 PAPER CONTAINERS,NES 44300 FLUORINATED HALOGEN HYDROCARBO22600 OFFICE AND STATIONERY SUPPLIES 44400 HYDROCARBONS&THEIR DERIVATIVES22700 PAPER END PRODUCTS 44500 METHYL ALCOHOL22800 NEWSPAPERS,MAGAZINES&PERIODICA 44600 PROPYL AND ISOPROPYL ALCOHOLS22900 BOOKS,PANPHLETS,NAPS&PICTURES 44700 BUTYL AND ISOBUTYL ALCOHOLS23000 BANXNOTES,BONDS,DRAFTS ETC 44800 PENTAERYTHRITOL23100 OTHER PRINTED MATTER 44300 ALCOHOLS AND THEIR DERIVATIVES23200 ADYERTISING,PRINT MEDIA 45000 PHENOL23300 SRECTAI.flED PLIBLISHLNG SERVICE 45100 P44EN0LS,PHEN.ALC0NOLSDERIVATV23400 PRINTING PLATES,SET TYPE ETC. 45200 ETHERS,ALCOHOL PEROXIDES,ETC23500 FERRO-ALLOYS 45300 METYL-ETHYL, ALDEHYDE-FUNCT IONS23600 IRON, STEEL INGOTS 45400 ACETONE23700 STEEL BLOOMS,BILLETS & SLABS 45500 ACETIC ACID23800 STEEL CASTINGS 45600 ACETIC ANHYDRIDE23900 STEEL BARS AND RODS 45700 ADIPIC ACID24000 STEEL PLATES, NOT FABRICATED 45800 CITRIC ACIDS24100 CARBON STEEL SHEETS NOT COATED 45900 STEARIC AND ORGANIC ACIDS24200 TINPLATE 46000 HEXANETHYLENEDIAMINE24300 GALVANIZED STEEL SHEET & STRIP 46100 SODIUM GLUTAMATE, MONO24400 RAILS&RLY TRACK MATERIALS, STEE 46200 DICYANDIANIDE24500 COAL TAR 46300 ORGANO-INORGANIC COMPOUNDS ETC24600 NAT.&SYN.SRAPHITE&CARBON PROD. 46400 ORGANIC CHEMICALS, NES24700 MECHANICAL STEEL TUBING 46500 TITANIUM DIOXIDE24800 OIL COUNTRY GOODS 46600 BLACK, ACETYLENE AND CARBON24300 LINE PIPE,TRANS.NAT.GAS & OIL 46700 PIGMENTS, LAKES & TONERS,PROPE25000 STEEL PIPES & TUBES NES 46800 IRON OXIDES25100 GRINDING BALLS,INSOT NOUI.DS ET 46300 FERTILIZER CHEMICALS25200 CAST&UROU6HT IRON PIPE&FITTINS 47000 SYNThETIC RUBBER25300 NICKEL IN PRIMARY FORMS 47100 ANTIFREEZE COMPOUNDS25400 COPPER&COPPER ALLOYS,PRIME.FOR 47200 ADDITIVES FOR MINERAL OILS,NES25500 LEAD,PRINARY FORMS 47300 GLYCERINE, CRUDE25600 ZINC&ZINC ALLOYS PRIMARY FORMS 47400 RUBBER&PLASTICS COMPOUNDING AG25700 ALUMINUM&ALUNINU$ ALLOYS PRIME 47500 EXPLOSIVES, FUSES AND CAPS25800 TIN&TIN ALLOYS PRIMARY FORMS 47600 AMMUNITION, NOW-MILITARY25900 PRECIOUS NETAL&ALLOYS PRIME.FO 47700 AMMUNITION & ORDNANCE, NILITAR26000 OTH.NON-FERROUS BASE METALS 47800 PYROTECHNIC ARTICLES & FIREVOR26100 ALUMINUM FLUORIDES&SCDIUM ALUM 47900 CRUDE YES. MATERIALS & EXTRACT26200 INOR6ANIC BASES&MET.OXIDES,NES 45000 PMTHRLIC ANHYDRIDE89APPENDIX A.6 OUTLINE OF SUBSUMED COMMODITIESINDUSTRY/COMMODITY W—LEVEI. NUMBER AND COMMODITY BROUP31) OTHER MANUFACTURES 26300 SCRAPIIASTE MATERIALS NES 48100 ASRICULTURAL CHEMICALS26400 ALUM1NUMALUNINUM ALLOYS, CAST 48200 ADHESIVES26500 COPPER PROD.CAST,ROL.LED&EXTRUD 48300 AUTOMOTIVE CHEN. El. ANTIFREEZ26600 COPPER ALLOY PROD.CAST,ROLLIEI 48400 CONCRETE ADDITIVES26700 LADLEAD ALLOY PROJD.CAST,R&E 4800 BOILER CNEHICALS26800 NICKEL&NICKEL ALLOY FALNATERI 48600 COMPOUND CATALYSTS26300 TIN&TIM ALLOY FAD. MATERIALS 48700 METAL WORXING COMPOUNDS27000 ZINC DIE CASTIN6OTh. ZINC MAT. 48800 PRINTINS AND OTHER INKS27100 SOLDERS INC.BLOCX,RODS,WIRE,ET 48900 TEXTILE SPECIALTY CHEMICALS27200 PIATES, STEEL, FABRICATED 49000 PDLISHES,WAIES,COMPOUNDS ETC27300 TANKS 49100 WAXES,ANINAL 0 VEGETABLE, OTHE27400 POWER BOILERS 49200 ESSENTIAL OILS, NATURAL OR SYN27500 BOILERS, MARINE TYPE 43300 TANNINS MATERIALS AND DYESTUFF27600 BEAMS AND OTHER STRUCT. STEEL 49400 FATS AND CHEMICAL. MIXTURES27700 SCAFFOLDING EQUIP., DENQUNTABI. 49500 EMBALMING CHEN.& PREPARATIONS27800 PREFAB.BLD6S&STRUCT. ,MAINLY ME 49600 MATCHES27900 METAL PRODUCTS NES 43700 AIRCRAFThNAUTICAL INSTRUNENIS28000 STEEL SHEET&STRIP COATED DR FA 49800 LAB&SCIENTIFIC APPARATUS ETC23100 CULVERT PIPE, CORRUGATED METAL 49900 NISC.NEASURE&CONTRQL INSTRUMEN28200 METAL BASIC PROD.&RANGE BOILER 50000 HEDICAL&RELATED INSTRUMENTS El28300 METAL PIPES,FITTINGS k SIDINGS 50100 IND.NILITARY&CIVIL DEF.SAFETY28400 METAL AWNINGS,ASH CANS,PAILS E 50200 WATCHES,CLOCKS,CHRONOMETERS ET28500 KITCHEN UTENSILS 50300 PHOTOGRAPHIC EQ&SU?PL.INCL.FIL28600 CONTAINERS&BOTTLE CAPS OF META 50400 JEWELRY,FINDINSS,MET.&BEM STON28700 WIRE & WIRE ROPE, OF STEEL 50500 PLATED&SILVERWARE,CUTLERY, ETC28800 WIRE FENCINS,SCREENING&NETTIN6 50600 BROOMS,BRUSHES,MOPS&OTH.CLEAN.28900 CHAIN,EX.AUTO TIRE&POWER TRANS 50700 BICYCLES,CHILDREN’S YEL&PARTS29000 RODS,WXRE&ELECTRODES,WELDIN6 50800 SPORTING,FISHIN6&HUNTIN6 EQUIP29100 SPRINGS FOR UPHQLSTERY&MISC.VE 50900 TOYS AND GAME SETS29200 BULTS,NUTS,SCREWS,WASHERS ETC. 51000 FABRICS, INPRES.EX.RUBBER-COATE29300 BUILDERS’ HARDWARE 51100 TILING, RUBBER, PLASTIC29400 FITTINSS,FURN.CABINETS&CASKETS 51200 ADVERTISING GOODS29500 BASIC HARDWARE,NES 51300 SHADES&BLINDS29600 CUTTINS&FORMINB TOOLS 51400 FUR DRESSING & DYEING SERVICES29700 MEASURINS,EDGIHG,MECHANIC’S TO 51500 CUSTOM WORK, MISCELLANEOUS29800 SCISSORS,RAZQR BIADES,IND.CUTL 51600 ICE29900 DOMESTIC EQUIPMENT, NES 51700 ANIMAL HAIR,FEATHERS,QIJILLS,ET30000 HEATING EQ,HOT WATER&STEAM ETC 51800 MISC.FAB.MAT.INCL.BRISTLES ETC30100 HEATING EQ,WARM AIR ELPIPES&E 51900 BUTTONS,NEEDLES,PINS&MISC.NOTI30200 UNIT&WATER TANK HEATERS NON—El. 52000 PHONO RECORDS AND ARTIST MATER30300 FUEL BURNING EQUIPMENT 52100 HOUSEHOLD ORNAMENTAL OBJECTS&A30400 CON. APPLIANCES, COOK&WARNING FU 52200 REPAIR CONSTRUCTION30500 CUSTOM METAL WORKING 52300 RESIDENTIAL CONSTRUCTION30600 FORGINGS OF CARBON&ALLOY STEEL 52400 NON—RESIDENTIAL CONSTRUCTION30700 VALVES 52500 ROAD HIGHWAY AIRSTRIP CONST.30800 PIPE FITTINGS,NOT IRON & STEEL 52600 GAS AND OIL FACILITY CONST.30901) GAS METERS AND WATER METERS 52700 DAMS AND IRRIGATION PROJECTS31000 FIRE FI6}IT&TRAFFIC CONTROL LOU 52800 RAILWAY TELEPHONE TELEGRAPH CO31100 TAXILPARK METERS,BL.OCKS&LADDER 52900 OTHER ENGINEERING CONSTRUCTION31200 FIREARMS & MILITARY HARDWARE90APPENDIX A.7 OUTLINE OF SUBSUMED COMMODITIESINDUSTRYICONMODITY W—LEVEI. NUMBER AND COMMODITY GROUP31 SERVICES 53000 AIR TRANSPORTATION 56000 GOYT.ROYALTIES ON NAT. RESOURC53100 DINER TRANSPORTATION 56100 EDUCATION SERVICES53200 SERV. INCIDENTAL TO TRANSPORT N 56200 HOSPITAL SERVICES53300 WATER TRANSPORTATION 56300 HEALTH SERVICES53400 SERY. INCIDENTAL TO WATER TRANS 56400 NOTION PICTURE ENTERTAINMENT53500 RAILWAY TRANSPORTATION 56500 OTHER RECREATIONAL SERVICES53600 TRUCK TRANSPORTATION 56600 SERVICES TO BUSINESS NANA6EMEN53700 BUS TRANSPORT, INTERURBAN RURA 56700 ADVERTISING SERVICES53800 URBAN TRANSIT 56800 LAUNDRY,CLEANIM66PRESSING SERV53900 TAXICAB TRANSPORTATION 56900 ACCOMMODATION SERVICES54000 PIPELINE TRANSPORTATION 57000 MEALS54100 HIGHWAY AND BRIDGE MAINTENANCE 57100 SERV.MARG.ON ALCOHOLIC BEVERAG54200 STORAGE 57200 PERSONAL SERVICES54300 RADIO & TELEVISION BROADCASTIN 57300 PHOTOGRAPHIC SERVICES54400 TELEPHONE TELEGRAPH 57400 SERVICES TO BLD6S. & DWELLINGS54500 POSTAL SERVICES 57500 RENTAL DATA PROCESSING EQUIP.54600 ELECTRIC POWER 57600 OTHER SERY.TO BUSINESSES&PERSO54700 GAS DISTRIBUTION 57700 RENTAL OF AUTOMOBILES ti TRUCKS54800 COKE 57800 TRADE ASSOCIATION DUES54900 WATER AND OTHER UTILITIES 57900 RENTAL AD MACH&EQ.INCL.CONST.M55000 WHOLESALING MARGINS 58000 SPARE PARTSNAINT.SUPPL.NACHDE55100 REPAIR SERVICE 58100 OFFICE SUPPLIES55200 RENTAL OF OFFICE EQUIPMENT 58200 CAFETERIA SUPPLIES55300 RETAILING MARGINS 58300 TRANSPORTATION MARGINS55400 IMPUTED SERVICE, BANKS. 58400 LABORATORY EQUIP. AND SUPPLIES55500 0TH REAL EST (NON—RENTThFIN.SE 58500 TRAVELLING AND ENTERTAINMENT55600 INSURANCE ? W.C.B. 58600 ADVERTISING & PROMOTION- 55700 IMPUTED RENT OWNER OCPO. IIWEL. 59400 UNALLOCATED IMPORTS & EXPORTS55800 CASH RESIDENTIAL RENT 59500 GOVERNMENT GOODS SERVICES55900 OTHER RENT 60300 ALL OTHER21 PRIMARY RESOURCES 59600 COMMODITY INDIRECT TAXESZ2 59700 SUBSIDIESZ3 59800 OTHER INDIRECT TAXES14 59900 WAGES AND SALARIES25 60000 SUPPLEMENTARY LABOUR INCOMER 60100 NET INCOME UNINCORP BUSINESS60200 OTHER OPERATING SURPLUS91APPENDIX 81: COMMODITY INPUTS (1986 SNILLIONS) AND CENSUS DISTRIBUTION KEYS1001002003003 004004007005008006010007014008015009 0160010 0180011 03700120390013 06100140640015 085001608600170870018 0880019089002010000211030022 1270023 1650024 1660025 2870026 28800273140028 3150029 33300303760031 3940032 3950033396003439700353990036 40300374080038 4230039 4690040 471004148100424830043522004454600455490046 5510047555004855600495590050 5750051577005257900535800054 59800555990056 600005760200CATTLE AND CALVESHOGSPOULTRYWHEAT,UNMILLEDBARLEY, OATS, RYE,CORN, GRAIN,NESEGGS IN THE SHELLVEGETABLES, FRESHHAY,FORA6E, AND STROllSEEDS EL OIL AND SEED GRADESOIL SEEDS,NUTS AND KERNELSCOALNATURAL 605FEEDS OF ANIMAL ORIGIN )IESCUSTOM WORK NEAT FOODPRIMARY OR CONCENTRATED FEEDSFEED FOR COMMERCIAL LIVESTOCKFEEDS, GRAIN ORIGIN, N.E.S.FEEDS OF VEGETABLE ORIGIN NESPET FEEDSBEET PULPOILSEED,$EAL & CAKETIRES & TU9ES,N.E.S.BALER AND BINDER TWINEOTHER CORDAGE, TWINE & ROPEWIRE & WIRE ROPE, OF STEELWIRE FENCINS,SCREENING&NETTINGTRACTORS, FARE & BAROEIL TYPEOTHER AGRICULTURAL MACHINERYNODIFICATIONS,CONVERSIONS, SERYLINEAVIATION GASOLINEMOTOR GASOLINEFUEL OILLUBRICATING OILS AND GREASESBUTANE, PROPANE&OTN.L10.PET.60SFERTILIZERSPHARMACEUTICALSAMMONIA, ANHYDROIJS AND AQUAFERTILIZER CHEMICALSANTIFREEZE COMPOUNDSAGRICULTURAl. CHEMICALSAUTOMOTIVE CHEJI. El. ANTIFREEZREPAIR CONSTRUCTIONELECTRIC POWERWATER AND OTHER UTILITIESREPAIR SERVICE6TH REAL EST (NON-RENfl&FIN.SEINSURANCE & W.C.B.OTHER RENTRENTAL DATA PROCESSING EQUIP.RENTAL OF AUTOMOBILES & TRUCKSRENTAL AD MACH&E0.INCLCONST.MSPARE PARTS&MAINT. SUPPLNACH&EOTHER INDIRECT TAXESWAGES AND SALARIESSUPPLEMENTARY LABOUR INCOMEOTHER OPERATING SURPLUS348 0.014511 0.000412 0.0005135 13 0.00611250 132 0.057816 0.0006I 45 0.00192208 3 0.092489 0.003747 0.00193 2 0.000213 12 0.00100.0000ii 0.0004195 0.00811579 0.066051 0.00214 0.000115 0.00062 0.000098 0.00407 30 0.001527 0.00118 0.000311 37 0.002013 11 0.0010-65 0.003213 67 0.00332 2 0.000113 0.00052 2 0.000173 375 0.0187148 547 0.029023 114 0.005726 32 0.00241014 0.042426 0.001068 0.0028616 0.02571 6 0.0002583 0.02431 0.0000111 96 0.0086169 120 0.01201 16 0.000735 114 0.006218 20 0.00154 52 0.002336 657 0.02891 2 0.00013 3 0.00023 3 0.000259 528 0.0245385 597 0.0410762 933 0.070827 39 0.00271345 3676 0.2100W—LEYEL COMMODITY AGGREGATE 1 AGRIC 2 A6RIC XSHARE DISTRIBUTION SOURCE(1981 Census of Agriculture, except as noted 1986)Cattle and calves, total, 1986pigs, total, 1986hens and chickens, total, 1986feed & supplements I areas under cultivationfeed & supplements I areas under cultivationhens and chickens, total, 1986feed & supplements / areas under cultivationfeed & supplements I areas under cultivationseeds & seedlingsseeds & seedlings, or areas under oil seedsfuel, oil and lubricantsfuel, oil and lubricantsfeed & supplementsfeed & supplementsfeed Ic supplementsfeed & supplementsfeed & supplementsfeed Ic supplementsfeed Ic supplementsfeed & supplementsfeed & supplementsnumbers of autos, trucks, tractorsnumbers of balersnumbers of balersrepairs, maintenance of buildingsrepairs, maintenance of buildingsnuibrs 0-f tratørsvalue of machinery and equipment, 1986repairs, maintenance of buildingsfertilizer and lime expends.fuel, oil and lubricantsfuel, oil and lubricantsfuel, oil and lubricantsfuel, oil and lubricantsfuel, oil and lubricantsfertilizer and lime expends.value of livestock and poultry, 1986fertilizer and lime expends.fertilizer and lime expends.agricultural chemicalsagricultural chemicalsnumbers of autos, trucks, tractorsrepairs, maintenance of buildingselectricity expenditureselectricity expendituresrepairs, maintenance of machineryvalue land Ic buildings, 1986hired plus on labourcash rent; or areas rented, 1986numbers of computers, 1986machine rental and custom yorkmachine rental and custom yorkrepairs, maintenance of machineryvalue land Ic buildings, 1986cash vages, 1986days of oil—work, 1986total revenue minus total expenses, 1986subtotal 9270 10816 0.840192APPENDIX B2: COMMODITY INPUTS (l986 IMILLIONS) AND DISTRIBUTION SOURCES (THOMASSIN AND ANDISON)I ASRIC 2 A6RIC ZENARE DISTRIBUTID SOURCE (Thoiassin and Andison)102300 SERY. INCIDENTAL TO A6R.FOREST 141 164 0.0127 ext. vet and custo. york (TA)2 02400 LO6S AND BOLTS I 1 0.0000 other expenses (T&A)304400 SALT 2 2 0.0001 other expenses (TkA)4 04500 PEATMOSS 3 0.0001 other expenses (TA)5 05000 STUNE,CRUOE 7 6 0.0005 other expenses (T&A)6 13600 ftASTIC CDNTAINERSbBOTTLE CAPS 5 0.0002 other expenses (ThA)717500 TEXTILE CONTAINERS 4 0.0001 other expenses (flA)8 19900 CONTAINERS, CI.OSURES&WOOD PAILE 8 0.0003 other expenses (TIA)922100 PAPER CARTOMS,BA6S,CANS&BOTTI.E 5 0.0002 other expenses (T&A)10 22500 PAPER CONTAINERS,NES 6 0.0002 other expenses (T&A)1153200 SERY. INCIDENTAL TO TRANSPORT N 8 0.0003 transport .argins (ThA)12 53600 TRUCK TRANSPORTATION 7 6 0.0005 transport largins (ThA)13 54000 PIPELINE TRANSPORTATION 8 11 0.0007 pipeline largin (T&A)14 54200 STORAGE 40 0.0016 storage .argin (ThA)1554400 TELEPHONE TELEGRAPH 43 44 0.0036 other expenses (T&A)16 54500 POSTAL SERVICES 5 4 0.0003 other expenses (ThA)1754700 6ASDISTRIBIJTION 3 2 0.0002 gasurgin(T&A)18 55000 NHOLESALING MARGINS 194 442 0.0266 iiholesale largin (TtA)19 55300 RETAILING MARGINS 29 45 0.0030 retail .argin (T&A)20 55400 IMPUTED SERVICE, BANKS 85 92 0.0074 distribution of interest costs (ThA)21 56600 SERVICES TO BUSINESS MANAGEMEN 17 31 0.0020 other expenses (T&A)22 57600 OTHER SERV.TO BUSINESSESPERS0 I 1 0.0000 other expenses (ThA)23 57800 TRADE ASSOCIATION DUES 8 0.0003 est. vet and custos sort (T&A)24 58100 OFFICE SUPPLIES 8 7 0.0006 other expenses (T&A)25 58200 TRANSPORTATION MARGINS 72 191 0.0110 transport largins (T&A)26 58500 TRAVELLING AND ENTERTAINMENT 1 0.0000 other expenses (TtA)2759500 IOVERNMENT 600051 SERVICES 6 24 0.0012 subsidies by fan, type (ThA)28 59600 COMMODITY INDIRECT TAXES 22 103 0.0052 indirect conodity tax (T&A)29 59700 SUBSIDIES —680 —1528 —0.092 subsidies by far. type (T&A)30 60100 NET INCOME UNINCORP BUSINESS 1651 2464 0. 1721 net jocose by fan, type (ItO)subtotal 1639 2183 0.1598INDUSTRY TOTAL USE 10909 12999

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0086546/manifest

Comment

Related Items