Bureaucracies and the Political Economy of Protection Reflections of a Continental European Patrick A. Messerlin WORLD BANK STAFF WORKING PAPERS Number 568 ISN=178127 A010 DATENT : 1983-10-23 A020 DATMOD 1998-08-25 A030 ENTBY : SE A040 LEVDES : INT A050 LSCNTN : 8829 A060 CATSOU R A100 APVALU : P A120 TYPE : SWP Staff Working Paper A160 SCODE : 9 B020 REPOSI IDU B050 SCHSER NONE B070 RNUM SWP568 B090 EXTNUM ISBN 0-8213-0198-5 C010 PREPBY MESSERLIN, P. EPD C030 ORIG ECS C100 TITLE : Bureaucracies and the political economy of protection reflections of a continental European C120 COTITL : Staff working paper ; no. SWP 568 C210 DATINT 1983-08 C260 PHYDT 0047 DOIO DESCRI Bureaucracy Protectionism Trade policy D030 ABSTR : A project to establish the shares of industrial country markets held by developing countries, changes in such shares in the 1970s, and reasons for variations among industry groups and countries aims to assist developing and industrial countries improve their policies through a better understanding of trade patterns and protectionist measures. This paper looks at the role of the bureaus (bureaucrats) in the supply of protection in France. Three factors appear to influence the bureaus' observed tendency toward protection: (i) its position as a "partial dictator" with respect to the industry it supervises; (ii) inability of a bureau to capture the profits that result from protection; and (iii) costs to a bureau of information about the impact of protection. The study suggests several approaches to negating the protectionist tendency of bureaus and states that more attention to and study of the relations between trade policy and administrative structure are necessary. D130 CR : FR FRANCE D160 SECTOR : MT Trade Policy Reform E101 ITMID : E E109 IFLAG : 01284748 FIO RESPUN : ECS D400 MSECT : MX Multisector D410 SECT : MT Trade Policy Reform PLEASE SELECT FUNCTION ?E ISN 178127 ** STATUS (STAT, F270) MUST BE PRESENT ** ERROR RUNNING USER EXIT AT SUBSCRIPT 2 ** ** DID NOT MEET USER EXIT CONDITIONS..RE-EDIT RECORD (M568) ** USER RECOVERABLE WRITE FAILURE..RE-EDIT RECORD (M569) A Set of Related WORLD BANK STAFF WORKING PAPERS Public Subsidies to Industry The Case of Sweden and Its Shipbuilding Industry Number 566 The Political Economy of Protection in Italy Some Empirical Evidence Number 567 Bureaucracies and the Political Economy of Protection Reflections of a Continental European Number 568 Economics and the Politics of Protection Some Case Studies of Industries Number 569 Public Assistance to Industries and Trade Policy in France Number 570 The Structure of International Competitiveness in the Federal Republic of Germany An Appraisal Number 571 WORLD BANK STAFF WORKING PAPERS Number 568 Bureaucracies and the Political Economy of Protection Reflections of a Continental European Patrick A. Messerlin The World Bank Washington, D.C., U.S.A. Copyright i 1983 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. First printing August 1983 All rights reserved Manufactured in the United States of Ainerica This is a working document published informally by the World Bank. To present the results of research with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The publication is supplied at a token charge to defray part of the cost of manufacture and distribution. The views and interpretations in this document are those of the author(s) and should not be attributed to the World Bank, to its affiliated organizations, or to any individual acting on their behalf. Any maps used have been prepared solely for the convenience of the readers; the denominations used and the boundaries shown do not imply, on the part of the World Bank and its affiliates, any judgment on the legal status of any territory or any endorsement or acceptance of such boundaries. The full range of World Bank publications is described in the Catalog of World Bank Publications; the continuing research program of the Bank is outlined in World Bank Research Program: Abstracts of Current Studies. Both booklets are updated annually; the most recent edition of each is available without charge from the Publications Distribution Unit of the Bank in Washington or from the European Office of the Bank, 66, avenue d'Iena, 75116 Paris, France. Patrick A. Messerlin, at the University of Lille I, is a consultant to the Economic Analysis and Projections Department of the World Bank. Library of Congress Cataloging in Publication Data Messerlin, Patrick. Bureaucracies and the political economy of protection. (World Bank staff working papers ; no. 568) Bibliography: p. 1. Europe-Commercial policy. 2. Free trade and protection-Protection. 3. Bureaucracy--Europe. I. Title. MI. Series. HF1531.M47 1983 382.7'3#094 83-17037 ISBN 0-8213-0198-5 ABSTRACT This report is part of an inquiry undertaken by the World Bank in conjunction with scholars from 12 industrial countries into the penetration of the markets of industrial countries by exports of manufactures from developing countries. The project sought to establish the shares of industrial country markets held by the developing countries, changes in such shares in the 1970s, and why they vary among industry groups and countries. The aim is to assist developing and industrial countries to improve their policies through a better understanding of trade patterns and protectionist measures. An often overlooked but very important element in the political economy of protection has been "the bureaus" (bureaucrats). This paper looks at the role of the bureaus in the supply of protection in France. The tendency has been for the bureaus to advocate protection, even beyond that advocated by politicians. This tendency seems contradictory to the presumed public interest goals of bureaus, which would be to favor free trade. Moreover, the bureaus have often resorted to complex subsidy programs or hidden non-tariff barriers that are hard to measure, rather than to transparent protection measures. Three factors that relate in large part to the bureaus' roles and relationships appear to influence their tendency toward protection. One is a bureau's position as a "partial dictator" with respect to the industry it supervises. A second is the inability of a bureau to capture the profits that result from protection. A third is the costs to a bureau of information about the impact of protection. The study suggests several approaches to negating the protectionist tendency of bureaus. The larger a bureau's mandate, the more likely it is to favor free trade. More accurate information about the impact of protection on the consumer will lead to more emphasis on free trade; that information is best obtained by institutions that can examine the full range of final goods. Finally, more attention to and study of the relations between trade policy and administrative structure are necessary. ACKNOWLEDGEMENTS I would like to thank J. P. Martin, A. 0. Krueger, S. P. Magee and J. Waelbroeck for their encouragement and many helpful comments on previous drafts. I am also indebted to H. Hughes and the World Bank project on the penetration of the markets of industrial countries by exports of manufactures from developing countries, to S. Easton, B. Lenz and J. J. Rosa, and to an anonymous referee of Weltwirtschaftliches Archiv. I am, of course, solely responsible for any errors which remain. TABLE OF CONTENTS I INTRODUCTION ........*............**Pse* 1 II ADMINISTRATTVE MECHANISMS ...... ...................... 4 Bureaus and Trade Policy in France ................. 4 The Workings of the Bureaucratic System ............ 6 III BUREAUS AND THE CONSERVATIVE SOCIAL WELFARE FUNCTION ...................................... 13 The Partial Dictatorship and the CSWF .............. 14 The Spread of Protection ........................,.. 20 IV BUREAUS VS. POLITICIANS . ............ . .. 24 Bureaus Vs. Politicians: The Case of One Means of Protection ... ......... o ............. 25 Bureaus Vs. Politicians: The Case of Several Means of Protection ...... ......@ 32 V INFORMATION COSTS AND THE BIAS TOWARD PROTECTION ..... 36 Bureaus as Information Producers .* ................. 39 Bureaus as Information Users ..................... 40 VI CONCLUSIONS ...................................... 44 REFERENCES ..................................... 46 LIST OF FIGURES Page 1 A Bureau with a Conservative Social Welfare Function .............. . .. .. .. . 16 2 Bureaucrats Vs. Politicians .......................... 28 3 Equilibrium Level of Tariff Changes in a Partial Equilibrium Framework . . ............................ . . 38 Chapter I INTRODUCTION To date, the human factors considered in the literature on the political economy of protection have been almost exclusively politicians and voters, while the role played by bureaucrats has been largely ignored. A well-known exception is Breton,l- whose approach, however, is rather different from that contained in this paper. While Breton's approach may be adequate for the U.S. case, it does not seem to fit the European situation well because of constitutional and institutional rules at both the national and EC levels. It is not excessive to argue that, in many respects, top-level bureaucrats have been responsible for determining the level of protection at least as much as politicians have, although obviously there are substantial differences among the European states. This paper analyzes the role of the bureaucracy on the supply side of the market for protection 21 If the bureaus had been pursuing public interest goals, they should have exhibited behavior that favored free trade. However, this does not appear to have been the case. On 1/ A. Breton, The Economic Theory of Representative Government, Chicago: Aldine Publishing Co., 1974. 2/ For the sake of simplicity, "bureau and bureaucrats" will be used here interchangeably. "Bureaus" are broadly defined as any institution in which civil servants are a non-negligible proportion of the personnel. No clearcut distinction is made between an individual bureaucrat and a bureau. (A more sophisticated approach would be to recognize the differences in the objectives of the members of a bureau, along the lines of the theory of teams a la Marschak-Radner.) Finally, it is assumed that bureaus are not corrupted by the industries they supervise. -2- the contrary, the evidence is that the bureaus have tended to advocate protection. Although some bureaus have been more concerned with helping export- than import-competing firms, they have not necessarily been free trade-oriented. Friendliness toward exporters is not usually perceived of as being inconsistent with a quite protectionist attitude toward other sectors. Moreover, where an industry was hurt by a new foreign competitor, the bureaus rarely advocated transparent protective measures, but, rather, supported complex subsidy policies (which are hard to measure fully) or hidden non-tariff barriers (NTBs). Within the bureaucratic framework, the implementation of measures appears to have been the result of a quite decentralized institutional framework, involving more than just politicians. Three arguments will be presented to explain the protectionist behavior of the bureaus. Following Chapter II, which provides some background to bureaucratic policy-making, Chapter III presents the argument that a bureau in charge of trade policy may be considered as a 'partial dictator" which will define the "rights to be protected from foreign competition" in favor of the factors of production it supervises. Further, a bureau tends to think of protective measures as "goods" and to overproduce them relative to the optimum level. In Chapter IV, politicians and bureaus are compared as producers of protection. The main conclusion is that bureaus are likely to produce more protection than politicians are. In Chapter V, it is argued that information costs will increase the bias of bureaus toward protection. All three arguments provide strong support for the idea that bureaucrats favor relatively greater protection than do politicians. At first glance, this conclusion seems paradoxical, since it is often believed that bureaucrats will be more "neutral" than politicians in determining trade policies because they cannot derive any political profit from such policies. However, it will be shown that bureaucrats do not get such political profits. The fact that bureaucrats tend to be more protectionist than politicians raises some doubts about the validity of the theoretical arguments that suggest that subsidies are superior to tariffs as instruments of commercial policy. It will be argued, to the contrary, that these theories are misleading because they do not fully take into account that these policy instruments are typically administered by public bodies which are protection-prone. Finally, some ways of limiting the protectionist bias of bureaus are discussed in Chapter VI. - 4 - Chapter II ADMINISTRATIVE MECHANISMS As stated above, there are differences among the European states in the role of the bureaucracy in carrying out commercial policy. Some countries such as Denmark, and perhaps the United Kingdom, are still somewhat similar to the United States, in that it is appropriate to consider only the role of politicians. In others, such as the Federal Republic of Germany and above all France, there is strong evidence that bureaus make decisive choices in the field of commercial policy. Finally, there is the EC Commission, a bureaucratic body that plays an essential role in controlling commercial policy in its member states.- Bureaus and Trade Policy in France It is interesting to look at some historical examples of how bureaus have functioned as decision-makers in the field of commercial policy in France. French commercial policy during the 1920s and the 1930s, which was imitated at the time by several European countries and was maintained more or less through the '50s, is one of the best illustrations of the dominant role of the bureaus. In the thoroughly mystifying French Tariff Law of 1920, Parliament voted to re-establish the specific duties existing in 1913. The governmental bureaus were to I/ J. Waelbroeck and E. Verreydt, "European Community Protection Against Manufactured Imports from Developing Countries: A Case Study in the Political Economy of Protection," in J. Bhagwati, ed., Adjustment and Protection, Chicago: University of Chicago Press, 1980. compute (and enforce by decrees) the "coefficients" to be applied to those duties, taking into account the wartime and postwar inflation and economic changes. The coefficients in fact were set at a level that provided real protection three and four (and sometimes seven or eight) times the old duty rates. The predominant role of the bureaucracy was also clearly evident under the Tariff Laws of 1926-1927, through which the Parliament gave the government the power to establish a new schedule of tariff rates (called a "consolidation"). It reached its peak after 1931, with an incredibly detailed system of quotas. An interesting feature of the system, one that is frequently observed in bureaucratic systems of protection, is that the bureaus tried to push discrimination between countries as far as possible. For example, they attempted to conclude bilateral trade agreements and to promote the "conditional" version of the Most Favored Nation clause. Indeed, it is quite clear that the bureaus developed trade controls as far as they could, limited only by their physical and human capacities to produce those types of services. That those capacities were limited is suggested by the fact that it was precisely the technical difficulties of managing the quota system (especially after 1934) that drove the bureaus themselves to ask the French government to limit the expansion- of the administrative trade controls and, in 1939, even to reduce the list of quotas. There is a host of examples of the dominant role that some bureaus currently play. At the EC level, the Commission's management of the agricultural sector and the steel cartel are well-known cases. At - 6 - national levels, the textile sector certainly exhibits the widest range of subtle trade controls, enforced by the bureaus. Devices such as France's so-called "surveillance statistique" allow the government to increase the costs of imports and to slow them temporarily; two outcomes are an increase in cheating and uncertainty about future imports. The Workings of the Bureaucratic System The obvious question is how the European bureaus (specifically, the top-level bureaucrats) have been able to become so autonomous, to dominate decision-making on commercial policy and, in part, to displace the politicians. The first reason is to be found in the constitutional basis for most European civil services, i.e., in the statutes establishing civil servants (including top-level civil servants). The statutes guarantee that a civil servant cannot be fired except for serious incompetence or illegal acts. A crucial point is that this guarantee relates to the "career" and not the specific position held by the civil servant /L Further, during the last 30 years, the practice has been to restrict as much as possible a politician's ability to fire a civil servant on grounds of incompetence or illegal acts. On the other hand, the statute does allow for a civil servant's transfer to another position when judged to be inefficient or "undesirable" in the current position. However, this threat of 1/ It should be noted that European civil services differ fundamentally from the U.S. civil service precisely because they are based on the career (rank-in-man system) rather than the job (rank-in-job system). - 7 - displacement is the only pressure on a civil servant. When a top-level bureaucrat is involved, even this option may be limited by the fact that the transfer could easily appear to be political and could trigger pressures from political parties and/or economic groups. It is argued here that because of the statutes, the cost of making decisions for a top-level bureaucrat desiring to become a dominant decision-maker is lower than it would be under other circumstances. Politicians, for their part, have been doing their best to adapt to this evolution. For example, cabinets within ministries tend to grow larger. Obviously, an important part of the job of a civil servant nominated to such a cabinet is to counter the influence of the top-level bureaucrats in the Ministry and to introduce some kind of flexibility and political flavor into decision-making. 1/ An additional point is that, since the beginning of the 1960s, more and more European politicians have come from the bureaucracy 2/ Whatever the reasons for this trend, there have been two effects. First, the bureaucrat-turned-politician tries to use the personal debts accumulated during his years as civil servant. Frequently, he is biased 1/ An example is the Secretaire d'Etat aux Postes, T6l&communications et la T4l4diffusion (State Agency for the Post, Telecommunications and Broadcasting), which is strongly involved in the telecommunica- tions sector. See Le Monde, February 26, 1981. 2/ As a brief illustration, 55 percent of the French ministers came from the top-level bureaucracy. There was also a massive entry of bureaucrats into the crucial Interministerial Committees at the highest levels in West Germany, France and Italy, and even in the United Kingdom, which so far seems to have been most reluctant to follow this trend. -8- toward solutions that use his knowledge of the bureaucratic process or that give some role to the bureau from which he came. Second, the "road to politics" is easier for a top-level bureaucrat who has had opportunities to show his capacity as a decision-maker while managing a bureau. It is also true that it is easy to find examples of top-level bureaucrats who are (were) considered as the "true" Minister because they have (had) the confidence of the President or Prime Minister. It should be noted that the statutes for the civil service in the United Kingdom are substantially different from those in continental Europe, a fact that may explain the differences in the roles played by the bureaus in the latter. Another factor is also important. As commercial policies cause all kinds of problems -- financial, industrial and social -- the choice of a trade policy, especially if fundamental decisions have to be taken (e.g., freer trade vs. protection), always involves several bureaus. At first glance, it would seem that some kind of fair competition among the bureaus would result. However, that does not appear to be the case. Indeed, two well-established practices within European bureaucracies have to be considered. First, one bureau will have the major role in defining the trade policy for a given industry, and its role is more or less officially accepted by the other bureaus. This "division of tasks" reduces the competition among the different bureaus, especially when, as is often the case, the principal bureau has played this role for a long time. This long-run stability exists because changing the principal bureau is usually a costly process. It also implies a redefinition of the roles of all the other bureaus, a - 9 - situation that could lead to lengthy log-rolling within the bureaucracy l/ On the other hand, the second well-established practice modifies the first one, and sometimes substantially. The bureaucratic system is not as centralized as is often believed. Instead, it may be viewed as a set of bureaus with weak inter-relationships, primarily because of information costs. Consequently, when no formal hierarchical relationship exists between the principal bureau and the others -- in the most common case because all the bureaus are located in different ministries -- a bureau which disagrees with the policy enforced by the principal bureau may make its own policy choices, which it then enforces itself. Such competition often leads to innovations, with each bureau seeking to "invent" new ways to enforce its objectives. It tries to use techniques that are sufficiently secret (e.g., subsidies) that it can escape accurate monitoring by other bureaus. Nevertheless, the long-run stability is clearly a decisive factor in explaining the role of bureaus in trade policy. This stability allows a bureau to accumulate information on any industry and to gain experience with it. Slowly it is able to develop a comparative advantage relative to politicians for solving the trade problems of that industry. 1/ This long-run stability is often reinforced by the stability of the civil servants working within the bureau. As a result, relationships between a bureau and the industry it supervises often involve the same people over a long period of time. - 10 - It is interesting to note that the bureaus which are con- sidered as principal relative to trade policy are not generally dependent on the Ministere du Commerce Ext6rieur (Ministry of Foreign Trade) (in France it has been argued that this dependence has never existed; see the Pisani report).- Indeed, the principal bureaus have generally depended on the Ministry of Finance and/or Industry and/or Agriculture, the so-called bureaus "de tutelle" (administrative agencies) which have an extensive knowledge of the industries the bureaus supervise. The long-run stability not only creates favorable conditions for the supply of protection, but also effects the demand for protection. Indeed, at times it may induce lobbying industries to invest more heavily in their relationships with the bureaus than with politicians, as the bureaus, because of their stability, might be seen as providing more lasting benefits than politicians would. This situation is even more likely when relationships with the bureaus involve fewer persons thanT relationships with politicians, or when the political game entails many political parties, some of them small, though possibly politically strategic. Two final remarks are in order. First, given the importance of lobbying, it is possible for an industry to "capture" the bureau that supervises it, just as it is said that regulated industries in the 1/ E. Pisani, "Rapport d'Information sur les Enjeux et Conditions des Equilibres Exterieux de la France" [Information Report on the Stakes and Conditions of France's External Equilibria], SMnat [Senate], Paris, 1979. - 11 - United States sometimes capture their regulators.I' A host of examples of this "capturing" can be supplied. This situation has nothing to do with corruption. An industry may "capture" a major bureau in two quite legitimate ways. The first relates to the preparation of decisions concerning trade policies. An industry may give the bureau data and review the pros and cons of all the possible technical means of protection. Casual observation (of the textile industry at the end of 1976-beginning of 1977, both at the national and EC levels, and the steel industry two years after) suggests that "captures" may result when fast, strong and relatively unexpected increases in penetration ratios occur. The second way relates to the enforcement of decisions by the bureau. Some of the more subtle measures of trade control (such as the verification of the origin of imports) require some input from industry, as illustrated recently by the protective measures for the textile sector in France 2/ A final point is that the bureaucratic case amounts to a special instance of the rent-seeking process first analyzed by Krueger 31 The inclusion of bureaus, however, makes the rent-seeking 1/ G. J. Stigler, "The Theory of Economic Regulation," Bell Journal of Economics and Management Science (Spring 1971):3-21; and G. J. Posner, "The Theory of Economic Regulation," Bell Journal of Economics and Management Science (Spring 1971):22-50. 2/ For instance, French customs are currently helped by people in the textile industry. 3/ A. 0. Krueger, "The Political Economy of the Rent-Seeking Society," American Economic Review (June 1974):291-303. - 12 - process very complex and leads to an ambiguous result. On the one hand, it is easy to understand why bureaus might themselves be interested in rent-seeking .l1 For instance, the survival of the industry can make the survival of the bureau supervising it more probable. On the other hand, it may also be observed that bureaus will try to stop the rent-seeking process, at least in part, if that process brings about demands for protection that exceed their production capacities. This important kind of phenomenon was mentioned in the discussion of French trade policy at the end of the 1930s, when the bureaus exerted pressure to reduce the list of quotas. 1/ The fact that bureaus are rent-seekers does not mean that the individual bureaucrats working in those bureaus can enjoy the "collective" rents (although wages for top-level bureaucrats are sufficiently differentiated as to incorporate some elements of the bureau's quasi-rents). - 13 - Chapter III BUREAUS AND THE CONSERVATIVE SOCIAL WELFARE FUNCTION Not surprisingly, the concept of a social welfare function (SWF) will be helpful in analyzing bureaucratic behavior. It is necessary to examine carefully the conditions under which a bureau is able to develop its SWF since, as emphasized by Buchanan.', it must act within the limits imposed by its mandate. What, then, are the main limits to be considered? First, a bureau has the power to make decisions only on a specific set of issues. Note that a politician, to the contrary, may do so for a whole range of economic issues. This difference is crucial, since it gives politicians opportunities for survival that are unavailable to bureaus. It is easier for a politician to compensate for the difficulties on one issue by promoting another: if a tariff problem is politically messy, he can focus on other issues that are politically easier to handle but have the same expected (allegedly) redistributive consequences for voters. Further, each bureau has only a specific range of means at its disposal, which may not have been defined logically relative to the issues. Rather, they may have been inherited or are the consequence of the task specialization between bureaus. Politicians, on the other hand, normally have access to all legal measures. A politician who was not able to get a tariff to protect an industry in his district can then try to obtain public contracts, military camps or whatever and will be able to argue convincingly that these measures 1/ J. Buchanan, "The Coase Theorem and the Theory of the State," Natural Resources Journal 13(4)(October 1973):579-94. - 14 - would increase the level of activity of the depressed regiond.' More important still, it is easy to see that politicians have more latitude than bureaucrats for capturing the profits created by protection, as will be discussed in Chapter IV. It is therefore appropriate to define a bureau as only a "partial dictator" because of the limitations on its sphere of issues and means. This concept of a partial dictatorship will allow more concrete substance to be given to the concept of the conservative social welfare function (CSWF) suggested by Corden2/ and to develop some new arguments on the issue of the optimal tariff. The Partial Dictatorship and the CSWF This subsection focuses on the differences between the dictator as generally considered in the orthodox theory of protection -- the so-called "benevolent despot" -- and the partial dictator of a bureaucracy. The "dictator" of economic theory is able to maintain a Pareto-optimal solution whatever the shape of his SWF because he can use the whole range of redistribution measures to enforce his choices. Even if the dictator has a "conservative" SWF -- according to Corden, "any 1/ This point is very important and suggests that the assumption usually made in presentations on the political economy of protection that tariffs are the only instrument of protection is not correct (although relaxing this assumption introduces analytical difficulties). Instead, politicians have to be judged on an average rate of success. This interpretation makes more plausible the view that politicians can be considered as monopolists on the tariff issue (see Chapter IV). 2/ W. M. Corden, Trade Policy and Economic Welfare, Oxford: Clarendon Press, 1974. - 15 - absolute reduction in real income in any section of the economy should be avoided" -- he may still choose a free trade solution combined with an adequate system of lump sum transfers. In other words, a CSWF is not a sufficient condition for protection, given the usual framework but assuming a dictator. The problem may be easily illustrated using the extreme case of a small economy where some factors are specific to the production of imported and exported goods. A policy change from autarky to free trade will shift the utility possibility frontier outwards. This new frontier is shown by TF in Figure 1. Call A the initial situation under autarky, and F the final one under free trade. As a policy change toward free trade will unambiguously hurt all the specific factors employed in the import-competing sector, the aggregated utility of all these factors, as measured along the UM axis, decreases when the economy moves from A to F. This result may be unacceptable to the dictator if he has a CSWF. Consequently, he has to move along TF to reach acceptable points. The point FM corresponds to the- first acceptable situation. The segment FmFX of the frontier is the locus of points which is acceptable to the dictator. However, this definition does not exclude some convexity for the indifference curves situated to the right of FMAFX. The optimal point for the dictator will be located somewhere between FM and Fx, at F', for instance, implying some level of lump sum transfers. Now assume a bureau, BM, in charge of import-competing industry M, and another bureau, Bx, in charge of export industry X. - 16 - Figure 1: A BUREAU WITH A CONSERVATIVE SOCIAL WELFARE FUNCTION ux T. B F F' L F Ax A _ - ___X _ 0 AM UM - 17 - Nothing new will occur if the bureaucratic organization is ideal, which means there will be full cooperation between BM and Bx at zero transaction cost. In other words, all the points between FM and Fx remain attainable, including Fx, the point at which bureau BM appro- priates all the gains from trade using transfers. Further, BM is giving all the gains from free trade to the factors of production hired by industry M (industry X is assumed to be indifferent between A (autarky) and Fx). There is, in reality, no such ideal bureaucratic organization. Cooperation between bureaus is generally difficult to achieve as it involves transaction costs. To avoid such costs, bureau BM may prefer to use the limited range of means it dominates to implement the "rights to be protected" that it decides to impose. Since a bureau in charge of commercial policy seldom has the jurisdiction to organize lump sum transfers, it will not consider the TF frontier to be attainable, since TF is outside its set of possible means of redistribution. Alternatively, the bureau will only consider the utility possibility frontiers available with the redistribution schemes it can effectively enforce (possibly with minimal cooperation from Bx). Therefore assume the simplest case, in which bureau BM can redistribute in favor of industry M only by imposing a tariff; under such a trade policy, the attainable utility frontier is illustrated by FT'. However, where cooperation between the bureaus is poor, a chain reaction develops -- after BM has made decisions aimed at improving the situation of the import-competing industry, Bx moves to avoid a decrease in the output share of the export industry. - 18 - This case will be returned to later when the spread of protection is considered. In the meantime, assume that each bureau is small enough to take as given the decisions of the other bureaus. Under this assumption, the only utility function which plays a role in the above framework is the utility function of bureau BM, which is now specified. Although the relationships between a bureau and the industry it supervises are complex, the survival and the prestige of a bureau are as a rule correlated with the existence and size of the industry it supervises.-. Consequently, the bureau may be permanently induced to identify its own utility with that of the factors of production hired by the industry it supervises. This evolution is strengthened by the tendency of a bureau to make decisions in a partial equilibrium framework that is restricted to the industry it supervises. 2/ This analysis can be expanded somewhat by taking into account the "prestige" component of BM's utility, derived from its power and effectiveness. To gain prestige, BM may try to protect the import- competing industry vigorously: it will seek to increase UM more than UX (starting from any point to the left of FMAAm, the "limit" indifference 1/ It may be added that a bureau may link its survival and prestige to the existence of "national" firms. Foreign firms, even if they produce in the domestic economy, are presumably associated with a decrease in the role of the bureau. 2/ One of the most illustrative examples is the amazing role of sectoral trade balances as a guide for policy. See, for example, the campaign for the "Reconquete du marche interieur" [Recapturing the domestic market] in France since mid-1979, which is mainly based on evidence concerning trade balances. This bureaucratic myopia is reinforced by the vertical structure of the bureaucracy and by the great difficulties the bureaus have in working together (sharing information, defining common goals, etc.). - 19 - curve when BM has a CSWF). In other words, the complete CSWF of BM may be illustrated by a set of indifference curves similar to B which exhibit a relatively "flat" gradient in any point. As a result, bureau BM will "optimally" favor the tariff solution, which allows a shift from A to some point L on FT'. Point L will be close to M if prestige plays an important role in the BM's behavior. Reaching point L depends on several factors: the rights to be protected that are enjoyed by the import-competing industry, the bureau's means of enforcing these rights, the prestige component of the bureau's utility, and the imperfect cooperation between the bureaus involved. Further, point L will remain stable only under two assumptions: the small bureau one, according to which the only utility function involved is BM's, and the small country one, according to which bureau BM imposes a level of protection higher than the optimal one (which is zero, i.e., free trade). It is important to note, as well, the effect of administrative organization on trade policy. If the division of means between bureaus does not change, the trade liberalization observed between A and some point on the frontier FT' (say L) will probably stop definitively at that point (all other things being constant). If, however, a bureau finds itself with new means of making transfers efficiently, its feasibility utility frontier will shift toward TF through a freer trade policy. In other words, a broad definition of the means available to a bureau may foster a trade liberalization process. This point may constitute part of the answer to the question raised by Waelbroeck and - 20 - Verreydtl-. namely, why "decisions taken at higher decision levels tend to be more liberal than the blow by blow decisions of the bureaus." It is argued here that people at the higher decision levels have available to them more means than do the basic bureaus and therefore may consider utility possibility frontiers more favorable than that of the bureau. The Spread of Protection Imperfect cooperation between bureaus, and the possibility of something like a chain reaction, were suggested above. It should be noted that the consequent instability is not inconsistent with the hierarchical organization of a bureaucracy. Consider the following. It might be argued that any conflict between BM and BX will automatically be solved through arbitration at a higher level of decision. However, as shown by Finger, Hall and Nelson 2/ to involve the highest public authorities on specific problems will not necessarily mean stable decision-making, since, at the highest political levels, "the clearest pressures may be away from a decision, not toward either a yes or a no." Consequently, the conflict between bureaus may persist even within a hierarchical organization. Still more important, the imperfect cooperation between bureaus will play a crucial role in an important phenomenon, called here the "spread of protection." Indeed, an interesting feature in the history of protection is the observed spread of protective decisions over all industries during some periods in a given country. 1/ Op. cit., Waelbroeck and Verreydt. 2/ J. M. Finger, H. K. Hall and D. R. Nelson, "The Political Economy of Administered Protection," American Economic Review 72(3) (June 1982): 452-466. - 21 - A first explanation might be found in some international process of retaliation against "beggar-my-neighbor" policies. Although sometimes present, this explanation does not seem to be either the unique or even the main one. Rather than relying on external reasons, the answer must be sought internally -- with the bargaining situation. The two hypotheses discussed below are alternative explanations -- one assumes perfect information on the part of those concerned about protectionist decisions, while the other assumes there is no information. The first hypothesis, which is also discussed by Waelbroeck and Verreydt 1/ assumes "perfect" information by all about the results of lobbying for protection. In such a case, any concession to a protectionist pressure group will result in familiar responses by other pressure groups, responses somewhat dreaded by bureaucrats: "If industry X gets protection, why not me?" In other words, the concession may be viewed as creating an externality which decreases the cost of lobbying for other interest groups. The decrease in the cost will be quite large if the industries asking for protection are supervised by different bureaus for the following reason -- log-rolling between bureaus lessens the ability of each to resist industry pressures.-2 1/ Op. cit., Waelbroeck and Verreydt. 2/ It is considerations such as this, on the other hand, that motivated the French delegation during the negotiation of the Treaty of Rome (those involved were top-level civil servants). The French tried to minimize the "exceptions" to the general rules of European trade liberalization precisely to help top-level French bureaucrats in confronting lobbying industries at the end of the 1950s. The "no exception" rule was a crucial ingredient in the success of the Treaty of Rome. - 22 - A good illustration of this hypothesis may be observed in the current French and European situations. The automobile sector in France got relatively fast and strong support against Japan -- Japanese cars were limited to a 3 percent market share. Additionally, 19 Japanese car models were not, and still are not, officially cleared for sale by the appropriate bureau in Ministere de l'Industrie (Ministry of Industry). Measures against Japanese cars subsequently spread to the entire EC. The success of the automobile industry helped other industries such as machine tools and consumer electronics to obtain stronger protection. The second hypothesis assumes a kind of non-cooperative competition between a country's free trade advocates and its protectionists. Bureaus presumably play a crucial role in this competition. Consider two scenarios. In the first, which has been observed in the past in some European countries, both importing and exporting industries lobby for more protection. Such behavior may seem puzzling, since protection works by changing the relative prices of imports and exports. Export subsidies and import tariff changes could even cancel out, leaving the domestic price ratio equal to the international one. These scenarios illuminate the contradiction which may exist between the economic rationality of agents acting as individuals, or as members of a cooperating group. In the economic literature, this has come to be called "the prisoner's dilemma," following the presentation of Rapoport and Chammah. .1/ However, the problem is a general one. It 1/ A. Rapoport and A. M. Chammah, Prisoner's Dilemma, Ann Arbor, Mich.: University of Michigan Press, 1965. - 23 - arises whenever it is in the interest of each member of a group, acting alone, to make decisions which are harmful to other members. It is clear that in such cases, cooperative behavior will often be far preferable to each member of the group than that which would be obtained if everyone acted selfishly. However, that cooperation will be very hard to sustain, because it is in the interest of each agent to cheat on the collectively optimal rule of conduct. - 1/ Chapter II presented some institutional motives and examples of the difficult communications among bureaus. A typical example is the case of the Fonds de Developpement Economique et Social (FDES) (Fund for Economic and Social Development). The FDES is the main bureau that gives subsidies to industry. Since 1977, politicians have tried unsuccessfully to get detailed figures on these subsidies. The Conseil d'Etat [State Council] claimed that the figures it provided (which are difficult to interpret) were a "sufficient piece of information" for Parliament. - 24 - Chapter IV BUREAUS VS. POLITICIANS In the preceding section, it was argued that bureaus probably impose a higher level of protection than the theoretically optimal one. This section compares how the two main groups in charge of trade policy in democracies -- namely, politicians and bureaucrats -- operate I/ Consider two situations. In the first, politicians are dominant, with the bureaus unable to pursue their own objectives during decision-making; this is the situation described by U.S. economists. In the second case, bureaucrats are dominant, with politicians playing only the role of the "driving belt" between industries and bureaus, as may be the case in France. The question here is whether one of these decision systems is more likely to be systematically biased toward protection than the other. The preceding sections provide some insight. Politicians can be expected to have utility functions correlated with those of the 1/ It may seem like an exaggeration to discuss politicians and bureaus as opposing each other; it might be more accurate to speak of "bundles" of politicians and bureaus, with different relative endowments depending on the country. This distinction permits more subtle and perhaps more accurate reasoning in terms of the relative rivalries between politicians and bureaus. Moreover, such an approach takes into account "private bureaucracies" -- for example, the U.S. steel producers, who complain about European dumping and carry out very well the job that might be played in Europe by bureaus. However, for the sake of simplicity, these nuances have been ignored by considering a transformation surface which would allow all the different kinds of factors in the production of protection to be taken into account. Finally, competition between politicians and bureaus is excluded. *. - 25 - voters and may appear a priori to be less dependent on one industry than bureaus are. Normally, they may also have access to a broader set of means than the bureaus do, and consequently may be more easily considered as a "collective dictator." Such arguments tend to interpret politicians as more liberal than bureaus. There are exceptions, of course. For example,, if there is a regional concentration of certain industries or firms, local politicians are going to be very sensitive to that. It therefore seems reasonable to focus on bureaus and politicians in terms of their two crucial potential impacts as producers of 1/ protection.- A first question is whether the ability of each institution to claim a share of the profit generated by its activity has any impact on the level of protection produced for industry i. To simplify the analysis, it is first assumed that only one means of protection is available. Then, to examine how the rivalry between bureaus and politicians affects protection, a situation in which there are several protection devices is considered. In both cases, as is shown below, the bureaus prove to be more biased toward protection than politicians do. Bureaus Vs. Politicians: The Case of One Means of Protection This case assumes only one device for protection (for instance, tariffs). Consider an industry observing (or expecting) 1/ In this section, no normative statements are made; the approach is to produce a purely neutral analysis of these two institutions, which, in the spirit of the public choice theory, may be considered as firms producing a (peculiar) public good. - 26 - increases in the domestic market penetration of foreign competitors. Assume that it wants to minimize this penetration and therefore lobbies for -- and gets -- a tariff which provides the desired decrease. The money and effort invested in the lobbying can be expressed as a function of the percentage points of the market share to be recovered, as tariff increases will raise the profits of domestic producers. There will be a relevant range over which the marginal benefits to the industry can be considered as a decreasing function of protection.1. Over this range, the relation, which has negative slope, can be interpreted as the demand curve for protection. To simplify this case, assume that industry expresses the same demand for protection whether it is politicians or bureaus in charge of deciding the level of protection. This assumption is plausible if the demand for protection is understood as a derived demand in the sense that the import-competing firms are demanding from the factors of production (bureaus or politicians) the services (protection) that the firms believe are necessary in order to recover their former market shares. Turning to the supply side, assume that the same cost for production of protection (tariffs) will prevail in both cases. This assumption is realistic, since an essential part of these costs is the result of implementing the decisions, and this depends little on whether politicians or bureaucrats are in charge. Finally, assume temporarily 1/ Such a range will eventually be reached as the protected domestic suppliers saturate the domestic market (assuming competition). - 27 - that the involved politicians or bureaucrats experience no rivalry from other politicians or bureaucrats respectively (a situation considered in the next subsection). The next step is to define the supply behaviors of the bureaus and politicians. Consider first the bureau. The amount of protection it will produce will obviously depend on its assumed objective -- to expand its power as far as possible through its actions. Such a case can be analyzed using Niskanen's approach 1/ a bureau will try to maximize its budget because it cannot capture the profits generated by its decisions. This is because the civil service regulations are designed to prevent a bureaucrat's access to the wealth of the bureau. This point is crucial, since it means that the maximization of profits (wealth) by bureaucrats is unattractive and that there must be another objective such as budget-maximizing. In the case of top-level European bureaucrats, this inability to capture the profits seems to be supported by the fact that they do not generally leave the civil service for industry 2/ Thus, if X is the demand curve, as illustrated in Figure 2, a Niskanen-type bureau will produce an amount of protection equal to QN, where the marginal cost is equal to the so-called "all-or-nothing" 1/ W. Niskanen, Bureaucracy and Representative Government, Chicago: Aldine-Atherton, 1971. See also Breton, op. cit. 2/ Less than 15 percent of top-level French civil servants leave the government for private firms, according to generally accepted estimates. - 28 - Figure 2: BUREAUCRATS VS. POLITICIANS Cost or benefits of security protection ($ per percentage point of tariff) Mc OP PN aM Tariff in percent- aCepit - 29 - demand curve, DF ./ This DF curve shows the maximum price the import- competing industry is willing to pay for a given amount of protection (or the acceptable tariff level), assuming that the bureau offers industry two alternatives: free trade or the given amount of protection. It may be askgd how a bureau can have such monopolistic power. One reason is -that the bureau could deprive the industry of valuable services (for instance, the collection and provision of data on domestic market penetration, efficient controls at customs stations, subsidies, etc.) unless the industry were paying the full price. An alternative objective for a bureau is as follows. Instead of trying to maximize its budget and accepting the implied strong constraints on the amount and use of its managerial discretion-2/ a bureau may seek a "quiet life" and accept some X-inefficiency. There 1/ In this case, some care has to be taken over who is expressing the demand. In the familiar Niskanen model, demand is expressed by the politicians who sponsor the bureaus. This view is accurate if it is assumed that the demand for protection as expressed by the politicians is nothing more than the demand expressed by the factors of production of the import-competing industry and transmitted by the politician without any change. Is such a role plausible for politicians? The answer is yes, at least during some periods. In an institutional system in which the bureaus are effectively in charge of policy-making, politicians may have only an "informative" role and may have no real motive to do costly (for them) arbitrating between the buyers of protection -- the factors of production in the import industry -- and the consumers. However, in such a case, it is likely that the "informative" role of the politicians will rarely be known by the voters except when they themselves are the factors, so that this role has no negative consequences in terms of votes for politicians. 2/ J. L. Migue and G. Belanger, "Towards a General Theory of Managerial Discretion," Public Choice XVII(Spring 1974):27-43. - 30 - are many ways to do that. One of the most plausible is for the bureau to obtain from its regulatory authority (politicians and top-level bureaucrats) some recognition in setting its budget of the "fair costs" of its operations, i.e., including such factors as prestige, power, direct and indirect revenues, etc., which the bureaucrats desire and can 1/ appropriate.- In such a case, the bureau will work as any regulated public utility that enjoys a "fair rate of return." That is, it will produce tariffs until the average cost intersects with the Marshallian demand for protection, DM, i.e., it will produce QM. It can be seen that QN > QM, a result that depends on the U-shaped cost curves. Indeed, empirical evidence suggests unambiguously that the average cost of enforcement increases in the relevant range. Consider now the politician's case. There are two scenarios. The first is that, unlike bureaucrats, politicians are able to capture some of the differences between the costs incurred and the services produced, although their right to claim the profits generated by their decisions is probably limited. They may receive personal gain, or secure campaign contributions or other types of support. This ability is reinforced by the fact that, as a rule, a lobbying industry will not invest in just any politician, but rather will choose "political stars" or at least "experienced" politicians. (This difference in their ability to capture quasi-rents is the only 1/ An alternative way would be to examine the "managerial efforts" of the bureau. This could be made within the framework suggested in Section II of J. P. Martin, "X-Inefficiency, Managerial Effort and Protection," Economica (45)(August 1978):273-286. - 31 - difference between politicians and bureaucrats considered here.) Provisionally, politicians are assumed to be monopolists. Although it is true that, with respect to voters, incumbent politicians are in a competitive market, the evidence suggests that, for "experienced" politicians, the level of competition is not as high as it is sometimes said to bed.' Consider first a politician who has bargaining power as a discriminating monopolist, so that the Marshallian demand curve, DM, is his marginal revenue curve. This politician produces a level of protection, PN 21 Clearly PN < QN. Thus, bureaus are more biased toward protection than politicians, who are discriminating monopolists. Now assume the second scenario -- a politician without strong bargaining power. He will act as a monopolist, producing at PM. When non-discriminating politicians and bureaus are compared, it is again apparent that bureaus are biased in favor of protection, since PM < QM- 1/ Long political careers might be seen either as the product of efficiency or a lack of competition. This second explanation cannot be ignored in Europe. The parties are highly organized, and a European politician who has not been elected as a representative can play an important role in national political life as long as he plays an important role in his party. Second, there are familiar techniques for suppressing competition between politicians belonging to the same party and who are at first sight close substitutes: they involve "electoral fiefs" or the assignment of distinct roles within the party. 2/ *For a similar problem, see R. G. Holcombe and E. 0. Price, "Optimality and the Institutional Structure of Bureaucracy," Public Choice l(33)(1978):55-59. - 32 - The conclusion, then, is that given identical conditions of discrimination, a bureau will choose a higher level of protection than will a politician. This is because of a basic (and unavoidable) characteristic of bureaus -- their inability to capture quasi-rents. Bureaus Vs. Politicians: The Case of Several Means of Protection So far it has been assumed that the politicians and bureaus in charge of commercial policy act as pure monopolists. However, it is clear that rivalry (or competition) sometimes exists among politicians, as well as among bureaucrats.1! Brock and Magee have interpreted the rivalry among politicians as generating competitive pressure. They have suggested some interesting results in the case of a duopolistic market for politicians. In particular, they showed that "the more competitive the political system, the more likely it is that distribution questions will dominate efficiency considerations." What is the nature of the rivalry between bureaus? In Chapter III, it was argued that a bureau is only a partial dictator in that it has a limited range of means. However, a bureau can always expand this range somewhat or use its means extensively in order to have some influence over problems which were "a priori" outside its jurisdiction. For instance, a bureau in charge of social affairs may lobby for an early retirement scheme in an import-competing industry, a benefit that could have a strong impact on commercial policy. For example, consider the first Davignon Plan (of 1978). To solve the 1/ A. Brock and S. Magee, "Tariff Formation in a Democracy," in J. Black and B. Hindley, eds., Current Issues in Commercial Policy and Diplomacy, New York: St. Martin's Press, 1980. - 33 - problem of the excess labor force, the bureaus of social affairs pushed the idea of a fifth team in steel plants. Obviously, this "fifth team" would have had a crucial impact on the EC's international steel policy by raising production costs, probably making it more protectionist. Thus one aspect of the rivalry among bureaus is attempts by new bureaus to enter into a problem involving one or several traditional bureaus by defining and using new means, or by using old means in a new way. Clearly, this entry is motivated by a desire to be involved in that problem. There is a large range of motives, among them, for example, to maximize prestige or secure the survival or growth of the bureau. However, the main point is the crucial role that the means available to a bureau play in the competition among bureaus. A bureau may also try to escape the competitive pressures by a policy of "differentiating" the product it produces; it will choose a l"specific" means of protection in order to increase its monopolistic power over the commercial policy it defines for the industry it supervises.l/ Obviously, such a policy of differentiation means that bureaus will be strongly motivated to prefer NTBs to tariffs, since monitoring of the former by others is more difficult.21 To illustrate this point, consider such NTBs as quality control for agricultural imports, the 1/ The word "specific" has the same significance as in the expression "specific" factors. 2/ For instance, subsidies given by bureaus in the Ministry of Industry can be easily monitored by bureaus in the Ministry of Finance (Treasury). Indeed, NTBs might be seen as the best means for the "technical" ministries (industry, transportation, agriculture, etc.) to escape the influence of the powerful Ministry of Finance. - 34 - definition of industrial norms, or public purchases. Such NTBs introduce enormous problems of measurement and evaluation that require technical skills not present within the bureaus trying to solve the problem, even the principal bureau, leaving room for discretionary judgements. Notice that this trend is not especially novel. Bureaus have understood since early on that ad valorem tariffs are more difficult to monitor than specific tariffs, since they are based on price evaluations for imported goods; those evaluations can be very controversial, requiring technical skills and information. The main consequence of this policy of differentiation may be the result pointed out earlier -- a bias of bureaus toward protection, a result that is more likely to occur with differentiated means of protection than without them. That outcome is attributable to the fact that differentiation is much easier for bureaus to work with than for politicians, since NTBs require both time and technical skills, resources that politicians cannot easily provide at a sufficient level. Thus, in addition to their well-known "qualities" from the external point of view -- namely, that NTBs are more difficult to control by other countries and are less subject to international agreements than tariffs are -- NTBs also present internal advantages, at least for bureaus. With respect to the initial problem -- the different impact that bureaucrats and politicians have on the level of protection -- some doubts may be raised about the validity of the usual theoretical - 35 - conclusion1/ that subsidies are preferable to tariffs. This conclusion depends on the assumption that the protection effects of the devices considered are equal. However, that assumption is in doubt if, for legal reasons, tariffs are more clearly the responsibility of politicians through legislative procedures than are other barriers to imports. It might be better to choose tariffs rather than subsidies, simply because they are less responsive to manipulation by bureaus, whose protectionist biases have been discussed. Are NTBs always favorable to the industry the bureau supervises? This question remains open. The industry may expect better control through NTBs than through a tariff system because the former involve technical observations and questions that really can only be provided by the firms or professional associations. This gives them some scope to control and orient the bureau. However, it should be stressed that the splintering of trade policies into tariffs and multiple NTBs makes it very difficult for firms to measure the effective rates of protection they face. The result may be errors of judgement, and even an over-investment in lobbying. In addition to this basic problem, it may be asked if the power of the industry to monitor and control NTB protection is not transitory. The industry cannot avoid giving crucial information to the bureau. Over time, the bureau will have accumulated substantial know- ledge and may become harder to convince. At some time the bureau may even discover and seek to rectify abuses in a way the industry dislikes. 1/ J. Bhagwati, "The Generalized Theory of Distortions and Welfare," in J. Bhagwati, et al., eds., Trade, Balance of Payments and Growth, Amsterdam: North Holland Publishing Co., 1971. - 36 - Chapter V INFORMATION COSTS AND THE BIAS TOWARD PROTECTION Information costs were suggested earlier as a basic explanation for protection.- However, the costs analyzed related specifically to those experienced by consumers or voters, and it was noted that consumers may not be able to relate observed increases in prices to protectionist measures. Even if they could do so, it might not be worth their while to oppose the measures if only a small fraction of their total expenditure is involved. A different issue was investigated here: do information costs increase or decrease the bias of bureaus toward protection? It is argued that they are likely to increase it, presumably strongly. Begin by considering the familiar theoretical analysis of a change in the tariff level, such as a decrease, in a partial equilibrium framework. The decrease will generate two opposing forces: gains for consumers, to be estimated by the present value of the familiar deadweight gains, and the social costs of decreased protection for the factors of production hired by the import-competing industry. How can these social costs be defined, if there is no awareness of the loss of tariff revenue? A first definition was suggested by Magee 2/ ae-. the social costs are to be estimated as the present value of "the costs of moving 1/ A. Downs, An Economic Theory of Democracy, New York: Harper and Row, 1957. 2/ S. Nagee, "The Welfare Effects of Restrictions on U.S. Trade," Brookings Papers on Economic Activity (3)(1972):660. - 37 - resources out of import-competing production and into other areas." Baldwin, Mutti and RichardsonL/ showed that the existence of adjustment costs (as defined by Magee) "does not necessarily result in a net welfare cost of tariff reduction to the economy." Casual observation suggests that bureaus generally use something like the Magee definition, which is easy to develop in a partial equilibrium framework. A bureau will favor freer trade only when the net marginal welfare gains (estimated in terms of a partial equilibrium) remain positive. As soon as they become negative, the bureau will favor protection. Figure 3 represents the situation as the bureau perceives it. At point Tel the net marginal welfare gains, assuming perfect information, are zero. The tariff change ATe will give a new equilibrium tariff, Tel to be imposed by the bureau; it is obtained by reducing the initial tariff T by a proportion ATe* This tariff change is optimal for the bureau since it maximizes the net welfare gains from trade liberalization&2/ Is Te optimal? This depends on whether the curves in Figure 3 are properly perceived by the bureau. 1/ R. E. Baldwin, J. H. Mutti and J. D. Richardson, "Welfare Effects on the United States of a Significant Multilateral Tariff Reduction," Journal of International Economics 10:405-423. 2/ Obviously, AO (i.e., a tariff change that results in a new tariff of zero) is the optimal tariff change in a general equilibrium framework for a small country. The tariff change ATe is the result of the limited rationality of the bureau and its willingness to organize compensation between the gainers and losers from free trade within a partial equilibrium framework. - 38 - Figure 3: EQUILIBRIUM LEVEL OF TARIFF CHANGES IN A PARTIAL EQUILIBRIUM FRAMEWORK (AO - the initial value of the tariff) Marginal net welfare gains (+) or losses TarI '-~~~~~~~I 0.) A / Tarif f O // // // - 39 - Bureaus as Information Producers Will the bureau produce directly its own estimates of the social gains and losses created by the tariff change? It may be that, because of practical problems, the costs of obtaining that information are high. Each firm does not use exactly the same technology, so that the effective rates of protection will be different for each. Further, protection induces smuggling, errors, briberies and fraud of all kinds. Thus, in practice the bureau may not be able to provide such estimates easily. The information costs therefore may not be negligible, and as they have to be deducted from the net welfare gains, the tariff, T', imposed by the bureau will be higher than the equilibrium tariff, Te. It is likely that a bureau can more easily collect data on the costs for factors of production than on the gains for consumers. As a result, their assessment of the latter will be too low, leading to an underestimation of the net welfare gains. Here, T' will be imposed instead of Te (that is, the tariff change AT' will be chosen instead of ATe) )/ 1/ A simple example may be given. During the late 19th and early 20th centuries, the debates between protectionists and free traders in France were dominated by the controversy about trade balances. As A. Arnaume ("Le Commerce Ext6rieur et les Tarifs de Douane" [Foreign Trade and Customs Tariffs], Paris: F. Alcan, 1911) reported, imports were probably slightly undervalued (3 percent), exports highly undervalued (5-15 percent). Consequently, over the 34 years between 1880 and 1913, France had, according to official statistics, between 11 and 14 years of trade surpluses instead of 1 year. (H. D. White, The French International Accounts (1880-1913), Harvard Economic Studies No. 40, Cambridge, Mass.: Harvard University Press, 1933). The French "Direction des Douanes" (DGD) (Customs - 40 - Bureau as Information Users The second and more likely possibility is that a bureau chooses to rely instead on estimates provided directly by the industries and persons involved, i.e., the factors of production and consumers. It is probable that, for a majority of goods, the amount of information a bureau obtains from consumers will be low and/or of poor quality. Consumers do not work as a group to estimate their gains and losses. In part this is because of the familiar free rider problem: as Olson'/ has shown, the probability of an organized consumer group is rather low. Another likelihood is that the bureau will ask for information about very specific products that represent only a small part of a consumer's total expenditures. As suggested by Downs 2/ it would be unwise for individual consumers to react to price changes which are unimportant to them. 1/ (cont.) Agency) never invested the time, money or effort to obtain accurate information on the movements of imports/exports, although costly fines for undervaluations were promulgated in 1863. One possible explanation is that, since no duties were generally levied on exports, only a minimal effort would have been expected from the civil servants. However, in addition, all those years that officially showed trade deficits (in reality trade surpluses) were between 1896 and 1909. Announcing the trade deficits was beneficial for the DGD, as it emphasized the necessity of greater watchfulness over imports, justifying higher budget appropriations for the DGD. Indeed, around 1900 the DGD published several reports strongly recommending more elaborate trade classifications (see Arnaume, op. cit., p. 39). 1/ M. Olson, The Logic of Collective Action, Cambridge, Mass.: Harvard University Press, 1965. 2/ Op. cit., Downs. - 41 - The quantity and quality of information will be superior where intermediate goods are involved if the group of consumers -- in this case, firms -- is small and if the goods involved are an important part of the firm's budget. In addition, better information can be expected from consumers if the bureau asks for information on many goods instead of on specific products. An example would be a basket of goods incorporating the main expenses of consumers, versus a single product. This is the situation when a tariff package resulting from a trade round such as the Tokyo Round is to be approved. The story is quite different where the factors of production are concerned. Presumably, organized groups such as producer federations and unions are quite able to estimate fairly easily the costs and benefits of changes in protection. The interesting problem this situation raises is that the "suppliers" of information -- the factors or their representives -- know the quality of the information, while the "buyer" -- the bureau -- does not. Obviously, it is very difficult for the bureau to monitor the information. This is the typical case of "asymmetrical information." _/ 1/ Further, bureaus with a strong "survival" objective will not necessarily try to monitor the information seriously. As a result, they will impose a tariff level much higher than the level implied by perfect information, T . An extraordinary counterexample is the case of the "Conseil Superieur du Commerce" of 1860, which was in charge of determining the new (lower) tariff rates. It had a very peculiar feature in that it was an "ad hoc" transitory committee without any survival objective. This council undertook hearings for French entrepreneurs but monitored their assertions by holding hearings for English entrepreneurs as well (see A. Dunham, The Anglo-French Treaty of Commerce of 1860 and the Progress of the Industrial Revolution in France, New York: Octagon Books, 1930). - 42 - To illustrate this point, assume that a bureau wants to estimate the social costs. This may be easier to achieve for the mobile factors, for instance, unskilled labor, than for the factors specific to the import-competing industry, where it is far more difficult to separate "rents" from returns, to estimate real returns and/or the average length of unemployment, etc.i' This may lead the bureau to accept some overestimations of the costs for labor to offset the suspected undervaluation of the costs for the other factors employed by the import-competing industry 2/ Obviously, any compromise of this kind will distort T' from Te As a result of the difficulty in monitoring the information produced by the factors of production, a bureau will be induced to encourage some competition between the sources of information, or try to verify the information provided (call on consulting firms, for instance). A variant of the first would be competition between bureaus themselves (if they do not agree on the diagnosis of the trade problem) or a cross-check using information supplied by other organizations. 1/ Notice that the estimates made by academics generally include only the "labor value cost," which has to be interpreted as the costs for the mobile factors. See, for example, Magee, op. cit., and M. Szenberg, J. W. Lombardi and E. Y. Lee, Welfare Effects of Trade Restrictions: A Case Study of the U.S. Footwear Industry, New York: Academic Press, 1977. 2/ Such behavior by bureaus is indeed favorable to the lobbies, since it makes it easier to present interest groups as having "social" rather than egotistical motives: capitalists and skilled workers seem to take care of the non-skilled workers' future. The hypothesis suggested here may explain the oft-alleged "labor-bias" of bureaus: this bias would be an indirect way of allowing far less for the other (non-labor) factors. - 43 - However, competition between the sources of information is not necessarily possible. Consequently, the second means appears to be the most common. A frequent device for upgrading information is to license a chosen professional association in the industry as the sole source of information for the bureau. This has two familiar characteristics: a monopoly and a long-term contract, so that any false information given by the association would get negative returns sooner or later. However, some bias is bound to remain, and the possibility of "capture of the bureau" by those who benefit from protection increases. - 44 - Chapter VI CONCLUSIONS To summarize, several findings have emerged concerning protection by bureaus. First, if the bureau supervising a given industry is a partial dictator, it will choose a protective level higher than the theoretically optimal one. Second, if, as expected, bureau- crats have fewer opportunities than politicians to capture the profits (wealth) resulting from protective activities, they will favor higher protection. Third, information costs will probably increase a bureau's bias toward protection. All these points lead to a basic conclusion -- any ranking among protective devices which does not fully identify the decision-makers who will implement them. 1, As regards institutions, the important finding is that administrative structure may have a crucial impact on a bureau's inclination toward free trade or protection. This could occur in several ways. It could change the results of the partial dictatorship (Chapter III), modify the conditions of competition between bureaus in the area of protective devices (Chapter IV) and, finally, determine the information costs and the "property rights" of bureaus and industries concerning trade policies (Chapter V). The following conclusions follow from these points. First, the broader the scope of a bureau's responsibilities with respect to 1/ Some empirical evidence seems, at a first glance, to corroborate this result. Deviations from the trend of trade liberalization occurred in France during periods when the bureaus were considered to be powerful. - 45 - protection, the more likely it is to favor free trade. Examples can be seen in the contrast between the EC Commission and national bureaucracies or between the West German and state bureaus (although this second case appears to have become more ambiguous in the last two years). Second, information about consumers' losses resulting frot protection can be correctly produced only by institutions which can examine the whole range of final goods and which are not limited to some category of goods. This is the only way to overcome the free rider problem noticed by Downsl/ and to create a bureaucratic organization which can be free trade-oriented. The recent evolution of the Australian Industries Assistance Commission,-/ and perhaps, of the U.S. International Trade Commission, are interesting examples of such a development. Finally, more effort should be devoted to studying the relations between trade policy and administrative structure. 1/ Op. cit., Downs. 2/ K. Anderson, "The Political Market for Government Assistance to Australian Manufacturing Industries," Economic Record 56(153) (June 1980): 132-144. - 46 - REFERENCES Anderson, K. "The Political Market for Government Assistance to Australian Manufacturing Industries." Economic Record 56(153) (June 1980): 132-144. Arnaume, A. Le Commerce Ext&rieur et les Tarifs de Douane [Foreign Trade and Customs Tariffs]. Paris: F. Alcan, 1911. Baldwin, R. E., J. H. Mutti and J. D. Richardson. "Welfare Effects on the United States of a Significant Multilateral Tariff Reduction." Journal of International Economics 10:405-423. Bhagwati, J. "The Generalized Theory of Distortions and Welfare." In Trade, Balance of Payments and Growth. Edited by J. Bhagwati, et al. Amsterdam: North Holland Publishing Company, 1971. Breton, A. The Economic Theory of Representative Government. Chicago: Aldine Publishing Company, 1974. Brock, A. and S. Magee. "Tariff Formation in a Democracy." In Current Issues in Commercial Policy and Diplomacy. Edited by J. Black and B. Hindley. New York: St. Martin's Press, 1980. Buchanan, J. "The Coase Theorem and the Theory of the State." Natural Resources Journal 13(4)(October 1973):579-594. Corden, W. M. Trade Policy and Economic Welfare. Oxford: Clarendon Press, 1974. Downs, A. An Economic Theory of Democracy. New York: Harper and Row, 1957. Dunham, A. The Anglo-French Treaty of Commerce of 1860 and the Progress of the Industrial Revolution in France. New York: Octagon Books, 1930. Finger, J. M., H. K. Hall and D. R. Nelson. "The Political Economy of Administered Protection." American Economic Review 72(3) (June 1982): 452-466. Holcombe, R. G. and E. 0. Price. "Optimality and the Institutional Structure of Bureaucracy." Public Choice 1(33)(1978):55-59. Krueger, A. "The Political Economy of the Rent-Seeking Society." American Economic Review (June 1974):291-303. Magee, S. "The Welfare Effects of Restrictions on U.S. Trade." Brookings Papers on Economic Activity (3)(1972):645-707. - 47 - Martin, J. P. "X-Inefficiency, Managerial Effort and Protection." Economica 45(August 1978):273-286. Migu6, J. L. and G. Belanger. "Towards a General Theory of Managerial Discretion." Public Choice XVII(Spring 1974):27-43. Niskanen, W. Bureaucracy and Representative Government. Chicago: Aldine-Atherton, 1971. Olson, M. The Logic of Collective Action. Cambridge, Mass.: Harvard University Press, 1965. Pisani, E. 1979. "Rapport d'Information sur les Enjeux et Conditions des Equilibres Exterieurs de la France" [Information Report on the Stakes and Conditions of France's External Equilibria]. Senat [Senate]. Paris, 1979. Posner, G. J. "The Theory of Economic Regulation." Bell Journal of Economics and Management Science (Spring 1971):22-50. Rapoport, A. and A. M. Chammah. Prisoner's Dilemma. Ann Arbor, Mich.: University of Michigan Press, 1965. Stigler, G. J. "The Theory of Economic Regulation." Bell Journal of Economics and Management Science (Spring 1971):3-21. Szenberg, M., J. W. Lombardi and E. Y. Lee. Welfare Effects of Trade Restrictions: A Case Study of the U.S. Footwear Industry. New York: Academic Press, 1977. Waelbroeck, J. and E. Verreydt. "European Community Protection Against Manufactured Imports from Developing Countries: A Case Study in the Political Economy of Protection." In Adjustment and Protection. Edited by J. Bhagwati. Chicago: University of Chicago Press, 1980. * World Bank Publications of Related Interest Adjustment Policies and Britain's Pattern of Special- Capital-Importing Oil Problems in Developed ization in Manufactured Exporters: Adjustment Countries Goods with Developing Issues and Policy Choices Martin Wolf 'Countries and Trade Alan H. Gelb World Bank Staff Working Paper No. Protection A background study for World 349. August 1979. 236 pages (includ- Vincent Cable and Development Report 1981. Uses a ing references). Ivonia Rebelo simple two-sector model involving big references). Rebelo ~~~~~~~~traded and nontraded goods as a Stock No. WP-0349. $10.00. World Bank Staff Working Paper conceptual framework to compare the ,No. 425. October 1980. 61 pages evolution of critical macro and sec- (including 3 appendixes). toral variables for a number of oil Adjustment to External Stock No. WP-0425. $300 economies after 1974 and discusses Shocks. in Developing govemment responses to the oil Economies crisis and the effects of these Bela Balassa iresponses of the nonoil economies. A background study for World World Bank Staff Working Paper No. Development Report 1981. Analyzes 475. August 1981. 38 pages (including adjustments to external shocks, in 9 tables). the form of changes in the terms of Stock No. WP-0475. $3.00. trade and the slowdown in foreign export demand, in twenty-eight developing economies, classified according to the character of extemal shocks, the level of industrial devel- opment, and the policies applied. World Bank Staff Working Paper No. 472. July 1981. 31 pages (including appendix). Stock No. WP-04 72. $3.00. NEBW The Export Experience of Oxford University Press. 1982. Developing Countries 224 pages (including index). The Developing Countries Barend A. de Vries LC 82-6309. ISBN 0-19-520211-2, and International Shipping The English-language edition is $22.50 (W050) hardcover. (A specially Harald Hansen out of print. priced edition will be available in India from Ox,ford Universiti Press branches.) Considers whether developing coun- French: Resultats obtenus en matiere tries can benefit from investments in d'exportation par les pays en voie de intemational shipping and discusses developpement. Dunod Editeur, 24-26, Industrial Country Policy the circumstances under which such boulevard de l'Hopital, 75005 Paris, and Adjustment to Imports investment might be favorable. France. 1969. from Developing Countries World Bank Staff Working Paper No. LC 67-28942, 10 francs. J. M. Finger 502. November 1981. iii + 148 pages Saih aeprecad o ass Abcgon td o ol (including 12 annexes, 38 tables, Spanish: L expenencia de los paises A.acgountud or Rld bibliography). en desarrollo en materia de exporta- Development Report 1981. Reviews ciones. Editorial Tecnos, 1969. and interprets recent analyses of the Stock No. WP-0502. $5.00. policies established by industrial 200 pesetas. countries in response to increasing imports from developing countries. Effects of Nulon-Tariff B8XS Export Promotion Policies World Bank Staff Working Paper No. riers to Trade on Prices, Barend A. de Vries 470. July 1981. ii + 20 pages (includ- TEmpCasenofthe SwedImprsh World Bank Staff Working Paper No. ing references). The Case of the Swedish 313. January 1979. v + 75 pages. Stock No. WP-04 70. $3.00. Textile and Clothing Industry Stock No. WP-0313. $3.00. Carl Hamilton Italian Commercial Policies World Bank Staff Working Paper No. NEW In the 1970s 429. October1980. ii + 61 pages Enzo R. Grilli (including appendLx, bibliography). On Exports and World Bank Staff Working Paper No. Stock No. WP-0429. $3.00. Economic Growth 428. October 1980. 47 pages. Gershon Feder Stock No. WP-0428. $3.00. Energy, International Trade, An analytical framework is developed Energy, International TlraLde, to analyze the sources of growth dur- and Economic Growth ing the period 1964-73 for a group of On the Political Economy of Alan S. Manne and semi-industrialized developing coun- Protection in Germany Sehun Kim tries. Discusses the relationship H. H. Glismann and A background study for World between export performance and 1 D. Weiss Developmnt Repot 1981.economic growth and concludes that Development Report 1981. growth can be generated not only by World Bank Staff Working Paper No. Constructs a small-scale intemational increases in the aggregate levels of 427. October 1980. 30 pages (includ- trade model that focuses on issues labor and capital but also by the ing bibliography). related to energy and economic reallocation of existing resources growth in order to determine the from the less-efficient nonexport sec- Stock No. WP-0427. $3.00. extent to which increasing energy tor to the higher-productivity export prices impose constraints on sector. economic growth. * Output and Employment World Bank Staff Working Paper No. World Bank Staff Working Paper No. Changes in a "Trade Sensi- 4 74. August 1981. 30 pages (including 508. fpebrua,rye192.r2npaes)(. cu tive" Sector: Adjustment in 2 appendixes, references). ing appendix. references). the U.S. Footwear Industry Stock No. WP-0474. $3.00. Stock No. WP-0508. $3.00. John H. Mutti and Malcolm D. Bale European Community NEW World Bank Staff Working Paper No. Protection against 430. October 1980. 21 pages Manufactured Imports India's Exports (including footnotes, references). Manoufactured Imports Martin Wolf Stock No. WP-0430. $3.00. A Case Study in the Pounti Despite improved performance, the growth of India's exports continues to cal Economy of Protection lag behind need, potential, and the E. Verreydt and achievements of several of its com- J. Waelbroeck petitors. This study examines India's overall export performance in the World Bank Staff Working Paper No. 1960s and 1970s, with emphasis on 432. October 1980. 25 pages. the central role of incentives. The Stock No. WP-0432. $3.00. major problems and policies are discussed, -as well as current strategic options. NEW Prospects for Partnership: The Structure of Protection Industrialization and Trade in Developing Countries Patterns of Barriers to Policies in the 1970s Bela Balassa and others Trade in Sweden: A Study in Helen Hughes, editor The Johns Hopkins University Press, the Theory of Protection The Johns Hopkins University Press, 1971, 394 pages (including 5 appen- Lars Lundberg 1973. 310 pages. dixes, index). This report is part of an inquiry being LC 72-12369. ISBN 0-8018-1498-7, LC 77-147366. ISBN 0-8018-1257-7, undertaken by the World Bank in con- $20.00 (£12.00) hardcover. $25.00 (i15.00) hardcover junction with scholars from twelve industrial countries into the penetra- ISBN 0-8018-1500-2, $6.95 (£4.25) Spanish: La estructura de la protecci6n tion of the markets of industrial paperback. en paises en desarrollo. CEMLA, Depar- countries by exports of manufactures tamento de Publicaciones, Durango 54. from developing countries. Spanish: Las perspectivas del comercio Mexico 7, D.F., Mexco. 1972. internacional: industrializaci6n y World Bank Staff Working Paper No. politicas comerciales en la decada de 494. October1981. 35 pages (including los 70. Editorial Tecnos, 1974. The Tokyo Round: Results 3 appendixes). ISBN 84-309-0497-2, 575 pesetas. and Implications for Stock No. WP-0494. $3.00. Developing Countries Ria Kemper The Political Economy of NEW World Bank Staff Working Paper No. Protection in Belgium O Prtcins in372. February 1980. iii + 35 pages R K. M. Tharakan nOn Protectionism in (icluding annex). the fNetherlands Stock No. WP-03 72. $3.00. World Bank Staff Working Paper K.A. Koekkoek, J. Kol, and No. 431. October 1980. 22 pages L.B.M. Mennes (including statistical appendix, This report is part of an inquiry being Trade A4justment Folides references). undertaken by the World Bank in con- and Income Distribution In Stock No. WP-0431. $3.00. junction with scholars from twelve Three Archetype Developing industrial countries into the penetra- Economies tion of the markets of industrial Jaime de Melo and NEW countries by exports of manufactures Sheman oban from developing countries. Sherman Robinson The Political Market for World Bank Staff Working Paper No. World Bank Staff Working Paper No. Protection in Industrial 493. October 1981. ii + 68 pages 442. December 1980. 91 pages Countries: Empirical (including 3 annexes, references). (including appendixes, references). Evidence Stock No. WP-0493. $3.00. Stock No. WP-0442. $3.00. Kym Anderson and Robert E. Baldwin Trade among Developing This report is part of an inquiry being NEW Countries: Theory, undertaken by the World Bank in con- Policy Issues, and junction with scholars from twelve Shadow Prices for Trade Polic Isu Trends industrial countries into the penetra- Strategy and Investment ncipa tion of the markets of industrial Planning m Egypt Oli Hiavrylyshyn and countries by exports of manufactures John Page, Jr. Martin Wolf from developing countries. This paper presents estimates of effi- A background study for World World Bank Staff Working Paper No. ciency and social accounting prices Development Report 1981. Presents 492. October 1981. 28 pages (including for commodities and factors of pro- the results of empirical work on trade references). duction in Egypt that are appropriate among developing countries. Based for the eriod 199-1985.on data derived from a sample of Stock No. WP-0492. $3.00. for the period 1979-1985. thirty-three developing countries World Bank Staff Working Paper No. that account for about 60 percent of 521. October 1982. 212 pages. developing countries' exports to ISBN 0-8213-0009-1. $5.00. one another. World Bank Staff Working Paper No. 479. August 1981. iv + 112 pages Structural Change in Trade (including 2 appendixes, references). in Manufactured Goods Stock No. WP-04 79. $5.00, between Industrial and Developing Countries Bela Balassa World Bank Staff Working Paper No. 396. June 1980. 46 pages. Stock No. WP-0396. $3.00. IYEW Trade Policy for Developing REAr'S Countries Can Developlig-Country Exports Keep Donald B. lieesing Growing in the 1980s? Trade and Employment Donald B. Keesing Helen Hughes and Jean Waelbroeck Policies for Industrial World Bank Staff Working Paper No. World Bank Reprint Series: rlumber 194. Development 353. August 19 79. vii + 264 pages Reprinted from The World Economy (June Keith Marsden (including references). 1981).127-47. In the last decade, the developing Stock No. WP-0353. $10.00. Stock No. RP-0194. Free of charge. countries have proved that they can compete internationally in exporting Questions on International Trade In manufactured goods, as well as pri- Trade Policy Issues for the TestWles and Clothing mary products and services. This Developing Countries in Donald B. Keesing and Martin Wolf paper examines three sets of issues: the 1980s World Bank Reprint Series: Nwumber 202. (a) whether good export performance Reprinted from The World Economy, volt. 4 (March is attributable to special characteris- Isaiah Frank 1981):79-101. tics of the most successful countries Explores the relation between trade Stock /Yo. RP-0202. free of charge. or whether their success can be policy and "industrial policy" and readily replicated in other countries; calls attention to points of conflict (b) whether the penetration of the and compatability. Reviews the Trade Policy and Resource markets of industrial countries has /recently completed multilateral trade Allocation In the Presence of Product reached, or will soon reach, a limit; negotiations and assesses the policy Differentiation and (c) whether trade in manufactures significance for the developing coun- Jaime de Melo and Sherrnan Robinson among the developing countries can tries during the 1980s. Takes a new Reprinted from The Review of Economics and expand further. Concludes with a look at industrial policy and struc- Statistics, vol. 63, no. 2 (May 1981).169-177. discussion of the contribution of tural adjustment, fair labor stan- Stock No. RP-0214. Free of charge. small enterprises to the creation of dards, trade among the developing employment and the alleviation of countries, and trade in services. poverty, World Bank Staff Working Paper No. 1982. vi + 64 pages (including annex). 478. August 1981. 52 pages. ISBN 0-8213-0017-2. $5.00. Stock No. WP-0478. $3.00. Trade in Non-Factor Worker Adjustment to Services: Past Trends and Liberalized Trade: Costs Current Issues and Assistance Policies Andre Sapir and Graham Glenday, Ernst Lutz Glenn P Jenkins, and World Bank Staff Working Paper No. John C. Evans 410. A ugust 1980. iii + 13 7 pages World Bank Staff Working Paper. No. (including 4 annexes). 426. October 1980. i + 86 pages Stock No. WP-0410. $5.00. (including 2 appendixes, bibliography). Trade in Services: Economic Stock N WP0426. $5.00. Determinants and Develop- ment-Related Issues World Trade and Output of Andre Sapir and Manufactures: Structural Ernst Lutz Trends and Developing A background study for World Countries' Exports Development Report 1981. Finds that Donald B. Keesing trade theories can help explain the World Bank Staff Working Paper No. patterns of trade in services in spite 316. January 1979. v + 69 pages of varying and often substantial (including statistical annex). degrees of protectionism. Represents (nldn ttsia ne) the second stage of a research Stock No. WP-0316. $3.00. project on trade in services. World Bank Staff Working Paper No. 480. August 1981. 38 pages (including appendix, references). Stock No. WP-0480. $3.00. WORLD BANK PUBLICATIONS ORDER FORM SEND TO: WORLD BANK PUBLICATIONS WORLD BANK PUBLICATIONS PO. BOX 37525 or 66, AVENUE D'IENA WASHINGTON, D.C. 20013 75116 PARIS, FRANCE U.S.A. Name: Address: Stock or ISBN # Author, Title Qty. Price Total Sub-Total Cost: Postage & handling fee for more than two free items ($1.00 each): Total copies: Air mail surcharge ($2.00 each): TOTAL PAYMENT ENCLOSED: Make checks payable: WORLD BANK PUBLICATIONS Prepayment on orders from individuals is requested. Purchase orders are accepted from booksellers, library suppliers, libraries, and institutions. All prices include cost of postage by the least expensive means. The prices and publication dates quoted in this Catalog are subject to change without notice. No refunds will be given for items that cannot be filled. Credit will be applied towards future orders. No more than two free publications will be provided without charge. Requests for additional copies will be filled at a charge of US $1.00 per copy to cover handling and postage costs. Airmail delivery will require a prepayment of US $2.00 per copy. Mail-order payment to the World Bank need not be in U.S. dollars, but the amount remitted must be at the rate of exchange on the day the order is placed. The World Bank will also accept Unesco coupons. The World Bank Headquarters European Office Tokyo Office U 1818 H Street, N.W. 66, avenue d'Iena Kokusai Building Washington, D-C. 20433, U-SA 75116 Paris, France 1-1 Marunouchi 3-chome C Telephone: (202) 477-1234 Telephone: (1) 723-54.21 Chiyoda-ku, Tokyo 100, Japan Telex: WUI 64145 WORLDBANK Telex: 842-620628 Telephone: (03) 214-5001 RCA 248423 WORLDBK Telex: 781-26838 Cable Address: INTBAFRAD WASHINGTONDC L ISSN 0253-2115/ISBN 0-8213-0198-5