Tariffs, Quotas, and the Corrupt Purchasing of Inappropriate Technology

Transcription

1 nternational Journal of Business an Economics, 25, Vol. 4, No. 1, 1-9 Tariffs, uotas, an the Corrupt Purchasing of nappropriate Technology Neil Campbell Department of Applie an nternational Economics, Massey University, New Zealan Abstract This paper evelops a simple moel where a manager of a firm in a Less-Develope Country (LDC) has the choice of whether or not to purchase an inappropriate technology in return for a bribe (kick-back) from the supplier of the technology. Provie that the manager achieves some imum level of profit, the manager has a positive probability of not getting caught taking the bribe. The actual size of the bribe is etere by Nash axiomatic bargaining between the manager an the supplier. An interesting an not immeiately obvious result is that, uner certain circumstances, if the protective instrument is change from a quota to an equivalent tariff the manager will switch from not acting corruptly to acting corruptly. Key wors: kick-backs; corruption; managerial iscretion; borer protection JEL classification: F13; F23; L21 1. ntrouction A well-known general conclusion from the international trae literature is that a tariff is preferable to the equivalent import quota. For a book-length treatment of the relative inefficiency of quotas see Anerson (1988). One of the least attractive aspects of quotas is the link with corruption. n her famous article Krueger (1974) iscusses the rent-seeking activities associate with the quota rents. She makes it clear that rent-seeking can inclue illegal activities involving corruption. From this it woul be natural to conclue that a shift from quotas to tariffs woul necessarily result in a reuction in corruption. n this paper argue that simply switching to tariffs, rather than reucing overall protection, may not necessarily be helpful in reucing all forms of corruption. Receive April 12, 24, revise April 13, 25, accepte April 15, 25. Corresponence to: Department of Applie an nternational Economics, Massey University, Private Bag 11222, Palmerston North, New Zealan. n.a.campbell@massey.ac.nz. The author wishes to thank Martin Richarson, Kunal Sen, Jong-Shen Wei, two anonymous referees, an participants at a University of Canterbury epartmental sear for helpful comments.

2 2 nternational Journal of Business an Economics The efinition of corruption use here inclues the private sector as well as the public sector. The more commonly iscusse corruption in the public sector is generally taken to refer to both theft of public assets by public officials an public officials accepting bribes in return for taking, or not taking, certain actions. Private sector corrupt behaviour is taken to refer to a situation where a ecision maker within a firm makes a particular choice as a result of being bribe that is not in the interest of the owner(s). The situation consiere here is that of a manager of an import-competing firm locate in a Less-Develope Country (LDC). The manager has to choose whether or not to purchase an inappropriate technology in return for a bribe (kick-back) from the foreign supplier of the technology. While the economics of corruption literature has grown substantially over the last ecae or so, see Jain (21), it has not focuse heavily on the corrupt purchasing of technology. One exception to this is containe in the masterly iscussion of corruption issues by Shleifer an Vishny (1993). They give the example of a bottle-making factory in Mozambique. The issue was the replacement of three archaic bottle-labelling machines with one moern bottle-labelling machine. The sensible choice woul have been a machine that cost about $1,. This type of machine coul have been purchase from any one of a large number of foreign suppliers using foreign ai money. However, the manager of the factory wante to buy a sophisticate $1, machine that was only available from a single foreign supplier. Shleifer an Vishny (1993) explain this perverse choice in terms of the greater opportunity for corruption when the machine is unique rather than generic. That is, the honest price of the unique sophisticate machine is more or less what the supplier says it is because there are not irect competitors. Thus, the supplier coul charge a price of $13, an then secretly pay the manager $3, as a rewar for having chosen to purchase this machine. Shleifer an Vishny use this example to make the point that the esire for secrecy associate with corruption can lea to ramatic misallocations of resources. This bottle factory story provies us with a possible neat explanation for the folk belief that a substantial problem for LDCs is the use of inappropriate technologies, see Stiglitz (1988, p. 149). A number of authors affiliate with the MF have empirically investigate the iea that resources are misallocate so as to allow the ecision maker to inulge in corrupt behaviour. Specifically they consier whether in relatively corrupt societies government expeniture ecisions are istorte towars items like military aircraft an consequently away from items like eucation. Mauro (1998) performe a crosssectional multi-country stuy where corruption was measure by a corruption inex evise by a private firm calle Political Risk Services, nc. Both government expeniture on eucation as a proportion of GDP an government expeniture on eucation as a proportion of total government expeniture were trie as epenent variables. n both cases corruption was shown to be a significant explanatory variable with a negative relationship to government expeniture on eucation. Mauro (1998) explains this result in terms of government isters an officials having limite opportunities to obtain kick-backs from the suppliers of eucation sector inputs. This is sai to be because typically this sector oes not require inputs

3 Neil Campbell 3 that are high-technology an supplie by oligopolists. While personal computers may be regare as high-technology, there is not the same opportunity to secretly inflate the price as there is with some customise high-technology weapons system. Also with regar to computers in schools the reference year for this stuy was 1991, presumably when there were limite purchasing opportunities for computers in schools in LDCs. n a similar stuy, Gupta et al. (21) use cross-sectional an panel regression techniques to show a positive association between corruption an military expeniture. The methoology use in this paper is reasonably stanar; see Aes an Di Tella (1997). The size of the bribe is etere by axiomatic Nash bargaining between the payer of the bribe (the foreign supplier) an the accepter of the bribe (the manager of the omestic firm). f the manager ecies to act corruptly an accept the bribe then there is a probability that the manager will be caught an incur the associate cost. The new technology will reuce the omestic firm s marginal cost but not sufficiently to ever offset the cost of the investment. Thus the ecision to make the investment is synonymous with eciing to act corruptly. While, in the example of the Mozambique bottle factory there is a profit-improving new technology available to aopt, incluing a goo new technology woul not a anything to this analysis. The reason the manager stans a chance of getting away with corruption is because his performance is imperfectly monitore by the owners. The type of situation being consiere is the traitional story from the managerial theories of the firm literature; see Williamson (1964). n this literature the owners o not have the ability an/or the incentive to actively monitor the various iniviual actions of the manager. Rather, the owners only take action against the manager if profit falls below some imum level. This imum profit limit woul presumably be etere by some rough comparison to the profitability of some imperfectly comparable firms. n this literature the stanar explanation for this substantial separation between ownership an control is that the owners are numerous an only have a small proportion of their total wealth investe in one particular firm. However, in the context of a LDC this ownership structure is not particularly plausible. A more appealing thir worl explanation for imperfect monitoring of the manager woul be state ownership of the firm. State ownership of the firm fits in comfortably with the use of the stanar assumption, use in the literature on public sector corruption, that both the cost of being caught an the probability of being caught are exogenous. With state ownership there is typically little concern by public officials an politicians about returns an managerial ecisions unless isastrous results occur that become a political issue. Hence the cost an probability of being caught are the result of arbitrary circumstances rather than some sophisticate incentive scheme. The literature on state-owne firms emphasises the soft nature of the buget constraint; see Megginson an Netter (21, p. 331). There is a imum profit constraint being use but this is very much to be thought of as soft because before it bites there can be substantial scope for the manager to pursue his own agena.

4 4 nternational Journal of Business an Economics f the manager oes ecie that the firm will buy the new technology, he will want the firm to pay as high a price as possible so that the supplier s surplus, from which the bribe comes, is as large as possible. The highest price the manager can set is the ifference between the firm s profit, gross of the price (the investment cost), an the imum profit constraint. Obviously a price that was any higher than this woul break the imum profit constraint. So the moel this paper evelops makes clear that the size of the potential bribe is epenent upon the size of the profit gross of investment cost. Given that there is a probability of being caught an there is a cost associate with this, the higher the size of the profit gross of investment cost, the stronger is the tenency for the manager to act corruptly. So just as protection can encourage rent-seeking behaviour, it can also encourage managers to corruptly purchase inappropriate new technology. Once the basic framework has been set out in Section 2, it is use to show in Section 3 that merely changing from quotas to equivalent tariffs will not eter the manager from choosing to accept bribes. That is, in Section 3, it is shown that, for certain values of the cost of being caught, the manager will choose not to accept the bribe if the firm is protecte by a quota but will accept the bribe if the firm is protecte by a tariff. Thus changing from a quota to the equivalent tariff will never inuce the manager to switch away from taking the bribe. 2. The Basic Moel Consier an inustry where there is a single omestic firm. This firm competes (Cournot fashion) with imports from a single foreign firm. The inverse eman function which both of these firms face is p = α β ( x + y), where x an y refer respectively to the omestically prouce an foreign prouce quantities sol. Both the omestic an foreign firms have constant unit costs which are respectively enote c an c. Now there is a foreign supplier that can provie the omestic firm with a piece of capital equipment that will reuce its marginal cost from c to c 1. The cost the supplier incurs in proucing this equipment is enote C S. The price pai by the omestic firm for this equipment is enote. We are intereste in the case where making this investment will never represent profit-maximising behaviour an will only occur because of corrupt behaviour on the part of the manager. Thus we assume that even if the omestic firm only pai C S for the piece of equipment it woul not be worth it. The manager of the omestic firm is pai a fixe wage that is normalise to zero. The manager has iscretion how he behaves provie the firm s profit oes not fall below π. f the manager is caught taking a bribe b, instea of receiving the bribe, the manager incurs the cost m. The probability of being caught is enote θ. The bribe uner consieration is a kick-back from the supplier of the capital equipment. The bribe is pai so as to rewar the manager for having the firm purchase the piece of capital equipment at a high price. The constraint on how high this price can be set is, effectively, the omestic firm s imum profit constraint.

5 Neil Campbell 5 Hence we can write the price as = π, where is the omestic firm s profit gross of investment costs. The total surplus which the supplier woul have prior to paying the bribe is Z = CS. How much of this is pai to the manager of the omestic firm in the form of a bribe is etere by Nash bargaining between the manager an the supplier. The Nash bargaining problem can be written as or max [(1 θ ) b θ m][ C b] b max [(1 θ ) b θm][ π C b]. b S Rearranging the first-orer conition gives us an expression for the bribe: S ( 1 θ )( π C ) + θ m S b =. (1) 2(1 θ ) Clearly we are intereste in what the conitions are for the manager not being corrupt. Obviously if m is sufficiently high, the expecte cost of being caught offsets the expecte benefit from the bribe. t is simple to fin m, the level of m that leaves the manager inifferent between making the investment (corruption) an not making the investment (non-corruption): (1 θ ) b θ m = (1 θ) (1 θ) m = ( π CS) = ( CS). (2) θ θ Here m is large enough to offset the expecte value of being pai a bribe equal to the supplier s total surplus. Thus, if m is larger than m, the manager will efinitely not make the investment even if the supplier is prepare to pay a bribe equal to its total surplus. 3. The uota versus Tariff ssue We now use our framework to show that, if a quota rather than a tariff is use to assist the omestic firm, this can affect whether the manager engages in corrupt behaviour. To compare these two alternative instruments, the tariff is set so that the omestic firm s level of output woul be equal to, the maximum volume of imports allowe if a quota is use. That is, the tariff an quota are equivalent if no cost-reucing investment is mae. They are equivalent in the sense that the quantities sol, the market price, an the profits are the same no-matter which instrument is use. f the cost-reucing investment is mae, then the profit associate with the tariff case will be higher. This result is looke at in more etail

6 6 nternational Journal of Business an Economics in Campbell (1998). To show it we begin by consiering the omestic firm s problem when there is a quota in place: max = x[ α β( x+ )] c x. x We are assug that the quota is bining an hence the foreign firm s best response is to set y =. Rearranging the first-orer conition we get the following expression for output in terms of the parameters: x α β c =. (3) 2β f a tariff is use we can obtain the expression for output an profit by simply solving the stanar Cournot problem (see the appenix) to obtain: α 2c + c + t x t = 3β (4) 2 t ( α 2c + c + t) =. 9β (5) Here t refers to the tariff levie on each unit importe. We equate (3) an (4) so as to obtain an expression for the tariff which results in a quota-equivalent output: α 3β + c 2c t = 2. (6) Substituting (6) into (5) gives the profit associate with this tariff level: 2 ( α β c ) =. (7) 4β Now we go on to show that if the piece of capital equipment is purchase, then t > assug neither nor t were change. Let us start with 2 ( α 2 ) t c + c + t =. (8) 9β Here we nee to remember t stays at the amount shown in expression (6). So we substitute (6) into (8): 2 α β t (3 4c + c 3 ) =. (9) 36β To compare with this we write it in the following manner:

7 Neil Campbell 7 2 α β (3 3c 3 ) =. (1) 36β t Now 4c + c > 3c since c > c. Therefore we can say that >. f we look back at (2) it is easy to see that m coul be at a level such that: where m < m < m, m m t t (1 θ ) = ( π C ) S θ (1 θ ) t = ( π CS ). θ Thus we have a result that at first glance may seem counterintuitive. That is, when a quota is use, a lower m (cost of being caught) is require to eter corrupt behaviour by the manager. When thought about carefully it is clear enough how this result comes about. The corrupt behaviour being consiere is the manager being inuce to spen a large amount on an unprofitable investment that nevertheless reuces marginal cost. f this investment is mae, the firm s profit (gross of the investment cost ) will be lower if the protective instrument use is a quota rather than a tariff. Thus with the quota, the amount pai to the supplier (the investment cost ) will be lower. Remember the constraint on is the requirement that the manager satisfies the firm s imum profit constraint π. The lower associate with the quota means that the supplier s surplus Z = CS will be smaller. Therefore the maximum amount the supplier is prepare to bribe the manager is smaller, an hence the cost of being caught, necessary to eter corrupt behaviour, is also smaller. This result is, of course, epenent on the assumption of a nonchanging tariff an quota. This is a stanar assumption use when emonstrating the ynamic non-equivalence of tariffs an quotas; see Vousen (199, pp ). t is certainly a non-controversial assumption since it seems highly unlikely that a policy maker woul have the information an sophistication to ajust a tariff or a quota in response to the firm making an investment. 4. Concluing Comments We can see from this paper that formalising the iea of a manager taking kickbacks provies us with useful insights. t shows clearly the link between the size of the supplier s surplus Z = CS, where = π, an the corruption-eterring cost of being caught m. Obviously a reuction in protection woul result in a lower supplier s surplus an hence the m woul be lower. So, potentially, a reuction in protection coul cause a switch from corrupt to non-corrupt behaviour. The result that has been attaine comparing equivalent tariffs an quotas is certainly not

8 8 nternational Journal of Business an Economics obvious. A policy implication that follows from this result is that, while a switch from a regime of quotas to a regime of tariffs may iish rent-seeking activities, it will not iscourage managers from accepting kick-backs in return for investing in inappropriate technology. t shoul be strongly emphasise that this is in no way an argument for quotas to be preferre over tariffs. Rather it is a result that emphasises that merely swapping between protective instruments is no substitute for actually reucing the level of protection. Appenix To erive expressions (4) an (5), suppose the omestic firm chooses its output to maximise = x[ α β ( x + y)] cx. The first-orer conition is α 2βx βy c =. We rearrange this to get the omestic firm s reaction function α βy c x = or 2β α 2βx c y =. β The foreign firm chooses its output to maximise t = y[ α β ( x + y)] c y ty. The first-orer conition is α βx 2βy c t =. We rearrange this to get the foreign firm s reaction function α βx ( c + t) y =. 2β Now we equate the reaction functions an solve for x to get α 2c + c + t x t =, 3β which is Equation (4). To obtain y t we substitute the above expression for x t into the reaction function for the foreign firm

9 Neil Campbell 9 α + c 2 ( c + t) y t =. 3β Price is simply obtaine using the inverse eman function α + c ( + c + t) p t =. 3β Finally we can use our expressions for price an quantity to obtain an expression for the omestic firm s profit 2 t ( α 2c + c + t) =, 9β which is Equation (5). References Aes, A. an R. Di Tella, (1997), The New Economics of Corruption: A Survey an Some New Results, Political Stuies, 45(3), Anerson, J. E., (1988), The Relative nefficiency of uotas, Cambrige: MT Press. Campbell, N. A., (1998), Can We Believe in Col Showers? Journal of Economic ntegration, 13, Gupta, S., L. e Mello, an R. Sharan, (21), Corruption an Military Spening, European Journal of Political Economy, 17, Jain, A., (21), Corruption: A Review, Journal of Economic Surveys, 15, Krueger, A. O., (1974), The Political Economy of the Rent-Seeking Society, American Economic Review, 64(2), Mauro, P., (1998), Corruption an the Composition of Government Expeniture, Journal of Public Economics, 69(2), Megginson, W. L. an J. M. Netter, (21), From State to Market: A Survey of Empirical Stuies on Privatization, Journal of Economic Literature, 39, Shleifer, A. an R. Vishny, (1993), Corruption, uarterly Journal of Economics, 18, Stiglitz, J. E., (1988), Economic Organization, nformation, an Development in Hanbook of Development Economics, Volume, H. Chenery an T. V. Srinivasan es., Amsteram: North Hollan. Vousen, N., (199), The Economics of Trae Protection, Cambrige: Cambrige University Press. Williamson, O. E., (1964), The Economics of Discretionary Behavior: Managerial Objectives in a Theory of the Firm, New Jersey: Prentice-Hall.