Policy Consideration on Privatization in a Mixed Market

Transcription

1 Poliy Consideration on Privatization in a Mixed Market Sang-Ho Lee * Abstrat This paper onsiders a mixed market where the publi firm ompetes with private firm and examines the welfare effet of the industrial poliy reform in the proess of privatization. In partiular, fousing on the ost effiieny gap between two different organizations, we investigate how the fundamental effiieny trade-off in privatization poliy will be hanged under ompetition poliy with or without foreign ompetitor. The analysis finds out that the suffiient ost effiieny gap improves welfare in postprivatization when proper ompetition poliy is aompanied. Otherwise, industrial poliy reform might not give a welfare-improving result. Several important eonomi and industrial poliy impliations on the issues of liberalization are also disussed. keywords: industrial poliy, foreign ompetitor, mixed market; privatization; JEL lassifiations: L5; D42 1. Introdution Publi firm is one of the eonomi instruments utilized by the government to orret market failures and to reah an improvement in soial welfare. Theoretially, the possibility of operating a publily owned enterprise to ahieve the publi goals arises under the ondition of imperfet ompetition in the industries or publi goods provision. However, the poor eonomi and finanial performanes of many publi firms and the ases of suessful privatization in some developed ountries have been used as arguments for the industrial poliies of privatization and ompetition. 1 * Professor of Eonomis Department, College of Business Administration, Chonnam National University, 3 Yongbong-dong, Bukgu, Gwangju 5-757, S. Korea. sangho@honnam.a.kr. This paper was presented at China-Korea onferenes with Jilin University (Changhun, China, 27.8, Nanjing University (Nanjing, China, 27.9, and Fudan University (Gwangju, Korea, I would like to thank to Prof. Wang for his valuable omments. 1 Laffont and Tirole (1993, Visusi, et al. (1995, Hart, et al. (1997, and Laffont and Martimort (22 disussed the important regulatory poliy issues of publi firms. Wang and Xu (27 extended their analysis into a three parties game among firms, government, and onsumers.

2 In the eonomis literature there is onventional wisdom on the osts of state ownership. As provided in Lee and Hwang (23, for example, the managers of a publi firm may mismanage its assets in suh a way that they ause (managerial prodution ineffiienies. First, they may invest too little or too muh, sine they are not given the stoks options that would enourage them to take a long-term perspetive: the absene of apital market monitoring. Thus, one investments are sunk, the government may use these investments for purposes they were not intended to serve. Seond, a publi firm is not subjet to takeovers and its managers are therefore less onerned about losing their jobs: the absene of labor market monitoring. Third, a publi firm is not subjet to the disipline of the bankrupty proess beause the government will always bail it out in ase of diffiulty: the soft budget onstraint problem. Fourth, governments are subjet to the pressure of interest groups to diret the behavior of a publi firm so as to enhane the welfare of these groups: the prinipal-agent problem. The managers of publi firms at to maximize politial support and extend their job tenure. Thus, in order to make their work environment more pleasant, the managers may hoose to minimize labor strife by providing higher wages. All these fators redue managerial inentives taking are of effiient prodution, so that a publi firm with greater managerial disretion ats more ineffiiently than a private firm. There have been also many empirial studies investigating the relative performane of publi and private firms operating in a omparative environment. Parris, et al (1987 ontains a omplete olletion of data illustrating the relative importane of private and publi firms in various markets in some western European industries. Vikers and Yarrow (1988, 1991 and Megginson and Netter (21 used empirial researh to assess the effets of privatization as an industrial poliy. They onluded that enterprises operating under publi ownership will be less effiient than their private setor ounterparts. However, although other developing ountries have followed the experiene of poliy reform in developed ountries and moved on the path of liberalization poliy suh as privatization and ompetition, and although several governments have shown a tendeny to require publi firms to at as private firms and to try to maximize profit, there are still several instanes of mixed market, whih is haraterized by the oexistene of profit maximizing and welfare maximizing firms. In fat, nowadays, the form of mixed eonomies where publi firms interat with private firms an be observed quite often, espeially in Asian, European, Latin Amerian, and former ommunist transitional eonomies suh as China and Russia. This indiates that the importane of the publi firm in the mixed industry in every eonomy remains obvious despite a reent trend towards privatization.

3 One example is the television market, in whih profit maximizing produers ompete with the state network, whih is finaned by taxation and required to pursue some publi goals of eduational and ultural programs. Other example is the parel division of the publi post offies, whih ompetes with private parel delivery servie in every ountry. For instane, even USA, there are examples of mixed market like pakage and overnight-delivery industries. Furthermore, ompetitions between private and publi firms in network industries suh as banking servies, railways, eletriity, natural gas, and teleommuniations are also popular in the reent industrial eonomies exept USA. In addition, private ompetitors in mixed market are not always domesti. For example, the government of New Zealand deided to found a state-owned publi banking firm ompeting against private foreign firms. When the government of Brazil bargained with Rohe, it used a publi medial institution as a potential ompetitor in the medial market. Many mixed industries with foreign firms in the former ommunist ountries inluding China and Russia are also observed in reent years. Then, we an raise several important questions. For example, what are the differenes in the ownership strutures and their performanes? What are the interations between publi ownership and private ownership? How is the performane of eah ownership struture under the ompetition? Speifially, what onditions does privatization promote the welfare? Therefore, it is needed to examine the welfare effet of privatization poliy under the ompetition, espeially with or without foreign firms, and find the relevant eonomi and industrial poliy impliations to improve soial welfare in post-privatization. In this paper, we review the liberalization poliy of privatization in aordane with the ompetition poliy, by fousing on the ost effiieny gap between private firm and publi firm. We onfine our analysis into the imperfet ompetition market, natural monopoly or duopoly market, where full ompetition poliy is neither desirable nor attainable. In this ase, it is needed to balane the industrial poliy reform between privatization poliy and limited ompetition poliy. We examine the welfare effet of these industrial poliies in a mixed duopoly market and point out the onventional effiieny trade-off in the poliy implementation. In partiular, we first model the mixed market in whih publi firm ompetes with domesti firms in the proess of privatization and investigate how the welfare onsequenes in the effiieny trade-off will be hanged beause of the eonomi effet of ompetition poliy. We then extend the analysis into the ase that the government faes a foreign ompetitor, investigate the effiieny of the industrial poliy, and ompare the welfare onsequenes with domesti ompetitor. The analysis finds out that the suffiient ost effiieny gap improves welfare in post-privatization when proper ompetition poliy is aompanied.

4 Otherwise, industrial poliy reform might not give a welfare-improving result. Several important eonomi and industrial poliy impliations on the issues of liberalization poliy inluding privatization and ompetition are also disussed. The final setion gives onluding remarks. 2. The Basi Model of Monopoly Market Consider a monopoly firm that faes an inverse demand funtion of P( Q = a bq and onstant prodution ost of. Then, the profit funtion for the monopolist and the onsumer surplus are given by, respetively, π = P( Q Q Q, (1 Q ( (. (2 CS = P v dv P Q Q Hene, the soial welfare, whih is defined as the simple sum of onsumers surplus and firm s profits, is given by Q (. (3 W = P v dv Q We assume that the government (or the publi firm as an ageny maximizes soial welfare while the private firm fouses on profits in the following analysis. This implies that the monopolist will maximize soial welfare in (3 when it takes the form of publi enterprise in pre-privatization, while it will maximize its profit in (1 in postprivatization. That is, the welfare-maximizing output is determined at P( Q = or a Q = with publi firm, and profit-maximizing output is determined at b a determined at P( Q bq= or Q1 = with privatized firm. 2b For the interior solutions and reasonable omparisons, we assume the ost differene between publi and private firms as follows: Assumption 1: a 1 a 1, where the ost of publi firm is defined as and that of private firm is defined as 1. Then, we have 1. This assumption on ost effiieny gap between private firm and publi firm implies that in the ageny relationships the prinipal and the agent will inur positive monitoring osts, and in addition there will some divergene between the agent's/deisions and those deisions, whih would maximize the welfare of the

5 prinipal. That is, the publi firm might inur self-interested and ineffiient expenditures, suh as pure waste in the form of goldplating, (aounting and managerial ross-subsidies to the other businesses, exessive employee ompensation, et. Suh `wasteful' expenditure an arise from politial reasons, lak of manager's inentives to eonomize, or prinipal-agent problems in general. 2 Therefore, the welfare effet of privatization depends on the effiieny gap between the publi firm and private firm. If it is insignifiant, however, the privatization poliy under monopolisti market always dereases the welfare after all. Speifially, from (3, the welfare level of eah ase is as follows: 2 Q ( a = ( = W P v dv Q (4 2b 2 Q1 ( a 1 W1 = P( v dv 1Q1 = (5 4b where W denotes the welfare with publi firm and W 1 denotes the welfare with privatized firm. Therefore, the welfare in post-privatization inreases only if 2( a ( a or ( a. Otherwise, the welfare under privatization poliy dereases. On the other hand, if we onsider the ase where foreigners have the share of the privatized firm s profit with the portion of (1-s, then the welfare with privatized firm 2 ( a 1 will be W1 π1= (1 s, whih is smaller than the welfare with privatized 4b firm in (5. However, the following Proposition will still hold. Proposition 1. (Fundamental Trade-Off in Privatization Poliy The welfare in postprivatization inreases only if the ost effiieny gap between the publi firm and private firm is suffiiently large. Proposition 1 shows that the welfare hange under privatization poliy depends on the relative size of the ost effiieny gap between the publi firm and private firm. 3 2 In the regulatory eonomis literature, the managerial ineffiieny is defined as ``waste'' and ``abuse'' in Sappington (198, Laffont and Tirole (1993, and Laffont and Martimort (22. 3 De Fraja and Delbono (1989 onsidered an inreasing ost funtion and shown that there is a trade-off in privatization and nationalization. Furthermore, De Fraja (1991 onsidered the X-ineffiieny in the less-prodution ase and also shown that there is a trade-off in the proess of privatization.

6 Notie that the welfare will always derease if = 1, where there is no ost effiieny gap between publi firm and private firm. Therefore, there is a trade-off in privatization poliy in monopoly market. 3. Duopoly Market Competition with Domesti Ownership In reent years, in developing areas, suh as Asian, European, and Latin Amerian ountries, the government ativated both privatization and ompetition poliies at the same time. This omposes the mixed market where the publi firm ompetes with private firms. We then analyze the industrial poliy on the mixed market where a publi firm ompetes with private firms. For the purpose of omparison, we assume that both firms produe goods that are perfet substitutes and thus, they fae a same market demand funtion of P( Q = a bq But, two firms have different, onstant prodution ost of for publi firm and 1 for private firm. Then, the profit funtion and the welfare funtion are the same as in (1 and (3, with the modifiations of Q= qi + qe where q i is inumbent output, q if it is publi firm or q 1 if it is privatized firm, and q e is privately managed entrant firm. We also assume that the inumbent ompetes against the entrant, whih has a onstant prodution ost of 1, with Cournot-type output ompetition Mixed Duopoly Market We first onsider the mixed duopoly ase where the government employs a simple ompetition poliy only and thus, the publi firm (inumbent maximizes the welfare in (3, while the private firm (entrant maximizes its own profit. Then, the equilibrium ( a ( 1 1 outputs are q = and qe =. Then, total market output is b b a Q = q+ qe =, whih is the same as the nationalized monopoly output, Q. b That is, publi firm still sets its prie at its marginal ost,. It represents that the

7 market prie is the same with the nationalized monopoly ase (publi firm. However, the output level of inumbent will be redued. This substitution effet of output hange will ause the welfare hange. Finally, the welfare level in ompetition poliy is as follows: 2 2 Q ( a + 2( 1 W = P( v dv q 1q e = (6 2b Notie that the welfare level in a mixed market with the ompetition poliy is greater than that of nationalized publi firm, i.e., W > W. This indiates that the ompetition poliy will inrease the welfare beause the output of ineffiient inumbent will be replaed with that of effiient entrant. However, if we onsider the ase where foreigners have the share of the private domesti firm s profit or/and privatized domesti firm s profit, the result will depend on the foreigner s portion of share on domesti firms Pure Duopoly Market We next onsider the pure duopoly ase where the government employs both ompetition and privatization poliies together, where the privatized firm (inumbent and private firm (entrant maximize its own profit, respetively. Then, the equilibrium ( a 1 2( a 1 outputs are q1 = q1 e = and total market output is Q1 = q1 + q1 e =. 3b 3b Finally, the welfare level in the ombined poliy of ompetition and privatization is as follows: 2 Q1 4( a 1 W1 = P( v dv 1q1 1q 1e = (7 9b A few remarks are in order. First, as usual, the output level of inumbent in a pure duopoly is smaller than that in the privatized monopoly ase, but total output level in a pure duopoly is greater than that in the privatized monopoly, Q 1. It represents that the market prie will be lower in the pure duopoly ase. Therefore, the welfare level in a pure duopoly is greater than that of privatized monopoly, i.e., W1 > W1. Again, this implies that the ompetition poliy will inrease the welfare beause of ompetitive pressure of the entrant. Seond, total output level in a pure duopoly is greater than that in the nationalized monopoly (publi firm when a 2( 1. Thus, if the ost effiieny gap is

8 large, the pure duopoly will inrease total output. Notie that this is a suffiient ondition for privatization poliy to inrease the welfare. Therefore, the welfare omparison between pure duopoly and nationalized monopoly depends on the relative size of the ost effiieny gap between the publi firm and private firm. In partiular, the welfare of pure duopoly is greater than that of nationalized monopoly only if 9( a 8( a. Otherwise, the ombined poliy of privatization and ompetition will worsen the welfare. Finally, total output level in a pure duopoly is greater than that in a mixed duopoly under simple ompetition poliy when a 2( 1. Thus, if the ost effiieny gap is large, the pure duopoly will inrease total output. In that ase, the output level of effiient private inumbent in a pure duopoly is greater than that of ineffiient publi inumbent in a mixed duopoly. Notie that this is also a suffiient ondition for privatization poliy to inrease the welfare. Therefore, the welfare omparison between pure duopoly and mixed duopoly also depends on the relative size of the ost effiieny gap between the publi firm and private firm. In partiular, the welfare of pure duopoly is greater than that of mixed duopoly only if ( a 1 9( a ( 1. Otherwise, ombined with ompetition poliy, 18 privatization poliy will worsen the welfare. In sum, we an summarize the findings. First, if 9( a > 8( a or ( a, we have the following orders in terms of welfare: 1 W > W > W > W. It indiates that if the ost effiieny gap between the publi firm 1 1 and private firm is small, simple ompetition poliy will be the first-best than any other poliies whih aompanies privatization poliy. That is, privatization poliy is harmful to the soiety. Notie that this result inludes the ase of = 1, where there is no effiieny gap between publi firm and private firm. Seond, if 9( a < 8( a, i.e., the ost effiieny gap between the publi firm and private firm is not so small, W 1 is the greatest welfare level among four alternatives if ( a 1 9( a ( 1, while W is otherwise. Therefore, the 18

9 first-best poliy will be simple ompetition poliy or ombined poliy of ompetition and privatization, depending on the relative ost effiieny gap between the publi firm and private firm. 4 Proposition 2. The welfare in post-ompetition poliy will be the highest if the ost effiieny gap between the publi firm and private firm is relatively small. However, if the ost effiieny gap between the publi firm and private firm is relatively large, putting privatization poliy into ompetition poliy will inrease the welfare. Proposition 2 in a duopoly market gives the same eonomis insights of Proposition 1 in a monopoly market, that is, privatization is soially benefiial only when the ost effiieny gap between two firms is large. Therefore, privatization poliy will not always give a welfare-improving result. The only benefit of a duopoly market is that there exists the welfare-inreasing effet of ompetition poliy. Therefore, the government should improve the ompetitiveness of the market in privatizing the publi firms. For improving the ompetitiveness of the market, it is important to onsider not only the degree of ompetition in the market, but also the harateristis of the market. For example, if the market is haraterized by its nature of natural monopoly, beause of sunk and fixed ost, it is not benefiial to introdue the entrants into the market in many ases. 4. Duopoly Market Competition with Foreign Ownership We now extend the analysis into the ompetition situation with a foreign firm. 5 Speifially, we onsider the ase where the government an hoose privatization poliy on nationalized publi firm and/or ompetition poliy to invite new foreign entrant into the domesti monopoly market. Again, for the interior solutions and reasonable omparisons, we assume the linear demand and onstant prodution ost, and ost differene between domesti firm and foreign private firm is as follows: Assumption 2: a 2 a 1 4 However, if the ompetition poliy is not available beause of politial or tehnologial reasons, the seond-best poliy should be determined based on Proposition 1. 5 Several researh inluding Pal and White (1998 and Matsumura (23 allowed foreign firms into the analysis in the eonomis literature.

10 where the ost of foreign private firm is defined as 2. Then, we have 1 2. This assumption on ost effiieny gap between domesti and foreign private firms implies that the foreign firm might have tehnologial and managerial advantages in prodution. This suggests that ombined poliy with ompetition with a foreign firm and privatization on the publi firm will yield a spill-over effet in the market through leaning on tehnologial and managerial skill, so alled neighboring effet in learning from its neighbor. Finally, the firms ompete with the Cournot-type pattern Mixed Duopoly Market with a Foreign Firm We first onsider the ase of mixed duopoly market where the government employs a simple ompetition poliy with foreign firm. Then, the publi firm maximizes the following welfare: f Q f f f f f f = 2 e e 2 e W P( v dv q q ( P( Q q q. f Q f f f P ( v dv q P ( Q q e (7 = Notie that the profit of foreign firm, P( Q q 2q f f f e e, should not be inluded in the domesti welfare, whih is the objetive funtion of the publi firm. Then, the f ( a ( 2 f 2 equilibrium outputs and total output are q =, qe =, and b b f a Q =. Therefore, ompared with the nationalized monopoly, the market prie b is the same, but the output level of publi firm will be more redued. Finally, the f ( a welfare level in ompetition poliy with foreign firm is W 2b the welfare level in a mixed duopoly market with foreign firm is the same with that of 2 =. Notie that f nationalized publi firm, i.e., W = W, and thus it is smaller than the mixed duopoly market with domesti firm. This is so beause the output of ineffiient inumbent will be replaed with that of effiient foreign entrant, whih is not onsidered into the welfare. This implies that the ompetition poliy with foreign firm will not be neessary to inrease the welfare. However, if we onsider the ase where domesti people have the share of the foreign firm s profit, the result will depend on the foreigner s portion of share on the foreign firm.

11 4-2. Pure Duopoly Market with a Foreign Firm We next onsider the ase of pure duopoly market where the government employs both ompetition poliy with a foreign firm and privatization poliy on publi firm together. Then, both privatized domesti firm and private foreign firm maximize its own profit, respetively. Then, under the assumption of equal prodution ost in postprivatization, the equilibrium outputs are q1 f f ( a 2 = q1 e = and total market 3b f 2( a 2 output is Q1 =. Finally, the welfare level in the ombined poliy of 3b ompetition and privatization is as follows: f Q f 1 f f f f f 1 = e 1 1e 2 1e W P( v dv q q ( P( Q q q (8 Thus, the welfare level in post-privatization and ompetition with foreign firm is 2 f 3( a 2 W1 =. 9b A few remarks are in order. First, as usual, the welfare level in a ombined poliy is f greater than that of privatized monopoly firm, i.e., W1 > W1 beause of ompetition effet. Seond, the welfare omparison between pure duopoly and nationalized monopoly depends on the relative size of the ost effiieny gap between the publi firm and private firm. Notie that the welfare level of nationalized monopoly is the same with f that of mixed duopoly, i.e., W = W. In partiular, the welfare of pure duopoly is greater than that of mixed duopoly if 3( a 2( a. Otherwise, the ombined poliy of privatization and ompetition with foreign firm will worsen the welfare. Therefore, the ompetition poliy with foreign firm will inrease the welfare only if the ost effiieny gap between publi firm and private firm is suffiiently large. In sum, we an summarize the findings as follows. First, if 3( a > 2( a, f f we have the following orders in terms of welfare: W = W > W1 > W1. It indiates that if the ost effiieny gap between the publi firm and private foreign firm is not suffiiently large, neither privatization poliy nor ompetition poliy inrease the

12 welfare. Notie that this result inludes the ase of = 2, where there is no effiieny gap between publi firm and foreign firm. Seond, if 3( a < 2( a, i.e., the ost effiieny gap between the publi f firm and private foreign firm is suffiiently high, W 1 is the greatest welfare level among four alternatives. Therefore, the first-best poliy will be ombined poliy of ompetition with foreign firm and privatization on the domesti publi firm. Proposition 3. Assume that the ompetitor is a private foreign firm. Then, neither ompetition poliy nor privatization poliy inrease the welfare if the ost effiieny gap between the publi firm and private foreign firm is small. However, if the ost effiieny gap is large, the welfare in post-ombined poliy of ompetition and privatization will be the highest. Proposition 3 is different with Proposition 2 in that ompetition poliy with foreign firm might not be soially benefiial if ost effiieny gap is small. That is, ompetition poliy will not always give a welfare-improving result. Therefore, very areful attention on the industrial poliy in the proess of privatization and ompetition should be taken into onsideration. In partiular, under the ompetition with foreign firm, the government should improve the ost effiieny of the publi firm before privatizing the publi firms. For example, the inentive mehanism to redue the X- ineffiieny from the priniple-agent relationship, suh as moral hazard problem, ould be benefiial to improve the ost effiieny of the publi firm. 5. Disussions on Poliy Impliations Sine 198s, industrial poliy reform of privatization and ompetition in the transitional eonomy, from planned eonomy into market eonomy, is a remarkable histori event of institutional evolution. In partiular, there is a stark ontrast between the eonomi performane of the Russian-style and the Chinese-style industrial poliies. That is, Chinese-style gradualism sueeded to the extent that the original governane remained, while the Russian-style radial reforms lead to governane failure. We have also observed that the governments in developing ountries, suh as Afrian, Asian, European, and Latin Amerian ountries, have ativated or have planned to ativate liberalization poliy of privatization and ompetition in mixed markets.

13 In priniple, the most important poliy aspet of privatization for the government is to indue the publi firm to ahieve ost effiieny by reduing its managerial ineffiieny. This implies that the poliy makers tend to believe that an ineffiient prodution ost level of the privatized publi firm would be dereasing, and finally, will be equal to the ost level of the independent private firm in post-privatization. Therefore, what we have learned from the above analysis is that the welfare impliations of industrial poliy reform depend primarily upon the relative ost effiieny between private firm and publi firm. However, in omparing the trade-off in privatization poliy, there are many other important poliy aspets that we have abstrated from for reasons of tratability and simpliity. We then present and disuss some industrial poliy-relevant impliations in the eonomis literature, whih inlude the issues of ompetition pattern, industrial struture, ageny problems and politial inentives, and partial ownership Competition Patterns and Produt Quality We have onfined our analysis into the simple duopoly market with homogeneous produt. However, most industries produe a large number of similar but not idential produts using the different ways of prodution. It is needed to examine the effet of the degree of produt differentiation on the welfare hange in post-privatization and ompetition. Matsushima and Matsumura (23, for example, investigated a mixed market with differentiated produts and shown that herd behavior by private firms our where private firms adopt very similar strategies within the market but those strategies are ompletely different from those of publi firm. In addition, the welfare onsequenes of a privatization poliy depend on the ompetition patterns of between the firms. In partiular, we an also onsider the different ompetition patterns in the duopoly market, suh as Bertrand ompetition or Stakelberg leadership ompetition. As shown in Vives (199 in a produt differentiated market, the symmetry between Cournot-type quantity ompetition and Bertrand-type prie ompetition an be established with the duality argument. However, in general, the ompetition patterns of Cournot or Stakelberg depend on the strategi eonomi environments among firms. In the eonomis literature, therefore, many studies have onsidered the order of play in a game and ompared the results of Cournot and Stakelberg. 6 De Fraja and Delbono (1989, 199, for example, show that if the publi firm is able to at as a leader to indue Stakelberg ompetition in a mixed market, it is possible to inrease its soial welfare. However, the sequential entry with 6 On this point, see De Fraja and Delbono (1989, 199, and Vives (1999, among others.

14 or without entry regulation should be also examined. As far as onerned on entry regulation, the effet of fixed ost on entry deision will be important under the endogenous market struture. Furthermore, as addressed by Hart et al. (1997 and Wang and Xu (27, there might be a trade-off in effiieny between ost and quality, i.e., attempts to lower ost will jeopardize quality. In general, publi firm has insuffiient inentive to lower ost while private firm has an inentive in ost redution. However, there might be a loss in produt quality due to ost redution under private ownership, whih will be fully absorbed by the soiety. Therefore, a omparison between effiieny of publi firm and private firm is in essene a ost and benefit analysis for various inentives in multi-dimensions Vertial Struture and Network Industry Most industries produing differentiated produts are vertially onentrated, where the upstream firm vertially inter-related with the downstream firms. Then, from the different poliy perspetive on vertial struture, if the publi firm is not only related with the upstream firms but also with downstream firms, the industrial poliies on the market struture of networked industry and on the strategi behaviors of bottleneked firm are also important. Lee (26, for instane, have examined the welfare onsequenes of privatization in vertial struture and pointed out the importane of the degree of ompetition and open aess poliy in the welfare tradeoff. For example, it might be sometimes tehnologially and/or politially ineffiient to divide the vertial market into two separating markets. That is, if high-tehnology an be applied to the integration between the different servies through o-utilizing failities and human resoures, then it will be effiient to integrate two markets vertially sine tehnologial eonomies or/and dynami investments an be realized. Furthermore, politially it might lead to large wasteful soial osts to separate the historial publi firm into different firms if the government an not lead the market. Therefore, on the matter of integration or separation, it is neessary to hek if there are tehnologial or politial linkages between two markets and to onsider how the regulatory ageny is able to treat these problems without large soial osts by using privatization poliy on publi firm. Therefore, Sappington and Stiglitz (1987 argue that privatization affets the transations osts of government intervention in enterprise 7 As addressed in Hart et al. (1997, even thought privatization redued ost, the deterioration of servie quality and abuse of prisoners beame ommon after the privatization in Ameria.

15 deision-making. This is so beause privatization does not entail the transfer of all deision-making authorities to the privatized firm. As a matter of fat, the Korean government announed plans to privatize its eletriity power utility (Korea Eletri Power Corporation, whih was a governmentinvested monopolist that supplied eletri power in Korea. During 21, the Korean eletri power industry underwent major hanges as its power generation unit was separated into six subsidiaries and the Korea Power Exhange was inaugurated. In addition, the power generation subsidiaries are supposed to be privatized and it is preparing separate power distribution units. But there still remain many debates between the government and employees. Therefore, it is expeted that there are ongoing soial osts in the privatization proess of separation and ompetition. Finally, when government ativates privatization poliy, there is a onern about market forelosure in whih the profit-maximizing, privatized integrated firm may have an inentive to disriminate the downstream ompetitors. Therefore, in the proess of privatization poliy in the vertially networked industry, the regulatory ageny needs to onstrut an appropriate anti-ompetition regulation for the privatized firm. In reality, even post-privatization, the government keeps the regulatory power to ontrol the strategi behavior of privatized firm by organizing independent regulatory institutions, suh as OFTEL in UK and FTC in Korea Ageny Problems and Politial Inentives In the analysis, we have assumed that the objetive funtion of the publi firm is defined as the soial welfare in (3, and that of private firm as the profit in (1. This omes from the assumption that the publi firm maximizes the objetive of the government and the managers of private firm maximize the objetive of the owners. However, these assumptions on publi firm and private firm should be reexamined from the viewpoint of ageny relationship. For the ase of publi firm, there are two fundamental assumptions in suh disussions. First, the objetive of government is to maximize the soial welfare, i.e., the assumption of benevolent government. Seond, the government has omplete information and absolute authority to ontrol the publi firm, i.e., the assumption of no ageny problem. However, as a matter of fat, there is a disrepany between the reality and these assumptions. Espeially, the objetive of government should be based on the aptured inentive under the politial environments. Levy (1987 suggests that organizational ineffiieny and waste may arise if managers of a publi firm reeive onfliting instrutions, partiularly if they may be exaerbated by a hange of government following general eletion. He also pointed out that the notion that the government is

16 the prinipal and the publi firm is the agent might be misleading sine the government is not a single organization. That is, the government ats through a variety of ministers, legislators and ivil servants, who are themselves agents of the general publi. It indiates that their goals are rarely stated expliitly and trade offs among them are not agreed. Thus, different agents give the publi firm onfliting parallel ommands, i.e., multilevel prinipal-agent relationship. 8 For example, Cook and Fabella (22 onsidered the model in whih the state-owned enterprise maximizes an unspeified objetive funtion, and examined the theoretial treatment of the welfare and politial eonomy dimensions of the hoie between publi ownership and privatization. On the other hand, due to inomplete information or ostly monitoring and inomplete ontrat, the objetive of publi firm will differ from that of government. Then, the theoretial treatment of the ownership effets dedued from the property rights and prinipal-agent perspetives should be onsidered to examine the effiay of the inentive system that is designed to maximize the efforts of the agents. De Fraja (1991, for example, used ageny theory to model managerial effort and to analyze the X-ineffiieny of the publi firm. Similarly, private firm onfronts the same ageny problem in the deision making proess, even though it might be less serious than the ageny problem in managing publi firm. As a seminal paper, Jensen and Mekling (1976 investigated the nature of the ageny osts to develop a theory of ownership struture in a private firm, and showed the Pareto optimality of their existene. Hene, based on the prinipal-agent theory, we need to inorporate the existene of information asymmetry to examine the non-profit-maximizing strategies of private firm. In sum, the inentive mehanism design is more important than the hoie of ownership strutures, whih is related with residual rights of ontrol. Thus, we need to make a multidimensional omparison between the effiieny of publi firm and private firm from the perspetive of inentives. In partiular, publi firm ought to adopt more expliit ompetitive (performane-based inentive mehanism in order to inrease managerial effiieny, while private firm sould be indued to use impliit ooperative inentive mehanism in order to derease managerial distortions Partial Ownership and Governane Struture On the path of privatization, the government may be able to manage the ativities of the privatized firm by ontrolling its portion of shares. In other words, there is a 8 For more disussion, see the artiles in the editorial book by Jones (1982 and the summary in De Fraja (1991.

17 possibility that full privatization, where the government sells all its shares in a publi enterprise, is not fulfilled at one. Thus, the government might determine the degree of privatization instead of the extreme full privatization. Lee (26, for example, pointed out the effiieny gap under the vertial network struture and showed that the welfare hange in post-privatization depends on the effiieny gap. In the publi eonomis literature, De Fraja and Delbono (1989 and De Fraja (1991 examined the effiieny of full privatization in an oligopoly market. In addition, Matsumura (1998 and Lee and Hwang (23 onsidered the possibility of partial privatization in a mixed model and showed that it is optimal for the government to sell some but not all of its shares in publi firm when there exist produtioneffiieny effets of partial privatization. Finally, Matsumura and Shimizu (27 extended the analysis into the ase of sequential privatization waves and examined its welfare onsequenes. [Table 1] The degree of liberalization in EU eletriity market Full Liberalization (1% Partial Liberalization (6-9% Austria, Denmark, Finland, Germany, Ireland, Netherlands, Norway, Portugal, Spain, UK Belgium, Frane, Greee, Italy, Luxembourg As we have seen in [Table 1], these partial privatization proesses are very popular in other ountries. In the Korean eletri power industry, for instane, its power generation unit was separated into six subsidiaries and the Korea Power Exhange was inaugurated, as mentioned in the previous disussion. In this privatization proess, the power generation subsidiaries have been privatized suessively and it is preparing separate power distribution units in a sooner year. However, when we onsider the ase of foreigners ownership on the domesti privatized firms, the welfare onsequenes might be ompliated. As we examined, one simple way of analyzing this ase is to inorporate the governane degree of partial ownership in the weighted welfare funtion in the analysis. But, as a matter of fat, the relative portion of share for ownership might not be real matter for the ontrol of publi firm. As we observed in the privatization proess of the U.K. and New Zealand, a sheduled governmental privatization ontrats suh as speial share an be used. Even if there were so, we might need to investigate the internal governane struture of the privatized firm, beause the privatized firm might suffer from a onflit of interests between regulators and interest groups. Therefore, the privatization should be reexamined on the base of before, in, and after to get the general onlusion on the government ontrol rights over the

18 firm. To determine the relative mix between publi firm and private firm, we must take aount of the subtle interations among the firms. In this ase, the role of Non-Profit Organization (NPO, a mixed form of organization between publi and private organizations, ould be examined. In the U.S.A, for example, NPO plays a great part in the soial eonomi life, inluding mediare and eduation servies. 6. Conluding Remarks This paper onsidered a mixed market with different prodution osts between two different organizations, and examined the welfare effet of the industrial poliy reform in the proess of privatization. The main result of the simple analysis is that the presene of relative ost effiieny gap between publi firm and private firm may be the key fator to determine the improvement in the overall welfare of the industry. On the issue of privatization and ompetition with a foreign firm, the analysis highlights how it may be ounter-produtive in terms of welfare level, depending upon the relative ost effiieny gap. Far from privatization improving welfare and reduing waste expenditure, it may also yield the opposite results of a redution in domesti prodution, whih omes from substitution under the imperfet ompetition with foreign firms. Therefore, it is not neessarily the ase that the improvement in the effiieny of publi firm through industrial poliy of privatization and ompetition will have a benefiial effet, as long as the ost effiieny gap is not suffiiently large. However, in pratie, there exists a series of industrial poliy impliations for privatization, whih is espeially important to north-eastern Asian ountries inluding China and Korea. First, even though the ost effiieny gap between domesti publi firm and foreign private firm is small, privatization poliy with foreign ompetitors might be benefiial to the soiety from the perspetive of ompetition. Therefore, mixed market with ompeting foreign firms might be effiient than a monopolized publi firm if the market an be sustainable with the ompetition in the long-run. Seond, the ompeting domesti firm an adopt advaned skills from the ompeting foreign firm in a long-term dynami ompetition. This is the ase where there is a strong spill-over effet in the proess of market ompetition, suh as learning and training on managerial and tehnologial skills by ompetition. Therefore, under the ompetition with foreign firms, full privatization or fully authorized private firm might not be soially benefiial in the long-term relationship. In this ase, a sheduled governmental privatization ontrats should be organized, for example as like the experienes in the U.K. and New Zealand. However, if the market is suffiiently ompetitive, this problem of ownership struture will not matter in industrial poliy, but the problem of fair ompetition will matter.

19 Finally, it is important to take other advantage of publi firm in the real eonomy. For the suess of industrial poliy, the informational requirement is very serious for the government to implement the proper poliy. In partiular, these requirements amount to knowledge of the ost of private firms. It is usually the ase that the private firm has no inentive to reveal truthfully the information they possess. Therefore, mehanism design under inomplete ontrat should be onstruted. In designing an inentive ompatible mehanism in a proess of privatization and ompetition, the publi firm might be an approximation for the features of the market in whih it operates beause publi ownership an provide more information than is obtainable in a private firm. Referenes Cook, P. and R. V. Fabella, 22, The Welfare and Politial Eonomy Dimensions of Private versus State Enterprise. The Manhester Shool 7, De Fraja, G., 1991, Effiieny and Privatization in Imperfetly Competitive Industries. Journal of Industrial Eonomis 39, De Fraja, G. and F. Delbono, 1989, Alternative Strategies of a Publi Enterprise in Oligopoly. Oxford Eonomi Papers 41, De Fraja, G. and F. Delbono, 199, Game Theoreti Models of Mixed Oligopoly. Journal of Eonomi Surveys 4, Grilo, I, 1994, Mixed Duopoly under Vertial Differentiation. Annels D eonomie et de Statistique 33, Hart, O, Shleifer, A., and R. Vishny, 1997, The Proper Sope of Government: Theory and an Appliation to prisons,'' Quarterly Journal of Eonomis 112, Jensen, M. C. and W. H. Mekling, 1976, Theory of the Firm: Managerial Behavior, Ageny Costs, and Ownership Struture,'' Journal of Finanial Eonomis 3, Jones, L.P., 1982, Publi Enterprise in Less Developed Countries, Cambridge University Press. Laffont, J.-J. and D. Martimort, 22, The Theory of Inentives: The Prinipal-Agent Model, Prineton University Press. Laffont, J.-J. and J. Tirole, 1993, A Theory of Inentives in Prourement and Regulation, The MIT Press. Lee, S. H., 26, Welfare-Improving Privatization Poliy in the Teleommuniations Industry. Contemporary Eonomi Poliy 24, Lee, S. H. and H. S. Hwang, 23, Partial Ownership for the Publi Firm and Competition. Japanese Eonomi Review 54,

20 Levy, B., 1987, A Theory of Publi Enterprise Behavior. Journal of Eonomi Behavior and Organization 8, Matsumura, T., 1998, Partial Privatization in Mixed Duopoly. Journal of Publi Eonomis 7, Matsumura, T., 23, Stakelberg Mixed Duopoly with a Foreign Competitor. Bulletin of Eonomi Researh 55, Matsushima, N. and T. Matsumura, 23, Mixed Oligopoly and Spatial Agglomeration. Canadian Journal of Eonomis 36, Matsumura, T. and D. Shimizu, 27, Privatization Waves, working paper, University of Tokyo, Japan. Megginson, W. L. and J. M. Netter, 21, From State to Market: A Survey of Empirial Studies on Privatization. Journal of Eonomi Literature 39, Pal, D. and White, M.D., 1998, Mixed Oligopoly, Privatization, and Strategi Trade Poliy. Southern Eonomi Journal 65, Parris, H., Pestieau, P. and Saynor, P., 1987, Publi Enterprise in Western Europe, Croom Helm, London. Sappington, D. E. M., 198, Strategi Firm Behavior Under a Dynami Regulatory Adjustment Proess," Bell Journal of Eonomis 11, Sappington, D. E. M., and J. Stiglitz, 1987, Privatization, Information, and Inentives," Journal of Poliy Analysis and Management 6, Vikers, J. and G. Yarrow, 1988, Privatization: An Eonomi Analysis, Cambridge, MA: The MIT Press. Vikers, J. and G. Yarrow, 1991, Eonomi Perspetives on Privatization. Journal of Eonomi Perspetives 5, Visusi, K., J. Vernon. and J. Harrington, 1995, Eonomis of Regulation and Antitrust, The MIT Press. Vives, X., 1999, Oligopoly Priing: Old Ideas and New Tools, The MIT Press. Wang, Y. and H. Xu, 27, Soial Heterogeneity and Optimal Mix between Publi and Private Provision of Publi Goods,'' working paper, Fudan University, China.