coase v the coaseans

COASE VERSUS THE COASIANS*

EDWARD GLAESERSIMON JOHNSONANDREI SHLEIFER

Who should enforce laws or contracts: judges or regulators? Many Coasians,though not Coase himself, advocate judicial enforcement. We show that the incen-tives facing judges and regulators crucially shape this choice. We then compare theregulation of nancial markets in Poland and the Czech Republic in the 1990s. InPoland, strict enforcement of the securities law by a highly motivated regulator wasassociated with a rapidly developing stock market. In the Czech Republic, hands-offregulation was associated with a moribund stock market.

I. INTRODUCTION

At the heart of economists traditional skepticism about gov-ernment regulation is the Coase theorem [Coase 1960]. Thetheorem states that when property rights are well dened andtransaction costs are zero, market participants will organizetheir transactions in ways that achieve efcient outcomes. Whenthey can do so, it is not necessary for the government to engage incorrective actions through taxes, regulations, or even legalrules. Financial markets are often used to demonstrate the Coasetheorems case against regulation. Advocates of the regulation ofthese markets point to a variety of potential failures, such as theability of security issuers to expropriate both potential and exist-ing investors through misrepresentation or prot diversion. Inves-tors fear of such expropriation prevents rms from raising externalfunds, and keeps efcient projects from being undertaken.

Not so, reply the Coasians. They point out that most securi-ties transactions take place between sophisticated adults, andthat both the buyers and the issuers of securities have availableto them a vast range of private arrangements to achieve ef-ciency, including contracts such as corporate charters, certica-tion by intermediaries, and various forms of bonding. Such con-tracts render most laws and regulations unnecessary [Stigler1964; Easterbrook and Fischel 1991].

* We thank Alberto Alesina, Simeon Djankov, Oliver Hart, Louis Kaplow,Lawrence Katz, Rafael La Porta, Florencio Lopez-de-Silanes, Gerry McDermott,Randall Morck, Raghuram Rajan, Mark Ramseyer, Steven Shavell, Peter Temin,and three anonymous referees for helpful comments and the National ScienceFoundation and the Massachusetts Institute of Technology EntrepreneurshipCenter for nancial support.

2001 by the President and Fellows of Harvard College and the Massachusetts Institute ofTechnology.The Quarterly Journal of Economics, August 2001

853

On the face of it, the Coasians argument is powerful. Yet itcrucially relies, among other assumptions, on the possibility ofeffective judicial enforcement of complicated contracts. Judgesmust be able, and more importantly willing, to read complicatedcontracts, verify whether the events triggering particular clauseshave actually occurred, and interpret broad and ambiguous lan-guage. These requirements on the judges apply as strongly to thejudicial enforcement of laws, where the interpretation and appli-cation of particular statutes requires signicant investment. Inreality, courts in many countries are undernanced, unmoti-vated, unclear as to how the law applies, unfamiliar with eco-nomic issues, or even corrupt. Such courts cannot be expected toengage in costly verication of the facts of difcult cases orcontingencies of complicated contracts. Indeed, even when con-tracts are restricted by statutes, the courts may not have theresources or incentives to verify whether or how particular stat-utes apply.

Financial contracting illustrates these problems. When is theinformation that a rms manager fails to disclose to shareholdersmaterial, and hence has to be disclosed because of a statute ora contract? When does a corporation abuse minority sharehold-ers, as opposed to just following the managers best businessjudgment? When does a broker fail to engage in honest tradingin executing customer orders? When does a manager trade oninside information rather than simply happen to be lucky? Theinterpretation of the contracts or statutes involving such terms isexpensive, and requires powerful incentives to motivate an adju-dicator to invest in understanding the case. Absent such incen-tives, courts often postpone decisions, or simply let go the poten-tial violators of rules and contracts.

An alternative strategy is the enforcement of legal rules byregulators as opposed to judges. In our view, the crucial distinc-tion between judges and regulators is that the latter can be moreeasily provided with incentives to punish violations of particularstatutes.1 Judges, in contrast, are by design more independentand therefore harder to motivate. The stronger incentives of theregulators have the benet of bringing about more aggressiveenforcement than can be achieved through courts. Yet these

1. The classic reference on the incentive of law enforcers is Becker and Stigler[1974], to whose work we return below. A recent survey of public enforcement oflaw by Polinsky and Shavell [2000] scarcely pays attention to the incentives of theenforcers.

854 QUARTERLY JOURNAL OF ECONOMICS

incentives also have the potential cost of excessively aggressiveenforcement when regulators motivated to nd violations penal-ize innocent suspects. There is thus a trade-off between enforce-ment by judges facing relatively weak but unbiased incentivesand enforcement by regulators facing stronger but possibly biasedincentives.2

We present a theoretical model that sheds light on this trade-off, and identies the circumstances under which enforcement byjudges or regulators is preferred. The model shows that, relativeto judges, regulators may be better motivated to invest in under-standing the laws and circumstances of a case, but also morelikelyif overmotivatedto reach politically desirable decisionsat the expense of doing justice. The model also shows how reduc-ing the costs of the investment in information by law enforcerscan improve enforcement efciency.

We then illustrate the model by comparing the regulation ofsecurities markets through corporate and securities laws in Po-land, the Czech Republic, and to a lesser extent Hungary. Inthese transition economies, nancial regulation was designedessentially from scratch, and hence we can compare both thedesigns of laws and regulations and their consequences. Themodel bears in particular on the design of securities laws, sincethese laws shape the incentives of market regulators as well asthe costs of information acquisition by the enforcers.

We show that, in its securities law, Poland adopted a morestringent regulatory stance than did the Czech Republic. Thisdifference was reected not just in the general philosophies ofregulation, but in the statutes and the mechanisms of law en-forcement. In contrast to the Czech Republic, Poland adoptedlegal rules highly protective of investors, mandated extensiveinformation disclosure by securities issuers and intermediaries,and created an independent and highly motivated regulator toenforce the rules. We nd that this approach to regulation inPoland has stimulated rapid development of securities markets,and enabled a number of rms to raise external funds. Theexpropriation of investors has been relativelymodest. In contrast,the lax regulations in the Czech Republic, enforced by an unmo-tivated ofce in the nance ministry, have been associated with

2. Coase [1988, pp. 117118] recognized that regulation may be preferred tojudicially enforced contracts as a method of regulating some types of conduct:There is no reason why, on occasion, such governmental regulation should not bean improvement on economic efciency.

855COASE VERSUS THE COASIANS

security delistings and a notable absence of equity nancethrough a public market by either new or existing rms. Expro-priation of investors has been rampant, and has acquired a newCzech-specic name, tunneling [Coffee 1996, 1998, 1999; Pistor1999; Johnson et al. 2000b]. Starting in 1996, the Czech govern-ment tightened its regulations. Hungary adopted an intermediateregulatory stance, and has shown an intermediate level of nan-cial development.

II. A MODEL OF ENFORCEMENT INCENTIVES

A. Basic Model

We consider a situation in which the government wishes topunish particular conduct creating negative externalities, such asnondisclosure of material information by a manager or marketmanipulation by a broker. This task is assigned to an enforce-ment ofcial (an adjudicator). The question we address is whetherthe government wants this adjudicator to be a judge or a regula-tor. In the case of a judge, we focus on the inquisitorial legalsystem of civil law countries, where the judge must himself un-dertake an investigation into the facts of the situation and thelaw. The model we present focuses on the case where there is alegal rule or law that restricts certain conduct. The question ofwho should adjudicate, however, equally well applies to a situa-tion in which two private parties such as an investor and a brokercontractually agree on their conduct and have a dispute onwhether this contract was followed.

Our general assumption is that the society does not have fullcontrol over the incentives facing law enforcement ofcials. Its abil-ity to reward them for enforcing the law is limited because doingjustice is largely unveriable. Many of the rewards that theseofcials receive for doing justice are intangible, including self-es-teem and the respect of ones peers. On the other hand, the govern-ment does have the ability to politicize the enforcement of particularlegal rules by rewarding the enforcers for certain outcomes such asnding violations. We are interested in the conditions under whichthe government would choose such politicization.

We consider an adjudicator (who can be a judge or a regula-tor) examining a possible violation of a legal rule. For a cost c >0, this adjudicator can undertake an investigationwhich forsimplicity we call searchand nd out for sure whether a viola-


tion had taken place. We think of c as a personal cost to theadjudicator, which includes the time he might otherwise spendworking on other matters. The adjudicator has complete discre-tion as to whether to penalize the potential violator, and candecide to do so without searching and incurring the cost c.

The adjudicator derives a payoff of b from following the law,or doing justice, which here means punishing a violator of the ruleand letting go an innocent person. We can think of b as self-esteem or long-run respect of the peers, which evidently mattersto judges [Posner 1995]. We assume that, in the short run, thegovernment cannot increase b, since it cannot verify whetherthe adjudicator actually searches or makes correct decisions.Training judges and building up their prestige presumably raisesb, but such policies may take decades to pay off.

In addition, the adjudicator derives the payoff a from eachsuspect he punishes whether or not this suspect actually violatedthe rules. If a = 0, this adjudicator is only interested in justice,and is not motivated by politics or short-run career concerns. Ifa > 0, this adjudicator has a personal interest in nding viola-tions. This can be so for a number of reasons. The state may beconcerned with nding violators of particular rules to achieve itsbroader political goals, such as ghting drugs or persecutingparticular ethnic minorities. More narrowly, only successful pun-ishments of violators may be recorded by the superiors of anenforcer, and hence his future career or budget may be deter-mined by the number of penalties he metes out. Still anotherimportant reason why adjudicators may wish to achieve certainoutcomes is that these improve their career opportunities follow-ing government service [Glaeser, Kessler, and Piehl 2000]. Inprinciple, law enforcement can be heavily politicized, and a couldbe a lot higher than b. We can also imagine the case where a 0, it never pays theadjudicator to sink the cost c and then ignore the information heobtains and be either lenient or abusive. If he searches, he alwayspunishes the violators and lets go the innocent. But before search,it may pay the adjudicator to be either lenient or abusive, de-pending on the magnitudes of a, b, c, and p.

To analyze the adjudicators incentives for enforcement, werst consider his payoffs to the three strategies he can pursue:leniency, abuse, and search. These payoffs are given by

(1) Leniency: (1 2 p)b;(2) Abuse: a + pb;(3) Search: b + pa 2 c.

These payoffs dene the optimal strategies of the enforcer, assummarized in

PROPOSITION 1. Fix b and p. The following strategies are followedfor respective parameter values:

Leniency: a # (1 2 2p)b and c $ (a + b)p;Abuse: a $ (1 2 2p)b and c $ (b 2 a)(1 2 p);Search: c # (a + b) p and c # (b 2 a)(1 2 p).

These conditions divide the space of parameter values into threeregions, as shown in Figure I.3

The interpretation of these conditions is straightforward. Forlow-powered punishment incentives and high cost of search, theadjudicator chooses leniency. For high-powered punishment in-centives and high cost of search, the adjudicator turns to abuse.He only searches for the truth as long as the cost of investigationis low enough that, for low as, he prefers search to leniency and,for high as, he prefers search to abuse.

Even this simple analysis in Figure I has several implica-

3. Note that if a > b, the only equilibrium outcome is abuse.

TABLE I

Not Punish Punish Probability

Innocent b a 1 2 pGuilty 0 a + b p


tions. First, we can think of c as a measure of the efciency of thejudicial system, the cost to the adjudicator of obtaining informa-tion. In principle, c can be reduced through legal and regulatoryreform. In the context of nancial markets, for example, c can bereduced by improving accounting systems and disclosure by issu-ers and intermediaries. The model implies that reductions in thelevel of c always lead to increases in search. For high levels of c,search may not be achievable. Increasing career or nancial in-centives of the enforcers only moves the system from leniency toabusea risk that a society may not wish to take if it prefers theformer to the latter. Put differently, a relatively efcient legalsystemwhich could potentially be designed using appropriatelegal rulesis necessary for achieving just outcomes; without it,it may be better to settle for leniency.

Second, for moderate and low levels of c, increasing incen-tives for punishment may indeed have the effect of moving theadjudicator from leniency to search. Even here, however, signi-cant increases in a move the adjudicator out of search and intoabuse. This analysis cautions against the Becker-Stigler [1974]

FIGURE IA Simple Model of Incentives for Enforcement

The adjudicators incentive for enforcement divides the space of parametervalues into three regions: leniency, abuse, and search.


enthusiasm for the high-powered enforcement incentives as itshows the risk for abuse, particularly in inefcient legal systems.

We can use this model to provide further comparative staticsresults, summarized in

PROPOSITION 2. Assume that b > a and that p < 12 . An increasein adjudicator professionalism, b, always 1) strictly reducesthe region of abuse, 2) strictly increases the region of search,and 3) diminishes leniency for low asto favor searchandexpands leniency for high asat the expense of abuse (Fig-ure II). An increase in the fraction of suspects who are guilty,p, always 1) reduces the region of leniency, 2) expands theregion of abuse, and 3) expands search for low asat theexpense of leniency and reduces it for high asto favorabuse.

The intuition behind these results is straightforward. Anincrease in the adjudicators concern for justice raises his aver-sion to both letting the guilty go (resulting from leniency) and

FIGURE IIComparative Statics: Adjudicator Professionalism (b)

The adjudicators incentive for enforcement divides the space of parametervalues into three regions: leniency, abuse, and search. Increasing adjudicatorprofessionalism (b) reduces the region of abuse.


punishing the innocent (resulting from abuse). As a consequence,for a broader range of parameter values, he conducts a search.Since with p < 12 most suspects are innocent, a higher b makesleniency more attractive relative to abuse, further shrinking thelatter region.

An increase in the guilty share of the population, p, obviouslyexpands the range of abuse and contracts the range of search. Forlow incentives, the attractiveness of search rises relative to thatof leniency and hence the scope of search expands. For highincentives, the attractiveness of search falls relative to that ofabuse, and hence the scope of search contracts.

B. An Extension

In the basic model we assume that the fraction of violators, p,is independent of the strategy the adjudicator pursues. Moregenerally, we expect a behavioral response by the potential vio-lators: fewer of them would violate the legal rule if the adjudica-tor searches than if he is either lenient or abusive. In this sub-section we briey consider such a behavioral response.

Suppose that there are many adjudicators, so that the deci-sions of a particular adjudicator have no effect on the pool ofpotential violators. Denote by P the fraction of actual violators inthe population in the equilibrium where all the adjudicators areeither lenient or abusive. This P must be the same in the lenientand the abusive equilibrium, since in both cases the action of thepotential violator has no effect on his fate. Denote by Q < P thefraction of actual violators in the population in the equilibriumwhere all the adjudicators search. If breaking a rule entails costs,the likelihood of violations falls. An adjudicator chooses betweenleniency, abuse, and search taking the behavior of other adjudi-cators, and therefore P and Q, as given. In equilibrium, thechoices of the adjudicators must be consistent with the choices ofthe potential violators.

Figure III presents the structure of equilibria in this modelfor different parameter values. There are now six regions. Asbefore, the area of high search costs and low incentives, denotedby L, has leniency as the only equilibrium. The area of highsearch costs and high incentives, denoted by A, has abuse as theonly equilibrium. The area of low search costs, denoted by S, hassearch as the only equilibrium. In area X, there is a unique mixedstrategy equilibrium, in which the fraction of actual violators isgiven by p* = c/(a + b), adjudicators are indifferent between


search and leniency, and choose them in proportions that makep* be the optimal response by the potential violators. In area Y,there are three equilibria, including pure search, pure abuse, anda mixture of the two with the fraction of actual violators given byp** = 1 2 c/(b 2 a). The reason for multiplicity is that, startingwith the mixed strategy equilibrium in this region, a decision byone adjudicator to become more abusive can increase the incen-tive of the potential violators to break the rule, making abuserather than search more attractive for other adjudicators. Fi-nally, in area Z, there are also multiple equilibria, including pureabuse.

The addition of the behavioral response introduces the pos-sibility of multiple and mixed strategy equilibria (alternatively,different adjudicators do different things). Nonetheless, the gen-eral thrust of the results, including our principal point that pro-viding adjudicators with incentives is desirable for moderate lev-els of investigation costs, is preserved.

C. Implications

What does this analysis imply for the choice of optimal en-forcement incentives? To begin, we can think of a = 0 as the case

FIGURE IIIIncentives for Enforcement with Behavioral Response by Potential ViolatorsThere are six different regions of equilibria.


of true justice, which is perhaps provided by judges truly inde-pendent of the government.We can alternatively think of high asas regulators or prosecutors whose careers and budgets dependnot only on doing justice, but also on nding violations. Onefurther difference between judges and regulators might be thegreater specialization of the latter, leading to lower search cost c,but one can of course imagine specialized judges, as in the casesof bankruptcy or family law. The intermediate as may perhapscorrespond to civil law judges, who are part of the civil service andhence may be dependent on the government, but who at the sametime have less of an incentive to nd violations than regulators do[Ramseyer and Rasmusen 1997]. Using this interpretation, thequestion becomes: Who should enforce a particular legal rule?

The model illustrates the costs and benets of enforcementby judges and regulators. The government must choose the in-centives of an enforcer, namely a (so long as career concerns arenot dominated by outside opportunities), to achieve two objec-tives. The rst is to stimulate search, as opposed to leniency, andthereby to punish the violators (this is the problem that Coasianslargely ignore). The second objective is to achieve justice by notpunishing the innocent (this is the problem that the advocates ofgovernment regulation usually ignore). Increasing a has the bene-t of stimulating search relative to leniency, and thereby makingit more likely that the violators are punished, but also the costof increasing the likelihood of abusethe punishment of theinnocent as well as the violators without search. Put differently,turning the enforcement of a legal rule over to an apolitical judgehas the benet that the innocent would be rarely punished, but ajudgeespecially a judge with a low bwould also tend towardleniency. In contrast, politicizing the system and turning theenforcement to a regulator moves it away from leniency (providedthat this regulator is not captured, i.e., a > 0), but risks abuse.

In principle, the government would wish to have judges withvery high bsa very professional and motivated judiciary whichhas both sufcient incentives to investigate and a strong interestin justice. But this may not be possible. In this event, the modelsuggests that the best enforcement strategyparticularly wheninvestigations are personally expensive (though not prohibitivelyexpensive)may be to have a regulator with a high enough a toget some search but not so high as to risk abuse. How high an athe government chooses would depend on how much it caresabout punishing the violators relative to avoiding punishing the


innocent. Presumably, in the cases where punishing the innocentis particularly expensive to the society, such as criminal law, thecosts of abuse are sufciently high that most governments wouldstill set a low and allocate adjudication to judges. In civil situa-tions, however, the case for regulation is stronger, at least whenc is moderately high. The other way of looking at this is thatenforcement reforms which lower c are likely to stimulate searchand lead to more efcient outcomes, regardless of whether a judgeor a regulator handles the enforcement.

These predictions of the model relate to the case for securitiesmarkets regulationmade by James Landis [1938], the architect ofsuch regulation in the United States and one of the rst SECcommissioners. Landis was skeptical that the courts were moti-vated enough to punish dishonesty in security issuance and trad-ing in a world where the opportunities for promoters and insidersto expropriate investors were extensive. He thought that an in-dependent and highly motivated SEC, whose only objective wouldbe to assure the integrity of nancial markets, could do thisbetter. He also argued that using regulators as adjudicators is abetter strategy because they face lower costs of investigation.Lower costs encourage search and make abuse less likely for agiven level of incentives. The model can thus account for somebasic intuitions for when regulation might be preferred to judicialenforcement.

In the following sections we examine the implications of themodel for nancial regulation in Poland and the Czech Republic(and to a lesser extent Hungary). We examine the reform in twocrucial areas governing nancial markets: corporate law andsecurities law. Corporate law deals in particular with the rela-tionship between corporate insiders and shareholders, and istypically enforced through private litigation. Securities law regu-lates nancial markets. As such it also deals with some aspects ofshareholder protection. In addition, securities law species thestatus and the powers of the securities regulator and deals withdisclosure of information by securities issuers and intermediar-ies. Variation in the securities laws, therefore, can be interpretedas variation in a and c in the model: a more motivated regulatorwould have a higher a, and greater disclosure would correspondto a lower c. We show that Poland and the Czech Republic haveadopted very different strategies toward shareholder protection,especially in their securities laws, and that these strategies canbe interpreted in light of the model. Our evidence suggests that


the greater success of nancial development in Poland than in theCzech Republic might be related to the more appropriate regula-tory stance in Poland, in line with the predictions of the theo-retical analysis.

III. INITIAL CONDITIONS

In broad terms, Poland and Czechoslovakia share similarhistories over the past 50 years. Both countries turned commu-nist and became Soviet satellites shortly after World War II, andspent the next 40 years building socialism. In 1989 the twocountries spearheaded the anticommunist revolution. In Poland,Solidarity won overwhelming support in the June 1989 elections,and by September 1989 was able to form a government. InCzechoslovakia the communists gave up their leading role inthe country in the face of massive protests in November 1989, andthe communist President resigned in December. Free elections inJune 1990 completed a sequence of events that came to be knownas the velvet revolution.

At the beginning of reforms, Poland had a larger populationof 38 million people, compared with 10.3 million in the CzechRepublic. The Czech Republic in 1989 had per capita income of$5727 in constant 1995 U. S. dollars compared with Polands$3045. Both countries were fully industrialized, with an indus-trial structure largely shaped by decades of Soviet-style centralplanning. Both countries border on Western Europe and in par-ticular Germany, although Warsaw is 569 miles from Frankfurtwhile Prague is only 261 miles away.

Both countries initiated economic reforms immediately aftershedding communism. In Poland critical legislation on liberaliza-tion was passed in the fall of 1989, and the key measures cameinto effect on January 1, 1990. Small-scale privatization began inMay 1990, although large-scale privatization started with a whis-per in 1991, ran into political obstacles, and spread over most ofthe 1990s. In Czechoslovakia reforms were also initiated in early1990, with the devaluation of the currency, budget cuts, andbanking reform. The formal reform package, including price in-creases, started on January 1, 1991. The law on large-scale pri-vatization was adopted on February 1, 1991. Privatizationthrough vouchers took place in two waves: in 1992 (completed in


mid-1993) and 1993 (completed in 1994). Most rules of privatiza-tion, including those on Investment Privatization Funds, weredeveloped in 1991 [Coffee 1996].

Moreover, both countries were virtually nished withthese basic reforms by 1994. They received virtually identicalscores on every World Bank indicator of the pace of transition[de Melo, Denizer, and Gelb 1996]. The European Bank forReconstruction and Development also ranked them veryclosely (see Table II). Although the Czech Republic moved morerapidly on large-scale privatization and so had a somewhathigher share of its GDP generated in the private sector, inmatters such as small-scale privatization, governance and re-structuring, price and trade liberalization, competition policy,banking reform, and nancial institutions, the countries are

TABLE IICOMPARISON OF ECONOMIC REFORM POLICIES BY THE EBRD

PolandCzech

Republic PolandCzech


Republic

Transitionindicators 1997



Private sectorshare of GDP 65 75 60 75 60 70

Large-scaleprivatization 3+ 4 3 4 3 4

Small-scaleprivatization 4+ 4+ 4* 4* 4* 4*

Governance andrestructuring 3 3 3 3 3 3

Price liberalization 3 3 3 3 3 3Trade and foreignexchange system 4+ 4+ 4* 4* 4* 4*

Competition policy 3 3 3 3 3 3Banking reformand interest rateliberalization 3 3 3 3 3 3

Securities marketand nonbanknancialinstitutions 3+ 3 3 3 3 3

Scale is from 1 (no reform) to 4+ (full reform).Source: European Bank for Reconstruction and Development [1997, 1996, 1995].


neck and neck and very far advanced.4 In short, both countrieswere rapid and thorough reformers in their emergence fromcommunism, especially in comparison with other transitioneconomies.

There are, however, two differences which we come back tobelow. First, the Czech large-scale voucher privatization wasfaster and more extensive than privatization in Poland, whichover time utilized a variety of methods from direct sales to sharetransfers to mutual funds. As a consequence, the number ofpublicly held companies in the early 1990s was signicantlyhigher in the Czech Republic than in Poland. Second, during thisperiod Poland grew faster but also had higher ination than theCzech Republic. The assessments of growth rates depend onexactly how they are calculated. The level of GDP in Poland in1997 stood at 110 relative to 100 in 1989, whereas in the CzechRepublic it stood only at 90. Using constant 1995 dollars, how-ever, Polands advantage is smaller.5 During 19921997 theCzech ination averaged 13.9 percent per annum, while Polishination was signicantly higher at 26.5 percent.

In legal development, the two countries again appear similar.In the universe of transition economies, both get perfect or nearlyperfect scores, although these scores have only been kept after1995. The European Bank for Reconstruction and Developmentevaluates transition economies on the extensiveness of laws(since 1996), effectiveness of laws (since 1996), and overall legaldevelopment (since 1995). Table III, Panel A, presents the scoresfor Poland and the Czech Republic, which again are close to eachother and as high as those of any transition economy.6 The legalsystems of the two countries, however, lagged behind those of richmarket economies. Freedom House generates an index of equal-ity of citizens under the law and access of citizens to a non-discriminatory judiciary. In 19951996 both Poland and the

4. In 1997 the EBRD gave Poland a 3+ relative to the Czech Republics 3 onsecurities markets and nancial institutions. We argue below that the differenceshould have been larger.

5. The World Bank reports the level of real GDP using constant 1995 pricesbut calculates growth rates using the GDP deator. Given the large changes inrelative prices during reforms, it is hard to know which measure is better. Onevery available measure, however, Poland has had more growth since 1989, andgrew signicantly faster during the 19951998 period.

6. Pistor [1995] assesses the extent of legal development in a number oftransition economies. She gives Poland and the Czech Republic the same score,the highest (shared with Hungary) among all the transition economies shestudies.


Czech Republic received scores of 5 out of 10, compared with 7.5or 10 for the rich industrial countries.7 The 1997 World Competi-tiveness Yearbook [IMD 1997] in its question on the legal frame-work, gave Poland 4.16 out of 9 and the Czech Republic 4.66. Thiscompares with 8.46 for the world leader, Singapore (and overeight generally for rich industrial countries) and the low of 2.35for Venezuela. Finally, the 1996 Global Competitiveness Report[World Economic Forum 1996], in its question on condence inthe fair administration of justice, gives 2.93 out of 6 to the CzechRepublic and 2.92 to Poland. This compares with the high of 5.78for New Zealand, and the low of 1.77 for Russia. All the surveys,then, treat the judicial systems of the two countries as aboutequally advanced, ahead of world laggards yet far behind the richindustrial countries.

These results are echoed by the concerns of knowledgeableobservers about the state of the judicial system in the two coun-tries in the early stages of reform [Gray et al. 1993]. With respect

7. These numbers come from Economic Freedom of the World 1997, by JamesGwartney and Robert Lawson, a publication of The Fraser Institute, a conserva-tive think tank in Canada.

TABLE IIILEGAL ENVIRONMENT

Panel A PolandCzech




Republic

EBRD 1997 1996 1995

Extensivenessof laws 4 4 4* 4 n.a. n.a.

Effectivenessof laws 4+ 4 3 4* n.a. n.a.

Overall 4 4 4 4 4 4

Panel B

Wall Street JournalCEER survey

December 1997January 1998



February1995

Rule of law/legalsafeguards 9 8.7 9 8.8 9.1 9.1 n.a. n.a.

Legal framework 9.8 9.8

Scale for legal extensiveness and legal effectiveness is from 1 (no reform) to 5 (full reform).Scale for rule for law/legal safeguards, and legal framework is from 1 to 10 (the highest/best score).Source: European Bank for Reconstruction and Development [1997, 1996, 1995], and Central European

Economic Review, a supplement of the Wall Street Journal Europe (issues indicated in table).


to Poland, Gray et al. [p. 109] write: Many of the newly appointedjudges lack experience. . . . Developing such expertise will taketime. Lack of experience and expertise creates uncertainty in thebusiness population . . . With respect to the Czech Republic,Gray et al. [p. 59] note: As in other Central and East Europeancountries, judicial institutions in the Czech Republic are ill pre-pared to cope with the rapidly emerging challenges of the marketeconomy . . . Incapacity in the court system is likely to be aconstraint for some time to come.

In summary, the economies and the economic policies ofPoland and the Czech Republic share some remarkable similari-ties during the 1990s. The two countries emerged from socialismwith a need to massively reorganize their economies and pro-ceeded to do so both rapidly and effectively. In many crucialrespects, they followed similar policies toward this goal, andachieved similar results, especially compared with other, lesssuccessful, transition economies.

IV. COMPANY LAW

Recent research shows that investor protection through com-pany laws and commercial codes is an important deterrent ofexpropriation of outside investors, and as such a key determinantof the development of securities markets across countries [LaPorta et al. 1997, 1998, 1999, 2000; Johnson et al. 2000a]. Beforefocusing on securities regulations, therefore, it is important tocompare Poland and the Czech Republic along this dimension.8

La Porta et al. [LLSV 1998] propose six dimensions to evalu-ate how well a commercial code (or company law) protects mi-nority shareholders against expropriation by the insiders, andcombine them into an index of shareholder protection. Table IV,Panel A, presents and explains this index and its components forPoland and the Czech Republic, based on their rst postreform

8. Polands law dates back to the code of 1934, which was modied repeatedlythrough the communist era and in the early 1990s. The Polish commercial codehas both German and French inuences [Gray et al. 1993; Pistor 1999]. Althoughthe Czech Republic also had a commercial code from the 1930s, its laws weremore thoroughly abrogated than those of Poland during communism, and itaccordingly adopted a new commercial code on January 1, 1992 [Gray et al. 1993].The principal inuence on the Czech commercial code was German. In this andthe following sections, we examined the laws adopted in the early 1990s, whichare relevant for nancial development during the 1990s. Toward the end of thedecade, the laws have been revised in both countries, particularly in the CzechRepublic.


TABLE IVCOMPARISON OF LLSV DIMENSIONS

SHAREHOLDER RIGHTS FROM COMMERCIAL CODES

Panel A

Poland CommentLLSVscore Czech Comment

LLSVscore

Proxy-by-mail No Article 405 (proxyin person isallowed)

0 No Article 185 0

Shares blockedbefore generalmeeting ofshareholders

Yes Article 399 (oneweek ahead ofmeeting)

0 Yes (one week aheadof meeting)

0

Oppressedminoritymechanism

Yes Articles 409 and414

1 Yes Can protestdecision ofgeneralassembly

1

Shareholders havepreemptive rightto new issues

No Not mentioned inPolish law

0 No Can be excludedby Articles ofAssociation(Article204(2))

0

Percent of votesneeded to callextraordinarygeneral meeting

10% Article 394 1 10% Article 181 1

Cumulative voting Yes Article 379A combination of

shareholderswith at least20% of theshare capitalcan elect aboard member

1 No Articles 186 and200

51% of the votesis enough toappoint allthe directors.13 of seats goto employeesif at least 500workers

0

Anti-DirectorRights index,calculated as inLLSV

3 2


Denitions used in Panel A(from LLSV [1998]):

One share-one vote Equals one if the company law or commercial code ofthe country requires that ordinary shares carryone vote per share, and zero otherwise.Equivalently, this variable equals one when thelaw prohibits the existence of both multiple-votingand nonvoting ordinary shares and does not allowsetting maximum number of votes per shareholderirrespective of the number rms of shares owned,and zero otherwise.

Proxy by mail allowed Equals one if the company law or commercial codeallows shareholders to mail their proxy vote to therm, and zero otherwise.

Shares not blockedbefore meeting

Equals one if the company law or commercial codedoes not allow rms to require that shareholdersdeposit their shares prior to a generalshareholders meeting, thus preventing them fromselling those shares for a number of days, andzero otherwise.

Cumulative voting orproportionalrepresentation

Equals one if the company law or commercial codeallows shareholders to cast all their votes for onecandidate standing for election to the board ofdirectors (cumulative voting) or if the companylaw or commercial code allows a mechanism ofproportional representation in the board by whichminority interests may name a proportionalnumber of directors to the board, and zerootherwise.

Oppressed minoritiesmechanism

Equals one if the company law or commercial codegrants minority shareholders either a judicialvenue to challenge the decisions of management orof the assembly or the right to step out of thecompany by requiring the company to purchasetheir shares when they object to certainfundamental changes, such as mergers, assetdispositions, and changes in the articles ofincorporation. The variable equals zero otherwise.Minority shareholders are dened as thoseshareholders who own 10 percent of share capitalor less.

Preemptive rights Equals one when the company law or commercialcode grants shareholders the rst opportunity tobuy new issues of stock, and this right can bewaived only by a shareholders vote; equals zerootherwise.

Percentage of sharecapital to call anextraordinaryshareholders meeting

The minimum percentage of ownership of sharecapital that entitles a shareholder to call for anextraordinary shareholders meeting; it rangesfrom 1 to 33 percent.


commercial codes. Neither country allows proxy-by-mail (scorezero), each requires that shares be blocked before the annualmeeting of shareholders (score zero), and neither gives sharehold-ers a preemptive right to new share issues (score zero). They eachrequire 10 percent of the votes to call an extraordinary share-holder meeting (score 1), and each provide the minority share-holders with some opportunities to protest certain majority deci-sions (score 1). The two laws differ in one important dimensionusing this classication: the Polish law allows a signicant (20percent and in some cases less) minority shareholder to elect adirector. Under the Czech law, 51 percent of the votes are enoughto appoint all directors. Overall, Poland ends up with a score of 3out of 6 on anti-director rights, and the Czech Republic with ascore of 2.

To put these scores in perspective, the highest actual share-holder rights score in the LLSV [1998] sample of 49 countries is5. Several common law countries, such as the United States, theUnited Kingdom, and Canada receive this score. Belgium is thelowest in the sample, with a score of 0, but several countriesincluding Italy, Jordan, and Mexico get a score of 1. The averagein the sample is 3. Thus, Poland is average in the world inprotecting shareholder rights through the company law, while theCzech Republic is below the average.

Some additional rules in the commercial codes, not studiedby LLSV [1998], are also more protective of minority shareholdersin Poland (Table IV, Panel B). Poland gives important rights tosignicant minority shareholders (those with either 20 percent ofthe votes or 20 percent of share capital). In Poland, but not in theCzech Republic, this group can demand the appointment of anadditional board of auditors, and not just a seat on the supervi-sory board. This group can also check who attended the generalshareholders meeting, thus keeping the management from ma-nipulating the total number of the available votes. Both countriesgenerally require supermajorities for important decisions, suchas the change in the objectives of the company. Poland grants ashorter term in ofce to directors (three years) than does theCzech Republic (ve years). In one interesting regard, the Czechlaw is more protective of minority shareholders. Article 185 of theCzech 1992 Commercial Code requires that a quorum of 30 per-cent of the total possible votes be present at a general meeting ofshareholders. The Polish Commercial Code does not set any suchquorum (Article 401).


TABLE IV(CONTINUED)

Panel B

Poland Czech Republic

Further rights ofshareholders Oneshare-one vote (forordinary shares) andno limits on votes pershareholder

No Art. 404: canlimit votesof largeshareholders

No Can set max votesper shareholder(Article 180)

Supervisory board andmanagement boardboth elected byshareholders meeting

Yes Articles 377and 366

Yes Articles 194 and200

Shareholdersrepresenting at leastone-fth of shares candemand an additionalboard of auditors

Yes Article377(3)

No Not mentioned inCzech law

Shareholders with 10%of share capitalrepresented atgeneral meeting cancheck the list ofattendance

Yes Article 403 No Article 185

Two-thirds majority ofgeneral assembly orvotes cast needed forlarge purchases (overone-fth of sharecapital) within twoyears of registrationof company

Yes Article 389 No Not mentioned inCzech law

Two-thirds majority ofgeneral assembly orvotes cast needed tochange articles ofassociation or objectsof company

Yes Article 409each sharehas onevotewithoutpreferencesorrestrictions

Yes Article 187

Term of board ofdirectors(management board)

3 years Article 366and 381

5 years Article 194

Bearer shares allowed Yes Article 345 Yes Article 155 and156

Preference sharesallowed (possiblywithout voting rights)

Yes Article 357 Yes Article 159

Quorum of votes neededto be present

None Article 401 30% Article 185


In summary, Polands company law is somewhat more pro-tective of minority shareholders than the Czech law. These dif-ferences in themselves, however, do not appear to be signicantenough to account for the differences in nancial developmentdocumented below.

V. SECURITIES LAW AND REGULATION

Despite the many crucial similarities, the two countries fol-lowed different approaches to reform in terms of the governmentsinterest in regulatory intervention. This difference did not escapethe early observers of the two countries, who viewed Czech eco-nomic policy as more laissez-faire than Polish economic policy.For example, in each of the three years 1994 1996, the conser-vative Heritage Foundation gave the Czech Republic a perfect(from its perspective) score of 1 and Poland a mediocre score of 3on its measure of regulationthe extent to which governmentrestricts economic activity. Along similar lines, Euromoney con-sidered Poland to be riskier for foreign investment and lendingthan the Czech Republic, in part because property rights wereless secure from government intervention.

These observers had every right to form such opinions basedon the pronouncements about markets and market reform comingfrom economic ofcials in the two countries. Vaclav Klaus, theCzech Finance Minister and later Prime Minister, was both tre-mendously articulate and unabashedly antigovernment in hisvision of reforms: We knew that we had to liberalize, deregulate,privatize at a very early stage of the transformation process, evenif we might be confronted with rather weak and, therefore, notfully efcient markets . . . Conceptually it wasat least for merather simple: all you had to do was to apply the economic phi-losophy of the University of Chicago [Klaus 1997, from a 1995speech]. Leszek Balcerowicz, the champion of Polish reforms,was more cautious: The capacity of the state to deal with variousproblems varies, mainly because of varying informational re-quirements. On this basis, one can distinguish on the one hand,the sphere of the states natural competence (legislating andenforcing the law, dealing with other states, for example) and, onthe other hand, its sphere of natural incompetence (a massive anddetailed industrial policy, for example) [1995, p. 176].

These differences revealed themselves most clearly in theregulation of capital markets. The Polish Law of Public Trading


in Securities and Trust Funds was adopted on March 22, 1991,and became effective in early April 1991. The Czech SecuritiesAct was adopted in 1992, and became effective on January 1,1993. Although this Act was passed after privatization hadstarted, nancial institutions, such as Investment PrivatizationFunds (IPFs), apparently did not lobby for or against it. In fact,the Czech rules were established before privatization started andbefore the IPFs existed, and only codied later [Coffee 1996].They were a product of the governments economic philosophy,not lobbying.

In our analysis of securities laws, we focus especially on twoissues. First, we show that there were signicant differences inthe institutions of securities regulation in the two countries,particularly with respect to the independence and the power ofsecurities regulators. We interpret the greater independence andpower of the regulator as an increase in the parameter a in themodel: the incentives of the adjudicator. Second, we show that theissuers and the intermediaries in the two countries faced radi-cally different disclosure requirements, so that the regulators hadvery different access to information. We interpret the greatermandatory disclosure, and the use of intermediaries to enforce itas reductions in the parameter c in the model: the cost of search.

From this perspective on regulation, an examination of secu-rities laws in Poland and the Czech Republic reveals profounddifferences. To begin, the two laws differed in the identity of thegovernment body supervising securities markets. In Poland itwas an independent Securities Commission. In the Czech Repub-lic such a commission was not established initially, and marketswere supervised by the Capital Markets Supervisors Ofce of theMinistry of Finance. The Ministry of Finance during this periodwas rst under Klaus, and later, when he became Prime Minister,remained indifferent to regulating securities markets. Both su-pervisory bodies received the power to generate regulations, toissue and revoke licenses, and to impose nes for violations ofsecurity laws and regulations, but had to refer criminal cases tothe public prosecutor. The criminal channel was scarcely used ineither country. The fact that the Polish Securities Commissionwas independent, and charged solely with supervision of securi-ties markets, is likely to have provided it with greater incentivesto nd violations than those faced by the Czech Ministry ofFinance, with its much broader agenda.

A key difference in the structure of securities laws in the two


countries is in the emphasis on the regulation of intermediaries.The idea of focusing the regulation of securities markets on in-termediaries is sometimes credited to James Landis [Landis1938; McCraw 1984], who reasoned that the U. S. SEC couldmonitor neither the compliance with disclosure, reporting, andother rules by all listed rms, nor the trading practices of allmarket participants. Rather, the SEC would regulate intermedi-aries, such as brokers, accounting rms, investment advisors,etc., placing on them the burden of assuring compliance withregulatory requirements by issuers and traders. By maintainingsubstantial administrative power over the intermediaries, includ-ing the power to issue and revoke licenses, the Commission couldforce them to monitor market participants. Moreover, the inter-mediaries would be relatively few in number, and more concernedwith their own reputations with the SEC compared with most ofthe issuers. By privatizing part of the enforcement of disclosure tothe intermediaries, the regulator could reduce the share of theenforcement costs he had to bear himselfa reduction in c in ourmodel.

Table V compares the two laws from the perspective of theregulation of nancial intermediaries. In the regulation of indi-vidual brokers, Poland instituted relatively elaborate licensingrequirements, accompanied by tests. Brokers were supposed toengage in honest trading as interpreted by the Commission, andcould lose their license. The Czech Republic had much more proforma licensing of brokers, with easy exams, no warning concern-ing honest trading and evidently no real power of the Commis-sion to revoke licenses. The Polish Commission used the broadhonest trading requirement, and its own power to interpret it,to discourage brokers practices that might not have served theinterests of clients.

Brokerage rms were also licensed in both countries butfaced considerably stiffer regulations in Poland. For example, theregulator received the right to access and inspect the books ofbrokerage rms, and these rms had to disclose their ownershipstructure, stay away from trading in the securities issued by aparent or a subsidiary company, and retain organizational andnancial separateness from banks which owned some of them.These regulations did not exist in the Czech Republic. It is clearthat the Czech Republic adopted a very hands-off stance towardbrokers and brokerage rms, in contrast to Poland.

The Czech Securities law contained no regulation of invest-


TABLE VREGULATION OF INTERMEDIARIES


Individual brokers

Licensed by securitiesmarket regulator

Yes Articles 18.2and 14.1

Yes Section 49

Must pass examadministered bysecurities marketregulator

Yes Article 14.1(4) No Section 49

Required to engage inhonest trading and actin the interest of clients

Yes Article 17.1 No Section 49

License can be suspendedor revoked by SecuritiesCommission

Yes Article 16.2and 16.3

Yes Section 49

Brokerage enterprises


Yes Article 18.2 Yes Section 45

Securities market regulatorhas right of access andinspection

Yes Article 26 No Sections 4548

License can be suspendedor revoked by securitiesmarket regulator

Yes Article 25.3 Yes Section 48(2)


Yes Article 25.2(3) No Sections 4548

Must not conduct otherbusiness with the samename

Yes Article 18.6 No Sections 4548

Must report who has morethan 5 percent of votingrights at general meetingof shareholders


Must report any change ofvoting rights for oneperson above 2 percent


Bank engaged in brokerageoperations must haveorganizational andnancial separateness ofdepartment for publictrading in securities


Must not trade securitiesissued by parent orsubsidiary company



TABLE V(CONTINUED)


Investment advisers(rms engaged in advisory activity in the eld of public trading)


Yes Article 33 No Not mentioned inthe Czech law

Must pass exam set bysecurities marketregulator

Yes Article 33.3 No Not mentioned inthe Czech law

Securities market regulatorhas right of access andinspection


License can be suspendedor revoked by securitiesmarket regulator




Must not conduct otherbusiness with the samename


Must report who has morethan 5 percent of votingrights at general meetingof shareholders


Must report any change ofvoting rights for oneperson above 2 percent


Bank engaged ininvestment advisoryoperations must haveorganizational andnancial separateness ofdepartment for publictrading in securities


Must not trade securitiesissued by parent orsubsidiary company


Sources: Poland: Act of Trading in Securities and Trust Funds, 1991; Czech: Securities Act 1992.


TABLE V(CONTINUED)

Poland Czech

Stock markets

Trading must take place ona stock exchange Yes Article 54.1 No

Section 50 of theSecurities Law

Securities regulatorcontrols stock exchangerules Yes No

Not mentioned inCzech law

Securities exchange shouldensure a uniform market Yes Article 57(1) No


Securities exchange shouldensure dissemination ofuniform information onthe value of securities Yes Article 57(3) No


Agreements among anygroups to articiallyraise or lower the priceof securities areprohibited Yes Article 64.3 No


Mutual funds

Mutual funds may beadministered solely bymutual fund companies Yes Article 89.2 No


Mutual fund companies arelicensed by securitiesregulator Yes Article 89 Yes Section 8

Mutual fund company canbe dissolved by securitiesregulator Yes Article 98 Yes Section 37

Mutual fund companiesmust be joint stockcompanies Yes Article 90.1 No Section 2

Only registered shares areallowed in mutual fundcompanies (no bearershares) Yes Article 92.2 No


Closed-end funds areallowed No Article 104 Yes

Founder limited to 10% ofshare capital Yes Article 93(1) No


Founder not allowed to beon Management Board Yes Article 93(1) No


Publicly traded securitiesor governmentobligations Yes Article 107 No Section 17


ment advisors; the Polish law contained substantial regulations,including licensing. The Polish law restricted trading to takeplace on a stock exchange, and regulated these exchanges to

TABLE V(CONTINUED)

Poland Czech

No more than 5% of thefunds assets can be insecurities issued by oneissuer Yes Article 108 No Section 17

Custodian banks (for mutual funds)

All fund assets must beentrusted to a trusteebank Yes Article 112.1 Yes Section 31

Trustee bank must makesure that sale andretirement ofparticipation units in thefund are consonant withthe law and house rulesof the fund Yes

Article112.2(2) No


Trustee bank mustcompute the net worth ofthe funds assets Yes

Article112.2(3) No


Trustee bank must notexecute instructions thatare in conict with thelaw or house rules of thefund Yes

Article112.2(4) No


Trustee bank must makesure income of the fundis made public Yes

Article112.2(6) No


Trustee bank may not be afounder of the mutualfund company, or a buyerof its securities, or theadministrator of thecompany Yes Article 113.1 No


Mutual fund company maynot buy securities issuedby the trustee bank or arelated company Yes Article 113.2 No


Source: Polish Act of Trading in Securities and Trust Funds, 1991; Czech Investment Companies andInvestment Funds Act April 1992 and Stock Exchange Act 1992.


ensure some transparency in trading. The Czech law did notinclude such regulations. The Polish law contained detailed regu-lations of mutual funds, and in fact for several years the entryinto this activity was severely limited. The Czech law took a muchmore lenient approach again. Finally, the Polish law containedstringent regulations of custodian banks, which are an importantcheckpoint for changes in ownership that might facilitate tunnel-ing. The Czech law again was less restrictive.

Finally, the Polish Securities law, to a much greater extentthan the Czech law, established administrative procedures en-abling the securities market regulator to discipline the interme-diaries without recourse to the judicial system. The intermediariescould then appeal the decisions of the regulator to administrativecourts, but then they, rather than the regulator, had to face thedelays and the inefciency of the judicial system. Because the judi-ciary in neither country is corrupt, the regulators had little fear oftheir lawful decisions being overturned.

Table VI compares the two original laws from the perspectiveof the regulation of security issuers, especially in the area ofdisclosure. Recall that greater disclosure of nancial informationcan serve to reduce the cost of information acquisition by a regu-lator or a judge. In Poland the introduction of securities to publictrading required both permission of the regulator and a prospec-tus. The Czech law required neither. The Polish law requiredmonthly, quarterly, semiannual, and annual reporting of nan-cial information; the Czech law only the annual results. ThePolish law required disclosure of all material information; theCzech law only that of signicant adverse developments.

Financial results are one area where disclosure may be im-portant; ownership structure is another. The Polish law requireddisclosure of substantial minority shareholdings; the Czech lawdid not. Indeed, under the original Polish law, a shareholdercrossing 10, 20, 33, 50, 66, and 75 percent ownership stakes hadto publicly disclosure his ownership. The lack of disclosure ofminority shareholdings has been seen as a problem in several WestEuropean countries, since it enables anonymous large shareholdersto collude with management and expropriate minority shareholders[European Corporate Governance Network 1997]. Finally, the orig-inal Polish law also required a mandatory bid for the remainingshares when a 50 percent ownership threshold was reached; theCzech law did not. Such mandatory bids, combined with disclosureof ownership, are intended to prevent the expropriation of minority


TABLE

VI

REGULATIO

NOFLISTEDCOMPANIES

Polan

dCzech

Republic

Regulation

oflisted

compa

nies

Introductionof

securities

into

publictrad

ingrequ

ires

perm

ission

ofthesecurities

regu

lator

Yes

Article

49No

Not

men

tion

edin

Czech

law

Introductionof

securities

into

publictrad

ingrequ

ires

aprospectus

Yes

Article

50.2

No

Not

men

tion

edin

Czech

law

False

statem

entin

prospectusisforbidden

Yes

Article

118

Yes

Section

79Mon

thly

reportingof

nan

cial

inform

ation

Yes

Reg.of

Sec.C

omm.an

dStock

Exchan

geNo

Not

men

tion

edin

Czech

law

Qua

rterly

reportingof

nan

cial

inform

ation

Yes

Reg.of

Sec.C

omm.an

dStock

Exchan

geNo

Not

men

tion

edin

Czech

law

Sem

iannual

reportingof

nan

cial

inform

ation

Yes

Reg.of

Sec.C

omm.an

dStock

Exchan

geNo

Not

men

tion

edin

Czech

law

Annua

lreportingof

nan

cial

inform

ation

Yes

Reg.of

Sec.C

omm.an

dStock

Exchan

geYes

Section

80

Obligationto

publishallmaterialinform

ation

Yes

Reg.of

Sec.C

omm.an

dStock

Exchan

geNo

Section

80just

sign

icant

adversedevelopm

ents

Constraints

onpu

rchasers/potential

controlling

shareh

olders

Transparency

ofow

nership

requ

irem

ent

Yes

No

CentreforSecurities

canchan

geow

nership

withoutdisclosure

Thresholdat

whichmustdeclarestak

e(percent)

None

10Yes

Article

72No

Not

men

tion

edin

Czech

law

20Yes

No

Not

men

tion

edin

Czech

law

33Yes

No

Not

men

tion

edin

Czech

law

50Yes

No

Not

men

tion

edin

Czech

law

66Yes

No

Not

men

tion

edin

Czech

law

75Yes

No

Not

men

tion

edin

Czech

law


Form

ofdisclosure

requ

ired

toSecurities

Com

mission

Yes

No

Not

men

tion

edin

Czech

law

ToAnti-Mon

opolyOfce

Yes

No

Not

men

tion

edin

Czech

law

Tocompa

ny

Yes

No

Not

men

tion

edin

Czech

law

Com

panymust

announce

whoow

nsmorethan

10%

Yes

In2national

Polishnewspap

ers

No

Not

men

tion

edin

Czech

law

Thresholdat

whichmustmak

egeneral

offer

Must

mak

eofferifintendto

pass

specied

threshold

forow

nership

stak

eYes

Anyperson

whointend

sto

acqu

ire

shares

inonecompa

ny,

once

orby

way

ofrepeated

tran

sactions,

becomingwithin

12mon

thsthe

holderof

shares

inan

amou

nt

that

guaran

tees

him

reachingor

surpassing33

percentof

votes

atthegeneral

meeting,

shallbe

obligedto

doso

solely

byway

ofpu

blicinvitation

tosubscribefor

thesale

ortheexchan

geor

shares

...(Article

73)

No

Not

men

tion

edin

Czech

law

Must

mak

eofferifactual

ownership

stak

epa

sses

specied

threshold

Yes

Any

person

who

hasbecomeaholder

ofshares

inonecompany

representing

over

50percentofthe

votesat

thegeneralm

eeting,shall

beobliged,prior

toexercising

any

powersresultingfrom

therightto

vote,toannoun

cean

invitation

tosubscribeforthesaleor

exchange

oftheremaining

shares

inthat

company

(Article87).

No

Not

men

tion

edin

Czech

law


TABLE

VI

(CONTIN

UED)

Polan

dCzech

Republic

Tenderofferrules

Not

allowed

tohidebehind

arelatedcompa

nyYes

Can

not

hide

behinddepen

dent

subject(Article

72(2),Article

73(2))

No

Not

men

tion

edin

Czech

law

Transactionsin

shareon

thestockexchan

geshou

ldbe

suspen

ded

Yes

Alltran

sactionsin

this

shareon

the

stockexchan

geshould

besuspended

No

Not

men

tion

edin

Czech

law

Tim

elimitforsubscribing

Yes

25da

ysNo

Not

men

tion

edin

Czech

law

Must

buyalltheshares

offered

Yes

No

Not

men

tion

edin

Czech

law

Speciedpriceforpu

rchase

...the

Com

mission

may

forbid

the

announcingof

theinvitation

if...the

priceofferedin

theinvitation

islower

by10

percentthan

theaverage

marketpricedu

ring3months

immediately

precedingthe

announcem

entoftheinvitation.

(Article74)

The

priceoffered...can

notbe

lower

than

thehighestpricepaid

therebyfor

theshares

inthelast

12months,or

whereno

such

pricewas

paidthe

averagemarketpricein

thelast

30days

before

thean

noun

cementof

the

invitation.T

hepriceshallalso

beregarded

asthevalueofthings

orrigh

tsintend

edto

begivenby

the

inviting

person

inexchan

geforshares

Conditionsunderwhichcangoprivate

Securities

Com

mission

must

approve

Yes

Not

specied

Sources:P

olishan

dCzech

SecuritiesLaw

s.


shareholders in tender offers, since they force an acquirer to buy outminority shareholders when he gains control.

This evidence shows that Poland chose to regulate its secu-rities markets more stringently than the Czech Republic. In linewith the model, its law provided for extensive disclosure of nan-cial and ownership information, a way to reduce c and thus tofacilitate regulation (as well as private governance). The Polishreliance on nancial intermediaries to ensure nancial disclosurecan also be seen as a reduction in c. Also in line with the model,Poland relied on administrative control over markets by a moti-vated securities regulator, an increase in a relative to judicialenforcement. This could, in principle, motivate the regulators tobecome informed and reduce the likelihood of leniency. We nextconsider whether this approach worked.

VI. OUTCOMES

A. Qualitative Assessments

Stable prices, rapid privatization, and openness to the Westcombined to generate favorable initial assessments of the Czecheconomic reforms. By 1996, however, there was mounting evi-dence of systematic expropriation of minority shareholders byInvestment Privatization Funds and company insiders colludingwith them. Coffee [1996], who rst presented his paper in 1994,drew attention to such expropriationwhich came to be knownas tunneling. In a typical scheme, the managers of an IPF holdinga large stake in a privatized company would agree with themanagers of this company to create a new (possibly off-shore)entity, which they would jointly control. The IPF might then sellits shares in the company to this entity at below market price,thereby expropriating the shareholders of the IPF. The companycould also sell some of its assets or its output to the new entity,again at below fair value, thereby expropriating its own minorityshareholders. These arrangements between corporate managersand their large shareholders (IPFs) enriched them at the expenseof minority investors in both the rms and the IPFs (see Coffee[1996, 1998] for a discussion of tunneling in the Czech Republic).

The laxity of the securities law accommodated tunneling.First, since transactions did not need to take place on an ex-change, large blocks of shares could change hands off the ex-change at less than the prevailing market price. Even on an


exchange, there was no guarantee of price uniformity. Moreover,brokers and brokerage rms had no restrictions on facilitatingsuch transactions, nor did the custodian banks have any regula-tory duty to stop them. Second, since there was no requirement ofownership disclosure, the acquirers of large blocks could remainsecret. Third, without a mandatory bid, these acquirers had noobligation to buy out the remaining minority shareholders.Fourth, the IPFs appear to have been under no restrictions inpursuing such transactions, since their management did not oweany clearly regulated duty to their investors let alone to theminority shareholders of the companies they tunneled. Fifth,there was no reason to disclose any nancial transactions be-tween the new owner of shares and the company, since suchtransactions were generally allowed and did not need to be dis-closed except perhaps in the annual report several months later.Finally, the minority shareholders had virtually no legal recoursein stopping such expropriation except in a very few cases whenthe oppressed minority mechanism came into play, and evensubstantial minority shareholders could not elect their own di-rectors to represent their interests.

During the mid-1990s, the heyday of tunneling in the CzechRepublic, the regulators did very little to stop it. Part of theproblem was the weakness of the laws. But equally importantwas probably the lack of interest of securities regulators com-bined with judicial ineffectiveness.

By 1996, it became widely believed that something had gonewrong with the regulation of the Czech nancial markets. InMarch 1996 the Central European Economic Review, a publica-tion of theWall Street Journal, surveyed assorted brokerages andfund managers on corporate governance in four transition econo-mies. The survey asked respondents to comment on the disclo-sure of large shareholdings, transparency of markets, quality ofreporting, protection of small shareholders, and insider trading.The Polish market came out as the best of the four, followed bythe Hungarian market. The Czech market came third, ahead ofthe Russian market, which received the lowest score on everydimension. The Polish market outscored the Czech market onevery dimension, with large spreads on the disclosure of owner-ship and transparency. Consistent with this general assessment,the International Federation of Stock Exchanges admitted theWarsaw Exchange as a full member as early as 1994 on thegrounds that the regulation of securities markets met its stan-


dards. As of this writing, the Prague Stock Exchange still had notbeen admitted even as an associate member.

Financial scandals in Poland were typically less egregiousthan those in the Czech Republic, and often invited an aggressiveregulatory response. The best known Polish scandal involves afailure of a large conglomerate, Elektrim, to reveal in a prospec-tus an existing agreement to sell some shares in a valuablesubsidiary to a third party at below market price (allegedly as apayment for services). When the agreement came to light, Elek-trims shareholders complained, and the Securities Commissionquickly referred the case to a public prosecutor. The top managerof Elektrim was forced to step down. The Elektrim case illus-trates the crucial interaction between the corporate and securi-ties law in the enforcement of investor rights. The failure by thecompany to disclose possibly material information in a prospectuswas the source of the Commissions investigation under the se-curities law. This failed disclosure also brought about an effort bythe outside shareholders to change the board of directors usingthe commercial code, which ultimately brought down the topmanager. This interplay between the securities law and the com-pany law appears in other countries as well: the securities lawforces disclosure, which in turn invites shareholder activism us-ing the provisions of company law.

The Polish regulator has also been aggressive in its admin-istrative oversight of the intermediaries. In 1994 Bank Slaski,one of the largest Polish banks which owned the largest broker atthe time, was privatized. In response to the evidence that thebrokerage arm of the bank favored the insiders in allocatingshares in this privatization, the regulators took away its broker-age license. This was done against opposition from the Ministry ofFinance.

The available evidence shows that the Polish regulators re-lied on the actual legal rules to protect investors; it was not justtheir ideology that made a difference. In the cases we examined,they relied on specic rules to promote disclosure that did notexist in the Czech law, consistent with the view that reductions inc matter. The Polish regulator was also evidently motivated topolice the market aggressively, consistent with the view that alevel of a above that of the judges may be benecial. Importantly(and in line with the model), the Polish regulator also had thepower, and not just the motivation, to punish the violations ofrules.


B. Quantitative Assessments

Table VII presents basic indicators of stock market develop-ment in Poland and the Czech Republic. In terms of capitaliza-tion, the Czech market in 1994 was twice as large as the Polishmarket, thanks to the more than 1500 rms listed on the Praguestock exchange as a result of privatization. As a share of GDP, theCzech market in 1994 was ve times larger than the Polishmarket. By 1998, the valuation of the Polish market increasedalmost sevenfold. The valuation of the Czech market increaseduntil 1996, but then fell, and the market ended up at roughlydouble its 1994 value. Over this period the Polish market rose to14.1 percent of GDP, although the Czech market capitalizationremained a larger share of GDP, at 24.2 percent.

Table VII also presents the number of listed companies inPoland and the Czech Republic. It separates the Czech companiesinto those trading on the main market (most liquid), those trad-ing on the secondary market (with more limited disclosure andoccasional trading), and those listed on the free market (withhardly any disclosure and infrequent trading). The listed Polishcompanies are separated into those trading on the main marketand those trading on the parallel (again, less liquid) market. The

TABLE VIISTOCK MARKET SIZE IN POLAND AND THE CZECH REPUBLIC

Marketcaptialization

MarketCap./GDP Number of issues listed

PolandCzech


Republic Poland Czech Republic

US $m, end of year End of year End of year

Mainmarket

Parallelmarket Total

Mainmarket

Secondmarket

Freemarket Total

1991 144 0.19% 9 0 91992 222 0.26% 16 0 161993 2706 3.15% 22 0 22 3 0 966 9691994 3057 5938 3.30% 14.9% 44 0 44 34 0 990 10241995 4564 15664 3.84% 30.8% 65 0 65 62 6 1630 16981996 8390 18077 6.23% 32.0% 83 0 83 42 51 1535 16281997 12135 12786 8.95% 24.6% 143 0 143 45 58 217 3201998 20461 12045 14.10% 24.2% 198 20 218 10 94 179 283

Sources: Polish numbers are from the International Finance Corporation 1998 and 1999 and includeNational Investment Funds; Czech numbers are from the Prague Stock Exchange webpage and the Inter-national Finance Corporation 1997 and 1999.


vast majority of Czech companies barely traded, and most of therms trading on the free market were delisted by the late 1990s.The number of rms on the main market, having risen to 62 in1995, fell all the way down to 10 by 1998, with most of the rmsbeing transferred to the less liquid secondary market. By 1998,most listed Czech rms had been either delisted or transferred toan exchange with only limited liquidity. In contrast, despite amuch lower initial level, the number of listed Polish rms rosesteadily over time, and hardly any rms were transferred to theparallel market.

Figure IV reports the number of Czech and Polish stocks overtime included in the IFC Investable Index compiled by WorldBanks International Finance Corporation, the maker of standardemerging market indices. The IFC Investable Index generallyincludes only the stocks liquid enough that foreign investors canpractically take positions in them. This Index for Poland startedout with 9 stocks in 1992 and rose to 34 stocks in 1998. In theCzech Republic the Index included ve stocks in 1993 and onlythirteen in 1998. Almost all of these thirteen stocks were eithergovernment or foreign controlled. The value of the IFC InvestableIndex in Poland, having started below that in the Czech Republic,has by the end of 1998 far surpassed it (Figure V).

Perhaps the most signicant indicator of success of a nan-

FIGURE IVNumber of Stocks in IFC Investable Index in Poland and the Czech Republic


cial market is how effectively it enables rms to raise capital.Table VIII presents data on the number of initial public offerings(for cash as opposed to vouchers) in the Czech Republic and

FIGURE VMarket Capitalization of Stocks in IFC Investable Index in Poland and the

Czech Republic

TABLE VIIIINITIAL PUBLIC OFFERINGS (FOR CASH)

Czech Republic initialpublic offerings

Poland initialpublic offerings

Issued aspart of

privatization

Issued byprivate

companies

Issued aspart of

privatization

Issued byprivate

companies

1991 0 0 9 1992 0 0 5 21993 0 0 4 21994 0 0 8 141995 0 0 6 151996 0 0 3 151997 0 0 10 361998 0 0 5 52Total 0 0 50 136

Figures do not include the National Investment Funds that were listed on the Warsaw Stock Exchange,or public issues for vouchers in the Czech Republic.

Source: Polish data are from the Warsaw Stock Exchange; Czech data are from Pioneer investment fund.


Poland. It also distinguishes between offerings of shares in pri-vatizing companies coming into public ownership through ota-tion, and offerings by new private companiesthe latter beingperhaps a more effective indicator of a markets effectiveness.Between 19911998 no Czech company sold equity for cash aspart of initial privatization, whereas 50 Polish companies did.9

This is not surprising, since the Czech Republic has followed anoncash privatization strategy. At the same time, the data showthat no private Czech company had done an IPO on the Pragueexchange. By comparison, 136 nonprivatizing companies hadgone public on the Warsaw exchange. This is perhaps the stron-gest evidence of the differential effectiveness of the two markets.

What about the total amount of capital raised on the stockexchange, by both already listed and newly admitted companies?Table IX presents the data since 1996. These numbers are moredifcult to interpret since there have been several rights offeringsin the Czech Republic, for which data are not available. The dataagain show that no new or already listed Czech company raisedequity funds on the exchange through a public offering. In con-trast, the Polish data show rapidly growing equity nancing byboth new and already listed rms. In 1998 over U. S. $1 billion ofnew equity funds was raised on the Warsaw exchange.

9. A foreign-controlled mobile phone company, Ceske Radiokomunikce,raised $134m in 1998 by issuing Global Depositary Receipts in London.

TABLE IXNEW LISTINGS AND EQUITY CAPITAL RAISED

NEW EQUITY CAPITAL RAISED BY DOMESTIC COMPANIES MILLIONS OF U. S. DOLLARS

CzechRepublic Poland

CzechRepublic Poland

Total capitalraised through public

issues

Total capitalraised throughpublic issues

Total capitalraised throughpublic issues

Alreadylisted

companies

Newlyadmittedcompanies

Alreadylisted

companies

Newlyadmittedcompanies

On PragueStock

Exchange

On WarsawStock

Exchange

1996 0 0 319 67.5 0 386.51997 0 0 547.4 438.6 0 9861998 0 0 818.6 328 0 1146.6

Source: International Federation of Stock Exchanges (FIBV), www.bv.com.


This evidence is consistent with both the reading of the lawsand the qualitative assessments. The regulated Polish stock mar-ket grew faster, maintained greater liquidity, and has been abetter source of capital for rms than the less regulated Czechmarket.

After a period of hostility toward any criticism of its policiestoward the stock market, the Czech government introduced anumber of reforms starting in 1996. These included disclosure ofblockholdings, greater regulation (through disclosure and other-wise) of investment funds, restrictions on trading off the ex-change, some separation of investment and commercial banking,and nally, in April 1998, the creation of an independent Secu-rities Commission.

VII. DISCUSSION

The quantitative and the qualitative evidence both point tosignicant problems in the Czech nancial system. Still, we onlyhave one comparison, and our analysis of even this case is subjectto alternative interpretations. Here we discuss some of theseinterpretations.

To begin, our assessment of the Czech situation may be undulyharsh. Overall, the Czech economy performedwell during the 1990sas the transition indicators show. Did the Czech rms simply avoidthe stock market and raise capital elsewhere?

Although we have no data showing that the lack of equitynance has undermined investment by Czech rms, there is noevidence of effective substitute sources of external nance. TheCzech banks have lent predominantly to the largest rms, andhave themselves been subject to governance problems and tun-neling, as evidenced by their huge nonperforming loans. If any-thing, the banking problems exacerbated rather than cured thelack of equity nance. The venture capital industry is also moredeveloped in Poland than in the Czech Republic.

Industrial production grew faster in Poland than in theCzech Republic. Between 1991 and 1998 the index of industrialproduction fell from 113.3 to 109.7 in the Czech Republic, androse from 73.6 to 127.4 in Poland. Much of that growth in Polandcame from new rms, often relying on external equity nance.More generally, the available evidence from other countries sug-gests that stock market development is associated with faster


economic growth and better resource allocation [Levine and Zer-vos 1998; Beck, Levine, and Loayza 2000; Wurgler 2000].

A second concern holds that the Czechs may have only fol-lowed a different strategy of stock market development: oat allthe companies you can through privatization, and then see whichones survive market selection. As part of becoming private, rmscould always commit, in the Coasian style, to charters that wouldobligate them to treat investors well, and thus facilitate externalnance in the future. Even if the Czech strategy resulted inmassive delisting of shares and investor losses in many rms, thettest rms, with most market-friendly charters, survived. SuchDarwinian arguments were made by the Czech reformers in theearly 1990s. Indeed, the Czech market still has more listed com-panies than Polands, and their aggregate value is still higher asa share of GDP. Is this approach to market development obvi-ously inferior?

We believe that it is. In the 1990s the Czech market saw notsome Darwinian selection of the ttest rms, but rather tunnel-ingthe diversion (both legal and illegal) of assets from both goodand bad rms. Coasian contracts bonding rms to treat investorswell did not materialize. The survival of the theft-proof rms isnot an efcient mechanism of economic selection. The most ef-cient rms might be the most attractive to tunnel, making themthe least rather than the most likely to survive. Such tunnelingwas not expected either by the Czech reformers or the investors inthe Czech rms. Moreover, the cost of tunneling has been theinability of both new and existing rms to raise equity capital,which is perhaps the markets main function. It is hard to con-sider this outcome a success even if the government had expectedDarwinian selection.

Even if one agrees that the Czech stock market has not func-tioned admirably, one can still object to our inference that the lack ofregulation is to blame. The leading alternative culprit is mass pri-vatization in theCzech Republic, which led to the listing of hundredsof rms on the stock market. But privatized rms have generallyoutperformed those remaining in government hands in the CzechRepublic [Claessens 1997; Claessens and Djankov 1999; Frydman etal. 1999] as elsewhere [Megginson and Netter 2001]. If anything,privatization could have jump-started nancial markets. Moreover,if the regulation had focused on intermediaries as it did in Poland inthe 1990s, the large number of listed rms would not have been aproblem so long as the number of intermediaries was small. Indeed,


much of the tunneling in the Czech Republic was perpetratedby relatively few IPFs and related intermediaries, which could inprinciple have been regulated with limited resources. Ultimately,the arguments come back to the Czech failure to protect investorsthrough the regulation of intermediaries.

Assuming that the regulation of nancial markets was in-deed inadequate in the Czech Republic, and that this inadequacyhad some costs, why has the system not adapted in other ways tothis regulatory failure? Several such adaptations come to mind.Perhaps private associations of market participants could havebeen created to enforce good conduct among members. Perhapsthe Prague Stock Exchange could design its own, private regula-tions to protect investors, similar to those of Germanys NeuerMarkt [Johnson 2000]. Perhaps the Czech companies could optinto more protective legal regimes, including those abroad,thereby committing themselves to good conduct and accessingexternal nance. Perhaps the Czech companies could individual-ize their corporate charters and incorporate good standards ofconduct into these charters: in principle, a Czech company couldagree to ad

coase v the coaseans

Documents

coase theorem coase

strict enforcement

coasians argument

regulation ofthese markets

securities law

market participants

issuers of securities

developing stock market