UNIVERSITY OF VIRGINIA SCHOOL OF LAW
Legal Studies Working Papers Series
The Common Law and Economic Growth:
Hayek Might be Right
Paul G. Mahoney
Working Paper 00-8
January 2000
This paper can be downloaded without charge from the
Social Science Research Network Electronic Paper Collection
http://papers.ssrn.com/paper.taf?abstract_id=206809
The Common Law and Economic Growth:
Hayek Might be Right
Paul G. Mahoney
University of Virginia School of Law
580 Massie Road
Charlottesville, VA 22903
(804)924-3996; fax (804)924-7536
pgm9h@virginia.edu
I thank Kevin Davis, Ronald Gilson, Barry Ickes, Ross Levine, Katharina Pistor, and participants
at the University of Virginia Legal Studies workshop and the Corporate Governance: Lessons
from Transition Economy Reforms conference at the University of Michigan. I am also grateful
to Ross Levine for access to some of the data used in the paper.
December 28, 1999
1
The Common Law and Economic Growth: Hayek Might be Right
Paul G. Mahoney
Abstract
Recent finance scholarship finds that countries with legal systems based on the common law
provide better investor protections and have more developed financial markets than civil law
countries. These findings echo Hayek’s claims of the superiority of English to French legal
institutions. In this paper, I present evidence that common law countries experienced faster
economic growth than civil law countries during the period 1960-1992. I suggest that the
difference reflects the common law’s greater orientation toward private economic activity and the
civil law’s greater orientation toward government intervention.
2
“[T]he ideal of individual liberty seems to have flourished chiefly among people where, at least for
long periods, judge-made law predominated” (Hayek, 1973, p. 94).
I. Introduction
Recently, economists have produced evidence that financial markets contribute to
economic growth and legal institutions contribute to the growth of financial markets. King and
Levine (1993) demonstrate that the rate of increase in per capita gross domestic product (GDP) is
greater in countries with more developed financial markets. La Porta, Lopez-de-Silanes, Shleifer
and Vishny (1998; 1997) (hereafter LLSV, 1998 and LLSV, 1997) find evidence that the extent
to which a country’s corporate laws protect the interests of minority investors is an important
determinant of the cost of external capital. They also, interestingly, find that countries whose
legal systems are derived from the common law tradition provide superior investor protections on
average, particularly in comparison to the French civil law tradition. Levine (1999) and Levine,
Loayza and Beck (1999) find that better investor protections are associated both with more
developed financial markets and faster economic growth.
Levine (1999), Levine Loayza and Beck (1999) and LLSV (1998) treat legal origin as an
instrumental variable for financial development. Legal origin is well suited to the purpose. It is
largely exogenous, as most countries have had their legal systems imposed by colonization or
conquest. Legal origin also correlates strongly with policies (such as creditor and minority
shareholder protections) that on the basis of theory and empirical results should lead to greater
financial market development. The authors do not, however, provide an explanation for the
correlation. Corporate law seems an unlikely place to find a systematic difference between
3
common and civil law countries. Compared to other areas of commercial law, such as contracts
or commercial paper, corporate law has been largely code-like in the common law countries from
a very early date. This raises the question whether the tendency toward more efficient rules of
corporate law in common law countries is a coincidence that might disappear or reverse in other
areas of commercial law.
This paper explores an alternative possibility–that legal origin is not merely an instrument
for financial development, but is causal in its own right. The motivation for the hypothesis is
Hayek’s (1960) argument for the superiority of English to French legal traditions. Hayek viewed
the decentralized, judge-made common law as an example of spontaneous order, and a
considerable amount of his later work returned to this idea (Hayek 1960, 1967, 1973; see Ogus,
1989 for a critical survey). Because, in Hayek’s view, the spontaneous order represented by the
common law is more consistent with individual liberty than the more rationalist and constructivist
(and therefore more interventionist) tendencies of the civil law, the common law is associated
with fewer government restrictions on economic and other liberties. If common law countries
indeed provide greater freedom to their citizens, they should experience more rapid economic
growth.
Hayek’s views are correct as a matter of legal history. Although legal systems are most
often acquired involuntarily, they were an object of conscious choice in both England and France.
English common law developed as it did because landed aristocrats and merchants wanted a
system of law that would provide strong protections for property and contract rights and limit the
crown’s ability to interfere in markets. French civil law, by contrast, developed as it did because
the revolutionary generation, and Napoléon after it, wished to disable judges from thwarting
4
government economic policies. Thus, quite apart from the substance of legal rules, there is a
sharp difference between the ideologies underlying common and civil law, with the latter notably
more comfortable with a centralized and activist government (Merryman, 1985).
The more complex question is whether the sharp differences in origin and ideology
translate into institutional differences that could affect economic outcomes today. I suggest that
there are structural differences between common and civil law, most notably the greater degree of
judicial independence in the former and the lower level of scrutiny of executive action in the latter,
that provide governments more scope for alteration of property and contract rights in civil law
countries. The law-and-development literature to date, by focusing on differences in the
substance of specific legal rules, has missed this larger picture.
I then report results of cross-country regression analyses showing an association between
the common law and higher rates of real per capita growth in gross domestic product (GDP). I
also provide evidence, derived from Clague, Keefer, Knack and Olson’s (1995) concept of
contract-intensive money, that citizens in common law countries demonstrate greater confidence
in the mechanisms of property and contract enforcement than those in civil law countries.
Section II provides theoretical background by drawing a link between judicial
independence and economic growth. Section III draws on the history of the common and civil
law traditions to show that the two differ sharply in attitudes toward judicial independence and
notes ongoing institutional effects. Section IV reports the results of cross-country growth
regressions. Section V provides additional evidence that the association between the common
law and growth is a consequence of the greater security of property and contract rights from
government interference, and Section VI concludes.
5
II. Theoretical Background
Why should legal origin matter? One possibility is that the quality of legal rules varies
systematically by origin. LLSV (1997, 1998) provide evidence that common law countries have
better systems of corporate law on average. Nevertheless, it is difficult to make out a strong case
for the superiority of the rules produced by the common law or the civil law across the board.
Early scholarship in law and economics contended that the process of litigation should lead to the
survival of efficient rules in a common law system (Posner, 1972; Rubin, 1977; Priest, 1977;
Priest & Klein, 1984). The unspoken implication was that statutory law would be generally less
efficient than judge-made law. More recently, however, claims about the comparative efficiency
of the common law have nearly ceased. Posner appears to have recanted his belief in the
superiority of judge-made to legislative rules (Backhaus 1997). Tullock (1997) provides a heated
refutation of the notion of an “efficient” common law and argues for the superiority of civil law
codification.
Another possibility is that the average quality of rules is similar, but the common law
provides greater stability and predictability. The common law tradition includes two
features–respect for precedent and the power of an appellate court to reverse the legal
conclusions of a lower court–that should result in more predictable outcomes (Manne, 1997).
These features are nominally lacking in the civil law. Only the code itself–not prior judicial
decisions or the pronouncement of a superior tribunal–counts as binding law in the civil law
tradition. Legislatures, unlike common law courts, are not bound by precedent. It is debatable,
however, whether the difference is a sharp in practice as it is in theory. Civil law courts appear in
practice to consult precedents and the decisions of higher courts (Merryman, 1985).
6
The strongest reason to suspect that legal origin would not be important is that both
traditions do well on the most fundamental questions, providing for enforcement of property and
contract rights and requiring compensation for certain wrongful (tortious) acts. As North (1981)
notes, the creation of a system of enforceable property rights is one of the most important
institutional prerequisites to economic growth. The substantive rules of common and civil law
provide redress for private actors’ interference in property or contracts.
The story might end here except that would-be appropriators of wealth can also use the
machinery of government to obtain their ends. Thus Weingast (1993) poses the “ central political
dilemma of an economic system : a government strong enough to protect property rights is also
strong enough to confiscate the wealth of its citizens.” Combined with the theory of public
choice, this insight shows that governments will often respond to the demands of rent-seeking
groups by modifying property or contract rights (Olson, 1982). A growing literature seeks to
identify in political and legal institutions forms of credible commitments not to modify property
and contract rights and thereby to encourage investment. For example, Weingast (1993) and
others identify federalism as a means of decentralizing power and raising the cost of rent-seeking.
Separation of legislative, executive, and judicial powers is another means to the same end.
Persson, Roland and Tabellini (1999) model public finance outcomes under two regimes, a
parliamentary regime and a presidential/congressional regime, in which the latter has a higher
degree of separation of executive and legislative power. The model predicts that the latter regime
will be associated with lower taxation and less redistribution, but at the cost of under provision of
public goods. The paper also provides some confirmatory empirical results. The intuition is
straightforward; it is more difficult for politicians to coordinate in a system of separated powers.
To anticipate a question that might arise in considering this paper’s empirical results, 1
there is not an association between the common law and presidential/congressional regimes;
indeed, there is a mild negative correlation. The two forms of separation of powers may be
substitutes.
7
Separation of powers thus produces less redistribution but also fewer public goods. 1
Judicial independence reinforces the separation of powers between the legislative and
executive branches, on the one hand, and the judiciary, on the other, and we would therefore
expect it to increase the cost of redistribution. Compared to Persson et al.’s argument for
separation between legislative and executive power, the argument for judicial independence is
more complicated and controversial. Landes and Posner (1975) argue that an independent
judiciary (that is, one that does not make decisions based on the current desires of the legislature)
makes interest group deals more secure. The reason is that most legislation is not self-enforcing;
to be effective, it must be enforced by the judiciary. By enforcing the will of the enacting
legislature, the judiciary insulates deals from tampering by future legislatures. The future
legislature can, of course, enact a new statute, but this is more costly than simply putting pressure
on a judge to rule in favor of the legislature’s preferred party. Thus, Landes and Posner conclude,
judicial independence increases the value of legislative deals by making them more durable.
Anderson, Shughart and Tollison (1989) argue that the Landes/Posner analysis does not
adequately explain why a judge would be motivated to enforce the will of the initial legislature. A
truly independent judiciary would not feel constrained to follow the legislature’s intent. Indeed,
there is some evidence that independent judges behave independently–that is, they consult their
own policy preferences when there is no clear answer from the text of a statute or other
Ashenfelter, Eisenberg and Schwab (1995), by contrast, find that a judge’s ideology 2
(proxied by the party of the President who appointed him or her) do not affect the outcomes of
cases at the trial court level in a sample of civil rights cases. They note that in a substantial
number of the cases in the sample, the applicable legal rule was reasonably clear. Thus there is no
necessary conflict between their findings and those of Revesz, who focused on appellate cases
that are more likely to involve the interpretation of facially ambiguous statutes.
8
authoritative text (Revesz, 1997). In that event, judicial independence should function similarly 2
to the separation of legislative and executive power. Independence introduces another actor with
policy preferences that are potentially different from those of the legislature and makes
coordination among those actors difficult. Independence thereby increases the risk that an interest
group will expend resources attempting to obtain favorable policies but come up empty-handed.
Indeed, administrative agencies seem much more likely than courts to play the role Landes
and Posner describe. Many regulatory statutes entrust an administrative agency with the
administration and enforcement of the regulatory program. Civil service laws and statutes
creating these agencies may protect their personnel against termination by the executive based on
policy disagreements. Moreover, those personnel may desire jobs with the regulated industry
after their government service, making them a potent force for protecting the rents created by
regulatory statutes. Here again an independent judiciary, unless it can be counted on to uphold
the agency’s decisions, will decrease the predictability of enforcement and the value of favorable
legislation.
The degree of formal separation between the judiciary and the other branch or branches is
ordinarily greater in common law than civil law countries. In the common law tradition, the
judiciary is a formally separate branch of government and any ordinary judgeship is a prestigious,
well-compensated post. A judge is appointed, typically as the culmination of a successful career
9
as a lawyer, to a specific court in a specific location. As Klerman (1999) notes, common law
judges surely desire promotion to higher, more prestigious courts. However, the ex ante
probability of any given judge advancing is quite small. Many judges on the U.S. federal courts of
appeals, for example, are appointed to those posts as their first (and almost certainly last) judicial
office rather than by promotion from a trial court or state court. The likelihood of advancing to
one of the nine slots on the Supreme Court is so small that it cannot be, for a rational judge, an
important motivator.
By contrast, the new civil law judge is typically a recent law school graduate who has
passed a qualifying examination for entry into a minor judgeship. To remain in the entry-level
post for an entire career would be a clear mark of failure. Prestige is gained by promotions and
postings to more desirable geographical locations. The civil law judge, then, has greater
motivation than the common law judge to gain the favor of the executive branch.
Although there has been very little empirical study of judicial independence, the existing
evidence suggests that the executive uses this motivation to its advantage in civil law countries.
In 1879, the French government dismissed or forced the resignation of thirty-eight administrative
judges who were deemed insufficiently loyal to the government (Brown and Bell, 1998). Using
recent data, Ramseyer and Rasmusen (1997) find evidence that Japanese judges who decide cases
against the government receive less attractive postings than those who find in favor of the
government. By contrast, Salzberger and Fenn (1999) fail to find evidence that the government
of the United Kingdom uses promotions as a tool to reward judges who rule in favor of the
government.
10
III. Historical and Ideological Distinctions
A. Individual versus collective liberty
The common law and civil law both evolved from a combination of Roman law concepts
and local practices and share many substantive traits. The common law and civil law also played
important roles in the creation of the modern English and French constitutional systems. Those
roles were sharply divergent, however, and as a consequence each system has an ideological
content distinct from the substance of particular legal rules.
During the seventeenth century, English common law became strongly associated with the
idea of economic freedom and, more generally, the subject’s liberty from arbitrary action by the
crown. The seventeenth century witnessed the completion of a centuries-long process in which
England’s large landowners pried their land loose from the feudal system and became in practice
owners rather than tenants of the king. Because landowners served as local justices of the peace
and the landowning nobility as judges of last resort, the judges unsurprisingly developed legal
rules that treated them as owners with substantial rights. The common law they created was
principally a law of property. Thus the first of Sir Edward Coke’s (1979 [1628]) Institutes of the
Laws of England is an extensive treatise on the law of real property, structured as a commentary
on Littleton’s (1903 [1481]) earlier treatise that itself is devoted entirely to property law.
Blackstone (1979 [1765]) describes the Court of Common Pleas, which resolved disputes
between subjects, as “the grand tribunal for disputes of property.”
Under the Stuart kings, both landowners and merchants were threatened by claims of royal
prerogative. As described by North and Weingast (1989), the Stuarts faced persistent revenue
shortfalls. The crown responded by coercing merchants to grant it loans, using claims of feudal
See Darcy v. Allen (The Case of Monopolies), 11 Co. Rep. 84b, 77 Eng. Rep. 12603
(1603). The story of the Case of Monopolies is, in fact, somewhat more complicated. As noted
by Corre (1996), the court did not provide such a sweeping condemnation of royal grants of
monopoly as Coke’s report presents. Coke did not publish the report of the case until 1615.
Moreover, Coke, in his capacity as Attorney-General, argued the case for the government,
supporting the validity of the monopoly. After his appointment as Chief Justice, however, Coke
resisted James I’s claims about the extent of royal power, which undoubtedly influenced Coke’s
after-the-fact publication of the case report. Whether Coke’s 1615 report is an embellishment of
what the court actually said in 1603 is not of great moment for present purposes. The key facts
are that by 1615, Coke was willing to declare that the common law trumped royal prerogative,
and that assertion quickly acquired a life of its own.
11
rights to appropriate land and goods, and selling monopoly rights. Disputes over the security of
property and executive intervention in the economy played a central role in England’s two
seventeenth century revolutions.
In those disputes, the common law courts and Parliament (whose members were drawn
from the landowning and merchant groups) took the side of economic freedom and opposed the
crown. The Court of King’s Bench decided, in the Case of Monopolies , that the king’s sale of
monopoly rights violated the common law. The courts, led by Chief Justice Coke, were3
accordingly drawn into confrontation with James I, who insisted that the absolute royal power
trumped the common law. Coke’s insistence in a variety of cases that the common law bound
even the king led James I to dismiss him and like-minded judges. Coke and later common law
judges thereby came to stand for the protection of the rule of law and economic rights against
royal power.
The French experience was very different. Judges, who were heroes in English
constitutional development, were villains in French constitutional development. While security of
economic rights was the motivating force in the development of English common law, security of
executive power from judicial interference was the motivating force in the post-Revolution legal
12
developments that culminated in the Code Napoléon .
The highest courts in pre-Revolutionary France, the Parlements , differed dramatically
from the common law courts in England. They were part court, part legislature, and part
administrative agency. They decided cases, promulgated regulations, and had partial veto power
over royal legislation. As a practical matter, judicial offices were salable and inheritable. The
purchase of a judgeship or other royal office automatically conveyed noble status and qualified the
purchaser and his descendants for entry into the Parlements (Stone, 1986). The return on the
investment was straightforward; in addition to obtaining prestige and various exemptions from
taxation that accompanied noble status, judges enforced the rigidly controlled system of guilds
and monopolies that characterized Bourbon France (Ekelund and Tollison, 1997).
Like the Stuarts in seventeenth century England, the Bourbons faced a fiscal crisis in
eighteenth century France. Having sold monopoly rights over nearly every trade possible and
raised taxes on the peasantry to levels that could not easily be sustained, continuance of royal
consumption and war making required new sources of revenue. Louis XV’s and Louis XVI’s
ministers attempted to address the situation by increasing the role of royal administrators, the
intendents , in the profitable business of enforcing guild and monopoly rights at the expense of the
parlements . This was partly successful; Stone (1986) reports that the prices of judicial offices
declined on average over the course of the century. The crown also attempted to increase the tax
base by eliminating some aristocratic privileges. The parlements , not surprisingly, strongly
resisted these strategies, and the resulting conflict between king and parlements helped ignite the
Revolution.
A central goal of post-Revolution legal reform, then, was to prevent a return of
Discours préliminaire prononcé par Portalis, le 24 thermidor an 8, lors de la présentation 4
du prjet arrêté par las commission du gouvernement, in Fenet (1968 [1827]), p. 465: “le désir
exalté de sacrificier violemment tous les droits à un but politique, et de ne plus admettre d’autre
considération que celle d’un mystérieux et variable intérêt d’état.” The text uses Hayek’s (1960,
p. 195) translation.
13
“government by judges” (Merryman, 1985). A law of 1790 forbade the judiciary to review any
act of the executive (Brown and Bell, 1998). The parlements themselves were shortly thereafter
abolished and replaced with a court of drastically reduced authority. The Civil Code, therefore,
was much more than a simplification and codification of legal rules. As the Code’s principal
drafter explained, it was also the expression of an “ardent resolve to sacrifice violently all rights to
a revolutionary aim and no longer to admit any other consideration than an indefinable and
changeable notion of what the state interest demands.” This assertion of the primacy of political4
over judicial power later dovetailed nicely with Napoléon’s goal of centralizing power in the
executive.
The English experience was that dispersion of authority to judges helped to secure
desirable political and economic outcomes. The French experience was just the opposite. The
authority of the Parlements stalled needed reforms in ancien régime taxation, and the lesson
drawn was that economic and political progress required the centralization of power. The civil
law and common law, then, are closely connected to the more centralizing tendency of French
political thought and the decentralized, individualistic tradition of English political thought.
Hayek (1960) argued that English and French concepts of law stemmed from English and French
models of liberty, the first (derived from Locke and Hume) emphasizing the individual’s freedom
to pursue individual ends and the second (derived from Hobbes and Rousseau) emphasizing the
government’s freedom to pursue collective ends.
14
In this, Hayek echoes many nineteenth and early twentieth century writers. Writing in the
revolutionary year 1848, when these issues were on many minds, Lieber (1881) argued that
Gallican liberty is sought in the government , and according to an Anglican point of view, it
is looked for in a wrong place, where it cannot be found. Necessary consequences of the
Gallican view are, that the French look for the highest degree of political civilization in
organization , that is, in the highest degree of interference by public power. The question
whether this interference be despotism or liberty is decided solely by the fact who
interferes, and for the benefit of which class the interference takes place, while according
to the Anglican view this interference would always be either absolutism or aristocracy. . .
In short, the French concept of law is more congenial to economic intervention and redistribution
as acts of the “general will” and less concerned for the individual’s economic rights.
B. Ongoing Effects
The history just recounted supplements the more abstract discussion of separation of
powers in several respects. The literature on government structure and growth focuses, quite
rightly, on institutional features that constrain self-interested government actors. But it is
sometimes useful to note, as does Buchanan (1975), that “the normative abstractions that we
impose on the actual workings of an imperfect political world are important themselves in placing
limits on action.” Normative abstractions about the nature of law and government differ sharply
between lawyers trained in the common and civil law traditions. The civil law tradition–because
of its historical associations, not its substantive rules–assumes a larger role for the state, defers
more to bureaucratic decisions, and elevates collective over individual rights. These distinctions
may play a role given the prevalence of lawyers in government in many countries.
15
A second important lesson is that the security of property and contract rights can be at risk
from multiple types of actors. Private parties may steal or refuse to honor contracts. The
legislature may enact statutes that alter existing rights. Administrators may use or threaten to use
their authority to extort or redistribute wealth. Much of the development literature, following
North (1981), focuses on the first of these risks, which looms large in a developing country with a
low level of law enforcement. Legal scholars in the United States focus most often on the second,
as the most significant innovation of American constitutionalism was to limit the power of the
national legislature. During the formative centuries for the common and civil law traditions,
however, the third problem was the most important. As Hayek (1960, p. 202) observes, the
concept of the “rule of law” developed during a period in which “an elaborate administrative
apparatus rather than a monarch or a legislature was the chief agency to be restrained.” This is
equally true of many modern states, particularly those with highly-regulated economies.
By the end of the seventeenth century, common law judges in England did not hesitate to
review actions of the executive if alleged to violate the rights of a subject. Among the many
writs, or forms of action, recognized by the royal courts in England were several prerogative writs
that could be issued to government officials. As described by Holdsworth (1903), these writs
were in theory a command by the king to one of his officials who had exceeded his authority.
However, as time passed, the courts began to issue writs such as habeas corpus , mandamus , and
quo warranto at the behest of subjects aggrieved by administrative action. This trend became
pronounced after 1688 and provided subjects with recourse against the arbitrary acts of
government officials.
In France, by contrast, a post-Revolution statute (still in effect) declared “It shall be a
Davy v. Spelthorne Borough Council, [1983] 3 All Eng. Rep. 278, 285 (opinion of Lord5
Wilberforce).
16
criminal offence for the judges of the ordinary courts to interfere in any manner whatsoever with
the operation of the administration, nor shall they call administrators to account before them in
respect of the exercise of their official functions.” France eventually developed a system of
specialized administrative courts authorized to review administrative decisions. The contrast to
common law practice, however, is substantial. The French administrative courts are under the
direct supervision of the executive. Its judges are trained at the administrative schools alongside
the future civil servants whose decisions they will oversee (Szladits, 1974).
These institutional features are supported by a body of substantive administrative law that
insists that the courts intrude as little as possible in the administration’s pursuit of the public
interest (Brown and Bell, 1998). The civil law draws a sharp distinction between private law
(applicable to disputes between citizens) and public law (applicable to disputes involving
government actions). The strong emphasis on property and contract that characterizes the former
gives way in the latter to concern for preserving the government’s freedom to pursue collective
ends (Szladits, 1974). The common law, by contrast, does not recognize a formal distinction
between private and public law. As described by the House of Lords, actions at common law
against public actors do not involve different principles, but different remedies, compared to
actions against private actors. 5
It is accordingly easier for administrators in civil law countries to play the role of enforcers
of interest group deals. Differences in the way common law and civil law courts review
administrative action make it less likely that courts will overturn administrative decisions at the
Grundgesetz, Article 97.6
For a description of the Penn World Tables, see Summers and Heston (1991).7
17
behest of the losing interest group and make interest group deals more secure.
C. The Problem of Germany and Scandinavia
Germany and Scandinavia developed distinct civil law traditions separate from France.
The substance of legal rules is similar but the political associations are not. Codification was not
part of a general upheaval, but rather a gradual process in the various German states and in
Scandinavia from the time of re-discovery of Roman law in the Middle Ages. Germany’s
constitution, moreover, provides for the independence of judges, who cannot be reassigned
without their consent. German courts resemble their French counterparts, however, in the6
fragmented system of jurisdiction and the lower prestige of judges relative to their common law
counterparts. There is accordingly some question how to treat German and Scandinavian civil
law countries given the concerns that motivate my analysis.
IV. Law and Growth: Cross-Country Evidence
In the tradition of cross-country growth studies, I examine differences in average annual
growth in real per capita GDP. The sample consists of 102 countries covered by the Penn World
Tables, Mark 5.6. The data cover the period 1960-1992, and I eliminate any country for which7
real per capita GDP data are missing for more than 3 years of that period. Following LLSV
(1998, 1997), Levine (1999) and Levine, Loayza and Beck (1999), I rely on Reynolds and Flores
(1989) for the descriptions of legal systems. I eliminate countries that do not fit either within the
common law or civil law traditions. These include socialist countries, some Middle Eastern
countries whose legal systems are almost entirely based on Islamic law (such as Saudi Arabia and
18
Oman), and a few countries whose legal systems have been largely free of European influence
(such as Ethiopia and Iceland). Cameroon is excluded because it is an amalgam of former English
and French colonies whose legal system draws almost equally from both traditions. For the time
being, I group all civil law countries together, regardless of French, German or Scandinavian
origin.
These decisions clearly involve some judgment calls. The most difficult involve Africa.
Many of the former British, French and Belgian colonies achieved independence just before or
early in my sample period. Most of the newly independent states went through a period of
socialist or military one-party government during some portion of the sample period. I excluded,
however, only Angola, Mozambique, and Zaire, which had socialist governments for nearly the
entire post-colonial portion of the sample period. Each is a civil law jurisdiction and each
experienced negative per capita GDP growth during the sample period, so their exclusion cannot
have biased the results in favor of the common law. None of the results in the paper are sensitive
to the wholesale exclusion of sub-Saharan Africa except as noted.
I test the effect of the common law using OLS regressions. My dependent variable in each
case is the average annual rate of real per capita GDP growth (GROW). The independent
variable of interest is a dummy (COMMONLAW) that takes on the value 1 for common law
countries and 0 otherwise. The regressions employ conditioning variables drawn from the
economic growth literature. Barro (1991) provides evidence of convergence in economic growth
during the period 1960-1985; countries with initially small economies grew faster than those with
larger economies. I therefore include in the conditioning set the level of real per capita GDP in
1960 (RGDP60). It is also accepted that growth is affected by investment in human capital.
19
Following Barro (1991) and others, I use the rate of enrollment in secondary education in 1960
(SEC60) as a proxy for investment in human capital. Two other common controls are the
average annual rate of population growth during the sample period (GPO) and the average
investment share of GDP over the sample period (INV). Levine and Renelt (1992) survey the
cross-country growth literature and find that RGDP60, SEC60, GPO and INV are the variables
most frequently found in the conditioning set. Those four variables, in addition to
COMMONLAW, are used in a “base” regression. Table 1 provides descriptive statistics for each
variable employed in the base regression.
The first column of Table 2 reports results for the base regression (Model 1). All of the
conditioning variables enter with signs as predicted from theory and prior empirical studies. Initial
per capita GDP and the rate of population growth are both negatively related, and initial
enrollment in secondary education and average investment share of GDP are positively related, to
growth. Only the first and last coefficients, however, are statistically significant, and the first is
not economically significant. The coefficient on the common law dummy variable is both
economically and statistically significant. Controlling for the other variables, the common law
countries grew, on average, 0.64% per year faster than the civil law countries (p = .014).
The use of cross-country growth regressions has attracted some criticism. Levine and
Renelt (1992), for example, note that a substantial number of variables have been found in at least
one regression to be positively and significantly related to growth. They note, however, that most
of these results are not robust to changes in the set of conditioning information.
I accordingly add additional variables drawn from the growth literature using a variant of
extreme-bounds analysis as described by Leamer (1983). The additional variables are taken from
20
Levine and Renelt (1992). They are the rate of primary school enrollment (PRI), the average
government share of GDP (GOV), the inflation rate, defined as the average annual rate of change
of the GDP deflator (INFLATION), the black market premium, defined as the average end of
year ratio of the black market exchange rate to the official exchange rate, minus one (BMP), and
the export share of GDP (EXPORT).
I add the variables in two stages. The first consists of PRI, GOV, and INFLATION. I
estimate seven different regressions that include the base regression variables and the seven
possible combinations of one or more of the three new variables. I then estimate three regressions
that include the base regression and stage one variables and the three possible combinations of
one or both of the two new variables. The variation of the estimated coefficient from one
regression to another provides a test of the robustness of the estimate to changes in the
conditioning set. The motivation for the two stages is that the final two variables appear to
present a greater possibility of endogeneity with the common law variable. I hypothesize that
governments operating under a civil law system will tend to intervene in private economic activity
more than their common law counterparts. The black market premium is partly a measure of the
extent of government restrictions on dealings in currency. Similarly, the extent of exports is
partly a measure of government trade policy. The government share of GDP, by contrast, is a
noisier measure of intervention because regulation is a substitute for direct government spending.
Models 2 and 3 in Table 2 report the results of the regressions that use, respectively, all
three of the stage one variables and both of the stage 2 variables. The coefficients on all of the
new variables are of the expected sign. As found in past studies, initial per capita GDP and the
investment share of GDP are the most robust predictors. The common law dummy, however,
One might also question whether the prevalence of ex-British colonies in East Asia biases 8
the results in favor of the common law. I re-estimated the regression described in the text adding
in a dummy for East Asian countries. The estimated coefficient on the new dummy was positive
but not statistically significant, and the inference regarding the common law dummy was
unchanged.
21
performs quite well. In all seven of the first stage regressions, the estimated coefficient on the
dummy is statistically significant at the 5% level or better, ranging from a high of 0.737 (p = .005)
to a low of 0.585 (p = .029). The same is true for the second stage regressions, the coefficient
ranging from a high of .642 (p =.014) to a low of 0.603 (p = .041).
Table 3 reports the results of regressions that attempt to meet two possible objections to
the analysis thus far. Sub-Saharan Africa and Latin America were notably poor performers during
the period of interest. Latin America consists almost entirely of civil law countries. Any omitted
variable causing low growth in Latin America could, therefore, lead to a mistaken conclusion that
the civil law is to blame. Sub-Saharan Africa contains a mix of common and civil law countries,
but as noted above, many of these countries experienced a period of socialist government during
some part of the sample period. I accordingly estimate the base regression adding in dummy
variables for sub-Saharan Africa and Latin America. The results are reported as Model 4 in Table
3. The inference regarding the common law variable is unaffected. 8
One might also wonder whether common law versus civil law origin, for part of the world,
is merely a proxy for Protestant versus Catholic religious heritage. The package of endowments
received by many former colonies includes, along with the common or civil law, the English,
French, Spanish, or Portuguese language and Protestantism or Catholicism. Weber (1958 [1904-
1905]) argued that Protestant (particularly Calvinist) doctrine encouraged vigorous worldly
pursuits as a means of demonstrating one’s faith and thereby unleashed a “heroic age” of
22
capitalism. I therefore estimate the base regression along with three new variables, the percentage
of the population that practices some form of Protestantism, Roman Catholicism, or Islam,
respectively. The results are reported as Model 5 in Table 3. Although the estimated coefficient
on the Protestant dummy is significant (but negative), the magnitudes of all three estimated
coefficients on the religion variables are very small. Once again, the inference regarding the
common law variable is unaffected.
I also tested the sensitivity of the results to outliers. Removing from the sample the 2
common law countries with the highest growth rates (Singapore and Hong Kong) does not
eliminate the result, nor does removing the 2 civil law countries with the lowest growth rates
(Chad and Madagascar). There is only one standardized residual with an absolute value greater
than 3. Zambia, a common law country, has a strongly negative standardized residual and
therefore is a potential outlier, but its removal would of course strengthen the result. The
association between the common law and growth is not a consequence of outliers.
As discussed above, the German and Scandinavian civil law families can be viewed as
distinct from the French law tradition. There are not enough German and Scandinavian civil law
countries to include separate dummies for each and expect significant results. I did, however,
estimate all of the regressions using a dummy for French civil law in place of the common law
dummy. The absolute values of the coefficients were slightly higher on average compared to the
common law dummy. The result, although far from conclusive, is consistent with the notion that
German and Scandinavian law operate differently from French law.
In short, the common law is associated with additional growth in the range of .59 to .74%
over the 1960-92 period, and the result is robust to a variety of changes in the set of conditioning
23
information. While the link is not statistically as strong as that between investment and growth, it
is a striking result.
V. Testing the intervention hypothesis
It is not straightforward to rank nations based on their governments’ respect for economic
rights. All non-socialist legal systems, for example, include a concept of property and contract.
Both common law and civil law systems include a right of redress for administrative deprivation
of these rights. The differences, if any, must come from differences in the way these rights are
understood and enforced and the effect they have on the level of governmental redistribution.
Clague, Keefer, Knack and Olson (1995) suggest an elegant way of testing the strength of
institutions that enforce property and contract rights. They define “contract intensive money”
(CIM) as that portion of the broad money supply (M2) that does not consist of currency. Citizens
may hold money either in the form of currency or in the form of bank deposits and similar non-
physical media. Clague et al. argue that CIM differs from currency in two important respects.
First, CIM itself represents a contract right–the right of the payee of a check to obtain money
from the drawee bank, for example. Second, although currency is well-suited to simultaneous
exchange, long-term contracting more frequently relies on CIM. Thus, they contend, the ratio of
CIM to the total broad money supply is a measure of the confidence citizens have in the
mechanisms that enforce property and contract rights.
Another feature of particular relevance for present purposes is that CIM is visible to public
actors. Citizens who are concerned about expropriation by government actors or by private
actors using the powers of government will tend to keep as much of their wealth as possible in
forms that cannot easily be monitored. Bank deposits are unattractive in such a setting because
24
governments closely monitor bank activity. For the same reason, currency is the medium of
exchange in the informal economy that seeks to evade regulatory restrictions. As Clague et al.
note, “[w]hen many trades are prohibited, or allowed only at prices that do not clear markets, or
subjected to regulations that induce those with small transactions or limited literacy to operate
outside the formal economy, currency has the advantage of permitting discreet transactions.”
Clague et al. define a CIM ratio as (M2-C)/M2, where C is currency held outside banks.
They provide the average CIM ratio over the period 1969-1990 for 93 of my sample countries (34
common law and 59 civil law). Using this data, I compare the CIM ratio (CIMR) in common and
civil law countries. As reported in Table 4, the average ratio for the common law countries (0.81)
is greater than that of the civil law countries (0.75) and the difference is significant at the 5%
level.
This result is suggestive but must be interpreted with caution because citizens of wealthier
countries may tend to hold more of their wealth in the form of bank deposits and other non-
currency forms, holding the quality of legal institutions constant. Clague et al. suggest a means of
separating the wealth effect from the institutional quality effect. They define the portion of the
CIM ratio that is independent of wealth as the residual from an estimated regression of the CIM
ratio against the log of starting per capita GDP. Table 4 also compares this measure (CIMR2) for
common and civil law countries and finds a stronger relationship. The average residual for the
common law countries is 0.058, while that for civil law countries is -0.029 (p=.000).
In addition to the comparison of means for common and civil law countries, I estimated
my base regression substituting CIMR and CIMR2 for GROW as the dependent variable. In both
cases the estimated coefficient on the common law dummy was positive and significant at the 1%
25
level.
Thus citizens of common law countries, controlling for the size of the economy, hold a
greater portion of their wealth in forms whose value depends on the quality of contract
enforcement and that is useful only for visible, “above-ground” transactions. This suggests a
higher level of confidence both that contracts will be enforced against private actors and that
governmental actors will not interfere excessively in transactions or appropriate wealth.
Berkowitz, Pistor and Richard (1999) consider measures of “legality” or “rule of law”
derived from commercially available country risk guides. They find that legal origin is not a
strong determinant of legality, although the way in which a country acquired its legal system is a
determinant. I obtained, for my sample, the average value of the “law and order” measure from
International Country Risk Guides over the period 1984-1999. As found by Berkowitz et al.,
there is no significant difference between the common law and civil law countries.
A possible explanation for this result is that the idea of the “rule of law” is notoriously
imprecise. Briefly put, the key issue is whether the rule of law is simply a measure of whether
government actors respect the limits put on them, or whether the substance of those limits also
matters. Legal positivism, for example, holds that “law” is simply the command of the sovereign,
made in accordance with established procedures. Under this definition, a government can alter or
abrogate economic rights extensively, but so long as it follows scrupulously the applicable
constitutional procedures, it is behaving in accordance with the rule of law. Others would argue
that “legality” includes some measure of respect for individual political and economic rights, such
as the right of private property. Dicey (1885) notoriously argued that France did not possess the
“rule of law” because ordinary courts were not permitted to review administrative action,
Indeed, the “law and order” measure in International Country Risk Guides is focused in 9
part on the danger of a violent change of government leading to repudiation of government debt.
See Knack and Keefer (1995).
26
touching off a debate that continues to the present day (Brown and Bell, 1998). This confusion is
a problem for measures that necessarily draw on subjective perceptions of legality. It is
particularly an issue for commercial risk guides, whose target market is likely interested
principally in the strength of political institutions and attendant stability. The CIM ratio, by9
contrast, provides a measure of the citizenry’s expectations regarding the security of property and
contract rights.
A somewhat more objective measure is the “economic freedom index” defined by
Gwartney and Lawson (1998). The index is based principally on measurable factors such as
inflation rates, top marginal tax rates, the size of government enterprises as a share of GDP, and
the percentage of total deposits in government-owned banks. It also, however, includes some
survey-based information. I obtained the 1997 economic freedom score for all 97 non-communist
(or recently-communist) countries. The scores differ markedly by region, with east Asia, Europe
and North America scoring highest and Africa lowest. Controlling for region through an OLS
regression using regional dummy variables, the common law countries scored higher than the civil
law countries and the result was significant at a 1% level. Gwartney, Lawson and Block (1996)
provide economic freedom scores for 1975, 1980, 1985 and 1990 for a smaller sample of 88 non-
communist countries.
Using the same procedure for the average of these scores, the common law countries
again score higher, but the result loses significance. My inclusive approach to sub-Saharan Africa
is the reason. Several of the lowest scores come from Uganda, Zambia and Tanzania (each a
27
former British colony) during those countries’ flirtation with socialist rule (or, in the case of
Uganda, during a period in which the common law was suspended). Removing these countries
restores the result. By the time of the 1997 measure, by contrast, every African country used in
the analysis is a multiparty democracy. On that ground, the 1997 measure is a much better test of
the relative effects of a functioning common law and a functioning civil law system.
VI. Conclusion
Law, as economists increasingly recognize, makes a difference. Legal enforcement of
property and contract rights sets the stage for investment and growth. The design of
governmental institutions, embodied in constitutional and administrative law, influences the
relative returns to productive activity and rent-seeking and therefore the rate of growth.
Common and civil lawyers have long debated the relative merits of the two legal
traditions. This paper provides evidence bearing on the debate. Over the period 1960-1992,
common law countries experienced, on average, a bit more than half a percent greater real per
capita GDP growth per year than did civil law countries, controlling for starting per capita GDP,
secondary school enrollment, population growth, investment, and other factors.
This result is not a definitive vindication of the common law. Common and civil law
origin typically come bundled with various other endowments as a consequence of colonization.
Although I have tried to control for some of these (region, religion, educational attainment, and
trade), it is possible that some other attributes of “Englishness” drive the result.
That said, the result is consistent with a longstanding critique of English and French
theories of the role of law in society. Hayek brought this critique into the economics literature
and argued for the superiority of the common law. His preference for the common law should not
28
be surprising, as the ideas underlying common law are more congenial to individualism. This may
be accidental; common law and civil law played their respective roles in English and French
political development as a result of possibly chance associations between judges and economic
and political reforms. Whether by chance or not, however, the two legal traditions have become
associated with different concepts of political liberty.
Historians will perhaps be receptive to the idea that the ideological associations of the
common and civil law affect the behavior of governmental and private actors. Economists will
likely view such a rationale as ad hoc and demand institutional explanations. The two legal
traditions have in fact generated different institutional arrangements. Judges are invested with
greater prestige and insulated more from political influence in common law systems.
Administrative bodies are insulated more from judicial influence in civil law systems. These
differences should lead to more rent-seeking and redistribution in civil law countries. The
evidence in this paper supports that view.
29
References
Anderson, Gary M., William F. Shughart and Robert D. Tollison, 1989. On the Incentives of
Judges to Enforce Legislative Wealth Transfers. Journal of Law and Economics 32, 215-228.
Ashenfelter, Orley, Theodore Eisenberg and Stewart J. Schwab, 1995. Politics and the Judiciary:
The Influence of Judicial Background on Case Outcomes. Journal of Legal Studies 24, 257-281.
Backhaus, Jürgen G., 1997. “Efficient statute law.” Peter Newman, ed., The New Palgrave
Dictionary of Law and Economics, Volume 2. London, Macmillan Reference Ltd.
Barro, Robert, 1991. Economic Growth in a Cross Section of Countries. Quarterly Journal of
Economics 106, 407-43.
Berkowitz, Daniel, Katharina Pistor and Jean-Francois Richard, 1999. Economic Development,
Legality, and the Transplant Effect (working paper).
Brown, L. Neville and John S. Bell, 1998. French Administrative Law. Oxford, Clarendon Press.
Blackstone, William, 1979 [originally published in 1765]. Commentaries on the Laws of England,
Volume 1. Chicago, IL, The University of Chicago Press.
Buchanan, James, 1975. Comment (on Landes and Posner, 1975). Journal of Law and
Economics 18, 903-905.
Clague, Christopher, Philip Keefer, Stephen Knack and Mancur Olson, 1995. Contract-Intensive
Money: Contract Enforcement, Property Rights, and Economic Performance (working paper).
Coke, Edward, 1979 [originally published in 1628]. The First Part of the Institutes of the Laws of
England. New York, NY, Garland Publishing, Inc.
Corre, Jacob I, 1996. The Argument, Decision and Reports of Darcy v. Allen. Emory Law
Review 45, 1261-1327.
30
Dicey, A.V., 1885. Lectures Introductory to the Study of the Law of the Constitution. London,
Macmillan and Co.
Ekelund, Robert B., Jr. and Robert D. Tollison, 1997. Politicized Economies: Monarchy,
Monopoly, and Mercantilism. College Station, TX, Texas A&M University Press.
Fenet, P.A., 1968 [originally published in 1827]. Recueil Complet des Travaux Preparatoires du
Code Civil. Osnabrück, Otto Zeller.
Goring, Rosemary (ed), 1994. Larousse Dictionary of Beliefs & Religions.
Gwartney, James, Robert Lawson and Walter Block, 1996. Economic Freedom of the World,
1975-1995. Vancouver, B.C., Fraser Institute.
Gwartney, James and Robert Lawson, 1998. Economic Freedom of the World Interim Report:
1998/1999. Vancouver, B.C., Fraser Institute. Data retrieved from
http://www.freetheworld.com.
Hayek, Friedrich A., 1973. Law, Legislation and Liberty. Chicago, IL, The University of
Chicago Press.
________, 1967. Studies in Philosophy, Politics and Economics. Chicago, IL, The University of
Chicago Press.
________, 1960. The Constitution of Liberty. Chicago, IL, The University of Chicago Press.
Holdsworth, William, 1903. A History of English Law. London, Methuen & Co. Ltd.
King, Robert G. and Ross Levine, 1993. Finance and Growth: Schumpeter Might be Right.
Quarterly Journal of Economics 108, 717-737.
Klerman, Daniel, 1999. Nonpromotion and Judicial Independence. University of Southern
California Law Review 72, 455-463.
31
Knack, Stephen and Philip Keefer, 1995. Institutions and Economic Performance: Cross-Country
Tests Using Alternative Institutional Measures. Economics and Politics 7, 207-227.
Landes, William M. and Richard A. Posner, 1975. The Independent Judiciary in an Interest-
Group Perspective. Journal of Law and Economics 18, 875-901.
La Porta, Rafael, Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert W. Vishny, 1998. Law
and Finance. Journal of Political Economy 106, 1113-1155.
_______, 1997. Legal Determinants of External Finance. Journal of Finance 52, 1131-1150.
Leamer, Edward E., 1983. Let’s Take the Con Out of Econometrics. American Economic
Review 73, 31-43.
Levine, Ross, 1999. Law, Finance, and Economic Growth. Journal of Financial Intermediation 8,
8-35.
Levine, Ross, Norman Loayza, and Thorsten Beck, 1999. Financial Intermediation and Growth:
Causality and Causes (working paper).
Levine, Ross, and David Renelt, 1992. A Sensitivity Analysis of Cross-Country Growth
Regressions. American Economic Review 82, 942-963.
Lieber, Francis, 1881. “Anglican and Gallican Liberty.” Daniel Coit, ed., Miscellaneous Writings
of Francis Lieber. Philadelphia, J.B. Lippincott.
Littleton, Thomas, 1903 [originally published in 1481]. Littleton’s Tenures in English. Eugene
Wambaugh, ed. Washington, DC, John Byrne & Co.
Manne, Henry G., 1997. What is the “Law” in Law and Economics?: The Judiciary and Free
Markets. Harvard Journal of Law and Public Policy 21, 11-37.
Merryman, John Henry, 1985. The Civil Law Tradition: An Introduction to the Legal Systems of
32
Western Europe and Latin America. Stanford, CA, Stanford University Press.
North, Douglass C., 1981. Structure and Change in Economic History. New York, Norton.
North, Douglass C., and Barry R. Weingast, 1989. Constitutions and Commitment: The
Evolution of Institutions Governing Public Choice in Seventeenth-Century England. The Journal
of Economic History, 64, 803-832.
Olson, Mancur, 1982. The Rise and Decline of Nations: Economic Growth, Stagflation, and
Social Rigidities. New Haven, Yale University Press.
Ogus, Anthony I., 1989. Law and Spontaneous Order: Hayek’s Contribution to Legal Theory.
Journal of Law and Society 16, 393-409.
Persson, Torsten, Gerard Roland and Guido Tabellini, 1999. Comparative Politics and Public
Finance (working paper).
Posner, Richard A., 1972. Economic Analysis of Law. Boston, MA, Little, Brown and
Company.
Priest, George L., 1977. The Common Law Process and the Selection of Efficient Rules. Journal
of Legal Studies 6, 65-82.
Priest, George L. and Benjamin Klein, 1984. The Selection of Disputes for Litigation. Journal of
Legal Studies 13, 1-55.
Ramseyer, J. Mark and Eric B. Rasmusen, 1997. Judicial Independence in a Civil Law System:
The Evidence from Japan. Journal of Law, Economics and Organization 13, 259-286.
Revesz, Richard L., 1997. Environmental Regulation, Ideology, and the D.C. Circuit. Virginia
Law Review 83, 1717-1772.
Reynolds, Thomas H. and Arturo A. Flores, 1989. Foreign Law: Current Sources of Codes and
33
Basic Legislation in Jurisdictions of the World. Littleton, CO, Fred B. Rothman & Co.
Rubin, Paul H., 1977. Why is the Common Law Efficient? Journal of Legal Studies 6, 51-63.
Salzberger, Eli and Paul Fenn, 1999. Judicial Independence: Some Evidence from the English
Court of Appeal. Journal of Law and Economics 42, 831-847.
Stone, Bailey, 1986. The French Parlements and the Crisis of the Old Regime. Chapel Hill, NC,
The University of North Carolina Press.
Summers, Robert and Alan Heston, 1991. The Penn World Table (Mark 5): An Expanded Set of
International Comparisons, 1950-1988, Quarterly Journal of Economics 106, 327-368.
Szladits, Charles, 1974. The Civil Law System, in René David, ed., International Encyclopedia of
Comparative Law, Volume II, Chapter 2, pages 15-76. Tübingen, J.C.B. Mohr.
Tullock, Gordon, 1997. The Case Against the Common Law. Durham, NC, Carolina Academic
Press.
United States Central Intelligence Agency, 1998. The World Factbook 1998. Washington, DC,
Government Printing Office.
Weingast, Barry R., 1993. Constitutions as Governance Structures: The Political Foundations of
Secure Markets. Journal of Institutional and Theoretical Economics 149, 286-311.
Weber, Max, 1958 [originally published in 1904-05]. The Protestant Ethic and the Spirit of
Capitalism. New York, NY, Charles Scribner’s Sons.
34
Table 1
Descriptive statistics, common and civil law countries 1960-1992
Panel A–Full Sample (n = 102)
Mean Standard Deviation Minimum Maximum
GROW 2.06 1.71 -2.00 7.03
RGDP60 2421.98 2309.97 313.00 9895.00
SEC .21 .21 .00 .86
GPO 2.11 1.00 .17 4.08
INV 16.34 7.94 1.40 34.43
Panel B–Common Law Countries (n = 38)
Mean Standard Deviation Minimum Maximum
GROW 2.44 1.81 -.83 6.67
RGDP60 2274.89 2504.96 313.00 9895.00
SEC .22 .23 .01 .86
GPO 2.18 .96 .17 3.71
INV 15.83 7.66 1.47 31.18
Panel C–Civil Law Countries (n = 64)
Mean Standard Deviation Minimum Maximum
GROW 1.83 1.61 -2.00 7.03
RGDP60 2509.31 2201.92 367.00 9409.00
SEC .20 .20 .00 .74
GPO 2.07 1.03 .24 4.08
INV 16.64 8.14 1.40 34.43
35
Table 2
Common law and growth, 1960-1992
Variable Model 1 Model 2 Model 3
COMMONLAW .638 .673 .603
(.256) (.249) (.290)
[.014]** [.008]*** [.041]**
RGDP -.0004 -.0005 -.0005
(.000) (.000) (.000)
[.000]*** [.000]*** [.000]***
SEC 1.702 1.271 1.322
(1.212) (1.157) (1.220)
[.163] [.275] [.282]
GPO -.298 -.203 -.126
(.188) (.182) (.197)
[.116] [.266] [.525]
INV .149 .111 .102
(.022) (.024) (.028)
[.000]*** [.000]*** [.000]***
GOV -.039 -.032
(.020) (.022)
[.059]* [.154]
PRI 1.663 1.984
(.589) (.658)
[.006]*** [.003]***
INFLATION -.003 -.003
(.002) (.002)
[.098]* [.171]
BMP -.0019
(.003)
[.473]
EXPORT .785
(.754)
[.301]
Adjusted R : .479 .545 .5602
(standard errors in parentheses) [p-values in brackets]
* significant at the 0.10 level, ** significant at the 0.05 level, *** significant at the 0.01 level
The dependent variable for all regressions is GROW. For variable definitions, see Appendix B.
36
Table 3
Sensitivity: Region and Religion
Variable Model 4 Model 5
COMMONLAW .627 .629
(.246) (.292)
[.012]** [.034]**
RGDP -.0004 -.0004
(.000) (.000)
[.000]*** [.000]***
SEC .342 1.297
(1.167) (1.228)
[.782] [.294]
GPO -.263 -.379
(.173) (.183)
[.132] [.041]**
INV .125 .157
(.021) (.021)
[.000]*** [.000]***
AFRICA -1.469
(.349)
[.000]***
LATINAM -1.014
(.323)
[.002]***
PROT -.020
(.009)
[.009]***
CATH -.005
(.005)
[.329]
MUSLIM .001
(.006)
[.816]
Adjusted R : .558 .5252
(standard errors in parentheses) [p-values in brackets]
* significant at the 0.10 level, ** significant at the 0.05 level, *** significant at the 0.01 level
The dependent variable for all regressions is GROW. For variable definitions, see Appendix B.
37
Table 4
Legal origin and Contract-Intensive Money
Mean, Mean, t-statistic p value
Common law countries Civil law countries
CIMR .81 .75 2.12 .037(.12) (.14)
CIMR2 .058 -.029 4.40 .000(.091) (.092)
(Standard deviations are in parentheses under means)
CIMR is the average ratio of M2 less currency to M2 for the period 1969-1990. CIMR2 is that
part of CIMR that is orthogonal to the log of 1969 per capita GDP. See Clague, Keefer, Knack
and Olson (1995).
38
Appendix A
Sample Countries
Algeria Indonesia Singapore
Argentina Iran South Africa
Australia Iraq Spain
Austria Ireland Sri Lanka
Bangladesh Israel Suriname
Barbados Italy Swaziland
Belgium Jamaica Sweden
Benin Japan Switzerland
Bolivia Jordan Syria
Botswana Kenya Tanzania
Brazil Korea, South Thailand
Burkina Faso Lesotho Togo
Burundi Liberia Trinidad & Tobago
Canada Luxembourg Tunisia
Central African Republic Madagascar Turkey
Chad Malawi Uganda
Chile Malaysia United Kingdom
Colombia Mali United States
Congo Malta Uruguay
Costa Rica Mauritania Venezuela
Cote d'Ivorie Mauritius Zambia
Cyprus Mexico Zimbabwe
Denmark Morocco
Dominican Rep. Nepal
Ecuador Netherlands
Egypt New Zealand
El Salvador Nicaragua
Finland Niger
France Nigeria
Gabon Norway
Gambia Pakistan
Germany Panama
Ghana Papua New Guinea
Greece Paraguay
Guatemala Peru
Guinea-Bissau Philippines
Haiti Portugal
Honduras Rwanda
Hong Kong Senegal
India Sierra Leone
39
Appendix B–Variable Definition and Sources
Variable name Definition Source
AFRICA Dummy for Sub-Saharan African countries Oxford Atlas of the
World
BMP Average ratio of black market to official exchange LR
rate, minus 1, 1960-89
CATH Roman Catholics as % of population LR, CIA, Goring
CIMR Average ratio of broad money (M2) less currency CKKO
to M2, 1969-1990
CIMR2 Portion of CIMR that is orthogonal to log of 1969 CKKO
per capita GDP
COMMONLAW Dummy for common law origin RF
EXPORT Average export share of GDP, 1960-89 LR
GOV Average government share of GDP, 1960-92 PWT
GPO Average annual population growth, 1960-89 LR
GROW Average annual growth in real per capita GDP, PWT
1960-92
INFLATION Average rate of change of GDP deflator, 1960-89 LR
INV Average investment share of GDP, 1960-92 PWT
LATINAM Dummy for Latin American countries Oxford Atlas of the
World
MUSLIM Muslims as % of population LR, CIA, Goring
PROT Protestants as % of population LR, CIA, Goring
PRI Gross enrollment rate in primary education, 1960 LR
RGDP Real per capita gross domestic product, 1960 PWT
SEC Gross enrollment rate in secondary education, 1960 LR
CKKO: Clague, Keefer, Knack and Goring: Goring (1994)
Olson (1995). LR: Levine and Renelt (1992)
CIA: United States Central PWT: Penn World Tables, Mark 5.6
Intelligence Agency (1998). RF: Reynolds and Flores (1989)
40
