Economic Freedom and Political Freedom

Economic Freedom and Political Freedom
Wenbo Wu
Assistant Professor of Public Policy
Faculty of Arts & Social Sciences
National University of Singapore
Otto A. Davis
William W. Cooper University Professor of Economics and Public Policy
Department of Social and Decision Sciences
Carnegie Mellon University
Forthcoming in Encyclopedia of Public Choice
Please do not cite without permission
Historical evidence speaks with a single voice on the relation between
political freedom and a free market. I know of no example
in time or place of a society that has been marked by a large
measure of political freedom that has not also used something
comparable to a free market to organize the bulk of economic
activity. (Friedman, 1962: 9)
History suggests only that capitalism is a necessary condition for
political freedom. Clearly it is not a sufficient condition (Friedman,
1962: 10)
Economic freedom refers to the quality of a free private market in which
individuals voluntarily carry out exchanges in their own interests. Political
freedom means freedom from coercions by arbitrary power including the
power exercised by the government. Political freedom consists of two basic
elements: political rights and civil liberties. Sufficient political rights allow
people to choose their rulers and the way in which they are ruled. The
essence of civil liberties is that people are free to make their own decisions as
long as they do not violate others’ identical rights. Friedman (1962) points
out the historical fact that economic freedom and political freedom are inextricably
connected. However, the relationships among economic freedom,
civil liberties, and political rights are complex (Friedman, 1991).
In a free private market, individuals have the freedom to choose what to
consume, to produce, and to give. The invisible hand leads free economic
agents to pursue their own interests and voluntarily cooperate with others
(Smith, 1776). Economic freedom and civil liberties are clearly related.
A society whose civil liberties are incomplete is unlikely to sustain a free
private market since civil liberties and economic freedom have in common
the freedom from coercions by other individuals or governments. A free
private market is characterized by voluntary transactions among individuals
who are left alone to pursue their own ends for their economic objectives.
1
The value of political freedom to economic freedom exactly lies in the fact
that civil liberties are defined as including guarantees to limit governmental
power and to protect individual autonomy. Human freedom embedded in
civil liberties is the means through which economic freedom is realized.
The importance of political rights to economic freedom, however, is less
clear. Friedman (1992) points out that ”political freedom, once established,
has a tendency to destroy economic freedom.” He basically believes that the
process of political competition, as determined by political rights, may generate
policies that negatively affect economic freedom. Public choice scholars
have long argued that competitively elected politicians and their agents in
the bureaucracy are self interested and may intervene and disturb the free
market to please their constituencies and sponsors. Individuals enjoying
political rights use democratic forms of government to redistribute wealth
from others often by interfering with the free market, by restricting competition
or limiting sales through the manipulation of prices, or otherwise
creating rents. The misuse of political freedom in democracies has caused
an expansion of services and activities by governments far beyond the appropriate
scope in which economic and human freedoms are protected and
maintained. Inefficiencies of democracy fundamentally impose constraints
on the workings of a free private market and hamper the full realization of
economic freedom (see Buchanan and Tullock (1962), Buchanan. Tollison
and Tullock (1980), Rowley, Tollison and Tullock (1989) and Tullock, Seldon
and Brady (2000)).
A democracy, once established, tends to limit economic freedom to some
degree. More surely, an authoritarian regime is less likely to positively promote
economic freedom. Suppose that a country develops a political system
with a hierarchically structured bureaucratic organization that gives privileges
to an elite class. In such a country, political freedom must be restricted
to serve the elite minority. Even if a market exists, it must not be a true free
private market. Individuals are merely agents of the state and cannot be
2
truly competitive. Moreover, political authorities in an authoritarian regime
tend to distort the market by allocating resources by coercion. The elite
class controls a large part of the resources and effectively controls the entire
spectrum of economic decisions. Economic freedom develops and evolves by
accident, and never by design (Hayek, 1944).
Historically and logically, it is clear that economic freedom is a condition
for political freedom. A core ingredient of economic freedom is private
property which is fundament in supporting political freedom. Without secure
private property and independent wealth, the exercise of political rights
and civil liberties loses its effectiveness. Hayek (1944) maintains that ”Economic
control is not merely control of a sector of human life which can be
separated from the rest: it is the control of the means for all our ends.”
People who depend on the government for their employment and livelihood
have little capacity to oppose the government as they exercise their political
rights. Without rights to own and utilize their properties as they want,
people cannot operate a free media, practice their religions, and so forth.
In the long term, economic freedom leads to and sustains political freedom.
It is no doubt that a free private market is most conducive to wealth
creation (Smith, 1776). A system of economic freedom is superior to any
system of planning and government management (Hayek, 1944). The market
process is a spontaneous order in which resources are efficiently allocated
according to individual needs voluntarily expressed by people. Without any
coercions and deliberate designs, a free private market brings about economic
efficiency and greater social welfare. The wealth effects of economic
freedom create necessary social conditions for political freedom. Fundamentally,
an authoritarian regime that represses political freedom cannot survive
alongside a free private market in the long run. A free private market not
only is a process for achieving the optimal allocation of resources and creating
wealth, which provides material foundations for political freedom, but
also provides an environment for learning and personality development that
3
constructs behavioral foundations for political freedom. However, whether
or not these conditions would lead to democracies depends on other complex
factors, especially, the strategic interactions among various political groups
(Przeworski, 1992).
Relations among economic freedom, civil liberties, and political rights are
complex theoretically and historically. To empirically assess these relations,
we face a hurdle of measuring economic and political freedom. Fortunately,
several serious efforts on measuring the freedoms have recently been made.
Particularly impressive are those measurements with regular upgrades. Rich
panel data sets of economic and political freedoms make it possible to test
various hypotheses developed in the vast theoretical literature on the two
freedoms and to inspire future theoretical development based on insights
derived from empirical analyses.
1 Measuring Economic Freedom
The first attempt to measure economic freedom was undertaken by Gastil
and his associates at Freedom House (Gastil, 1982). Economic freedom
rankings were compiled to complement Freedom House’s political freedom
rankings. Scully and Slottje (1991) later explored a more academically oriented
economic freedom measurement. Soon two major economic freedom
indexes were published by the Heritage Foundation and the Fraser Institute
(Johnson and Sheehy, 1995; Gwartney, Lawson, and Block 1996). The Fraser
Institute and the Heritage Foundation have updated their indexes regularly.
We focus on comparing the Fraser Index and the Heritage Foundation Index.
Both the Fraser Index and the Heritage Foundation Index attempt to
obtain an overall economic freedom ranking for each country during a particular
year based on raw scores on a variety of factors relevant to economic
freedom. They follow a similar procedure that contains the following elements:
defining economic freedom; selecting component variables; rating
4
component variables; combining component ratings into the final overall
rankings of economic freedom.
The Fraser Index defines core ingredients of economic freedom as personal
choice, protection of private property, and freedom of exchange. In the
Heritage Foundation Index, ”economic freedom is defined as the absence of
government coercion or constraint on the production, distribution, or consumption
of goods and services beyond the extent necessary for citizens
to protect and maintain liberty itself.” (O’Driscoll, Holmes and O’Grady,
2002). Both definitions reflect the essence of a free private market. They
represent an ideal state in which a limited government focuses on protecting
private property rights and safeguarding the private market for individuals
to freely engage in exchanges.
Guided by their definitions of economic freedom, the two Indexes identify
areas that are relevant to economic freedom. The Fraser Index selects
twenty-one components under seven areas: size of government; economic
structure and use of markets; monetary policy and price stability; freedom
to use alternative currencies; legal structure and security of private ownership;
freedom to trade with foreigners; and freedom of exchange in capital
markets. The Heritage Foundation Index chooses fifty variables in ten factors:
trade policy; the fiscal burden of government; government intervention
in the economy; monetary policy; capital flows and foreign investment; banking
and finance; wages and prices; property rights; regulations; and black
market activity. Apparently, the two Indexes attempt to cover essential features
of economic freedom: protection of private property; reliance on the
private market to allocate resources; free trade; sound money; and limited
government regulations.
The two Indexes differ significantly in ways in which they rate on components
of economic freedom. The Heritage Foundation Index uses a five-level
grading scale to determine scores for each factor based on information collected
on pertinent factor variables. However, not all factor variables are
5
individually graded. Therefore, raw data on factor variables are combined
into factor grades in a relatively subjective way. There are no explicit formulas
for summarizing information on factor variables into factor grades.
The Fraser Index directly assigns scores to all component variables on a
0-to-10 scale. For continuous component variables, the Fraser Index applies
explicit and fixed formulas to convert original data on component variables
into scores. For categorical component variables, subjective judgments are
applied to obtain scores. Areas of economic freedom in the Fraser Index are
rated solely on the scores of pertinent component variables. In comparison,
scores on the factors in the Heritage Foundation Index seem to be obtained
in a more subjective way than scores on areas in the Fraser Index. However,
this practice allows the Heritage Foundation Index more liberty in using a
wider range of information sources. The Heritage Foundation Index not only
has more factor variables, but also covers more countries and time periods.
The two Indexes also differ in their weighting schemes for combining
component ratings into their final overall rankings of economic freedom.
The Heritage Foundation Index simply weights factor ratings equally. The
Fraser Index uses principle component analysis to construct weights for each
component variables in calculating the final scores of economic freedom. As
pointed out in the Appendix of Wu and Davis (1999b), principal component
analysis allows one to obtain a measure of economic freedom which is statistically
objective in the sense that the final ratings of economic freedom are
directly derived from the data of component variables. The method suits
well for the exercise of measuring economic freedom since the final overall
scores of economic freedom are derived from components that are assumed
to reflect some aspects of the concept of economic freedom. The weights are
based on the principal component that explains the maximum variations
in the original data of component variables among all standardized linear
combinations of the original data.
The Fraser Index claims to develop an objective measure of economic
6
freedom. It is transparent and objective in scoring component variables and
weighting component ratings in the final index of economic freedom. Nevertheless,
these solid steps in the procedure do not change the qualitative
nature of a measure of economic freedom. An economic freedom measure is
not quantitative data such as national income that can be truly objectively
measured. The usefulness of an economic freedom measure lies in the fact
that it provides rankings of different countries over time. In other words,
final scores of economic freedom are ordinal in nature. The Heritage Foundation
Index applies four categories of economic freedom: Free, Mostly Free,
Mostly Unfree, Repressed. The Fraser Index does not provide qualitative
categories like this. Final ratings in the Fraser Index range from 0 to 10,
with a higher number indicating higher degree of economic freedom. We
ought not to mistake these ratings as continuous and cardinal data. The
only valid and usable information in these ratings is the relative degrees of
economic freedom indicated by the scores. For example, in 1997, the Fraser
Index gives Hong Kong a score of 9.4, Albania 4.3, and Chile 8.2. From
these numbers, we can only conclude that Hong Kong is freest economically
among the three, Chile second, and Albania third. The difference between
any two scores cannot be interpreted numerically. For example, we cannot
say that Hong Kong is 119% freer than Albania.
The two Indexes share some similarities and some differences in their
methods of measuring economic freedom. We are interested in whether the
different methods would lead to differences in their final ratings of economic
freedom. To statistically compare the two Indexes, we compile a data set
which includes 238 country-years in 1995 and 1999. The two Indexes overlap
in these two years.1 We use data of the four categories in the Heritage
Foundation Index, and accordingly collapse the inherent ten rankings (based
1In the Heritage Foundation’s annual report (say, in year ”n”) on economic freedom,
authors claim that data in the current reports generally cover the last half of year ”n-2”
and the first half of year ”n-1.” However, it is reasonable to assume that data in the annual
report of year ”n” are representative of situations in year ”n-3” (Cummings, 2000).
7
upon the 0-10 scale) into four categories based on the final rankings in the
Fraser Index.2 Table 1 shows a cross-classification of economic freedom in
the Heritage Foundation Index by that in the Fraser Index. Table 1 clearly
demonstrates a pattern in which observations classified as economically freer
in the Heritage Foundation Index are also classified as economically freer in
the Fraser Index.
Table 1: Cross-Classification of Heritage Foundation Index by
Fraser Index
Heritage Foundation Index
Fraser Index Repressed Mostly Unfree Mostly Free Free
Repressed 5 10 0 0
Mostly Unfree 8 56 14 0
Mostly Free 0 27 79 3
Free 0 0 14 22
The two Indexes classify observations similarly. Among the observations
in Table 1, the total number of concordant pairs3 is: C = 5×(56+14+27+
79+3+14+22)+10×(14+79+3+14+22)+8×(27+79+3+14+22)+
56 × (79 + 3 + 14 + 22) + 14 × (3 + 22) + 27 × (14 + 22) + 79 × 22 = 13223.
The number of discordant pairs of observations is: D = 10 × 8 + 14 ×
2To make the economic freedom ratings explicitly ordinal, it would be a good practice to
assign a few broad categories of economic freedom in the Fraser Index. We construct four
categories of economic freedom as follows (original final rankings in the Fraser Index are
in the parentheses): Free (8-10); Mostly Free (6-7.99); Mostly Unfree (4-5.99); Repressed
(0-3.99).
3In this case, a pair is concordant if the country ranking higher in the Heritage Foundation
Index also ranks higher in the Fraser Index. A pair is discordant if the country
ranking higher in the Heritage Foundation Index ranks lower in the Fraser Index. For
example, consider a pair of observations, one of whom is classified in the cell (Repressed,
Repressed) and the other in the cell (Mostly Unfree, Mostly Unfree). This pair is concordant,
since the second observation is ranked higher than the first both by the Fraser Index
and by the Heritage Foundation Index. Each of the 5 observations in the cell (Repressed,
Repressed) form concordant pairs when matched with each of the 56 observations in the
cell (Mostly Unfree, Mostly Unfree), so there are 5x56=280 concordant pairs from these
two cells. The 5 observations classified as Repressed by both Indexes are also part of a
concordant pair when matched with each of the other (12+27+79+3+14+22) observations
ranked higher in both Indexes.
8
27 + 3 × 14 = 500.
Of the concordant and discordant pairs, 96.36% are concordant and only
3.64% are discordant. The difference of the corresponding proportions gives
a gamma (= (C − D)/(C + D) = 92.72%).
The number indicates that the Heritage Foundation Index and the Fraser
Index are highly correlated. We can further explore the relationship between
the two Indexes by testing the null hypothesis of independence between the
two categorical variables. We simply use a Pearson chi-squared test and a
likelihood-ratio chi-squared test to analyze the cross-classification as shown
in Table 1. The Pearson chi-squared statistic is 218.43 which yields a P-value
less than 0.0001, and the likelihood-ratio chi-squared statistic is 197.49 with
a P-value less than 0.0001 (based on degree of freedom =9). There is very
strong evidence of association between these two measures even if we ignore
the category orderings of the variables.4
2 Measuring Political Freedom
The concept of political freedom and democracy is much debated. Our task
here is not to provide an exhaustive review of the debates, but to point out
empirical conceptions of political freedom and democracy that cover its main
features in the modern world that are relevant to statistical analyses. These
narrow definitions of political freedom and democracy enable us to identify
empirical cases of democracies and non-democracies. Dahl (1971) provides a
useful definition of democracy by emphasizing the procedural characteristics
of the political system. Democracy, as an institutional arrangement, ought
to ensure the following conditions:
1. Freedom to form and join organizations;
4We can exploit the ordinality of the two measures by using the so-called uniform
association model that assigns scores to the rows and columns with a coefficient that
describes strength of association. The assigned scores reflect category orderings and can
be modeled as equal-interval. As expected, the uniform association model predicts a
greatest departure from independence of these two measures.
9
2. Freedom of Expression;
3. Right to vote;
4. Eligibility for public office;
5. Right of political leaders to compete for support (Rights of political
leaders to compete for votes);
6. Alternative sources of information;
7. Free and fair elections;
8. Institutions for making government policies depend on votes and other
expressions of preferences.
When these conditions are met, the elected government is judged to be
responsive to citizens’ preferences, and the democracy and political freedom
are established
Dahl’s eight conditions describe the core of a modern democracy. For
the purpose of empirically measuring political freedom, however, we need to
further condense the definition and establish rules that can help categorize
observations unambiguously. There exist several empirical measurements,
and there is no agreement among scholars regarding the ways of actually
measuring democracy (Bollen, 1980, 1993; Vanhanen, 1990; Przeworski, Alvarez,
Cheibub, and Limongi, 2000; Mainwaring, Brinks, and Perez-Linan,
2001). However, Przeworski et al. (2000) point out
. . . even if regime classification has been the subject of some controversies,
alternative definitions of ”democracy” give rise to almost
identical classifications of actual observations.
We want to compare two representative measurements.5 One is by Freedom
House and the other by Przeworski, Alvarez, Cheibub and Limongi
5Mainwaring, Brinks, and Perez-Linan (2001) provide a trichotomous ordinal classification
of democracy for Latin American countries from 1945 to 1999. They argue that such
10
(2000) (PACL, hereafter). The two measurements are both rule-based in
the sense that both apply pre-determined criteria in identifying democracies.
Nonetheless, Freedom House and PACL appear to represent the two
”extremes” of measuring political freedom and democracy. Freedom House’s
political freedom rankings are based on raw scores assigned by experts, and
hence, seem to be subjective. PACL’s political regime classification exclusively
relies on observables, and attempts to avoid subjective judgments.
Freedom House first differentiates two basic dimensions of political democracy:
political rights and civil liberties. The former mainly refers to the
electoral process. Elections should be fair and meaningful (choices of alternative
parties and candidates, and a universal franchise). The latter implies
freedom of the press, freedom of speech, freedom of religious beliefs, and the
right to protest and organize. The Freedom House measure is comprehensive
and related to multiple dimensions of a modern democracy. Its focus is not
the form of government itself, but upon political rights and the freedom of
citizens caused by the real working of the political system and other societal
factors. To contrast, the PACL measure is concerned with political regimes
as forms of government, and focuses on contestation as the essential feature
of democracy. The authors intentionally exclude political freedom from
their measurement. The narrow definition by PACL is aimed to avoid using
different aspects of democracy (e.g., as defined by Dahl (1970)) that the
authors believe to be of little use. The authors argue ”Whereas democracy
is a system of political rights - these are definitional - it is not a system that
necessarily furnishes the conditions for effective exercise of these rights.”
(Przeworski et al., 2000) Whether or not democracy as narrowly defined
by PACL is associated with political rights and other desirable aspects of
a classification achieves greater differentiation than dichotomous classifications such as
PACL regime classifications, and needs much less information that a fine-grained measure
such as the Freedom House political freedom rankings would require. For our purposes,
however, we need comprehensive rankings for a majority of countries in the world. That is
why we focus on PACL and the Freedom House classifications. Furthermore, Mainwaring
et al.’s measure is highly correlated with the PACL and Freedom House measures.
11
democracy is a question for empirical testing.6
Different concepts and scopes of political freedom put forward by Freedom
House and PACL underpin their rules and criteria for classifying democracies.
Freedom House assigns each country the freedom status of ”Free,”
”Partly Free,” and ”Not Free” based on their ratings in political rights and
civil liberties. To rate political rights and civil liberties in a country, Freedom
House employs two series of checklists for these two aspects of democracy.
For political right ratings, Freedom House uses eight checklist questions and
two discretionary questions. These questions are not only related to formal
electoral procedures but also other non-electoral factors that affect the real
distribution of political power in a country. Freedom House’s civil liberties
checklist includes four sub-categories (Freedom of Expression and Belief,
Association and Organizational Rights, Rule of Law and Human Rights)
and fourteen questions in total. Freedom House maintains that it does not
mistake formal constitutional guarantees of civil liberties for those liberties
in practice.
While the civil liberties component is broadly conceived, the political
rights dimension of the Freedom House measure is more compatible with
PACL’s rules for regime classification.7 These rules exclusively deal with
electoral contestation and government selections. The three basic rules are
labeled as ”executive selection,” ”legislative selection,” and ”party.” The
idea is to identify democracies as regimes in which the chief executive and
the legislature are elected in multi-party elections. The great majority of
6For the sample described below, the gamma (a statistic measuring association between
two ordinal variables) for political rights and civil liberties ratings is 0.91 and the
Pearson correlation is 0.92. Both statistics indicate a strong association between these
two dimensions of democracy in Freedom House’s surveys.
7For the sample described below, the gamma for political rights rating and PACL
regime classification is -0.98, and the Pearson correlation is -0.85. The gamma for civil
liberties rating and PACL regime classification is -0.96, and the Pearson correlation is
-0.80. So the associations between PACL regime classification and the two components
of Freedom House’s democracy ratings are very strong, and the Freedom House political
rights rating is more closely related to PACL regime classification.
12
cases (91.8% percent of country-years in PACL’s sample) are unambiguously
classified by the three rules. PACL further introduces an additional
rule (”alternation”) for those ambiguous cases. The ”alternation” rule is
used to classify countries that have passed the three basic rules. In these
countries, the same party or party coalition had won every single election
from some time in the past until it was deposed by force or until now. For
these cases, we face two possible errors: excluding some regimes that are
in fact democracies from the set of classified democracies (type I error); including
some regimes that are not in fact democratic in the set of classified
democracies (type II error). PACL seeks to avoid type-II errors. Therefore,
there are some regimes that meet the three basic criteria which are
disqualified as democracies.
PACL rigidly and mechanically applies these four rules, but Freedom
House rates countries with discretion. To be ”unbiased”, PACL only needs
to strictly adhere to their rules while Freedom House needs to consciously
maintain a culturally unbiased view of democracy and utilize the broadest
range of information sources. The PACL measure is necessarily consistent
because it exclusively relies on observables and objective criteria. Freedom
House’s checklists and ratings procedures are consistent. However, the ratings
themselves could be inconsistent because of variation in information
sources, raters’ expertise and so on.
The distinctive spirits of the Freedom House and PACL measures are
nicely reflected in their timing rules. Both measures observe countries in a
period of a year. PACL codes the regime that prevails at the end of the
year. Information about the real situation before the end of the year is not
relevant. For example, a country that has been a democracy until the last
day of a year is classified as a dictatorship in the PACL measure. For the
same country, Freedom House would treat it differently and consider the
political development during the whole year and assign appropriate scores.
The information lost in the PACL measure is utilized in the Freedom House
13
ratings, and the loss of information is significant for some cases, especially
for many countries in political transition.
The categorization of political regimes in the PACL measure could be
nominal in the sense that there is no ordering between democracy and dictatorship,
or ordinal that a transition from dictatorship to democracy means
some improvement. PACL’s further classifications of democracy and dictatorship
are nominal in nature. The difference between parliamentary, mixed,
and presidential democracies is meaningful if merely qualitative and definitional.
However, the difference between bureaucracy and autocracy could
be quantitative. PACL classifies a dictatorship with a legislature as a bureaucracy,
and not as an autocracy. Freedom House’s freedom rankings are
explicitly ordinal. The overall statuses of ”Free,” ”Partly Free,” and ”Not
Free” reflect different degrees of political freedom in countries. The base
scores of political rights and civil liberties ratings are themselves ordinal.
Freedom House uses a seven-point scale for the two dimensions of democracy,
with 7 indicating the highest degree and 1 the lowest degree. This
measurement implicitly assumes that there exists a continuum of political
democracy. The two poles are the fully democratic regime (with 1 for both
political rights and civil liberties) and a full autocratic regime (with 7 for
both political rights and civil liberties). The underlying continuum of political
regimes makes it easier to describe and analyze the rich phenomena of
political transitions. Freedom House rankings make it possible to analyze
the intermediate cases of semi-democracies or semi-dictatorships, and the
complicated nature of democratization or reverse-democratization. Under
the PACL classification, there are only two possible transition modes: from
democracy to dictatorship, and from dictatorship to democracy. The simplified
transition modes are less capable of capturing the transitional nature
of political development.
To quantitatively compare the Freedom House and PACL measures, we
use a data set that consists of 2584 country-years during the period from
14
1972 to 1990.This data set includes all observations that have both the
PACL regime classifications and the Freedom House ratings.8 Table 2 shows
a cross-classification of political regime (PACL classifications) by freedom
status (the Freedom House overall ratings).
Table 2: Cross-Classification of Political Regime by Freedom Status
Political Regime
Political Freedom Dictatorship Democracy
Not Free 931 1
Partly Free 711 106
Free 66 769
From Table 2, we observe ”Not Free” is mostly associated with ”Dictatorship”,
and ”Free” with ”Democracy”. For countries with a ”Partly Free”
ranking, more are ”Dictatorship” than ”Democracy”. This result probably
reflects the cautious stance taken by the PACL measure that tries to
avoid type-II errors. Overall, these two measures are quite similar. Among
the observations in Table 2, concordant pairs number at 1,361,384, and the
number of discordant pairs of observations is 7,773. Thus, 99.43% are concordant
and only 0.57% are discordant. The sample gamma is 98.86%. This
confirms that a low degree of political freedom occurs with non-democratic
regimes and high degree of political freedom with democratic regimes. The
Freedom House rankings are highly correlated with PACL political regime
measures.9
We can further explore the relationship between the Freedom House and
PACL measures by testing the null hypothesis of independence between
the two categorical variables. The Pearson chi-squared statistic is 1896.62,
8South Africa is excluded. Freedom House rated separately for ”White” and ”Black”
in South Africa during the sample period.
9Other measures of ordinal association further confirm the conclusion. For example,
Kendall’s tau-b is 0.742. If we ignore the ordinality of political freedom and political
regime data, a high Pearson correlation (0.793) indicates a very high degree of association
between these two measurements.
15
and the likelihood-ratio chi-squared statistic is 2201.61 (based on degree of
freedom =2). Both statistics give P-values less than 0.0001. There is very
strong evidence of association between these two measures even if we ignore
the category orderings of the variables.
The apparent ”anomalies” are cases that are reflected in up-left and
bottom-right corner cells. The only one observation that is classified as
”Not Free” and ”Democracy” is Guatemala in 1981. The Freedom House
1981 volume’s description of Guatemala (p. 352) reports: ”Most opposition
parties are now heavily repressed . . . . . . . Military and other security
forces maintain decisive extra-constitutional power at all levels: those politicians
who oppose them generally retire, go into exile, or are killed.” Then
the 1982 edition begins its report with the sentence (p.296): ”Until a 1982
coup Guatemala was formally a constitutional democracy on the American
model.” The PACL measure seems to have classified Guatemala as a democracy
in 1981 while Freedom House observers clearly judged the government
to be categorized by repression which was confirmed by the 1982 coup.
There are sixty-six cases that are classified as ”Free” and ”Dictatorship”
as shown in Table 3. Among these cases, two third (66.7%) are classified according
to ”alternation” rules. As pointed out above, the ”alternation” rules
risk type-I error. So those country-years that are classified as dictatorship
could actually be democracies. Using more information and discretion, Freedom
House gives these observations a ”Free” ranking. Botswana is a typical
example. Political stability characterizes Botswana’s political landscape.
Botswana’s Democratic Party has been ruling the country until the present.
PACL ”alternation” rules require Botswana during the period from 1972 to
1990 to be classified as a dictatorship. However, Freedom House rates it
as ”Partly Free” in 1972 and ”Free” in 19 years after that, based on their
information sources and survey methodology.
It is interesting to note that all those forty-four cases, in which ”alternation”
rules are applied, are classified as bureaucracies in PACL’s more
16
Table 3: Cases That Are Classified as Dictatorship and Free
Country Period Regime Alternation Rule or Not
Botswana 1973-1990 Bureaucracy Yes
Burkina Faso 1978, 1979 Bureaucracy Yes
Djibouti 1977, 1978 Bureaucracy No
El Salvador 1972-1975 Bureaucracy Yes
Fiji 1972-1986 Bureaucracy No
Gambia 1972-1980, 1989, 1990 Bureaucracy Yes
Ghana 1981 Autocracy No
Guyana 1972 Bureaucracy Yes
Malaysia 1972, 1973 Bureaucracy No
Nigeria 1983 Autocracy No
Seychelles 1976 Bureaucracy No
Sri Lanka 1977-1982 Bureaucracy Yes
West Samoa 1989, 1990 Bureaucracy Yes
detailed regime classification. Actually, the great majority (64 out of 66) of
cases rank ”Free” in Freedom House surveys are classified as bureaucracies
by PACL. Bureaucracies in the PACL measure are those dictatorships with
legislatures, and certainly more likely to be ranked at ”Free” by Freedom
House than those autocracies. Among the sixty-six observations, there are
only two cases in which Freedom House ranks them at ”Free” and PACL
classifies them as ”autocracies.” In the case of Nigeria, the ”anomaly” is
due to the timing rules used by PACL. According to Freedom House, Nigeria
changed from a multiparty democracy which began in 1979 and began
to change after 1982 in a series of coups, rather than a single event, which
by 1984 had placed the government under the control of a military command.
The judgmental nature of the Freedom House rules caused a slower
reclassification.
There was a military intervention in Ghana in 1979 that led to political
executions. However, the 1981 Freedom House review gives the ”free” rating
and begins with the sentence (p. 350): ”Since Fall 1979 Ghana has been
ruled by a parliament and president representing competitive parties.” The
17
1982 report changed the rating to ”not free” and noted that the country was
being ruled by a military faction. It also noted that there had been some
political detentions and police brutality before the 1981 coup, but ”. . . such
denials of rights have subsequently increased.”(p. 295) In this instance the
observers from Freedom House seem to have recognized the institution of
democracy and classified the Country as ”free” in 1981 but in 1982 they not
only changed their rating to ”not free” but also implicitly corrected their
observation of the previous year.
3 Empirical Analyses on Economic Freedom and
Political Freedom
With comprehensive data on economic and political freedoms becoming
available, there is a surge of empirical studies on the two freedoms and
the relationships between freedom and other economic and social variables.
There already existed a vast literature on the influence of political freedom
on economic growth before measurements of economic freedom were
published. The findings of these empirical studies are conflicting (Weede,
1983; Pourgerami, 1988; Scully, 1988; Glahe and Vorhies, 1989; Weed, 1993;
Przeworski and Limongi, 1993; Paster and Sung, 1995; Haan and Siermann,
1996). The empirical results range from positive to negative influences of political
freedom on economic growth. The contradictions of the results could
be attributed to contrasting model specifications and empirical measurements
of political freedom. Some authors argued that the freedom which
really matters in economic growth is economic freedom (Scully, 1992; De
Vanssay and Spindler, 1994; Brunetti and Wedder, 1995; Knack and Keefer,
1995; Barro, 1996).10 This line of investigation was energized by the pub-
10These studies were published before the Heritage Foundation Index and the Fraser
Index were compiled. Scully, De Vanssay and Spindler used a measure of economic liberty
developed by Scully and Slottje (1991), and others used proxies. Barro used government
consumption, the black-market premium and the rule-of-law index (measures of market
distortions). Knack and Keefer used country risk evaluators, including evaluations of
18
lication of the Heritage Foundation Index and the Fraser Index. The Cato
Journal published a special issue on ”Economic Freedom and the Wealth of
Nations” in 1998. Positive effects of economic freedom on growth are also
reported in a variety of empirical studies (Easton and Walker, 1997; Ayal
and Georgios, 1998; Dawson, 1998; Haan and Siermann, 1998; Gwartney,
Lawson, and Holcombe, 1999; Wu and Davis, 1999a; Wu and Davis, 1999b).
Economic freedom is also used to explain other aspects of economic development
including income equality and human well-being (Berggren, 1999;
Esposto and Zaleski, 1999).
Economic freedom as an independent variable in explaining economic
growth and development is robust in numerous studies. We conclude that
the arguments heard down through the centuries - a reliance upon the market
place and unrestrained competition in the allocation of a society’s resources
is the best policy to promote economic growth - has largely been established
by the experiences of the countries of the world as these have been analyzed
by numerous researchers working with these new measurements. It is also
the same case for economic development as a significant explanatory variable
for democracy. In Lipset’s classic statement, ”Perhaps the most wide
widespread generalized linking political system to other aspects of society
is related to the state of economic development” (Lipset, 1959). Economic
development, as an independent variable, has survived in a variety of rigorous
empirical tests on determinants of democratic development (Diamond,
1992; Przeworski, Alvarez, Cheibub, and Limongi, 2000).
The linkages depicted in these empirical studies relate economic freedom
to economic growth, and economic development to political freedom. As
noted above, these two links seem to be well established. Moreover, these
linkages suggest a less well established empirical the influence of economic
freedom on political freedom. Such a link between economic freedom and
political freedom would certainly confirm certain theoretical insights in the
contract enforceability and risk of expropriation.
19
literature. We judge there to be a need for further studies.
Empirical analyses on the possible reverse relationship between political
freedom and economic freedom are largely lacking. We will not speculate
upon the reasons for the relative scarcity of studies. We simply note
that it is important to learn whether the careful study of our measured
history of freedoms will confirm the theoretically based arguments and observations
that democracy ”inherently” acts to constrain economic freedom,
and whether the precise nature of such constraints can be illuminated. Further,
these additional analyses are important to verify theoretical arguments
about possible conflicting effects of civil liberties and political rights upon
each other as well as on economic freedom. Is there a tendency for a free
electoral system to work so as to limit civil rights if there is not a developed
constitution and independent judiciary to prevent such an action? Do
guarantees of civil rights mean that through time an electoral system will
necessarily be established? Which, if not both, might be a factor in limiting
economic freedom? Finally, will further empirical analyses be able to establish
that there exists an endogenous relationship between economic freedom
and political freedom?
There are many questions waiting to be answered and empirically established.
It may be that the possible endogeneity between economic and
political freedom is one of the most intriguing and perhaps the most important.
If a demonstration of endogoniety can include a specification of the
mechanism, if the establishment of more economic freedom really tends to
lead to the development of democratic forms of government, then there are
urgent reasons to hope for and expect additional studies.
20
References
[1] E. B. Ayal and K Georgios. Components of economic freedom and
growth: An empirical study. Journal of Developing Areas, 32:327–38,
1998.
[2] R.J Barro. Democracy and growth. Journal of Economic Growth, 1:1–
27, 1996.
[3] N. Berggren. Economic freedom and equality: Friends or foes? Public
Choice, 100(3/4):203–23, 1999.
[4] K.A. Bollen. Issues in the comparative measurement of political democracy.
American Sociological Review, 45:370–90, 1980.
[5] K.A. Bollen. Liberal democracy: Validity and method factors in crossnational
measures. American Journal of Political Science, 37:1207–30,
1993.
[6] A. Brunetti and B. Wedder. Political sources of growth: A critical note
on measurement. Public Choice, 82:125–134, 1995.
[7] J.M. Buchanan, R.D. Tollison, and G. Tullock. Toward a theory of the
rent-seeking society. Texas A & M University Press, College Station,
1980.
[8] J.M. Buchanan and G. Tullock. The calculus of consent: Logical foundations
of a constitutional democracy. University of Michigan Press,
Ann Arbor, 1962.
[9] J. Cummings. Economic freedom indices: Their use as tools for monitoring
and evaluation. SCS Working Paper, 00/01, 2000.
[10] R.A. Dahl. Polyarch: Participation and opposition. Yale University
Press, New Haven, 1971.
21
[11] J.W. Dawson. Institutions, investment, and growth: New cross-country
and panel data evidence. Economic Inquiry, 36(October):603–19, 1998.
[12] L. Diamond. Economic development and democracy reconsidered.
American Behavioral Scientist, pages 450–99, 1992.
[13] S.T. Easton and M.A. Walker. Income, growth, and economic freedom.
American Economic Review, 87(May):328–32, 1997.
[14] A. Esposto and P Zaleski. Economic freedom and the quality of life.
Constitutional Political Economy, 10:185–197, 1999.
[15] M. Friedman. Capitalism and freedom. University of Chicago Press,
Chicago, 1962.
[16] M. Friedman. Economic freedom, human freedom, political freedom.
Speech delivered at The Smith Center for Private Enterprise Studies,
California State University, Hayward, 1991.
[17] R.D. Gastil and followers. Freedom in the world. Freedom House, New
York, 1972-2002.
[18] F. Glahe and F. Vorchies. Religions, liberty and economic development:
An empirical investigation. Public Choice, 62:201–15, 1989.
[19] J. Gwartney and R. Lawson. Economic freedom of the world, 2001
Annual Report. Fraser Institute, Vancouver, 2001.
[20] J. Gwartney, R. Lawson, and R. Holcombe. Economic freedom and
the environment for economic growth. Journal of Institutional and
Theoretical Economics, 155(4):1–21, 1999.
[21] J.De. Haan and C.L.J. Siermann. New evidence on the relationship
between democracy and economic growth. Public Choice, 86:175–198,
1996.
22
[22] J.De. Haan and C.L.J. Siermann. Further evidence on the relationship
between economic freedom and economic growth. Public Choice,
95:363–80, 1998.
[23] F.A. Hayek. Economic freedom of the world, 2001 Annual Report. University
of Chicago Press, Chicago, 1944.
[24] B.T. Johnson and T.P. Sheehy. The index of economic freedom. The
Heritage Foundation, Washington, DC, 1995.
[25] S. Knack and P. Keefer. Institutions and economic performances:
Cross-country tests using alternative institutional measures. Economics
and Politics, 7 (November):207–227, 1995.
[26] S. M. Lipset. Some social requisites of democracy: Economic development
and political legitimacy. American Political Science Review,
53:69–105, 1959.
[27] S. Mainwaring, D. Brinks, and A. Perez-Linan. Classifying political
regimes in latin america, 1945-1999. Studies in Comparative International
Development, 36(Spring):37–65, 2001.
[28] G. P. O’Driscoll Jr., K.R. Holmes, and M.A. O’Grady. The 2002 index
of economic freedom. Heritage Foundation and TheWall Street Journal,
Washington DC and New York, 2002.
[29] J. Pastor and J.H. Sung. Private investment and democracy in the
developing world. Journal of Economic Issues, 29:223–43, 1995.
[30] A. Pourgerami. The political economy of development: A cross-national
causality test of development. Public Choice, 58(2):123–41, 1988.
[31] A. Przeworski. The games of transition. In Issues in democatic consolidation:
The new south American democracies in comparative perspective.
University of Notre Dame Press, Notre Dame, 1992.
23
[32] A. Przeworski, M.E. Alvarez, J.A. Cheibub, and F. Limongi. Democracy
and development: Political institutions and well-being in the world,
1950-1990. Cambridge University Press, New York, 2000.
[33] A. Przeworski and F. Limongi. Political regimes and economic development.
Journal of Economic Perspectives, 7:51–69, 1993.
[34] C.K. Rowley, R.D. Tollison, and G. Tullock. The political economy of
rent seeking. Kluwer Academic Publishers, Amsterdam, 1989.
[35] G.W. Scully. The institutional framework and eocnomic development.
Journal of Political Economy, 96:652–62, 1988.
[36] G.W. Scully. Constitutional environments and economic growth. Princeton
University Press, Princeton, 1992.
[37] G.W. Scully and D.J. Slottje. Ranking economic liberty across countries.
Public Choice, 69:121–152, 1991.
[38] A. Smith. An Inquiry into the Nature and Causes of the Wealth of
Nations. Random House, New York, morden library edition, 1937.
Reprinted.
[39] Seldon A. Tullock, G. and G.L. Brady. Government: Whose obedient
servant? The Institute of Economic Affairs, London, 2000.
[40] T. Vanhanen. The process of democratization: A comparative study of
147 states, 1980-88. Taylor and Francis, New York, 1990.
[41] X.De. Vanssay and Z.A. Spindler. Freedom and growth: Do constitutions
matter? Public Choice, 78:359–72, 1994.
[42] E. Weede. The impact of democracy on economic growth: Some evidence
from cross-national analysis. Kyklos, 36:21–39, 1983.
24
[43] Wenbo Wu and Otto A. Davis. The two freedoms in a growth model.
Journal of Private Enterprise, 14(2):115–43, 1999a.
[44] Wenbo Wu and Otto A. Davis. The two freedoms, economic growth
and development: An empirical study. Public Choice, 100(1/2):39–64,
July 1999b.
25