Working Paper
Colonizer Identity, Settlement Conditions, and Early Mercantilism:
Do They Matter for Modern Development?
Short Description
This paper examines the causal effect of economic institutions and policies on development among
former European colonies, utilizing an instrumental variable approach that accounts for heterogeneous
formation of institutions and policies among the various colonizers.
Abstract
This paper combines two institutional theories of comparative post-colonial development to examine
the importance of legal-colonial origins and settlement conditions on the development of economic and
political institutions that impact current per capita income. The instruments for institutional
development utilized by Acemoglu et.al. (2001, 2002, 2005), namely settler mortality rates and
population density in 1500, are adjusted to account for the identity of the colonizer (Klerman et.al.,
2011). This procedure allows one to account simultaneously for both the impact of settlement
conditions and differences among European colonizers as sources of variation in institutional
development and their potential causal effect on modern per capita incomes. The British in particular
exhibited less mercantilist economic policies than other European nations prior to colonial expansion,
suggesting that inclusive British settlements would have been more likely than other European
settlements to develop institutions and follow policies more consistent with the protection of property
rights and market allocation. To account for heterogeneous post-colonial institutional development, we
utilize a more comprehensive measure of economic institutions and policies than has been previously
used by researchers in this area, the Fraser Institute’s Economic Freedom of the World Index. The
findings indicate that: (1) the colonizer-adjusted instruments are a better predictor of contemporary
institutions and income per capita than the unadjusted instruments, (2) British influence exerted a more
favorable impact on institutional development than other European colonizers, ceteris paribus, and (3)
the Fraser measure of economic institutions exerts a stronger impact on income per capita than
constraints on the executive, the preferred institutional measure of Acemoglu and his co-authors.
JEL Codes: I25, O11, O43, P14, P16, P51
Keywords: Growth, Development, Economic Institutions, Political Institutions, Economic Freedom,
Mercantilism, Human Capital
1
I. Introduction
The transition from the Malthusian income per capita stagnation to an era of sustained growth, marked
by the onset of the Industrial Revolution, induced a remarkable tenfold increase of the income per
capita (Quamrul Ashraf and Oded Galor 2011). This remarkable growth has not benefited all nations
equally, as data from the World Bank (2012) for 181 countries indicates that there is 60-fold difference
in the level of income per capita between the richest and poorest twenty-five percent of countries.1
Despite substantial progress in our understanding of the causes behind the unparalleled contemporary
growth and the inequality in living standards between nations, an overall consensus on the still proves
elusive. This is evidenced by the emergence of three major theories of development in the economics
literature.
First is the neoclassical growth theory, and its extensions, which stress the accumulation of physical and
human capital and technological change as the ingredients for economic growth (e.g. Evsey Domar,
1946; Oded Galor, 2005; Roy Harrod, 1939;; Robert Lucas, 1988; Paul Romer, 1986, 1990; Robert Solow,
1956). Next is the geographic endowment theory of development that is most closely associated with
contributions by Jared Diamond (1997), John Gallup, Jeffery Sachs and Andrew Mellinger (1998), Sachs
(2001, 2003) and Sachs and Andrew Warner (1997). This theory suggests that factors such as a
temperate climate, resource endowments, access to coastal waterways and proximity to major markets
are necessary to ignite development. Tropical climates are said to retard development by lowering
worker productivity levels as well as providing a fertile environment for life-threatening diseases such as
malaria. Third is the institutional theory of development which contends that institutional arrangements
determine the incentive structure of an economy and are thus directly responsible for economic
performance. Notable contributions include Daron Acemoglu, Simon Johnson and James Robinson,
hereafter AJR, (2001, 2002), Robert Hall and Charles Jones (1999), Douglas North (1981, 1990, 1991) and
Mancur Olson (1996, 2000).
The three major theories are not necessarily contradictory, but are more likely to be complimentary in
gaining a more comprehensive theory of economic development. William Easterly and Ross Levine
(2001) for instance document that factor accumulation is important for growth but total factor
productivity (TFP) explains most of the observed variation in cross-country income levels. Augmented
growth models have typically assumed that institutions influence development through the TFP channel
(e.g. Stephen Knack and Philip Keefer, 1995; Jakob de Haan, Susanna Lundström and Jan-Egbert Sturm,
2006), but these studies tend to underestimate the impact of institutions on growth because institutions
also exert an influence on the level and productivity of investment in an economy (e.g. John Dawson,
1998; James Gwartney, Randall Holcombe and Robert Lawson, 2006; Joshua Hall, Russell Sobel and
George Crowley, 2010; Paulo Mauro, 1995). Furthermore, geographic endowments may influence the
development of institutions as suggested by Kenneth Sokoloff and Stanley Engerman (2000) and Easterly
and Levine (2003).
Our work constitutes a contribution to the institutional theory of comparative development and is most
closely related to two related lines of research. The first emerges from the seminal contributions of AJR
(2001, 2002), who argue that settlement conditions determined European settlement strategies in the
colonies. The hypothesis advanced by AJR is that Europeans were likely to settle in large numbers and
invest in the replication of inclusive European institutions to protect private property and constrain the
1
Income per capita, in constant 2000 $USD, among the 45 richest and poorest nations is $28,946 and $449,
respectively.
2
powers of government in colonies with low mortality rates and sparse indigenous populations. In
colonies with high mortality rates and dense indigenous populations, on the other hand, European
colonizers would have sought to establish an extractive state to transfer resources from the colony back
home. Because institutions are persistent, early institutional differences set the colonies on divergent
development paths and largely explain large disparities in income per capita levels among the former
colonies. The second line of research related to our paper follows from the seminal contributions of
Rafael La Porta et.al. (1997, 1998, 1999), who argue that a country’s legal traditions were largely
imparted through the colonization process and that differences in legal origins in large part explain
differences in contemporary laws and regulations that influence economic outcomes. In particular,
countries with English common law origins tend to have better economic performance relative to those
with French civil law origins (La Porta et.al., 2008). Klerman et.al. (2011) illustrate the imperfect
correlation between legal origin and colonial history in suggesting that the identity of the colonizer is a
more important determinant of modern development than legal origins because the former captures
more of the diversity in colonial polices that matter for development.
Thus two institutional theories of comparative post-colonial development exist in the current literature.
Both the settlement strategy and the legal-colonial origins theories have helped shape our
understanding of how institutions have exerted an effect on comparative development among former
colonies, but we contend that the two theories in isolation are incomplete and attempt to bridge the
two into a more cohesive theory that better reflects historical evidence that the British began to
liberalize their economic and political institutions prior to the colonization period, whereas nocuous
mercantilist institutions and policies remained intact in continental Europe. The current work offers
two major innovations to the institutional theory of comparative development literature. First, we
integrate the two lines of research in developing an identification strategy that simultaneously accounts
for the impact of settlement conditions and heterogeneous colonial policies among the European
colonizers as a means to better capture variation in the development of early institutions. We do so by
interacting colonial settlement conditions, as measured by population density in 1500, with the identity
of the colonizer to provide an instrument that is a stronger determinant of contemporary institutions
than has previously been used, allowing us to more accurately estimate the causal impact that
institutions exert on modern per-capita income levels.
The underlying theory motivating our identification strategy relies on evidence that the British exhibited
less mercantilist policies than their counterparts in continental Europe prior to colonial expansion.
Existing mercantilist policies would have been transplanted by the colonizer to its colonies, suggesting
that inclusive British settlements would have been more likely than other European settlements to
develop institutions and follow policies more consistent with the protection of property rights and
market allocation. The second innovation is the use of an institutional measure, the Fraser Institute’s
Economic Freedom of the World Index, hereafter EFW, which has not been previously used in this line of
inquiry. The EFW index is constructed to provide a comprehensive measure of the degree to which a
nation’s economic institutions and policies reflect the protection of private property and market
allocation, the antithesis of mercantilism.
Next, we demonstrate that EFW is a stronger and more robust institutional determinant of modern
development than constraints on the executive, the professed preferred institutional measure of
Acemoglu and his co-authors (AJR, 2001; AJ, 2005). Section three offers a more comprehensive review
of the existing literature in laying out the theory underlying our identification strategy. Empirical results
of the causal impact of institutions on income per capita are presented in section four, followed by a
serious of robustness checks in the penultimate section. Section six concludes.
3
II. Institutional Determinants of Development
Empirical analyses of the impact that institutions exert on comparative post-colonial development have
primarily made use of two measures of institutions, Polity IV’s constraints on the executive and the
International Country Risk Guide’s risk of expropriation (AJR, 2001, 2002; AJ, 2005). Both are singledimensional institutional variables, with the former a measure of checks and balances placed on the
authority of the chief government executive and the latter a measure of de facto protection of private
property. Edward Glaeser et.al., henceforth GLLS (2004), provide two criticism regarding the use of
these two measures to establish evidence of the primacy of institutions in promoting growth.
First, GLLS (2004, 272) challenge the “close to an intellectual consensus that the political institutions of
limited government cause economic growth” in suggesting that authoritarian political regimes, hardly
limited governments, sometimes pursue a development strategy through liberalization of the economy,
as has been the case recently in countries such as Chile, China, Taiwan and Singapore. Robert Barro
(1996, 2) clarifies this apparent dichotomy in suggesting that “dictators come in two types, one whose
personal objectives often conflict with growth promotion and another whose interests dictate a
preoccupation with economic development.” While there is a large body of literature (e.g. de Haan and
Sturm, 2003; Lundström, 2005; Hans Pitlik and Steffen Wirth, 2003; Pitlik, 2008; Martin Rode and
Gwartney, 2012) suggesting that political democracy lays the foundation for economic liberalization,
studies by Dawson (2003), Manuel Vega-Gordillo and José Álvarez-Arce (2003), and José Aixalá and
Gema Fabro (2009) demonstrate that political freedom granger causes economic freedom and that
economic freedom granger causes growth. Similarly, Richard Jong-A-Pin and De Haan (2008, 2011)
provide evidence that economic liberalization precedes growth acceleration. Related is the HayekFriedman hypothesis, which suggests that societies with high levels of political freedom must also have
high levels of economic freedom, but the relationship need not work the other way around. Lawson and
J.R. Clark (2010) present empirical evidence supportive of this hypothesis.
Second, GLLS argue that the risk of expropriation and constraints on the executive variables reflect
policy choices or outcomes rather than measures of permanent and durable institutions, suggesting they
are outside the periphery of North’s (1981, 1991) widely acknowledge definition of institutions. There is
some merit to this criticism in that there is a weak correlation between de jure and de facto property
rights protections (Feld and Voigt, 2003), a point that GLLS (2004, 276) themselves recognize by
indicating that the “rules on the books are very different from what actually takes place in a nation.” For
instance, government officials in some countries such as the Czech Republic refrain from expropriating
private property without a constitutional provision requiring them to do so, while the constitutions in
other countries such as Russia and Moldova offer property rights protections that are nonetheless often
circumvented in practice (Bjørnskov, 2012). Regarding the difference in property rights protections
between Russia and China, Rodrik, Subrmanian and Trebbi , henceforth RST (2004, 157) indicate that
“Credibly signaling that property rights will be protected is apparently more important than enacting
them into law as a formal private property rights regime.” North’s definition of institutions also includes
informal customs and norms such that if a government is credibly committed to preserving private
property rights without a constitutional or legislative mandate, then such policy would fall under the
spectrum of institutions. The same can be said of other persistent economic policies, a point that GLLS
appear to agree given their recommendation to use economic policies and rules that can be
4
manipulated by a policy maker, regardless of whether an authoritarian or democratic regime is in place,
to estimate the effect of policies and institutions on economic performance. 2
We heed to the advice of GLLS in adopting the Frasier Institute’s Economic Freedom of the World (EFW)
index as our measure of economic institutions and policies. While the protection of private property
rights may very well be the core institutional characteristic of a healthy market economy that facilitates
growth, the comprehensive economic environment exerts an impact on the development path that an
economy will follow such that failure to account for additional institutional and policy characteristics
may overestimate the impact of any single variable such as property rights protections. Our preferred
institutional variable, EFW, provides a much more comprehensive measure of a nation’s economic
institutions and policies that are likely to exert an impact on a country’s development than the singledimensional risk of expropriation variable (cf. Dawson, 2003; de Haan, Lundström and Sturm 2006;
Gwartney, Holcombe and Lawson, 2006).
The EFW index is designed to measure the degree to which a country’s institutions and policies are
consistent with personal choice, voluntary exchange coordinated by markets, freedom to enter and
compete in markets, and protection of persons and their property from aggression by others, and is on a
0-10 scale that is increasing in the level of economic freedom. It is comprised of five main areas: size of
government, legal system and property rights, sound money, freedom to trade internationally and
regulation of credit, labor and business. Each of the five areas of the index is weighted equally and is
derived from a number of underlying components that are predominantly objective, data-driven
measures obtained from external sources such as the International Monetary Fund, World Bank and
World Economic Forum. A total of 42 variables are utilized in the construction of the index (Gwartney,
Lawson and Hall, 2012). The comprehensiveness of EFW captures a “broad cluster of institutions” that
are mutually reinforcing in the development process, a desirable feature for measures of institutions
(AJR, 2001; Acemoglu, 2005).
Some researchers have been hesitant to utilize measures of economic freedom, suggesting that indices
such as that of the Frasier Institute are ideologically-biased and therefore unreliable measures of good
governance or institutions (e.g. Raphael Auer, 2013). This criticism is invalid for several reasons. First,
the EFW index is constructed to measure the extent to which a nation’s economic policies and
institutions are consistent with property rights protections and market allocation, rather than a measure
of the quality of governance. Second, Martin Paldam (2003) describes the potential moral hazard risk
associated with the construction of an economic freedom index by free market advocates, but argues
that the reputation of the scholars involved and objectiveness of the data used make it highly unlikely
that the EFW index be manipulated in a fashion favorable to its creators. Additionally, the methodology
used is highly transparent and all of the data are publicly available such that the index, or alternative
ones, can easily be reproduced. Lastly, as de Haan, Lundström and Sturm (2006) and Nathan Ashby and
Sobel (2008) indicate, the ideological-persuasion of the organizations producing the index ensures that it
measures what it intends. Thus, we conclude that EFW is an appropriate measure of a nation’s economic
institutions and policies for the current study, particularly given that it is constructed in a manner that is
antithetical to mercantilism, as will be discussed in more detail in section III below.
2
At a theoretical level the issue of institutions is complex. Jones and Romer (2010) claim that a major challenge of
the new growth theory is to endogenize in parsimonious models variables accounting for ideas, population, human
capital and institutions. The unified growth theory is an attempt to endogenize human capital, technological
advancement and population (Galor, 2005), but little progress has been made thus far to endogenizing institutions
in theoretical growth models.
5
As described above, most empirical studies in the institutional growth literature have implicitly assumed
that institutions exert a direct influence on development through the TFP channel. They have typically
done so using an augmented neoclassical growth model, but often without formally specifying how
institutions theoretically enter the growth equation. Following the recommendation of Samuel Bazzi and
Michael Clemens (2013) to ground growth regressions within a theoretical model , we adopt a humancapital augmented neoclassical production function with constant returns to scale (Lucas, 1988; Gregory
Mankiw, David Romer and David Weil, 1992; Romer, 1990), given by equation 1,
where
and represent TFP, the capital stock, level of human capital and population for
country i, respectively.
(1)
TFP is heterogenous across economies and is modeled as
. As such, country i’s TFP is a
function of its distance from the ideal institutional environment,
, and a homogenous level of
productivity, , that can be thought of as the production possibilities frontier (PPF). With an ideal
institutional environment, productive entrepreneurship, investments in human and physical capital and
the division of label are incentivized in a manner necessary to foster innovation and economic growth
(William Baumol, 1990; Randall Holcombe 1998). TFP is here structured to serve as a production deflator
for a country whose institutions are less than ideal,
, which can be thought of as operating at a
point inside the production possibilities frontier.
(2)
Substituting the TFP term into equation (1), normalizing it to per-capita terms such that
, and
logging the equation yields equation (2), where
is a constant and reflects the
productivity level of a country with the worst possible institutions. This equation can be used to
estimate the direct effect of institutions on the level of income per capita, and differenced to examine
how institutional change affects economic growth. This equation likely understates the total effect of
institutions on the level of income per capita though, as institutions also exert an indirect effect on
development through their influence on the level and productivity of human and physical capital
(Dawson, 1998; Gwartney, Holcombe and Lawson, 2006; Hall, Sobel and Crowley, 2010). Given this
limitation of the model, we initially estimate the effect of institutions on the level of per capita income
by equation (3), which omits the human and physical capital variables from the regression and includes,
a vector of additional covariates that are believed to exert an influence on the development process, .
(3)
Table 1 reports standardized coefficients from OLS estimates of equation (3) using the three different
measures of institutions discussed above. The income per capita data are from the Penn World Tables
version 7.1, henceforth (PWT). The results indicate that our preferred and comprehensive measure of
economic institutions, EFW, exerts an effect on per-capita income levels that is similar to that of the risk
of expropriation measure, and a much larger impact than constraints on the executive. These results are
robust to the selection of year and measurement of income per-capita, as well as the timing of
institutional measures that have been used in the literature. The constraints on the executive measure
are used in columns (1), (4), (7) and (10). Risk of expropriation and economic freedom are the
institutional measures used in columns (2), (5), (8) and (11), and (3), (6), (9) and (12), respectively.
Columns (1)-(3) provide estimates for all countries for which the respective data are available for each
institutional measure. Columns (4)-(6) provide estimates for the base sample of former colonies for
6
which population density in 1500 data are available. The remaining columns introduce additional
covariates to the base sample regressions.
[INSERT TABLE 1 HERE]
[INSERT TABLE 2 HERE]
As indicated in the introduction, one theory of development suggests that geography and climate exert
a direct impact on economic performance (Niccolò Machiavelli, 1519; Diamond, 1997). Although several
variations of the geography-climate theory exist, the major underlying argument is that tropical climates
are developmentally handicapped relative to temperate climates (David Landis, 1998). Proponents of
this view argue that labor is less productive in tropical climates because (1) hot temperatures in the
tropics reduce the energy levels of workers, and (2) the disease environment is more fertile in the
tropical climates such that the populations suffer from poor health conditions and high mortality rates
(e.g. Sachs 2001, 2003; Sachs and Warner, 1997). Some researchers have also argued that proximity to
major economic markets is an essential ingredient for development, suggesting that countries located in
remote tropical locations are destined for underdevelopment (e.g. Gallup, Sachs and Mellinger, 1999).
Other researchers however have argued that climate only affects development indirectly through its
influence on the development of institutions, with the basic premise being that geographic and climactic
endowments create a natural environment for the establishment of different types of institutional
arrangements (AJR, 2001; Easterly and Levine, 2003; Easterly, 2007; RST, 2004; Sokoloff and Engerman,
2000). We elaborate on this matter further in section III. Beginning with column (7), we follow the
literature by including latitude, as measured by the absolute value of the distance from the equator and
scaled to take value between 0 and 1, to control for the direct effect that climate and/or geography may
exert on development (La Porta et.al., 1999; AJR 2001, Gallup, Sachs and Mellinger, 1999; Hall and
Jones, 1999; RST, 2004).3
Ethno-linguistic fragmentation, an index that measures the probability that two randomly selected
individuals from a country are from different ethno linguistic groups, is included in columns (10)-(12) to
control for political instability resulting from ethnic and linguistic heterogeneity among a nation’s
populace. Samuel Huntington (1968) argues that governments in countries with a more fractionalized
population tend to implement policies that benefit the winning minority at the expense of groups not
represented in government. Further, Mauro (1995) adds that divided countries are prone to greater
political instability and are associated with more corruption and lower growth because bureaucrats
engage in ethnocentric behavior favoring members of their own group and attempt to take as many
bribes as possible given the uncertainty about their tenure in office. Accordingly, this variable has been
identified as an exogenous source of government predisposition to political infighting, to enact
redistributive policies, weakening institutional quality and consequently hindering economic prosperity
(Easterly and Levine 1997, 2009; Hall and Jones, 1999; Mauro, 1995). Moreover, Mark J. Roe and Jordan
Siegel (2011) find that ethno linguistic fractionalization has a significant impact on political instability. A
theoretical model developed by Marina Azzimonti (2011) is consistent with uncovered evidence showing
growth retarding effects of polarized societies. The data are taken from the dataset of La Porta et al.
(1999).
In panel A, PPP-adjusted log GPD per capita in 1995 is regressed separately on the three institutional
measures. AJR (2001, 2002) and AJ (2005) used the average constraints on the executive over the period
3
The latitude data are taken from the dataset of La Porta et.al. (1999), but the Central Intelligence Agency’s CIA
World Factbook (1996) is the original source of the data.
7
1990-1999 and average risk of expropriation over the period 1985-1995. We add the average EFW over
the period 1985-1995.4 The estimates of columns (1)-(3) for the all countries samples indicate that a
one standard deviation increase in constraints on the executive, risk of expropriation and economic
freedom measures is associated with a 0.50, 0.79 and 0.75 standard deviation increase in log GDP per
capita, respectively, suggesting that the latter two measures of economic institutions exert a larger
effect than the former, a measure of political institutions. The respective adjusted R2 values for these
regressions are 0.24, 0.62, and 0.56, indicating that risk of expropriation and economic freedom have
considerably more explanatory power than constraints on the executive.
The estimates for the base sample of former colonies in columns (4)-(6) reiterate these findings,
although the performance gap between constraints on the executive and the two measures of economic
institutions narrows. The standardized coefficient increases from 0.50 to 0.60 for the constraints on the
executive, drops from 0.79 to 0.77 for risk of expropriation and remains unchanged at 0.75 for economic
freedom. A similar pattern emerges for the explanatory power of the respective regression, as the R2
value in column (4) increases from 0.24 to 0.35, while that of column (5) drops slightly from 0.62 to 0.59
and remains unchanged at 0.59 in column (6). The magnitude of the estimated effects of each
institutional measure on income per capita declines slightly when additional covariates are included,
although all remain highly significant. As expected, the coefficient on the latitude and ethno-linguistic
fractionalization measures are positive and negative, respectively. Latitude is only marginally significant
in columns (7) and (8), and significant at the 5 percent level in (9); however, it become insignificant in
columns (10)-(12) when ethno-linguistic fractionalization is included also included.
The dependent variable in panel B is PPP-adjusted log GDP per capita in 2000. For this set of regressions,
we follow GLLS (2004) in using the long-run average constraints on the executive over the period 19602000, and provide a longer-run measure of EFW than was used in panel A by taking the average over the
period 1985-2000.5 The results are similar to those obtained in panel A. The standardized coefficients on
both the constraints on the executive and economic freedom measures are higher in panel B than they
were in panel A for all specifications, perhaps indicative of the long-lasting effects of institutions that are
reflected in the longer-run average used for these two measures. Meanwhile, the coefficients for risk of
expropriation remain little changed. In panel C, PPP-adjusted income per capita in 2010 is the
dependent variable. The constraints on the executive and economic freedom institutional measures
from panel B are updated accordingly to reflect the long-run averages through 2010.6 Using the more
recent data, the results are nearly identical; suggesting that they are relatively robust to the timing of
the institutional measures and the period in which income per capita is measured.
Taken as a whole, the results from Table 1 indicate that measures of economic institutions are a
stronger determinant of development than constraints on the executive, a measure of political
institutions. Approximately 60 percent of the variation in income per capita across countries is
associated with differences in economic institutions. As argued above, the EFW index provides a much
4
EFW data is available in five-year intervals over the period 1970-2000 and annually after 2000. Country coverage
is limited prior to 1985, with much more comprehensive coverage of countries beginning in 1990. As such, the
average over the period 1985-1995 is used here with the stipulation that a country have data for at least two of
the three five-year periods. The chain-linked index is used.
5
The chain-link EFW index scores are used. A country must have data for at least 2 of the 4 five-year periods. The
average risk of expropriation score over the period 1985-1995 is again used because we do not have access to
more recent data.
6
Again, the chain-link EFW scores are used. A country must have data for at least 4 of the 6 five-year periods.
8
more comprehensive assessment of a nation’s economic institutions and policies than the singledimensional risk of expropriation measure, making it a more suitable measure of economic institutions.
Nevertheless, the regressions using either the risk of expropriation or economic freedom measures
always have greater explanatory power and exert a larger effect on log income per capita than those
using constraints on the executive, the preferred institutional measure of Acemoglu and his co-authors
(AJR, 2001; AJ, 2005). These results hold for both the all country and base samples for all three time
periods examined, and remain robust after controlling for latitude and ethno-linguistic fractionalization.
Following the advice of Rati Ram and Secil Ural (2013) to caution in drawing strong conclusions from the
results due to the differences in income per capita measure by the PWT and World Bank’s World
Development Indicators, hereafter (WDI), we demonstrate that the results are robust to the measure of
output in appendix A1, which repeats the regressions from Table 2 using the WDI Development
measures of income per capita in lieu of the PWT ones.7
As an illustration of the estimated impact of economic freedom on the level of per-capita income, the
partial effect of 0.78 reported in column (3) of panel C suggests that the difference in economic freedom
between Costa Rica and Ghana, which are approximately in the 75th and 25th percentile of nations with
economic freedom values of 6.9 and 5.5, respectively, should translate into a 1.07 log-point difference
(or approximately a 2-fold difference . The actual income per capita difference
between the two countries is 1.70 log–points (a 4.5-fold difference), suggesting that differences in
economic freedom, if causal, account for substantial amount of the output gap between the two
nations.8
III. Theoretical Foundations for Our Identification Strategy
The OLS results reported in Table 2 suggest that contemporary institutions are highly correlated with
modern levels of log income per capita. The two measures of economic institutions, risk of expropriation
and economic freedom, were shown to be stronger determinants than constraints on the executive, a
measure of political institutions. The estimates were however derived from equation (3) and the grand
transition view of institutions and development suggests that institutions and per-capita income evolve
simultaneously (Paldam and Erich Gundlach, 2008). This would indicate that institutions are endogenous
]
to the equation, violating the exogeneity assumption, [
, necessary for the estimates to be
consistent. If contemporary institutions are endogenous, then we need an instrumental variable that is
correlated with , but uncorrelated with , in order to consistently estimate the causal impact that
institutions exert on development.
7
Johnson et.al. (2013) identify problems of variability and valuation with earlier versions of the PWT income-per
capita measures. The issue of variability is not pertinent to the current study given that it only uses the level of
income per capita in a given year. As indicated by Alan Heston, Robert Summers and Bettina Aten (2012),
improvements were made for PWT version 7.1 to provide better information about prices and validation, reducing
the problem of valuation. Nearly identical results are obtained when using the WDI measures. [DEVELOP TABLES
TO INCLUDE IN APPENDIX]
8
Using the WDI data in lieu of the PWT, the actual difference in the income per capita levels between Costa Rica
and Ghana is 1.96 log points, or a 6.1-fold difference. The partial effect of 0.76 reported in column (3) in panel C of
appendix A1 predicts there to be a 1.06 log point, or 1.9-fold difference, between the two nations.
9
A. Colonization as an Exogenous Institutional Shock: A Brief Literature Review
Hall and Jones (1999) recognized the potential endogeneity of institutions and used latitude and the
share of the population speaking a European language as instruments for their index of social
infrastructure, a measure of the quality of institutions. 9 The theory underlying their identification
strategy is that a country’s institutions are largely a function of the extent to which it was influenced by
Western Europe. The instrumentation strategy used by Hall and Jones (1999) has been criticized for not
distinguishing between good and pernicious Western European influence, as well as having weak
theoretical foundations for exclusion of the climate and linguistic diversity variables from estimate
based on all available countries (AJR, 2001, 2011; Acemoglu, 2005).
AJR (2001, 2002, 2011) exploit European colonization as a natural experiment in history in developing an
identification strategy that allows European influence to be either beneficial or harmful for institutional
development, depending on the colonization strategy pursued by the European colonizer. The
colonization strategy in turn depended on the feasibility of permanent settlement, as determined by the
settlement conditions in the colony. AJR contend that two types of settlement strategies were pursued.
Colonies in which settlers experienced high mortality rates and/or were densely populated by
indigenous persons provided unfavorable settlement conditions. When settlement conditions were
poor, AJR allege that the Europeans pursued an extractive strategy that involved mass expropriation of
resources from the colony, often through coercion of the native populations, to be shipped home and
enrich the kingdom. On the other hand, when settlement conditions were favorable, as indicated by low
settler mortality rates or sparse indigenous population, the European colonizers were more likely to
settle permanently and invest in inclusive institutions similar to those existing in their native European
nation. Because institutions are durable and history is path dependent, AJR argue that the divergent
development paths experienced by the former colonies is largely a function of the type of institutions
that emerged during the colonization process, with inclusive settlements resulting in the development
of growth-promoting institutions protective of property rights and limiting the power of political leaders,
and extractive settlements resulting in growth-retarding institutions that protect the power and wealth
of the political elite at the peril of the remaining population.
A related line of research argues that a nation’s legal origins are intrinsically linked to the development
of a wide range of institutions. The seminal contributions of La Porta et.al. (1997, 1998) link English
common law origins to the development of institutions protective of financial investors, and hence
greater financial development, than rules emanating in countries with civil law origins inherited from
France and other continental European nations. La Porta et.al. (2008, 286) indicate that subsequent
research suggests that “civil law is associated with a heavier hand of government ownership and
regulation than common law,” more formalism of judicial procedures and less judicial independence,
which are in turn are linked to less secure property rights and weaker enforcement of contracts.
9
The social infrastructure index of Hall and Jones (1999) consists of two sub-indices: (1) an index of antidiversionary government policies such as law and order, bureaucratic quality, corruption, risk of expropriation and
government repudiation of contracts; and (2) the Sachs-Warner index of trade openness. The Sach-Warner
measure is the average of the number of years that a country is open over the period 1950-1994. For a given year,
a country is classified as open, and assigned a value of one, if all of the following conditions are met: (a) nontariff
barriers cover less than 40% of trade; (b) average tariff rates are less than 40%; (c) black market premiums during
the 1970s and 1980s were less than 20%; (d) nation not classified as socialist by Kornai (1992); and (e) the state
does not monopolize major exports (Sachs and Warner, 1995). The EFW index captures many of the same
elements as the social infrastructure index, but has a greater scope and is a better approximation of the broad
cluster of mutually-reinforcing economic, political, social and legal institutions advocated by Acemoglu (2005).
10
Interpreting the preponderance of evidence, La Porta et.al. (2008, 286) contend that legal origin
represents a “style of social control of economic life,” and argue that “common law stands for the
strategy of social control that seeks to support private market outcomes, whereas civil law seeks to
replace such outcomes with state-desired allocations.”10
Legal origin has been used as an instrument for institutions in the comparative post-colonial
development literature given that legal systems were transplanted throughout the world through the
colonization efforts of a small number of Western European nations. Thus, it is a valid exogenous
instrument for institutions so long as only former colonies that inherited a legal system from their
colonizer are included in the sample (Klerman et.al, 2011). Klerman et.al. (2011, 380) contend however
that the identity of the colonizer is a better instrument than legal origins despite the high correlation
between the two because the colonial powers transplanted not only legal systems, but also differences
in policies related to “education, public health, infrastructure, European immigration, and local
governance.” 11 Studies by Graziella Bertocchi and Fabio Canova (2002), Robin Grier (1999) and Jacek
Rostowski and Bogdan Stacescu (2006) find that the former British colonies exhibited higher growth
rates than French colonies. Klerman et.al. (2011, 405) provides additional evidence that British colonies
experienced greater growth than French and other continental European colonies, but also finds
evidence that the identity of the colonizer is a “better predictor of post-colonial growth rates than legal
origin.”
Thus two institutional theories of comparative post-colonial development – settlement strategy and
legal-colonial origins – exist the current literature. While sympathetic to both views, we believe that in
isolation each is incomplete and attempt to bridge the two into a more comprehensive theory that
better reflects historical evidence.12 We argue that heterogeneous mercantilism among the European
colonizers was transplanted to the colonies when large-scale settlements occurred, influenced by the
settlement conditions of the colony, such that an inclusive strategy pursued by the British vis-à-vis any
of the continental European colonizers would have resulted in the development of less mercantilist
institutions and policies in the former, and hence an environment more favorable for growth and
development. We will elaborate on the details of our theory below, but first present evidence to justify
our position that England was less mercantilist than the continental European colonial powers.
B. European Mercantilism Manifested in the Colonies
We define mercantilism as an economic system in which property rightfully acquired by citizens is
subjected to depredation and erosion by the economic and political elites through brutal force, coercion
or the law. Our notion of private property is comprehensive enough to include not only land, but also
other forms of property including commercial enterprise, financial assets, and individual labor. Given
10
Paul Mahoney (2001, 505) makes a similar distinction in noting that legal origins reflect “different views about
the relative role of the private sector and the state.”
11
Klerman et.al. (2011) indicate that French civil law was transplanted by not only the French, but also the
Belgians, Dutch, Portuguese and Spanish, noting institutional and developmental differences between former
colonies of the respective colonizers. They also point out that several British colonies have mixed legal systems
today.
12
Auer (2013) argues that by failing to account for the influence that geographic endowments may have exerted
on colonization strategy, AJR (2001) overestimate the influence of institutions on economic growth, while La Porta
et.al. (1999) underestimate the importance of colonial-transplanted legal origins on the development of
institutions.
11
our definition, examples of mercantilist policies are monopolies extended to colonial companies to
exploit a given territory, trade barriers, slavery, forced labor, medieval guilds, punitive taxation, complex
regulations, and intentional impediments fostered by the elite to make health and education widely
inaccessible which includes monopolies in the market of ideas that do not promote diversity of
economic views. Inflation and currency devaluations are also inherently mercantilist to the extent that
vast segments of the population are robbed of the fruits of their labor to the benefit of the political and
economic elites. Similarly, widespread ownership of government enterprises, although socialist by
modern accounts, is also conceivably mercantilist to the extent that citizens who are potential
entrepreneurs and/or shareholders are precluded from becoming owners of these companies with a
legal property title.
We contend that there is a spectrum of mercantilist institutions such that it is unlikely that two
economies exhibit identical forms of mercantilism. At one extreme is the extractive state described by
AJR (2001) that is governed by an unconstrained class of economic and political elites that enjoy
protection of their property and reap all of the benefits from economic activity. The means of
production and enterprises are monopolized by the state or a few elite citizens who pay homage to the
state for the protection of their monopoly. Normal citizens are often forced into slave labor and
deprived of basic health and opportunities to become educated, start an enterprise, acquire property or
participate in the political process. One might construe of such states as oppressive totalitarian regimes.
At the other extreme is a free market capitalistic system that offers equal protection of persons and
their property before the law. All citizens enjoy the right to voluntarily pursue an education and health,
start an enterprise, enter into voluntary contracts, sell their labor, reap the entire rewards of their
economic activities, and participate in the political process, so long as their efforts do not impinge on
the rights or inflict harm on their fellow citizens. Under such a laissez faire system, prices are freely
coordinated by supply and demand factors, and the allocation of resources is dictated by market forces.
Of course there is a continuum of economic systems that lay somewhere in between these two
extremes, and indeed, almost all economies in practice fall within the endpoints, engaging in various
forms of mercantilist policies that are at odds with the aforementioned definition of economic freedom.
Our view is thus consistent with that of La Porta et.al. (2008), who suggest that a variety of capitalist
institutions exist in practice. Indeed, even the English practiced a form of mercantilism spanning the
period 1500-1750, although as William Grampp (1952, 495) argues, “instead of the classical policy of
laissez faire, the [English] mercantilists proposed a policy which could utilize the market wherever
possible, supplement and control it where not, and which would have full employment as its proximate
objective.” Curtil Nettels (1952, 106) adds: “In England, the system invoked the initiative and enterprise
of private citizens. It encouraged the merchants, shippers, and manufacturers by conferring benefits
upon them and by identifying their private interests with the highest needs of the state. So close was
this identification that one may properly regard the theory of mercantilism as a rationalization of the
special interests of dominant groups of the time. The mercantilist policy was an expression of an accord
between landowners and merchant-capitalists in alliance with the Crown.”
[THIS SUBSECTION IS STILL IN PROGRESS…COMMENTS/ RECOMMENDATIONS WELCOME]
C. Settlement Conditions and Colonizer Identity as Determinants of Early Institutions
Although the selection of target colonies by the various European powers for settlement may not have
been random and instead based on factors such as climate, geography or the presence of native
12
populations (Auer, 2013; Klerman et.al, 2011), European colonization nonetheless provides a natural
experiment for the exogenous imposition of institutional arrangements in the colonies by the
Europeans. As such, contemporary institutions are modeled as a linear function of initial institutions,
and other factors, , that may have exerted an influence on the development of institutions such as
geography, and is described by equation (4). Because institutions are durable and path dependent
(North, 1991), the institutional shock corresponding to European colonization would have altered the
trajectory of institutions and hence economic development in the colonies, depending on the type of
institutions implemented.
,
(4)
The establishment of highly mercantilist, extractive institutions would have led to a poor institutional
environment and sluggish if not counterproductive growth prospects going forward, while the
transplantation of a more modest form of mercantilism such followed by the French or Spanish, would
have likely laid the foundation for a similar form of mercantilism in the future and modest, albeit
inefficient, growth opportunities. Meanwhile, the imposition of more inclusive, milder forms of
mercantilism, such as that practiced by the British, would likely have fostered an environment
incentivizing market competition, entrepreneurship and innovation, and hence one conducive to
persistent economic growth (e.g. Baumol, 1990; Holcombe, 1998; Israel Kirzner, 1973, 1985). Thus, using
initial colonial institutions as an instrument for contemporary institutions would provide a theoretically
sound exogenous source of variation to enable consistent estimation of the causal impact that
institutions exert on the level of development.
Unfortunately, adequate early institutional measures for the colonies are unavailable.13 Following
Acemoglu et.al. (2001, 2002, 2005), we postulate that settlement conditions faced by colonizer i in
region j,
, influenced the settlement strategy and the development of early institutions by the
European colonizers. They use two variables to proxy for settlement conditions: log of population
density in 1500 and log of settler mortality rates. The line of reasoning suggests that Europeans would
have been deterred from permanent colonial settlement and the establishment of institutions similar to
those in their home country in regions with high indigenous population densities and where settlers
experienced high mortality rates. Instead, they would have sought to extract the local resources through
coercion of the indigenous populations. In regions with low population densities and settler mortality
rates on the other hand, the European colonizers would have been much more likely to establish
permanent settlements and expend effort to establish institutions resembling those from home such as
widespread property rights protections and restraints on the power of political leaders. Indeed, risk of
expropriation and constraints on the executive are the two measures that have been used by Acemoglu
and his co-authors, although they profess the latter to be their preferred institutional measure (AJR,
2001; AJ, 2005).
Our theory is consistent with that advanced by Acemoglu and his co-authors in that European colonizers
followed a similar settlement strategy such that they would have chosen to establish highly mercantilist
(extractive) colonies in regions with high indigenous population densities and/or settler mortality, but
settled permanently and established less mercantilist institutions and policies similar to those from their
home country in regions with low indigenous population densities and/or settler mortality rates. It
differs however in two respects. First, we do not challenge that settler mortality rates would have
13
Klerman et.al. (2011) use levels of school and life expectancy in 1960 as proxies for colonial policies. In our view,
these two variables are development outcomes that may be correlated with early institutions and policies, but are
not per se a good measure of them.
13
influenced the decision whether to establish permanent settlements and implement inclusive or
extractive institutions; however, as Olson (1996) suggests, the presence of large native populations
would have limited the ability of the colonizers to adopt institutions and policies resembling those in
their home country if the natives comprised a significant proportion of the total population and had
previously established their own set of institutions and policies. In such circumstance, the colonizers
would represent a weak minority, limiting their ability to implement radical institutional change
peacefully. This would have been the case even in regions in which colonizers experienced low mortality
rates and were able to establish inclusive institutional arrangements. As such, we contend that
indigenous population density is a more appropriate proxy for colonial settlement conditions than is the
settler mortality rate. This rationale, combined with recent controversy surrounding the settler mortality
rate data (c.f. Albouy, 2012 and AJR 2011), is the basis for our use of the population density in 1500 as
the main instrument for institutions in the analysis to follow, although we do report analogous results
using the settler mortality rates.
Next, the settlement strategy theory assumes that home country institutions and policies were
homogenous across Europe during the colonial era. As discussed above, this was not the case,
particularly with respect to economic institutions and policies. Specifically, the British were decidedly
less mercantilist and a common law system was emerging by the colonial era, whereas most of
continental Europe embraced a stronger form of mercantilism and operated under a civil law system
such that one would expect early institutions and policies to be less mercantilist among British than
other European colonies.14 Our identification strategy allows for a differential impact on the
development of institutions in colonies settled by the British and is given by equation (5) in which the
initial post-colonial institutional environment is a linear function of settlement conditions and
settlement conditions interacted with a dummy variable equal to one if the colonizer was Britain.15 We
use the colonial identification classifications of Klerman et.al. (2011).The main results reported below
use population density in 1500 as the proxy for settlement conditions.
(5)
One of Albouy’s (2012) criticisms of the settler mortality rate data used by Acemoglu and his co-authors,
who used a log transformation of the variable that reduces the variance, is that outliers were driving the
result that institutions are a strong and robust causal determinant of economic performance.16 Even
though we make use of the population density in 1500 data as our proxy for colonial settlement
conditions that to our knowledge have thus far avoided a similar critique, the variable does nonetheless
have a large standard deviation due to the presence of several right-skewed outliers. In an effort to
14
Glaeser and Andrei Schleifer (2002) argue that the divergence of English common law and French civil law
th
th
institutions originates with the advent of the first national court systems in the 12 and 13 centuries, with the
former adopting a jury adjudication process and the latter a process by which professional judges adjudicated
disputes. Klerman and Mahoney (2007) dispute this finding, arguing instead that the divergence of the two
th
th
systems occurred in the 17 and 19 centuries as a result of political choices made following revolutions in the
two countries.
15
Note that a constant term is omitted in equation 3, suggesting that initial institutions and policies were
dependent entirely on the settlement conditions and identity of the colonizer. We also assume that additional
covariates such as climate and geography factors that may influence the development of both modern institutions
and development are excluded from this equation on the basis that these factors are accounted for in the
settlement conditions variables.
16
Acemoglu, Johnson and Robinson (2011) replied to Albouy’s (2012) criticism, suggesting that limiting the effect
of outliers actually strengthens their results rather than weakening them.
14
mitigate the potential for a similar condition along with a reasonable theoretical conjecture, we adopt a
different transformation metric that rescales population density in 1500 to a 0-1 scale that is decreasing
using the formula
, where and are the adjusted and nominal population
densities in 1500 for colony , respectively, and
̅
. This transformation rescales the
variable in a relative sense such that sparsely populated regions have values approaching one, while the
most densely populated areas receive a value approaching zero.17 This has the benefit of simplifying
both our theory and the interpretation of point estimates. A one unit increase in population density in
1500 is equivalent to the difference between an uninhabited and the most densely populated regions.
The rescaled metric assumes that that negative effect of indigenous population density on institutional
development fails to exert a differential impact beyond a certain density level. In other words, the
marginal effect of indigenous population density on institutional development is zero above
.
We anticipate a positive relationship between the rescaled population density in 1500 and
contemporary institutions variables, and the effect to be greater among colonies settled by the British.
Figure 1 illustrates this relationship by plotting initial mercantilist institutions and policies against
population density in 1500. Regardless of the identity of the colonizer, it is postulated that the
colonizers would have attempted to establish highly mercantilist and likely purely extractive institutions
and policies in regions with the highest indigenous population densities. In less populated regions, early
institutions and policies are predicted to be less mercantilist among British than other European
colonies.
IV. Empirical Results
A. Determinants of Current Institutions: OLS Results
Substituting equation (5) into equation (4) yields the recursive instrument, , given by equation (6)
where
and
are the composite coefficient and error terms, respectively. OLS
estimates of equation (6) using population density in 1500, henceforth PD1500, as the proxy for
settlement conditions are given in Table 3. Panels A, B and C report the results using three different
measures of contemporary institutions: average constraints on the executive over the period 19602010, average risk of expropriation over the period 1985-1995 and average economic freedom over the
period 1985-2010, respectively.
(
)
(6)
Column (1) omits the PD1500 adjusted from British colonization term and thus does not allow for
settlement conditions to exert a differential impact on the development of institutions in former British
colonies, resulting in an overestimation of the effect that population density in 1500 exerts on
contemporary institutions. Nonetheless, the estimates obtain from column (1) indicate that PD1500
exerted more influence on economic freedom than it did on the other two institutional measures. The
regression in panel C has a much higher adjusted R2 value and F-statistic than those of panels A and B.
Additionally, the point estimates indicate that a unit increase in PD1500 (a decline from the most to
17
Using
̅
results in Burundi, Bangladesh, Egypt, India, Morocco, Pakistan, Rwanda, Sri Lanka
and Sudan being assigned
for population density in 1500. Applying the same transformation to the settler
mortality rate data results in Cote d’Ivoire, Ghana, Guinea, The Gambia, Madagascar, Mali, Niger, Nigeria, Sierra
Leone and Togo being assigned
. Similar results are obtained when setting
equal to 0.5, 0.75, and 1
standard deviations above the mean, respectively, for each variable.
15
least densely populated region) is associated with a 0.98 standard deviation rise in economic freedom,
but only a 0.91 and 0.73 standard deviation increase in constraints on the executive and risk of
expropriation, respectively.
Low
Level of Early Mercantilism
𝜹𝟏
𝜹𝟐
𝜹𝟏
High
0
Max
1
Population Density in 1500
Uninhabited
FIGURE 1: EARLY MERCANTILISM VS. SETTLEMENT CONDITIONS
Column (2) adds the PD1500 adjusted for British colonization term, increasing the explanatory power
and joint significance of the regressors substantially relative to column (1) for all three institutional
measures, although panel C remains the specification with the highest R2 and F values. Despite the
added explanatory power and increased joint significance, the unadjusted PD1500 term is only
statistically significant in panel C, suggesting that our identification strategy is more appropriate when
economic freedom is the measure of institutions. The estimates suggest the British colonization exerted
a significant differential impact on the development of economic freedom, as a unit increase in PD1500
is associated with a 2 standard deviation increase in modern economic freedom among former British
colonies, but only a 0.67 standard deviation rise among non-British colonies.
[INSERT TABLE 3 HERE]
Column (3) includes latitude as an additional covariate to control for the effect that climate may exert
on the development of institutions, and column (4) adds ethno-linguistic fractionalization as a means to
control for the effect of ethnic heterogeneity on institutional development. The signs of these two
additional covariates are positive and negative, as anticipated, and both are statistically significant at the
5 percent level in all three panels. The coefficients on the main variables of interest, PD1500 and
PD1500 adjusted for British colonization, remain qualitatively similar to the estimates obtained in
column (2), although the magnitudes are moderately lower. Controlling for the impact of climate and
ethnic heterogeneity, a unit increase in PD1500 is associated with a 1.87 and 0.58 standard deviation
increase in contemporary economic freedom among former British and non-British colonies.
16
Following AJR (2001, Table 4), column (5) excludes from the sample the four nations described by AJR as
the “Neo-European” countries: Australia, Canada, New Zealand and the United States. This exercise has
the benefit of checking sensitivity of the estimates to the inclusion of these four former British colonies,
which we refer to as the “Neo-Englands,” that all belong to the OECD and are among the most
economically free and developed countries in the world today.18 The explanatory power of each
regression is reduced substantially with the exclusion of the Neo-Englands, but appears to be driven by
the reduction in both the magnitude and statistical significance of latitude as a determinant of economic
freedom. Excluding the Neo-Englands from the sample modestly enhances the findings that settlement
conditions mattered for the development of institutions and that British colonization exerted a positive
differential influence, as the partial effects suggest that a unit increase in PD1500 is associated with a
1.91 and 0.62 standard deviation rise in contemporary economic freedom among former (non NeoEngland) British and non-British colonies, respectively. Rather than discrediting our theory that
settlement strategy and the identity of the colonizer are important determinates of the post-colonial
development of economic institutions, excluding the Neo-Englands from the sample strengthens it. In
addition, the estimates suggest that climate is less important as a determinant of economic institutions,
as latitude loses statistical significance.
Congruent with AJR (2001, Table 4), column (6) excludes the African countries from the sample, and
column (7) includes continent dummies for Africa and Asia as additional covariates to control for
potential regional fixed effects on the development of institutions.19 Excluding the African countries in
column (6) reduces the sample size to 38 nations, but only modestly reduces the explanatory power,
joint significance of the regressors and estimated impact that settlement conditions exerted on
contemporary economic freedom, relative to column (4). The inclusion of continent dummies in column
(7) only trivially influences the results, as the African and Asian fixed regional effects are in general
insignificant with controls for climate and culture already in place.20 Holding climate, culture and fixed
continent effects constant, the estimates from column (7) indicated that a unit increase in PD1500 is
associated with a 1.76 and 0.74 standard deviation increase in contemporary economic freedom among
former British and other colonies, respectively.
Taken as a whole, the OLS estimates from Table 3 provide evidence in support of our theory. Settlement
conditions in place prior to or at the time of colonization do appear to have exerted an influence on the
development of institutions, particularly economic ones, with British colonization having a positive
differential impact, ceteris paribus. Extractive institutions that evolved into highly mercantilist ones
appear to have been established in colonies densely populated with indigenous persons, even among
former British colonies. Meanwhile more inclusive institutions were established in relatively uninhabited
colonies, with the British colonies adopting relatively less mercantilist institutions and policies as is
exhibited by high levels of contemporary economic freedom. Similar results are found when we also
allow for Spanish colonization to exert a differential impact on the development of institutions by
adding PD1500 interacted with a dummy variable equal to one for Spanish colonization to equation (6).
18
22 of the 64 countries in the sample are former British colonies.
We depart slightly from AJR (2001, Table 4). They also included an “other continent” dummy variable that was
coded as a 1 for Australia, Malta and New Zealand. We exclude this variable as it appears to be an ad hoc grouping
of countries, effectively re-categorizing these countries with the remaining “America” as an “other than Asia or
Africa” dummy that is represented as the base. Including AJR’s “other continent” dummy does not materially affect
the results.
20
There are two exceptions: Africa in negative and highly significantly associated with constraints on the executive,
and Asia is positive and marginally significantly related to economic freedom.
19
17
These results are reported in Table 4. In addition, we obtain similar results when settler mortality rates
are used as the proxy for settlement conditions in lieu of PD1500, with these results reported in Table
5.21
[INSERT TABLE 4 HERE OR APPENDIX]
[INSERT TABLE 5 HERE OR APPENDIX]
B. Mercantilist Institutions and Economic Performance: 2SLS Results
Instrumenting with , two-stage least squares (2SLS) estimates of equation (1) are presented in Table
6. The dependent variable in all columns is the Penn World Tables measures of 2010 log GDP per capita.
Panel A, B and C use the average constraints on the executive over the period 1960-2010, average risk
of expropriation over the period 1985-1995, and average economic freedom over the period 1985-2010,
as the measure of institutions, respectively. The first stage results for each column are given in the
corresponding columns of Table 3.
The OLS results reported in Tables 3 and 4 suggest that our identification strategy of using PD1500,
adjusted for the identity of the colonizer, performs better when economic freedom is used as the
measure of institutions rather than the single-dimensional measures of constraints on the executive and
risk of expropriation. Tests of weak instruments reinforce these results. We report the Kleibergen-Paap
Wald first-stage F-statistics and Stock-Yogo maximal Wald test size distortion critical values in Table 6.
Maximal bias critical values are not reported because the number of instruments is sufficiently small
such that that Staiger and Stock (1997) rule of thumb that instruments be deemed weak if the first-stage
F-statistic is less than ten reasonably approximates a 5% test that the worst relative bias is 10% or less
(Stock and Yogo, 2002). Test results from column (1), which includes unadjusted PD1500 as the sole
instrument, are highly suggestive that the regressions in panels A and B have weak instruments, while
the null of weak instruments is rejected for panel C. When an additional instrument, PD1500 adjusted
for British colonization, is included in column (2), an F-statistic of 9.6 in panel A is less than the StaigerStock rule of thumb and falls between the Stock-Yogo critical values for a Wald test size distortion of 15
and 20 percent, suggesting that the two instruments are moderately weak for constraints on the
executive. The F-statistic of 11.7 in panel B suggests that the instruments are slightly stronger for risk of
expropriation. In panel C the F-statistic is 17.8, indicating that the instruments are much stronger for
economic freedom. Similar results emerge when additional covariates are included and sample
restrictions are imposed. Using the size of the F-statistic as a measure of the relative strength of the
instruments, for each specification, the instruments are strongest when economic freedom is the
measure of institutions. The results from these tests are highly indicative that the instruments are valid
for economic freedom, with the only sign of modestly weak instrumentation occurring in column (5),
which omits the 4 Neo-England nations, as the F-stat is 8.6, less than the Staiger-Stock rule of thumb and
20 percent Wald test size distortion critical values.
21
We use the transformation metric described above to rescale the nominal settler mortality rate (SMR) data to a
0-1 scale that is decreasing in SMR. The results obtained using SMR as the proxy for settlement conditions at first
appear to be most favorable when constraints on the executive is the institutional measure. Difference in means
tests of the null that the coefficient on the SMR adjusted for British and Spanish colonization terms are not
statistically different suggest that we in general cannot reject the null when constraints on the executive is the
measure of institutions, but we do in general reject it in panels B and C. This has implications for the channels
through which settlement conditions exert an influence on institutional development.
18
Using the Sargan-Hansen test for over-identification, we easily fail to reject the joint null that the
instruments are uncorrelated with the error term and that the excluded instruments are correctly
excluded in all specifications in panel C, for which economic freedom is the measure of institutions. We
fail to reject the null in all specifications except column (6) in panel A, for which constraints on the
executive is the institutional measure. For panel B, which uses risk of expropriation, we reject the null in
all but column (7), suggesting that over-identification is a problem when risk of expropriation is the
institutional measure. P-values for the over-identification tests are reported for each regression. We
also report p-values for the Kleibergen-Papp rk under-identification test, and reject the null hypothesis
that the equation is under-identified at the 5 percent level in all but panel B of column (1).
Statistical tests for the validity of the instruments strongly suggest that our identification strategy is valid
when economic freedom is the institutional measure, as weak instruments are suspect when either
constraints on the executive and risk of expropriation are used, and over-identification may be an issue
for the latter. Given these results, the remaining discussion will focus predominantly on the results
reported in panel C that use economic freedom as the measure of institutions.
The 2SLS estimate of the impact of economic freedom, instrumented for by PD1500, on income per
capita in column (1) is 1.35, highly significant with a heteroskedastic-robust standard error of 0.25. The
point estimate is more than a third greater than the corresponding OLS estimate of 0.99, reported as ̂ .
The corresponding point estimates for column (1) of 1.02 and 1.66 in panels A and B are approximately 2
and 2.5 fold increases, respectively, relative to their OLS counterparts, suggesting that attenuation bias
attributable to measurement error is less pronounced for economic freedom than the other two
institutional variables. This is likely attributable to economic freedom being derived from a
comprehensive set of objective measures of economic institutions and policies.
After allowing for the differential effect of settlement conditions by the British on institutional
development, the magnitude of the impact of economic freedom on income per capita is reduced
slightly to 1.15 in column (2). This estimate remains highly significant with a robust standard error of
0.14, and is one-sixth greater than the OLS estimate. Recall the 1.4 point difference in economic
freedom measures between Costa Rica and Ghana. The estimated partial effect of 1.15 suggests that
there should be a 1.58 log point (approximately 4-fold) difference in the income levels of the two
countries, less than the 1.79 log point (approximately 5-fold) difference observed in practice. Thus, the
estimates of Table 4 imply that differences in the institutional environments between countries can
result in substantial, but not implausibly large, differences in economic performance.
Including latitude in column (3) only modestly diminishes the impact of economic freedom on income
per capita, as the point estimate is reduced slightly to 1.15 and remains highly significant with a robust
standard error of 0.17. Latitude, which has a positive and highly significant relationship with economic
freedom in the first stage estimates, has the anticipated positive sign in the second stage but is not
significant. This result is consistent with the findings of AJR (2001) and RST (2004), and even though it is
tempting to conclude that climate only influences economic performance indirectly through its influence
on institutional development , Sachs (2003) notes that latitude may not be the best proxy for the effect
that climate may exert on the development process. This result should therefore be interpreted with
caution and will be addressed further in the next section.
Column (4) adds ethno-linguistic fractionalization to the regression. It has the expected negative sign,
and the -0.89 coefficient is significant at the 5 percent level (s.e. = 0.35). A significant coefficient of -0.79
in the first stage regression, representing the indirect effect, indicates that the total effect of ethnic
19
heterogeneity on economic performance is fairly large. A one unit increase in ethno-linguistic
fractionalization, equivalent to the difference between the most and least fractionalized nations, is
associated with a decrease in income per capita of 1.8 log points among former colonies, after
controlling for the influence of institutions. The estimated impact of economic freedom on economic
performance is unchanged when ethno-linguistic fractionalization is included in the regression, although
the structure of the model assumes that the latter is exogenous to the system. As demonstrated by AJR
(2001, appendix A), if ethno-linguistic fractionalization is endogenous to the system then the coefficient
on economic freedom is biased downwards.22
Columns (5) and (6) document that the result that economic freedom exerts a significant impact on
income per capita is not sensitive to either the inclusion of the four Neo-England nations of Australia,
Canada, New Zealand and the United States, or the inclusion of the nations of Africa, respectively. The
institutional impact for the former actually increases to 1.28, and remains highly significant with a
robust standard error of 0.23, when the Neo-England nations are excluded. The coefficient does
diminish modestly to 1.07 with the exclusion of the 26 African nations from the sample, although it also
remains highly significant (s.e. = 0.17). The coefficient on ethno-linguistic fractionalization remains
negative and significant at the five percent level for both equations, and latitude remains insignificant.
Column (7) includes continent dummies for Africa and Asia, both of which have a negative and
significant coefficient; however, given that ethno-linguistic fractionalization is now insignificant and it is
fairly well correlated (
) with the Africa dummy, this result may be attributable to multicollinearity rather than a systematic direct negative effect of being located on the African continent.
Nonetheless, including continent fixed effects does not materially alter that finding that economic
freedom exerts a strong and positive impact on income per capita, as the coefficient of 1.01 remains
highly significant with a standard error of 0.14.
Overall, the results in Table 6 suggest that institutions associated with economic freedom exert a large
and significant effect on economic performance. They also indicate that our identification strategy that
allows settlement conditions to exert a differential impact in British colonies on the development of
institutions is valid when economic freedom is used as the measure of institutions, but that the
instruments are weak when either constraints on the executive or risk of expropriation are used as the
institutional measure. Furthermore, statistical tests point to over-identification for the regressions in
panel B. The results are nearly identical when we use the World Development Indicator 2010 income per
capita as the dependent variable. These results are reported in Table 7. Very similar results are obtained
when we also include PD1500 adjusted for Spanish colonization as an additional excluded instrument,
although this improves the strength of the instruments for the constraints on the executive measure in
panel A. These results are reported in Table 8. We investigate the robustness of these results below.
[INSERT TABLE 7 HERE OR APPENDIX]
[INSERT TABLE 8 HERE OR APPENDIX]
22
Given that the coefficient for economic freedom obtained in column (2) is identical to that in column (4),
combined with the time lag involved with the measurement of ethno-linguistic fractionalization, we are fairly
confident that any bias attributable to endogeneity of ethnic heterogeneity is relatively small. Mauro (1995) and
Auer (2013) find that ethno-linguistic fractionalization only influences growth through its effect on the
development institutions. They employ different measures of institutions, possibly explaining the different results.
20
V. Robustness
[THIS SECTION IS STILL IN PROGRESS…COMMENTS/ RECOMMENDATIONS WELCOME]
VI. Concluding Remarks
This paper integrates the two major institutional theories of comparative-post colonial development –
settlement strategy (Acemoglu et.al. 2001, 2002, 2005) and legal-colonial origins (e.g. La Porta et.al.,
1999, 2008; Klerman et.al., 2011) –to form a unified theory that accounts simultaneously for both the
impact of settlement conditions and differences in colonial history among the European colonizers as
sources of variation in institutional development, and their potential causal effect on modern levels of
development. We pursue an identification strategy that allows the impact of settlement strategy on the
development of institutions to vary based on the identity of the colonizer. Specifically, inclusive British
settlement exerted a substantial positive differential impact on the quality of institutions, promoting
stronger long-run economic growth, relative to inclusive settlement by the other European nations,
because the British practiced a much milder form of mercantilism than that present in France, Portugal,
and Spain, the other major European colonizing powers.
We define mercantilism to cover a broader cluster of institutions and policies that influence the
allocation of resources, entrepreneurial activity and investment decisions. Many of the mercantilist
institutions and policies are negatively encapsulated by the Frasier Institute’s Economic Freedom of the
World, making it an ideal measure of institutions to test our hypothesis. The results are supportive of
our theory, indicating that: (1) the colonizer-adjusted instruments are a better predictor of
contemporary institutions and income per capita than the unadjusted instruments, (2) British influence
exerted a more favorable impact on institutional development than other European colonizers, ceteris
paribus, and (3) the Fraser measure of economic institutions exerts a stronger impact on income per
capita than constraints on the executive, the preferred institutional measure of Acemoglu and his coauthors.
[THIS SECTION IS STILL IN PROGRESS…COMMENTS/ RECOMMENDATIONS WELCOME]
21
References
Acemoglu, Daron. “Constitutions, politics, and economics: A review essay on Persson and Tabellini’s The
Economic Effects of Constitutions.” Journal of Economic Literature 43 (2005): 1025-1048.
Acemoglu, Daron, Simon Johnson and James A. Robinson. “The colonial origins of comparative
development: An empirical investigation.” The American Economic Review 91, no. 5 (2001): 1369-1401.
Acemoglu, Daron, Simon Johnson, and James A. Robinson. "Reversal of fortune: Geography and
institutions in the making of the modern world income distribution." The Quarterly Journal of Economics
117, no. 4 (2002): 1231-1294.
Acemoglu, Daron, Simon Johnson and James A. Robinson. “Hither thou shalt come, but no further: Reply
to ‘The colonial origins of comparative development: An empirical investigation: Comment.” NBER
working paper, no. 16966 (2011).
Acemoglu, Daron and Simon Johnson. “Unbundling Institutions.” The Journal of Political Economy 133,
no. 5 (2005): 949-995.
Acemoglu, Daron and James A. Robinson. “Economic backwardness in political perspective.” The
American Political Science Review 100, no. 1 (2006a): 115-131
Acemoglu, Daron and James A. Robinson. “De facto political power and institutional persistence”
American Economic Review 96, no. 2 (2006b): 325-330
Acemoglu, Daron and James A. Robinso.n (2008) “persistence of power, elites, and institutions”
American Economic Review 98, no. 1 (2008): 267-293
Acemoglu, Daron and James A. Robinson. Why Nations Fail: The Origins of Power, Prosperity, and
Poverty. New York: Crown Business (2012).
Aixalá, José and Gema Fabro. “Economic freedom, civil liberties, political rights and growth: a causality
analysis.” Spanish Economic Review 11 (2009): 165–178.
Albouy, David Y. “The colonial origins of comparative development: An empirical investigation:
Comment.” American Economic Review 102, no. 6 (2012): 3059-3076.
Ashby, Nathan J. and Russel S. Sobel. “Income inequality and economic freedom in the U.S. states.”
Public Choice 134 (2008): 329-346.
Asraf, Quamrul and Oded Galor. “Dynamics and stagnation in the Malthusian epoch” American
Economic Review, 101, no.5 (2011): 2003-41
Auer, Raphael A. “Geography, institutions, and the making of comparative development.” Journal of
Economic Growth 18, no. 2 (2013): 179-215.
Azzimonti, Marina. “Barriers to Investment in Polarized Societies” American Economic Review, 101, no. 5
(2011): 2182-2204.
22
Barro, Robert J. “Democracy and growth” Journal of Economic Growth 1, no. 1 (1996): 1–27
Baumol, William J. “Entrepreneurship: Productive, Unproductive, and Destructive.” The Journal of
Political Economy 98, no. 5 (1990): 893-921.
Bertocchi, Graziella and Fabio Canova. “Did colonization matter for growth? An empirical exploration
into the historical causes of Africa´s underdevelopment” European Economic Review 46 (2002): 18511871
Bjørnskov, Christian. “Constitutional property rights protection and economic growth: Evidence from the
post-communist transition.” The Ratio Institute, no. 189 (2012).
Collins, John Churton. Voltaire, Montesquieu and Rousseau in England. London: Eveleigh Nesh (1908).
Dawson, John W. “Institutions, investment and growth: New cross-country and panel data evidence.”
Economic Inquiry 36 (1998): 603-619.
Dawson, John W. “Causality in the freedom-growth relationship.” European Journal of Political Economy
19 (2003): 479-495.
De Haan, Jakob and Jan-Egbert Sturm. “Does more democracy lead to greater economic freedom? New
evidence for developing countries.” European Journal of Political
Economy 19 (2003): 547–563
De Haan, Jakob, Susanna Lundström, and Jan‐Egbert Sturm. "Market‐oriented institutions and policies
and economic growth: A critical survey." Journal of Economic Surveys 20, no. 2 (2006): 157-191.
Diamond, Jared. Guns, Germs and Steel. New York: W.W. Norton & Company (1997).
Domar, Evsey D. “Capital expansion, rate of growth, and employment.” Econometrica 14, no. 2 (1946):
137-147.
Easterly, William.“Inequality does cause underdevelopment: Insights from a new instrument” Journal of
Development Economics 84 (2007): 755-776.
Easterly, William and Ross Levine. “Africa’s growth tragedy: politics and ethnic divisions.” The Quarterly
Journal of Economics 112, no. 4 (1997): 1203-1250.
Easterly, William and Ross Levine. "Tropics, germs, and crops: how endowments influence economic
development." Journal of Monetary Economics 50, no. 1 (2003): 3-39.
Feld, Lars and Stefan Voigt. “Economic growth and judicial independence: Cross-country evidence using
a new set of indicators.” European Journal of Political Economy 19, no. 3 (2003): 497-527.
Gallup, John Luke, Jeffrey D. Sachs, and Andrew D. Mellinger. "Geography and economic development."
International Regional Science Review 22, no. 2 (1999): 179-232.
23
Galor, Oded. “Unified growth theory: From stagnation to growth.” In Philippe Aghion and Steven N.
Durlauf (Eds.), Handbook of Economic Growth, Amsterdam: Elsevier North Holland (2005): 171-293.
Glaeser, Edward L., Rafael La Porta, Florencio Lopez-De-Silanes, and Andrei Shleifer. “Do Institutions
Cause Growth?” Journal of Economic Growth 9 (2004): 271-303.
Glaeser, Edward L. and Andrei Shleifer. “Legal origins.” The Quarterly Journal of Economics 17, no. 4
(2002): 1193-1229.
Grampp, William D. “The liberal elements of English mercantilism.” The Quarterly Journal of Economics
66, no. 4 (1952): 465-501.
Grier, Robin M. “Colonial legacies and economic growth.” Public Choice 98 (1998): 317-335.
Gwartney, James D., Randall G. Holcombe, and Robert A. Lawson. "Institutions and the impact of
investment on growth." Kyklos 59, no. 2 (2006): 255-273.
Gwartney, James, Robert Lawson, and Joshua Hall. Economic Freedom of the World Annual Report. The
Fraser Institute, 2012.
Hall, Joshua C., Russell S. Sobel and George R. Crowley. “Institutions, capital and growth.” Southern
Economic Journal 77, no. 2 (2010) 385-405.
Hall, Robert E. and Charles I. Jones. “Why do some countries produce so much more
output per worker than others?” Quarterly Journal of Economics 114, no. 1 (1999): 83–116.
Harrod, Roy F. “An essay in dynamic theory.” The Economic Journal 49, no. 193 (1939): 14-33.
Heston, Alan, Robert Summers and Bettina Aten. Penn World Table Version 7.1, Center for International
Comparisons of Production, Income and Prices at the University of Pennsylvania (Nov 2012).
Holcombe, Randall G. “Entrepreneurship and Economic Growth.” The Quarterly Journal of Austrian
Economics 1, no. 2 (1998): 45-62.
Huntington, Samuel. Political Order in Changing Societies, New Haven & London: Yale University Press
(1968).
Johnson, Simon, William Larson, Chris Papageorgiou and Arvind Subramanian. “Is newer better? Penn
World Table revisions and their impact on growth estimates.” Journal of Monetary Economics 60, no. 2
(2013): 255-274.
Jones, Charles I. and Paul M. Romer (2010) “The new Kaldor facts: Ideas, institutions, population, and
human capital.” American Economic Journal: Macroeconomics 2, no. 1 (2010): 224-245.
Jong-A-Pin, Richard and Jakob De Haan. “Growth accelerations and regimen changes: A correction” Econ
Journal Watch 5, no.1 (2008): 51-58
24
Jong-A-Pin, Richard and Jakob De Haan. “Political regime change, economic liberalization and growth
accelerations” Public Choice 146 (2011): 93-115
Kirzner, Israel M. Competition and Entrepreneurship. Chicago: University of Chicago Press (1973).
Kirzner, Israel M. Discovery and the Capitalist Process. Chicago: University of Chicago Press (1985).
Klerman, Daniel and Paul G. Mahoney. “Legal origin?” Journal of Comparative Economics 35 (2007): 278293.
Klerman, Daniel M., Paul G. Mahoney, Holger Spamann and Mark Weinstein. “Legal Origin or Colonial
History?” Journal of Legal Analysis 3, no 2 (2011):379-409
Knack, Stephen and Philip Keefer. “Institutions and economic performance: Cross-country tests using
alternative institutional measures.” Economics and Politics 7, no. 3 (1995): 207-227.
Kornai, Janos. The Socialist System: The Political Economy of Communism. Princeton, N.J.: Princeton
University Press (1992).
La Porta, Rafael; Florencio Lopez-de-Silanes, Andrei Shleifer and Robert Vishny. “Legal determinants of
external finance.” Journal of Finance 52, no. 3 (1997): 1131-1150.
La Porta, Rafael; Florencio Lopez-de-Silanes, Andrei Shleifer and Robert Vishny. ”Law and Finance.”
Journal of Political Economy 106, no. 6 (1998): 1113-1155.
La Porta, Rafael; Florencio Lopez-de-Silanes, Andrei Shleifer and Robert Vishny. “The quality of
government”, The Journal of Law, Economics & Organization 15, no.1 (1999): 222-279.
Landes, David. The Wealth and Poverty of Nations, Why Some Are So Rich and Some Are So Poor. New
York: W.W. Norton and Company (1998).
Lawson, Robert A. and J.R. Clark. “Examining the Hayek-Friedman hypothesis on economic and political
freedom.” Journal of Economic Behavior & Organization 74, no. 3 (2010): 230-239.
Lucas, Robert E. Jr. “On the mechanics of economic development.” Journal of Monetary Economics 22
(1988): 3-42.
Lundström, Susanna. “The effect of democracy on different categories of economic freedom.” European
Journal of Political Economy 21 (2005): 967–980.
Machiavelli Niccoló. Discourses on Livy, Chicago: University of Chicago Press, [1519] 1996.
Mahoney, Paul G. “The common law and economic growth: Hayek might be right.” Journal of Legal
Studies 30 (2001): 503-525.
Mankiw, N. Gregory, David Romer and David N. Weil. “A contribution to the empirics of economic
growth.” The Quarterly Journal of Economics 107, no. 2 (1992): 407-437.
25
Mauro, Paolo. “Corruption and Growth” The Quarterly Journal of Economics, 110 (1995): 681-712.
Nettels, Curtis P. “British mercantilism and the economic development of the thirteen colonies.” The
Journal of Economic History 12, no. 2 (1952): 105-114.
North, Douglass C. Structure and Change in Economic History New York: Norton and Company (1981).
North, Douglas C. “Institutions.” Journal of Economic Perspectives 5, no. 1 (1991): 97-112.
North, Douglas C. Institutions, Institutional Change, and Economic Performance. Cambridge: Cambridge
University Press (1991).
Olson, Mancur Jr. “Big bills left on the sidewalk: Why some nations are rich, and others poor.” Journal of
Economic Perspectives 10, no. 2 (1996): 3-24.
Olson, Mancur Jr. Power and Prosperity: Outgrowing Communist and Capitalist Dictatorships. New York:
Basic Books (2000)
Paldam, Martin. “Economic freedom and the success of the Asian tigers: An essay on controversy.”
European Journal of Political Economy 19 (2003): 453-477.
Paldam, Martin and Erich Gundlach. “Two views on institutions and development: The grand transition
vs the primacy of institutions.” Kyklos 61, no. 1 (2008): 65-100.
Pipes, Richard. Property and Freedom. New York: Alfred A. Knopf (1999).
Pitlik, Hans. “The impact of growth performance and political regime type on economic policy
liberalization.” Kyklos 61 (2008): 258–278.
Pitlik, Hans and Steffen Wirth. “Does crises promote the extent of economic liberalization? An empirical
test.” European Journal of Political Economy 19 (2003): 565–581
Roe, Mark J. and Jordan Siegel (2011) “Political instability: Effects on financial development, roots in the
severity of economic inequality.” Journal of Comparative Economics 39, no. 3 (2011): 279-309,
Ram, Rati and Secil Ural. “Comparison of GDP per capita data in Penn World Table and World
Development Indicators.” Social Indicators Research (2013), forthcoming.
Rode, Martin and Sebastian Coll. “Economic freedom and growth: Which policies matter the most?”
Constitutional Political Economy 23, no. 2 (2012): 95-133.
Rode, Martin and James D. Gwartney. “Does democratization facilitate economic liberalization?”
European Journal of Political Economy 28 (2012): 607-619.
Rodrik, Dani, Arvind Subramanian, and Francesco Trebbi. "Institutions rule: the primacy of institutions
over geography and integration in economic development." Journal of Economic Growth 9, no. 2 (2004):
131-165.
26
Romer, Paul M. “Increasing returns and long-run growth.” Journal of Political Economy 94, no. 5 (1986):
1002-1037.
Romer, Paul M. “Endogenous technological change.” Journal of Political Economy 98, no. 5 (1990): S71S102.
Rostowski, Jacek and Bogdan Stacescu. “The wig and pith helmet: The impact of ‘legal school’ versus
colonial institutions on economic performance (second version).” CASE Studies & Analyses, no 300
(2006).
Sachs, Jeffrey D. “Tropical Underdevelopment.” NBER working paper no. w8119 (2001).
Sachs, Jeffrey D. “Institutions don’t rule: Direct effects of geography on per capita income.” NBER
working paper no. 9490 (2003).
Sachs, Jeffrey D. and Andrew Warner. “Economic reform and the process of global integration.”
Brookings Papers on Economic Activity 1 (1995): 1-118.
Sachs, Jeffrey D. and Andrew M. Warner. “Sources of slow growth in African economies.” Journal of
African Economies 6, no. 3 (1997): 335-376.
Solow, Robert M. “A contribution to the theory of economic growth.” The Quarterly Journal of
Economics 70, no. 1 (1956): 65-94.
Sokoloff, Kenneth L. and Stanley L. Engerman. “Institutions, factor endowments, and paths of
development in the new world.” Journal of Economic Perspectives 14, no. 3 (2000): 217-232
Stock, James H. and Motohiro Yogo. “Testing for weak instruments in linear IV regression.” NBER
technical working paper no. 284 (2002).
Uslar, Arturo. Medio Milenio de Venezuela. Caracas: Cuaderno Lagoven, Refolite C.A (1962).
Vega-Gordillo, Manuel and José L. Álvarez-Arce. “Economic growth and freedom. A causality study.”
Cato Journal 23 (2003): 199–215.
Voltaire, Francois-Marie Arouet. Letters concerning the English nation. In Letter VIII. New York: Oxford
University Press (1926).
World Bank. World Development Indicators (2012).
Zweigart, Konrad and Hein Kötz. An Introduction to Comparative Law. Third edition. New York:
Clarendon Press (1998).
27
28
29
30
31
32
33
34
35
36