New articles in this journal are licensed under a Creative Commons Attribution 4.0 United States License.
This journal is published by the University Library System, University of Pittsburgh as part of
its D-Scribe Digital Publishing Program and is cosponsored by the University of Pittsburgh Press.
JOURNAL OF WORLD-SYSTEMS RESEARCH
Research Note
Measuring the Impacts of Colonialism: A New Data Set for the Countries of
Africa and Asia
Patrick Ziltener
University of Zurich, Switzerland
ziltener@soziologie.uzh.ch
Daniel Künzler
University of Fribourg, Switzerland
André Walter
University of St. Gallen, Switzerland
Abstract
We present a new dataset with 15 indicators for the political, economic and social impact of colonialism. This dataset and our four
indices for the impact of colonialism create for the first time the opportunity to compare directly the levels of colonial transformation
for a sample of 83 African and Asian countries. Some of our exploratory findings on the interrelation of the dimensions show that
in British colonies political domination was in general less direct and less violent. Plantation colonies experienced more investment
in infrastructure and more violence during decolonization. The correlations between indicators for economic distortion (trade
policy, trade and FDI concentration) show that the economic re-direction of some colonies towards a more exclusive exchange with
the metropole country was an interdependent process. In general, a more intense political domination came along with a higher
level of economic transformation. If an area was transformed economically, however, a social transformation was likely to take
place too, but these processes should not be confounded. In areas that were politically united for the first time under colonialism,
economic distortion and social transformation were more profound.
Keywords: Colonialism; political, economic and social impacts; Africa; Asia
ISSN: 1076-156X | Vol. 23 Issue 1 Pages 156-190 |
DOI 10.5195/JWSR.2017.683 | jwsr.org
Vol. 1 | DOI 10.5195/JWSR.1
http://www.library.pitt.edu/
http://www.pitt.edu/
http://www.library.pitt.edu/articles/digpubtype/index.html
http://upress.pitt.edu/
https://creativecommons.org/licenses/by/4.0/
mailto:ziltener@soziologie.uzh.ch
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
157
Introduction and Background1
Research from a world-system perspective stresses the importance of long-term historical factors,
but it has been confined to indirect measuring of historically earlier factors than those present in
the postcolonial era. Influential was the article by Lenski and Nolan (1984), focusing on the long-
term effects of social-evolutionary development levels on the postcolonial economy and society.
However, both precolonial and colonial factors could only be measured indirectly. A remedy to
the first flaw was the Atlas of Precolonial Societies (Mueller, et al. 1999). Based on this new data
set, Lenski and Nolan’s findings could be confirmed on a more elaborate level and with a larger
sample (Ziltener and Mueller 2007). However, the second flaw persisted, as the impacts of
colonialism were not measured with the same level of sophistication. The authors admitted that
the problem had remained unsolved and that “it should be addressed by future quantitative-
empirical research” (Ziltener and Mueller 2007:400). Since colonialism has shaped core-
periphery-relations (cf. Sanderson 2005:186-187), it is rather surprising that, in contrast to the
situation of postcolonial dependence, “colonization has not been a central concern to world-system
theory” (Boswell 1989:180). As a consequence, we simply do not know which aspects of
colonialism have been relevant in terms of long-term effects—the “open veins” (Galeano 1973),
i.e. the level of exploitation of resources, or investment, taxes, and trade policies, direct or indirect
rule, schooling and proselytization—all of them or only certain ones?
From a world-system perspective, the modes of integration of the postcolonial economies in
the periphery into the world economy are crucial. The discussion of how foreign direct investment
(FDI) in form of transnational corporations and their affiliates affect economic and social
development in their host countries since independence started the 1970s (Bornschier, Chase-
Dunn, and Rubinson 1978; Chase-Dunn 1975), and this has been a topic ever since (Herkenrath
and Bornschier 2004). We go a step back and have a look at the different levels of economic
dependence of colonies by differentiating between those with a strong unilateral dependence on
the metropole country regarding trade and investment and those without. We contribute to a wider
and more thorough consideration of colonial factors and their relevance for postcolonial
development.
New indicators to measure length, depth, and different impacts of colonization have to be
developed in order to define the different “legacies” of colonialism. As a first step, the temporal
boundaries of colonialism were coded for 83 Countries of Africa and Asia (Ziltener and Künzler
2013:292-296). A dimensional analysis within the same research survey (Ziltener and Künzler
1 The research on which this article is based has been funded by a grant by the Swiss National Science Foundation
(Statecraft in the non-Western World: A Multifactorial Model to Explain Persisting Differentiation, No.
101512112495). The authors would also like to thank Harold Kerbo for a critical review of this paper.
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
158
2013:297-305) revealed a complex web of interrelations of political, economic and social factors,
a challenge for any attempt to measure the impacts of colonialism. We have analyzed the existing
body of literature in our research survey (Ziltener and Künzler 2013). Until recently, empirical-
quantitative research has dealt inadequately with the problem of the multidimensional impact of
colonialism, typically using simple country-dummies (British vs. French colonies etc.). Some
authors argued that colonial powers introduced different kinds of institutions (legal systems,
extractive institutions etc.) which ultimately determined the post-colonial development of the
former colonies (cf. Acemoglu, Johnson, and Robinson 2001, 2002; Grier 1999; Mahoney 2003).
In research practice, this usually takes the form of one-dimensional concepts, completely
neglecting different levels of colonial transformation. We see this article as a contribution to
dealing with the impacts of colonialism in a more complex manner. We analyze the variation in
the forms of colonial domination as the first step in our effort to make the different aspects
measurable.
Table 1. Dimensions of colonial transformation (CT)
1. Political transformation 2. Economic transformation 3. Social transformation
1.1. Form of domination
1.2. Violence
1.3. Instrumentalization of
ethnolinguistic/religious
cleavages
1.4 Gradualism in the transfer of
administration
2.1. Trade policy
2.2. Trade concentration
2.3. Investment concentration
2.4. Investment in infrastructure
2.5. Plantations
2.6. Mining
2.7. Gold/silver/diamonds
3.1. Colonial immigration
3.2. Success of missionary
activities
3.3. Work immigration
3.4. Partition
We propose to use four indicators for the political dimensions of colonial domination, seven for
the economic and four for the social dimension (see table 1). The first three sections discuss the
background and construction of these indicators and present some descriptive statistics, following
the structure of table 1. Based on these 15 indicators, we constructed indices for the political,
economic and social impact of colonialism which we discuss in part four, followed by discussion
of a combined index for the impact of colonialism. This new data set creates for the first time the
opportunity to compare directly the levels of colonial transformation of the countries of Africa and
Asia. The full set of precolonial and colonial variables is available at worlddevelopment.uzh.ch,
and the dataset may be found at http://dx.doi.org/10.7910/DVN/UQZFYA; links provided
throughout the text add additional illustrative tables.
http://www.worlddevelopment.uzh.ch/en.html
http://dx.doi.org/10.7910/DVN/UQZFYA
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
159
Indicators for Political Transformation by Colonialism:
Form of Political Domination (DOMFORM)
Beyond the mere length of time of colonialism, the intensity of political domination is a crucial
aspect: without a significant reduction of the level of political sovereignty, we would not even
speak of colonialism (Ziltener and Künzler 2013:292). From a sociological viewpoint, the
historical pattern of foreign control does not depend on the official claim or de jure control. We
treat all so-called “mandates” as colonies. On a scale of five increasing levels of political
domination, we coded for every country in the sample the maximal impact:
0 = no colonial domination / not applicable
1 = semi-colonialism
2 = indirect rule with little interference in internal affairs
3 = indirect rule with strong interference in internal affairs
4 = direct rule
No colonial domination was coded only from Mongolia, which was in the 19th and 20th century
mostly under Chinese supremacy and thus not under ‘modern’ colonialism. All other countries of
our sample had an experience of being dominated by one or several colonial powers.
We call the “weakest” form of colonial domination semi-colonialism, characterized by the
reduction of political sovereignty through “unequal treaties” or similar arrangements. These
treaties usually were forced at gunpoint or were the results of wars. They were non-reciprocal and
included elements like
• the opening of ports to trade, the ending of state monopolies and of certain
interdictions of commerce (as in the case of opium in China)
• special trade/tariff concessions, most-favored-nation status
• the change of diplomatic traditions and communication rules according to European
standards, e.g. exemption of Europeans from court rituals deemed to be humiliating,
acceptance of a European language as diplomatic language
• the ending or lowering of travel and residency restrictions,
• extraterritoriality (geographical zones, administration, jurisdiction)
• freedom of religion, i.e. the right to preach Christianity openly and to exert missionary
activities, while giving the country’s administration the duty to protect persons and
facilities involved
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
160
• a de facto partition of the country into mutually exclusive spheres of influence (as in
the case of Persia between Britain and Russia, or of China as coordination among
several colonial powers).
These instruments were used mainly in dealing with the pre-colonial states and empires in Asia,
where “the presence of large, antique, yet still death-defying political structures meant that
indigenous authorities could not be taken over without promoting internal disorder, incurring
massive expense and risking international conflict“ (Cain and Hopkins 1993:397). Examples are
the Bowring Treaty (1855 Britain-Siam), the Tianjin Treaty (1858 Britain-China)2 and the Harris
Treaty (1858 US-Japan). The group of (present) countries coded as semi-colonial consists of
Afghanistan, Iran (Persia), Japan, Saudi Arabia, Thailand and Turkey (as the core area of the
Ottoman Empire), i.e. 7% of the cases in the sample.
With our concept “five increasing levels of political domination” we overcome most of the
problems that approaches working with a “col/non-col” binary are confronted with.3 We do not
consider it appropriate to code “more technologically advanced non-European countries”
including Turkey, Iran, China, and Japan “as never colonized by Europeans” (Ertan, Fiszbein, and
Putterman 2016:182, also Heldring and Robinson 2012:2). In some cases, diplomatic relations
became successively more unequal, as in the cases of the series of treaties of Britain with the small
emirates at the Persian Gulf, which makes it difficult to define the exact onset of semi-
colonialism/colonialism. The next level of intensifying political domination is defined by claims
of exclusive rights over the foreign relations of an area/country, the posting of an advisor or
resident at a court, the defense against third countries or by bringing parts of the administration
under direct control. This form of colonial domination is usually called “indirect rule,” because
traditional rulers stay in power and institutions remain at least formally unchanged.4 However,
“indirect rule” had a strong impact on the political sphere, as a Furnivall (1941:9) argues in the
case of Dutch rule in Indonesia who “subverted the native principle of rule, and drained the vitality
of social life.” Because the category “indirect rule” covers a large number of colonies, we
2 We base this on Auslin’s (2004:21) assessment: “The crucial change in Sino-Western relations came with the 1858
Tianjin treaty imposed by Britain after the Arrow War. This new treaty was vastly more punitive and intrusive than
the Treaty of Nanjing and materially reduced Chinese sovereignty.”
3 Cf. Ertan, Fiszbein, and Putterman (2016:170) who consider colonization to have begun “once 20% or more” of the
territory of a today's country is deemed by sources to have been “largely under the control of the colonizing power”.
4 In general, the caveat by Trocki (1999:80) applies: “While the legal status of the prior institutions could be taken as
a formal guide to the presence of a system of direct or indirect rule, in practice there was often little to distinguish the
systems. During this period, legal niceties were respected only at the convenience of the colonial power, though such
technical points once again became of importance when questions arose regarding the structure of the post-colonial
state. (...) indirect rule often signified no more than a transitory stage between the inception of a colonial presence and
the development of a more comprehensive administrative machine.”
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
161
differentiate between indirect rule with little interference in internal affairs (level 2) and indirect
rule with strong interference in internal affairs (level 3). This is somewhat arbitrary, but the
experience of coding the impact on traditional institutions and processes clearly confirmed the
importance and practicability of this distinction.
In cases where the area of today’s country was politically divided and experienced different
forms of domination over a longer period, as in the case of the Malay peninsula with directly ruled
“settlements,” federated and unfederated Malay states with British residents, but different degrees
of interference in internal affairs, we coded level 3 in view of the experience of the majority of the
population. We “balanced” the coding in cases of several groups of the population with
significantly different experiences, e.g. level 3 in a case with two major groups with direct ru le
and level 2 in one with indirect and little interference. One example for this is Cameroon, where
in the British part and the northern French part administration was based on local authorities as the
Lamibé and Sultans, while in the southern French part the administrative was clearly direct.
We did not code traditional empires (Mughal, Ottoman, Omani etc.) as colonialism (cf.
Ziltener and Künzler 2013:292). Here we are confronted with the problem how to code
areas/regions that were part of one of these empires and therefore were politically dominated
already before “modern” colonialism. The partition of the Ottoman Empire by European powers
led to new entities for which new political “centers” had to be created, and which led to economic
fragmentation: “The creation of these new political entities entailed the creation of new national
economies as well, each with its own boundaries, and each with its own centralized system of laws,
taxes, and fiscal management. One result was the growth of separate national markets (...), another
the development of barriers to the large movement of goods and labor that had been such a
prominent feature in the last years of the Ottoman Empire” (Owen and Pamuk 1999:51).
The problem gets even more complicated because the political reality of these empires often
did not match the symbolic discourses and rituals of imperial unity. For instance, in the case of the
Ottoman Empire, the Egyptian regime of Muhammad Ali (reg. 1805-48) was “virtually
autonomous” (Lewis 1995:301). Tunis and Morocco also enjoyed a certain degree of autonomy.
It was only in these areas that the option of “indirect rule” through traditional institutions and
dynasties was available (and therefore level 2 or level 3-codings). In others, colonialism not only
created new political entities, but also founded completely new “dynasties.” These cases were
coded as direct rule (level 4), as was the creation of chieftaincies without historical precursors.
Eighteen countries of 83 were coded level 2 (22%), 27 level 3 (33%) and 31 as level 4 (37%). This
means that more than two thirds of the countries in our sample experienced strong colonial political
intervention, with or without the dissolution of pre-colonial political institutions. As will be
discussed later, the intensity of political domination correlates with level of colonial violence as
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
162
well as with some indicators of economic transformation. British colonies were in general ruled
less directly (annex 1).5
Level of Colonial Violence (VIOLTOT)
Colonization was confronted with a broad variety of resistance, triggering often, but not
exclusively, violent reactions by the colonial power. We must distinguish between defensive wars
fought against the incoming colonialists and later protracted rebellions. In many instances the
initial war was fought with the traditional military system (cf. Ranger 1969:312). This primary
resistance was motivated by the goal of preserving autonomy and sovereignty and was often led
by traditional elites. However, in this context we are less interested in the character or amount of
resistance than the varying impact of colonial violence against these resistance movements,
measured with our variable (VIOLCOL).
Without denying possible links between different resistance movements, we divided
subsequent forms of resistance into two forms: resistance against colonial domination and
exploitation (VIOLRES)—sometimes referred to as “secondary resistance,”—and violence during
decolonization (VIOLIND). VIOLRES is an indicator for the resistance against measures of
colonial domination such as taxation, expropriation, forced cultivation, forced recruitment or
forced labor, which was—even when seen economically ineffective—only slowly and unevenly
phased out.6 This form of resistance is often carried out by peasants, workers and/or women or,
quite common in sub-Saharan Africa, messianic movements. It is often labeled as “rebellion,”
“insurrection,” “revolt,” “uprising” or “mutiny.” It has a time lag to the colonial conquest, although
the distinction is not always clear cut: What for some groups is primary resistance can for other
groups be resistance against domination, and a variety of motives may converge into one anti-
colonial movement. In cases were the centers of resistance were under colonial rule for some time,
such as in the case of the famous “mutiny” in British India (1857/58), we coded VIOLRES,
otherwise VIOLCOL.
The third form of violence (VIOLIND) occurs during decolonization. Having a varying
degree of rural and urban mass participation, decolonization movements tend to be led by younger
5 There is one article with a comparable aim at coding different levels of political domination in the colonial era (Lange
2004). However, Lange measures the extent to which British colonial rule depended on customary legal institutions
for the regulation of social relations, by dividing the number of colonially recognized customary court cases by the
total number of court cases in 1955. This is an indicator very different from our coding, and our samples overlap only
in 22 cases (British colonies). There is no statistically significant correlation between his indicator and DOMFORM.
On the other hand, the correlation between his later estimates of the “level of colonialism” (Lange, Mahoney, and vom
Hau 2006:1414), defined as “the extent to which a colonizing power installs economic, political, and sociocultural
institutions in a colonized territory“, and DOMFORM is highly significant (0.71, N= 23 British colonies).
6 While it was in the French colonies abolished by law in 1945, it lasted even longer in the Portuguese empire (Austin
2015:10).
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
163
educated elites. Some movements took up armed struggle for independence, while others relied on
other means as strikes and demonstrations. Yet others opted for peaceful resistance. We coded for
the three variables different levels of colonial violence: no colonial violence (level 0), moderate
colonial violence (level 1) and massive colonial violence (level 2). For VIOLCOL and VIOLRES,
moderate colonial violence is characterized by small numbers of causalities among the local
population, limited length of resistance and limited economic effects for the local population. In
contrast, massive colonial violence caused heavy casualties and persisted for a considerable time.
It is often characterized by the purposeful destruction of the sources of livelihood (crops, domestic
animals, and/or infrastructure), resulting in years of hunger and misery for a significant part of the
population. For violence during decolonization (VIOLIND), the armed suppression of isolated acts
of violence and/or of riots, strikes or demonstrations were coded as moderate colonial violence,
while armed wars of independence were coded as massive colonial violence. Peaceful resistance
is thus not measured, neither is the psychological impact of colonial violence which is emphasized
by Ranger (1969), but not measurable in a comparative perspective. As we are interested in the
impact of colonial violence, we are neither considering the casualties and expenditures among the
colonial troops nor wars among colonial powers, as during World Wars I or II.
The total level of colonial violence (VIOLTOT) was measured by adding the values of the
three violence-related variables and can thus vary between 0 and 6. Examples for countries with
maximal violence are Algeria, Angola or Vietnam. Fiji, Jordan or Kuwait were not exposed to
significant direct, open colonial violence. Most countries are in between. The total level of colonial
violence correlates significantly and positively with the form of political domination (table of
relations between political domination and violence available here). It also correlates in general
positively with the length and intensity of political domination as well as with some indicators of
economic transformation (annex 2). However, violence during decolonization seems not to be
influenced by directly political factors, but it correlates with the existence of a plantation economy.
As is documented in annex 1, the level of violence was in general lower in British colonies than in
others and the onset of colonialism was usually more violent in sub-Saharan Africa compared to
North Africa and Asia.
Colonial Instrumentalization of Ethnolinguistic/Religious Cleavages (ETHNFUNC)
Colonial powers instrumentalized ethnolinguistic and/or religious cleavages mainly in three areas:
army/police, administration/education, and in the economy. In all colonies, the upper layers of the
administration were staffed by officials from the metropole. The colonial powers however used
ethnolinguistic and religious cleavages to different degrees to occupy the lower layers and special
functions (“martial races,” administrators etc.). As important as this aspect of colonial domination
is, it is difficult to measure its impact. Going beyond a simple differentiation between a
http://www.worlddevelopment.uzh.ch/research/upl/Relations_between_Political_Domination_and_Violence.pdf
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
164
low/moderate impact (level 1) and a significant/strong impact (level 2) would be hazardous. For
more than one third of our cases, we found indications that there was a strong impact of colonial
policy on occupational specializations along ethnolinguistic and/or religious lines. For another 15
cases, there was a low/moderate impact.
Related to this variable is the coding of colonially induced work immigration (for the army,
infrastructure project, plantations etc.) because the colonial policy in this regard followed similar
considerations. The difference is that with the variable ETHNFUNC we measure the degree to
which colonial policies intentionally as well as unintentionally created and/or reinforced such
occupational specializations along ethnolinguistic and/or religious lines while WORKIM mainly
catches the quantitative aspect of colonially induced immigration.
As significant bivariate correlations show, the instrumentalization of ethnolinguistic and/or
religious cleavages usually took place in more directly ruled colonies, in colonies with a higher
dependence on investment from the metropole, with more successful missionary activities and,
unsurprisingly, in colonies with a higher level of ethnic heterogeneity and work immigration
(annex 2). It was more often applied in sub-Saharan Africa than in Asian and North African
countries (annex 1). This might be related with the lower levels of ethnic heterogeneity in the latter
countries.
Gradualism in the Transfer of Administration (INDTRANS)
In September 1958, Guinea was the only French territory in West Africa overwhelmingly rejecting
the French constitutional project and opting for immediate independence: “France was quick to
retaliate, withdrawing civil servants, severing aid, trade, and development agreements, absconding
with government files, and sabotaging whatever the departing officials could not take with them”
(Schmidt 2007:178). This discontinuity in the administrative institutions was a heavy burden for
the newly independent Guinea. A similar institutional discontinuity has in other cases been caused
by massacres targeting expatriates. In Mozambique or the Belgian Congo, most European civil
servants fled or were chased away immediately before or after independence. In contrast, there are
countries with an extreme gradual transfer of administrative authority to local civil servants
accompanied by training and meritocratic promotion, resulting in considerable state capacity. In
Botswana, efficiency of administration was clearly put before Africanization (Künzler 2004:162-
163). Just after independence in Malaysia still 44.5% of civil servants were British, in Pakistan
34.2%, in Ceylon (Sri Lanka) 18.9%, in India 7.3%, while in Burma they all left with independence
(Braibanti 1966). According to Kerbo (2005a, 2005b), the way the colonial power left the country
(such as the extent of planning and preparation for independence) was indeed crucial for the further
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
165
development in East and Southeast Asia.7 While our main concern is the transfer of bureaucracy,
we also considered the transfer of the judicial system and the security forces, but not the transfer
of legislative.
The transfer of administration is an important aspect for the development of post-colonial
institutions that has been somewhat neglected so far. We therefore introduce a variable
(INDTRANS) measuring the gradualism in the transfer of administration:
0= not applicable/planned, coordinated and very gradual transfer of administration (52
cases)
1= short initiation, no conflicts (16 cases)
2= very quick, disruptive, disorderly transfer (15 cases).
Of course, a planned and coordinated transfer of administration is not possible in the case of armed
conflicts/war of independence. INDTRANS correlates highly with indicators for the level of
violence, but also a more direct rule seems to have impeded a gradual transfer of administration
(annex 2). Of all economic indicators, only the trade-related ones (TRADEPOL, TRADECON)
have a significant relationship with INDTRANS: The closer the colonial economy was related to
the metropole, the less likely a gradual transfer (annex 2). British colonies experienced a gradual
and orderly transfer of administration more often than others (annex 1).
Indicators for Economic Transformation by Colonialism
Political domination did not lead automatically to economic changes. To make profit beyond
occasional plundering and traditional ways of taxation, a mise en valeur (capitalizing on the
colony/valorization) was necessary, which proved to be difficult in many cases. Because of their
ubiquity and the variety of forms applied, we were not able to code the main aspects of economic
colonialism: the degree of expropriation (especially of land), and of financial exploitation. Only
recently attempts have been undertaken to compare colonial taxation systems based on historical
records (Frankema and Waijenburg 2014). We consider this a major gap in our coding of the
impacts of colonialism. However, we measure these effects indirectly: The colonial population
had to pay for higher costs if there were more wars, more violence, more investment in
7 Kerbo (2005b:147): “It is particularly important within East and Southeast Asia to compare how each country was
treated by the colonial powers and how these colonial powers left the country. The end of British colonialism, for
instance, was at first comparatively tame for both Malaysia and Burma. In Malaysia, the British stayed on for several
years to provide advice on running the newly independent nation and to fight communist insurgents. Burma, however,
became mired in conflict after the British abruptly left the country. Primarily because of internal conflicts the British
helped rekindle as they left the country, Burma immediately degenerated into political violence and remains in
suspension with a military dictatorship repressing any needed reforms.”
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
166
infrastructure or more trade distortions. We focus on the seven indicators mentioned in table 1 to
measure the colonial impact on the economy.
Trade Policy (TRADEPOL)
Colonies were seen as a way of ensuring markets. Trade flows have always been closely monitored
by colonial governments and trade policies were changed frequently in order influence them. In
the Great Depression of the 1930s, all colonial empires tried to introduce protectionist measures.
Therefore, it is difficult to code colonial trade policies 1850-1950 beyond a simple “general
opening” vs. “more distortive.” We followed the definitions of Mitchener and Weidenmier (2008),
but made no distinction between “preferential tariff policy” and “tariff assimilation/customs
union”:8
0= not applicable (5 cases in which no colonial influence in trade policy was discernible)
1= predominantly ‘open door’ policy (43 cases)
2= preferential tariff policy, policy of tariff assimilation/customs union with metropole
country (35 cases, of which 21 were French colonies, 3 Portuguese and 3 Japanese)
French colonies more often experienced a distortive trade policy than British ones (annex 1). Some
other relations with colonial indicators become evident (annex 2): The more direct colonial rule
(DOMFORM), the more distortive the colonial trade policy, and the more likely anti-colonial
resistance (VIOLRES). Beyond that, trade policy (TRADEPOL) also correlates significantly with
foreign presence (FORPRES), nature of authority transfer (INDTRANS) and trade concentration
(TRADECON). Mitchener and Weidenmier (2008:2) emphasize for the period 1870-1913: “Being
in an empire roughly doubled trade relative to those countries that were not part of an empire.”
However, trade policies were not always effective, and these “colonial ties” have other causes than
trade policies as well. Therefore, an indicator for the factual impact of colonial domination on
trade flows is needed.
8 Mitchener and Weidenmier (2008:18-19): “The policy of tariff assimilation is a policy regime where the tariff rates
on goods are the same in the metropole and the colony. Under this arrangement, the metropole and colony form a
customs union. A preferential tariff system describes a trade policy where colonies and the mother country have
differential tariffs, but non-empire goods are generally taxed at a higher rate. An ‘open door’ trade policy refers to a
tariff regime where there is no distinction made between the products of the mother country and non -empire trading
partners. In other words, a colony or metropole with an ‘open door’ trade policy does have not a preferential tariff
policy or trade agreement (i.e., customs union) with some of its trading partners. The open door trade policy should
not be confused with a free trade policy, however. Many countries with ‘open door’ policies levied duties to protect
local industries or to raise revenue for the fiscal authority.”
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
167
Trade Concentration (TRADECON)
Trade concentration measures the colony’s trade with the metropole as proportion of total trade.
The main source for this indicator is the International Historical Statistics Data Set (Mitchell 2007).
Additional sources were the United Nations Commodity Trade Statistics (COMTRADE) Database
(https://comtrade.un.org/) and, in very few cases, the literature on colonial/national histories. Due
to the poor data quality for many colonies, we decided to classify the cases in five categories:
0 = trade concentration under 25%, no colonial trade effect is discernible, three quarter of goods
exchanged with third countries/areas:
1 = trade concentration 25-40%, a moderate trade effect
2 = trade concentration 40-50%, a significant trade effect
3 = trade concentration 50-66.5%, a strong trade effect
4 = trade concentration 66.6-%, a very strong trade effect
Unsurprisingly, for the semi-colonial countries (except Persia) no colonial trade effect is
discernible; these countries were usually opened for trade under ‘most favored nation’ clauses.
Also, for many colonies with an ‘open door’ policy, trade with the metropole was not particularly
strong. Together, 22 cases (27%) of our sample had no colonial trade effect (level 0), measured by
a trade concentration of less than 25%. For the same number of cases we find a moderate trade
effect (level 1) with a trade concentration of 25-40%. In 12 countries, trade concentration was 40-
50%, representing a significant trade effect (level 2). A strong trade effect (level 3) with a trade
concentration of 50-66.5% was experienced by 18 cases, and another 9 experienced a very strong
trade effect (level 4), where more than two thirds of the trade went to the metropole. Among the
27 countries with a strong or very strong trade effect (one third of our sample), there are five British
colonies, 14 French, two Italian as well as the Japanese colonies (Korea, Taiwan).
Again, French colonies experienced significantly more often a distortive trade effect than
British ones (annex 1). In general, a high level of trade concentration was to be found in colonies
with a high level of colonial investment and a disorderly transfer of administration (annex 2). There
is a statistically significant relation between the intensity of political domination and trade
concentration (table of relations between political domination and trade concentration available
here).
Investment Concentration (FDICON)
Similar to trade concentration, we categorize colonies regarding their dependence on direct
investment from the metropole. The data situation, however, is very poor for the colonial period
itself. As a proxy, we relied on the data provided by the OECD (1972), again complemented by
https://comtrade.un.org/
http://www.worlddevelopment.uzh.ch/research/upl/Political_Domination_and_Trade_Concentration.pdf
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
168
sources such as colonial/national histories and estimates. As in the case of trade concentration, we
classified the cases in five categories:
0 = investment concentration under 25%, no colonial investment effect discernible, three
quarter of investment from third countries/areas
1 = investment concentration 25-40%, a moderate investment effect
2 = investment concentration 40-50%, a significant investment effect
3 = investment concentration 50-66.5%, a strong investment effect
4 = investment concentration 66.6-%, a very strong investment effect
In 12 cases (15%), we found no colonial investment effect (level 0), in 8 cases a moderate
investment effect (level 1), and in 9 cases a significant investment effect (level 2). Thus, in two
thirds of the cases, more than 50% of investment came from the metropole: 10 had a strong
investment effect (level 3) and 44 a very strong investment effect (level 4).
It seems that colonial domination strongly relates to investment flows. Countries in sub-
Saharan Africa are more likely to have a high direct investment concentration than the ones in Asia
and North Africa (annex 1). As annex 2 shows, there are significant correlations two indicators of
political transformation (DOMFORM, ETHNFUNC), with several other indicators of economic
transformation (TRADECON, GOLD, INVEST, PLANTAT) and two indicators of social
transformation discussed below (FORPRES, MISSION).
Investment in Infrastructure (INVEST)
The development of the infrastructure is both a prerequisite for and consequence of the mise en
valeur of a colony. Harbors were crucial, as well as the transport links (railways, streets) into the
interior, followed by local transport systems such as tramways, telecommunications, and gas,
electric, and water works. Because of the different situations and needs for infrastructure, we
decided not to refer to specific historical data (railway miles etc.) but to code more general colonial
investment in infrastructure as “not significant/modest/huge:”
0 = no significant colonial investment in infrastructure / not applicable
1 = moderate colonial investment in infrastructure (e.g. some streets and bridges, not more
than one railway line to the coast, only isolated irrigation measures)
2 = huge colonial investment in infrastructure (e.g. a railway net and streets linking major
parts of the colony, sea harbor(s), and/or channels/irrigation measures in a significant
part of the colony)
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
169
The means of financing these investments (public/private) does not matter here. Investment in
infrastructure by non-colonial governments/agencies, e.g. the Ottoman railway system built before
WWI, are not considered. Mitchell (2007) has been an important source for the coding of
infrastructure, controlled by colony-specific sources and complemented for countries/areas not
mentioned.
Apart from the semi-colonial countries and Mongolia, we found no significant colonial
investment in infrastructure (level 0) in Bhutan, Nepal, three British colonies at the Persian Gulf,
four French controlled areas in Africa as well as in Portuguese Guinea-Bissau and Italian Somalia.
We coded a moderate colonial investment in infrastructure (level 1) were e.g. some streets and
bridges were built, but not more than one railway line to the coast and only isolated irrigation
measures. This is the case for one third of our sample (27 colonies), mainly landlocked African
countries, areas in Western Asia that were colonially dominated only for a short period such as
Syria and Lebanon as well as some Pacific island colonies. Finally, 39 countries (47%)
experienced a huge colonial investment in infrastructure (level 2), measured e.g. by a railway net
and streets linking major parts of the colony, sea harbor(s), and/or channels/irrigation measures in
a significant part of the colony.
As expected, this indicator for the level of economic transformation correlates with many
others, such as plantations, gold, mining, trade policy, trade concentration, investment concent-
ration but also with foreign presence (FORPRES) (see annex 2). Also not surprising is the fact that
colonial investment was higher in more directly ruled colonies (DOMFORM) that were held for a
longer period of time (COLYEARS). Investment in infrastructure also seems to have triggered
more anti-colonial resistance (VIOLRES), probably because it required a higher level of resource
mobilization out of the colony. Regarding the level of colonial investment in infrastructure, there
are no statistically significant differences between sub-Saharan Africa and Asian and North
African countries and French and British colonies (annex 1).
Plantations (PLANTAT)
It is common to distinguish between smallholder farms and plantations. According to
Osterhammel:
Two diametrically opposed forms of entreprise, the farm household and the
plantation were and still are the most efficient bases of the agrarian export
economy. The farm household cultivates its own and/or rented land with family
members and perhaps a small number of paid workers. The plantation is a large-
scale enterprise, often remotely situated. Its construction requires substantial capital
investments in land, machines, and plants and is kept in operation by often
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
170
incompetent wage laborers under the direction of a foreign management.
Plantations are often owned by foreign stock corporations, which also manages the
processing and marketing of the products. (Osterhammel 1997:76)
We followed this definitions, but coded not only the typical large-scale, foreign owned rubber,
sugar, coffee etc. plantations as such, but also the large cattle, tobacco, corn etc. farms in Africa
(including the French vineyards in North Africa). These were usually owned by permanent resident
families (of European descent), produced for export and for local/regional markets, and employed
many workers, but in smaller proportion to the total expenditure (Osterhammel 1997:77-78).
Special cases are concessions allowing private companies the exploitation of wild-growing
resources as timber or wild rubber. In both the French and the Belgian Congo, these companies
used forced labor to collect rubber, and the impact of this system was similar to the impact of
large-scale plantations. However, the central buying of wild-growing resources collected by
individual gatherers has a different impact and is thus not considered as a form of plantation.
Historical data sets provide data for most important cash crops including plantation products
but do not allow us to quantify the significance of the plantation economy in the countries of our
sample. In almost all colonies, setting up some kinds of plantations has been tried, often
unsuccessfully. Therefore, we code the presence of some small, isolated plantations as zero.
Beyond that, we again apply a very simple distinction between countries with plantation
economies of moderate and those with significant size. The latter does not necessarily mean that
they dominate the economy, and in many countries a significant plantation economy co-existed
with the smallholder agriculture.
0 = no or a some small, isolated plantations / not applicable
1 = plantation economy of moderate size (regarding areas occupied and share of total
export)
2 = plantation economy of significant size (regarding areas occupied and share of total
export)
We identify the following regions/countries as having a significant legacy of a colonial plantation
economy:
• in sub-Saharan Africa: South Africa, Swaziland, Zimbabwe, Equatorial Guinea,
Liberia, Mozambique, Angola, Kenya, Cameroon, Zaire;
• in North Africa: Morocco, Algeria, Tunisia, Egypt;
• in South Asia: Sri Lanka;
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
171
• in Southeast Asia: Malaysia, Indonesia, Vietnam, Cambodia, Philippines;
• in Oceania: Fiji, Papua New Guinea, Solomon Islands, Vanuatu.
These are 24 cases out of 83 (29%); 19 cases (23%) had a plantation economy of moderate size.
A plantation economy usually developed in areas that were colonially dominated for a long time,
in which significant investment in infrastructure was done and in which mining took place (annex
2). Furthermore, in plantations economies investment concentration was high, there were work
immigration and missionary activities and decolonization was usually a difficult and violent
process. There are no statistically significant differences between the regions (sub-Saharan Africa
vs. Asia/North Africa) and British/French colonies (annex 1).
Extraction of Natural Resources (GOLD, MINING)
Gold and diamonds have a high value in proportion to their weight and are thus easily exported
(or looted) without special infrastructure. The exploitation of oil, coal or metals as copper, iron,
tin, zinc and others need more investment in infrastructure, especially when going beyond small-
scale traditional techniques. We coded these natural resources thus in one variable (MINING) and
gold, diamonds and silver in another (GOLD). For both variables we used trichotomies.
0= not applicable
1= moderate colonial extraction
2= extensive colonial extraction.
We considered an extraction of large quantities over a lengthy period of time as “extensive colonial
extraction,” while an extraction of medium or low quantities over a lengthy period or a large
quantities over a short period (for example a short “gold rush”) was considered as “moderate
colonial extraction.” The UN Statistical Yearbooks and Mitchell (2007) proved to be valuable but
not exhaustive data sources. Fifteen countries in our sample scored high on both variables, while
for 25 countries there was no evidence of colonial extraction at all. These cases have been carefully
checked, like Syria, where petroleum was discovered after independence and no colonial
involvement can be found. Botswana is among the countries where the colonial extraction was
moderate. Again, an important commodity (diamonds) was discovered only after independence.
During the colonial period, there has been only a short gold boom around Tati and some small
mining (Künzler 2004:179).
Unsurprisingly, GOLD and MINING are related to a high level of colonial investment into
infrastructure (annex 2). Extraction took usually place in colonies where also a plantation economy
existed. MINING (but not GOLD) signified usually a higher presence of foreigners from the
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
172
metropole. The extraction of bullion (but not MINING) correlates with a higher level of violence
during colonial domination (VIOLRES) and more missionary activities (MISSION). There are no
statistically significant differences between the regions (sub-Saharan Africa, Asia, North Africa)
and British/French colonies (annex 1).
Indicators for Social Transformation by Colonialism
Regarding the social dimension of the colonial transformation of societies, we defined four
processes as significant and operationalizable, the immigration of people from the colonizing
country (FORPRE) and other colonies or areas for labor market reasons (WORKIM), the extent
of successful missionary activities (MISSION), and the colonial drawing up of borders
(BORDERS). Due to lack of comparable and consistent data, we could not define indicators for
the colonial investment in health and education systems (e.g. colonial government expenditure
data, or medical doctors per population). Life expectancy and literacy cannot be taken as indicators
for colonial impact because they are determined by several other factors as well (cf. Ziltener and
Mueller 2007). Also for more general effects of colonial domination, such as alienation (Fanon
1963; Césaire 2000) we could not find appropriate indicators measuring different levels of impact.
Colonial Immigration (FORPRE)
Colonies can be distinguished regarding the presence of people from the colonizing country. While
all colonies had officials, clerks, administrators, doctors, traders, and some military and religious
personnel, they differed significantly in regard to settlers. Undoubtedly, a larger presence of
colonial population has a stronger social transformation effect than a smaller one. Heldring and
Robinson (2012:6) argue that in two sorts of colonies there is “a clear case to be made for
colonialism retarding development,” one of them being those with “white settlement.” Lange,
Mahoney, and vom Hau (2006:1426) even use the size of the European population as statistical
proxy for the “level of colonialism.”
As many others, we defined the situation at the end of the colonial period as crucial, except
in those cases where a big proportion of settlers had left some years earlier (usually due to wars,
such as in Libya, where WWII drove most Italian immigrants home before independence in 1951).
Colonial immigration, not citizenship, is decisive. To measure the presence of Europeans
(Americans, Japanese) in the years before independence, we consulted the UN Statistical
Yearbooks (various years), the Demographic Survey of the British Colonial Empire (Kuczynski
1977), the Dictionnaire de la colonisation française (Liauzu 2007), Encyclopedia of Western
Colonialism since 1450 (Benjamin 2007), the Handbuch der Dritten Welt (eight vol., ed. by
Nohlen and Nuscheler 1992ff), and Maddison (2007:137) and we added colony-specific sources.
Following the construction of other variables, we transformed the data into a pentatomy:
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
173
0 = colonial population under 1%
1 = colonial population 1-2%
2 = colonial population 2-5%
3 = colonial population 5-10%
4 = colonial population over 10%.
Only three cases of our sample had a colonial population over 10%, namely Algeria, Libya and
South Africa. Also, all four level 3-countries are in Africa: Djibouti, Tunisia, Morocco and Angola.
At level 2, we find among 11 cases (13%) also non-African colonies: Fiji, Vanuatu and the
Japanese colonies Korea and Taiwan. Six cases had between 1-2% colonial population (Senegal,
Equatorial Guinea, Republic of Congo, Mozambique, Egypt and Papua New Guinea), which
makes the overwhelming majority of the cases in our sample (58, 70%) level 1-countries, with a
very low number of immigrants from the colonizing country in relation to the dominated
population.
Colonies with a higher level of investment in infrastructure, plantations and mining activities
(INVEST, PLANT, MINING) as well as more protectionist trade policy (TRADEPOL) had more
immigration from the metropole than from other countries (annex 2). There are no statistically
significant differences between sub-Saharan Africa and Asian/North African countries and
French/British colonies (annex 1).
Colonial Mission (MISSION)
Because missionaries and missionary organizations had to report regularly on their activities, the
data situation for this indicator of social transformation is better than for others. However, there
are the recurring problems of the areas reported not corresponding to actual state boundaries.
Furthermore, the number of converts reported by different churches is sometimes exaggerated. In
certain cases the numbers combined exceed the (estimated) total population of the area. Most data
come from the World Christian Handbook (Grubb and Bingle 1949, 1952), the World Christian
Encyclopedia (Barrett, Kurian, and Johnson 2001), the Dictionnaire de la colonisation française
(Liauzu 2007), Encyclopedia of Western Colonialism since 1450 (Benjamin 2007), and the
Handbuch der Dritten Welt (eight vol., ed. by Nohlen and Nuscheler 1992ff) and were checked by
colony-specific sources and complemented for countries/areas not mentioned. The data (number
of Christians per population) have to be controlled for older religious groups which have no
relation with colonial missionary activities, e.g. the Copts in Egypt and Ethiopia, for the number
of settlers (e.g., almost all of the 100,000 Catholics in Libya before WWII were immigrants from
Italy) and for Christian work immigrants (e.g. in some oil-producing countries).
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
174
In general, missionaries of any Christian church were protected by the colonial government.
In some areas, a protestant colonizing power followed a Catholic one (e.g., in Sri Lanka and parts
of the Southeast Asian archipelago). Therefore, we coded all converts to Christianity as “impact
of colonialism.” Again, we defined a year before independence as reference, because in some cases
decolonization led to emigration of Christians. Therefore, we coded “percentage of Christians
(Catholics, Protestants), years before independence.” Following the construction of other
variables, we transformed the data into a pentatomy:
0 = no colonial missionary activities / not applicable
1 = colonial missionary activities with little effect (Christian population under 2%)
2 = colonial missionary activities with significant effect (Christian population 2-7%)
3 = colonial missionary activities with big effect (Christian population 7-50%)
4 = majority converted (Christian population over 50%)
Obviously, Japanese colonies (Korea, Taiwan, Manchuria) have to be coded differently. There
were Christians under Japanese rule, in Taiwan mainly proselytized by American Presbyterians,
while in Korea a process of “self-proselytization” had taken place in parts of the élite since late
18th century. In the 19th century, before the onset of Japanese colonialism, there were around
20,000 Christian converts in Korea (Kim 2005:100-103; Seth 2006:220-221). Under Japanese rule,
coercive measures to assimilate (culturally at least) parts of the population were taken, including
the participation at state-Shinto rituals. However, these had little effect on everyday religious
practices. Therefore, Korea and Taiwan were coded as level 1-countries.
Thirty-seven countries in our sample (45%) did not experience a significant religious
transformation (either with or without missionary activities; level 1 or 0), among them all Islamic
North Africa and West/South Asia as well as the semi-colonial countries. There are 12 level 2-
countries (15%), among them India, several Southeast Asian countries (Indonesia, Malaysia,
Burma/Myanmar, Vietnam, Cambodia), five African countries (Burkina Faso, Ivory Coast, Sierra
Leone, Senegal, Liberia) as well as Lebanon. Among the 25 level 3-countries (30%) are only two
non-African countries, namely Sri Lanka and Papua New Guinea. Among the nine countries with
the strongest impact of missionary activities (conversion of more than 50% of the population) are
the Philippines, three Pacific Islands (Vanuatu, Fiji and the Solomons) and five African countries
(Lesotho, Equatorial Guinea, Zambia, Swaziland, South Africa).
Consequently, there is a statistically significant, positive relationship between MISSION and
sub-Saharan Africa respectively a negative one for Asia and North Africa (annex 1). As
documented in annex 2, the religious impact of colonialism was stronger in areas that were longer
under colonial rule (COLYEARS), that were economically more transformed (positive
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
175
correlations with FDICON, INVEST, PLANT, GOLD) and that experienced colonial measures
related to ethnic heterogeneity (ETHNFUNC, WORKIM, BORDERS).
Labor Immigration (WORKIM)
Beyond some special cases (such as Malaya and Southern Africa), there are no systematic data for
colonially induced labor immigration. We had to make rather crude estimations based on colony-
specific sources:
0 = no indicators for colonially induced labor immigration / not applicable
1 = low level of colonially induced labor immigration
2 = high level of colonially induced labor immigration (high figures, big groups; also cases
in which some functional groups completely consisted of foreign laborers).
“Colonially induced” means that work immigration has not only been tolerated but encouraged,
independently of permanent settlement in the colony. Work immigration has been encouraged by
means such as opening markets, recruitment/deployment from abroad, providing permits/licenses
for agencies, etc. Therefore, the migration of individual Lebanese traders to Africa or Chinese
traders to Southeast Asia were not “colonially induced,” but the immigration of Chinese and Indian
laborers into the tin mining areas of British-Malaya were.
For 42 cases (51%) in our sample we did not find any evidence for colonially induced labor
immigration. In 20 colonies, some work immigration took place; 15 of these cases are in Africa,
two at the Persian Gulf, three in Southern Asia (Indonesia, India, Vietnam). Areas with a high
influx of work immigrants (21 cases) typically had strong mining activities and/or numerous
plantations, in some cases, colonial governments deemed the local population as not willing or
capable of certain economic activities. Among these are Cambodia, Laos (French-Indochina); the
British colonies of Malaysia, Sri Lanka and Myanmar; Fiji and Vanuatu in the Pacific; the oil
producing Qatar, Bahrain and Kuwait; in Africa 11 cases linked to colonial administration or
infrastructure construction, or to plantation or mining work (Ivory Coast, Cameroon, Congo Rep,
Ghana, Guinea-Bissau, Equatorial Guinea, Kenya, Senegal, Uganda, South Africa, Zimbabwe).
As expected, colonially induced labor immigration took place in colonies that were longer
under colonial rule (COLYEARS) and that had a significant plantation economy, a high degree of
trade concentration and a strong impact of missionary activities (annex 2). It also relates to colonial
instrumentalization of ethnolinguistic and/or religious cleavages (ETHNFUNC). There is no
statistically significant correlation with the colonizing country (British vs. French) or the region
(sub-Saharan Africa vs. Asia/North Africa) (annex 1).
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
176
Unification and Partition (BORDERS)
The present situation of ethnic, linguistic and religious heterogeneity in African and Asian
countries is a product of coercive division and/or unification, in which colonialism played a major
part. The importance of ethnolinguistic fragmentation for the explanation of post-independence
economic growth rates is established by the work of Easterly and Levine (1997) using the index
of ethnolinguistic fractionalization popularized by Mauro (1995). Englebert, Tarango, and Carter
(2002:1099) measure for an African sample the suffocation by borders, or in other words “their
propensity for bringing together peoples that historically lived under different, if not inimical,
systems.” This measure is strongly related to group grievances and the likelihood of civil war. The
forced unification is, according to Oliver (1991), more formative than the division of Africa by
colonial borders. It is however already included in measures of ethnolinguistic fractionalization
usually used in multivariate studies.
The second aspect of artificial borders, the “partition of preexisting political groupings,” is
called “dismemberment” by Englebert, Tarango, and Carter (2002:193). They measure it as the
“unweighted average proportion of groups astride a border” and it “equals one half the sum of the
percentages of national populations that these groups represent on both sides of border.” 9 For a
sample of 48 African countries, dismemberment is positively associated with the likelihood and
intensity of boundary disputes. Alesina, Easterly, and Matuszeski (2006:8) go a long way to
measure the fractal dimension of borders based on the assumption that “borders which are coast
lines or squiggly lines (perhaps meant to capture geographic features and/or ethnicities) are less
likely to be artificial.” Among the “most artificial” states identified by these authors (Alesina,
Easterly, and Matuszeski 2006:13) there are countries like Chad, Mali, Niger or Sudan among
others. These countries lie in the Sahara-Sahel belt of Africa and indeed have some ruler-straight
borders. This is misleading, however, as these borders cut across areas thinly populated by mostly
nomadic groups. A border line could be as squiggly as possible but would still cut across these
groups. Furthermore, a border can be squiggly and follow a geographic feature as a river, but
nevertheless divide, as the same people live on both sides of the river.
The examples of the Bakongo on the lower Congo River or the Lao along the Mekong show
clearly that borders following a river are not automatically “natural” and less problematic than
other borders. It is therefore advisable to rely on empirical observations in measuring the partition
effect by borders. This is also the approach behind the second measure used by Alesina, Easterly,
and Matuszeski (2006:12), “the percent of the population of a country that belongs to a partitioned
group.” This measure is not correlated to the fractionalization measure, but is significant for the
explanation of a number of political and economic variables. As there are no consistent data for
9 http://www.politics.pomona.edu/penglebert/Borders%20appendix.doc
http://www.politics.pomona.edu/penglebert/Borders%20appendix.doc
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
177
our sample available, we followed this approach and measured the partition by colonial borders as
follows:
0= not applicable
1= low/moderate colonial partition effect
2= significant/strong colonial partition effect
For Africa, the Atlas of Precolonial Societies by Mueller et al. (1999) was a useful tool for the
coding. If more than 50% of the population belongs to groups partitioned by colonial borders the
impact is considered strong; otherwise it is moderate unless there was no dismemberment as in the
case of an island like Madagascar. In areas with strong traditions of state- and empire-building,
colonial borders could not be drawn as arbitrarily as in those without. Most political units could
be only dominated entirely or not at all. Colonial partition is therefore a rather rare experience in
these areas, except for the non-sedentary or semi-nomadic ethnic groups (e.g. the so-called
“mountain tribes” and “sea nomads”) which successively had to accept a new type of borderline
enforcement.
In 19 cases (23%) of our sample, we did not find evidence that colonial borders were
especially “artificially” drawn in the interests of colonialists, cutting traditional areas of settlement
or of close interaction. These cases are mainly semi-colonial countries and islands, or cases in
which the actual borders cannot be traced back to colonialism. Thirty-seven cases (45%) were
coded as 1, i.e. a moderate impact of colonialism regarding partition; mainly the states in the
Northern part of Africa as well as, Iran, Pakistan, Thailand, and the Philippines. In the clear
majority of cases (46; 55%), colonial borders did have a distortive impact. This concerns the Asian
part of the former Ottoman Empire (Iraq, Lebanon, Syria, Jordan), where—as mentioned above—
the creation of new political entities entailed the creation of new national economies and legal
systems (Owen and Pamuk 1999:51). In many cases, territories were arbitrarily detached,
sometimes “bartered” among colonial powers, integrated into existing colonies and reconfigured
(e.g. between British-Malaya and Dutch Indonesia or in French-Indochina with Cambodia and
Laos). Most cases of “artificial” borders, however, are found in sub-Saharan Africa (37 cases).
French colonies experienced significantly more often distortive borders than British ones, as did
countries in sub-Saharan Africa, compared to those in Asia and North Africa (annex 1). In areas
with “artificial” borders, occupational specializations along ethnolinguistic and/or religious lines
developed more often (ETHNFUNC) and missionary activities were more successful (annex 2).
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
178
The Construction of the Indices
In this section, the indicators for the three dimensions were combined into separate indices for
political transformation, economic transformation and social transformation. Many indicators have
different scales of measurement. Therefore, we standardized all indicators between 0 and 1, giving
them equal weighting in the process of constructing the indices. To do so, we used the following
formula:
xk is the value of one country for any of our indicators. xmin and xmax then denote the minimum and
maximum for all country observations for the same indicator, respectively. δ represents the
standardized values for each indicator. In the next step, we use the mean to combine the indicators
according to table 1 to get our three indices:
To make interpretation in quantitative studies easier in terms of one unit change/one percent
change, we standardized the indices between 0 and 100 using the first formula again:
Table 2 shows the resulting values for these three indices for the colonies in our sample. The three
indices correlate significantly:10
• the Political Transformation (PT)-Index with the ET-Index 0.50,
• the PT-Index with the Social Transformation (ST)-Index 0.44,
• the Economic Transformation (ET)-Index with the ST-Index 0.51.
This means that, in general, political domination came along with economic and social
transformation. However, the correlations are not that strong, which indicates that the three
dimensions should be measured separately. Colonies in sub-Saharan Africa were more likely to
experience a higher level of transformation, politically, economically and socially (correlations
with PT-Index 0.37, ET-Index 0.38, ST-Index 0.58), than the Asian and North African ones.11 The
smallest difference to be found is regarding political transformation. That means that, although
political domination was not much less intensive in Asia and North Africa, these economies and
10 All correlations mentioned in this section are significant at the 1%-level, 2-tailed (Pearson correlation coefficients).
11 Graphs (available here) ranking the countries of Africa (red) and Asia/North Africa (blue) according to their levels
of political, economic and social transformation show the different effects of colonialism.
http://www.worlddevelopment.uzh.ch/research/upl/Levels_of_colonially_induced_political_economic_and_social_transformation_in_Africa_and_Asia.pdf
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
179
societies were more difficult to transform through colonialism. There are no highly significant
differences between British and French colonies regarding the level of colonial transformation.
However, British colonies seem to have been transformed a little less than others, politically and
economically.12
Table 2. Indices of Colonial Transformation – Country Values
Code
World
Bank
Country Political
trans-
formation
Economic
trans-
formation
Social
trans-
formation
Colonial
trans-
formation
AFG Afghanistan 32 0 25 20
AGO Angola 88 79 63 82
ARE Unit Arab Emirates 29 14 13 20
BDI Burundi 85 54 44 65
BEN Benin 65 57 44 58
BFA Burkina Faso 38 57 38 48
BGD Bangladesh 82 43 6 46
BHR Bahrain 32 14 25 25
BTN Bhutan 24 0 25 17
BWA Botswana 24 50 44 47
CAF Centr. African Rep. 94 57 56 74
CHN China 53 29 6 26
CIV Ivory Coast 76 75 63 74
CMR Cameroon 56 82 69 72
COG Congo Republic 74 61 75 77
DJI Djibouti 71 46 50 59
DZA Algeria 71 86 44 71
EGY Egypt 74 50 25 53
ETH Ethiopia 76 79 13 53
FJI Fiji 26 89 63 64
GAB Gabon 53 61 56 68
GHA Ghana 76 71 69 77
GIN Guinea 47 68 31 54
GMB Gambia 65 36 56 56
GNB Guinea-Bissau 100 46 81 81
GNQ Equatorial Guinea 82 64 81 81
IDN Indonesia 91 82 50 79
IND India 94 79 25 70
IRN Iran 15 21 19 17
IRQ Iraq 59 29 31 42
JOR Jordan 71 14 31 41
JPN Japan 15 7 0 8
KEN Kenya 91 82 69 86
KHM Cambodia 79 54 63 70
12 Correlation British/PT-Index -0.28; British/ET-Index -0.27; British/CT-Index -0.26 (significant at 5%-level; N= 75,
without semi-colonial countries and shared colonies).
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
180
Table 2. Indices of Colonial Transformation (Continued)
KOR Korea Republic 59 86 19 58
KWT Kuwait 18 21 50 32
LAO Laos 79 29 56 58
LBN Lebanon 82 21 38 50
LBR Liberia 79 79 63 79
LBY Libya 59 32 25 41
LKA Sri Lanka 82 54 44 64
LSO Lesotho 35 25 50 47
MAR Morocco 44 54 38 47
MDG Malagasy Republic 65 75 56 62
MLI Mali 74 50 44 59
MMR Myanmar 94 64 38 69
MNG Mongolia 0 0 0 0
MOZ Mozambique 65 64 63 68
MRT Mauritania 68 43 19 46
MWI Malawi 53 39 56 53
MYS Malaysia 79 50 63 68
NER Niger 88 39 44 65
NGA Nigeria 47 64 56 60
NPL Nepal 29 4 0 14
OMN Oman 38 11 44 28
PAK Pakistan 88 43 19 53
PHL Philippines 53 100 38 68
PNG Papua New Guinea 32 71 38 51
PRK Korea Dem. Rep. 59 86 19 58
QAT Qatar 18 14 25 20
RWA Rwanda 79 39 44 58
SAU Saudi Arabia 9 0 0 3
SDN Sudan 88 54 31 61
SEN Senegal 71 64 69 70
SLB Solomon Islands 24 57 25 38
SLE Sierra Leone 62 75 38 62
SOM Somalia 85 32 38 57
SWZ Swaziland 24 64 63 57
SYR Syria 76 25 31 47
TCD Chad 74 39 69 57
TGO Togo 74 57 44 59
THA Thailand 15 14 13 12
TUN Tunisia 32 57 38 46
TUR Turkey 18 7 19 15
TWN Taiwan 65 57 44 64
TZA Tanzania 71 68 44 65
UGA Uganda 74 54 56 65
VNM Vietnam 88 71 25 65
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
181
Table 2. Indices of Colonial Transformation (Continued)
VUT Vanuatu 38 46 63 53
ZAF South Africa 94 86 100 100
ZAR Zaire (DR Congo) 68 82 44 69
ZMB Zambia 65 46 63 62
ZWE Zimbabwe 65 86 81 83
Source: Ziltener and Künzler Dataset (2008).
To measure the total impact of colonialism, the three indices were combined to our final result:
the index for the colonial transformation. Again, table 2 shows the value for this index for all
colonies in our sample. As the ranking of the countries of Africa (red) and Asia and North Africa
(blue) according to their total level of colonial transformation in graph 1 shows, they have been
transformed to very different degrees.
Figure 1. Total level of colonially-induced transformation in Africa and Asia
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
182
Discussion and Conclusion
We started by admitting that the problem of colonization as “intervening variable” remains
unsolved in recent research, including our own. The huge variety and diversity of colonial
experiences that we found mentioned in the research literature and that we can confirm on the
basis of our evaluation (Ziltener and Künzler 2013) is a challenge to all attempts at coding the
factual impact of colonialism and therefore its “legacies.”
In our operationalization, we defined indicators for the main dimensions of the colonial
impact, in the political sphere as well as in economy and society. Confronted with numerous data
problems—including the recurring major challenge that the areas reported do not correspond to
actual state boundaries—we decided to make sometimes crude estimates rather than exclude
important aspects of colonialism. Apart from (more or less exploitative) forms of colonial financial
organization, we were able to find or create data for all important aspects of colonialism for which
variance has been reported in the previous research. As result, we present, to our knowledge for
the first time, a multidimensional measure of the impact of colonialism. This opens up new avenues
for comparative research, qualitative as well as quantitative.
Some of our exploratory findings on the interrelation of the dimensions confirm the results
of research done by others, e.g. that in British colonies political domination was in general less
direct, less violent and the path to independence more orderly. Plantation colonies experienced
more investment in infrastructure and more violence during decolonization. The correlations
between indicators for economic distortion (trade policy, trade and FDI concentration) show that
the economic re-direction of some colonies towards a more exclusive exchange with the metropole
was an interdependent process. In general, a more intense political domination came along with a
higher level of economic transformation. If an area was transformed economically, however, a
social transformation was likely to take place too, but these processes should not be confounded.
Some findings relate to the research from a social-evolutionary perspective. In areas that were
politically united for the first time under colonialism, economic distortion and social
transformation were more profound. This might explain why colonialism in sub-Saharan Africa
has left a heavier burden than in most Asian countries. This, however, has to be tested with
multivariate methods elsewhere.
Our main result is that there is not one colonial experience. There is an enormous variance
regarding the levels of distortion that countries and regions in Africa and Asia experienced.
Violence and political domination did not lead necessarily to economic or social transformation.
Many Asian societies displayed a strong resilience against the colonial inroads, which might well
have to do with the fact that many of these have been part of the Eurasian world system for
centuries before European hegemony. Only in a few countries of our sample, namely parts of sub-
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
183
Saharan Africa and the Pacific region, European colonialism meant first-time integration into the
world system.
In many regards, European colonialism between 1860 and 1960 left much deeper marks than
traditional, usually tributary forms of domination. Where a long-term mise en valeur was to be
realized, infrastructure investment was huge, and labor migration from near and far significant.
Some of the countries in our sample truly are creations of colonialism. There is no question that
these “legacies” shape our world.
Overall, our data echoes the heightened interests in the study of colonialism among scholars
from various disciplines over the last two decades. More specifically, we consider our data to
advance existing research in three areas. First, the data cater to the dominant interest of scholars
to investigate the effects of colonialism on economic development and trade relations (cf.
Acemoglu, Johnson, and Robinson 2001; Grier 1999; Head, Mayer, and Ries 2010; Lange,
Mahoney, and vom Hau 2006; Sousa and Lochard 2012) but also on democratization (cf. Lange
2004; Olsson 2009), conflict and state building (cf. Blanton, Mason, and Athow 2001; Lange and
Dawson 2009, Wucherpfennig, Hunziker, and Cederman 2016), and policy outcomes (cf. Schmitt
2015). Second, recent research has also started to investigate the determinants of colonialism such
as the timing or the mode of domination (cf. Ertan, Fiszbein, and Putterman 2016; Gerring et. al.
2011). More generally, our data provide the first opportunity to study how geographical conditions
and pre-colonial societies affected different impacts of colonialism in a larger sample of countries.
Third, we argue that the data enable scholars to create meaningful typologies of colonialism for
their research questions, thereby overcoming the simple French vs. British dichotomy that has
guided most empirical research about colonialism since Crowder’s article about indirect rule in
1964.
About the Authors
Patrick Ziltener, born in St.Gallen/Switzerland, studied in Basel, Berlin (Free University) and
Zurich Sociology, History, and Economics; Ph.D. from University of Zurich, 1997; 2003-05
research project Regional Integration in East Asia; 2008-09 research project Statecraft in the non-
Western World: A Multifactorial Model to Explain Persisting Differentiation; since 2015 Assoc.
Prof. at the University of Zurich.
Daniel Künzler studied sociology, political science and media science in Zurich (Switzerland) and
received a Ph.D. from Zurich University for his study on statehood and development in sub-
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
184
Saharan Africa. He worked as a post doc researcher in Benin. Since 2009, he is working at the
University of Fribourg, where he is currently lecturer in sociology, social policy and social work.
His current publications are mainly on his research on social policy in Eastern Africa.
André Walter, born in Erfurt/Germany, studied in Jena and Zurich Sociology, Political Science,
Anthropology and Gender Studies. Research Projects on Welfare State Development in Federalist
States , the Introduction of Income
Disclosure Statement
Any conflicts of interest are reported in the acknowledge section of the article’s text. Otherwise,
author has indicated that she has no conflict of interests upon submission of the article to the
journal.
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
185
References
Acemoglu, Daron, Simon Johnson, and James A. Robinson. 2001. “The Colonial Origins of
Comparative Development: An Empirical Investigation.” American Economic Review
91(5):1369-1401.
Acemoglu, Daron, Simon Johnson, and James A. Robinson. 2002. “Reversal of Fortune:
Geography and Institutions in the Making of the Modern World Income Distribution.”
Quarterly Journal of Economics 117(4):1231-1294.
Alesina, Alberto, William Easterly, and Janina Matuszeski. 2006. Artificial States. National
Bureau of Economic Research (NBER) Working Paper No. 12328. Cambridge, MA: National
Bureau of Economic Research. Retrieved February 11, 2017 (www.nber.org/papers/w12328).
Auslin, Michael R. 2004. Negotiating with Imperialism. The Unequal Treaties and the Culture of
Japanese Diplomacy. Cambridge, MA: Harvard University Press.
Austin, Gareth. 2015. The Economics of Colonialism in Africa. The Oxford Handbook of Africa
and Economics. Vol. 1: Context and Concepts. Oxford Handbooks Online. Retr.Feb. 11, 2017
(http://www.oxfordhandbooks.com/view/10.1093/oxfordhb/9780199687114.001.0001/oxfor
dhb-9780199687114-e-4)..
Barrett, David B., George T. Kurian, and Todd M. Johnson. 2001. World Christian Encyclopedia:
A Comparative Survey of Churches and Religions. 2nd ed. New York, NY: Oxford University
Press.
Benjamin, Thomas, ed. 2007. Encyclopedia of Western Colonialism since 1450. Detroit, MI:
Thomson Gale.
Blanton, Robert, T. David Mason, and Brian Athow. 2001. “Colonial Style and Post-Colonial
Ethnic Conflict in Africa.” Journal of Peace Research 38(4):473-91.
Bornschier, Volker, Christopher Chase-Dunn, and Richard Rubinson. 1978. “Cross-National
Evidence of the Effects of Foreign Investment and Aid on Economic Growth and Inequality:
A Survey of Findings and a Reanalysis.“ American Journal of Sociology 84(3):651-683.
Boswell, Terry. 1989. “Colonial Empires and the Capitalist World-Economy: A Time Series
Analysis of Colonization, 1640-1960.” American Sociological Review 54(2):180-196.
Braibanti, Ralph. 1966. “Concluding Observations.“ Pp. 643-675 in Asian Bureaucratic Systems
Emergent from the British Imperial Tradition, edited by R. Braibanti. Durham, NC: Duke
University Press.
Cain, P.J., and Anthony G. Hopkins. 1993. British Imperialism: Innovation and Expansion 1688-
1914. London: Longman.
Césaire, Aimé. 2000. Discourse on Colonialism. New York, NY: Monthly Review Press.
Chase-Dunn, Christopher. 1975. “The Effects of International Economic Dependence on
Development and Inequality.” American Sociological Review 40(6):720-738.
http://www.oxfordhandbooks.com/view/10.1093/oxfordhb/9780199687114.001.0001/oxfordhb-9780199687114-e-4
http://www.oxfordhandbooks.com/view/10.1093/oxfordhb/9780199687114.001.0001/oxfordhb-9780199687114-e-4
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
186
Crowder, Michael. 1964. “Indirect Rule: French and British Style.” Africa 34(3):197-205.
Easterly, William and Ross Levine. 1997. “Africa’s growth tragedy: Policies and ethnic divisions.”
Quarterly Journal of Economics 112(4):1203-1250.
Englebert, Pierre, Stacy Tarango, and Matthew Carter. 2002. “Dismemberment and Suffocation.
A Contribution to the Debate on African Boundaries.” Comparative Political Studies
35(10):1093-1118.
Ertan, Arhan, Martin Fiszbein, and Louis Putterman. 2016: “Who Was Colonized and When? A
Cross-Country Analysis of Determinants.” European Economic Review 83:165–184.
Fanon, Frantz. 1963. The Wretched of the Earth. New York, NY: Grove Press.
Frankema, Ewout, and Marlous van Waijenburg. 2014. “Metropolitan Blueprints of Colonial
Taxation? Lessons from Fiscal Capacity Building in British and French Africa, c. 1880–
1940.” Journal of African History 55:371-400.
Furnivall, John S. 1941. Progress and Welfare in Southeast Asia. A Comparison of Colonial Policy
and Practice. New York, NY: Institute of Pacific Relations.
Galeano, Eduardo. 1973. Open Veins of Latin America: Five Centuries of the Pillage of a
Continent. New York, NY: Monthly Review Press.
Gerring, John, Daniel Ziblatt, Johan Van Gorp, and Julián Arévalo. 2011. “An Institutional Theory
of Direct and Indirect Rule.” World Politics 63(3):377-433.
Grier, Robin. 1999. “Colonial Legacies and Economic Growth.” Public Choice 98(3):317-335.
Grubb, Kenneth G, and Ernest J. Bingle, eds. 1949. World Christian Handbook. London: World
Dominion Press.
Grubb, Kenneth G, and Ernest J. Bingle, eds. 1952. World Christian Handbook. London: World
Dominion Press.
Head, Keith, Thierry Mayer, and John Ries. 2010. “The Erosion of Colonial Trade Linkages after
Independence.” Journal of International Economics 81(1):1-14.
Heldring, Leander, and James A. Robinson. 2012. Colonialism and Economic Development in
Africa. National Bureau of Economic Research (NBER) Working Paper No. 18566.
Cambridge, MA: National Bureau of Economic Research. Retrieved February 11, 2017
(www.nber.org/papers/w18566.pdf).
Herkenrath, Mark, and Volker Bornschier. 2004. “Transnational Corporations in World
Development: Still the Same Harmful Effects?” Journal of World-Systems Research 9(1):105-
139.
Kerbo, Harold. 2005a. “Foreign Investment and Disparities in Economic Development and
Poverty Reduction: A Comparative-Historical Analysis of the Buddhist Countries of
Southeast Asia.” International Journal of Comparative Sociology 46:425-459.
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
187
Kerbo, Harold. 2005b. World Poverty. Global Inequality and the Modern World System. Boston,
MA: McGraw-Hill.
Kim, Djun Kil. 2005. The History of Korea. Westport, CT/London: Greenwood Press.
Kuczynski, Robert R. 1977. Demographic Survey of the British Colonial Empire, Vol. 1-3.
Fairfeld, NJ: A.M. Kelley.
Künzler, Daniel. 2004. Wo die Elefanten tanzen, leidet das Gras. Staat und Entwicklung in Afrika.
Münster: Lit.
Liauzu, Claude, ed. 2007. Dictionnaire de la colonisation française. Paris: Larousse.
Lange, Matthew K. 2004. “British Colonial Legacies and Political Development.” World
Development 32(6):905-922.
Lange, Matthew, and Andrew Dawson. 2009. “Dividing and Ruling the World? A Statistical Test
of the Effects of Colonialism on Postcolonial Civil Violence.” Social Forces 88(2):785-817.
Lange, Matthew, James Mahoney, and Matthias vom Hau. 2006. “Colonialism and Development:
A Comparative Analysis of Spanish and British Colonies.” American Journal of Sociology
11:1412-1462.
Lenski, Gerhard E., and Patrick D. Nolan. 1984. “Trajectories of Development: A Test of
Ecological-Evolutionary Theory.“ Social Forces 63:1-23.
Lewis, Bernard. 1995. The Middle East. 2000 Years of History from the Rise of Christianity to the
Present Day. London: Weidenfeld and Nicolson.
Maddison, Angus. 2007. Contours of the World Economy, 1-2030 AD. Essays in Macro-Economic
History. Oxford: Oxford University Press.
Mahoney, James. 2003. Long-Run Development and the Legacy of Colonialism in Spanish
America. American Journal of Sociology 109(1):50-106.
Mauro, Paolo. 1995. “Corruption and Growth.” The Quarterly Journal of Economics 110:681-712.
Mitchell, Brian R. 2007. International Historical Statistics. Africa, Asia & Oceania, 1750-2005.
5th ed. Basingstoke/New York, NY: Macmillan.
Mitchener, Kris J. and Marc Weidenmier. 2008. Trade and Empire. National Bureau of Economic
Research (NBER) Working Paper No. 13765, Cambridge, MA: National Bureau of Economic
Research. Retrieved February 11, 2017 (www.nber.org/papers/w13765).
Mueller, Hans-Peter, Claudia Kock, Eva Seiler, and Brigitte Arpagaus. 1999. Atlas vorkolonialer
Gesellschaften. Sozialstrukturen und kulturelles Erbe der Staaten Afrikas, Asiens und
Melanesiens. Berlin: Reimer.
Nohlen, Dieter, and Franz Nuscheler, eds. 1992ff. Handbuch der Dritten Welt. 8 vol. Bonn: Dietz.
OECD. 1972. Stock of Private Direct Investments by D.A.C. Countries in Developing Countries
End 1967. Paris: Organisation for Economic Co-operation and Development OECD.
Oliver, Roland. 1991. The African Experience. London: Weidenfeld & Nicolson.
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
188
Olsson, Ola. 2009. „On the Democratic Legacy of Colonialism.“ Journal of Comparative
Economics 37(4):534-51.
Osterhammel, Jürgen. 1997. Colonialism. A Theoretical Overview. Princeton, NJ: Markus Wiener
Publishers.
Owen, Roger, and Sevket Pamuk. 1999. A History of Middle East Economics in the Twentieth
Century. Cambridge, MA: Harvard University Press.
Ranger, Terence O. 1969. “African Reactions to the imposition of colonial rule in East and Central
Africa.” Pp. 293-324 in Colonialism in Africa, 1870-1960. Volume 1. The History and Politics
of Colonialism 1870-1914, edited by L.H. Gann and P. Duignan. Cambridge: Cambridge
University Press.
Sanderson, Stephen K. 2005. “World-Systems Analysis after Thirty Years.“ International Journal
of Comparative Sociology 46:179-213.
Schmidt, Elisabeth. 2007. Cold War and Decolonization in Guinea, 1946-1958. Athens, OH: Ohio
University Press.
Schmitt, Carina. 2015. “Social Security Development and the Colonial Legacy.” World
Development 70:332-342.
Sousa, José de, and Julie Lochard. 2012. “Trade and Colonial Status.” Journal of African
Economies 21(3):409-439.
Seth, Michael J. 2006. A Concise History of Korea. From the Neolithic Period through the
Nineteenth Century. Lanham, MD: Rowman & Littlefield.
Trocki, Carl A. 1999. “Political Structures in the Nineteenth and Early Twentieth Century.“ Pp.
75-126 in The Cambridge History of Southeast Asia, Vol. 2, Part 1: From c. 1800 to the 1930s,
edited by N. Tarling. Cambridge: Cambridge University Press.
United Nations. Various years. Statistical Yearbook. New York, NY: Statistical Office of the
United Nations.
Wucherpfennig, Julian, Philipp Hunziker, and Lars-Erik Cederman. 2016. „Who Inherits the
State? Colonial Rule and Postcolonial Conflict.“ American Journal of Political Science
60(4):882-898.
Ziltener, Patrick, and Daniel Künzler. 2013. “Impacts of Colonialism – A Research Survey.”
Journal of World-Systems Research 19(2):290-311.
Ziltener, Patrick, and Hans-Peter Mueller. 2007. “The Weight of the Past – Traditional Technology
and Socio-political Differentiation in African and Asian Societies: A Quantitative Assessment
of their Impact on Socio-economic Development.” International Journal of Comparative
Sociology 48(5):371-415.
Journal of World-System Research | Vol. 23 Issue 1
jwsr.org | DOI 10.5195/JWSR.2017.683
189
Annexes
Annex 1. Interrelations between colonizer, regions and impact
UK F Sub-Saharan Africa Asia/North Africa
colyears13
violcol + —
violres —
violind
violtot —
domform —
ethnfunc + —
indtrans —
gold
mining
tradepol — +
tradecon — +
fdicon + —
invest
plantat
forpres
workim
mission + —
borders + + —
+ = positive correlation, - = negative correlation, no indication = no correlation.
British vs. French colonies: N= 75 (without semi-colonial countries and shared colonies).
Sub-Saharan African vs. Asian/North African colonies: N=83.
All indicated correlations significant at the 0.01 level, 2-tailed. Source: Ziltener, Künzler and Walter Dataset (2008).
13 The variable COLYEARS is introduced in Ziltener and Künzler (2013) and measures the length of colonial
domination. We find no significant correlation between the colonizing country (British vs. French) and COLYEARS
for the countries of our sample. Also, there is no significant difference between the length of colonialism in sub-
Saharan African and Asian/North African countries. But the length of colonial domination is related to some economic
and social indicators of colonial transformation (annex 2): A longer colonial period means more colonial violence,
more investment in infrastructure and more plantations, more work immigration and more religious conversions.
Journal of World-System Research | Vol. 23 Issue 1 | Research Note
jwsr.org | DOI 10.5195/JWSR.2017.683
190
Annex 2. Interrelations Between Types of Colonial Impacts
c
o
l
y
e
a
r
s
v
i
o
l
c
o
l
v
i
o
l
r
e
s
V
io
li
n
vi
ol
to
t
d
o
m
fo
r
m
et
hn
fu
nc
in
dt
ra
ns
g
ol
d
mi
ni
ng
tr
a
d
e
p
ol
tr
a
d
e
c
o
n
f
di
c
o
n
in
v
es
t
pl
a
nt
at
fo
rp
re
s
w
or
ki
m
m
is
si
o
n
bor
der
s
colyears — X X X X X
violcol — X X
violres — X X X X X X X
violind — X X X
violtot X X X X — X X X X X
domform X X X — X X X X X X
ethnfunc X — X X X X
indtrans X X X X — X X
gold X X — X X X X X
mining X — X X X
tradepol X X X X — X X
tradecon X X X X — X X
fdicon X X X X — X X X X
invest X X X X X X X X — X X X
plantat X X X X X X — X X
forpres X X X X —
workim X X X X — X
mission X X X X X X X — X
borders X X —
X = Positive correlation between the two variables, significant at the 0.01 level, 2-tailed. Source: Ziltener and Künzler
Dataset (2008).
Journal of World-Systems Research
Vol. 1 | DOI 10.5195/JWSR.1