Friday, July 8, 2011

Impacts of Colonization on India, Pakistan and Bangladesh by Emily de la Reguera and Jessi Waters


I.               Introduction
Throughout history, imperial colonization has had lasting impacts on the territories and colonies that were ruled by foreign powers. European empires increased their power through the occupation of these areas all over the world by gaining access to their local trade, resources and labor. In South Asia, the British Empire controlled the area of what is now India, Pakistan and Bangladesh until the region’s independence in 1947.
With the independence movements in the 20th century and essentially an end to imperial colonization, scholars began studying the degree to which foreign rule influenced and shaped the development of modern states.  Daron Acemoglu, Simon Johnson and James Robinson studied the impacts of colonization in African territories. Their analysis led them to believe that the type of colony – extractive or settlement – during the period of colonization would determine their post-colonial economic performance and development. Furthermore, they argue that an extractive colony will have lower levels of development in terms of income per capita while a settlement colony will experience higher levels of development and therefore achieve lower income levels. This paper looks to apply Acemoglu, Johnson and Robinson’s thesis to India, Pakistan and Bangladesh, asking the question does the type of colony that was implemented in the region by the British determine the development levels and economic factors in these three countries? We argue that since India, Pakistan and Bangladesh share the same colonial history under British imperial rule then they will also have similar patterns of development. Furthermore, if we discover that the British established a settlement colony in colonial India, then India, Pakistan and Bangladesh will have high development levels marked by high Gross National Income per capita levels. However if the British set up an extractive colony, then we expect to find that India, Pakistan and Bangladesh will have low development levels. Therefore in testing the thesis of Acemoglu, Johnson and Robinson we expect that the type of colony that was set up in the region will produce similar patterns of development and furthermore determine their level of development after independence.

II.              Literature Review
As states began to shape themselves after gaining independence from foreign occupation, studies began analyzing the factors that influence the direction and level of development. Along with Acemoglu, Johnson and Robinson who examined the impact of the type of colony in African countries, scholars have also tried to discover other possible impacts of colonization. Ertan and Putterman suggest in “Determinants and Economic Consequences of Colonization: A Global Analysis”, that colonization has had a significant impact on the development of countries.  They argue that there are three explanatory factors: differences in pre-1500 – pre-Neolithic Revolution – levels, proximity to the colonizing power, and the disease environment that the colonizers faced.  They find that it is not a history of colonization that has a direct effect on development levels and rates of economic growth, rather that “the share of the population that migrated from the places of greater pre-modern development had a positive impact on the current level and growth rate of income and on the quality of institutions in the newly settled places” (Ertan and Putterman, abstract).  In addition to migration, Lange and Rueschemeyer examine factors including settlement, economic disease and pre-colonial populations to see the variation and form of British colonialism.  Rather than looking at settlement versus extractive colonies, Lange and Rueschemeyer categorize British rule into direct and indirect forms of colonialism. However their results were very similar to those of Acemoglu, Johnson and Robinson. Lange and Rueschemeyer discovered that direct rule in which the British set up bureaucratic institutions involving colonial officials, like settlement colonies, provided a more positive effect on post-colonial development. On the other hand, indirect rule, similar to extractive colonies, consisted of poor infrastructure and leadership domination that prevented post-colonial areas from achieving successful political, economic and social development.
While many scholars have focused on the impacts of colonialism by focusing on the period of colonialism itself, Engerman and Sokoloff examine the initial factor endowments of the colonies prior to the foreign occupation. They suggest that differences in development patterns among post-colonial states are a result of “the degree of inequality in wealth, human capital, and political power” which are rooted in the “differences in the initial factor endowments of the respective colonies” (Engerman and Sokoloff, p. 34). The initial factor endowments prior to colonialism attracted different foreign empires which looked to set up different types of colonies. Thus it was the factor endowments that existed in the regions which determined the type of colonialism that would be established which consequently determined the level of political, economic and social inequality which finally impacted the post-colonial development.
In addition to examining the possible factors that might explain the impact of colonialism on the post-colonial development of the state, scholars also look to compare the types of colonialism executed by different foreign empires. Lange, Mahoney, and vom Hau compare British and Spanish forms of colonialism, and argue that Spain tended to colonize countries that were more heavily populated and more developed, whereas Britain tended to colonize countries that were less densely populated and more underdeveloped.  As a result, while both Spanish and British colonization affected development in their respective colonies, British colonialism had more positive effects than the Spanish. 
            These studies all look to analyze the impacts of colonialism on the post-colonial states and explain the different patterns of development based on the different types of colonies that were set up. While these scholars have tried to account for different factors of colonialism that impact the colony’s development including time periods, proximity to the colonizing power, diseases, population, institutions, factor endowments, income, capital and political factors, they all tend to follow the thesis put forward by Acemoglu, Robinson and Johnson that the type of colonialism determines their post-colonial level of development.

III.            Theory
            The main theory that we are testing in this paper is that of Acemoglu, Robinson and Johnson.  According to Acemoglu, Robinson and Johnson, colonies that were set up as settlement colonies will experience greater levels of development after independence because settlement colonies have witnessed and experienced forms of government via their colonizers and are therefore structured to perform better than extractive colonies. Extractive colonies on the other hand, were used by the colonizers solely for their resources and thus the extraction of their resources and the lack of any advanced government institutions and frameworks leads to them performing worse relative to settlement colonies, as defined in terms of income per capita.
Based on this framework, we will look to examine the impacts of British imperialism on colonial India – which is now India, Pakistan and Bangladesh. Given the fact that these three countries each share the same colonial history under the British Empire, we will test the thesis of Acemoglu, Johnson and Robinson by analyzing whether or not these post-colonial countries also have similar patterns of development. We will also look to analyze whether or not colonial India was set up as an extractive colony or as a settlement colony to test whether or not the type colony that was established yields high or low development levels. In order to measure the level of development, we will be looking at Gross National Income per capita (GNI), the same determinant of development used by Acemoglu, Johnson and Robinson, starting in 1973 through 2009. While Bangladesh gained independence in 1971, 1973 marked the first year that all three countries were able to report a GNI per capita. After collecting this data, we will then take the average of the GNI per capital between 1973 and 2009 to determine the average GNI for the current country. We will be using the World Bank Database to gather these statistics for each country. We will also be comparing these statistics with those of other countries with similar development levels based on the World Bank classifications to gauge the development levels of the three countries with the rest of the world.
In accordance with the thesis of Acemoglu, Johnson and Robinson, we expect to find that as a result of their shared colonial history, India, Pakistan and Bangladesh will have similar patterns of development. Furthermore if the British set up a settlement colony in the region, we expect to find high GNI levels in all three countries to demonstrate high patterns of development. If the British set up an extractive colony then we expect to find low GNI levels to demonstrate low patterns of development. Based on these expectations we presume that Acemoglu, Johnson and Robinson were accurate in theorizing that a similar colonial history would yield similar patterns of development and more specifically that settlement colonies will generate a stronger level of development in the post-colonial country whereas extractive colonies will generate a lower level of development in the post-colonial country.

IV.            Data and Evidence
            Based on the thesis of Acemoglu, Johnson and Robinson, we expect to find that India, Pakistan and Bangladesh all have similar development levels given that they share the same colonizer. However, according to The World Bank Data, we find that India and Pakistan are categorized as “lower-middle income” countries and Bangladesh is categorized as a “low income” country.  
We also discovered that the British did not settle in British India at one time.  They gradually conquered territories across British India and depending on the time frame, some states were set up as extractive states whereas others were set up as settlement states.  Therefore India, Pakistan and Bangladesh did not all share the same colonial history because different types of colonies were set up in the region. In order to determine which areas were set up as settlement colonies and which areas were extractive colonies, we have used the cut-off date of 1858 as the determinant for whether a state is considered an extractive one or a settlement one. We chose to use 1858 because it marks the year when control over colonial India transferred from the East India Company to the British Crown. The East India Company looked to extract as much wealth from the colony as possible and therefore set up extractive states. However once the British Crown assumed control, it was more likely for settlement colonies to be established. Therefore areas that were occupied by the British before 1858 will be considered extractive colonies and areas that were occupied after 1858 will be considered settlement colonies.
Table 1 lists the Indian states under British Rule, the time period that the British acquired it, the type of colony it was, and the current country to which the state belongs.  Based on the current country that the state belongs to, the final column of the table indicates the average GNI per capita for that current country as representative of the development levels for each country. Table 2 is a legend of the three average GNIs per capita for Bangladesh, India, and Pakistan.
Based on this data, our findings indicate that there were only 2 states that would be classified as "settlement" based on the post-1858 date – Baluchistan which is now part of modern day Pakistan and Tipperah now part of modern day Bangladesh.  We also found that there were some states that were neither considered extractive or settlement states because they were considered “Princely States.” These “Princely States” fall outside of our definitions of extractive and settlement states because while they were considered part of British India, they were not technically annexed by the British. Therefore they are labeled as “other.”  Every other state would be labeled as extractive, meaning that all of India would be considered "extractive" while parts of modern day Pakistan and Bangladesh would be labeled as "settlement". 






Table 1
States under British Rule
Time Period of British Acquisition
Type of Colony (Settlement or Extractive)
Current Country
Average Gross National Income per capita (between 1947 and 2009)
Punjab
Before 1858
Extractive
India/Pakistan

Sind
Before 1858
Extractive
Pakistan
453.5135
Sirhind
Before 1837
Extractive
India
424.3243
Rohilkhand
Before 1805
Extractive
India
424.3243
Oudh
Before 1837
Extractive
India
424.3243
Kumaon
Before 1837
Extractive
India
424.3243
Bihar
Before 1805
Extractive
India
424.3243
Bengal
Before 1805
Extractive
India/Bangladesh

Sikkim
Before 1858
Extractive
India
424.3243
Chota Nagpur
Before 1805
Extractive
India
424.3243
Assam
Before 1837
Extractive
India
424.3243
Orissa
Before 1805
Extractive
India
424.3243
Bhonsle
Before 1837
Extractive
India
424.3243
Berar
Before 1858
Extractive
India
424.3243
Ajmer
Before 1837
Extractive
India
424.3243
Gujarat
Before 1805
Extractive
India
424.3243
Northern Circars
Before 1805
Extractive
India
424.3243
Madras
Before 1805
Extractive
India
424.3243
Mysore
Before 1837
Extractive
India
424.3243
Coorg
Before 1837
Extractive
India
424.3243
Malabar
Before 1805
Extractive
India
424.3243
Cochin
Before 1805
Extractive
India
424.3243
Tinnevelly
Before 1805
Extractive
India
424.3243
Baluchistan
After 1858
Settlement
Pakistan
453.5135
Kashmir
Princely State
Other
India/Pakistan

Rajputana
Princely state
Other
India
424.3243
Kathiawar Peninsula
Princely state
Other
India
424.3243
Tipperah
After 1858
Settlement
Bangladesh
287.027
Travancore
Princely state
Other
India
424.3243
Nizam’s Dominions
Princely state
Other
India
424.3243


Table 2
INDIA-424.3243
BANGLADESH-287.027
PAKISTAN-453.5135

            Graph 1 shows the income trends for India, Bangladesh, and Pakistan.  Based on the GNI per capita for each of the three countries between 1973 and 2009, the graph below represents the average trends for the three countries.
Graph 1









Graph 1 shows a positive slope and increasing trend for all three countries’ average GNI per capita between 1973 and 2009.  This positive trend indicates that their GNI per capita has consistently increased each year since 1973, and in using GNI per capita to define development, this therefore also shows that the three countries have increased their development levels between 1973 and 2009.  While the three countries have had a gradually increasing GNI per capita trend, it seems that 2003 marks a significant increase in each of the three countries GNI per capita. 
According to the graph, Bangladesh has consistently had the lowest average GNI per capita, except for 1975, when Bangladesh and India both had an average GNI per capita of 190 and Pakistan had a GNI per capita of 150.  For most of the time between 1973 and 2009, Pakistan had a higher GNI per capita than India; specifically from 1977 to 2003.  In 2004, India and Pakistan share the same GNI per capita of 640 and in 2005, India’s GNI per capita surpasses Pakistan’s and has continued to do so.

V.            Comparative Analysis
            As Table 1 demonstrates, the British actually set up many different types of colonies within the region. Therefore modern day India, Pakistan and Bangladesh experienced different types of colonization despite the fact that they were all part of colonial India under the control of the British. All of the areas occupied by the British that now make up India were set up as extractive states while parts of modern day Pakistan and Bangladesh were also extractive states aside from Tipperah and Baluchistan. Based on Acemoglu, Johnson and Robinson’s thesis we would expect that the regions that are labeled as “extractive” colonies to have similar development (GNI per capita) levels. However, according to the World Bank, Bangladesh is classified as “low income” and India and Pakistan is classified as “lower middle income.”  More specifically, if the areas predominantly in India are defined as "extractive" colonies, then we would expect based on Acemgoglu, Johnson and Robinson’s thesis that India would have lower income levels than Bangladesh and Pakistan, but in actuality India’s development level is categorized as higher than Bangladesh’s, who is considered to be low. Pakistan and Bangladesh on the other hand had areas that are defined as “settlement” colonies. With this in mind, we would expect Pakistan and Bangladesh to have a higher development (GNI per capita) level than extractive India, but Pakistan is actually categorized in the same income level group as India and has roughly the same GNI per capita as India while Bangladesh is categorized in a lower income level group and has a significantly lower GNI per capita than both India and Pakistan.  Although parts of modern day Bangladesh and Pakistan were also established as “extractive” colonies, we would still expect that these countries would have higher development levels than those of India because of the presence of some “settlement” colonies whereas all the regions in India were set up as “extractive” colonies. However our results indicate that this was not the case because India had similar or higher GPI per capita levels in comparison to Pakistan and Bangladesh. This shows that there is another factor outside of colonial history that must explain the difference in development levels.
            Graph 1 also reveals that Pakistan and India have similar post-colonial development levels. Although Pakistan had higher income levels from 1977 to 2003, in 2004 the two countries shared the same income levels while after 2004, India experienced higher income levels than Pakistan. This is an anomaly because given that Pakistan is defined as partly a settlement state and India comparatively was more of an extractive state, we would expect given Acemoglu, Johnson and Robinson’s thesis that Pakistan would have had the higher development level and thus a higher GNI per capita level.  One thing that may contribute to this anomaly is the War on Terrorism.  Because of the predominance of Islamic fundamentalists in Pakistan, and the United States’ War on Terrorism that began after September 11, 2001, it is likely that a drop in foreign aid and investment in Pakistan may contribute to a lower GNI per capita.  Similarly, American businesses have expanded in India since 2000 with the post millennium increase in technology and technological services, India’s GNI per capita has increased.
            Our results, however indicate that Acemoglu, Johnson and Robinson’s thesis does not perfectly hold in the case of colonial India under the control of the British. We expected that because India, Pakistan and Bangladesh all shared the same colonizer, that they would have similar patterns of development after independence. However, the British set up different types of colonies in different regions and therefore the impacts of colonization on the three countries were also dissimilar. We also expected that if the British had set up a settlement colony that we would discover high levels of development, represented by high GNI per capita levels in the three countries. Conversely if the British had set up an extractive colony, we expected to find low levels of development, or low GNI per capita levels in the three countries. Based on our findings, we discovered that the regions that make up modern day India were all extractive states, while parts of modern day Pakistan and Bangladesh were both extractive and settlements. Thus we would expect Pakistan and Bangladesh to have higher patterns of development marked by higher GNI per capita levels. However Bangladesh actually has the lowest average GNI levels and was categorized as a low-income country by the World Bank, whereas Pakistan had the highest average GNI levels but was very close to the average GNI levels of India and after 2004 experienced lower GNI levels relative to India. Therefore Acemoglu, Johnson and Robinson’s thesis does not hold in the case of colonial India because India as an extractive state experienced higher levels of development than Bangladesh and similar levels of development with Pakistan and both Bangladesh and Pakistan had extractive and settlement states.

V.             Conclusion

In comparing the GNI per capita levels of India, Pakistan and Bangladesh, we discovered that India and Pakistan had much higher levels of development compared to Bangladesh. However when we compared the three countries to other countries in the world, we found that India and Pakistan were categorized as lower-middle income countries while Bangladesh was categorized as a low-income country. Furthermore in comparison to other lower-middle income countries, India and Pakistan have relatively low development levels defined by low GNI per capita levels. Bangladesh, on the other hand, has relatively high income levels in comparison to other low-income countries, but if we were to compare Bangladesh’s income levels with other lower-middle income countries like India and Pakistan, we would find that Bangladesh has significantly lower income levels and therefore lower development levels. This data also refutes the thesis of Acemoglu, Johnson and Robinson because their thesis claims that settlement colonies will have higher levels of development in comparison to extractive colonies. However Bangladesh and Pakistan both had “settlement” colonies during the period of British colonialism and the development levels of these countries today, represented by their average GNI per capita levels between 1937 and 2009 are in fact similar to or lower than those of India, which only consisted of extractive states. Furthermore, when we compare the development levels of Bangladesh and Pakistan, both of which had areas that were “settlement” colonies, to other countries that were once British “settlement” colonies like the U.S., Canada, Austalia and New Zealand, we find significant differences in GNI per capita levels. The U.S., Canada, Australia and New Zealand are all classified by the World Bank as high-income countries. Thus we would expect that since parts of Bangladesh and Pakistan were established as “settlement” colonies, we would find similar patterns of development to those of the U.S., Canada, Australia and New Zealand. However, as seen by our results, Bangladesh and Pakistan have very low levels of development in comparison to these other “settlement” colonies. Given these disparities along with our results which indicate that Bangladesh and Pakistan – both of which had areas that were established as “settlement” colonies – have lower income levels than those of India, which was established as an “extractive” colony, the thesis of Acemoglu, Johnson and Robinson does not hold in the case of British India.
In theorizing why this may be the case, research indicates that factor endowments, trade barriers, institutions, and investment climate may help to explain the differences. David Dollar, Mary Hallward-Driemeier and Taye Mengistae in their paper “Investment climate and International Integration” examine the relationship between investment climate and international integration.  They define good “investment climate” in terms of the regulatory systems that firms have to operate within the country.  More specifically, if the infrastructure and regulation is corrupt and inefficient so that firms cannot operate within a dependable regulatory framework then the investment climate would be considered low and vice versa.  Based on surveys from firms in Bangladesh, India, Pakistan, Brazil, China, Honduras, Peru, Nicaragua, they find a clear relationship between low investment barriers corresponding to higher levels of international integration.  Specific to this paper, according to Dollar, Hallward-Diemeier and Megistae, India, Pakistan and Bangladesh have hardly increased development as defined by income.   India and Bangladesh’s international integration has increased between 20% and 45% while Pakistan has had no increase in international integration.  Contrastingly, China has experienced tremendous growth, which they contribute to both investment climates and local governance.  
In their article “Institutions, Trade and Growth” David Dollar and Aart Kraay argue that both institutions and greater participation in international trade tend to also correspond with economic growth, as defined by GDP, in the long run.  However, using cross-country regressions, they find that in the short run, trade is more likely to increase economic growth, while in the long run both higher levels of trade in conjunction with sound institutions contribute to higher levels of economic development. 
Stanley Engerman and Kenneth Sokoloff suggest that the initial factor endowments of the respective countries explain the difference of New World economies’ levels of GDP per capita.  More specifically, they refer to differences in the degree of inequality in wealth, human capital and political power.  While they focus almost exclusively on Canada and the United States as examples, we find that it is still relevant as the United States was a British settlement colony and was therefore more likely to experience earlier and more sustained economic growth.  They suggest that this is because the factor endowments in the United States as compared to those of Latin America “encouraged the evolution of more equal distributions of wealth and more democratic political institutions…..and more growth oriented policies” (Engerman and Sokoloff, 4).
It is difficult to pinpoint a specific causal factor that explains different development levels in India, Pakistan, and Bangladesh.  According to World Bank data, Acemoglu, Robinson, and Johnson’s claim that shared colonial history and whether or not the colony was settlement or extractive does not seem to hold.  Whether the causal link is that of investment barriers, factor endowments, levels of trade or institutions is questionable.  Certainly they all play an important role, but determining which one is the causal link is more unclear.  What does seem clear however is that the economic development of India, Pakistan, and Bangladesh is less determined by their past colonial history and more with the governance and policies that have occurred post colonialism, showing that their economic success is more within their control than their colonizer’s. 





Appendix A:

Map of the British Acquisition of Indian Territories 




































References

Acemoglu, Daron, Simon Johnson, and James A. Robinson. 2001. “The Colonial Origins of Comparative
Development: An Empirical Investigation.” American Economic Review, 91: 1369–1401.

Dollar, David and Aart Kraay. (2003). “Institutions, Trade and Growth.” Journal of Monetary Economics,                 
                  50: 133-162.

Dollar, David, Mary Hallward-Driemeier and Taye Mengistae. (2006). “Investment Climate and                  International Integration.” World Development, 34(9): 1498-1516.

Engerman, Stanley L. and Kenneth L. Sokoloff. 1994.“Factor Endowments: Institutions and Differential                  Paths of Growth Among New World Economies: A View from Economic Historians of the United                  States.” The National Bureau of Economic Research, 99: 1-54.

Ertan, Arhan S. and Louis Putterman. 2007. “Determinants and Economic Consequences of Colonization:                  A Global Analysis.” Boston University Department of Economics, 1-34.

Lange, Matthew and Dietrich Rueschemeyer. States and Development: Historical Antecedents of                  Stagnation and Advance. Palgrave Macmillan: New York, 2005.

Lange, Matthew, James Mahoney and Matthias vom Hau. (2006). “Colonialism and Development: A                  Comparative Analysis of Spanish and British Colonies.”  AJS, 111(5): 1412-1462.

World Bank. “Bangladesh.” The World Bank. 30 Mar. 2011                  <http://devdata.worldbank.org/AAG/bgd_aag.pdf>.

World Bank. “India.” The World Bank. 30 Mar. 2011.                  <http://devdata.worldbank.org/AAG/ind_aag.pdf>.

World Bank. “Pakistan.” The World Bank. 30 Mar. 2011.
                  <http://devdata.worldbank.org/AAG/pak_aag.pdf>.