January 17, 2017

Summary - This paper pursues an inquiry into the relationship between ethnicity and development in the largest authoritarian country in the contemporary world, the People’s Republic of China. It engages the theoretical literature on ethnic diversity and development in general, but also pays special attention to political economy logics unique to authoritarian systems. Focusing on the western part of China over a decade since the launch of China’s Western Development Program (xibu da kaifa) in 2000, this paper utilizes the data from two censuses (2000 and 2010) together with nighttime streetlight imagery data to analyze the overall relationship between ethnicity and development provision. It also analyzes changes in such a relationship during this period. The paper finds that ethnic minority concentration negatively correlates with economic development in both the years 2000 and 2010 across the western provinces. It also finds that counties in non-autonomous provinces, which are historically more integrated with the rest of China than autonomous provinces, have a positive and systematic correlation between changes in ethnic minority concentration and changes in development during the 10-year period. The counties in autonomous provinces, on the other hand, show the opposite trend. Using three case studies of Tibet, Inner Mongolia, and Xinjiang, the paper concludes that although there is in general a tendency for ethnic minority concentrated areas to be less developed, ultimately which groups prosper more or less depends upon specific economic development and which political control logics the Chinese state implements.

Key words

ethnicity and development; 
China; 
western development program; 
authoritarian regimes and development; 
authoritarian regimes and ethnicity; 
East Asia
1. Introduction

How does ethnic integration explain the level of development we see in authoritarian states? While development in general entails economic growth and provision of public goods, ethnicity has also featured prominently in the literature for its role in shaping and contextualising outcome processes. When economic growth is concerned, for example, the lack of ethnic diversity has been considered a powerful indicator in many parts of the world (Alesina et al., 2003, Alesina and Ferrara, 2000 and Easterly and Levine, 1997). In terms of public goods provision, ethnicity has also been pointed out as a crucial explanatory valuable (Alesina et al., 1999 and Berman and Laitin, 2008). The common assumption is that provision of public goods is easier in ethnically homogenous societies than in ethnically diverse ones, because in ethnically diverse societies, minorities’ precarious political status and the tendency toward in-group favoritism among politically connected members of the ethnic majority lead to discrimination in public goods provision against minorities. Ethnically diverse societies may have coordination problems in providing public goods, and different communities may also have divergent preferences (Habyarimana, Humphreys, Posner, & Weinstein, 2007). In addition, much of the literature on the relationship between ethnicity and development often focuses on democracies, where the electoral process specifically incentivizes development along ethnic lines, although certainly democracies often tend to protect minority rights as well (Brown and Mobarak, 2009 and Hassler et al., 2007). Few, if any, studies have explored empirically how authoritarian regimes provide development in ethnically diverse settings (Tsai, 2007). Can there be similarities or differences in development provision in ethnically diverse societies when there is a lack of electoral process and rule of law?

This paper pursues an inquiry into the relationship between ethnicity and development in the largest authoritarian country in the contemporary world, the People’s Republic of China. It engages the theoretical literature on ethnic diversity and development in general, but also pays special attention to political economy logics unique to authoritarian systems. Empirically, the paper examines whether ethnic divisions between the majority Han Chinese and various other ethnic minorities have an effect on development throughout the western part of China, where the majority of China’s ethnic minorities reside. For this purpose, this paper utilizes data from two Chinese censuses in 2000 and 2010, together with nighttime streetlight imagery data, to analyze the overall relationship between ethnicity and development as measured by luminosity, as well as changes in such a relationship during the 10-year period. More importantly, the year 2000 also saw the official launch of “Open Up the West” (xibu da kaifa) initiative, also known as the Western Development Program (WDP) ( Lai, 2002). While economic development in Western part of China is likely conditioned upon several factors, including various local initiatives for industrialization, the 10-year span since the launch of the WDP creates a golden opportunity to study whether this state-led initiative, which involved the migration of the majority Han Chinese into the ethnic minority-dominated western provinces, had any implication on the ethnic dimension of development in China.

The structure of the paper is as follows. After reviewing the literature, the paper introduces the politics of ethnicity in China, the background of the WDP and its ethnic characteristics. It presents a couple of operationalizable hypotheses for empirical testing, explains the research design, and then describes our data. The results of our statistical analysis offer a set of nuanced findings. Overall, ethnic minority concentration negatively correlates with development in both the years 2000 and 2010 across the western provinces. Compared to western provinces designated as ethnic minority autonomous regions (Tibet, Inner Mongolia, Xinjiang, Guangxi and Ningxia), counties in the non-autonomous provinces (Chongqing, Gansu, Guangxi, Guizhou, Qinghai, Shaanxi, Sichuan, and Yunnan) have a positive and systematic correlation between changes in ethnic minority demographic concentration and changes in development during the 10-year period, when the Chinese state directed its attention at economic development in these provinces. The interaction term between the change in minority concentration and the dummy for autonomous province, on the other hand, is negative. This means that while increases in ethnic minority concentration are generally associated with increases in development in western provinces, this relationship does not hold in ethnic minority autonomous provinces. The autonomous provinces have instead benefited less from the WDP, as they have been predominantly inhabited by ethnic minorities and remained less integrated with the rest of China.

We also find that while the overall relationship between the minority concentration and development is negative for autonomous provinces and positive for non-autonomous western provinces, the relationship is much more complex among the autonomous provinces. In Tibet, there is indeed a negative relationship between changes in ethnic minority concentration and changes in the level of economic development. This means that counties with more growth of Tibetan population relative to Han population have experienced less development in comparison with those with more relative growth in Han population during the ten-year period. However, in Inner Mongolia, the relationship is positive, and the growth of ethnic Mongol population is positively correlated with economic development. We contend that because Inner Mongolia is the first ethnic minority region established in 1947, where more than 80% of the population are Han Chinese, the Chinese state perceives that the region is much better integrated than Tibet, where the Tibet autonomous region was only established in 1965 and where Han Chinese still only account for less than 10% of the local population. These findings suggest that we have to understand the ethnic dimension of economic development in an authoritarian system such as China through the lens of political control. This paper presents the two contrasting case studies of Tibet and Inner Mongolia in detail to illustrate such logics. In addition, we include a third case of Xinjiang, where rising violence from radicalized Uyghurs has rendered the Chinese government’s plan to encourage development through integration of Han Chinese less successful. In the following conclusion, we present further theoretical reflections on ethnicity and economic development in authoritarian systems in general.

2. Ethnicity, development provision, and authoritarianism

There has been a vast amount of literature written in economics and political science on the development externalities of ethnicity. Scholars have probed how development is dependent upon various ethnic factors in ethnically diverse societies. For example, there are many works concerned with the economic consequences of ethnic distribution in a given society. For some, ethnic diversity is shown to have a direct negative effect on economic growth (Alesina et al., 2003, Alesina and Ferrara, 2000, Easterly and Levine, 1997, Gisselquist, 2014 and Sala-i-Martin et al., 2004). For others, it is not ethnic fractionalization but rather ethnic polarization that is believed to retard economic development. The effect of polarization on economic growth can be explained through its impact on civil wars, the rate of investment, and the proportion of government consumption over GDP (Montalvo & Reynal-Querol, 2005). Relatedly, there is also discussion about the effect of regime type on ethnicity’s relations with economic growth, in that democracies prove to be able to ameliorate the negative effect of ethnic diversity (Collier, 1999).

A large body of literature related to the issues of economic development also exists; specifically, how ethnic diversity affects the provision of public goods. As one of the classic pieces on the topic by Alesina, Baqir, and Easterly argues, ethnically diverse societies tend to have fewer provisions of public goods than ethnically homogenous ones (Alesina et al., 1999). Although this study is based on data from the United States, the idea has been empirically tested elsewhere (Berman and Laitin, 2008, Cooray, 2014, Miguel and Gugerty, 2005 and Schündeln, 2013). For example, Baldwin and Huber, in their study of 46 African countries, demonstrate that although ethnic diversity can mean different things according to specific measurements, economic differences between ethnic groups are statistically and negatively correlated with public goods provision (Baldwin & Huber, 2010).

Additionally, scholars have probed why, instead of whether, ethnic diversity impedes public goods provision. As stated originally by Alesina, Baqir, and Eastly, the reasons why ethnically diverse societies have low public goods provision are mainly due to two mechanisms: either because people do not want to share with ethnically different others or because different ethnic groups tend to have non-aligned preferences when public goods provision is concerned (Alesina et al., 1999). Adding on to these mechanisms, Habyarimana et al. contend that public goods provisions are better provided in ethnically homogenous societies because co-ethnics are more likely to play cooperative equilibria. Therefore, the under-provision of public goods in ethnically diverse societies is not because of innately different preference systems across ethnic groups ( Habyarimana et al., 2007). On the other hand, Lieberman and McClendon instead argue that ethnicity is rather used as a group heuristic for evaluating public policies, which illustrates that the relationship between ethnicity and public goods provision is in fact a strategic and political one (Lieberman & McClendon, 2013). In addition, Wimmer contends that the relationship between ethnic diversity and public goods under-provision is spurious because “both contemporary ethnic heterogeneity and low public goods provision represent legacies of a weakly developed state capacity inherited from the past” (Wimmer, 2016).

Thrown in this mix is how the type of regime can contextualize the relationship between ethnic diversity and public goods provision. Similar to the literature on ethnicity and economic development, there currently exists more of a focus on how ethnic diversity affects public goods provision in a democratic setting. Baldwin, for example, argues that in Zambia people are more likely to vote with their traditional chief if they perceive that a strong relationship between chiefs and politicians can lead to better local provision of public goods (Baldwin, 2013). Indeed, such an ethnic favoritism argument is quite easy to understand in that politicians are believed to be more likely to favor their co-ethnics in redistributive politics—although Kasara’s study demonstrates the opposite may be the case in some settings (Kasara, 2007).

Most such studies tend to focus on public goods provision in democratic societies (Brown and Mobarak, 2009, Hassler et al., 2007, Nooruddin and Simmons, 2015 and Rosenzweig, 2015). There is scarcely any study that specifically looks at the dynamics of public goods provision in ethnically diverse but authoritarian societies (Tsai, 2007). In this context, one can argue that due to the lack of the democratic process of monitoring, ethnic minorities may be more likely to be discriminated against when it comes to development programs directed by the central state. Indeed, political or material goods are often provided by the ruling regime for some key sections of the population in return for their political support (Taydas & Peksen, 2012), similar to the distributive politics setting in democracies. For example, in a recent study on public goods provision and violence in the Syrian civil war, De Juan and Bank demonstrate that the risk of violence has been lower in sub-districts that have been favored by the ruling Assad regime in terms of preferential access to material goods such as electricity (De Juan & Bank, 2015).

Following the logic of political control, we can also conceptualize that an authoritarian state is more likely to foster development in areas in which it feels its control is more secure and state dominance is unquestioned. This means that in areas where the majority ethnic group has more dominance, the state is more likely to direct development to that area. For ethnic groups living in those areas, the state might also feel comfortable in encouraging development among these groups because their loyalty is considered more trustworthy. However, in areas that the majority ethnic group does not have such dominance, then the state would be more likely to encourage the migration of the majority group to that area through the incentives of economic development. For ethnic minority groups living in these less secure areas, economic development is less likely to be provided unless they demonstrate their loyalty toward the central state. The development change due to the WDP in this regard provides an invaluable case study for the authoritarian state’s policy impact on ethnic minority areas.

3. The politics of ethnicity and development in authoritarian China

China provides an ideal setting to study the relationship between ethnicity and economic development because it has an authoritarian political system with strong development records.1 The state has pursued provisions of necessary infrastructure such as roads, railways, and airports as means to promote local economic growth and, in turn, integrate ethnic minorities with the Han majority. It is therefore both theoretically and empirically interesting to explore how and whether, in the context of rapid economic development, ethnicity features in the “mind” of the authoritarian Chinese state.

As the ruling party of the largest authoritarian country in the world today, the Chinese Communist Party (CCP) has ruled the PRC for more than 60 years. For the past few decades, the country has also wholeheartedly pursued a “developmental state” strategy that has put an exceptional amount of emphasis on economic development (Knight, 2014 and Nordhaug, 2012). At the same time, China is also a country with many ethnic problems. According to China’s 2010 Census, about 91.5% of the Chinese population belong to the majority Han Chinese while the other 8.5% are various ethnic minority groups (National Bureau of Statistics of China, 2013). Although this percentage might be a relatively small one, the absolute number of ethnic minority people in China is more than 100 million, an unarguably significant number. More significant is the fact that the areas where various ethnic minorities have historically resided (and were subsequently designated as minority autonomous areas) constitute about half of China’s territory. In particular the areas of Tibet and Xinjiang, two autonomous regions where almost half of Tibetans and most Uyghurs reside and two regions who have ongoing disputes with the central state, are both extremely large in area size and of significant strategic importance (Bovingdon, 2010, Han, 2013, Millward, 2007, Sautman and Dreyer, 2006 and Shakya, 1999).

China’s policies toward its various ethnic minorities have many components. On the one hand, some of the Chinese state’s policies have often been described as discriminatory. Although Chinese society has undergone dramatic changes during the past few decades, the Chinese state still has shown little tolerance for political dissent from ethnic minorities. Any signs of resistance from these ethnic minorities, often interpreted as separatism by the Chinese state, have faced severe repression (Clarke, 2015, Han and Paik, 2014 and Pirie, 2013). Discriminatory policies toward ethnic minorities can also be observed in the job market, where they are significantly disadvantaged (Hannum and Xie, 1998, Hasmath, 2011 and Maurer-Fazio, 2012). One therefore wonders whether such discrimination against ethnic minorities is also present when economic development is concerned.

Those discriminatory aspects aside, the Chinese political system nonetheless provides “nominal” autonomy based on ethnicity. Institutionally, there are five autonomous regions jurisdictionally equivalent to provinces: Guangxi Zhuang Autonomous Region, Inner Mongolia Autonomous Region, Ningxia Hui Autonomous Region, Tibet Autonomous Region, and Xinjiang Autonomous Region. At sub-provincial levels, there are also autonomous prefectures and autonomous counties, most of them located in non-autonomous provinces. However, in reality, this autonomy translates to no genuine power for the minority groups. Instead, we can think of the provision of autonomous status as an indicator of a lesser presence of the majority Han Chinese, and subsequently as areas that the Chinese state, comparatively speaking, has less secure control than other non-autonomous provinces.

On the other hand, where ethnic minorities are concerned, the Chinese government sees economic development as the main solution for ethnic dissent (Barabantseva, 2009). Since China’s economic reforms began in the late 1970s, the government’s development strategy has focused on initially jumpstarting the economy on the eastern coast. Deng Xiaoping deemed it necessary to let some regions to growth their wealth first, with the understanding that the wealth would later trickle down and spread to the rest of the country (Vogel, 2011). In reality, the development gap between the eastern provinces and the western ones significantly widened throughout the 1980s and 1990s. Particularly worrying for Beijing is the fact that many of its western provinces are also home to the majority of China’s ethnic minority groups. This has created concern among the Chinese leadership as to whether the widening economic gap would hinder its national integration project, or worse still, contribute to further grievance on part of some ethnic minorities toward the central government (thus creating a national security problem). In fact, in much of China’s domestic discourse, problems of ethnic conflict have often been blamed on underdevelopment and poverty among certain ethnic minority groups (Barabantseva, 2009 and Fischer, 2005).

Due to concerns about widening economic gaps, the Chinese government launched the WDP around the year 2000.2 Theoretically at least, this meant that with this change of development focus, there would be not only more resources directed to the western provinces, but also more state intervention in economic development (Goodman, 2004). Concerns about creating a more equal development model aside, the other key mission of the WDP was to generate enough economic centripetal force to further integrate its peripheral regions in which various ethnic minority groups reside. Given this reason, developing the economies of the western provinces and further integrating them into the rest of the country had a nation-building function (Goodman, 2004).3 As far as the Chinese state is concerned, economic development has been necessary to prevent ethnic minority regions from estrangement, that is, “[I]f minority nationalities were not given better chances for economic development, it was argued, then social harmony, political stability and national security would be in danger” (Holbig, 2004).

Owing to this nation-building dimension, the WDP put emphasis on two strategic goals that aimed to facilitate better movement of goods and people between the western regions and the “core” of China. The first was the priority of infrastructure development (Perkins, 2004). This manifested not only in some landmark rail projects, such as the Qinghai–Tibet Railway, but also highways linking the west to the core and roads connecting rural and urban areas in the west. Similarly, many airports have been built to increase connectivity between western regions and the rest of China. The other focus was the implicit push for the migration of more Han Chinese to areas populated with ethnic minorities, although in reality the most migration of Han Chinese into ethnic regions tends to occur in urban areas. Historically, various countries have encouraged members of the ethnic majority to settle minority areas, in order to alleviate population density in ethnic core areas and solidify control over peripheral regions. (Fearon & Laitin, 2011). In the Chinese context, in the 19th century the Qing Dynasty government encouraged the migration of Han Chinese to Manchuria and Mongolia, in order to prevent encroachment upon those areas from the north and expanding Tsarist Russia (Jagchid, 1999).

Therefore, within the context of China, we have a somewhat complicated situation. On the one hand, the ostensibly authoritarian nature of the Chinese government comes with a history of discrimination and repression against ethnic minorities. Meanwhile, the recent strong development focus on the western provinces through the WDP also means that the Chinese state has an explicit goal of further integration of the ethnic minority areas to the rest of China through infrastructure development and migration of its Han Chinese population to ethnic areas. With the confluence of all of these different factors, how does ethnicity feature in terms of development change in China?

4. Research design and data

In order to examine the relationship between development and ethnicity in China, we conceptualize the empirical test to include two levels of analysis. One is to see cross-sectionally the overall relationship between ethnic concentration and development provision in 2000 and 2010. The other is to see temporal changes in development provision in the context of changing ethnic demographic distribution during the 10-year period of the WDP. The purpose is to examine the association between the change in economic development and the factor of ethnicity since the start of the WDP in 2000. Does the Chinese government provide more development in areas with larger shares of Han Chinese? Can we observe areas having an upsurge of such provisions alongside Han Chinese in-migration? Regarding these inquiries we have the following three main hypotheses.

H1.

Areas with higher percentages of Han Chinese residents get more provisions of development.

 

H2.

As a result of Han Chinese in-migration, areas with larger Han Chinese population growth get more of provisions development over the 10-year period.

 

H3.

Ethnic minority autonomous regions where the Chinese government has less control would report less development over the 10-year period.

 

 

The level of analyses of our paper is at the county level, and we confine our analyses to the western provinces in China targeted by the WDP, including Chongqing, Gansu, Guangxi, Guizhou, Inner Mongolia, Ningxia, Qinghai, Shaanxi, Sichuan, Tibet, Xinjiang, and Yunnan.

The data for our analyses come from two main sources. First, for our main dependent variable of development in China, we utilize nighttime streetlight images to gauge the level of electricity provision throughout China. In the Chinese context, the state grid monopolizes electricity provision, whereas almost no private provision exists (Ngan, 2010). Instead of using official statistics from China, this use of illumination data is not only more objective, but also captures the general level of electricity usage on the ground.4 There remain potential drawbacks of using luminosity as a measure of development; these include measurement errors in light density related to any cross-county differences in the use of nightlights, blooming and bleeding of light images, gas flares that incorrectly increase luminosity, and differences in light sensitivity across different satellites. When truly unbiased estimates of development activities are available, they should be preferred to the light data for the above reasons. In the absence of such, however, we argue that the luminosity score in China, averaged over a calendar year to minimize these potential measurement errors, serves our purpose as an arguably suitable proxy.

While we interpret the brightness scale as a measure of overall development of the region, the streetlight data feature prominently in both economics and political science literature, albeit in different ways. For example, numerous works in economics use the electricity data to overcome the lack of credible productivity measures at both national and subnational scales (Chen and Nordhaus, 2011, Ebener et al., 2005, Henderson et al., 2012, Michalopoulos and Papaioannou, 2011 and Sutton et al., 2007). In political science, total lumens are interpreted as functions of government provision of electrification, the allocation of scarce power (in countries with shortages), and government investment in streetlights. In a democratic context, scholars have also used the level of illumination as a measure of the extent to which public goods provision is driven by political favors (Agnew et al., 2008 and Golden and Min, 2013). Their doing so highlights the political incentives to distribute public goods depending on both a region’s political importance and leaders’ strategies for rewarding their constituents.

To measure illumination within each county, we collected data on light from human settlements detected by the Defense Meteorological Satellite Program's Operational Linescan System (DMSP-OLS). The DMSP-OLS “nighttime lights of the world” images are processed specifically for the detection of change, and are made available by the National Geophysical Data Center. In this paper, we only use lights from human settlements in cloud-free composite images produced using all the available archived satellite images of DMSP-OLS during a calendar year. These composites are scaled onto a geo-referenced 30 arc-second grid (approximately 1 km2) where each grid cell takes on a 6-bit scale digital number (DN), from 0 to 63. For each year, a grid cell with a value of zero can be interpreted as an area with zero nighttime light. On the other hand, the value of 63 is the saturation value and indicates the brightest area for each year.5 For each region we calculate the average DN and the difference in DN during 2000–10 as our key dependent variable.

Our main independent variable on ethnic minority population is derived from the 2000 and 2010 Chinese censuses. We use the reported percentages of the population at the county-level which are categorized as ethnic minorities in these two censuses. We also take a list of control variables. For the time-series analysis with 2000–10-differenced control variables, we include changes in population density and total county population, and changes in urban population as a percentage of county population. There are other variables that can influence the changes in streetlight provision over the time period, some of which are only available in the year 2000. Given that we cannot observe the changes in many of these variables over time, we do not use a standard difference-in-differences estimation approach but instead include a list of variables measured in 2000 as initial confounders; these include the mean luminosity score as well as the ethnolinguistic fractionalization (ELF) index score (Fearon, 2003), urban population, total population, population density, average number of years in school, and the percentage of population that is illiterate, all obtained from the China county population census data. We also include 2000 census data on road and railroad coverage by county in kilometers from the township population census data (China Data Center, 2009). In addition, we include several geographic controls, such as the mean county elevation, longitude, and latitude, as well as indicators for a county being an urban center, measured as whether it is where the provincial or prefectural government is located. Finally, we include controls for provincial fixed effects, and an indicator for counties belonging to one of the five ethnic autonomous regions—Guangxi, Inner Mongolia, Ningxia, Tibet, and Xinjiang. We argue that Chinese government’s control in those ethnic autonomous regions is less secure, and thus they would have an effect on development provision along ethnic lines. We will also look at the variation among such autonomous regions according to the level of control the Chinese government has, measured by how early the autonomous region was established and how much its population are Han Chinese.

In the following section, we present our findings with a potential caveat of a simultaneity bias in mind; that is, despite a series of demographic, geographic and initial condition controls, we may not fully exclude the possibility that Han Chinese moved to areas that are more developed over the years. One argument that could help with our empirical strategy is to assert that regions with development attracted not only Han migrants, but other types of migrants as well, i.e., local minority groups who moved from rural to urban areas in search for employment opportunities. Since all types of ethnic groups could move to urban areas, this would imply little evidence of higher luminosity necessarily leading to less minority concentration. However, given the lack of urban population data classified into ethnic groups, and in the absence of a clear instrumental variable for the minority concentration variable, we refrain from making any explicit causal claims here, and focus instead on finding a meaningful association between changes in minority concentration and development.

5. Empirical findings

We first present summary statistics for both autonomous provinces and non-autonomous provinces in Table 1a and Table 1b. It is evident from these tables that autonomous provinces continue to have significantly higher minority concentrations than other western provinces. We also find that for western provinces, both the urban population and population density increased over the time period. Our first cross-sectional analyses are conducted with each census year, and in Table 2, we can see that overall there is a statistically significant and negative correlation between ethnic minority population concentration and development provision. It appears that counties with more ethnic minority concentration generally tend to have fewer developments provided in the form of streetlights compared with counties with more Han Chinese population, irrespective of a set of demographic and geographic controls, as well as provincial fixed effects. The sets of results support our first hypothesis that in terms of development provision, there is a latent bias in the Chinese context, where ethnic minorities have been disproportionally under-provided with electricity.

Table 1a.
Summary statistics of autonomous provinces

Variables
(1)
(2)
(3)
(4)
(5)
 
N
mean
sd
min
max
Mean Luminosity in 2000
353
3.280
9.857
0
60.80
Mean Luminosity in 2010
353
6.184
13.71
0
63
Minority Conc. in 2000
353
49.90
37.10
0.390
99.78
Minority Conc. in 2010
353
50.00
36.56
0.190
99.78
Total Urban Pop. in 2000 (in millions)
353
0.0735
0.0972
0
0.506
Total Urban Pop. in 2010 (in millions)
353
0.107
0.139
0.000508
0.724
Total Population in 2000 (in millions)
353
0.216
0.211
0.00638
1.359
Total Population in 2010 (in millions)
353
0.237
0.229
0.00688
1.497
Population Density in 2000
343
639.5
2,831
0.102
35,590
Population Density in 2010
343
1,127
6,797
0.139
107,631
Mean Elevation in km
343
1.856
1.601
0.0312
5.156
Longitude
353
99.97
13.66
74.90
124.5
Latitude
353
36.22
8.150
21.45
51.62
Average Years in School in 2000
353
6.717
2.195
0.630
11.06
Percent of Population illiterate, 2000
353
17.93
19.22
0.550
86.22
State Road Coverage (km) in 2000
353
63.77
81.06
0
488.2
Provincial Road Coverage (km) in 2000
353
64.42
85.65
0
728.9
Railway Coverage (km) in 2000
353
15.94
46.65
0
470.5

Table options

Table 1b.
Summary statistics

Variables
(1)
(2)
(3)
(4)
(5)
 
N
mean
sd
min
max
Mean Luminosity in 2000
666
2.459
7.645
0
63
Mean Luminosity in 2010
666
5.895
11.88
0
63
Minority Conc. in 2000
666
26.49
33.78
0
99.11
Minority Conc. in 2010
664
26.75
33.76
0
98.86
Total Urban Pop. in 2000 (in millions)
666
0.103
0.147
0
1.262
Total Urban Pop. in 2010 (in millions)
666
0.153
0.188
0.00286
1.375
Total Population in 2000 (in millions)
666
0.378
0.289
0.00889
1.649
Total Population in 2010 (in millions)
666
0.381
0.283
0.0105
1.563
Population Density in 2000
666
1,036
7,625
0.246
131,858
Population Density in 2010
666
1,454
13,572
0.282
309,756
Mean Elevation in km
666
1.555
1.030
0.243
4.809
Longitude
666
104.6
3.324
92.61
110.9
Latitude
666
30.54
4.299
21.73
40.63
Average Years in School in 2000
666
6.458
1.552
1.470
11.71
Percent of Population illiterate, 2000
666
17.86
14.49
1.600
75.65
State Road Coverage (km) in 2000
666
35.53
60.92
0
1,039
Provincial Road Coverage (km) in 2000
666
57.84
58.94
0
582.0
Railway Coverage (km) in 2000
666
12.50
25.65
0
209.0

Table options

Table 2.
2000 vs. 2010 level analysis of western provinces

Variables
(1)
(2)
(3)
(4)
(5)
(6)
Year 2000

Year 2010

Minority Conc. in 2000
−0.053∗∗∗
−0.020∗∗∗
−0.016∗∗∗
 
 
 
 
(0.008)
(0.006)
(0.006)
 
 
 
Minority Conc. in 2010
 
 
 
−0.103∗∗∗
−0.035∗∗∗
−0.017∗
 
 
 
 
(0.012)
(0.009)
(0.010)
 
Observations
1,019
1,009
1,009
1,017
1,007
1,007
Dem.Controls
No
Yes
Yes
No
Yes
Yes
Geo.Controls
No
No
Yes
No
No
Yes
Provincial FE
Yes
Yes
Yes
Yes
Yes
Yes

 

Robust standard errors in parentheses.

***p