The Long-Term Effects of ProtestantActivities in China · 2013-11-14 · has generated significant...

1

The Long-Term Effects of Protestant Activities in China

Yuyu Chen Hui WangGuanghua School of Management Guanghua School of ManagementPeking University Peking [email protected] [email protected]

Se YanGuanghua School of ManagementPeking [email protected]

November 2013

Does culture, and in particular religion, exert an independent causal effect on long-termeconomic performance, or is it merely a reflection of the latter? We explore this issue by studyingthe case of Christianity spread in China during the late 19th and early 20th century. Even thoughChristianity never become a prevalent religion in China, its spread, especially among the lessdeveloped and isolated regions, open a new window for the local people to see the outside world,and encourage them to become more open-minded. Such cultural interactions generate persistentand profound impacts on regional development. Combining county-level data on Protestantpresence in 1920 and socioeconomic indicators in 2000, we find that the spread of Protestantismhas generated significant positive effects in promoting long-run regional economic prosperity. Tobetter understand if the relationship is causal, we exploit the fact that missionaries activelyconducted disaster relief to gain trust from local people and use frequencies of historical disastersas instruments for Protestant distribution. Our IV results confirm and enhance the OLS results. Wefurther investigate the transmission channels between history and today, and find thatimprovements in education and health account for about half of the total effects of missionaries’work on today’s economic outcomes, and the other half are possibly the contributions from thechanges in culture and people’s mindset.

Keywords: Christianity, Economic Growth, Education, Foreign Direct Investment, China

JEL Classification Numbers: I20, N15, N35, O11, O43, Z12

Preliminary; For Seminars and Conferences only; No Dissemination or Citation Please

2

“The missionaries …have been among the pioneers of civilization.”

--- President William McKinley, 19001

1. Introduction

The causal effects of Christianity, in particular Protestantism, on economicperformances are controversial. In his famous work, “The Protestant ethic and theSpirit of Capitalism”, Max Weber suggested that Reformed Protestantism, bynurturing stronger preferences for hard work and thriftiness, has led to greatereconomic prosperity (Weber 2001). Recent empirical studies of the role of religion ineconomic growth have found mixed results. While Cantoni (2010) argues that therewas no effect of Protestantism on economic growth, other studies find significantpositive effects, but suggest different mechanisms through wich religion may havecontributed to economic performances. Some of these studies suggest similarchannels to Weber’s argument that religion beliefs may have fostered certain moralcodes or social values that are conducive to economic growth (Lipset and Lenz, 2000;Stulz and Williamson, 2003; Barro and McCleary 2003, Barro 2004; McCleary andBarro 2006, Arruñada, 2010). But other studies argue that it is not religious beliefs butthe side effects that religious activities brought in that have contributed to economicgrowth (Glaeser and Glendon, 1998; Glaeser and Sacerdote 2008; Becker andWoessman 2009).

In this paper, we use the lens of history to better understand the long-run effectsof religion on economic growth. In particular, we want to find out whether and howthe spread of Protestantism in the late nineteenth and early twentieth centuriesgenerated persistent impacts on China’s economic achievements today. We construct adata set mapping historical Protestant presence and contemporary economic outcomesat county level, and find that the diverse economic performances across differentcounties in 2000 are positively correlated with the intensity of Protestant presence atthe beginning of the twentieth century. To our knowledge, our paper provides the firstempirical evidence on the long-term effects of religious activities on economicoutcomes in China.

However, such correlation might be subject to endogeneity issue due tounobservable county-level heterogeneities. It might be possible that more developedcounties attracted more Protestant activities in the past and continued to performbetter at present. We try to identify the causal effects of Protestant presence oncontemporary economic outcomes in three ways. First, historical archival recordssuggest that missionaries were usually more passionate to go to less developed areas.Second, we proxy historical economic prosperity by population density and land taxrevenues of Chinese counties, and find no positive correlation between historical

1 The president address for the Ecumenical Missionary Conference in the Carnegie Music Hall, New York City,April 21, 1900.

3

economic prosperity and Protestant presence. Third, we notice that missionariesactively conducted disaster relief in order to gain trust from Chinese people. In lightof this historical fact, we use frequencies of droughts and floods in the early twentiethcentury as the instruments for the intensity of Protestant presence across counties. OurIV coefficients are positive and significant, suggesting that more Protestant activitiesin 1920 resulted in better economic outcomes in 2000.

We further investigate the precise channels through which the Protestantactivities in 1920 survived the political turbulences in China persisted to the present.Our empirical tests suggest that higher intensity of Protestant presence results in bettereducation and health conditions in 2000. Historical evidences also corroborate ourfindings. While going to remote hinterland counties to spread the God’s message,missionaries also help local people build modern education, health systems anddiffuse western science and technology. Such efforts may have contributed toaccumulation of human capital in local communities and reshaped the social values oflocal people.

There are two reasons why studying the long-term effects of Protestantism inChina is valuable. First, China is a large country with homogeneous culture andinstitutions but large variations in local economic performances. China in history waslong dominated by Confucianism. Christianity was prohibited until in the latenineteenth century when China was defeated by western powers and forced to openup. Since then, foreign missionaries self-selected to go to different counties andspread the God’s message. Therefore, the transplant of Protestantism to a large andperipheral country such as China was to a large extent a quasi-natural experiment,which allows us to examine the within-country variations in religious activities andeconomic performances.

Second, since 1978, China’s economy has been keeping a stunning growth spurtwith an almost double-digit annual GDP growth rate. People tend to attribute thisso-called “China miracle” entirely to the Reform and Open-up, a radical institutionalchange beginning in 1978, and try to define a Chinese model” of growth or toestablish a “Beijing consensus” of development-enhancing policies (e.g. S. Philip Hsu,Yu-Shan Wang and Suisheng Zhao, 2011). Such thinking assumes that China’seconomic, political, and social circumstances sufficiently resemble prevailingconditions in other low-income nations so that application of Chinese policies mayproduce something akin to recent Chinese outcomes. But the deep historical rootssurrounding important features of China’s current institutions and the central role ofChina’s legacy of human capital undermine this approach. We argue in this paper thatChina’s Reform and Open-up since 1978 was to a large extent a continuation of themodernization movement since the mid nineteenth century, but was disrupted by warsand revolutions. Missionaries were in fact pioneers of the modernization movement.They disseminated western science, technology and ideology to even the most remotecounties in China, fostered the accumulation of certain types of human capital thatwere conducive to modern economic growth. Such human capital was suppressedduring the Cultural Revolution, but revived quickly since 1978, and paved the way for

4

the institutional changes and rapid economic growth. On this regard, our researchcontributes to a large body of literature of that study the persistent effects of historicalevents on current economic developments, such as Acemoglu, Johnson and Robinson(2002; 2005) and Nunn (2009).

In the next section, we provide some historical background on the spread ofChristianity during the late 19th and early 20th century. Section III describes the datawe use to study the long-term effects of the spread of Christianity. Section IVprovides OLS estimations to assess the relationship between the spread of Christianityand today’s economic performances and then use an instrumental variable approach toaddress the issue of causality. Section V explores the possible mechanisms for thelong-term effects. The final section discusses the implications of our findings andconcludes.

2. Historical Background

Christianity was never a major religion in China, a Confucianism-dominatedcountry. It began to spread in China as early as in the seventeenth century. However,Pope Clement XI forbade Chinese Christians to engage in Confucius-related activitiesin 1704, and this outraged Emperor Kangxi. As a result, he completely bannedChristian activities in 1720. This policy was reinforced by the successive emperors.Since the early nineteenth century, the trade between Europe and China reached ahistorically high record, and Christian missionaries attempted to penetrate Chinaagain. But as Christian activities were still banned by the Qing court, Christianpresence was still negligible up to the 1840s.

China was forced to open up to the western powers when China was defeated byBritain in 1842. As a result, missionaries as well as other foreigners were allowed tolive in China in 1846. Defeated again in 1860, the Qing government signed theConvention of Peking which granted freedom of religion in China and allowed themissionaries to own lands and build churches. A large number of Protestantmissionaries came to different parts of China from Europe and America. By 1900,there were more than 80 thousand Protestant converts (Wang, 1991).

Rapid expansion of Christianity accumulated mistrust and opposition betweenmissionaries and local residents. There are three reasons of mistrust and opposition.First, the monotheism of Christianity was not accepted by Chinese people whobelieve the polytheism and worship ancestors. Second, the egalitarian tradition ofChristianity clashed with the entrenched hierarchical tradition in the Chinese society.Lastly and more generally, western customs and lifestyle brought to China with theexpansion of Christianity were drastically different from Chinese culture. Mistrustand opposition led to frequent conflicts, and these conflicts culminated in the BoxerRebellion in 1900. In this tragic xenophobic conflict, more than 20 thousandChristians were killed, and three fourth of churches were destroyed. This rebellion ledto a war between the Qing government and the Eight Powers. China was defeated

5

again and was forced to sign the Boxer Protocol in 1901 in which the governmentguaranteed safety of the missionaries. Protestant missionaries and activities in Chinawere well protected since then.

The first two decades of the twentieth century witnessed an extraordinarily rapidexpansion of Christianity in China. 1,500 foreign Protestant missionaries arrived inChina in 1900. The number climbed to 3,445 in 1905 and 8,000 in 1927. There wereabout 80 thousand Protestant converts in 1900, and 130 thousand in 1904. In 1922 thisnumber soared to 402,539. By the time of 1920, there were more than 120 Protestantdenominations in China, and church activities penetrated nearly 70 percent of Chinesecounties (Wang, 1991).

The unprecedented expansion of Protestantism in China in the early twentiethcentury was not purely a consequence of improved security condition. The changedmissionary strategy after 1900 was a major cause of surging Christian activities inChina. Learning the lesson from the tragedies in the Boxer Rebellion, many Christianorganizations came to realize that the hostility toward Christianity was largely causedby their condescending way of conducting missionary work. As a result, almost all themajor denominations agreed to hold a historical conference, the Centenary MissionaryConference in 1907. In this conference, many missionaries resolved to take measuresto localize themselves and gain trust from local people, and agreed that an effectiveway of building trust was disaster relief.2 Therefore, the conference officially honoredthe past disaster relief services, and for the first time appealed for concerted efforts inthe ongoing relief work.3 Since this conference, disaster relief became one of thecentral strategies in Protestant missionary work in China.

China is a country frequently inflicted by natural disasters. In the long history offighting with disasters, the Chinese government had developed a set of quitesophisticated disaster relief mechanisms. However, as the central empire was fallingapart in the late nineteenth century, the government’s disaster relief mechanismsgradually collapsed. Noticing the absence of government support, missionaries beganto raise fund and bring food and other resources to people in drought-stricken areas. Acatastrophic drought hit five provinces of northern China between 1876 and 1879,followed by a famine in which about one million people died of hunger. Over 100missionaries went to those areas and helped those in need (Gu, 2004, p289). This wasone of the earliest famine relief operations by missionaries in China. Inspired by therelief work in this famine, some missionaries began to sporadically engage in otherfamines as well. After the Centenary Missionary Conference in 1907, the disasterrelief work became much more organized. Different denominations set up committees

2 For example, a renowned missionary, F. Harmon, said in the conference that “At the end of the famine reliefoperations, it was felt that the great spirit of inquiry and readiness to listen to the Gospel, which was manifested,and should be taken advantage of” (China Centenary Missionary Conference, Shanghai, 1907, p526). Timothy Lee,a famous missionary then, said that disaster relief was “an ideal way to reduce prejudices and prepare the ways forthe Chinese to accept Christianity” (Gu, 2010).3 For example, the conference records noticed that “the whole world is being called upon to send food and moneyto relieve the famine stricken millions in Central China. This conference is being appealed to almost daily forvolunteers to go to aid in the distribution of relief. No one questions the Christlikeness of this work” (ChinaCentenary Missionary Conference, Shanghai, 1907, p493).

6

to coordinate their relief operations in the major disasters in 1907, 1910-1912 and1917-1918. In 1920, northern China was struck by another severe drought. In order tocoordinate international disaster relief efforts, missionaries set up a nation-wideorganization, China International Famine Relief Commission and raised over 30million US dollars for the famine (Gu, 2004, p294).

Due to lack of modern medical knowledge and functioning public health system,cholera, plague and other epidemics decimated thousands of people in the disasters.Missionaries made great efforts to establish hospitals and public health system inthose areas trying to control diseases and save life. As Christian missionaries becamemore deeply engaged in local communities in disaster relief, they reported that manygovernment officials and gentry elites were experts of Confucius classics, but ignorantof modern science and technology. This ignorance was a big impediment of effectivedisaster relief. As a result, they tried hard to persuade Chinese government to buildmodern education system. Moreover, many missionaries put the idea of educationreform into practice, and established a large number of elementary schools, middleschools, and even universities. Many of these Christian schools soon becameexemplary models that secular schools learned from. These achievements improvedthe human capital of local people, and might also have changed the mindset and socialvalues of local people. Such effects survived the political turmoils in the ensuingyears and persisted to the present, and might have significantly contributed to theeconomic performances today.

.

3. Data

To examine the impacts of Protestant activities on long-run economicperformances, we assemble three county-level data sets: Christian activities andsocioeconomic conditions in 1920; socioeconomic conditions in 2000; andgeophysical conditions of all the counties in the data. In order to match historical datawith contemporary data, we have to handle the issue of changing county borders withGIS method. We explain the details in this section.

3.1 Historical Data

Our data of Christian activities in China in 1920 are obtained from a publishedstatistical report: The Christian Occupation of China: A General Survey of theNumerical Strength and Geographical Distribution of the Christian Forces in China("COC" afterwards). The report was compiled by the China Continuation Committee,the central organization of Christian churches in China. One main task of thisorganization was to coordinate and promote more effective evangelization in China.For that purpose, beginning from 1918, the Committee spent three years in conductinga county-level survey on the status quo of Christian activities in China and publishedthe survey results in “COC”. “COC” reports various measures on Christian activitiessuch as numbers of missionaries, converts, churches, and Christina schools and

7

hospitals in Chinese counties. It also collects and publishes other importantsocioeconomic information of surveyed counties from county gazetteers and varioussurvey materials by local governments, such as political orientation, population, andeducation condition. The Committee made special efforts to ensure the quality ofsurvey data in order to deliver a reliable and objective description of the situation ofChristian influence in Chinese counties. Therefore, COC is considered to be a goodsource to study socioeconomic conditions in Chinese counties in the early twentiethcentury.

[TABLE 1 ABOUT HERE]

In this paper we measure the intensity of Christian activities in each county bythree indicators: the numbers of Christian converts, churches and vicars for every onethousand people. The summary statistics are reported in Table 1. On average therewere 0.74 converts, 0.016 churches, and 0.024 vicars per 1,000 people in eachChinese county in 1920. Figure 1 shows the density of Christian converts acrosscounties in our sample. Darker grids are the counties with higher densities ofChristian converts. Although many coastal counties had higher density of Christianconverts, the figure suggests that there were also active Christian presence in quite anumber of inland counties. Therefore, our data suggest that the different intensitylevels of Christian activities were not simply driven by the distances of these countiesto the coastline.

[FIGURE 1 ABOUT HERE]

3.2 Contemporary Data

We assemble contemporary socioeconomic indicators of Chinese counties tostudy the long-term impacts of Christian activities. The original sources of these dataare various surveys conducted by the National Bureau of Statistics of China. The 2000county-level GDP, educational and health expenditure data, are obtained from theStatistical Materials of Public Finance of Cities and Counties. Demographic variables(e.g., population and children mortality rate4) and variables of education outcome (e.g.,years of schooling and literacy rate) are drawn from the Fifth National PopulationCensus (2000). The main dependent variable in our paper is GDP per capita in 2000.The mean of GDP per capita is 5,584 yuan (about $800). Figure 2 reports GDP percapita of all the counties in our data set, and each grid represents each county. Thedarker the grids are, the higher GDP per capita the counties have. Similar to Figure 1,it is evident that not all rich counties are coastal ones. Quite a number of inlandcounties also have high GDP per capita.


3.3 Climate Data

Our historical climate data are drawn from the Distribution Gallery of Droughts

4 2000 population census ask each women between age 16 and 60 about the total number of children they givebirth to, and the total number of children that passed away. “Children morality rate” is defined as the ratio of thelatter over the former.

8

and Floods in the Past Five Hundred Years of China. This gallery was compiled bythe Institute of Synoptic Meteorology and Climatology under the administration ofChina Meteorology Administration (“CMA” hereafter). The original sources ofhistorical climate data are county gazetteers and official archives. Combininghistorical and contemporary climate data, this gallery allows us to build a data set ofdroughts and floods for 120 stations from 1470 to 2000. For each station in each year,the gallery uses a 5-scale Drought-Flood Index (“DF-index” hereafter) to categorizelocal climate: 1 for serious flood, 2 for flood, 3 for normal condition, 4 for drought, 5for serious drought.5 This data set has been widely used in various studies, and itsconsistency and reliability have been carefully examined and confirmed by manymeteorologists (Yao, 1982; Ronberg and Wang, 1987).

In our paper, we count the incidence rate of DF-index=1 or DF-index=2 for astation over our sample periods and use it as a measure of flood frequency.6 Similarly,we construct drought frequency by counting the incidence rate of DF-index=4 orDF-index=5.7

Our climate data are collected from only 120 stations. We use a conventionalapproach, “inverse-distance-weighted method” (IDW), to convert station-leveldrought and flood frequencies into county-level variables. It assumes that a county’sclimate is an average outcome of the climates of all nearby stations, weighted by thedistances from this county to nearby stations. We report the details of processing theclimate data in Appendix 1.

3.4 Geophysical Data

We control for county-level geophysical characteristics in our multivariateanalyses. These data are drawn from the Surface Meteorological Database constructedby the CMA. The database contains annual precipitation and temperature for 754meteorological stations between 1990 and 2000. We use the same IDW method toconvert station-level variables to county-level ones.

3.5 County Matching

A big challenge in our study is that the territories of many counties changedbecause of regime changes over this long historical period. We solve the problem bycomparing the GIS data for both 1920 and 2000 and converting county-level variablesin 1920 to those in 2000 using overlapping area as weights. Historical GIS data areobtained from the China Historical GIS Project by Harvard University and FudanUniversity.8 County GIS data in 2000 (including longitudes and latitudes) areobtained from the ACASIAN Data Center at Griffith University in Brisbane, Australia.Average altitudes of these counties are obtained from the SRTM 90m DigitalElevation Data (Jarvis et al., 2008). As a result, we build a data set combining

5 See Zhang and Crowley (1989) for detailed description of this categorization method.6 In different specifications we test our results for three periods: 1800-1840, 1900-1920, and 1978-2000.7 Using incidence rate of DF-index=1 for flood and incidence rate of DW-index=5 fordrought does not alter themain results.8 The historical county-level GIS data can be downloaded at http://www.fas.harvard.edu/~chgis/.

9

historical Christian intensity data, historical and contemporary climate data,contemporary socioeconomic data and geophysical data for 1,743 counties in Chinaproper.

4. Empirical Results

4.1 Baseline Correlations

We first study the simple correlation between county-level Christian intensities in1920 and economic performances in 2000. Figure 3 shows the correlation between lognumbers of converts per 1,000 people in 19209 and log per capita GDP in 2000 for allthese counties. It is evident that these two variables have a strong positive correlation.


We further examine this relationship with OLS regressions controlling for othercounty-level characteristics. The baseline specification is given by:

ln(pcGDPi)= α + β‧ln(convertsi/populationi) + Xi'‧γ+ εi , (1)

where ln(pcGDPi) is the natural log of GDP per capita in county i in 2000;ln(convertsi/populationi) is the natural log of Christian converts per 1,000 people in1920; Xi is a vector of control variables including county locations, climates, andother geophysical characteristics; and εi is the error term.

Table 2 reports the results. Column (1) contains only the variables of interest.This result reflects the relationship shown in Figure 4. In Column (2) we include a setof regional fixed effects.10 While the magnitude of the marginal effect is reduced, thecorrelation between Christian activities and GDP per capita remains positive andsignificant at 1% level. In Column (3), we control for a set of county geophysicalvariables, including longitudes, latitudes and altitudes, and distances to provincialcapitals. We further control county climates by adding in precipitation andtemperature in column (4), and contemporary frequencies of extreme weather(1978-2004) in column (5). The coefficients on Christian activities remainsignificantly positive.


4.2. Establishing Causality: Instrument Variables Results

The positive correlation between Christian activities in 1920 and contemporary

9 For counties with zero convert (272 such counties in our data), their natural logs are not defined. Instead, we useln(1/populationi), pretending that these counties have 1 convert - the lowest value of the non-zero sample. We alsotry alternative methods to deal with the logarithm transformation of the zeros, including (1) assumingconverti/populationi = 0.001 for zero-convert counties – the lowest value of converti/populationi for among thenon-zero sample, and (2) conducting the analyses excluding the zero-convert counties. The main results remainrobust.10 China proper is divided into 7 regions: North China includes Beijing, Tianjin, Hebei, Shanxi; East Chinaincludes Shanghai, Jiangsu, Zhejiang, Shandong, Anhui; Northeast China includes Liaoning, Jinlin, Heilongjiang;Middle China includes Hubei, Hunan, Henan, Jiangxi; Southern China includes Guangdong, Guangxi, Hainan,Fujian. Southwest includes Sichuan, Chongqing, Guizhou, Yunnan. Northwest includes Shanxi, Gansu, Xinjiang,Qinghai, Ningxia.

10

economic performances in 2000 reported in Table 2 does not guarantee a causalrelationship. One major concern for causality is that selection of counties forpreachment in 1920 was not random. It is possible that missionaries preferred topreach in richer counties in 1920 and these counties are still richer today. In thissubsection we study the location choices of Christian preachment before 1920 fromdifferent perspectives.

First, historical evidence suggests that a considerable amount of missionariesactively worked in the less developed hinterland of China. Hudson Taylor, a famousBritish missionary, believed that, facing life hardship, people in poor and remoteregions were more likely to respond to the calling of God (Wang, 1997). Manymissionaries shared the same view and volunteered to live and work with local peoplein the poor and isolated counties. A good example is China Inland Mission (“CIM”hereafter), a Protestant denomination founded by Taylor in 1865. The goal of CIMwas to conduct missionary work in all the inland provinces. It was the first Protestantdenomination arriving in Shanxi (in 1876), Sichuan (in 1877), Guizhou (in 1877) andYunnan (in 1877) (Broomhall, 1901).11 CIM’s goal was supported by manymissionaries. It became the largest Protestant denomination in the early twentiethcentury. More than 1000 missionaries joined CIM and spread the God’s gospel inremote villages. Many other denominations followed CIM, and left footprint in almostevery part of this big county.

To empirically examine if missionaries were more active in the less developedcounties, one needs the economic indicators of those counties in 1920. However,direct measures of historical economic performances are unavailable. Instead, we usetwo proxies for historical economic conditions: population density and per capita landtax revenue. Population density is a commonly used, although quite crude, proxy forhistorical economic prosperity.12 Per capita land tax revenue can also largely reflectlocal economic conditions.13 Since both these two measures are relatively crude, weonly consider the results here suggestive. Figure 4 plots these two measures againstthe number of churches and vicars per 1,000 people for Chinese counties in 1920.14

All figures show significant negative correlations, suggesting that Christians weremore active in less developed regions. Given this evidence, it is unlikely that theobserved positive relationship between Protestant activities and current economicperformance is completely driven by selection. In fact, selection tends to bias the OLSresults towards zero.


11 Actually it is the only Protestant mission in province Guizhou and Yunnan until 1900.12 See Acemoglu, Johnson and Robinson (2002) for example. Our measure of county-level population also comesfrom the COC.13 Prefectural-level land tax data in 1820 are obtained from Liang (2008). Land tax in the Qing dynasty wascalculated based on size and quality of arable lands multiplied by a fixed tax rate. After normalized by population,it reflects the development of local agriculture sector, which to a large extent indicates overall economic conditionsin a traditional society.14 We use the number of missionaries and the number of churches in each county because we want to measure thelocation choices of Christian missionaries before 1920. As a robustness check, we also use the number of convertsand find similar results.

11

In order to rigorously examine the causality between historical Christianactivities and contemporary economic performances, we use instrumental variabletechnique to formally address the potential selection problems. As described inSection II, since the beginning of the twentieth century, Christian organizationsdecided to use disaster relief as one of the main methods to gain trust from Chinesepeople and better conduct missionary work. In light of this fact, we constructcounty-level incidence rate of droughts and floods between 1900 and 1920 and usethem as instruments for intensity of Christian activities.


We report the 2SLS estimates of equation (1) in Table 3. Column (1) reports theresults without any control variables. Column (2) includes regional fixed effects.Column (3) adds in county level geophysical characteristics. Column (4) and (5)further include contemporary climates and contemporary frequencies of droughts andfloods.

The second-stage estimation results are reported in Panel A. Coefficients onln(convert) remain positive and significant across different columns, ranging from0.16 to 0.19. These results are significant not only statistically but also economically.The standardized beta coefficients reveal that a one-standard-deviation increase inln(convertsi/populationi) leads to 0.44 to 0.52 standard deviation increase in log GDPper capita. According to the estimates in column (5) with full control, a county withthe mean of income at 5,584 yuan, one more convert per 1,000 people can lead to anincrease of 1168 yuan (about 180 US dollars) in per capita GDP. Moreover, estimatesof β in Table 3 though 2SLS are significantly higher than those in Table 2, whichconfirms the underestimation of impacts of historical Christian activities in OLSregressions.

The first-stage results are reported in Panel B. The coefficients on both flood anddrought frequencies are significantly positive, suggesting more Christian activities inthe counties with more disasters. The F-statistics of the instrument variables is higherthan the critical value suggested by Stock and Yogo (2005), which rules out concernsfor weak instruments. In an unreported placebo test, in addition to the disasterfrequency in 1900-1920, we also control for the disaster frequency during the adjacentperiods: 1860-1900 and 1920-1960. It turns out that only 1900-1920 variables havepositive and significant coefficients. The coefficients on disaster controls in otherperiods are mostly insignificant. The evidence here is consistent with the historicalfact that Protestant churches actively involved in disaster relieves in China as astrategy to gain trust from Chinese society.

One might worry about the validity of our instruments because disasters inhistory might correlate with disasters today and adversely affect today’s economicoutcome. In fact, this channel is unlikely because of the following reasons. First, all ofour specifications pass Sargan over-identification test at conventional statistical levels.Second, we find that the weather between 1900 and 1920 and that between 1920 and1940 are highly correlated, but their correlations with the weather between 1980 and

12

2000 are very weak.15 Third, we control for the current disaster frequencies incolumn (5) of Table 3, and the effect of historical Christian activities remains robust.Moreover, this specification delivers a much higher P-value in the over-identificationtest than other columns, which strongly supports the exclusive restrictions of ourinstruments.16

4.3. Robustness Checks

In Table 4 we conduct a number of robustness checks for our estimates. Column(1) of Table 4 addresses the concerns of potential outliers by omitting potentiallyinfluential observations.17 Compared with our preferred specification in Table 3, thecoefficient on convert density is unaffected.


Coastal counties were easier to reach than hinterlands in the past. Furthermore,coastal regions usually enjoy more preferred policies in today’s China. Therefore, onemight wonder if the long-term impacts of Christian activities are merely driven bycoast-hinterland disparity. In column (2) we examine this issue by taking out allcounties in coastal provinces. The effects of Protestant activities become evenstronger for the hinterland sample, suggesting that the mechanisms through whichmissionary work affected long-run economic outcomes were more effective for inlandcounties.

A small number of counties in our full sample are in fact big cities. For bothgeographical accessibility and economic policy reasons they could be positivelycorrelated with both Christian activities and present economic performance. Incolumn (3) we drop the urban counties and focus only on the rural ones. Ourestimation results on the long-term effects of Christianity are barely changed.

In addition to the application of instrument variables to tackle the selection issue,in column (4) of Table 4, we also add in control for historical economicheterogeneities across counties. We use land tax per capita to proxy for initialeconomic development of each county.18 The effects of Christian activities becomeslightly bigger than those in the main specification (column 5 in Table 3), which isreasonable due to the nature of selection bias discussed in Subsection 4.2.

Catholic missionaries were also quite active in China during the period we study.However reliable and detailed data on Catholic activities are not available. As a resultwe focus on the effects of Protestantism instead of Catholicism in this paper. As a

15 The correlation coefficients of droughts and floods between the period of 1900-1920 and the period of1980-2000 are only 0.19 and 0.095 respectively, but the correlation coefficients between the period of 1900-1920and the period of 1920 and 1940 are 0.4 and 0.5. The weak correlation between disasters in history and todaymight be resulted from improved irrigation systems.16 We should also keep in mind that, different from one hundred years ago, China’s economic performances todayare less sensitive to weather shocks because the main sector of Chinese economy today is manufacture instead ofagriculture. In addition, with vastly improved communication, transportation and weather forecasting techniques,the impacts of extreme weather on economic outcome today are much smaller.17 “Potentially influential observations” are identified using Cook’s distance, which measures how much each ofthe estimated coefficients change when each observation is deleted. If an observation has a Cook’s distance greaterthan 4/N, where N=1742 is our sample size, it is excluded from regression in Column (1) of Table 4A.18 Using population density in 1920 as a proxy for economic development delivers similar results.

13

robustness check, we include the number of prefectural-level Catholics missionstations as a control variable in order to rule out the possible effects of Catholicactivities.19 The result is reported in column (5) of Table 4. The coefficient on theProtestant converts is reduced but remains positively significant.

In the main specification we use the number of converts instead of churches orvicars to measure the intensity of Christian activities in each county. We do sobecause the number of converts captures the outcome of Christian preachment, whilethe numbers of churches and vicars captures the inputs. As robustness checks, incolumns (6) and (7) of Table 4 we report the 2SLS results using numbers of churchesand vicars, and find that our main results are not affected. We also find that thecoefficients of churches and vicars are bigger than that of converts. This is becauseone additional church or vicar should yield much larger effects than one additionalconvert.

5. The Effects of Protestant Activities on Education

Many historians notice that, while conducting disaster relief, Protestantmissionaries were actively engaged in helping local Chinese establish moderneducation and medical system in the early twentieth centuries.20 In the next twosections we discuss how their efforts in promoting education and medicine generatedprofound and long-lasting effects in the Chinese society.

Exactly as John K. Fairbank noted, “in the end the Christian influence wasprobably strongest in education” (Fairbank, 1974, p13). China completely rebuilt theeducation system in the early twentieth century, and missionaries played a critical rolein this rapid transformation. Christian missionaries soon realized the fundamentalflaw of traditional Confucius education after they came to China in the earlynineteenth century.21 In the process of disaster relief, missionaries found that a mainobstacle of their work was lack of modern science and technology among manyChinese people, including some elites, and the main cause was believed to betraditional Confucius education. Therefore, building modern education systembecame an imperative task. For instance, a well-known missionary, Timothy Lee,wrote a proposal to the governor of Shanxi Province when he was conducting disasterrelief in 1884, emphasizing the importance of reforming China’s education system.He said, “Education is the first priority for China. As the western countries keepdeveloping their education day-to-day, China will lose his chance to overtake them in

19 We thank Ying Bai and James Kung for generously providing us with the data on Catholic mission stations. Theoriginal information was shown in a crude map included in the appendix of Stauffer (1922), which contains noprefecture boundary information. Bai and Kung (2011) locate the data to all the prefectures by their relativepositions in the original map. Given the way of constructing the data, the results with this measure of Catholicactivities is only suggestive.20 For example, see Cohen, 1978, p548.21 E. C. Bridgman was believed to be the first American missionary working in China. As early as in 1830, hebegan to criticize the drawbacks of the traditional education and examination system and call for the establishmentof a modern education system (Barnett and Fairbank, 1985, p. 100).

14

10 years. In conclusion, education is the most significant and urgent thing for China”(Lee, 1889).

Christian missionaries determined to build new school system in China for thepurpose of not only educating Chinese, but also converting Chinese. They believedthat western education was crucial to dissipate mistrust and misunderstanding ofChinese people toward Christianity. Alvin Pierson Parker, chairman of theEducational Association of China, explicitly expressed this opinion in 1896, “as aChristian educators’ association, we should play a dominating role in China’seducation reform and satisfy the interests of Christianity” (Educational Association ofChina, 1896).

Driven by the motives of educating and converting Chinese, missionaries madebig achievements in revamping China’s education. The number of Protestant schoolsrose sharply from 347 in 1877 to 7,382 in 1922, almost tripling every twenty years(Gregg, 1946, p.16-17). Among these Christian schools, 6,599 (86%) were elementaryschools, 291 (7%) were middle schools, 16 (0.2%) were colleges, and 75 (1%) werevocational schools (China Educational Commission, 1922, p. 416).22

Primary schools are the foundation of modern mass education. To empirically testthe effects of Christian activities on the development of modern primary schools, weregress, at the county level, the number of students per 1,000 people in Christianprimary schools on the density of Christian converts,23 the latter of which wasinstrumented by the frequencies of historical droughts and floods. We report theresults in column (1) of Table 5. Because our dependent variable is censored at zero,we estimate a Tobit model instead of OLS.24 The IV result in panel A indicates astatistically and economically significant effect of Christian activities on theenrollment rate of the Christian primary school: transforming the coefficient intomarginal effect (evaluated at the sample mean) suggests that one standard deviation inconvert per 1,000 people increases the enrollment rate of Christian primary school by0.92 standard deviations.

[Table 5 is about here]

Christian missionaries were not only building more schools, but more critically,different schools. They introduced new curriculums which put a big emphasis on bothnatural sciences, such as mathematics, physics and chemistry, and social sciences,such as law and business. Practical subjects such as foreign language studies andengineering were also included. English, for example, was offered to 58% of thestudents in the 4th or 5th grade (Stauffer, 1922, p. 1075). Class lectures werecombined with practices and laboratory experiments. As a result, these new schoolsenjoyed big advantage over traditional schools in training students qualified forbooming industrial and commercial sectors.

22 The rests are special institutions, including orphanages and schools for the blind or deaf.23 If one assumes that portion of school-aged children is constant across counties, then this result can be viewed asclear evidence for the casual effects of Christian activities on the enrollment rate of the school-aged children inChristian elementary schools.24 The OLS results are not significantly different from the Tobit ones.

15

Another big advantage of Christian schools was funding. According to a report onthe status of Christian schools in China, over half of the funding of Christian schoolscame from foreign Protestant organizations, which were usually quite stable andsufficient (Stauffer, 1922, p. 1094).25 As a sharp contrast, because of frequentpolitical turbulences and civil wars, public schools were often plagued by insufficientfinancial support from the government. For example, in 1911, the education budgetaccounted for merely 1.5% of the total government spending (Stauffer, 1922, p.1068).26 In the early 20th century, teachers and administrators in the public schools,and even staffs of the Ministry of Education frequently protested for arrears in salary(China Educational Commission, 1922, p. 22).

Better schools educate better students. Take student promotion rate as an example.In 1920, 21% of the students in Christian junior elementary schools entered seniorelementary schools (as opposed to 10% for the public school students), and 10% ofthe students in Christian senior elementary schools entered middle schools (asopposed to 3.4% for the public school students) (Stauffer, 1922, p. 404). Moreover,Christian schools educated a large amount of professionals who were urgently neededby the society. Stauffer (1922, p. 409) reports a surveyed on 5,500 high schoolgraduates in 1918: 30% of them continued their study in college; 30% of them workedfor churches; 20% of them went to teach in other schools; and the other 20% went forbusiness or other professions such as doctor, nurse and lawyer. By contrast, 70% ofthe graduates from public middle schools had difficulty in finding jobs (ChinaEducational Commission, 1922, p. 19).

As Christian schools became so successful, they set an exemplary model forChinese public schools. Wang (1997) documents three spillover effects of Christianschools to China’s education system. First, combining western origins with Chinesereality, most textbooks edited by missionaries were well customized to suit the needsof Chinese students. Therefore, their textbooks were widely used in many publicschools (Wang, 1983, p112). Second, a large number of graduates from Christianschools became teachers in public schools and taught subjects that were in urgentneed of those schools.27 Third, both the teaching and the organization methods ofChristian schools greatly influenced Chinese officials and educators, who put similarmodels into practices later.28 As being said in one of the reports from a missionaryeducation association of China, “by perfecting and strengthening this arm of theservice (Christian schools), we increase the probability that the future governmentaleducational system of China will be largely influenced and molded by such superiorexamples.” (Silby, 1902, p 621)

Empirically, we corroborate the spillover effects of Christian schools in Chinawith two pieces of statistical evidence. First, we investigate, within a county, whether

25 31% came from tuition fees, and the rest of 18% came from domestic donations.26 60% of the government budget was diverted to military purpose and war indemnity.27 For example, graduates from Tengchow College, a famous school established by the American MissionaryCalvin Mateer, showed high competiveness in their job markets among Chinese schools (Wang, 1983).28 For example, when Zhang Zhidong, the Huguang Governor, planned to establish modern schools in the early1900s, he sent many of his officials to Boone Memorial School, a famous Christian academy established byAmerican Episcopal Church in 1871, to study its education and management model (Wang, 1983, p113).

16

more intensive Christian activities induced higher enrollment rate for publicelementary schools. The results in column (2) of Table 5 support this conjecture. Inparticular, our IV Tobit coefficient implies that one standard deviation in convert per1,000 people increases the enrollment rate of public elementary school by 0.70standard deviation. In other words, 1 more convert per 1,000 people increases theenrollment rate of public elementary schools by 0.58% (the mean value is 1.19%).

Next, we examine, across counties, whether the enrollment rate of publicelementary schools in one place was influenced by the Christian activities in thenearby counties. Specifically, consider the following specification:

ln(Student/pop1920i)= α +β1 ln(Christianityi)

+β2 ln(Christianity_Contiguousi) + Xi'γ+ εi , (2)

where the enrollment rate of elementary schools in county i (Student/pop1920i) isregressed on both the Christian activities in its own location (Christianityi) and theChristian activities in the neighboring locations (Christianity_Contiguousi). The latteris constructed by averaging the convert densities in all counties contiguous to county i,weighted by the distance of each of these counties to county i. The instrument for thisvariable, which is the average frequency of droughts and floods in the nearby counties,can be constructed in a similar fashion.

Column (3) in Panel B reports the OLS results of β1 and β2. They are bothsignificantly positive, and the effects of Christian activities in the own counties arelarger than those in nearby counties (β1>β2). However, once we estimate the modelwith instruments (column (3) in Panel A), the neighboring effects becomeinsignificant. We further investigate this issue by dividing the overall sample into twogroups: counties with Protestant converts and counties with no converts. Their resultsare reported in column (4) and (5) of Table 5, respectively. Notice that in column (5),only β2 can be estimated due to the construction of the subsample (Christianityi=0).The results show that the influences of Protestantism in neighboring counties are notsignificant for the counties with their own Protestant activities, but significant for thecounties without Protestant presence. Specifically, for the latter counties, theirelementary school enrollment rates increase by 0.97 standard deviation if the convertpopulation density in their contiguous counties increases by 1 standard deviation. Thatis to say, increasing 1 convert per 1,000 people leads to the increase in the enrollmentrate by 1.61% (the mean value is 1.06%).

6. The Effects of Protestant Activities on Health Care

6.1 Hospitals and Clinics

Herbal therapy was the main form of medical treatment in China until westernmedicine was introduced in China in the nineteenth century. Christian missionariesplayed a critical role in the introduction of western medicine into China. Peter Parker,

17

an American missionary, built up the first modern hospital in China in 1837. Morethan 2,000 patients received treatment in the first year of its operation (Bush, 1879).Since then modern medical system gradually flourished in China. By 1889, 61Christian hospitals and 44 clinics were operating in China.29 This number was morethan doubled in the first two decades of the twentieth century, reaching 326 and 244respectively (Stauffer, 1922, p96). The annual number of inpatients soared to nearly150 thousands and that of outpatients over one million (Stauffer, 1922, p623). By1937, over 300 Christian hospitals were established in China, providing with morethan 20 thousand beds. These hospitals were not only located in coastal provinces, butmore in hinterlands. Many hospitals offered free medical services to the poor and theneeded. John K. Fairbank, a renowned historian, commented on this, “ModernWestern medicine in China was to an important degree a consequence of missionarydemonstration and instruction” (Fairbank, 1983).

Figure 5 presents the correlation between Christian activities and the numbers ofhospitals and pharmacies at provincial level.30 The positive correlations are verystrong: 1% increase in the number of converts is associated with 1% greater in thenumber of modern hospital and 0.4% greater in the number of modern pharmacies.

[Figure 5 about here]

With strong support from Christian organizations, these Christian hospitals wereusually equipped with advanced facilities and were well financed. Dr. Harold Balme,the dean of the School of Medicine in Shandong Christian University, surveyed thesituation of 165 Christian hospitals in 1919. His study found that most of the surveyedhospitals owned medical laboratories, 75 hospitals had the capability to conductlaparotomy, and 24 had been equipped with X-ray machines (Balme and Stauffer,1920).

6.2 Public Health

What the missionaries brought to China was not only western hospital system, butmore importantly the idea of western public health. Whenever large epidemics brokeout in disaster-struck areas, missionaries were usually the foremost in organizingeffective remedial and preventive measures. For example, in 1872, cholera struckTianjin. Missionaries provided medical aid to people who were infected. In1911-1912, a highly deadly pneumonic plague broke out in northeast China. A groupof missionaries set up the Anti-Plague Bureau to fight with the plague.

In frequent disaster reliefs, Christian missionaries found that lack of public healthknowledge and consequent outbreaks of epidemics were main reasons of large deathtoll in the disasters. Disseminating knowledge of disease control became a priority inthe disaster relief. The prevention measures included sterilization of medical

29 Records of the General Conference of the Protestant Missionaries of China, held at Shanghai, May 7-20, 1890.(pp. 733), published by General Books LLC.30 The main reason we cannot conduct this analysis at the finer geographical level is that the number of hospital in1920 is only available at the provincial level.

18

equipment and tools, disinfection of food and drinking water, control of flies andmosquitoes, etc. Missionaries enforced quarantine in disaster struck areas andeffectively controlled expansion of contagious diseases. According to a study by Dr.Balme in early twentieth century, 69 Christian hospitals (42% of total) were equippedwith quarantine facilities (Balme and Stauffer, 1920). A good example is control ofleprosy. Local governments and missionaries collaborated to enforce quarantine lepers,and the number of leprosy cases reduced drastically in the early twentieth century(Stauffer, 1922, p437-438).

Health conditions of women and children were critical in public health.Missionaries worked hard to help improve health conditions of women and children.Female missionaries played a critical role in disseminating knowledge of obstetricsand gynecology among Chinese women. They made regular visits to pregnant womenand provided medical services in case of necessity.31 Because of enormous efforts offemale missionaries, obstetricians, midwives and nurses began to practice in manyplaces in China, resulting in significant decline of women and infant mortality in theearly twentieth century.

6.3 Medical Education

Missionaries were also pioneers in promoting modern medical education in China.They believed that “scientific medicine in China must not continue indefinitely to be a‘foreign doctrine’” and “the medical profession of China must become national if it isto be universally accepted” (Mac Alister, 1921). As a result, Christian organizationshad established 116 medical education institutions by 1920. Among these institutions,10 were medical colleges, which accounted for one third of all the medical colleges inChina (Stauffer, 1922, p.425). Most of these colleges continued to operate today andtrained a large amount of excellent doctors. The other 106 institutions were nurseschools (China Educational Commission, 1922, p. 416). Usually these schools werelarger and equipped with better facilities than public medical schools. In fact, theseChristian medical schools provided an advanced model of medical education thatpublic schools were trying to follow.

Setting up schools was not the only way that missionaries promoted medicaleducation. They translated a large number of western medical books to Chinese. Afamous missionary, John Glasgow Kerr, published the first medicine journal in Chinain 1868.32 A few years after that, the Medical Missionary Association of China, set up

31 For example, as early as in 1888, seven Canadian missionaries conducted a comprehensivesurvey of health conditions of women and children in Zhangde Prefecture, Henan Province.Meanwhile, with the fund raised in the home country, they set up gynecological clinics andprovided free medical services to local women and children. They also provided monthly trainingclass to women on the knowledge of obstetrics and gynecology. See Song (2008, p. 31) for details.32 John Glasgow Kerr (1824 – 1901) was a Presbyterian medical missionary. He came to Chinawith the American Presbyterian Mission in 1854, and soon became the head of the OphthalmicHospital in Canton and later the Guangzhou Boji Hospital (The Canton Hospital). He workedthere for 47 years and treated about 1 million patients. Besides his outstanding contribution inmedical care, he was also famous for his student, Dr. Sun Yat-sen.

19

in 1886, published its own journal in 1888 with a focus on transmitting westernmedical knowledge to the Chinese. Missionaries published the first Chinese medicinedictionaries in 1908, which standardized the Chinese medical terms (China MissionYear Book. 1912, pp. 267-268).

7. Persistent Impacts of Protestant Activities

The status of Christianity overturned since the People’s Republic of China (PRC)was established in 1949. The government regarded missionaries as a form of “westernimperialism”.33 Consequently, the central government established its official churchsystem and severed their connections with foreign organizations. By 1952, all foreignmissionaries left China. During the Cultural Revolution (1966-1976), all religiousactivities, including the government-regulated churches, were banned. Christianityrevived in China since the Reform and Open-up in 1978. However, it is still undertight regulation by the government. Religious services can only be practiced in thegovernment-sanctioned churches, and these churches do not maintain any officialconnection with foreign churches.

Soon after the Communist Party took power in 1949, the administration ofChristian schools and hospitals was handed over to the government. However, manyof these schools were still functioning, trained a large number of students, andpersistently contributed to local economic development. Most of Christian hospitalswere also functioning and benefiting local communities. Furthermore, themissionaries’ endeavors before 1949 promoted the consciousness of education andpublic health of Chinese people, and were greatly conducive to accumulation ofhuman capital in the years to come.

In Table 6 we empirically examine the long-run effects of historical Protestantactivities on the education and health outcome in 2000. Columns (1)-(3) report theresults for three education measures: county-level educational expenditure normalizedby population, average years of schooling and literacy rate. Similar to the previoussection, we use frequencies of droughts and floods from 1900 to 1920 as instrumentsfor county-level numbers of converts in 1920. The first stage results are not reportedhere because they are the same as that of the Column 5 in Table 3.


Panel A reports the second stage of the IV estimates.34 The results in Columns(1)-(3) confirm that the past Christian activities have positive and significant effectson a county’s long-run education achievements. Take years of schooling as anexample, evaluating at the mean, 1 standard deviation increase in converts per 1,000people in 1920 leads to 0.92 standard deviation increase in average years of schooling

33 This point was stressed many times by the Premier Zhou Enlai in the meetings with Chinese Christian leaders.34 For the same reason discussed in Section 4.2, 2SLS coefficients on Christian activities have higher magnitudethan OLS coefficients (reported in Panel B).

20

in 2000.

Christian organizations help local communities build hospitals, introduce modernmedical technologies, and improve public sanitation. We examine whether suchendeavors generated long-lasting effects in current health conditions, measured bychildren mortality rate in column (4) and public health expenditure in column (5).Regression results show that historical Christian activities had significantly negativeeffects on current children mortality rate. 1 standard deviation increase in converts perthousand people decreases the children mortality rate by 0.86 standard deviations.When evaluating at the sample mean, this amounts to 37% decrease in childrenmortality rate. In addition, counties with more intensive Christian activities tend tohave more government spending on public health, but the effects are not statisticallysignificant.

8. Other Possible Channels of Causality

Missionaries conducted a lot of undertakings in the early twentieth century. Wewonder if, aside from education and health care, there was any other channel that theirundertakings persistently affected today’s economic outcome. We control for currenteducation and health outcome in the GDP regression and examine whether the effectsof Protestant activities remain. In principle, we would like to estimate the followingmodel:

200021920222000 )/ln()ln( EdupopulationconvertspcGDP

2220002 XHealth , (3)

where ln(converts/population) is instrumented by historical drought and floodfrequency during 1900 to 1920. In this test, we experiment all the combinations ofdifferent education and health measures and report the results in Table 7. As expected,the coefficients of these controls are positive and significant. For example, column (2)of Table 7 reveals that 1% increase in years of schooling is associated with 0.71%increase in GDP per capita. This result is comparable to existing studies usingcross-country variations.35

More importantly, Table 7 shows that, after controlling for education and healthmeasures, the effects of Christian activities are lower than the results in our baselinemodel (Table 3), but remain positive and significant. For example, in column (8)when both years of schooling and children mortality rate are included, the coefficientof converts is reduced by 38% compared with the last column of Table 3.

[INSERT TABLE 7 HERE]

35 For example, Mankiw et al (1992) conduct cross-country analysis and find that 1% increase in years ofschooling is associated with 0.66% increase in GDP per capita.

21

A potential problem of estimating equation (3) is that current education andhealth conditions may also be endogenous. Unobservable county-level characteristicsaffecting education and health status can also affect economic outcomes, and there isno instrument variable for them in our context.

Following Becker and Woosman (2009), we bound the effect of Christianactivities on GDP per capita net of contemporary education and health outcomes. Thebounding procedure involves two steps. First, we run an OLS regression withequation (3). The results of this auxiliary regression are reported in Table A2. Wechoose years of schooling as a measure of education development and childrenmortality rate as a measure of health outcome. Using other measures delivers similarresults.

In the second step, we take out education and health outcomes, and estimate thenet effects of historical Protestant intensity on today’s GDP in the followingspecification:

22192022200020002000 )/ln()ln( XpopulationconvertsHealthEdupcGDP

, (4)

where and are based on the estimate values of 2 and 2 obtained from

Step 1 (column 5 of Table A2), and adjusted for potential biases reported by otherwell identified studies in the literature.

Card (1999) reviews the literature on the return to education, and concludes thatthe upward bias due to ignorance of ability accounts for about 10% of the OLSestimates. Moreover, studies using institutional changes to create exogenous variationin schooling actually receive estimates 20-40% higher than those of OLS.36 In lightof this, we bound the range of estimates of the economic return to years of schooling

( ) from 80% to 140% of its OLS estimate ( 2 ).

Studies evaluating the effects of health improvement on economic performancebased on regional variations are not quite conclusive (Jack and Lewis, 2009). Theyalways suffer from omitted variable bias and reverse causality (Mankiw, 1995,p303-304). Gallup and Sachs (2001), and Bloom et al (2004) are examples of veryfew studies that construct instrumental variables to solve these endogeneity issues.But their identification strategies are still in debate, and their 2SLS results are notsignificantly different from the OLS ones. As a result, we conservatively choose a

wide range for the economic return of improvement in children mortality rate( ),

from 80% to 200% of its OLS estimate ( 2 ).

36 The main reason is that institutional improvements usually affect people with low education outcomes, whotend to have higher marginal return to schooling. Similar downwards biases in OLS may apply to our context aswell.

22

Each cell in Table 8 reports an estimate of 2 in equation (4), where Christian

activities in 1920 are instrumented by historical disaster frequencies. The titles of

each column and row indicate the values we choose for and . In cell [1,1],

when both and are set to 0, we obtain the point estimate of Christian effect in

Table 3.

[INSERT TABLE 8 HERE]

Column (1) reports the results when we only take out the education channel in thedependent variable of the regression in equation (4). The coefficients on Christian

activities range from 0.1 to 0.14 depending upon the values we choose for , and are

both statistically and economically significant. Such results suggest that educationimprovements account for about 26-47% of the total effect of Christian activities.Row (1) reports the results when we only take out the health channel. With a wide

range of we choose, the coefficients on Christian activities vary from 0.15 to 0.18,

which are also statistically significant. Such results suggest that long-run healthimprovements contribute to about 5-21% of the total effects of Christian activities onGDP per capita today.37

With almost all the combinations of possible values of education and healthcoefficients, the effects of Christian activities are significant, ranging from 0.07 to0.18. Insignificant results only appear when the true effect of return to education is 40percent higher than that of the OLS estimate and the true effect of return to health ismore than 120% of the OLS estimate (Row 7 and Columns 4-7). Even in these cases,the magnitudes of β2 still have significant economic meanings.

The above analysis shows that improvements in education and health account forabout 50% of the long-run effects of Protestantism on economic outcomes. That is tosay, there are other unquantifiable channels through which historical Protestantismaffect today’s GDP per capita, and our studies find that they may contribute to another50% of the total effects. Indeed, aside from education and health care, Protestantismmay have generated a variety of profound effects on the Chinese society. For example,more Protestant activities may have resulted in more open attitudes toward new ideasand technologies, better work ethics, or more entrepreneurship.

An implication of this argument is that the long-run impact of Protestantism mayhave been most conspicuous since China began Reform and Open-up in 1978.Although most effects of Protestantism on attitudes and ethics were suppressed duringthe Revolution era, they revived and became indispensable for economic success in a

37 In fact, it is possible to calculate the threshold value of and beyond which β starts to become

insignificant. The value for β to lose significance at 5% level is 1.8 times of χ2, and the corresponding value is 4 times of δ2.

23

market-oriented environment. To test this, in Figure 6 we plot average GDP per capitabetween 1950 and 2010 for two groups of our sample provinces. Group 1 consists ofthe 11 provinces with the lowest measures of ln(convert/population) in 1920, andGroup 2 consists of the 12 provinces with the highest measures ofln(convert/population) in 1920.38

[INSERT FIGURE 6 HERE]

Two patterns are noticeable here. First, provinces with higher intensity ofChristian activities are on average richer than those with lower intensity. Moreinterestingly, the gap between these two groups widens rapidly in the late 1970s. TheGDP per capita in Group 1 almost doubled that in Group 2. These patterns areconsistent with our hypothesis that the unquantifiable effects of Protestantism onattitudes, ethics, mindset, and so on might have mattered in a central plannedeconomy, but mattered much more in the era of Reform and Open-up since 1978.

9. Conclusion

In this paper, we construct a data set mapping Protestant presence in Chinesecounties in 1920 and socioeconomic indicators for those counties in 2000. Byexploring this large within-country variation, we find that the spread of Protestantismhas generated significant positive effects in promoting long-run regional economicprosperity.

One major identification challenge is that, if counties with better geographical, orsocioeconomic conditions were more attractive to missionaries in the past, and thesecounties continue to have better economic performance in the present, we mightobtain a positive correlation between past Christian activities and present economicoutcomes, even though the former do not have any causal effect on the latter. Wepursue several strategies to tackle this challenge. First, historical and statisticalevidence suggests that missionaries were actually inclined to preach in less developedareas. Therefore, if there is any selection effect, our OLS regression should haveunderestimated rather than overestimated the true effects of Christian activities oneconomic development. Second, in light of the fact that Christian churches activelyconducted disaster relief to gain trust from local communities, we use historicaldisaster frequencies as instruments for Christian activities. Our 2SLS estimatesconfirm and further enhance the OLS results of the positive effects of Christianactivities. Our results are robust to different subsamples, different measures ofChristian presence, and different geographical and climate controls.

We then investigate the channels through which the effects of missionaries’ workpersisted over time. While spreading Protestantism in China, missionaries took activepart in building modern education and medical system. They set up a large number of

38 The Provincial-level GDP data are obtained from China Compendium of Statistics 1949-2008. We use theprovincial-level data because county-level GDP data before 1993 are not available.

24

western-style schools and hospitals, and helped build public health system in China.Tens of millions of Chinese people benefited from their pioneering work. Such effortsmight have contributed substantially to the accumulation of human capital in China inthe past century.

However, we realize that the effects of the spread of Protestantism could becomplex and profound. It is possible that missionaries’ work induced the changes inwork ethics, attitudes toward western culture, entrepreneurship, and so on. Suchchanges might also have boosted long-run economic growth. Our empirical resultssuggest that improvements in education and health account for about half of the totaleffects of missionaries’ work on today’s economic outcomes, and the other half arepossibly the contributions from the changes in culture and people’s mindset.

25

References

Acemoglu, Daron, Simon Johnson, and James Robinson, Yunyong Thaicharoen,“Institutional Causes, Macroeconomic Symptoms: Volatility, Crises and Growth”,NBER Working Paper No. 9124, Issued in August 2002.

Acemoglu, Daron, Simon Johnson, and James Robinson, “Institutions as theFundamental Cause of Long-Run Growth”, NBER Working Paper No. 10481,Issued in May 2004.

Arruñada, Benito, “Protestants and Catholics: Similar Work Ethic, Different SocialEthic”, The Economic Journal, Volume 120, Issue 547, September 2010, pp.890–918.

Bai, Ying, and James Kung, "Diffusing Knowledge While Spreading God's Message:Protestantism and Economic Prosperity in China, 1840-1920." Manuscript, April2011.

Balme, H. & Stauffer, M.T. (1920). An Enquiry into the Scientific Efficiency ofMission Hospitals in China (pp. 430)

Barnett, S. W., and J. K. Fairbank, eds, 1985, Christianity in China: early Protestantmissionary writings, Harvard studies in American-East Asian relations, vol 9.

Barro, Robert J. 2004. “Spirit of Capitalism,” Harvard International Review, 25(4):64-66.

Barro, Robert J, and Rachel M. McCleary, “Religion and Economic Growth acrossCountries”, American Sociological Review, Volume 68, No 5, October 2003, pp.760-781.

Becker, Sascha O, and Ludger Woessmann, “Was Weber Wrong? A Human CapitalTheory of Protestant Economic History”, The Quarterly Journal of Economics,Volume 124, Issue 2, 2009, pp.531-596.

Becker, Sascha O, Peter Egger, and Maximilian von Ehrlich, "Absorptive Capacityand the Growth Effects of Regional Transfers: A Regression DiscontinuityDesign with Heterogeneous Treatment Effects," CEPR Discussion Papers 8474,C.E.P.R. Discussion Papers, 2011.

Becker, Sascha O, and Ludger Woessmann, "Luther and the Girls: ReligiousDenomination and the Female Education Gap in 19th Century Prussia," StirlingEconomics Discussion Papers 2008-20, University of Stirling, Division ofEconomics.

Becker, Sascha O, and Ludger Woessmann, "Knocking on Heaven’s Door?Protestantism and Suicide," The Warwick Economics Research Paper Series(TWERPS) 966, University of Warwick, Department of Economics, 2011.

Becker, Sascha O, and Ludger Woessmann, “The effect of Protestantism on education

26

before the industrialization: Evidence from 1816 Prussia,” Economics Letters,Volume 107, Issue 2, 2010, pp. 224-228.

Cantoni, Davide, “The economic effects of the Protestant Reformation: testing theWeber hypothesis in the German Lands”,Working Papers, 2010.

Cheng, Leonard, and Yum Kwan, “What are the determinants of the location offoreign direct investment? The Chinese experience”, Journal of InternationalEconomics, 51, 2000, 379-400

China Continuation Committee, Special Committee on Survey and Occupation,Milton Theobald Stauffer, “The Christian occupation of China: A general surveyof the numerical strength and geographical distribution [sic] of the Christianforces in China, made by the Special Committee on Survey and Occupation,China Continuation Committee, 1918-1921”, Chinese Materials Center, 1922.

China Continuity Committee, 1914, The Christian Yearbook of China, p. 70 (inChinese)

China Educational Commission, 1922, Christian education in China; a study made byan educational commission representing the mission boards and societiesconducting work in China. New York City, Committee of Reference and Counselof the Foreign Missions Conference of North America

Cohen, Paul A., “Christian missions and their impact to 1900”, in the CambridgeHistory of China, Vol. 10, Late Ch’ing, 1800-1911, Part 1, edited by DenisTwitchett, and John K. Fairbank,. Cambridge University Press, 1978

Delacroix, Jacques, and Francois Nielsen, “The Beloved Myth: Protestantism and theRise of Industrial Capitalism in Nineteenth-Century Europe,” Social Forces,Volume 80, 2001, pp.509–553.

Educational Association of China, 1896, “Records of the Second Triennial Meeting ofthe Education Association of China,” Held at Shanghai, May 6-9, 1896, p. 41

Ekelund, Robert B., Jr. Robert F.Hébert, and Robert D. Tollison, “An EconomicAnalysis of the Protestant Reformation”, Journal of Political Economy, Volume110, No. 3, June 2002, pp. 646-671.

Elman, Benjamin A, 2006, A Cultural History of Modern Science in China.Cambridge, MA: Harvard University Press.

Fairbank, J. K., 1974, The Missionary Enterprise in China and America, HarvardStudies in American-East Asian Relations, Cambridge, MA: Harvard UniversityPress. P13.

Gallego, Francesco, and Robert D Woodberry, “Christian Missionaries and Educationin Former African Colonies: How Competition Mattered” Journal of AfricanEconomies, Volume 19, Issue 3, March 2010, pp. 294-329.

Glaeser, Edward L., and Spencer Glendon. 1998. “Incentives, Predestination and FreeWill,” Economic Inquiry, 36: 429–443.

27

Gu, Weimin, 2010. Christianity and Modern Chinese Society, Shanghai: ShanghaiPeople’s Press.

Guiso, Luigi, Paola Sapienza, and Luigi Zingales, “Does Culture Affect EconomicOutcomes?” Journal of Economic Perspectives, Volume 20, Issue 2, 2006, pp.23-48.

Huff, Toby. 2003. The Rise of Early Modern Science: Islam, China, and the West.Cambridge. UK: Cambridge University Press.

Lee, Timothy, (1889),“New rules for making Shanxi rich”. Globe Magazine. InChinese.

Lin, Justin Yifu. 1995. “The Needham Puzzle: Why the Industrial Revolution Did NotOriginate in China,” Economic Development and Cultural Change, 43(2):269-292.

MacAlister, Donald, 1921, Preface of “China and Modern Medicine: A Study inMedical Missionary Development”, by Haroll Balme, London, United Councilfor Missionary Education.

McCleary, Rachel M. and Robert J. Barro, “Religion and Economy,” Journal ofEconomic Perspective, Volume 20, Issue 2, 2006, pp.49-72.

McCleary, Rachel M. and Robert J Barro, “Religion and Political Economy in anInternational Panel,” Journal for the Scientific Study of Religion, Volume 45,Issue 2, 2006, pp. 149-75.

Nunn, Nathan, "The Long Term Effects of Africa's Slave Trades," Quarterly Journalof Economics, Volume 123, Issue 1, February 2008, pp. 139-176.

Nunn, Nathan, "Religious Conversion in Colonial Africa," American EconomicReview Papers and Proceedings, Volume 100, No. 2, May 2010, pp. 147-152.

Records of the General Conference of the Protestant Missionaries of China, held atShanghai, May 7-20, 1890. (pp. 733). General Books LLC.

Shepard D.,1968, A two-dimensional interpolation function for irregularly-spaceddata, Proceedings of the 1968 23rd ACM national conference, ACM, pp.517-524

Silby, Rev. J. A., 1902, “Appeal to Foreign Mission Boards for Trained Educators forChina,” Chinese Recorder, Vol 33： 619-621.

Sinnott, R.W. 1984. Virtues of the Haversine.Sky and Telescope, 68 (2). 158.

Stauffer, Milton T. 1922. The Christian occupation of China: a general survey of thenumerical strength and geographical distribution of the Christian forces in China.Shanghai: China Continuation Committee.

Stock, James H. and Motohiro Yogo. 2005. "Testing for Weak Instruments in IVRegression," in Identification and Inference for Econometric Models: AFestschrift in Honor of Thomas Rothenberg. Donald W. K. Andrews and JamesH. Stock, eds. Cambridge University Press, pp.80–108.

28

Stulz, R.M. and R. Williamson. 2003. “Culture, Openness and Finance,” Journal ofFinancial Economics, 70, p313-349.

Wang, Shenyin. (1983). A brief introduction of the Cheeloo University’s history. InCommittee of Cultural and Historical Data of the CPPCC (Eds.), SelectedHistorical Data (Volume 91, pp. 112-113). Beijing: Chinese Culture and HistoryPress. In Chinese.

Wang, Yousan., 1991, History of Religion in China, Page 985-986, Jinan,Shandong:Qilu Press.

Woodberry, Robert D. “Missions and Economics,” Encyclopedia of Missions andMissionaries. Jonathan J. Bonk (ed.). New York: Routledge, 2007.

Woodberry, Robert D. “The Social Impact of Missionary Higher Education.” pp.99-120 in Christian Responses to Asian Challenges: A Globalization View onChristian Higher Education in East Asia. Philip Yuen Sang Leung and Peter TzeMing Ng (eds.). Hong Kong: Centre for the Study of Religion and ChineseSociety, Chung Chi College, Chinese University of Hong Kong, 2007.

Woodberry, Robert D. “Religious Roots of Economic and Political Institutions,”Forthcoming in Oxford Handbook of the Economics of Religion. RachelMcCleary (ed.). Oxford: Oxford University Press, 2010.

Wylie, Alexander and William Gamble (1867). Memorials of Protestant Missionariesto the Chinese. Shanghai: American Presbyterian Mission Press.

Yang, Senfu. 1968. China History of Christianity (Zhongguo Jidujiao Shi). Taibei:Taiwan Shang Wu Press.

29

Data Appendix

1. Mapping historical information with current administrative units

Refer to Figure A1 as an illustrative example, where squares represent 1999’scounty boundaries and circles represent 1920’s county boundaries. For the latter wehave measures of Christian activities (Xi

1920, i=1,…,5), including number of converts,vicars, and churches. The corresponding 2000 county level measures (X1999), shownas the grey area, is calculated by

1999

~

1999 19201

1

1999 19205

5

...

k kk A E

AA

EE

x x

where

sx x

s

sx x

s

The implicit assumption is that the church activities in 1920 are equallydistributed with a county.

2. Convert station level information into county level

Given the climate measures at the station level (Yj, j=1,…,J), the county levelmeasures (Yc) are constructed by

1,c j ij

j J

Y Y w

1,...,

c j cjj J

Y Y w

where wij is the weight, which is constructed using Shephard’s method (Shepard,1968):

2

2

1,..,

cjcj

ckk J

distw

dist

Figure A1. An illustrative map

1 / 14

Figure 1.Converts per 1,000 People in 1920

Note: This map shows the distribution of converts (normalized by population) in year 1920 among counties in our sample. Each grid is a county with 2000 boundary. Converts information at the 1920 county level was transform into 2000 county level. Refer to Appendix for details. Darker color in the map means higher density of Christian converts per 1,000 people.

Convert per 1k people (by quantile)

.00 - .01

.01 - .10

.10 - .20

.20 - .30

.30 - .50

.50 - .80

.80 - 1.00

1.00 - 1.50

1.50 - 2.50

2.50 - 25.00

2 / 14

Figure 2. GDP per capita in 2000

Note: This map shows the distribution of GDP (normalized by population) in year 2000 among counties in our sample. Each grid is a county with 2000 boundary. Darker color in the map means higher GDP per capita.

GDP per capita in 2000(by quantile, in 1k yuan)

.00 - 1.80

1.80 - 2.50

2.50 - 3.00

3.00 - 3.60

3.60 - 4.20

4.20 - 5.00

5.00 - 5.80

5.80 - 7.20

7.20 -10.00

10.00 - 35.00

3 / 14

Figure 3. Correlation between Historical Christian Activities and Current Economic Performance

Note: This figure plots logarithm of converts per capita in 1920 against logarithm of GDP per capita in 2000. Each dot is a county in our sample. The solid line is the fitted regression line with slope equal to 0.13 and t statistics equal to 16.92. The correlation coefficient between these two variables is 0.38.

Code: D:\Religion\OLS\Tables_20131010 \ T1_SumStats.do

67

89

1011

GD

P p

er c

apita

(lo

g),

200

0

-7 -5 -3 -1 1 3Converts per capita (log), 1920

(coef = .13, s.e. = .01, N = 1742, R2 = .14)

4 / 14

Figure 4. Correlation between Historical Economic Conditions and Historical Christian Activities in 1920

Panel A

Panel B

Panel C

Panel D

Note: These figures plot Christian church activities in 1920 against historical economic performance. Historical economic performance is measured by the population density (normalized by geographic area) in panel A and B, and the land tax per capita in 1820 in panel C and D. Christian church activities is measured by number of churches per 1,000 people in panel A and C, and number of vicars per 1,000 people in panel B and D. All values are in logarithm. In each figure, the dot represents a county. The solid line is the fitted regression line with its slope and standard error report reported below the figure. The shaded area represents the 95% confident interval.

-20

24

68

Po

pula

tion

Den

sity

(lo

g),

19

20

-8 -6 -4 -2 0Churches per 1000 people (log), 1920

coef = -.21, se = .029

-20

24

68

Po

pula

tion

Den

sity

(lo

g),

19

20

-6 -4 -2 0Vicars per 1000 people (log), 1920

coef = -.117, se = .029

-10

-8-6

-4-2

Tax

pe

r ca

pita

(lo

g),

18

20

-8 -6 -4 -2 0Churches per 1000 people (log), 1920

coef = -.211, se = .029

-10

-8-6

-4-2

Tax

pe

r ca

pita

(lo

g),

18

20

-6 -4 -2 0Vicars per 1000 people (log), 1920

coef = -.118, se = .029

5 / 14

Figure 5. Correlation between Historical Christian Activities and Historical Western Medicine Development in China

Panel A. Number of Western Hopital

Panel B.Number of Phamacies

Note: These figures plots provincial level number of converts against historical Western medicine development in China, measured by number of western hospitals in 1920 (Panel A) and number of pharmacies in 1920 (Panel B). All values are in logarithm. In each figure, the solid line is the fitted regression line with its slope and standard error report reported below.

Figure 6. Path of Economic Development across Provinces since 1950

Anhui

FujianGuangdong

Guangxi

Guizhou

Henan

HubeiHunan

Jiangsu

Jiangxi

Shangdong

Yunnan

Zhejiang

Zhili

01

23

4ln

(# o

f Ho

spita

ls),

192

0

8.5 9 9.5 10 10.5 11ln(# of Converts), 1920

beta coef = 1.01, t-stat = 5.05, N = 14, R2 = 0.68.

Anhui

Fujian

Guangdong

Guangxi

Guizhou

Henan

Hubei

Hunan

Jiangsu

Jiangxi

Shangdong

Yunnan Zhejiang

Zhili

11

.52

2.5

33

.5ln

(# o

f Pha

mac

ies)

, 192

0

8.5 9 9.5 10 10.5 11ln(# of Converts), 1920

beta coef = .39, t-stat = 2.05, N = 14, R2 = 0.27.

6 / 14

Note: In this figure, 23 provinces in our sample are divided into two groups: one includes 12 provinces with higher converts/population in 1920; the other one includes 11 provinces with lower converts/population in 1920. The figure plots the weights average per capita GDP across years for each of the group. The weights are the provincial population in each year. Nominal values are deflated to 2000 price level.

0.5

11.

52

2.5

Ave

rage

rea

l per

cap

ita G

DP

(00

0s)

1950 1960 1970 1980 1990 2000 2010Year

Provinces with high convert/population in 1920Provinces with low convert/population in 1920

7 / 14

8 / 14

Table 2. Relationship between Historical Church Activities and Current Economic Performance

Dependent variable is log GDP per capita in 2000

(1) (2) (3) (4) (5)

ln(converts/population) 0.13 0.09 0.05 0.06 0.05

[0.01]*** [0.01]*** [0.01]*** [0.01]*** [0.01]***

Longitude 0.04 0.04 0.04

[0.01]*** [0.01]*** [0.01]***

Latitude -0.01 -0.01 -0.01

[0.01] [0.01]* [0.01]*

Distance to Provincial Capital -0.15 -0.15 -0.15

(in logs) [0.02]*** [0.02]*** [0.02]***

Altitude -0.10 -0.11 -0.11

(in logs) [0.02]*** [0.02]*** [0.02]***

Temperature -0.29 -0.38

(in logs) [0.11]** [0.12]***

Rain 0.10 0.20

(in logs) [0.11] [0.12]

Current Flood Frequency -1.14

(1980-2000) [0.37]***

Current Draught Frequency -0.61

(1980-2000) [0.28]**

Regional FE NO YES YES YES YES

R-sq 0.14 0.26 0.36 0.36 0.37

N 1742 1742 1742 1742 1742

Notes:

OLS estimates of euqation (1) are reported. Observationas are at the 2000 county level. The dependent

variable is the logarithm of GDP per capita in 2000, lny. The Chrisitian activities variable

ln(converts/population) is the logarithm of converts of each county in 1920 normalized by population.

Regional fixed effects are indicator variables for the 7 regions in China proper: north, east, northeast,

middle, southern, southwest, and northwest China.

Coefficients are reported with standard errors in brackets. ***, **, and * indicate significance at 1%, 5%

and 10% levels.

9 / 14

Table 3. Relationship between Historical Church Activities and Current Economic Performance

(1) (2) (3) (4) (5)

Panel A. Second Stage. Dependent variable is log GDP per capita in 2000

ln(converts/population) 0.20 0.20 0.21 0.21 0.20

[0.03]*** [0.06]*** [0.05]*** [0.05]*** [0.05]***

Location Controls NO NO YES YES YES

Climate Controls NO NO NO YES YES

Current Natural Disaster NO NO NO NO YES

Reg. FE NO YES YES YES YES

N 1742 1742 1742 1742 1742

Panel B. First Stage. Dependent is historical Christian activities, ln(converts/population)

Flood Frequency 1.41 1.64 1.86 2.12 2.04

(1900-1920) [0.51]*** [0.50]*** [0.49]*** [0.49]*** [0.50]***

Draught Frequency 5.84 3.52 4.26 4.13 4.27

(1900-1920) [0.55]*** [0.57]*** [0.56]*** [0.56]*** [0.58]***

Location Controls NO NO YES YES YES

Climate Controls NO NO NO YES YES

Extreme Weather Frequency NO NO NO NO YES

(1980-2000)

Regional FE NO YES YES YES YES

F-stat on IV 66.73 18.93 29.35 27.06 27.38

Sargan Test (p-value) 0.22 0.40 0.10 0.36 0.45

Notes:

This table reports estimates of equation (1), where historical Chiristian activities are instrumented by

historical frequency of extreme weather. Observationas are at the 2000 county level. Panel A reports

estimates of the second stage. The dependent variable is the logarithm of GDP per capita in 2000, lny. The

Christian activities variable ln(converts/population) is the logarithm of converts of each county in 1920

normalized by population. Panel B reports estimates of the first stage. The dependent variable is the

endogenous variable, ln(converts/population). Instruments include the flood and draught frequency

between 1900 to 1920.


and 10% levels.

10 / 14

11 / 14

Table 5. Relationship between Historical Church Activities and Development of Modern Education

Enrollment Rate

Christian

Primary

Schools

Public

Primary

Schools

Primary Schools in Total (2SLS)

(in logs) (IV

Tobit) (2SLS)

Overall

Sample Convert/Pop1920>0 Convert/Pop1920=0

(1) (2) (3) (4) (5)

Panel A. Second Stage

ln(converts/population) 1.47 0.39 0.54 0.58

[0.19]*** [0.08]*** [0.32]* [0.30]*

ln(converts/population in

the neighboring counties) -0.12 -0.15 0.42

[0.20] [0.17] [0.19]**

Controls YES YES YES YES YES

N 1742 1742 1701 1444 250

F-stat on IV 26.56 26.56


Panel B. OLS Estimates

ln(converts/population) 1.33 0.12 0.12 0.16

[0.04]*** [0.01]*** [0.01]*** [0.02]***

ln(converst/population in

the neighboring counties) 0.06 0.04 0.08

[0.02]*** [0.02]*** [0.04]**

Notes:

This table examines the effects of historical Christian activities on development of modern educational

system in the early 1920s of China. Observations are at the 2000 county level. The Christian Activities

variable log (converts/population) is the logarithm of converts of each county in 1920 normalized by

population. The across county spillover effects are captured by the weighted average of converts/population

in 1920 in its contiguous neighboring counties (in logs).

All specifications are estimated through 2SLS, where ln(converts/population) is instrumented by historical

frequency of extreme weather, including the flood and drought frequencies between 1900 and 1920. Christian

activities in the neighborhood counties are instrumented by the weighted average of historical Flood/Draught

frequency in the neighborhood counties. Column (1) is Tobit regression since 30% of the county has zero

number of primary student from Christian schools. Column (3)-(5) exclude counties with no converts and no

converts in their neighborhood.

Coefficients are reported with standard errors in brackets. ***, **, and * indicate significance at 1%, 5% and

10% levels.

12 / 14

Table 6. Effects of Church Activities on Current Education and Health Outcomes

(1) (2) (3) (4) (5)

Panel A. Second Stage

Dep. Var. (in logs)

Educational

Expenditure

(per Capita)

Years of

SchoolingLiteracy

Children

Mortality

Health

Expenditure

(per Capita)

ln(converts/population) 0.05 0.07 0.05 -0.38 0.05

[0.03]* [0.01]*** [0.01]*** [0.06]*** [0.05]

Location Controls Yes Yes Yes Yes Yes

Climate Controls Yes Yes Yes Yes Yes

Current Natural Disaster Yes Yes Yes Yes Yes

Reginal FE Yes Yes Yes Yes Yes

Number of Observations 1742 1742 1742 1742 1742


Panel B. OLS Estimates. Dependent Variable is Log GDP per Capita in 2000

ln(converts/population) 0.02 0.01 0.00 -0.04 0.03

[0.01]*** [0.00]*** [0.00] [0.01]*** [0.01]***

Notes:

The Christian Activities variable ln(Convert/Pop1920) is the logarithm of converts of each

county in 1920 normalized by population. Coefficients are reported with standard errors in

brackets. ***, **, and * indicate significance at 1%, 5% and 10% levels. The Hansen's J statistic

is used for the test of overidentifying restrictions in the presence of heteroskedasticity. The joint

null hypothesis is that the instruments are valid instruments, i.e., uncorrelated with the error

term, and that the excluded instruments are correctly excluded from the estimated equation.

13 / 14

Table 7. Alternative Channels

(1) (2) (3) (4) (5) (6)

Panel A. Second Stage. Dependent Variable is Log GDP per Capita in 2000

ln(converts/population) 0.15 0.12 0.13 0.17 0.17 0.11

[0.04]*** [0.05]** [0.05]*** [0.06]*** [0.05]*** [0.06]*

ln(Education exp. per capita) 0.59

[0.05]***

ln(Years of schooling) 0.71 0.72

[0.16]*** [0.16]***

ln(Literacy rate) 0.82

[0.13]***

ln(Children mortality rate) -0.03 -0.03

[0.03] [0.03]

ln(Health exp. per capita) 0.26

[0.02]***

Location Controls Yes Yes Yes Yes Yes Yes

Climate Controls Yes Yes Yes Yes Yes Yes

Current Natural Disaster Yes Yes Yes Yes Yes Yes

Reg. FE Yes Yes Yes Yes Yes Yes

N 1742 1742 1742 1742 1742 1742

Panel B. First Stage. Dependent Variable: ln(Converts/Pop1920)

1900-1920 Flood Freq. 2.15 1.75 2.01 2.09 2.22 1.74

[0.52]*** [0.51]*** [0.52]*** [0.51]*** [0.52]*** [0.52]***

1900-1920 Draught Freq. 4.35 4.04 4.4 3.97 4.38 3.08

[0.59]*** [0.58]*** [0.59]*** [0.59]*** [0.59]*** [0.60]***

F-stat on IV 27.53 24.92 28.46 22.79 27.83 13.43

Over Identification 0.03 0.99 0.95 0.52 0.03 0.87

(p-value)

Panel C. OLS Estimates. Dependent Variable is Log GDP per Capita in 2000

ln(converts/population) 0.04 0.04 0.05 0.05 0.04 0.04

[0.01]*** [0.01]*** [0.01]*** [0.01]*** [0.01]*** [0.01]***

Notes:

This table reports estimates of equation (3), where historical Chiristian activities are instrumented by the

flood and draught frequency between 1900 to 1920. Observations are at the 2000 county level. Panel A

reports estimates of the second stage. The dependent variable is the logarithm of GDP per capita in

2000. The Christian activities variable ln(converts/population) is the logarithm of converts of each

county in 1920 normalized by population. Panel B reports estimates of the first stage. The dependent

variable is the endogenous variable, ln(converts/population). Panel C reports the corresponding OLS

specifications of Panel A without instruments. Control Variables include location controls like

longitude, latitude, (log) distance to the capital city and (log) altitude, climate variables like temperature

and precipitation, current extreme weather frequency and regional fixed effect.


and 10% levels.

14 / 14

Table 8. Effect of Christian Activities on Economic Performance after Adjusting for Education and Health:

Bounding Analysis

ln(Children Death Rate)

(1) (2) (3) (4) (5) (6) (7)

0 80% δ 100% δ 120% δ 150% δ 180% δ 200% δ

ln(Y

ears

of

scho

olin

g)

(1) 0 0.19 0.18 0.17 0.17 0.16 0.16 0.15

[0.05]*** [0.05]*** [0.05]*** [0.05]*** [0.05]*** [0.05]*** [0.05]***

(2) 80% χ2 0.14 0.12 0.12 0.11 0.11 0.1 0.1

[0.05]*** [0.05]*** [0.05]*** [0.05]** [0.05]** [0.05]** [0.05]**

(3) 90% χ2 0.13 0.11 0.11 0.11 0.1 0.1 0.09

[0.05]*** [0.05]** [0.05]** [0.05]** [0.05]** [0.05]** [0.05]**

(4) 100% χ2 0.12 0.11 0.1 0.1 0.09 0.09 0.08

[0.05]*** [0.05]** [0.05]** [0.05]** [0.05]** [0.04]** [0.04]*

(5) 110% χ2 0.12 0.1 0.1 0.09 0.09 0.08 0.08

[0.05]** [0.05]** [0.05]** [0.05]** [0.04]* [0.04]* [0.04]*

(6) 120% χ2 0.11 0.09 0.09 0.09 0.08 0.07 0.07

[0.05]** [0.05]** [0.05]** [0.04]* [0.04]* [0.04]* [0.04]

(7) 140% χ2 0.1 0.08 0.08 0.07 0.07 0.06 0.06

[0.05]** [0.04]* [0.04]* [0.04] [0.04] [0.04] [0.04]

Notes:

This table reports estimates of equation (4). Each cell reports the results of a separate regression. They are the 2SLS

estimates of the coefficients on ln(converts/population) in 1920 that is instrumented by frequency of flood and

drought during 1900-1920. The dependent variable is ln(GDP per capita) in 2000. The returns to education and

health stem from OLS coefficients in an auxiliary regression of the ln(GDP per capita) on years of schooling,

children mortality rate, ln(converts/population), and the control variables (as reported in column (5) of Table 2). To

adjust for the potential bias of the OLS estimates, coefficient on years of schooling is multiplied by factors

indicated in titles of each row, and coefficient on children mortality rate is multiplied by factors indicated in titles

of each column.

Coefficients are reported with standard errors in brackets. ***, **, and * indicate significance at 1%, 5% and 10%

levels.

The Long-Term Effects of ProtestantActivities in China · 2013-11-14 · has generated significant...

Documents

Transcript of The Long-Term Effects of ProtestantActivities in China · 2013-11-14 · has generated significant...