Elections, Democratic Values, and Economic Development in Rural China

David Zweig is Chair Professor, Division of Social Science, and Director of the Center on China's Transnational Relations, The Hong Kong University of Science and Technology. He is author and editor of six books, most recently Internationalizing China: Domestic Interests and Global Linkages (Cornell University Press, 2002) and Globalization and China's Reforms, edited with Chen Zhimin (Routledge, forthcoming). His research covers Chinese who have studied overseas, rural China and China's resource-based foreign policy. Chung Siu Fung holds a Ph.D. from The University of Hong Kong in bio-statistics. She is currently director of the John Cathedral HIV Education Centre in Hong Kong. Most of her publications are in the field of public health.

This paper assesses several assumptions underlying the promotion of democracy and good governance in rural China. We draw on a 1999 survey of 120 villages in four counties, two in Anhui and two in Heilongjiang provinces (sample of 2,997 households, including villagers, cadres and entrepreneurs). First, we look at how institutionalized ‘democratic procedures', such as secret ballots, multi-candidate elections, public nominations, and village contracts, are in these villages. Then we analyze villager views on economic development and democracy, finding that villagers want more democracy, even if the economy is doing well. Third, we assess their views on the election process; do they see elections as efficacious, fair, and competitive, or do they feel that the local power elite manipulate outcomes? Fourth, we found that the richest people are less supportive of democracy, with the most democratic being middle-income households. Finally, elections have increased local elite turnover, cadres understand this, and therefore, democracy does lead to good governance.

Introduction

Within a decade of the decollectivization of rural China (1978-1983) and the enormous economic boom generated by decollectivization, rural China fell on hard times¹. Particularly after the urban reforms of 1984 and rapid urban economic growth in the mid- to late-1980s, rural China's economic conditions deteriorated. The price scissors between urban and rural goods, which had narrowed in 1978-1983, expanded significantly. As many as 20% of rural villages lacked real political authority. In the 1990s, local taxes and fees, imposed by cash-strapped rural cadres, created political hostility and social conflict that threatened the Communist Party's grip on power in the countryside². As a result, the number of rural protests increased almost four-fold between 1993 and 1999.³

In response to these challenges, China's leaders, egged on by younger reformers, have sought to introduce greater democracy and mechanisms of good governance into rural China.⁴ These include village elections, village contracts, village assemblies and transparent financial management within the villages. These reforms, they hope, will increase villagers' willingness to invest in capital construction projects, such as new roads, under the assumption that if their own elected officials call on them to do so, they will.⁵ Moreover, it is anticipated that more economic growth and democracy will promote political stability. Thus, their support for democracy remained highly instrumental and not an end in itself; so long as democracy strengthened state power and national economic growth, it would be supported.

Conceptual perspectives on democracy and economic development

This paper assesses several assumptions underlying the promotion of democracy and good governance in rural China.⁶ A most common argument in the political science literature is that economic development leads to democratization. Quantitative and qualitative studies generally confirm this relationship.⁷ For Huntington, ‘few relationships between social, economic, and political phenomena are stronger than that between level of economic development and existence of democratic politics'.⁸ And while Przeworski and Limongi tried to refute the argument that economic development leads to democracy,⁹ Boix and Stokes reconfirm that assumption,¹⁰ as did Londregan and Poole.¹¹ Nevertheless, the Latin American experience in the 1970s and 1980s suggests that, while middle levels of economic development are conducive to democratization, both poverty and higher levels of economic development could generate authoritarianism.¹²

In rural China, the relationship between economic development and democracy may also be curvilinear. In very poor communities, local governments cannot afford ballot boxes, and villagers may be too busy making ends meet to care about good democratic procedures.¹³ At the other end of the economic spectrum, the most industrialized and collectivized villages have the financial wherewithal to resist democracy. Oi and Rozelle found that the competitive nature of elections decreased in the top decile of villages in terms of their levels of industrialization. However, in the other 90% of villages, they found no relationship between economic development and electoral politics.¹⁴ In this vein, Guangdong Province was the last major province to introduce village elections. Perhaps middle-income regions, where numerous private businesses want less intrusive leaders and more transparent polities, are where we find the socio-economic roots of democratization and a positive relationship between economic development and democracy. Still, Hu's research in Fujian Province found greater participation and more competition in wealthier villages, which also did a better job of instituting various election laws.¹⁵

Our research questions¹⁶
In light of the above discussion, the key questions driving this study are as follows.

1. Is rural China democratic? Has the level of democracy increased, and have villagers become more aware of the opportunities democracy affords them? How institutionalized are ‘democratic procedures', such as secret ballots, multi-candidate elections, public nominations, and village contracts?
2. Are villagers satisfied with economic development alone, or do they believe that democratic reforms are necessary even if there is good economic growth? Do they see elections as efficacious, fair, and competitive, or do they feel that the local power elite manipulates outcomes? Do more democratic nomination systems increase government legitimacy?
3. What is the relationship between economic development and democracy? Are richer areas, or richer people, more supportive of democracy? Are cadres in wealthier areas more or less supportive of democracy? How does the emerging rural middle class see democratic development? If the answers to these questions are positive, continued economic growth could transform rural politics.
4. Finally, have elections increased leadership turnover? How do local elites perceive electoral politics? Does the ballot box discipline cadres so that democratically elected cadres are less prone to misbehave - i.e. does democracy lead to good governance?

Methodological issues

With the cooperation of a research centre in China, we drafted two detailed surveys, one for villagers and another for cadres. Due to financial and political limitations, however, the survey was carried out in 1999 in two provinces, Heilongjiang and Anhui, with two counties selected in each province and 30 villages selected in each county for a total of 120 villages. In each village we selected 20 villagers at random, for a total of 2,400 villagers. We also pinpointed 237 individuals who ran their own businesses - approximately two in each village - whom we categorize as entrepreneurs and treat as the emerging middle class. Finally we interviewed 360 cadres, including from each village the party secretary, the current director of the Villager Committee, and one other cadre who did not hold either post. About 20% of informants were women.

Unfortunately, we could not collect time series data on the elections. Election data only reflected the village's most recent election. Thus we cannot talk about how elections or better governance over time changed popular political attitudes. Moreover, the elections in Heilongjiang Province all occurred in 1996-1997 (with only one in 1998), before China's National People's Congress passed a law announcing its preference for direct nomination by villagers. Fortuitously, in Anhui, all the villages held elections in 1999, with about half of them following the ‘haixuan' system, or other forms of direct nominations by villagers.¹⁷ Thus, though we have no time series data, we can still try to assess whether better democratic procedures - in the form of more direct nominations - affected villagers' attitudes towards the government and political reforms. Still, we must recognize that we cannot show causality - it is possible that areas introduced more democratic forms of nominations because the people and/or local government supported greater democracy.

The context of the village study

All 120 villages were predominantly agricultural, with little variation in demographic or socio-economic characteristics of the two Anhui counties. Per capita income for the two counties as reported in Anhui Provincial Yearbook, 2000 was RMB 2,182 (Anhui A) and RMB 2,201 (Anhui B), while the provincial average was RMB 1,900.¹⁸ In Helongjiang, the per capita income in one county in 1999 (Heilongjiang A) was RMB 2,637, but the second county, Heilongjiang B, was significantly poorer, with a per capita income of RM 1,227,¹⁹ and significantly below the provincial average per capita income of RMB 2,165. In terms of per capita income as derived from the village records, the two Anhui counties and Heilongjiang B were relatively similar. Anhui A and B had a mean income of 1,983 and 1,827, respectively, while Heilongjiang B had a mean per capita income of 1,877. However, Heilongjiang A had a mean per capita income of 3,273.²⁰ This pattern also holds true for per capita income, measured by the number of household appliances. In the two Anhui counties, 79 and 84% of all households had three household appliances or less. In Heilongjiang B, 64% of villagers had three items or less, while in Heilongjiang A, the richest county, 58% of villagers reported having three household items or less.
Education levels also varied. The mean number of years of schooling of the Anhui counties was 5.1 and 5.7, while in Heilongjiang, the richer of the two counties had a mean score of 6.8, while the poorer county had a higher mean of 7.2. There was also greater inequality in the level of individual schooling in Anhui than in Heilongjiang, as measured by the standard deviation.
As outlined above, the political context of the survey varied between the two provinces as well. Both counties in Anhui held their elections in 1999, after the National People's Congress passed the new Organic Law, while most Heilongjiang elections came in 1996-1997. As a result, nominations in Anhui were far more democratic (Table 1).

Table 1. Measures of democratic procedures, by province, 1996-1999

	Percentage of villagers reporting
Procedures employed	Heilongjiang	Anhui
Voting rate	84	85
Villages with multi-candidate elections	66	74
Villages with more democratic nominations	23	47
Secret ballot	80	79
Election speeches	26	19
Village contracts	88	68

Notes: * Please see footnote 24.
Source: Survey in rural China, Summer 1999.

The practice of democracy in rural China

How well established are democratic institutions in these communities? Indicators of the level of democracy include voting rates, secret ballots, multi-candidate elections, a village contract, whether the villagers felt that village leaders were publicly sharing financial records with them, and how many times elections had been held in the community (Table 1).

Almost 70% of villagers reported that their community had multi-candidate elections,²¹ a much higher rate than that reported by numerous observers for the 1995-1997 period.²² However, our survey occurred in summer 1999, after the spring 1999 reform of the Organic Law on Village Elections by the National People's Congress. Over 79% of villagers reported that their village used secret ballots. Similarly, 91.4% of our villagers reported satisfaction with the level of transparency of the village's financial reports, while 78% of villagers reported that their leaders had signed a contract with the villagers, both important steps in improving local governance. On the other hand, candidates made speeches before the election in only 22% of the cases. According to Pastor and Tan, since villagers know the candidates intimately before the election, there is little need for campaigning.²³ Still, if elections were really competitive, candidates should want to make speeches to differentiate themselves on important policies from their opponents.

We also had results on how candidates were nominated, and broke those responses into two categories: more democratic nomination (MDN) systems and less democratic nomination (LDN) systems. The MDN category included villages using the haixuan system - which the Ministry of Civil Affairs posits as the fairest nomination system - those whose villagers nominated candidates directly, self-nominations, nomination by a unified group of villagers, or nominations by representatives of the households. Nominations by organizations, such as the party committee, the election committee, the villager representatives assembly, nominations by ‘powerful or privileged people in the village', and people ‘who did not know' how the nomination occurred, all fell into the LDN category.²⁴

In our overall sample, 43% of villagers reported that their village used more democratic nomination systems, but in Anhui Province, which held its elections in 1999, the share of MDN elections was significantly higher. In the paper, we compare the findings of these two cohorts, in part because we believe that MDN villages and villagers who live in MDN villages, represent a more democratic trend, and we wish to see how that trend affects attitudes and beliefs about local politics.²⁵

Our villages' voting rates were rather high - 86% of villagers reported voting in the villager committee elections - but this is the easiest medium through which people can participate. Also, 21% of people had encouraged others to vote, 13% nominated candidates, while another 10% of villagers voluntarily participated in the meeting of the Villager Assembly or the Village Party Committee. Only 10% of villagers reported not participating in any form of political activity at all.

Elections as a source of support for the government

What were the villagers' attitudes towards elections? If they saw elections as fair and competitive, and felt that the CCP or other administrative forces were not too influential in determining electoral outcomes, the elections may become legitimate and help stabilize the political system. However, if they felt that local power holders could manipulate electoral outcomes, democracy would remain formalistic and yield few positive political outcomes.
For example, people may feel that party officials control the electoral outcome through manipulating candidate selection. So, we asked villagers - ‘what do you think of the CCP's level of influence over elections?' Surprisingly, only 10% saw the party's influence as ‘very great'; 22% saw the CCP as having ‘a certain level of influence'; 22% saw it as having ‘not much influence'; while 21% saw it as having ‘no influence'. Finally, 21% said that they ‘did not know'. Even if we assume that those who selected ‘did not know' were afraid to say that the CCP had too much influence, 43% did not feel that the CCP had very much influence.
Similarly, we asked people to respond to the statement: ‘If a ministry or bureau wants to control the election, they need not control the voting but instead can control the selection of candidates'. The responses? ‘Totally agree' - 6.5%; ‘relatively agree' - 18%; ‘don't agree much' - 21%; ‘totally disagree' - 20%; ‘don't know' - 27%. Thus, 41% of respondents felt that outside forces do not control candidate selection or electoral outcomes. In Anhiu Province, villagers in MDN villages were even more likely to ‘totally disagree' (25%) or ‘disagree' (21%) than villagers in LDN villages, whose scores were ‘totally disagree' - 19% and ‘don't agree much' - 15% (total 34%) (MLA, p = 0.01).²⁶

Overall, villagers believed that elections were becoming more competitive and fair relative to the previous election. In our survey, 15% of villagers saw the level of competitiveness as being ‘very intense', 40% saw it as being ‘relatively intense', while 22% saw no change. Only 7% saw the recent elections as ‘not intense at all'. Similarly, 18% saw the most recent election as ‘fairer', 41% saw it as ‘a little fairer', and 21% saw no change. Only 7% saw the most recent election as ‘a bit less fair' or ‘far less fair' than previous ones.

These findings were particularly true in villages with more democratic institutions. Based on a multi-variate analysis, the best explanations for ‘level of perceived fairness' was whether villages allowed voters two or more ballots - villages that allowed some voters two or more ballots were seen to be very unfair - and whether there were more candidates than positions. In Anhui, when we asked rural inhabitants to rank the level of fairness of the electoral process, 70% of villagers whose village used more democratic nomination systems selected ‘fairer' or ‘a bit fairer', while only 52% of LDN villagers did so (p = 0.001). Therefore, villagers in villages with more democratic nomination systems, saw more fairness in the electoral process (see Table 2).

Table 2. Perceived level of fairness in MDN and LDN elections, Anhui Province, 1999

Perception of elections	MDN villages	LDN villages
Became fairer	18%	13%
Became a bit fairer	52%	39%
No change	10%	19%
Became a bit less fair	7%	6%
Became far less fair	3%	3%
Don't know/No response	10%	20%
Total	100%	100%

Notes: MLA, p = 0.001. When we calculate the MLA value, we put the ‘Don't know/No response' between ‘No change' and ‘Became a bit less fair' category.
Source: Survey in rural China, Summer 1999.
Question: Was the most recent election fairer than the previous election?

Similarly, villagers who reported that their village selected candidates through a more democratic nomination process also saw the electoral system as more competitive, another boost for the system's overall legitimacy (see Table 3). Thus 20% of MDN villagers saw the level of competition under the MDN systems as ‘extremely intense' versus 11% of villagers in LDN villages. And when we combine ‘extremely intense' and ‘comparatively intense', villagers in MDN villages holding these views comprise 78%, while 52% of villagers in LDN villages held similar views (p = 0.001). Given that the Anhui elections were held two or three years after the Heilongjiang ones and after the reform of the electoral nomination system, it is possible that more democratic procedures make people see the electoral system as more competitive and fair.

Table 3. Perceived level of competitiveness in MDN and LDN elections, Anhui Province, 1999

Perception of elections	MDN villages	LDN villages
Extremely intense	20%	11%
Relatively intense	58%	41%
No change	5%	15%
Not very intense	5%	6%
Not intense at all	1%	1%
Don't know/No response	11%	26%
Total	100%	100%

Notes: MLA, p = 0.001. When we calculate the MLA value, we put the ‘Don't know/No response' between ‘No change' and ‘Not very intense' category.
Source: Survey in rural China, Summer 1999.
Question: Was the level of competition in the most recent election more intense than the previous election?

Elections strengthen villager support for government policy. When asked to respond to the statement: ‘After several rounds of elections, the villagers are more supportive of government policy than before', some 23% of villagers strongly agreed with this statement, 43% agreed somewhat, 11% disagreed somewhat, and only 4% strongly disagreed (19% didn't know or didn't answer). Moreover, villagers who reported that their candidates had been nominated through more democratic systems were more likely to see a relationship between elections and support for government policy. Of villagers in MDN villages, 72% ‘strongly supported' or ‘relatively supported' this statement, while 58% of villagers in LDN villages held such a view (MLA, p = 0.001).

If villagers believed that elections brought bad influences, increased democracy could contribute to political or social instability, as some oligarchs assert. However, 61% of our villagers do not believe elections are harmful; only 11% feel this way, while 28% selected ‘don't know'.

Would villagers be satisfied with good governance - without greater democracy - and could good economic development ameliorate aspirations for greater democracy? In response to the statement ‘If the existing cadres are capable and trusted, there is no need for democratic elections', 12% agreed strongly and 24% agreed somewhat; yet, 33% disagreed somewhat, and 22% disagreed strongly (9% didn't know or had no response). Thus good governance alone is not sufficient for 55% of our villagers - they want democracy. Similarly, when asked to respond to the statement ‘as long as village economic development is stable, there is no need to increase the level of democracy', 33% strongly disagreed and 32% disagreed somewhat. Only 7% agreed strongly, while 14% agreed somewhat (14% didn't know). Most villagers cannot be bought off by capable leaders and economic growth; they want democracy.

Elections, democratic values and economic development
As mentioned in our theoretical discussion, political scientists tend to believe that economic development, without significant inequality, is highly correlated with democratic development. As more rural inhabitants, particularly in the wealthier coastal or suburban areas, become more prosperous, might the demand for democratic procedures and greater political participation increase? To answer this question, we need to know if wealthier villagers held stronger democratic values.

We combined villager responses to four questions about democracy, elitism, and electoral politics to create a variable called ‘the democratic idea' (see the Appendix for the list of questions and the villagers' responses to those questions). We stratified the responses into four categories, running from ‘not democratic' to ‘very democratic'. We also asked villagers to describe their income relative to other people in their village, selecting upper, upper-middle, middle, lower-middle and lower. Table 4 shows the cross-tabulation of these two variables.

Table 4. Democratic idea by perceived level of wealth, in percentages

	Perceived level of wealth
	Lower	Lower-middle	Middle	Upper-middle	Upper	Total
	(n=250)	(n=435)	(n=1,415)	(n=419)	(n=102)	(n=2,621)
Democratic idea	%	%	%	%	%	%
Not democratic	41	28	26	29	38	28
A bit democratic	42	50	47	41	43	46
Democratic	9	14	16	17	13	15
Very democratic	8	8	11	13	6	11
Total	100	100	100	100	100	100

Note: MLA, p = 0.003.
Source: Survey in rural China, Summer 1999.

Interestingly, the very wealthy (upper) are almost as non-democratic (38%) as the poorest or lower income group (41%), and comprise even a smaller share of the most democratic group (6%) than the poorest villagers (8%). The most democratic people are those who see their wealth as ‘middle' or ‘upper-middle', while those who see themselves as ‘lower-middle' fall in-between. These findings are statistically significant.

Figure 1 presents the same data in a visual format. Note that across all four aspects of the ‘democratic idea' we find a curvilinear relationship, where the poorest and richest villagers in general are less democratic and the middle levels of wealth tend to be more democratic. This shape reflects our hypothesis in the opening sections of this paper. In particular, the ‘not democratic' group shows the clearest curvilinear relationship.

The above findings employ the individual's wealth relative to villagers in their own village, rather than the village's overall wealth. But which communities are likely to be democratic - rich ones, poorer ones, or average ones? As mentioned above, Oi and Rozelle found a negative relationship between competitive elections and the level of industrialization. However, in their data, the incidence of contested elections decreased only in the wealthiest 10% of villages,²⁷ otherwise the level of economic development did not affect their measure of democratization.

Our data, though tentative, also suggest that richer villages, based on per capita income from village records, are less democratic. While 62% of villagers in wealthier villages reported that their most recent election had involved more candidates than positions, 72% of middle-income villages and 76% of poorer villages reported multi-candidate elections (p = 0.001). Similarly, while 64% of villagers in poorer villages saw the most recent elections as more competitive, only 44% of villagers in wealthier villages expressed such views.²⁸ Also, while speeches may not be that important in village elections, 26% of villagers in the poorer villages reported having had speeches, while only 21% of those in richer villages said that there had been speeches.²⁹

Also wealth did not correlate with democratic values. When asked to respond to the statement ‘so long as there is stable economic development, there is no need to promote greater democracy', villagers in richer villages were more likely to ‘strongly agree' and ‘agree' (24%) than people in poorer (17%) or middle-income villages (22%), and the finding was statistically significant (p = 0.006). Similarly, people in wealthier villages were more likely to believe that the wealthy should have more right to speak out on village affairs (p = 0.001) - what might be seen as an elitist view of democracy - and they strongly agreed with the statement ‘because I know what's happened in the village, I have the right to participate in village affairs'. Here, 34% of villagers in wealthier villages ‘strongly agree' with that statement, while only 22% and 17% of people in poor and middle-income villages, respectively, strongly agree. Again, the finding was statistically significant (p = 0.001), and again, middle-income villages were more democratic.
Cadres' attitudes towards democracy vary based on the village's per capita income (Table 5). Four questions assessed cadres' views on democracy and elections, and we broke their responses into three categories - non-democratic, democratic and very democratic. We also divided village income into richer, middle-income and poorer. In richer villages, 25% of cadres fell into the ‘non-democratic' category, while only 16% and 13% of cadres in poor and middle-income villages held ‘non-democratic' views, respectively. Similarly, only 24% of cadres in richer villages expressed ‘very democratic' ideas, while 38% and 40% of cadres in poor and middle-income villages, respectively, expressed ‘very democratic' values - a very big difference.³⁰ These findings suggest that richer villages breed less democratic cadres.

Table 5. Cadre attitudes towards democracy by village income per capita, in percentages

	Village income per capita
	Poor	Middle	Rich	Total
	(n=108)	(n=129)	(n=123)	(n=360)
Attitudes towards democracy	%	%	%	%
Not democratic	16	13	25	18
Democratic	46	47	51	48
Very democratic	38	40	24	34
Total	100	100	100	100

Note: MLA, p = 0.01.
Source: Survey in rural China, Summer 1999.

Finally, what kind of villages, and who within them, are most likely to demand good governance? We asked villagers: ‘if you heard that village cadres were considering instituting an unfair measure (cuoshi), what would you do?' Based on their selection from a list of options, including collective or individual action, as well as no action, we gave each villager a total score. The individual scores varied based on their relative income, showing that richer villagers (62%) were much more likely to consider action than poorer (44%) or middle-income ones (55%, MLA = 0.00). Similarly, when asked if they had taken actions to solve problems facing their village, richer villagers were again more active. Of the 14% of our entire sample who had taken some concrete action, richer villagers had acted more often (16%) than middle-income (14%) or poorer villagers (11%). The measure of linear association was 0.008. Therefore, perhaps it was the richer villagers - who are better connected, particularly with the local party leader - in poorer villages - where there were economic problems - who were most likely to consider taking action in response to unfair policies. This finding suggests that if cadres misbehave, wealthier villagers are most likely to challenge those actions. Still, although they are more likely to take action, remember that these richer villagers possess less democratic values.

Democracy and the emerging entrepreneurial class

The role of the middle class in promoting or supporting democracy is sometimes problematic. Barrington Moore argued that democracy depends on the emergence of a middle class,³¹ while O'Donnell and others argue that in Latin America, the business community supported ‘bureaucratic authoritarianism' and the repression of the working class.³² Payne, however, found that Brazilian industrialists were agnostic about the regime type - they supported any regime that could insure political stability and stable investment rules.³³ Conversations between one of the authors and private business owners in rural China in the early 1990s yielded a strong antipathy towards democratization and political liberalization. However, the growing influence of the rural private sector and the emergence of a new middle class might abet democratic development. Newspapers, magazines and some scholars report that villagers elect entrepreneurs because they believe they are most capable of promoting the village's economy. Therefore, if they feel that they can gain political power through the electoral process, the local business elite may support elections.

Our localities, however, had few factory owners, with only 2% of our entire sample reporting that their family's main income came from wage labor. Still, interviewers in each village identified two entrepreneurs, yielding a total of 237 individuals, whose views we can compare with non-entrepreneurial peasants. These people were significantly richer than other villagers, both in terms of their self-identification of their level of wealth (their mean score of 2.4 versus a score of 3.2 for non-entrepreneurial villagers reflected their greater relative wealth) and in terms of the number of household appliances they possessed (mean of 4.4 versus 2.8).
But, do their views and attitudes about politics differ from the average peasant? They are much more aware of political phenomenon, discuss politics more frequently, and are more willing to voice their opinion on issues than average villagers. In fact, 28% of them report ‘often' discussing political and economic issues, while only 14% of other villagers do so. This may be because they are more educated and more likely to have a member of the Communist Party in their family. They are also more willing than other villagers to discuss issues with the party secretary and to participate in the village representative assembly's meetings; but they are not more likely to discuss issues with the village director, perhaps due to his more limited influence.

In terms of values, they are no more democratic than other villagers, except for their response to one question - ‘as long as the village economy is stable, there is no need to increase the level of democracy'. The data in Table 4 above, about the relationship between the ‘democratic idea' and level of perceived wealth, reflect the position of entrepreneurs somewhat, in that 47 of the 237 entrepreneurs, or 20%, declared their income as upper level, which placed them in the least democratic group. Still, 88 or 37% of them fall into the upper-middle income group, which was the most democratic cohort.

Compared to other villagers, however, entrepreneurs show greater support for good governance and political procedures that limit the local elite's ability to manipulate the local economy and intervene in private economic affairs. They are more likely than regular villagers to think that publicly displaying the village's financial records is good for economic development,³⁴ and to think that a contract between the villagers and the cadres improves villager-cadre relations.³⁵ Therefore, while they may not favor greater democracy, they reflect similar views to Brazilian industrialists who care less about democracy but more about good governance and political stability.

Yet villagers do not want entrepreneurs to serve as village director.³⁶ When choosing possible candidates from a list - villagers could choose more than one - only 7% selected entrepreneurs, less than those who selected normal villagers (putong nongmin) (8%) or cadres (17%). Also, when asked to respond to the statement ‘entrepreneurs who are good at business and have money are the best candidates for village director', 71% of villagers ‘totally disagreed' (38%), or ‘disagreed' (33%), while only 13% ‘relatively agreed' (9%) or ‘totally agreed' (4%). Villagers see the agglutination of political and economic power as threatening to local democracy.³⁷ In these localities it will not be easy for entrepreneurs to break into local politics through the electoral process.

Democratizing the rural elite?

Have democracy and elections affected the local elite? Have elections led to significant turnover among elites and are the values of the new elite more democratic? And, do more competitive elections serve as a monitoring mechanism limiting cadre misbehavior?

Is there a ‘power transition' underway?

To what extent have local elections led to a restructuring of power in the countryside? In our villages, elite turnover increased significantly between 1996 and 1999. The percentage of village leaders taking office for the first time increased significantly year to year, from 37% in 1996 to 43% in 1997, reaching 53% in 1999.³⁸ Much of this turnover occurred in 1999 in our Anhui villages, where the introduction of more democratic nominations led 25%of cadres to decide not to run for re-election (Table 6). Also, when Anhui village directors ran again for office, they lost approximately 50%of the time. Rural elections are changing the people who hold political power.

Table 6. Getting re-elected, MDN and LDN village directors, Anhui Province, 1999

	MDN village directions			LDN village directions
	No.	%		No.	%
Previous director, ran for office and re-elected	7	25.0		15	46.9
Previous director, ran for office but not re-elected	14	50.0		13	40.6
Previous director, did not run for office	7	25.0		4	12.5
Total	28	100		32	100

Note: Chi-Square test, p = 0.173.
Source: Survey in rural China, Summer 1999.

Second, elections are pluralizing political authority in our villages. While the village party secretary in 1996-1997 was often also director of the villager committee, the 1999 election ended this phenomenon almost entirely. Thus in 1996 and 1997, the party secretary held both posts in 32% and 46% of the cases, respectively, but in 1999 the same person was elected to both posts in only one of 49 cases.³⁹

But even though a clear division of labor and authority between the two top posts in the village emerged, an authoritarian leadership strata still held power, making the division of authority somewhat pro-forma. In Anhui, the director of the village committee was usually the deputy secretary of the village party committee. While in Heilongjiang in 1996-1998, 20% of people elected as director of the village committee were also the vice-party secretary, in 1999, 53% of the newly elected directors in Anhui were also vice party secretary. So, even with different people holding the two posts, key decisions would still be made in the party committee run by the village party secretary.

Are non-CCP members elected to the villager committee, a source of democratic values within the local elite? According to Jakobson, in 1997, 44% of village committee members nationwide were not party members. She also quotes Wang Zhenyao to the effect that 30-50% of village committee directors were not party members when they won the election, but were recruited into the CCP within a few weeks.⁴⁰

In our survey, 25% of village committee directors reported that they had joined the CCP before becoming a cadre, while 54% of them had become a cadre and then joined the party. This latter group, then, was recruited into the CCP after attaining local elite status. However, the pattern for the village party secretaries was reversed: 48%of them had joined the CCP before becoming a cadre.⁴¹ The finding suggests a different mobility track for party secretaries and village committee directors - with the former having much deeper roots in the CCP - which could lead to different values and attitudes.

The values of local cadres

Despite the different mobility tracks, village party secretaries and village directors held similar views on most issues, reinforcing the argument that a village-level oligarchy exists. On our democratic scales there were no statistically significant differences between these two groups. We would have hypothesized that on issues such as democratization or elections, their differing experiences should have led to different views. Nevertheless, since many village directors in our survey were nominated by local organizations or elected through only partially democratic procedures, their views may not differ so much from local party elites who may have influenced their selection. In fact, they were no more worried about losing office if they could not resolve local problems than the party secretary, suggesting that they felt well protected by the local power structure.

Still, elite turnover has had some impact on the values of local leaders. Cadres who simultaneously had held both posts were less democratic on a number of dimensions than cadres who held either the post of village party secretary or director of the village committee. In terms of their democratic predilections, of the 22 officials who had held both posts, 32% fell into the ‘non-democratic category', while only 20% of party secretaries and 14% of village directors did so. Similarly, only 23% of those holding dual posts fell into the very democratic group, while 36% of party secretaries and 34% of villager committee directors did so. On the democratic process, these ‘village emperors' were also less democratic. They were much more likely to believe that wealthier people should have more right to speak out on public affairs; they were less likely to disagree with the statement that ‘if the economy is good then there is no need to promote democracy'; and, they were much more likely to agree that ‘only people with special skills should have the right to hold office'.

Nevertheless, the most democratic group was the local non-party elite who were members of the village committee. As Table 7 shows, they were least likely to believe that the wealthy should have more say in public affairs or to believe that only people with special knowledge should have greater influence on decisions. Also, they were less likely to believe that good economic development mitigates the need for greater democracy (question 2). If these villagers move into leadership positions, the future local elite could become more democratic. Still, will they be able to enter the elite if they maintain those attitudes or will they not change those attitudes after they gain real power?

Table 7. Attitudes towards democracy among different types of cadres

	Party secretaries	Village directors	Other cadres who are CCP members	Other cadres who are not CCP members
No.	(122) %	(97) %	(103) %	(38) %
(1) Wealthy people should have more say in public affairs than poorer people
Strongly agree	20	10	13	8
Agree	24	33	22	18
Disagree	21	24	35	32
Strongly disagree	33	31	30	40
(2) As long as development of the village economy is stable, there is no need to increase the level of democracy
Strongly agree	4	4	3	3
Agree	5	4	8	3
Disagree	23	21	24	24
Strongly disagree	67	69	65	71
(3) Only people with special knowledge and abilities have the right to speak at times of decision making
Strongly agree	24	12	18	16
Agree	25	32	36	21
Disagree	26	34	25	29
Strongly disagree	22	21	19	29

Source: Survey in rural China, Summer 1999.

Cadre attitudes towards elections, democratization and good governance

Central leaders frequently call on local cadres to be less rapacious and ‘lighten the burden of the peasants'. Given that the state has raised this plea almost annually since 1978, a new mechanism is needed to monitor cadre behavior. According to Li and O'Brien, the Fujian provincial Bureau of CivilAffairs convinced the provincial Discipline Inspection Committee to support village elections ‘when the commission realized that corruption tended to be lower where well-run village elections took place'.⁴² If generally true, elections could decrease rural instability that results from cadre malfeasance.

In our survey, both villagers and cadres believed that elections had a positive effect on cadre behavior. In Anhui, 48% of villagers from more democratic villages believed that village elections decreased cadre haughtiness and rudeness, while only 41% of villagers in less democratic villages held these views.⁴³ Villagers from more democratic villages were more satisfied with the level of transparency of the village's financial records, with 64% of them being either ‘very satisfied' or ‘satisfied', while only 47% of villagers in less democratic villages held these views (MLA, p = 0.001). This finding occurred because villages that introduced direct nominations were also more likely to make their financial records public. That meant that 29% of citizens in less democratic villages were ‘not satisfied' and 6% were ‘extremely dissatisfied'.

This more favorable attitude towards how officials allocated village finances was reconfirmed by other questions. MDN villagers expressed greater satisfaction with how funds that they contributed to the village were used - 44% being ‘very satisfied' or ‘satisfied' - while only 30% of villagers in LDN villages felt this way (MLA, p = 0.001). Importantly, 25% of villagers in LDN villages were flatly dissatisfied (the bottom scale on this question) as compared to only 15% of MDN villagers.⁴⁴ Still, villagers in MDN villages were not more likely to believe that elections could end the problems of ‘random taxes or assessments' (luan shou fei) or prevent cadre corruption, a position echoed by Bernstein and Lu.⁴⁵

Cadres say that elections make them more responsive to villagers' interests, a positive force for political stability. More than peasants, cadres believe that elections make them more likely to resist taxes and fees from the upper levels. While only 55% of villagers ‘completely agree' (14%) or ‘somewhat agree' (41%), 81% of cadres either ‘completely agree' (31%) or ‘somewhat agree' (50%). If true, however, cadre relations with higher-level officials become problematic if they respond to villagers' demands and protect the villagers' interests, a finding shared by Li Lianjiang.⁴⁶

Ironically, elections may make it more difficult, not easier, for village directors to promote economic development. Cadres in MDN villages were more likely to have faced villagers who refused to contribute to development projects (36%) than cadres in less democratic villages (19%) (MLA, p = 0.14).⁴⁷ We also asked the village directors to agree or disagree - on a four point scale - with the statement ‘Democratic elections make villagers more willing to contribute money to the village accumulation fund'. Directors in MDN villages tended to disagree with this statement (their mean score was 2.5), while directors in LDN villages were inclined to agree with it (mean score 2.1).⁴⁸ While the process of accumulating local capital may have become more democratic, democracy may not benefit economic development, despite the fact that local governments need capital to build roads and other infrastructure that promote economic development.

Conclusion

Like many other studies of democracy in rural China, this study will not make sweeping or definitive generalizations. Data from four counties simply do not permit that. Our findings remain suggestive. Part of the problem is that good indicators of democracy, stability, economic development or good governance are not easy to collect. Though few analysts confess to the limitations of their data, because it would call into question their findings, surveying rural China on political variables is costly and very difficult.⁴⁹ We, therefore, should not give up case studies, or in-depth interviews, although a research strategy involving case studies and surveys would be optimal.

Also, significant changes have occurred since 1999. Rural protests have increased dramatically, with land issues jumping to the forefront. Also, concern within the Chinese leadership over the conflict between two centers of power, the appointed party secretary and the elected - and therefore more legitimate - village director, has reportedly led to some backtracking on the electoral process. In some locations, the same person is again holding both posts. To that extent, our study catches rural China in a perhaps unique period of expanding democracy.

Still, democratic procedures based on our indicators were fairly well established, suggesting that democracy was strengthened in rural China in the late 1990s. As villages moved from having organizations or collectivities nominating candidates, increasing instead the role of individuals in this process, the democratic environment improved. Moreover, in villages with more democratic nomination systems, elections were also seen to be more legitimate, fairer and competitive.

But it is hard to say that wealthier localities are more prone to adopt democratic principles and procedures. Our data suggest that neither wealthier villagers nor villagers in wealthier villages are the strongest advocates of greater democracy or repositories of democratic principles; even though they are more likely to take action against cadre malfeasance. Similarly, cadres in the wealthiest villages were the least democratic of our entire sample. Instead, villagers in the upper-middle income bracket may be the most democratic citizens in rural China, while poorer villages were most active in trying to solve their problems.

One negative result of the electoral process was that elections made it more difficult for cadres to convince villagers to contribute to local projects, which may undermine economic development. Admittedly, local cadres often misuse local funds. But if they are disciplined by elections, then granting responsive or responsible officials the authority to create collective goods could promote the local economy.

Our data cannot confirm that elections constrain cadre misbehavior, though the direct nomination process in Anhui Province may have had that effect. Cadres said that elections kept them in line; but they knew the politically correct responses - that is how they got to be cadres. Moreover, few expressed great concerns that they might lose their jobs if they could not solve the villagers' problems. And, they are under great pressure to follow the directives of township leaders.⁵⁰ Still, elections may encourage them to resist the demands for fees and taxes from upper levels.

Finally, despite the increasing institutionalization of democracy, changes within the local elite suggest relative political constancy. While the people clearly favored more democracy, they were not really getting it. The old party secretaries, who also held the post of village director, were gone and a real division of authority emerged between the village director and the village party secretary. But as long as village directors are recruited into the CCP and are becoming the vice party secretary, the challenge to the CCP's authority posed by elections remains limited. We are not yet seeing the emergence of a 'dang wai', or the rise to power of any cohort of likeminded individuals outside of the ruling party who could coalesce into a stable opposition. Only if the non-party elite become village directors, all the while resisting pressures or blandishments to join the CCP, could this process materialize. But our data suggest that this type of major political transformation was not happening.

Appendix: Measuring the democratic idea

We asked villagers to respond to a series of questions which could tap into their democratic predilections. We made the following statement - ‘Only people with specialized knowledge and ability have the right to speak during periods of decision making'. Almost 45% of villagers disagreed with this statement - 28% ‘disagreed' and 17% ‘strongly disagreed', while over 30% ‘somewhat agreed', 12% ‘strongly agreed', and 9% selected ‘don't know'.⁵¹ A second question confirmed the views of Li and O'Brien on the use of petitions by villagers as a means of political expression. In response to the statement ‘If villagers disagree with local policies, they have the right to petition the higher levels', 81% ‘strongly agreed' (41%) or ‘agreed somewhat' (40.3%), while only 9% ‘disagreed' (6.4%) or ‘disagreed strongly' (2.4%).⁵² Only 7% of villagers had no opinion on this issue. The third statement focused on whether poor people had the right to speak out on village-level policies. It stated: ‘People with good economic conditions have more right to speak out on village policy as compared to people with poor economic conditions'. Only 7% of villagers ‘strongly agreed', and 18% ‘somewhat agreed'. Instead, 35% ‘disagreed somewhat' and 29% ‘totally disagreed'. Some 13% either did not know or did not answer. Clearly, people in these villages do not think that the rich should dominate the decision-making process. Finally, villagers strongly disagreed with the argument often proposed by politically influential Chinese that Chinese people accept non-democratic politics if they can make money. This is a common reason given in Hong Kong for the slow pace of democratization. But in response to the statement ‘As long as the village economy shows stable development, there is no need to raise the level of democracy', only 21% ‘strongly agreed' (7%) or ‘agreed somewhat' (14%), while 63% ‘disagreed somewhat' (32%) or ‘strongly disagreed' (33%). Finally, 10% ‘didn't know'. Clearly, the oligarchs are wrong; even if there is good economic development, villagers in these communities still wanted more democracy.

Since each statement can be a good indicator of democratic values, those who ‘strongly supported' any of the above statements received a score of -2 for each answer, while those who ‘agreed somewhat' received a score of -1. Those who ‘disagreed somewhat' received +1, while those who ‘strongly disagreed' scored +2. By combining each person's score on all four questions, we gave them an overall score on the ‘democratic idea' and positioned them on an anti- versus pro-democratic continuum.

¹ On rapid rural industrialization, see Jean C. Oi, Rural China Takes Off (Berkeley: University of California Press, 1999). For a more general overview of rural reform, particularly in Jiangsu Province, see David Zweig, Freeing China's Farmers: Restructuring Rural China in the Deng Era (Armonk, NY: M. E. Sharpe, 1997).

² For an excellent study of ‘peasant burdens', see, Thomas P. Bernstein and Xiaobo Lu, Taxation without Representation in Rural China: State Capacity, Peasant Resistance, and Democratization (Cambridge University Press, 2003).

³ Data from the Fourth Research Institute of the Public Security Ministry show a jump in the number of protests from 9,709 in 1993 to 32,000 in 1999. Li Lianjiang in his seminar, ‘Do Villagers Trust the Central Government', Hong Kong University of Science and Technology, 30 October 2002. By 2005, there were over 86,000 urban and rural protests.

⁴ Lianjiang Li and Kevin O'Brien, ‘The struggle over village elections', in Merle Goldman and Roderick MacFarquhar, eds, The Paradox of China's Post-Mao Reforms (Cambridge, MA: Harvard University Press, 1999).

⁵ Wang Zhenyao, ‘Village committees: the basis of China's democratization', in Eduard B. Vermeer, Frank N. Pieke and Woei Lien Chong, eds, Cooperative and Collective in China's Rural Development (Armonk, NY: M. E. Sharpe, 1998), pp. 239-256.

⁶ On China's democratic development see Suzanne Ogden, Inklings of Democracy in China (Cambridge, MA: Harvard University Press, 2002).

⁷ Kenneth A. Bollen, ‘Issues in the comparative measurement of political democracy', American Sociological Review 45(2), (1980); Dietrich Rueschemeyer, Evelyne Huber Stephens and John D. Stephens, Capitalist Development and Democracy (Chicago: University of Chicago Press, 1992); and Larry Diamond, ‘Economic development and democracy reconsidered', in Gary Marks and Larry Diamond, eds, Re-examining Democracy (Newbury Park: Sage, 1992).

⁸ Samuel Huntington, The Third Wave: Democratization in the Late Twentieth Century (Norman, OK: University of Oklahoma Press, 1991), p. 311.

⁹ Adam Przeworski and Fernando Limongi, ‘Modernization: theories and facts', World Politics 49, (January 1997); and Adam Przeworski, Michael E. Alvarez, Jose Antonio Cheibub and Fernando Limongi, Democracy and Development: Political Institutions and Well-Being in the World, 1950-1990 (London: Cambridge University Press, 2000).

¹⁰ Charles Boix and Susan C. Stokes, ‘Endogenous democratisation', World Politics 55 (July 2003), pp. 517-549.

¹¹ John B. Londregan and Keith T. Poole, ‘Does high income promote democracy?', World Politics 49 (October 1996), pp. 1-30.

¹² Guillermo A. O'Donnell, Modernization and Bureaucratic-Authoritarianism: Studies in South American Politics (Berkeley: University of California Press, 1979).

¹³ See Amy Epstein, ‘Village elections in China: experiments with democracy', in Joint Economic Committee, Congress of the United States, ed., China's Economic Future: Challenges to US Policy (Armonk, NY: M. E. Sharpe, 1997), p. 418; and Tianjian Shi, ‘Economic development and village elections in rural China', Journal of Contemporary China 8(22), (1999), pp. 425-442.

¹⁴ Jean C. Oi and Scott Rozelle, ‘Elections and power: the locus of decision-making in Chinese villages', The China Quarterly no. 162, (June 2000), pp. 513-539.

¹⁵ Hu Rong, Cunmin zizhi jiagou xia de cunmin weiyuanhui xuanju: Fujian sheng 1997 nian ge'an yanjiu [Village Elections within the Framework of Village Autonomy: Research on Case Studies from Fujian Province], City University, unpublished Ph.D. thesis, 1999.

¹⁶ Most of the funding for this research came from the United States Institute of Peace, with further support coming in the form of a Direct Allocation Grant, Hong Kong University of Science and Technology, with funds coming from the Research Grants Council, Hong Kong.

¹⁷ According to the ‘haixuan' system, anyone possessing their democratic rights can run for office. Then, through a series of ballots, the number of candidates is cut until only two or three remain. These candidates compete for Village Director.

¹⁸ See Anhui tongji nianjian, 2000 [Anhui Statistical Yearbook], p. 589, and Zhongguo tongji nianjian, 2000 [China Statistical Yearbook , p. 332.

¹⁹ Heilongjiang tongji nianjian, 2000 [Heilongjiang Statistical Yearbook], p. 439.

²⁰ Either the villages selected were not representative of the county overall, which we doubt, or reported income by the local governments is quite different from that reported by the statistical yearbooks.

²¹ In most villages, villagers gave mixed reports as to whether or not the election had been democratic. Some villagers, perhaps women or migrant workers, simply did not know what types of procedures had been used in the most recent elections. So any village where over 65% of the 20 informants said that there had been multi-candidate elections was considered as having undergone a multi-candidate electoral process.

²² O'Brien and Li cite estimates by the US State Department, Shi Tianjian's nationwide survey, and their own survey, with most estimates falling between 10 and 30%. See Kevin J. O'Brien and Lianjiang Li, ‘Accommodating "democracy" in a one-party state: introducing village elections in China', The China Quarterly no. 162, (June 2000).

²³ Robert A. Pastor and Qingshan Tan, ‘The meaning of China's village elections', The China Quarterly no. 162, (June 2000), pp. 490-512.

²⁴ Villagers in the same village had very different memories of how the election had transpired. So we used a 50% rule; if more than 50% of villagers reported nomination forms that fell into the more democratic category, we scored the village as MDN. In five villages, nine or ten villagers reported that their village had used ‘haixuan', but 10-11 people in the same village said that either they did not know, or that some local organization had nominated the candidates. Our view is that if 50% of villagers did not know what system was used then it was not a well functioning democratic system and therefore we scored it as a less democratic nomination system.

²⁵ For an in-depth analysis of these two cohorts, see David Zweig and Chung Siu Fung, ‘Do new institutions create new politics? The impact of "haixuan" or direct nominations for village elections on elite and villager attitudes', paper presented at the International Symposium: Villager Self-government and Rural Social Development in China, Ministry of Civil Affairs and the Carter Center, 2-5 September 2001, Beijing.

²⁶ MLA refers to the Mantel-Haenszel Test for Linear Association and is a good statistical test of linear relationships of Likert scales.

²⁷ Oi and Rozelle, ‘Elections and power', p. 537.

²⁸ It is worth noting that 41% of villagers in wealthier villages felt that there had been no change. Therefore, if their previous elections had been relatively competitive, this finding says nothing about the real level of competition.

²⁹ Part of these results may be due to the fact that one of our Heilongjiang counties was significantly richer than the Anhui counties, but had held its elections between two and three years earlier. Therefore, the county affect may explain this difference. However, there was very little variation in voting rates among the four counties, suggesting that wealth, rather than locality, had some influence. Similarly, more of Heilongjiang's villages reported having speeches than Anhui's.

³⁰ The relationship between these two variables was significant at the 0.003 level.

³¹ Barrington Moore, The Social Origins of Dictatorship and Democracy (Boston: Beacon Press, 1967).

³² O'Donnell, Modernization and Bureaucratic-Authoritarianism.

³³ Leigh A. Payne, Brazilian Industrialists and Democratic Change (Baltimore: Johns Hopkins University Press, 1994).

³⁴ While a smaller percentage believe that it ‘really speeds up economic development' (16% versus 18%), many more think that it ‘speeds up development' (62% versus 47%). The Mantel-Hanzsel Measure of Linear Association was 0.02.

³⁵ Of entrepreneurs, 76% believe that the charter has a ‘big influence' or ‘some influence' versus 65% for regular villagers. MLA = 0.00.

³⁶ Baogang He, who studied elections in wealthier parts of China, particularly Zhejiang Province, found that 15- 20% of village directors were entrepreneurs. See Baogang He, ‘Village elections, village power structure, and rural governance in Zhejaing', American Asian Review XX(3), (Fall 2002), pp. 55-89.

³⁷ Agglutination occurs when elites accumulate power on more than one dimension, i.e. political and economic power. See Robert D. Putnam, Political Elites (Englewood Cliffs, NJ: Prentice Hall, 1975).

³⁸ In 1998, there were only three elections in the villages we sampled and in two of them the director of the villager committee was elected for the first time.

³⁹ According to Jakobson, in 1997 only 2% of village directors nationwide were also party secretaries. Linda Jakobson, lecture at the Division of Social Sciences, Hong Kong University of Science and Technology, 2000. Perhaps our two Heilongjiang counties were unique. Recent reports, however, suggest that the CCP favors fusing these posts again because competition between these officials has created political instability. Personal communication, Kevin O'Brien, January 2004, Hong Kong.

⁴⁰ Jakobson, lecture at the Division of Social Sciences.

⁴¹ Unfortunately, we used the term ‘cadre' - rather than village director - in the questionnaire, so it is unclear what was entailed in the term ‘cadre'. If they held a village post before being elected village director, they might already have been a party member when they ran for village director.

⁴² O'Brien and Li, ‘Accommodating "democracy" in a one-party state', p. 481.

⁴³ Chi-square test, p = 0.017.

⁴⁴ Nevertheless, it is worth reporting that 53.1% of villagers in MDN villages and 57.7% of villagers in LDN villages selected ‘don't know' when asked who decides how the funds are spent in the village.

⁴⁵ Bernstein and Lu, Taxation without Representation in Rural China.

⁴⁶ Lianjiang Li, ‘Elections and popular resistance in rural China', China Information XV(2), (2001).

⁴⁷ We report this finding, even though it is not statistically significant. But given that we only have a total of 60 cases in Anhui, the value of the MLA is quite suggestive.

⁴⁸ Despite our small sample size of 60 villages, the finding was significant at the 0.08 level.

⁴⁹ A conference at George Washington University, organized by Bruce Dickson, entitled ‘Surveying China', addressed many of these issues, including the difficulty of finding good and reliable Chinese co-researchers, the high costs of surveys, getting local-level  support for carrying out the survey, and problems of measurement.

⁵⁰ Bjorn Alpermann, ‘The post-election administration of Chinese villages', The China Journal no. 46, (July 2001), pp. 45-67.

⁵¹ Selecting ‘don't know' is a common choice in surveys in the PRC. In fact, one can study these responses to explain who is more likely to select this option. In many cases it reflects a lack of political awareness. See M. Kent Jennings, ‘Missing data and survey research in China: problems, solutions, and applications', paper prepared for Conference on Surveying China, George Washington University, Washington, DC, 9-10 June 2000.

⁵² Even local cadres expressed strong support for the villagers' right to petition higher-level officials if they disagreed with ‘local policies'. The question, though, is how they defined ‘local policies'.

 

ISSN 1067-0564 print/ 1469-9400 online/07/500025-21 © 2007 Taylor & Francis
DOI: 10.1080/10670560601026728

The above does not necessarily represent the views of the Foundation.
Reproduction of the article requires written permission from the author.

 by Chung Siu Fung
David Zweig