Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Internal prosperity and external alienation: compounding effects of village elections in rural China
(USC Thesis Other)
Internal prosperity and external alienation: compounding effects of village elections in rural China
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Internal Prosperity and External Alienation:
Compounding Effects of Village Elections in Rural China
Zhenhuan Lei
Department of Economics and Sol Price School of Public Policy
University of Southern California
Degree conferral date: May 2016
A thesis submitted to the faculty of University Graduate School
in partial fulfillment of the requirement of the degree
Master of Arts in Economics
Abstract
Village elections in rural China have profound influence on China. The previous literature
has mainly focused on how village elections in rural China have brought about welfare
improvement within villages, such as more public goods provision, less income inequality,
lower tax burden, and etc. Yet, few, if not none at all, of them have discussed the interaction
between democratic villages and the broader nondemocratic administrative system in
authoritarian China. This paper utilizes a national data containing information about 961
villages from 22 provinces in China to examine both the internal impact of elections on
villages and the interaction between democratic villages and their upper-level governments.
By using instrumental variable method, this paper finds that village elections can (1)
promote public goods provision, (2) hold the village administration accountable, (3) reduce
the amount of financial support from upper-level governments, and ( 4) lead to distrust
between democratic villages and township governments. These results further confirm that
local democracy in China helps promote the internal prosperity within villages, but also
show that village elections can alienate rural China from the formal administrative system.
Key words: democracy, village election, China, public goods, accountability, inter
governmental relationship
Committee:
Jeffrey B. Nugent, PhD, Professor (chair)
Shui Yan Tang, PhD, Professor
Guofu Tan, PhD, Professor
1
Contents
I Introduction ......................................................................................................................................... 3
2 Institutional Background ..................................................................................................................... 5
2.1 The Institutional Context of Rural China ................................................................................. 5
2.2 Goals of Village Elections ........................................................................................................ 6
3 Hypotheses .......................................................................................................................................... 8
3. I Internal Prosperity .................................................................................................................... 8
3.2 External Alienation ................................................................................................................... 9
4 Data and Research Design ................................................................................................................. 11
4.1 Data Source ............................................................................................................................ 11
4.2 Measurement of Dependent Variables .................................................................................... 12
4.3 Measurement of Village Elections .......................................................................................... 13
4.4 Model Specification ............................................................................................................... 13
4.5 The Validity of the IV ............................................................................................................. 14
5 Main Results ...................................................................................................................................... 15
5. I Internal Prosperity .................................................................................................................. 15
5.2 External Alienation ................................................................................................................. 16
6 Implications ....................................................................................................................................... 17
6.1 Local Democracy Matters ...................................................................................................... 18
6.2 Incorporating Villages into the Formal Administrative System ............................................. 18
7 Conclusion ......................................................................................................................................... 19
References ............................................................................................................................................ 20
Acknowledgement
I am grateful to my thesis committee members - Jeff Nugent, Shui-Yan Tang, and Guofu
Tan - for advising my work. In particular, I wish to thank my principal advisor, Professor
Jeff Nugent, who has long been encouraging me to develop this topic into a full-fledged
paper since the time when I first wrote it in his graduate seminar on economic development.
In addition, the writing processes incorporate and reflect several discussions I had with
Professor Shui-Yan Tang on the public administration and governance in China. Professor
Tang always encourage me to delve into the fundamental issues of the governance in
contemporary China and this thesis is an attempt in this line of thought. My thanks also
extend to participants of the Second New Structural Economics Workshop at the Peking
University for their helpful comments on an earlier draft of the paper. The generous
fellowship support from the China Scholarship Council is also appreciated.
2
1 Introduction
For several years, many social scientists from different disciplines have tried to understand
the function and usefulness of village elections (VEs) in China (for instance, Brandt and
Turner, 2007; Luo et al, 2010; Martinez-Bravo et al, 20 14; O'Brien and Li, 2000; O'Brien
and Han, 2009; Oi and Rozelle, 2000; Zhang et al, 2004). These studies have mainly
focused on how VEs in China have promoted some aspects of villagers' welfare and local
governance. These benefits include lower income inequality, better public goods provision,
fair distribution within villages, lower taxes on villagers and others. Meanwhile, we have
also seen another line of literature committed to helping us understand rural China by
focusing on its informal institutions and practices. Most notable among them are Tsai (2007)
and Xu and Yao (2015). Both of them share the similar perspective that traditional practices
and informal institutions, such as clans, can help hold village heads accountable. These
authors have not found a clear relationship between village elections one the one hand, and
the promotion of public goods and holding village administration accountable on the other
hand.
Although these two lines of literature have helped us better understand the internal
welfare improvement within villages caused by VEs, they have not examined Chinese VEs
within the broader administrative system in China. To understand this, one should notice
that China, unlike western democracies, is an authoritarian regime, even though it has
decentralized many economic decision power since its economic reform in 1978 (Xu,
2011 ). This special identity of China means that every policy coming from above should
have its own function of, to some extent, upholding this authoritarian system. With this
understanding, one can see that it would be wrong to evaluate the Chinese VEs only in
terms of local governance and welfare promotion within villages; instead, one should go
beyond this and question what linkages have been formed between China's villages and
their upper-level government entities. In other word, one should consider also the external
effects of VEs.
In this paper, I will call the analysis of welfare promotion (i.e. the analysis of public
goods provision, income disparity, self-governance, tax, etc. within villages) "internal
prosperity." Consistent with most previous studies, I confirm in this paper that VEs can
indeed promote public goods provision. Unlike most previous literature using data from
one province or city, this paper takes the advantage of a national data set providing detailed
information about 96 1 villages in China in the year 1998 and 2002. With this data set, I
have tested three type of public goods provision: education, infrastructure construction and
3
agricultural services. The results show that villages with elections have higher level of
public spending on all three types of public goods, but among them, the effects on
agricultural services is the most significant and robust.
In order to overcome the problem of self-selection, I have employed instrumental
variable (IV) method. Consistent with Heckman Selection Model, this paper uses IV in the
first stage to predict the occurrence of VEs in 1998 and then uses the predicted value of
such occurrences to evaluate the effect of VEs on public goods provision and the
administration accountability. The IV used in this paper is the number of migrant workers
in 1990. This should be a good IV First, it is correlated with the occurrence ofVEs in 1998.
This makes sense intuitively since as more migrant workers worked outside of villages,
they would bring back more information about the country and progressive thoughts during
that time. Second, the number of migrant workers in 1990 should be quite exogenous. It is
hard to believe that migrant workers in 1990 made their decision because they wished to,
for instance, promote the public goods provision in their villages in 2002. Instead, these
villagers mainly went out in the early 1990s for their own personal or family benefits rather
than collective good. The endogeneity issue can be further controlled by adding other
socioeconomic control variables, such as income level, population size and others, to the
regression. In fact, it is this conditional exogeneity criterion (i.e. after controlling for the
socioeconomic factors, the dependent variable in the second stage and the IV should be
orthogonal) that the regressions need satisfy; thus, the various control variables used can
further ensure the acceptability of the IV
As this paper has replicated the results of previous literature, I will also study the other
side of the story - the external alienation. This concept refers to the fact that democratic
villages have difficulty to join in the broader administrative system in China. In the
empirical part, I have found that VEs lower the amount of transfer payments from upper
level governments and increase the number of inspections from above. These findings can
be best interpreted as the distrust between democratic villages and bureaucratic township
government. This may be a dangerous signal since villages are alienated from the broader
administrative system in China.
The remainder of the paper will be organized as follows. Section 2 introduces the
institutional background in rural China. Section 3 proposes four hypotheses which will be
tested in the empirical part. Section 4 discusses the data source and research design. Section
5 reports the empirical evidence. Section 6 further discusses the lessons and implications
from the empirical results. Section 7 concludes.
4
2 Institutional Background
2.1 The Institutional Context of Rural China
The first self-organized village committee (VC) came into being in the early 1980s in a
southern province, Guangxi. Endorsed by major political figures such as Peng Zhen, Bo
Yibo and others in the central government, VC s and VEs were later recognized by and
written in the 1982 Constitution. After this, another huge progress of village self
governance is the passage of the Organizational Law of Village Committee at the National
People's Congress in 1987. However, due to the controversial nature of rural self
governance, the 1987 version of Or ganizational Law of Village Committee did not clearly
specify the meaning of "village election." Depending on the interpretation by the people
who implement the law, VEs could mean either real competitive elections or selection [tui
xuan] - wherein the township or county governments appoint the village head (VH) but
still go through all the required legal procedures - or non-competitive election [deng 'e
xuan ju], in which case the number of candidates equals the number of positions being
elected.
The vague meaning of the term "VE" was later clarified in the revised version of the
Or ganizational Law of Village Committee which was passed by the National People's
Congress in 1998. In this version of the law, it is clearly stipulated that competitive
elections should be implemented in all villages and that every villager should have the right
to vote. Yet, passing the law is one thing; implementing it is another. It was not surprising
at all that even after the passage of the law some township governments still, in fact,
appointed VHs and made up the legal procedures in some way. For these township
governments, similar to replacing the production teams [ sheng chan dui] with VCs in early
1980s, implementing VEs is just another time to change a name and continue the business
as-usual governance mode in villages.
Due to the various implementation criteria, VEs are not organized in the same way
across China. For instance, some may allow free nomination of candidates, anonymous
ballots, public vote counts, and public speeches, while others do not.
1
In spite of these
manipulative characteristics, VEs across the country share some similarities. First, in most
cases, VCs and VHs have a three-year term and there is no constraint on how many times
they can be reelected. Second, the election is usually comprised of a primary and, after the
primary, a formal election. Villagers should have the right to nominate the candidates
before the primary, and the primary will only elect two to three final candidates based on
1 For a more detailed accollllt of the differences in village elections across the colllltry, one can refer to Tsai (2010). As
Tsai sununarized, by manipulating the key procedures of the election, township and collllty governments are capable of
intervening the village self-governance.
5
an open vote count to the formal election. The candidate gaining the most votes in the
formal election will be the VH in the next three years.
Although VHs and VCs are selected in VEs every three years, VHs and VCs are not
the only administrators in villages. For instance, the village CCP branch is a very powerful
political entity in the village. The law has been very unclear about the separation of power
between village CCP brunch and VCs, except for mentioning that village CCP brunch
remains its leading role in rural China. Some scholars have found that both village CCP
brunch and VCs are involved in major decisions such as projects investment (Martinez
Bravo et al, 20 14). Yet, who dominates the decision varies from case to case. Some Chinese
literature has implied that village CCP brunch has relatively stronger power (Xu, 2009).
Another active actor in rural governance is township governments. In China, the top
down administrative system follows this political order: central government -
province/municipality - city - county - township - village. According to this
administrative order, although VEs are stipulated by the central government, the central
government is not directly involved in implementing them. Similarly, the central
government does not deal with rural governance directly. Instead, villages need to interact
with local governments, which in most cases are township governments, for many
governance issues such as family planning, tax collection, business investment and others.
To be clear, it may never be sufficient financially or politically for a village to promote
economic development and provide public goods on its own. Villages need to rely on
township government for financial support (e.g. government transfer payment) and
political support (e.g. the permission for some investment projects). For these reasons, VHs
and VCs still need to follow the orders from above.
2.2 Goals of Village Elections
The intention of introducing local democratic institutions is not as simple as it seems to be.
In the existing literature, we can find at least two explanations for the initial appearance of
VEs in rural China.
To begin, some scholars argue that VEs were created from below. Because in the early
1980s the old local institutions began to break down, the VC was created to fill in the
political vacuum left by production teams [ sheng chan dui] and to provide the needed
collective goods, such as irrigation and infrastructure maintenance, to villagers (Xu and
Yao, 20 15). What the central government has done later is no more than a formal
recognition of the ubiquitous informal practice. This is very consistent with the classic
literature of collective action (for instance, Hayek, 1945; Ostrom, 1990; Ostrom, 2005) and
more recent sociological analysis of China 's economic reform (Nee and Opper, 20 12). The
central government, by empowering villagers and taking advantage of local knowledge,
6
minimizes its agent costs of managing rural China; hence, the central government did not
block such benign democratic practices.
The other line of reasoning focuses more on the top-down perspective. Instead of
seeing VEs as popping up from below, these scholars argue that VEs are planned political
reform (O'Brien and Li, 2000). These scholars believe that the central government
voluntarily gave democracy to villagers in the hope that VEs could combat the rampant
corruption at the local level. In essence, VEs should reinforce peop le's confidence in the
rule of CCP (by wiping out corruption at the local level) rather than only threaten it.
The coexistence of these two different explanations is actually not unusual. In China,
there is a long tradition of two trends competing one another. On the one hand, grassroots
leaders at the local level have long been interested in participating in local governance.
During its processes of modernization, one fundamental issue for China is to find way to
incorporate these grassroots participation into its formal system (Kuhn, 2002). Given the
passion of these local residents, it is not unusual that they invented VCs in the first place
to provide basic local governance when political vacuum existed and that they actually did
well in self-governance (e.g., by providing more public goods, less income inequality and
others). This is the internal prosperity.
On the other hand, the central government is also trying to penetrate into the most
local level of governance. Since the Qin Dynasty (from 221 to 207 BC), Chinese villagers
have been ruled under a centralized bureaucratic system and they have been positioned at
the bottom of this administrative line. Even though villages have existed in China for more
than two thousands of years, the centralized governance system rarely penetrates down to
the village level. The usual practice is to establish formal governance at the county level,
or at most, the township level. The leaders at the village level are not directly appointed by
the higher-level government but are, in most cases, selected by villagers themselves based
on their clan background, age, and reputation (Fei, 1939). As a consequence, public goods
and formal governance have rarely been delivered by the government in rural China, but
are usually organized and provided by grassroots leaders.
After the establishment of New China in 1949, the CCP managed to set up some kind
of governance at the village level. Every village set up a CCP branch to lead the governance
and production in this village. Later on, the name of the village governance organization
may have been changed, but the CCP maintained relatively stringent control over village
affairs before 1978. This formal control over village affairs disappeared after the 1978
economic reform. The VC replaced the production team [sheng chan dui] in 1982. However,
the VC was a weak formal institution whose functions and operations were not clearly
specified even with the passage of the revised Or ganizational Law of Village Committee in
1998.
In essence, VEs are just another round of attempts made by the authoritarian China to
7
establish formal governance at the village level, similar to its previous attempts in history.
This is, of course, a very important process in that both the grassroots leaders and central
government have a strong preference for a unified administrative system. The problem is
that this process is never easy to accomplish. How to incorporate democratic villages into
the authoritarian administrative system is a more challenging task. This incorporation
process can create alienation, rather than cooperation, between villages and the broader
authoritarian system, which I call "external alienation." In the next section, I will further
elaborate on the theoretical grounding for internal prosperity and external alienation.
3 Hypotheses
3.1 Internal Prosperity
One of the most intriguing issues in studying Chinese villages is how villagers, given their
periphery position in Chinese politics, have managed to organize themselves and produce
needed governance and public goods (Tsai, 2007). Because collective action is hard to
coordinate (Olson, 1965; Olson, 1982), villages without strong formal institutions should
have difficulty providing needed public goods and maintaining social order. How do
Chinese villagers manage to uphold the order and produce needed collective goods without
the aid of strong formal institutions?
To answer this question, both Tsai (2007) and Xu and Yao (2015) have emphasized
the informal institutions in villages, such as clans, religious groups and others. They believe
that informal institutions, especially clans, can produce strong moral constraint on VHs'
behavior, and so, VHs have the incentive to produce a satisfactory level of public goods.
Although empirical tests support this claim well, O'Brien and Han (2009) have argued that
clans may do harms to the efficacy of village democracy in the long run. Acting as the de
facto political power in the village, clans challenge the legitimacy and formality of VCs
and VEs. Hence, although clans may in the short term bring about benefits, such as
providing more public goods and social stability, they will hinder the long-term political
development. Moreover, strong clans can strategically interfere in the rural governance and
capture the village administration to serve the clan's interests (Bardhan, 2002). In general,
the influence of informal institutions such as clans is, at best, mixed.
Unlike the literature above focusing on informal institutions, this paper will argue that
the weak democratic institutions (i.e. VEs) still matter in terms of providing public goods
and good governance. As the previous scholarship focuses on informal institutions, the
weak democratic institutions are, by and large, ignored by students of rural governance and
economy in China. This problem has been gradually discussed in some more recent
8
literature (for instance, Luo et al, 20 10; Martinez-Bravo et al, 2014), showing that VEs can
increase the level of public goods provision.
The rationale behind this is simple. To begin, let us assume that both villagers and
township governments have an incentive to control village collective funds. For simplicity,
we can assume that when a village has elections, its villagers decide how to use collective
funds; on the contrary, if this village does not have elections, township government decides
how to use the collective funds. Because the township government does not contribute to
the public funding of the village, the township government may very well use the village
public funds in an inefficient way. Meanwhile, since villagers contribute to the public funds
themselves, they have the incentive to use public funds wisely, such as to build roads,
provide irrigation service, operate village primary schools, etc. Moreover, villagers know
better about their own village because of the local knowledge they have, which can help
increase the provision of public goods. Hence, one should expect to see that public goods
are allocated more efficiently in villages with elections, in which case villagers, not
township governments, decide on the usage of collective funds. Based on this reasoning,
we can conclude our first hypothesis:
Hypothesis 1. Villages with elections will provide more public goods to villagers
compared with those without village elections.
In addition to better public goods provision, VEs can also help hold VHs accountable.
This is exactly one of the major functions of elections. Vo ters use votes to elect more
capable and honest VH to serve the village. In fact, this is also the major reason why the
central government agreed to have VEs in the first place. Peng Zhen, the President of the
National People's Congress from 1983 to 1988 and one of the most important supporters
of VEs, argued that VEs are very helpful tools to wipe out incapable village officials and
CCP cadres, and this will in turn strengthen the rule of the CCP (O'Brien and Li, 2000).
This can lead to the second hypothesis:
Hypothesis 2. VEs can promote the accountability of village officials since they are
under the pressure of being rejected from their positions in the election.
3.2 ExternalAlienation
The more recent literature of local democracy has moved beyond the internal effects of
democracy; indeed it has begun to question how, especially in developing countries where
democracy is not a tradition, local democracy interacts with the broader institutional
context of the country. For instance, Baldwin and Mvukitehe (2015) have found that the
tribal elections in Liberia actually lowered the public goods provision level and led to
internal animosity among people, because the local democracy weakened, rather than
9
complemented, the functioning informal institutions such as traditional tribal practices.
This line of research shows that democracy may not work well if it does not fit into the
existing institutional context. This, in the context of rural China, means that the local
democracy may not work because it is at odds with nondemocratic institutions in China.
According to the past literature, the mainly two existing nondemocratic institutions
functioning in rural China are traditional village practices and the top-down administrative
system. The former refers to the informal practices, traditions, and social conventions that
rural China has been following for a very long time. This is part of the culture as these
practices have been passed down through generations and are enshrined in people's daily
life (Alesina and Giuliano, 2015). Although no systematic empirical research has been
conducted to examine the relationship between traditional practices in rural China and the
recent democratization in villages, there seems to be no strong substitution effects between
the two. For instance, as Xu and Yao (2015) have shown in their work, the traditional values
help hold VHs accountable and hence increase the public goods provision in rural China.
Indeed, these traditional values and practices are mostly correlated with community
welfare, and because of this, it is not very likely that these traditional values and practices
undermine the functioning ofVEs. Yet, due to our limited understanding of this topic, more
systematic examination is needed to confirm this proposition.
The other type of nondemocratic institutions is the top-down administrative system.
This seems to be more problematic. Many scholars believe that the fundamental reason
why VEs appeared in the first place is that the central government intended to use VEs to
wipe out corruption and malfeasance deeply rooted in the top-down administrative system
at the local level. Given this background, township and county governments have long been
quite hostile to grassroots democracy. The research on township governments' hostility to
active local democracy has been mainly limited to the suppression on individuals (O'Brien
1994; O'Brien, 1996) or rigged elections (Zhou, 20 10). Yet, it is not unthinkable that
township governments could make use of some more severe punishments on these
democratic villages. For one, since most villages still rely heavily on township
governments' financial and political support, township governments may very well choose
not to support those most active democratic villages as a punishment. One type of such
support is government transfer funds, which are financial grants made by township
governments to villages. Therefore, we should expect that democratic villages are less
likely to get these government transfer funds.
Hypothesis 3. Villages with elections should get smaller government transfers than
villages without elections.
In addition to punishment, township governments will also invest more time and
resources to control democratic villages. Because democratic villages do not answer to
township governments, township governments need to make sure that these villages do not
10
fail to deliver the tasks passed from above. For instance, one major task for township
governments is family planning. Due to stringent birth control in China, especially in rural
China, township governments usually work hard to ensure that no large-scale transgression
of the one-child policy occurs within their own jurisdiction. This, of course, needs the
cooperation from villages. Yet, when villages do not have the obligation to follow the
orders from the township, that is, when villages have elections and answer to voters,
township governments shall send more officials or working groups to enforce the policy.
Hence we see that township officials make more visits and inspections to democratic
villages to push the policies made from above. This is the fourth hypothesis.
Hypothesis 4. Township officials make more visits and inspections to democratic
villages than villages without elections.
4 Data and Research Design
4.1 Data Source
This paper uses a subset of 2002 Chinese Household Income Project (CHIP) data to
conduct empirical tests. The CHIP is a series of surveys mainly conducted at the household
level. In particular, it aims to compare the househo Ids in rural China and urban area of
China. This project has been conducted with the collaboration between Chinese and
international scholars and with the aid of the National Bureau of Statistics, P. R.C. Until
now, four waves of CHIP data have been available for year 1995, 1998, 2002 and 2007.
The data sets are available online. 2 For this paper, I use the 2002 CHIP data. In the 2002
CHIP data, there is a subset data of village administrative characteristics, which is not
included in other years' data sets. It contains the survey questions for 961 villages from 22
provinces in China (see Table 1).
There are several benefits ofusing this data set. First, this data set contains the answers
to most survey questions for both 1998 and 2002. The original intention of providing the
1998 data is for researchers or policy analysts to compare between the two years. With
these two years of data, we can see how elections in 1998 can affect the socioeconomic
variables in 2002. Moreover, the data set also contains the 1990 data for some survey
questions. This is very useful for constructing instrumental variables as this paper will do.
Second, the project has surveyed 961 villages in China, and to my knowledge, this
has been the largest sample size for Chinese village surveys. In addition, these 96 1 villages
are from 22 provinces in China; hence, the data set is very representative of the nation as a
2 To get the access to the data sets, please visit the website of the CHIP:
http://www .ciidbnu.org/chip/index.asp?lang=oEN.
11
whole.
Third, the data set contains various information and characteristics of the village. This
both helps to conduct impact analyses of elections, and also supplies a large pool of control
variables that can alleviate the omitted variable biases.
4.2 Measurement of Dependent Variables
To test the four hypotheses made in the previous section, there will be four sets of
dependent variables accordingly.
Measurement of Public Goods. In this paper, I mainly consider three types of public
goods: education, infrastructure construction, and agricultural services. I choose these
measures because they are considered as the major public goods provided by villages, not
by higher-level governments (Tsai, 2007). To measure education and infrastructure, I use
the proportion of public spending on these two categories. The larger proportion of public
budget on education and infrastructure, the more public goods of these two categories are
provided. Besides, in CHIP 2002 survey, each village has been asked whether it provided
the following six agricultural services/goods in year 1998 and 2002: (1) collective
irrigation, (2) mechanized cultivation and plotting, (3) prevention of plant diseases and
insects, ( 4) collective purchase of agricultural inputs, (5) collective plantation planning,
and (6) organizing collective labor. They are all recorded as dummy variables in the data
set. To measure the provision of agricultural public goods, I calculate the change of the
number of these six public service/goods provided between year 1998 and 2002. According
to Hypothesis 1, we should expect that villages with elections should, on average, increase
the number of public goods/services provided.
Measurement of Accountability. I use the relative size of public expenditure on
administrative spending as a proxy for accountability of the VC. Since administrative
spending is at the discretion of the VC and VH, it can be used for personal interests of
village officials and CCP cadres. Hence, it should be a good indicator of accountability. We
should see that villages with elections have a smaller size of administrative spending.
Measurement of government transfer funds. To measure the size of government
transfer funds of a village, I use the logarithm form of the transfer fund size in 2002. Based
on Hypothesis 3, we should expect that democratic villages have smaller size of
government transfer fund.
Measurement of the Distrust between the VC and the Township. According to
Hypothesis 4, I use the average number of inspections per month from the township
government to a village between 1998 and 2002 to represent the distrust between townships
and villages. The more inspections the township officials make, the less trust the township
government has over this village. In the empirical tests, we should expect to see that the
12
number of inspections is positively correlated with the occurrence ofVEs.
4.3 Measurement of Village Elections
The major explanatory variable in this paper is the existence of competitive election in
1998, that is, whether the election has more than one candidate in 1998. This excludes the
pseudo-elections that only have a name as elections. For instance, the single-candidate
election [deng 'e xuan ju] or selection [tui xuan] (i.e. appointed by the township in nature
but at the same time going through all the needed election procedures) will not be counted
as competitive elections in this paper 's analysis.
Scholars usually have different estimation over the implementation of VEs. This is
probably caused by the different criteria of elections and the different data source. For
instance, Xu and Yao (2015) have claimed in their paper that almost all the villages in their
data have implemented village election after 1998. Yet, they have not mentioned the criteria
of elections and their sample size is only 220 villages. In contrast, Tsai (2010) reports that
only around 20% of villages had VEs in the late 1990s (i.e. 1997 to 1999); meanwhile, she
also observes a rapid growth of the adoption of VEs. Tsai does not specify the criterion of
elections, and her data comes from 101 villages from five provinces.
Using the CHIP 2002 data set, I have found that 79% of the 961 villages had
competitive elections in 1998 and this percentage was increased to 94% in 2002. This
finding is consistent with Tsai (2010) that more villages are adopting VEs as time goes by.
Since the CHIP 2002 data set has a sample size of 961 and the villages are sampled from
22 provinces, the results reported above should be more compelling and accurate.
4.4 Model Specification
To test the four hypotheses developed in the previous section, I use the following model to
estimate the effect of competitive elections:
Yt = a
k
+ {J k
VE; + X;y
k
+ Ef
(1)
In this model, Y;
k
is the dependent variable, where k can be the public goods
provision in 2002, the share of the spending over administrative costs in 2002, the size of
government transfer fund in 2002, and the average number of inspections per month from
the township government during 1998 to 2002. These dependent variables are used to test
the hypotheses one to four respectively. X; is a vector of control variables, where i denotes
a village. VE; is the variable representing the existence of competitive election in a village
in 1998. a
k
is the constant term and Ef is the stochastic error term.
In this model, because I have considered the time lag of election, the main independent
variable of village election is the existence of competitive election in 1998, while the
13
dependent variables are either the change between 1998 and 2002 (for instance, the income
growth between these two years) or a variable from year 2002. The use of time lag can help
illustrate the logical direction. Usually, things happened earlier affect the results happened
later. Thus, the time lag can help resolve the problem of reversed causal reference.
Another methodological difficulty in this empirical study is that VEs are not randomly
assigned treatment. To overcome this self-selection bias, I will use instrumental variable
(IV) method. Following Heckman Self -Selection Model, I will use the IV in the first stage
to predict the occurrence ofVEs in 1998. Since the occurrence ofVEs is a dummy variable,
the first stage will be a pro bit regression. In the second stage, I will use the predicted value
of the probability of the existence of the VEs in 1998 as the explanatory variable, and run
the regressions according to the equation (1). The research questions, derived from the four
hypotheses, can then be rewritten as whether p
k
equals 0. More specifically, we would
expect to see that p
k
is significantly different from zero.
4.5 The Validity of the IV
The IV used in this paper is the number of migrant workers coming from a given village in
1990. Before presenting the empirical results, we should first ensure that this IV is valid.
This means that this IV should satisf y the following two criteria:
Cov(MW199 0, VE1998) =!=
0
(2)
and
(3)
Here, MW199
0
is the number of migrant workers in 1990; VE1998 is the existence of VE
in 1998; and
£�002
is the stochastic error term from the equation (1).
Equation (2) and (3) together mean that the IV should be correlated with the
instrumented variable but uncorrelated with the error term in the second stage. I will begin
with the discussion of the first criterion, the relevance of the IV
One very important preparation for local democracy is inf ormation sharing and civic
education. Migrant workers in early 1990s helped deliver such preparations. That is, more
migrant workers means that this village was more likely to get new information from
outside and receive the progressive thoughts in the 1990s. Because migrant workers went
back to villages periodically, they were the information bridge between the village and the
outside world. The larger size of the migrant workers means that there are a huger number
of information bridges; therefore, democracy is more likely to take place in the future.
Based on this assumption, the number of migrant workers in 1990 should be correlated
with the existence ofVEs in 1998.
This assumption can be further corroborated by regression analysis. In Table 2, I
14
present the first stage regressions in the Heckman Self-Selection Model. In order to test if
this IV satisfies the relevance criterion, i.e. the equation (2), I include several
socioeconomic control variables and mainly check if the coefficient of migrant worker
number in 1990 is positive and significant. The control variables used in Table 2 are thought
to be highly relevant to the likelihood of VEs in 1998. For instance, I use the income per
capita to capture the income level of the village; I use the existence of social organizations
to proxy for the level of social capital in a village; I also add the school dummies to
represent the level of human capital in a village. I include them as control variables because
the traditional wisdom implies that income level, social capital and human capital are
important factors promoting local democracy. Yet, even after controlling for these variables,
the number of migrant workers in 1990 is still positive and significant at 1 % level.
Moreover, the coefficient of the number of migrant workers in 1990 has not changed very
much across all these regressions. This shows that the number of migrant workers in 1990
satisfies the first criterion of IV.
The second criterion of IV, i.e. the exogeneity of IV, is even harder to test. Yet, we
should expect that the number of migrant workers in 1990 is quite exogenous. This is
because the dependent variables in the second stage are mostly measurements in the year
2002, which is 12 years after the year 1990. It is very hard to believe that the number of
migrant workers in 1990 have a very strong and direct influence on, for instance, the public
goods provision in 2002.
5 Main Results
5.1 Internal Prosperity
In this section, I present the empirical evidence for the first two hypotheses. Following Tsai
(2007), I have used Seemingly Unrelated Regressions (SURs) to estimate the effects of
VEs on public goods provision and administrative costs. I use this method because the
public spending on these categories are correlated with one another. Since the total public
budget is limited, more spending on, for instance, education will squeeze the spending on
other types of public goods. SURs address this issue. It estimates a system of regressions
(in our case, there are four equations) and allows the standard errors in different equations
to be correlated.
The results of SURs are reported in Table 3. It is clear that the existence of VEs is
positive and significant for increasing the spending on education, the spending on
infrastructure, and the provision of agricultural services. The effects on agricultural
services are the more significant than the other two. However, we do not find significant
15
effects of VEs on reducing administrative costs, although the coefficient is negative.
From Table 3, we also find that no control variable has as systematic effects on all
four types of public spending or public goods as VEs in 1998. But, poor town dummy has
a differentiated effects on education and infrastructure on the one hand, and agricultural
services on the other hand. Poor towns refer to those townships officially recognized by
the national, provincial or municipal governments as extremely poor. These townships
usually receive special funding and fa vorable policies from above for economic
development. Hence, the differentiated effects over education, infrastructure and
agricultural services may be related with these special funding and fa vorable policies. It
seems that these special funding and policies favor the investment in education and
infrastructure, and do not contribute to the agricultural services.
Although Table 3 presents some evidence supporting the first hypothesis, SURs do
not address the self-selection bias mentioned earlier. We should still use IV methods to
confirm the results we have got from Table 3. The results of IV method has been reported
in Table 4.
Using IV, we confirm that VEs still have a positive and significant effect on education
and agricultural services. In fact, from Table 4 we see that the coefficients of VEs in these
two regressions become even larger than those in Table 3, meaning that SURs may have
underestimated the effects ofVEs. However, in Table 4, we see that VEs have no significant
influence on infrastructure spending. Meanwhile, VEs have a negative and significant
coefficient in column (3), showing that VEs can reduce the size of administrative spending.
This is different from Table 3 's results. Hence, combining Table 3 and Table 4, we have
relatively strong evidence to support Hypothesis 1 that VEs promote public goods
provision and the effects are more robust on education and agricultural services. At the
same time, VEs can also help hold VHs accountable since VEs have a negative effect on
administrative spending, based on the results from two stage least squares (2SLS)
regressions in Table 4.
For control variables in Table 4, we have found consistent results with Table 3.
Villages in poor towns invest more in education and infrastructure, but have lower
investment in agricultural services. This further supports the assumption made earlier.
5.2 ExternalAlienation
I report the empirical evidence for Hypothesis 3 and 4 in Table 5. In the first two columns
of Table 5, the dependent variable is the average number of township government
inspections per month to a village during 1998 to 2002. Both OLS and 2SLS methods
report that the coefficient of VEs in 1998 is positive and significant at 5% level. The only
difference is that after using IV to address the self-selection bias, the effect ofVEs become
16
even stronger, because we see that the coefficient has changed from 0.682 to 5.667.
In addition, to ensure that no systematic omitted variable bias exists, I have controlled
for several socioeconomic factors. At least two of these controls have strong influence on
the number of inspections from township governments. First, larger villages (measured by
population scale) attract more inspections. This should not be hard to understand since
large villages will have more complex social problems and are more important for township
governments to deliver their policy goals. It is natural that the township government checks
more frequently with these large villages. The other important control variable is
investment project dummy. This variable measures if any business project existed in 2002.
As we should expect, investment projects attract inspections from above because they
provide tax and are important for township governments. Yet, after controlling for these
variables, the effects of VEs remain strong. These results together provide strong support
for Hypothesis 4.
Column (3) to (6) provide some evidence for Hypothesis 3. First, from column (3)
and ( 4), we find OLS and 2SLS produce similar results that VEs reduce the amount of
public revenue for a village in 2002, although only in 2SLS the coefficient of VEs is
significant. We also find that population and income level play an important role: Larger
population and higher income increase the scale of public revenue.
From column (5) and (6), we concentrate on a subcategory of total revenue, the
transfer fund from above. The results from these two columns are consistent with
Hypothesis 3. We see that VEs have a negative coefficient in both OLS and 2SLS, but the
coefficient of VEs is only significant in 2SLS specification. Similarly, this result is robust
when we control for several important socioeconomic factors such as population and
income level. In fa ct, transfer fund is a very important source of financial resources, since
from the data, the average mean for the share of transfer fund to total revenue is 0.45. This
means that on average around 45% of village budget comes from the transfer fund. These
results of total revenue and transfer fund provide evidence that villages with election have
less financial resources at hand and this is mainly because less financial support comes
from above.
To conclude, I have provided some evidence in support of the four hypotheses. In the
next section, I will further explain why these hypotheses are important knowledge to us.
6 Implications
The empirical tests have supported the four hypotheses made in this paper. These tests have
shown that even weak democratic institutions such as VEs in China work to provide better
governance to villagers, but VEs can also create external alienation especially with
17
township governments. However, beyond these impact analyses themselves, what general
lessons and implications can we learn regarding the processes of democratization and the
governance system in China? I will try to answer this question in this section.
6.1 Local Democracy Matters
Many Chinese people still think that elections in China are just appearance. This judgement
is not completely wrong, since these elections are, indeed, not very formal and competitive.
Hence, people gradually undermine the importance of elections. This understanding ofVEs
in China is consistent with the procedural definition of democracy that democratic
institutions are defined by their formal, fair and competitive procedures such as the free
nomination of candidates, the competitiveness of the election, secret votes, and others.
However, this definition implicitly ignores the other side of the story: democratization
is a long-term process. Even the implementation ofVEs is a long and gradual process which
takes up more than ten years before the formal regulation made in 1998 (Shi, 1999).
Recognizing the long-term nature of democratization, the positive return of earlier political
reform is highly important since it provides confidence in the institutional change we are
working on. For instance, as this paper has shown, VEs can help promote public goods and
hold VHs accountable, so people will be more confident in this institutional arrangement
and will support VEs in the future. As a result, this positive return of local democracy forms
a self-reinforcing institutional change path (Greif and Laitin, 2004; Greif, 2006). To better
understand this long-term path dependency characteristics of VEs in China, we still need
further empirical research on this topic.
In conclusion, the positive return of VEs in rural China should not be ignored. The
internal prosperity (i.e. more public goods and less corruption) is crucial to everyone's
daily life in Chinese villages and to people's confidence in democratic institutions. These
implications are especially important for authoritarian regimes whose political leaders and
citizens need more affirmative evidence that local democracy is good to society in general.
Moreover, this paper suggests that further research is needed to study how these positive
return of local democratic institutions in authoritarian regimes can (or cannot) reinforce a
long-term democratization process.
6.2 Incorporating Villages into the FormalAdministrative System
The other crucial factor to consider is the broader political context. In China, there is a long
tradition for the central government to penetrate its political clout down to the village level.
In other words, the central government in China has long been working on a solution that
can minimize the agency costs to an effective control over villages.
18
VEs are the latest attempt made by the central government to incorporate rural
governance into the formal administrative system. Yet, in this paper, we see that this
incorporation process is not smooth. There is a clash between democratic villages and the
authoritarian political system in China. This result is not good for both the village and the
township government. Villages still need the financial and political support from township
governments; meanwhile, township governments also need villages to help implement
policies and deliver various political mandates from above. This clash between township
governments and villages harms both of them by prohibiting potential transactions and
cooperation that can make both parties better off.
The true difficulty is just this. How to smoothly accommodate democratic local units
into the broader nondemocratic context awaits more systematic research. In fact, the
incorporation problems does not stop at the clash between democratic villages and external
administrative system. In other contexts, we may also see that democracy is not supported
by informal institutions, such as traditional practices and social conventions. Hence, a more
general lesson is that how local democracy can fit into the broader institutional
environment, formal or informal, is crucial to the democratization processes. This means
that grassroots leaders and policy analysts should focus on not only internal ramifications
for local democratic institutions but also the external environment that the local democracy
fits into.
7 Conclusion
In this paper, I have used a new data set to examine VEs in China. This new data set
contains information for 961 villages from 22 provinces, which makes the analysis of this
paper more representative and accurate. Taking the advantage of this data set, I have shown
that democratic villages (1) provide more public goods of education, infrastructure
construction and agricultural services and (2) can use elections to hold the village
administration accountable; at the same time, I report that these democratic villages are
more alienated from the formal administrative system since (3) they are less likely to get
financial support from the upper-level government and ( 4) they are closely monitored and
hence less trusted by the upper-level government.
These findings are consistent with previous research on Chinese villages that elections
help promote internal prosperity; meanwhile, this paper also contributes to recognizing the
potential risk that democratic villages are alienated from the external political system.
There is still a clash between democratic villages and nondemocratic top-down
administrative system. To incorporate democratic villages, together with their internal
prosperity, into the formal administrative system is the true challenge that China, as well
as other authoritarian regimes trying local democratization experiments, should seriously
19
consider.
References
Alesina, Alberto and Paola Giuliano. 20 15. "Culture and Institutions." Journal of
Economic Literature 53( 4): 898-944.
Baldwin, Kate and Eric Mvukitehe. 20 15. "Elections and Collective Action: Evidence from
Changes in Traditional Institutions in Liberia." World Politics 67( 4): 690- 725.
Bardhan, Pranab. 2002. "Decentralization of Governance and Developm ent." Journal of
Economic Perspectives 16( 4): 185-205.
Brandt, Loren and Matthew A. Turner. 2007. "The Usefulness of lmperfect Elections: The
Case of Village Elections in Rural China." Economics and Politics 19(3): 454-480.
Fei, Hsiao-Tung. 1939. Peasant Lif e in China: A Field Stud y of Count ry Lif e in the Yangtze
Valley. PhD Dissertation, University of London.
Greif, Avner. 2006. Institutions and the Path to the Modern Economy. New York, NY:
Cambridge University Press.
Greif, Avner and David D. Laitin. 2004. "A Theory of Endogenous Institutional Change.
American Political Science Review 98( 4): 633-652.
Hayek, F. A. 1945. "The Use of Knowledge in Society. " The American Economic
Review 35(4): 519-30.
Hu, Rong. 2005. "Economic Development and the Implementation of Village Elections in
Rural China." Journal ofContemporar y China 14(44): 427-444.
Kuhn, Philip. 2002. Origins of the Modern Chinese State. Stanford, CA: Stanford
University Press.
Li, Lianjiang. 2001. "Elections and Popular Resistance in Rural China (revised version)."
China Inf ormation 15: 1-1 9.
Li, Lianjiang and Kevin J. O'Brien. 1996. "Villagers and Popular Resistance in
Contemporary China." Modern China 22(1): 28-61.
Luo, Renfu, Liuxin Zhang, Jikun Huang and Scott Rozelle. 20 10. "Village Elections,
Public Goods, Investments and Pork Barrel Politics, Chinese-style." Journal of
Development Studies 46( 4): 662-684.
Martinez-Bravo, Monica, Gerard Padro, Miquel, Nancy Qian and Yang Yao. 2014.
"Political Reform in China: The Effect of Local Elections." Working paper.
Nee, Victor and Sonja Opper. 2012. Ca pitalism from Below: Markets and Institutional
Change in China. Cambridge, MA: Harvard University Press.
O'Brien, Kevin J. 1996. "Rightful Resistance." World Politics 49(1): 31-55.
O'Brien, Kevin J. 1994. "Implementing Political Reform in China's Village s." The
Australian Journal ofChineseA ffa irs 32: 33-59.
O'Brien, Kevin J. and Lianjian Li. 2000. "Accommodating 'Democracy' in a One-Party
20
State: Introducing Village Election in China." The China Quarterly 162: 465-89.
O'Brien, Kevin J. and Rongbin Han. 2009. "Path to Democracy? Assessing village
elections in China." Journal ofContemporar y China 18(60): 359-378.
Oi, Jean C. 1996. "Economic Development, Stability and democratic Village Self
Governance. " China Review: 125-144.
Oi, Jean C. and Scott Rozelle. 2000. "Election and Power: The Locus of Decision-Making
in Chinese Villages." The China Quarterly 162: 513-5 39.
Olson, Mancur. 1965. The Logic of Collective Action: Public Goods and the Theory of
Groups. Cambridge, MA: Harvard University Press.
Olson, Mancur. 1982. The Rise and Decline of Nations. New Haven, CT: Yale University
Press.
Ostrom, Elinor. 1990. Governing the Commons: The Evolution of Institutions for
Collective Action. New York, NY: Cambridge University Press.
Ostrom, Elinor. 2005. Understanding Institutional Diversity. Princeton, NJ: Princeton
University Press.
Shen, Yan and Yang Yao. 2008. "Does Grassroots Democracy Reduce Income Inequality
in China?" Journal of Public Economics 92: 2182-2 198.
Shi, Tianjian. 1999. "Village Committee Elections in China: Institutionalist Tactics for
Democracy. " World Politics 51: 385-4 12.
Smith, Graeme. 20 10. "The Hollow State: Rural Governance in China." The China
Quarterl y 203: 601-618.
Tsai, Lily L. 2007. "Solidary Groups, Informal Accountability, and Local Public Goods
Provision in Rural China." American Political Science Review 10 1(2): 355-372.
Tsai, Lily L. 20 10. "Governing One Million Rural Communities after Two Decades: Are
China's Village Election Improving?" In Jean C. Oi, Scott Rozelle, and Xueguang
Zhou (eds.), Growing Pains: Tensions and Op portunit y in China 's Trans f ormation,
143-166. Stanford: The Walter H. Shorenstein Asia-Pacific Research Center.
Xu, Chenggang. 2011 . "The Fundamental Institutions of China's Reforms and
Develop ment." Journal of Economic Literature 49( 4): 1076-1151.
Xu, Yiqing. 2009. "cun Zhuang xuan ju gei zhong guo nong cun dai lai le shen me[What
Does Village Election Bring about for Chinese Villages?]" Xin Zheng zhi jing ji xue
ping lun [New Political Economy Review]
Xu, Yiqing and Yang Yao. 2015. "Informal Institutions, Collective Action, and Public
Investment in Rural China." American Political Science Review 109(2): 371-391 .
Zhang, Xiaobo, Shenggen Fan, Liuxiu Zhang and Jikun Huang. 2004. "Local Governance
and Public Good Provision." Journal of Public Economics 88: 2857-2871.
Zhou, Xueguang. 2010. "Can a Falling Leaf Tell the Coming of Autumn? Making Sense
of Village Election in a Township ... and in China." In Jean C. Oi, Scott Rozelle, and
Xueguang Zhou (eds.), Growing Pains: Tensions and Op portunit y in China 's
Trans f ormation, 167-188. Stanford: The Walter H. Shorenstein Asia-Pacific Research
21
Center.
22
Table 1 Descriptive Statistics
Mean SD Obs
Income per capita in 1998 (Yuan) 2063 1230 959
Income per capita in 2002 (Yuan) 2453 1498 957
Competitive election in 1998 (I � YES, 0 �NO) 0.79 041 961
Competitive election in 2002 (I � YES, 0 � NO) 0.94 0.24 961
Investment project in 1998 (I � YES, 0 �NO) 0.18 0.38 961
Suburb dummy (I � YES, 0 �NO) 0.08 0.27 961
Existence of a primary school (I � YES, 0 �NO) 0.80 040 961
Existence of a middle school (I � YES, 0 � NO) 0. 10 0.30 961
Collective revenue in 1998 (Yuan) 21 70924 1163920 852
Share of education spending 0.11 0 17 653
Share of infrastructure spending 0.20 0.35 675
Share of administrative costs 0. 12 0.14 777
Av erage number of inspections fr om higher-level government between
2.84 4.75 827
1998 and 2002 per month
Existence of a social organization for organizing marriage, fun erals and
0.23 042 961
etc. (I � YES, 0 �NO)
Poor town (I � YES, 0 �NO) 0.32 047 961
Number of migrant workers in 1990 108.61 164.65 938
Collectively managed irrigation/evacuation in 2002 (I � YES, 0 �NO) 0.36 048 961
Collectively managed irrigation/evacuation in 1998 (I � YES, 0 �NO) 0.38 049 961
Collective mechanized cultivating in 2002 (I � YES, 0 � NO) 0 13 0.33 961
Collective mechanized cultivating in 1998 (I � YES, 0 �NO) 0. 15 0.36 961
Collective prevent and cure the plant disease and insect pests unified in
0 17 0.38 961
2002 (I � YES, 0 �NO)
Collective prevent and cure the plant disease and insect pests unified in
0.20 040 961
1998 (I � YES, 0 �NO)
Collectively Purchased agriculture inputs in 2002 (I � YES, 0 �NO) 0.08 0.27 961
Collectively Purchased agriculture inputs in 1998 (I � YES, 0 �NO) 0. 10 0.30 961
Plantation plan in 2002 (I � YES, 0 � NO) 0.28 045 961
Plantation plan in 1998 (I � YES, 0 �NO) 0.28 045 961
Does the collect ive organize the labor force to work out? (2002; 1 �
0.06 0.23 961
YES, 0 �NO)
Does the collect ive organize the labor force to work out? (1998; 1 �
0.05 022 961
YES, 0 �NO)
Change of the number of agricultural public goods -0 09 0.65 961
Planting area in 1998 (mu) 3.56 2.84 956
23
Table 2 First Stage
Dependent variables: Existence of village election in 1998
(I) (2) (3) (4) (5) (6)
Migrant worker number (1 990) 1.217
***
1.269
***
1.243
***
1.186
***
1.290
***
1.293
***
(0413) (041 8) (0428) (0427) (0438) (0433)
Log population y y y y y y
Log income per capita N y y y y y
Social organization N N y y y y
Suburb dummy N N N y y y
Distance (km) N N N y y y
Poor town dummy N N N N y y
Primary school dummy N N N N N y
Middle school dummy N N N N N y
Province dummy y y y y y y
N 938 928 928 922 922 922
R-squared 0.136 0 138 0.145 0 153 0 161 0.163
Note: The unit of the main variable, migrant worker number in 1990 is coded as thousand people. All
regressions include a constant term. All regressions are Probit regression. Heterosce dasticity-robust standard
errors in parentheses.
'
p < 0.1, "p < 0.05, "' p < 0.01.
24
Table 3 Seemingly Unrelated Regressions of Public Spending
(!) (2) (3) (4)
Dependent variables: Education Infrastructure Administrative Fee Agricultural public goods
Election dummy 0.0276' 0.0749' -0 00222 0.178"
(0.0147) (0.0425) (0.0131) (0.0790)
Log population 0.0194' -0 001 15 0.0323 -0 0355
(0.0105) (0.0299) (0.0254) (0.0562)
Log income per capita -0 0219 0.0461 -0 00000252 -0.189"
(0.0162) (0.0468) (0.00000847) (0.0870)
Social organization 0.0223 -0 0465 0.0308" 0.0397
(0.0157) (0.0452) (0.0140) (0.0842)
Land acre per capita 0.00516 -0 00927 -0 00121 0.0565"
(0.00474) (0.0137) (0.00408) (0.0254)
Plain dummy 0.00181 -0 0687 0.0113 -0 0547
(0.0146) (0.0420) (0.0129) (0.0782)
Suburb dummy 0.0367' -0 0949 -0 000287 -0.210'
(0.0214) (0.0618) (0.0192) (0.115)
Distance (km) 0.000621' -0 00105 0.0000966 -0 0000176
(0.00035) (0.00101) (0.000313) (0.00188)
Poor town dummy 0.000723 0.0886' 0.0162 -0.372
***
(0.0163) (0.0471) (0.0148) (0.0877)
Province dummy YES YES YES YES
Constant 0.00388 -0. 109 -0.111 14 63'
0.153 0441 0.186 0.822
N 515 515 515 515
R' 0.184 0.070 0.119 0. 185
Note: Heter osceda sticity-robust standard errors in parentheses. ' p < 0.1,
"
p < 0.05,
"'
p < 0.01.
25
Table 4 Two-Stage Least Square Regressions of Public Spending
(!) (2) (3) (4)
Dependent variables: Education Infrastructure Administrative Fee Agricultural public goods
Election dummy 0.1 56' 0.0604 -0.138' 0.755"
(0.0918) (0.219) (0.0781) (0.363)
Log population 0.0223' -0 00314 -0 00337 -0 0493
(0.0116) (0.0250) (0.0100) (0.0403)
Log income per capita -0 0183 0.0524 0.0173 -0 00112
(0.01 84) (0.0422) (0.0163) (0.0658)
Social organization 0.00779 -0 0392 0.0282 -0 0489
(0.0198) (0.0446) (0.0176) (0.0708)
Land acre per capita 0.00209 -0 0175 0.00758 0.0226
(0.00555) (0.0140) (0.00474) (0.0181)
Plain dummy 0.00299 -0 0408 -0 00480 0.0174
(0.0168) (0.0361) (0.0147) (0.0604)
Suburb dummy 0.0173 -0. 102 -0 0213 -0.213'
(0.0294) (0.0670) (0.0272) (0.115)
Distance (km) 0.0001 12 -0 000490 0.0000154 -0 000736
(0.000410) (0.000873) (0.000335) (0.00127)
Poor town dummy 0.0487" 0.0808' -0 0213 -0.116'
(0.0190) (0.0449) (0.0179) (0.0669)
Province dummy YES YES YES YES
Constant -0 0921 -0.131 0.259 -0.287
0.186 0473 0.163 0.687
N 625 652 746 918
Note: Heter osceda sticity-robust standard errors in parentheses. ' p < 0.1,
"
p < 0.05,
"'
p < 0.01.
26
Table 5 Inter-governmental Relationship
Dependent variables: Number of inspections per month Log revenue Log government transfer
(!) (2) (3)
(4)
(5)
(6)
OLS 2SLS OLS 2SLS OLS 2SLS
Election dummy 0.682" 5.667
***
-0.188 -3.311
***
-0. 169 -3.78 1
***
(0.292) (2.188) (0.128) (0.871) (0.185) (1. 133)
Log population 0458' 048 1' 0. 760
***
0. 750
***
0.306" 0.463
***
(0.236) (0.261) (0.0866) (0.109) (0.140) (0. 158)
Log income per capita 0.351 0.327 1.050
***
0.991
***
0.538
***
0.229
(0.611) (0441) (0.156) (0.1 84) (0.195) (0.300)
Investment project dummy 1.179" 1. 133
***
0 175 0.151 -0 0141 -0 0901
(0460) (0411) (0.129) (0.1 67) (0.150) (0.233)
Social organization I 490" 0.530 -0 0353 0.292 -0 0994 0.287
(0.584) (0470) (0.151) (0.203) (0.252) (0.290)
Suburb dummy -1.336' 0.315 0.0864 -0.384 0437 -0.106
(0.681) (0.811) (0.268) (0.310) (0.368) (0484)
Distance (km) -0.0213
***
-0 0107 0.000788 0.000579 0.00333 0.00420
(0.00712) (0.00844) (0.00234) (0.00358) (0.00325) (0.00456)
Poor town dummy -0 0187 0.396 -0 0394 -0.298' 0.151 -0 0888
(0.360) (0437) (0.1 14) (0.176) (0.151) (0.238)
Province dummy y y y y y y
Constant -2.779 -7.71 1' -1.699 14 32 2 014 6.657"
(5.331) (4487) (1.511) (1 .945) (1.896) (3.105)
N 812 790 851 837 489 487
Note: Heter osceda sticity-robust standard errors in parentheses. ' p < 0.1, "p < 0.05, "' p < 0.01.
27
Abstract (if available)
Abstract
Village elections in rural China have profound influence on China. The previous literature has mainly focused on how village elections in rural China have brought about welfare improvement within villages, such as more public goods provision, less income inequality, lower tax burden, and etc. Yet, few, if not none at all, of them have discussed the interaction between democratic villages and the broader nondemocratic administrative system in authoritarian China. This paper utilizes a national data containing information about 961 villages from 22 provinces in China to examine both the internal impact of elections on villages and the interaction between democratic villages and their upper-level governments. By using instrumental variable method, this paper finds that village elections can (1) promote public goods provision, (2) hold the village administration accountable, (3) reduce the amount of financial support from upper-level governments, and (4) lead to distrust between democratic villages and township governments. These results further confirm that local democracy in China helps promote the internal prosperity within villages, but also show that village elections can alienate rural China from the formal administrative system.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Essays on policies to create jobs, improve health, and reduce corruption
PDF
Essays on education and institutions in developing countries
PDF
A social accounting matrix multiplier analysis of a water supply shortage in the Los Angeles County water service area SEZ 386
PDF
Three essays on social policy: institutional development, and subjective well-being as a cause and consequence of labor market outcomes
PDF
The decisions of migration and remittances in rural China
PDF
A study of Chinese environmental NGOs: policy advocacy, managerial networking, and leadership succession
PDF
Essays in international economics
PDF
The social making of authoritarian environmentalism: protest-litigation nexus and policy changes in China
PDF
What leads to a happy life? Subjective well-being in Alaska, China, and Australia
PDF
Governing public goods: how representation and political power in local and regional institutions shape inequalities
PDF
Supply and demand of economic hierarchy: the Northeast Asia case
PDF
The crisis of potable water in Mexico City: institutional factors and water property rights as conditions for creating adequate metropolitan water governance
PDF
Three essays on economics of early life health in developing countries
PDF
Essays on fiscal outcomes of cities in California
PDF
Governance networks: internal governance structure, boundary setting and effectiveness
PDF
The essays on the optimal information revelation, and multi-stop shopping
PDF
The risks and rewards of city management: how city managers evaluate the nature of the job and compensation
PDF
The effects of interlocal collaboration on local economic performance: investigation of Korean cases
PDF
Synaptic integration in dendrites: theories and applications
PDF
You always say that: physical aggression perpetration, linguistic behavior, and conflict intensity in young adult dating couples
Asset Metadata
Creator
Lei, Zhenhuan
(author)
Core Title
Internal prosperity and external alienation: compounding effects of village elections in rural China
School
College of Letters, Arts and Sciences
Degree
Master of Arts
Degree Program
Economics
Publication Date
04/15/2016
Defense Date
03/25/2016
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
accountability,China,democracy,inter-governmental relationship,OAI-PMH Harvest,public goods,village election
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Nugent, Jeffrey B. (
committee chair
), Tan, Guofu (
committee member
), Tang, Shui-Yan (
committee member
)
Creator Email
reedlei@outlook.com,zhenhual@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c40-229414
Unique identifier
UC11277468
Identifier
etd-LeiZhenhua-4263.pdf (filename),usctheses-c40-229414 (legacy record id)
Legacy Identifier
etd-LeiZhenhua-4263.pdf
Dmrecord
229414
Document Type
Thesis
Format
application/pdf (imt)
Rights
Lei, Zhenhuan
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
accountability
inter-governmental relationship
public goods
village election