Corruption and Accountability in China's Rural Poverty Governance: Main Features from Village and Township Cadres

ABSTRACT

Corruption among village and township cadres is a serious problem in China’s governance of rural poverty. Based primarily on government websites, but also newspapers and interviews, the article analyzes categories, types and the degree of corruption as well as forms of accountability. The findings show that there is more corruption among village cadres than township cadres. The main form of corruption is embezzlement of poverty funds; there is more individual than group corruption; and political accountability is the most important form of accountability. Using an instrumental and a cultural approach, the characteristics of corruption and accountability among village and township cadres are explained. The conclusion is that, although the new anti-corruption policies and laws have been implemented, corruption among village and township cadres is still extensive and difficult to eliminate in the short term. Accountability also has some room for improvement.

KEYWORDS:

• Corruption
• poverty governance
• accountability
• local cadres
• instrumental approach
• cultural approach

Introduction

Corruption is nothing new. Anyone in a position to exercise public power may be tempted to use such power for purposes other than those prescribed by rules and norms (Lü, 2000). Especially since the end of the Cold War, corruption is generally recognized to be a major global problem and foreign policy issue (Banik, 2010; Collier, 2002). This applies to developed and developing, large and small, and market- and non-market-oriented countries (Tanzi, 1998). Corruption usually brings serious problems: it reduces investment and has a significant negative effect on growth (Egger & Winner, 2006; Haque & Kneller, 2015; Mauro, 1995; Pellegrini & Gerlagh, 2004). It also corrodes and harms democracy, or undermines the leaders’ legitimacy in non-democratic states (DeLeon & Holloway, 1993; Johnston, 2005; Porta & Vannucci, 1999; Rose-Ackerman, 1999; Rothstein, 2005; Thompson, 1995; Warren, 2004).

Corruption in China has a long history, but it has become a subject of intense debate and discussion among scholars, researchers, pundits, and commentators, especially since the early 1980 s (Ramirez, 2014). Corruption continues to exist in China and has been called systemic and dynamic (Wedeman, 2004). The current Chinese leadership is therefore conducting a major anti-corruption campaign. Because of the Chinese constitutional structure and political-administrative mechanisms, cadres have vast power. This often leads them “to abuse public office for private gain” (Warren, 2006). Therefore, the corruption of cadres is a central rationale for studying Chinese corruption.

Since 2013, Chinese governments have focused on rural poverty governance, systematically adopting policies to help vulnerable rural groups in the context of marketization and government-orientation. There are two phases of rural poverty governance. The first aims to eliminate poverty generally in rural areas. But, there are some sectors of the population whose poverty has proved difficult to alleviate. These are referred to as the “deep poor population” (shendupinkunqunti). The second phase is therefore devoted to eliminating poverty among the “deep poor population”. Although poverty has decreased among the rural population, corruption among village and township cadres (cunganbu and xiangzhenganbu) is extensive in rural poverty governance, and more and more corrupt village and township cadres have been identified.

The focus in this article is on corruption among village and township cadres in China’s rural poverty governance from 2016 to 2018. During these three years, corruption grew fast and it has become an important component of overall corruption in the Chinese system. It is therefore vital to understand why and how corruption has emerged in China’s rural poverty governance.

Accordingly, three research questions are asked about corruption:

• What characterizes corruption among village and township cadres in China, in terms of the extent, types and avenues of corruption, as seen by selected cases presented by anti-corruption public authorities? Do the two groups have a different corruption profile?

• What characterizes the typical Chinese accountability measures used to combat corruption and what are the criteria used?

• How can two approaches based in organization theory – an instrumental and a cultural one – contribute to analyzing poverty governance-related corruption?

An instrumental approach mainly focuses on structural features of public organizations that may lead to corruption, but also on individual, self-interested behavior (Graaf, 2007). A cultural approach focuses on historical traditions, path-dependency, and informal values and norms (Christensen et al., 2020; Selznick, 1957).

The analysis is based on a combination of in-depth qualitative and quantitative data on corruption and it looks at selected cases of corruption among village and township cadres. Information about these cases was drawn from the website of the Chinese Commission for Discipline Inspection Organs, which keeps national statistics on corruption, but supplemented with other data from the Chinese government and judiciary. Interviews with some regional and local public officials also provided important information. However, regulations restricting the disclosure of government information meant that it was not possible to obtain comprehensive data on corruption in rural poverty governance as a whole, but only on selected cases for research purposes.

In the following, the conceptual framework is presented and then the Chinese context and the methods are outlined. Last,the main results are presented and analyzed.

Analytical framework

Central concepts in a Chinese context

In the following, central concepts in corruption studies are presented in a Chinese context, and it’s shown how it’s a dynamic relationship between corruption types and the reactions from the party/authorities, i.e. the accountability system. Further, corruption types and accountability reactions are seen through two lens, meaning that they are either related to structural and/or cultural factors.

In contemporary China, corruption is defined broadly as the abuse or misuse of authority for personal gain (Rose-Ackerman, 1999). There has been a tendency to define corruption as “virtually any form of ‘improper’ behavior by either a state official or a member of the Communist Party” (Wedeman, 2004). It includes graft (tanwu), bribery (xinghui) and misappropriation of public property (nuoyonggonggongcaiwu), along with seeking illicit benefits for relatives and friends; neglecting official duties; nepotism and favoritism; shirking; retaliation; filing false reports; and the popular definition of corruption often also includes deceit, womanizing and “offending public morality” (Gong, 1994; Kwong, 2015).

Corruption among Chinese village and township cadres displays many of the above-mentioned characteristics. However, corruption among village and township cadres – i.e., low-level cadres – also has a distinctive character. The two categories of corruption are individual corruption and group corruption; the forms of corruption include formalism and bureaucratism, embezzlement of poverty funds, giving preferential treatment to relatives and friends, and receiving bribes.

Individual and group corruption

Traditionally, individual corruption is personal gain or benefits obtained by a public official in exchange for promoting private interests (Thompson, 2013). But, in China individual corruption among village and township cadres not only yields personal gain, violates the norms of public office and harms the interests of the public, but also promotes the interests of cadres’ families and friends or provides benefits to other people who bribe individual village and township cadres to act in their interest. Group corruption among village and township cadres is connected with guanxi. Guanxi describes a network of relationships among individuals who incur obligations to one another, leading to a continual exchange of favors, i.e. an institutional feature (Dunfee & Warren, 2001). These guanxi networks contribute to group corruption in China and embrace both village and township cadres as separate and related groups (Gong, 2002). Group corruption is called woan.

Four types of corruption

The first type is formalism and bureaucratism. Formalism seeks to conform superficially in order to satisfy inspections from superiors, and mainly consists of distortion and falsification of data and the superficial implementation of policy (Li, 2017). Bureaucratism refers to failure to respond to requests, inactivity, delays and other work-to-rule type of behavior. In Chinese it is described as “empty words” (jiadakong), “affecting the pose of an official” (baiguanjia), and “being disconnected from citizens” (tuoliqunzhong). Embezzling poverty funds is defined as grafting of funds, and the poverty funds are usually skimped, transferred and misappropriated. It’s a typical corrupt way among village and township cadres. Giving preferential treatment to relatives and friends, such as illegally defining them as eligible for poverty funds, is indirect corruption and can essentially be classified as nepotism.

Corruption in China bears the influence of historical traditions and culture, and is further influencing newer programs of poverty governance. Officials take bribes in exchange for helping poor villagers to become eligible for or to gain access to poverty alleviation projects and funds, and then utilize their power to extort money from them. Another form of corruption is when village and township cadres examine and approve certain poverty alleviation projects and funds and pretend that the villagers are poorer than they actually are in exchange for money from them.

Accountability and corruption

Accountability has served as a traditional anchor for the modern state since its emergence in the late Middle Ages (Dubnick, 2005). But, accountability is a very elusive concept, because it can mean different things to different people, in different contexts and at different times (Bovens, 2007). Historically, the concept of accountability is closely related to accounting, meaning the use of various kinds of assets (Bovens et al., 2014). In the modern sense, accountability means being called to account by an authority for one’s actions (Jones, 1992). Nowadays, accountability implies that certain superior actors have the right to hold other actors to a set of standards (Grant & Keohane, 2005). Accountability thus means having to answer for one’s action or inaction with respect to those standards, and potentially being subjected to sanctions for failure to comply (Oakerson, 1989).

In China, three accountability types are distinguished (Jiang & Ma, 2015): political accountability, administrative accountability, and legal accountability. Chinese political accountability is different from that in Western countries. Chinese political accountability is, however, based on Communist Party discipline and investigates the actions of those who violate political responsibility (cf. Bovens, 2007). Measures include warnings issued from within the Party and revoking internal Party positions.

Administrative accountability in China is more similar to the Western notion of the term and concerns the extent to which an administrative agency is answerable to its stakeholders for the tasks assigned to it (Wang, 2002). Administrative accountability includes administrative compliance with laws and policies and the efficient use of resources in administrative operations. Legal accountability means that malfeasance and illegal activities will be punished by law. It usually takes the form of detailed investigations, such as legislative oversight (McCubbins et al., 1984; West, 1995) financial or program audits, and reviews of employment grievances by external monitoring agencies (Romzek, 2000). It’s potentially more illusive in China since the judiciary is not independent from the ruling party.

Theoretical approach

There are two versions of an instrumental approach to corruption. First, it directs our attention towards formal arrangements, “bounded rationality” and “administrative man” (Simon, 1957). A formal public organization is here seen as an instrument for achieving goals, and channels and influences the models of thought and decision-making behavior of individual actors (Egeberg, 2012). With respect to corruption, the main question will be how public organizations, in this case those related to poverty governance, are structured and what opportunities this gives for corrupt behavior and control/accountability.

Second, according to the “economic man” theory, the behavior of individual actors in formal public organizations is rational and self-interested and is directed towards maximizing utility for themselves (Buchanan and Tillerson, 1984; Knott & Hammond, 2012). It’s supposes that individual actors their diverse own preferences (Ostrom & Ostrom, 1971); the known alternatives are ranked in a transitive manner (Sears et al., 1980); and individuals calculate the alternatives based on their preference scale (Simon, 1955). In reality, it is difficult to expect very profound rational calculation, but one could expect there to be some self-interested rational thought behind gaming and shirking as elements of corruption (Hood, 2006).

A cultural approach to corruption is based on the notion that public organizations gradually develop unique informal features as a result of adapting to internal and external pressure (Ott, 1989; Selznick, 1957). It emphasizes informal norms, path-dependency and a logic of appropriateness (March, 1994). The logic of appropriateness means that actors are internalizing prescriptions of what is socially defined as normal and good, without involving the calculation of pro and cons of instrumental consequences (March & Olsen, 1995). Organizational culture guides sense-making and actions of individuals in institutions (Scott & Davis, 2006). Every individual in an institution will adapt to the organizational culture and in order to advance certain interests, individuals will cooperate on the basis of common values.

Using such an approach to rural poverty governance, corruption and accountability, the focus is first on whether changes in poverty reduction policy accord with historical cultural norms or not. Further, it’s focused on whether the attitudes towards or actions related to corruption have a long historical tradition or not. Third, it’s asked whether control by and accountability towards the Party has followed the same path in recent years or whether there has been a break in that path (cf. Kingdon, 1984).

The Chinese context

Corruption has been a long-standing problem among village and township cadres in China, but in the late 1980s, it became really serious and villagers often complained about corrupt cadres (O’Brien, 2001). Although more than 30 years have passed since then, corruption among village and township cadres is still evident, but it has not increased so much, since more and stricter anti-corruption policies and laws have been implemented (Fan & An, 2007). Corruption in rural poverty governance is also rather evident. It can be characterized as a typical “wicked problem” (Harmon & Mayer, 1986; Head, 2008).

Since 1949, China has been characterized by a high level of centralization of the authoritarian state. Typical features are respect and deference for paternalistic, strong political leadership; the concentration of power and privilege in Beijing among a relatively small inner-circle elite, surrounded by its “court”; strong bureaucratic influence, with elements of both meritocracy and patronage in different parts of the administration; and a shifting balance or struggle between the central concentration of power and regional and local autonomy and influence (Christensen et al., 2008). Furthermore, leaders of local party committees and governments at all levels also enjoy strong power. Power tends to corrupt and absolute power corrupts absolutely (Acton, 1887). Some argue that a culture of corruption characterizes these systems of Party control and privilege (Lan, 2001; Yang, 2004).

The relationship between the Communist Party and the public administration apparatus is rather complex in China. In the late 1970s, Deng criticized the fact that the Party’s power was excessively centralized and that the Party leadership often became an individual leadership (Fang & Zhang, 2017). He proposed that the relationship should be separated. The new relationship was called “separation of Party and administration” (dangzhengfenkai). One obvious change was that leading Party groups (dangzu) were abolished in all local governments. Although separation of the functions of the Party and the government (dangzhengfengong) was implemented, following the Tiananmen incident Jiang deemed that the Party must give political, ideological and organizational leadership to the state organs of power. Consequently, leading Party groups tried to recover their power in all local governments. Now Xi has stressed that the Party must lead everything, so the relationship of the Party and the administration has become intertwined again. From Deng to Xi, whatever the relationship of the Party and the administration has been, the Party has always been the dominant power (Christensen et al., 2008). The problem is that at all the different levels Party organs and Party cadres have huge power over the administration, which has led to corruption in practice (Ling, 2019).

Township cadres are appointed by a superior department of the Party, while the village cadres are elected by villagers. According to China’s Organic Law of the Villager Committees (cunmin weiyuanhui zuzhifa), the relationship between township and village cadres is characterized by administrative guidance, meaning that village elections exert less political pressure on village cadres (Chen, 2015), and village cadres are less controlled by township cadres. In reality, the mutual supervision of village and township cadres is extremely limited. In recent years, in order to restrain corruption and implement accountability, supervision committees and disciplinary inspectors have been established (cunjijijianweiyuan) in the villages. It should be noted that the supervision committee of a village is based on villager autonomy, the disciplinary inspectors are elected by villagers, implementing villagers’ self-management and self-supervision, so it is regarded as an informal way to hold corrupt village cadres to account. Most village disciplinary inspectors are part-time jobs, but this type of control is more formal. Township cadres, on the other hand, have resources and authority and can influence the behavior of village cadres. Although this is non-mandatory, it increases the dependence of village cadres on township cadres.

Case sources and methods

The focus is mainly on corruption cases from 2016 to 2018, because of availability of data. The main quantitative data used is a sample from a total population of cases that was selected based on certain criteria. The cases are taken from the website of the Chinese Central Commission for Discipline Inspection and Chinese Local Commission for Discipline Inspection. From 2016 to 2018, there were about 200.000 cases of corruption in Chinese poverty governance (Zhang, 2017; Zhao, 2018; Xinhua News, 2019). The website has publicized almost 3500 of these cases, giving various levels of detail. 1325 cases of them were selected: 436 cases from 2016, 440 cases from 2017, and 449 cases from 2018.

The criteria and principles for selecting a more limited number of cases from all the cases presented on the website include three general aspects: first, cases were chosen that illustrated well the different kinds of corruption, divided into two categories: individual corruption and group corruption; four kinds, including formalism and bureaucratism, embezzlement of poverty funds, giving preferential treatment to relatives and friends, and taking bribes; and three degrees of corruption: petty corruption, routine corruption and aggravated corruption. The sample cases selected had to include these various aspects of corruption. The second aspect was continuity – i.e., the categories, kinds and degrees of corruption had to have existed over a longer period of time and could thus be classified as recurrent and serious. Third, strong representativeness, i.e., categories, kinds and degrees of corruption that existed in most provinces, autonomous regions and municipalities directly under the central government and that were also prevalent among village and township cadres.

To deepen our understanding of the cases from the website, it was supplemented by relevant research literatures including articles and monographs, and background information from media of different types. Adding to this, the basis for the qualitative part of the study is mainly that 25 cadres were interviewed – 15 village-township cadres and 10 cadres in municipal government. Of the latter group seven were poverty alleviation cadres and three had leadership positions in the Discipline Inspection Commission, mainly from Guizhou, but also some from Hunan. The respondents were asked questions directed towards their experience with the different types of corruption and accountability, based on their position on different levels in the hierarchy. The answers were written down and systematized in separate texts, and the respondents were promised anonymity.

There are obvious limitations with this set of data, for example, related to the problem of selection bias. First, the cases investigated by the authorities may not accurately reflect the underlying phenomena. Second, the cases published on the central websites may not accurately reflect the totality of investigated cases. Third, the cases selected by the authors may not accurately reflect the cases published on the websites. There are therefore no strong generalization claims, but the data is more a rather comprehensive snapshot, covering three years, of the current situation for poverty governance corruption, based on the selection criteria laid out.

The selected corruption cases in poverty governance 2016-2018

Table 1 shows the main results for number of corruption cases and number of cadres involved. First, the corruption involving village cadres is much higher than the one related to township cadres and increasing over time. This confirms a seemingly more long-term tendency (Fan & An, 2007). Second, the corruption of township cadres is still rather evident. In the following, the corruption of the village and township cadres will be treated together, but when their profile is different, it will be commented on.

Table 1. Corruption cases and corrupt cadres among village and township cadres. 2016–2018

Year	No. of corruption cases*	Type of cadres involved	No. of cadres involved	Percentage
2016	436	Village cadres	548	67
		Township cadres	270	33
2017	440	Village cadres	650	70
		Township cadres	279	30
2018	449	Village cadres	998	74
		Township cadres	351	26
Total	1325	Village cadres	1896	72
		Township cadres	752	28

*The number of corrupt cases each year is the basis for the percentages in following tables.

Individual corruption and group corruption are the two main categories of corruption. Obviously, individual corruption is the main category of corruption, by far, and Table 2 shows a sustained high tendency towards this in rural poverty governance, which is also confirmed by the main tendency from the interviews. Individual corruption mainly involves village cadres and has the highest share overall, while individual corruption among township cadres and group corruption among village and township cadres make up a much smaller share.

Table 2. Individual and group corruption

Year	Corruption type	Percentage	Total no. of cases
2016	Individual corruption	85	436
	Group corruption	15
2017	Individual corruption	80	440
	Group corruption	20
2018	Individual corruption	83	449
	Group corruption	17

Two examples of individual and group corruption could be mentioned from the interviews. First, from 2016 to 2017, X Liu who was deputy director of NW village in Guizhou province diverted collective funds and used 36700 yuan of poverty alleviation funds for personal industrial development expenditure and household expenditure. Second, during 2016, three village cadres in BZY village of Guizhou province, including XX Fu, XX Wei, and X Zhang, took advantage of their positions to obtain 93000 yuan of project funds and divided them privately.

The embezzlement of poverty funds is the most serious form of corruption and have reached a very high and stable level, ranging between 75–80% of all cases in our data, so it seems to be difficult to eliminate. Among the different types of embezzlement of poverty funds, gaining poverty funds illegally accounts for the highest percentage, followed by abusing poverty funds; results that are confirmed by the main tendency in our interview data (Table 3). While defrauding householders of renovation funds and illegally gaining house renovation funds are occurring much less frequent, indicating a stricter control system for these resources.

Our respondents agreed, as shown in the quantitative data, that the other three types of corruption – formalism and bureaucratism (Smolkov, 2000), taking bribes and giving preferential treatment to relatives and friends – overall are of much less importance than embezzlement of poverty funds, the latter one mainly because most villagers know the relationship between the village-township cadres and the local villagers. They give, anyhow, some examples of the three types respectively. First, in 2017 in BZ village of BJ city in Guizhou province, XX Wu who was the secretary of village party branch didn’t organize personnel to go to villages and households to carry out poverty alleviation acceptance and collect relevant information, but wrote the fake poverty alleviation data without investigation. Second, in 2017 in YX village of YY county in Guizhou province, X Yang, who was director of NW village, asked XX Yang, who in charge of the reconstruction of the village’s dilapidated houses, for 5000 yuan. Third, in 2017 in LL village of YL county in Hunan province, XX Xie who was director of LL village, illegally applied for poverty alleviation micro credit for his brother and got himself a bonus of 4000 yuan of those illegally obtained money.

Let us move on to the question of accountability. The Chinese Communist Party (CCP) and Chinese government have taken measures to punish village and township cadres found to be engaging in corruption. Thus, accountability measures – whether political, administrative or legal – are an important element in fighting corruption. The main tendencies are shown in Table 4.

Table 3. Different types of embezzling poverty funds

Years	Types	Percentage of this type	No. of cases
2016	Abusing poverty funds	38	327
	Illegally gaining poverty funds	38
	Defrauding householders of renovation funds	20
	Illegally gaining house renovation funds	4
2017	Abusing poverty funds	35	356
	Illegally gaining poverty funds	56
	Defrauding householders of renovation funds	7
	Illegally gaining house renovation funds	2
2018	Abusing poverty funds	42	359
	Illegally gaining poverty funds	43
	Defrauding householders of renovation funds	10
	Illegally gaining house renovation funds	5

Table 4. Political, administrative and legal accountability measures used to punish corrupt cadres

Year	Cadre type	Accountability type	Percentage of all cadres	Total no. of cadres involved*
2016	Village cadres	Political	80	548
		Administrative	6
		Legal	14
	Township cadres	Political	64	270
		Administrative	28
		Legal	8
2017	Village cadres	Political	90	650
		Administrative	2
		Legal	8
	Township cadres	Political	62	279
		Administrative	20
		Legal	18
2018	Village cadres	Political	88	998
		Administrative	1
		Legal	11
	Township cadres	Political	72	351
		Administrative	15
		Legal	13

*This number corresponds with the number in the second to right column in Table 1.

The table shows, first, that control measures related to political accountability are the most important and most common. Most village cadres and township cadres have been held politically accountable. Second, political accountability is much more important for village cadres than township cadres. Third, administrative accountability is much more important for township cadres than village cadres, while legal accountability is used more than administrative accountability for village cadres. Fourth, legal accountability is more common for township cadres than village cadres, while the opposite is the case for village cadres.

These are all tendencies that are confirmed in our interviews. One example given was from QM town in Hunan province in 2016, where XX Wang, who was the secretary of village party branch of XM village in QM town, diverted poverty alleviation funds to himself. He was then punished through political accountability and was given what is labeled ‘inner party warnings’.

Analysis of corruption and accountability in poverty governance

The main results show that individual corruption is occurring more than group corruption. Moreover, it mainly takes the form of embezzlement of poverty funds, it tends to be individual, and political accountability is the main form of control. The latter reflects the Chinese political system, because most village cadres are Communist Party members and they will be held politically accountable, in line with Party discipline if they are corrupt (Keliher & Wu, 2016; Ling, 2019).

How can one interpret the main features of corruption in Chinese poverty governance, including the accountability measures, based on an instrumental approach? (Christensen et al., 2020). The instrumental approach in its structural version focuses on formal arrangements and bounded rationality (March & Simon, 1958), as related to corruption. Furthermore, individuals in public organizations are limited in their rational calculation for cognitive and capacity reasons, limitations that in principle should be modified by an organizational rationality designed and controlled by the leaders (Dahl & Lindblom, 1953).

Looking first at “administrative man” and organizational rationality, the poverty governance system gives village and township cadres in particular access to a lot of resources, which they can use through their powerful positions in the hierarchy either solely to achieve public goals or else also to enrich themselves. Adding to this, the multi-level inspection system is structurally rather complex and uncoordinated and seems to have problems routing out corruption, as shown by the general increase in corruption. Overall, it is probably not surprising if poverty governance and anti-corruption measures are not easy to implement in the vast and complex Chinese political-administrative system, given that policies in other areas are implemented to varying degrees with big differences between levels and geographical areas (Christensen et al., 2008).

Taking an “economic man” approach to corruption means to focus on self-interest as instrumental to the individual’s attainment of valued goals (Knott & Hammond, 2012). Self-interested actors make decisions not only for others but also for themselves (Mansbridge et al., 2010). Rather often, however, individual rationality may lead to organizational irrationality (Allison, 1971), as evidenced by long-term corruption that is difficult to control.

According to the interviews in this study, self-interest is the key motive for corruption in rural poverty governance. The rational choices of corrupt individuals mainly entail behavior alternatives (Simon, 1955), and individuals seem to weigh up benefits against costs (Ostrom, 1998), in order to rank these alternatives. The different types of corruption may be seen as the result of choosing behavior alternatives that are judged to produce benefits and reduce costs. For instance, the best rational choice for some corrupt individuals is to embezzle poverty funds, because this will bring major benefits without incurring too much cost (to themselves).

For other individuals, giving preferential treatment to relatives and friends or taking bribes is the best rational choice, because it brings benefits to their relatives and friends and enables them to enrich themselves. Self-interest induces corrupt village and township cadres to act publicly in ways that maximize their material interests, whether or not they are inclined to act in the same way privately (Miller, 2001). Because cadres have a number of behavior alternatives, corruption can occur at any time. Each individual thinks that their individually oriented rational choice is the best and least risky one (Tversky & Kahneman, 1986).

Utility maximization does not relate only to the individual, but may also mean group utility maximization (Dunfee & Warren, 2001). Group corruption may reflect systematic collective preferences that maximize utility in the group, or a lack of individuals in the group that oppose it. The cases indicate that utility maximization usually embodies collective embezzlement and distribution of poverty funds. This makes corruption more complex and hence more difficult to control.

According to a cultural-institutional approach (cf. Selznick, 1957), an organizational culture may generate corrupt values and norms and individual corrupt behavior will be determined by these (Rabl & Kühlmann, 2008). The cultural norms and values formed in internal processes over a long period of time may institutionalize corruption. Alternatively, it may be based on societal cultural norms (cf. March & Olsen, 1989). These norms are often stable and may undermine the rules of formal organizations and the anti-corruption measures they take. Corruption as a cultural tradition may mean in certain organizational cultures that not being corrupt is tantamount to betraying the group (Jackall, 1988; Punch, 2000).

Cultural norms and values may play a fundamental role in a dynamic relationship with structural aspects related to corruption (Husted, 1999). Furthermore, a certain culture – or a certain group culture – may lead to a certain mindset, which in turn leads to corrupt behavior (Graaf, 2007). Formalism and bureaucratism in poverty governance are not a new problem, but represent a long cultural tradition in China. Cultural norms are the context in which corruption takes place. In other words, once an organizational culture (or country) is corrupt, anybody who comes into contact with it also runs a big risk of becoming corrupt. The persistent higher percentage of individual corruption reflects the fact that village and township cadres are affected by a corrupt organizational culture and therefore always feel bound to seek self-interest and utility maximization.

One further important question is whether the accountability measures used are effective for fighting corruption. First, Chinese accountability institutions represent a top-down, complete and dynamic structure (Li, 2017), and have rather specific methods. The purpose is to enhance accountability (Wang, 2002). Within highly institutionalized regimes, accountability is routine (Olsen, 2017). In other words, accountability should be effective based on the structure, method and routine. But, the figures for accountability from 2016 to 2018 show that corruption among village and township cadres has not been obviously reduced, despite the anti-corruption control measures. It reflects the fact that corruption is still serious and there is major pressure for accountability.

Second, although political accountability is an extremely important and dominant type of accountability (Bovens, 2007). and scores the highest percentages in China, it does not seem to yield results. Even though the period is rather short, the percentage remained high from 2016 to 2018, so perhaps political accountability does not impose serious sanctions for corruption, because these tend to be a matter of internal Party political disciplinary measures. In other words, those held politically accountable stand only to lose their political identity and political reputation but not to be put in prison.

Third, it seems to be difficult to use legal accountability to prevent corruption. The reason is probably not a lack of internal Party institutions and laws, but the fact that the volume of poverty funds obtained illegally or misappropriated falls below the threshold for punishment foreseen by legal accountability. According to China’s relevant judicial provisions, obtaining poverty funds illegally or misappropriating them must exceed 30000 RMB before legal accountability can be used. Most of the funds embezzled by corrupt village and township cadres are less than this, so legal accountability is hard to use. Generally, since there are many different levels, several disciplinary inspection organs, and accountable institutions and laws, it should in principle be possible to restrain corruption. But, this is not the case, for corruption is a “wicked problem” (Head, 2008).

Conclusion

Our analysis, based on a selected ‘snapshot’ of cases from the websites of the anti-corruption bodies, has shown a certain profile of corruption on lower levels in China. Individual corruption is more common than group corruption, embezzlement of poverty governance funds is the most common type, and political accountability is the most common form of control or scrutiny. The main features are explained in terms of instrumental and cultural approaches.

First, the structure of poverty governance on the regional and local level gives the village and township cadres a position in the hierarchy that enables them to misuse poverty funds (cf. Egeberg, 2012). Adding to this, seen from a self-interested point of view, the costs of individual corruption related to embezzling poverty funds seem to be lower than the benefits, and more so for village than township cadres (cf. Knott & Hammond, 2012). The importance of political accountability is a reflection of the one-party state and represents a less powerful punitive instrument than legal accountability may be in other systems.

Second, corruption in poverty governance is a reflection of path-dependency in broader society, i.e. it has become historically accepted and institutionalized (Banik, 2010). But internal cultural factors also play a role, meaning that over a long period of time a culture has developed on the lower levels of the Chinese administrative apparatus where individual corruption is accepted, something that reflect a relational aspect in the behavior of the street-level bureaucrats. Summing up, the self-interested behavior and political dependence of corruption results from organizational instruments and culture and from a failure to reform government regulations. It is therefore hard to predict and control and may occur at any time.

What is this study adding to the field of corruption studies? First, it adds one of the first comprehensive overviews of the occurrence of different types of corruption in China. Second, it adds to systematic approaches, taken from organization theory (Scott & Davis, 2006), that helps us to understand some of the main mechanisms behind corruptions. Third, as to implications and applications of our main results, they point to increased focus on individual corruption through embezzlement of poverty funds, either through increased control or transparency, but also through cultural measures. Further, may be political accountability is working too leniently, because the corruption is overall not going down, and more cases may be moved to the legal system.

Fourth, a wider point is that our main results from poverty alleviation show a pattern of corruption that seems to be the same in other policy areas (Zhou, 2010; Zhou et al., 2011). The central authorities seems to struggle to implement policy measures at the local level because of the influence from a long-term culture of corruption, the fact that local leaders having a lot of power and resources they potentially may misuse, the cadre selection system favoring leaders showing results that can be based in cheating, and an accountability system that is not working in practice (Kwong, 2015). Many of these factors are also working in other developing countries.

As alluded to in the method section, the main problem with our data is connected to problems of selection biases. First, anti-corruption authorities may have certain structural biases and mental maps when going out in society (Weick, 1995), making them to register corruption in biased and convenient ways, which should warrant deeper studies of their activities. Second, nothing is known about how the anti-corruption authorities select, from all cases, the cases they display on their websites, something that one may study deeper in the future. Do these authorities simple take a representative sample of all the hundred thousand cases handled, or do they systematically select cases that should symbolize political or administrative priorities? Third, of capacity reasons, slightly more than a third of all the cases displayed during three years on the anti-authorities websites were selected, and it’s not known how representative they are, even though they are many and selected in a systematic way and deplaying a similar empirical pattern for each year. So a more comprehensive research effort is needed, covering all the cases over several years.