The Policy and Practice of Microcredit in Rural China: Toward a Relational Understanding of Heterogeneous Implementation

Abstract

Much research on rural development interventions in China has found substantial variation in their implementation at the local level. These studies often analyze the reasons for this heterogeneity by identifying variables responsible for influencing (or distorting) interventions as they travel from central formulation to local implementation. Through an in-depth ethnographic examination of the local implementation of one of China’s largest microcredit programs in three townships in rural Jiangxi province, this article challenges the dominant depiction of external development interventions as reified and linear, and instead proposes a relational approach to understanding local heterogeneity. This approach reveals implementation outcomes to be emergent from negotiations between diverse sets of actors at different levels. At the same time, it becomes clear that microcredit implementation reflects the contradictions implicit in China’s political economy of development. This points to two levels of implementation “relationality,” which results in unpredictable semi-chaotic variation, while simultaneously reproducing wider developmental relations.

Keywords

microcredit rural development policy variation relational approach rural China

This article examines policy formulation and implementation involving one of China’s largest government-run microcredit schemes—the Employment Microcredit Program 再就业小额担保贷款 (hereafter EMP).¹ The primary purpose of this article is to contribute to our understanding of the processes shaping the well-documented phenomenon of heterogeneous local implementation of development interventions in rural China. Through an analysis of the EMP in three rural townships in Jiangxi province,² the article shows that the local implementation of microcredit was the result of relational dynamics between actors at different levels. This ultimately produced complex and emergent outcomes that were more than the sum of their parts, and, therefore, not easily predictable. At the same time, in all three townships the microcredit program also reflected the contradictions implicit in China’s contemporary political economy of development by necessitating the transfer of resources to more prosperous segments of local society—albeit in diverse ways, depending on the place and people involved.

Our findings stand in contrast to mainstream research on the implementation of development interventions—both in China and elsewhere—which often depicts development as the result of a “step-by-step process whereby policy [is] formulated, implemented and then followed by certain results” (Long, 2001: 25). This dominant perspective assumes local implementation outcomes to be the result of a top-down causal chain of events influenced by variables or determinants, and which can be standardized (and even optimized) as long as best practice is identified and adhered to (Hobbes, 2014; White, 2010). We challenge this linear and technocratic understanding of how external development interventions manifest themselves by proposing a relational approach that conceives of interventions as being embedded in, not external to, interlocking sets of developmental relationships at different levels. By focusing on relationships, rather than reified interventions, local variation ceases to be an exceptional result stemming from improper implementation, and instead must be conceptualized as a fundamental feature of externally instigated development interventions, as outcomes are the result of negotiations rather than mechanically predictive. At the same time, this means that external interventions serve to reflect and reproduce, rather than replace or smooth over, the relational dynamics and contradictions that are shaping China’s contemporary developmental terrain at different levels (Loubere, 2016a). Ultimately, this article aims to expose the relational roots of implementation, and thereby shed light on how interventions manifest themselves in complex, emergent, and even chaotic ways, while simultaneously reflecting the often paradoxical and contradictory processes that underpin socioeconomic development in contemporary China.

Background and Literature Review

Understanding the realities of rural development policy implementation in China has taken on increasing importance for academics, policy makers, and development practitioners in recent decades. This is primarily due to the rapidly increasing inequality between rural and urban China, which is reflected in the exceptionally high Gini coefficient (estimated to be over 0.5, indicating extreme income inequality), and highlighted by the fact that average urban incomes are three times larger than those in rural areas (Chen et al., 2010; Yeh, O’Brien, and Ye, 2013). This situation and its associated challenges have, unsurprisingly, caught the attention of the central government. Especially since the beginning of the Hu Jintao–Wen Jiabao administration in the early 2000s, the promotion of rural socioeconomic development and the reduction of urban-rural inequalities have become priority policy areas. In particular, issues associated with the so-called three rural problems 三农问题 have featured prominently in every Number One Document 中央一号文件 issued by the central government since 2004 (State Council, 2004, 2014), and have been the impetus for the formulation of a series of overarching (and often overlapping) policy frameworks, such as the Construction of a New Socialist Countryside 社会主义新农村建设 and Urban-Rural Integration 城乡一体化.

These overarching policy frameworks are composed of a variety of specific rural development interventions targeting various issues or problems that have been identified by governments at different levels. Currently, one of the most popular types of intervention in China is microcredit administered by government agencies. Particularly over the past two decades, microcredit has become one of the government’s primary strategies for promoting rural development, and there are now a variety of microcredit programs targeting different aspects of the “three rural problems.” The EMP, initiated in 2002 by the People’s Bank of China (PBC) and three government ministries as a response to the mass layoffs following the state-owned enterprise (SOE) reforms of the 1990s 国有企业改革, is one of the largest and newest microcredit programs in the country.³ Its goal was to provide loans at subsidized interest rates to laid-off workers, thereby allowing them to start their own microenterprises to sustain themselves. Since 2006, the program has been extended to rural areas in an attempt to lure unemployed migrant workers back to their rural hometowns with the promise of easy credit to facilitate entrepreneurship and modern wage labor in the countryside.⁴

That microcredit has become an integral element in China’s development planning is unsurprising, as the concepts underpinning microcredit as a development intervention are ideologically consistent with China’s adoption of “market socialism.” Indeed, microcredit is an inherently market-oriented development approach, as it is based on the notion that underdevelopment is the result of exclusion from the formal financial and economic systems, and that inclusion in these systems through access to credit—allowing marginal people to undertake entrepreneurial activities—is the antidote (Hickel, 2015; Weber, 2006). Therefore, the adoption of microcredit in rural China embodies the country’s shift toward capitalist market-centered modes of promoting socioeconomic development, which focus on the expansion of commercial enterprises and emphasize the need for development interventions to be financially sustainable—or even profitable. In this way, microcredit programs provide a lens through which to examine the fundamental features underpinning contemporary China’s development landscape, as well as the issues surrounding the design and implementation of development interventions in the country more generally.

Based on this, it is surprising that microcredit programs have received little attention in the large body of research on China’s transformative rural development and the local implementation of development interventions. Although there have, undoubtedly, been some excellent studies exploring the local workings of microcredit programs and exposing how their outcomes are negotiated by local actors, ultimately reflecting heterogeneous local sociopolitical and socioeconomic contexts (Bislev, 2012; Hsu, 2014; Tsai, 2004), this is the exception rather than the rule. Instead, the vast majority of research on microcredit in China has a top-down orientation, and is primarily interested in the institutional organization and structure of microcredit providers or the system as a whole, rather than the actual operation and implementation of microcredit at the local level (Du, 2008; He et al., 2009; OECD, 2003).

When variation in the local implementation of microcredit is discussed, it is often conceptualized negatively as mission drift and/or elite capture through the diversion and embezzlement of subsidized funds. This is depicted normatively as microcredit being provided and used improperly, which is invariably blamed on government interference through market-distorting subsidies and greedy officials/elites eager to take advantage of the programs for their own personal gain (Hofmockel, 2005; OECD, 2003; Ong, 2011; Park and Ren, 2001). For instance, in Rural Finance and Credit Infrastructure in China, a report published by the Organisation for Economic Co-operation and Development (OECD), Du Xiaoshan “suggests strengthening the regulation and supervision of existing programs, adopting flexible interest rate policies, and shifting from subsidy based to fully sustainable, financially self-sufficient programs” (OECD, 2003: 12). Implicit in this type of analysis is a view of microcredit as an inherently external intervention, which needs strict supervision in order to remain detached from the distortions caused by local development realities and to ensure that it operates on purely market principles in order to function optimally. What is missing, however, is in-depth analysis of the relational roots of the interference that shapes microcredit formulation and implementation, or acknowledgment of the fact that interference itself is often a reflection of the contradictions and paradoxes that define China’s contemporary development landscape.

However, while the literature on Chinese microcredit is relatively short on in-depth empirical or theoretical analysis of variation in implementation, this is not true of the wider literature on rural development in China more generally. Indeed, there has been much interest in the implementation of rural development policies at the local level.⁵ In-depth studies have explored transformation and variation in provincial-level development strategies (Donaldson, 2009), the campaign to Open Up the West 西部大开发 (Goodman, 2004; Lai, 2002), the Construction of a New Socialist Countryside (Ahlers and Schubert, 2013; Thøgersen, 2011; Trappel, 2016), the New Cooperative Medical System 新型农村合作医疗 (Brown, de Brauw, and Du, 2009), the Rural Tax and Fee Reform 农村税费改革 (Göbel, 2011), the abolition of agricultural taxes (Linda Li, 2007), earmarked transfers from central to local governments (Liu et al., 2009), the law on specialized farmer cooperatives 农民专业合作社法 (Augustin-Jean and Xue, 2011), and agricultural industrialization (Gao, 2011), just to name a few.

What emerges from these studies is a picture of local implementers as street-level bureaucrats who are intimately involved in shaping, and even defining, development policy through their everyday decisions (Lipsky, 1980; Zang, 2016). This in-depth focus on local agency yields valuable understandings of how and why implementers act as they do. Nevertheless, for the most part, policy implementation is still conceived of as a reified and largely linear process that is influenced by variables and determinants. For instance, in his insightful and detailed examination of provincial-level development strategies, John A. Donaldson argues that a combination of external constraints and opportunities, local conditions, and the personal backgrounds of provincial officials themselves shape development policy formulation and implementation. He then goes on to suggest that this creates a type of “path dependency” that steers policy in a certain direction (Donaldson, 2009: 423). Seen in this way, policy outcomes are influenced by a variety of variables, but once they have been set, the implementation follows a more or less linear trajectory. Hongyi Harry Lai sees the centrally mandated Open Up the West campaign as facing local obstacles related to “corruption, governmental inefficiency, ethnic division, and low economic returns” (Lai, 2002: 432). In other words, he conceptualizes the policy as a reified object that gets distorted (or made improper) by local conditions and implementers who fail to facilitate the manifestation of the policy in its true form.

Implementers are, therefore, seen as being in the driver’s seat when it comes to how policies ultimately unfold at the local level, but they are also, to a certain extent, depicted as operating in the mode of Homo economicus, attempting to manipulate reified policies by rationally maximizing personal gain and shifting their strategies based on external stimuli. Local variation is framed negatively as collusion (Tan, 2008; Zhou, 2010) or more positively as strategic innovation and pragmatism (Ahlers and Schubert, 2013, 2015; Thøgersen, 2011), with implementers playing the role of “wily defiers” or “principled agents” (O’Brien and Li, 1999). For this reason, much research on rural development policy implementation is primarily interested in analyzing the conditions (i.e., sets of variables) that lead local implementers to “successfully” implement policies, rather than distort them.

This large body of literature is undoubtedly invaluable in that it provides important insights into the decision-making processes of local-level officials in contemporary China. However, implicit in much of this research is a misleading conception of policies and interventions as reified entities that traverse a linear path—either from the center to the local or vice versa—being transformed by implementers along the way. This assumption is problematic for a number of reasons. First, rather than focusing on the act of local negotiation itself, this perspective assumes that interventions and their implementation are inherently external and can, therefore, be optimized as long as the different variables are identified and properly accounted for. Second, for the purpose of measurement, it implicitly detaches individuals, variables, and determinants from the complex socioeconomic and historical contexts they are embedded in. This linear perspective, therefore, tends to pay insufficient attention to how long-run relational dynamics shape implementation outcomes in unpredictable and emergent ways. Finally, by focusing primarily on the implementation process in an attempt to define success based on specific (and often normative) criteria, this perspective shifts attention away from the wider political economy of development that produced the developmental problems in the first place, and which invariably shapes how interventions are formulated, conceptualized, and ultimately implemented. This article addresses the limitations that result from these assumptions and proposes a relational stance, which focuses on complexity, chaos, local self-organization, and emergence.

Research Methodology, Data Collection, and Fieldwork Sites

Our argument is built on empirical data collected during in-depth ethnographic fieldwork in Jiangxi province in 2012 and 2013. The fieldwork was designed according to grounded principles to ensure flexibility and openness, and to facilitate the production of local-level theories and concepts emerging from the empirical data (Atkinson and Hammersley, 2007; Charmaz, 2006). We used a range of ethnographic methods, including over a hundred semi-structured interviews and unstructured conversations with a variety of local actors, participant and contextual observation, and the collection of primary documentary data—for example, visual data, policy documents, and local socioeconomic records. All data went through a process of systematic reflection and preliminary analysis while in the field in order to draw out emergent themes and identify important areas for further inquiry (Loubere, 2017). On completion of the fieldwork, data were coded and analyzed using the NVivo software package.

The vast majority of data collection took place in three townships, each located in separate counties in northern Jiangxi and representing distinct socioeconomic contexts. Jiangxi was chosen because it has been selected by the central government as a nationwide “model province” for the EMP, thus providing an opportunity to examine the vanguard of EMP policy development. The township was selected as the unit of analysis because it is the township Office of Human Resources and Social Security (HRSS) that is responsible for the implementation of the EMP in rural areas (see the next section). Therefore, by focusing on the township—rather than the village below or the county above—it is possible to directly analyze the implementation event. It is also worth noting that the township level is underrepresented in the literature on rural development policy implementation outlined above, which primarily focuses on the county level or higher. Townships located in separate counties were selected in order to analyze how EMP implementation emerged out of distinct sets of sociopolitical relationships and structures of power. We also selected three townships with diverse geographical, historical, and socioeconomic conditions, as well as different levels of access to formal financial services, allowing for a comparative analysis of different contexts. Ultimately, the diversity of these field sites provides the opportunity for an in-depth examination of how local negotiations and relationships shape implementation, and how implementation reflects both local particularities, as well as the wider development landscape.

In order to comply with ethical guidelines and to protect the anonymity of our interviewees, the three townships have been given descriptive pseudonyms based on their primary economic activities: Agricultural Township, Migrant Work Township, and Diverse Economy Township. Agricultural Township is located approximately 60 kilometers from a large urban center. However, there are no major transportation linkages, and the roads that do exist are narrow and largely in disrepair. The township center is small and is surrounded by flat agricultural land, which is primarily used for rice farming. Most households are engaged in agriculture as their primary livelihood. Farming income is supplemented by remittances from migrant work, with 30% of the population going out to work in urban areas. Of the three townships, Agricultural Township is the poorest.⁶

Migrant Work Township is 40 kilometers from a large urban center and is connected by a new high-speed road that provides easy access to the city. The township center is larger than that of Agricultural Township, and consists of a range of small, medium, and large shops and enterprises, most of which are engaged in businesses related to the construction of new houses and apartments. The primary source of income for households in this township is remittances from family members engaging in migrant work, and the overall proportion of residents working in urban areas fluctuates between 40% and 70%. Most households have at least one person working in an urban area, and those that do not are often relatively poor. Migrant Work Township is wealthier than Agricultural Township, but the allocation of resources is also more unequal.

Diverse Economy Township is the largest of the three townships and is located on a well-maintained road connecting two urban centers, both of which are approximately 40 kilometers away. The township center is much larger than in the other two townships, and has a diverse range of shops, restaurants, and enterprises of different sizes and scales. Additionally, many people still farm, so there is very little unused agricultural land. Approximately 40% of the local population is currently engaged in migrant work. Diverse Economy Township is the wealthiest of the three townships and also less unequal than Migrant Work Township.

The Policy Design of the Employment Microcredit Program

Microcredit as a development intervention was first attempted in China in the 1980s in conjunction with small-scale projects run by international development agencies and nongovernmental organizations (NGOs). In the 1990s, influenced by the global microfinance movement, domestic nonprofit organizations (NPOs) and NGOs began providing microcredit as well (He et al., 2009; Zhou and Takeuchi, 2010). However, the main source of microcredit in China has come in the form of government-subsidized programs—that is, the government organizes the program and pays part or all of the interest on loans. These microcredit programs have served as important components in China’s overarching rural development policy framework, which attempts to reduce historical rural poverty and mitigate the adverse effects of economic liberalization, such as the growing socioeconomic divide between rural and urban areas (Zhang and Loubere, 2015). Indeed, since the 1990s these government microcredit programs have been expanding alongside the global popularization of the discourse of “financial inclusion,” which advocates expanding the outreach of, and access to, financial service providers as a means of reducing poverty and promoting rural development (World Bank, 2014). This is most clearly demonstrated in the government’s 2014 Number One Document, which stresses the importance of expanding financial service provision to “excluded” segments of the society in order to modernize agriculture and promote rural-urban integration (State Council, 2014).

The EMP is one of the three largest government-run microcredit programs in the country, along with schemes aimed at reducing poverty and providing loans to farmers. It was initiated in 2002 by the PBC, the Ministry of Finance, the State Economic and Trade Commission, and the Ministry of Labor and Social Security (which has since been transformed into the Ministry of Human Resources and Social Security),⁷ as the main element in the government’s reemployment campaign (Loubere, 2016b). The EMP was initially only provided in urban areas to people with official documentation proving their status as laid-off workers. The goal of the program was to provide loans to those who could help themselves through entrepreneurial activity (Jiangxi Provincial Department of HRSS, 2009; Jiangxi Provincial Rural Credit Cooperative Union, 2002; PBC, 2002; State Council, 2002).

As with other development programs, the EMP was first modeled/piloted,⁸ and Jiangxi was selected as the pilot site partly because over one million workers had been laid off from provincial SOEs between 1998 and 2002 (Interviews with officials in the provincial Department of HRSS). In 2006 the EMP was expanded to rural areas, and the target group for receiving loans was expanded to include returning migrant workers, farmers, agricultural cooperatives, as well as other groups (Jiangxi Provincial Department of HRSS, 2012; JXP, 2009; Jiangxi Provincial People’s Government, 2009; State Council, 2006, 2014). The program has been deemed a success and is credited with being directly responsible for 10% of all new employment in the province. For this reason, Jiangxi has continued to expand the EMP in terms of total funding, target area, and target borrowers (Jiangxi Provincial Department of HRSS, 2009; JXP, 2009; Interviews with officials at the provincial, prefectural, and county levels).

In Jiangxi, the program is subsidized by the Ministry of Finance, which pays 100% of the interest directly to the financial institutions providing the loans (meaning that borrowers pay no interest) at a rate of three percentage points higher than the prevailing interest rate set by the PBC.⁹ While this is lower than the interest rate state-owned financial institutions are usually allowed to charge (which is 0.9–2.3 times the PBC interest rate), the interest payment is guaranteed by the Ministry of Finance, which is an attractive proposition for lenders (Jiangxi Provincial Department of HRSS, 2012; Jiangxi Provincial People’s Government, 2009; JXP, 2009; Ministry of Finance, 2008; State Council, 2008). Currently, EMP loans up to RMB 100,000 are available to individuals and up to RMB 4,000,000 to businesses, cooperatives, or projects that “promote employment” (e.g., by hiring the unemployed). The maximum loan term is two years; however, this is decided at the local level (county or township) and normally the term is only one year. Localities can also determine how to secure the loans, the most common method being through a guarantor 担保人 connected to the local government or an SOE.¹⁰ However, the local HRSS office is also allowed to accept credit ratings, mortgages, joint-liability loan groups, proof of assets, or nothing at all (Jiangxi Provincial Department of HRSS, 2012; JXP, 2009).

In rural areas, the application process for the EMP is complicated. Prospective borrowers need to first apply to the township HRSS office, and, if accepted, their documents are sent to the county Bureau of HRSS for approval. After the loan documents have been approved at both the township and county levels, they are then sent to a participating financial institution—usually at the county level—where the final decision is made on whether to issue the loan.¹¹ The financial institution’s decision is supposed to be based on an approach that prioritizes financial considerations: repayment and profitability. Considering the number of different actors and interests involved at different levels, this process can take many months, and can represent a huge expenditure of time for both applicants and administrators (Interviews with officials in the county and township branches of HRSS).

As with other centrally formulated development policies, there is significant room for local discretion built into the EMP, meaning that negotiations are crucial in shaping implementation outcomes at the local level. For instance, while the EMP has national guidelines, the provinces are ultimately responsible for designing their own specific policy guidelines.¹² In the case of Jiangxi, counties and townships have discretion to experiment within these guidelines without asking for permission from above (Interview with official in the provincial Department of HRSS).¹³ This local discretionary power is significant in that local implementers can choose how loans should be secured, whether there should be a quota, and which groups to target (households, businesses, cooperatives, or projects), thus giving local governments the ability to use loans to push forward their own development agenda (e.g., promoting agriculture, infrastructure projects, etc.). Most significantly, national and provincial policies provide a massive loophole for local governments by stating that implementation should always be “based on the actual [local] situation” 根据实际情况 (State Council, 2008), effectively creating an environment where anything goes as long as it works.¹⁴ Nonetheless, due to the “soft centralization” of some township agencies, it is not always clear which local-level groups actually have the discretion, since the Human Resources and Social Security apparatus is sometimes controlled at the county level (i.e., as a “vertical agency”) or at the township level (i.e., as a “devolved work unit”), meaning that the final decision makers differ across localities.¹⁵

Compounding this designed discretion are sets of contradictory imperatives underpinning EMP implementation, which reflect the contradictions implicit in contemporary China’s market-oriented developmental approach, which requires the creation of winners at the expense of losers. For instance, while the program was originally designed to target and assist the relatively marginalized and vulnerable laid-off workers through the provision of inherently risky loans as a social service, it has since evolved in such a way as to adhere to a more neoliberal “financial systems approach” to microcredit, which prioritizes repayment, financial sustainability, and even profitability.¹⁶ Especially since 2011, both the Ministry of Human Resources and Social Security and the participating financial institutions have been increasingly pressured to ensure that loans are repaid (Interviews with officials in the provincial Department of HRSS).

These paradoxical imperatives can be seen in the financial and social capital requirements that are built into the architecture of the EMP. For instance, prospective borrowers are explicitly required to have sufficient financial capital to ensure repayment (Jiangxi Provincial Rural Credit Cooperative Union, 2002), and the short term of the loans (one to two years) suggests that those without resources should not bother to apply. Indeed, it is very difficult to start a profitable enterprise from scratch and fully repay a loan within such a short time. Moreover, the loan policy explicitly states that borrowers must be “honest and trustworthy” 诚实守信 and have a good reputation and standing in the community (Jiangxi Provincial People’s Government, 2009). This means that local government agencies and financial institutions vet prospective borrowers (and their guarantors) based on socioeconomic status, effectively excluding many of the marginalized people and groups that the program is supposedly designed to assist. The EMP is, therefore, embedded in a set of developmental relationships (comprising the wider development landscape) that implicitly and explicitly require the production of inequality through the transfer of resources to more prosperous places and people (Loubere, 2016a).

A Tale of Three Townships: Implementing the Employment Microcredit Program in Rural Jiangxi

Because of the contradictory imperatives built into the EMP policy outlined above, the program unavoidably contributes to reproducing and reinforcing relational inequalities that have emerged from China’s contemporary market-oriented rural development approach. At the same time, however, because of the local discretion that is also a fundamental feature of the policy design, the implementation of the EMP that emerges from the negotiations between diverse actors, and ultimately reflects these inequalities at the local level, is highly heterogeneous and unpredictable. This section details the reality of EMP implementation, and how the program has been perceived and negotiated in Agricultural Township, Migrant Work Township, and Diverse Economy Township (see Table 1).

Table 1.

Implementation of the Employment Microcredit Program in Three Townships in Jiangxi, 2011–2013.

	Agricultural Township	Migrant Work Township	Diverse Economy Township
Level of administrative control	County	Township	Township
Quota	RMB 1,100,000 (set by the county)	RMB 600,000 (set by the county)	None
Borrowers	12-member farmer cooperative (all 22 loans)	Friends and family of township officials	2012—Any eligible applicant 2013—N/A (program halted)
Borrower selection	Social connection between the county HRSS and the leader of the farmer cooperative	Social connections between the borrowers and the head of the township HRSS	2012—Based on policy guidelines 2013—N/A (program halted)
Use	Build two modern vegetable greenhouses	Investment, consumption, informal on-lending	2012—Various 2013—N/A (program halted)
Loan security	Township government acts as guarantor	Each borrower must find an eligible guarantor (connection with local government or SOE)	2012—Any approved method 2013—N/A (program halted)
Loan amount (per person maximum)	RMB 50,000	RMB 50,000	2012—RMB 100,000 2013—N/A (program halted)
Loan term	1 year	1 year	2012—1 year 2013—N/A (program halted)
Perception (providers)	Pride—The county and township HRSS both hope this can become a “new model” for the Employment Microcredit Program	Indifference/annoyance—The township HRSS is under pressure to ensure 100% repayment from the county; the loan process is time consuming and the officials have many other more important tasks	Indifference/annoyance—The loan process is time consuming and the officials have many other more important tasks
Perception (borrowers)	Pride—The members of the farmer cooperative are proud to be involved in a “new model” and to be using technology from a “more developed” region (Jiangsu); they also believe that the vegetable greenhouses can increase their income and improve their lives	Indifference/apathy—The borrowers have many other sources of credit (both formal and informal) and significant amounts of personal capital; the loans represent a small amount of extra capital for investment or consumption	Indifference/apathy—Many other sources of credit are readily available (both formal and informal) in the township; most people think the loans are time consuming (particularly the training element) and represent a small amount of money

Note. HRSS = Human Resources and Social Security; N/A, not applicable.

Implementation in Agricultural Township

In Agricultural Township, the county Bureau of HRSS initiated the EMP in 2008, and set a quota of RMB 1,100,000—with a maximum loan size of RMB 50,000 over a one-year loan period—per township, per year.¹⁷ In this county the Human Resources and Social Security apparatus is controlled at the county level (as a “vertical agency”). County officials have taken the lead in the implementation of the EMP and have decided that the program should be used to support agriculture, increase farmer incomes, and help farmers become more entrepreneurial. Therefore, HRSS officials at both the county and township levels have been instructed to target borrowers who are able to help achieve these goals (Interviews with officials in the county and township branches of HRSS).

The director of the county Bureau of HRSS was introduced to a 12-member farmer cooperative 农民专业合作社 based in one of Agricultural Township’s villages through mutual friends of its leader (who is also the village party secretary) and officials in the township HRSS office. This local leader was described by both county and township officials as being a very “capable person” 能人 who is good at organizing the villagers. He had also gained technical knowledge of how to build vegetable greenhouses from a party-organized “model tourism” trip to Jiangsu province in 2006, which led to the construction of some small vegetable greenhouses in the village.¹⁸ Through this social connection, the county and township branches of Human Resources and Social Security agreed to support the farmer cooperative in an application for the entire township loan quota (RMB 1,100,000) to build two large modern greenhouses to scale-up the cultivation of more profitable vegetables year-round.

Rather than applying for the larger EMP loan for companies, cooperatives, or projects (which is only 50% subsidized), the members of the cooperative were instructed to apply for loans individually in order to receive the full subsidy (viz., 0% interest). The Human Resources and Social Security officials in both the county and the township then lobbied the township government to act as the guarantor for all the loans to ensure that the cooperative’s application for the EMP would not be deemed too risky by the county financial institution—in this case the county rural credit cooperative.¹⁹ The members of the farmer cooperative then gathered together the necessary documentation with the support of the township and applied for the EMP; however, since there were only 12 members in the farmer cooperative applying for the full quota of 22 loans, the household registration 户口 documents of other villagers were used in the loan application without their knowledge, which is effectively identity theft and means that the farmer cooperative added fictitious members in order to qualify for the loans.²⁰ Ultimately, the full quota of EMP loans was granted to the cooperative, and during our visit the greenhouses were under construction (Interviews with officials in county and township HRSS branches; interviews with members of the farmer cooperative; contextual and participant observation).

The people we spoke to who were directly involved with the EMP in Agricultural Township were proud of what had been accomplished, and were committed to ensuring that the program in the township became a success. The members of the farmer cooperative felt pride in the fact that they were using farming technologies and methods that come from “more developed” provinces (Jiangsu), and they were excited at the prospect of increasing their income and reducing the risks associated with a single crop (rice) (Interviews with members of the farmer cooperative). The implementers (officials in the county and township branches of HRSS) considered Agricultural Township’s experience to be a “new model” for EMP provision that allows for more efficient, convenient, and secure service provision through group loans with one guarantor—in this case the township government. They also felt that this model has the potential to promote cooperative and agricultural development, and they hoped that it would be adopted in other places.²¹ For instance, an official in the county Bureau of HRSS told us, “The EMP is one of the ‘calling cards’ [priority policies] of the provincial Department of Human Resources and Social Security. . . . We hope to make a breakthrough with this new model, which represents an innovation in EMP service, and is unique in the province. Of course, this innovation is within the permitted scope of the policy” (Interview with an official in the county Bureau of HRSS).

However, some of the other villagers who were not members of the farmer cooperative or directly involved in the EMP felt excluded and even resentful, and wished that they could join the cooperative to gain access to interest-free credit. For instance, one of the poorer households in the village consists of an elderly couple taking care of their three grandchildren (two boys and one girl) while their children work in the city. The household income comes from small remittances from the children’s migrant work and from the rice the grandparents cultivate even though they are nearly 70 years old. They told us they do not have enough money to send the granddaughter to high school (the little money they do have is for the two grandsons), and that they wished they could join the farmer cooperative to increase their income: “Now there are greenhouses made of steel, but we normal people can’t join the farmer cooperative and plant our crops in them. Only officials and party members can do this. We want to join the farmer cooperative but aren’t given a place” (Interview with a non-borrowing householder in Agricultural Township).

And it is not only the poor who are excluded. Even the arguably most entrepreneurial household in the village—which runs the largest shop and has a van they contract out for transporting rice and other agricultural products—told us they have no access to any formal credit or government microcredit programs in the township. For this reason, they have to travel to the county seat to borrow money for their businesses from financial institutions at standard (higher) interest rates and using collateral, such as their van. They also want easily accessible interest-free loans to support their entrepreneurial activities:

Of course we wish we could get this type of microcredit! If we had easy access, we could use this money to restock our store and the loans could also be used to buy pesticides and fertilizer for farming. . . . But, we haven’t even applied because we have no money, don’t know people, and have no social connections 关系, so they won’t accept our application. . . . If the process was more open and fair, we’d apply for the EMP. (Interview with a non-borrowing householder in Agricultural Township)

Implementation in Migrant Work Township

In Migrant Work Township, the county Bureau of HRSS initiated the EMP in 2008, and set a quota of RMB 600,000 per township per year, with loans ranging from RMB 10,000 to RMB 50,000 for a period of one year. The Human Resources and Social Security apparatus is controlled by the township (as a “devolved work unit”), with visibly less material and administrative support than its counterpart in Agricultural Township. The township branch of the HRSS is in an office with four people responsible for a number of “hanging sign work units” 挂牌子单位.²² In reality, there is only one local official dealing with the EMP. He also has a large number of other responsibilities for both the Office of HRSS and the other work units, many of which arise suddenly and unexpectedly—such as inspections from higher levels of government. Moreover, the county Bureau of HRSS has delegated full responsibility to the township for finding and vetting borrowers, helping with applications, and ensuring that all the requirements are met; but the county has retained final decision-making power over who gets the loans (Interview with the official in charge of implementing the EMP in Migrant Work Township).

The county Bureau of HRSS also set different developmental goals for the EMP. Officially the program was supposed to promote economic development—as opposed to the goal of agricultural development in Agricultural Township. Unofficially, however, the county ordered the township to achieve 100% repayment, as this would result in a monetary reward for the county from the province equaling 1% to 2% of the loan total. Due to the lack of time/resources and the top-down pressure to ensure repayment—and because there was no bottom-up pressure from cooperative groups or interaction between potential local borrowers and the county Bureau of HRSS like there was in Agricultural Township—the township official decided to target friends and family members that he knew he could trust to repay the loans, and who would be accepted by the Bureau of HRSS and financial institutions at the county level. During our fieldwork, the entire quota had been lent to this target group (Interview with the official in charge of implementing the EMP in Migrant Work Township).

In contrast to the pride expressed by those involved in the program in Agricultural Township, the implementation of the EMP in Migrant Work Township engendered feelings of exhaustion, annoyance, and/or indifference. The township official told us that the EMP causes a significant amount of stress. This is not only because it is getting increasingly difficult to find a sufficient number of new borrowers every year (each with a suitable guarantor) that he can trust to repay the loans, but also because the lack of engagement from the county Bureau of HRSS means that the entire process is time consuming, which makes the program unappealing to potential borrowers. However, he feels that he has no choice but to prioritize the program, because he fears his job promotion prospects will be damaged if he does not continue to ensure a 100% repayment rate. He described the situation by saying:

We have four people working in the office, and besides the EMP we’re also in charge of social security, health insurance, work placement, employment training, and so on. A lot of work, a multitude of things. . . . The application and decision process for the EMP is too long. It takes at least a month and sometimes even two or three months for a decision, and the loans are often held up by the county Bureau of Human Resources and Social Security or the rural credit cooperative, and so applicants can’t get the money for a long time. (Interview with the official in charge of implementing the EMP in Migrant Work Township)

The local borrowers we spoke to during our fieldwork were also largely indifferent. Some were using the money to add to other investments or for consumption purposes, and some were using it to informally lend to others in the community at up to 12% annual interest. However, overall the borrowers were not particularly interested in the EMP, and perceived the loans as simply representing a small amount of extra capital. In reality, most of the EMP borrowers in Migrant Work Township have little trouble gaining access to substantial loans from other formal and informal sources. Therefore, rather than feeling grateful toward the township official for providing them with interest-free loans, some of the borrowers actually felt as though they were doing him a favor because the application process was so time consuming (Interviews with EMP borrowers in Migrant Work Township).

Most of the non-borrowers we spoke to were also indifferent, and did not know about the EMP and/or were not interested in applying for any type of formal loan. This was usually because they had enough income from remittances to support themselves and could borrow informally from friends and family if necessary—for example, for medical emergencies or other crises. In the words of a members of a non-borrowing household: “We don’t do business so we don’t need a loan. Borrowing from the rural credit cooperative is very troublesome. . . . It’s too annoying and complicated to provide collateral or get a guarantor. . . . If I need money urgently, I won’t go to the bank; I’ll borrow from my friends or family” (Interview with a non-borrowing householder in Migrant Work Township). At the same time, there were also individuals and households who wanted loans, but were excluded because they did not meet the financial and social capital requirements. These individuals and households were often relatively poor and marginalized in the community. Usually they wanted the loans to deal with immediate needs, such as fixing structural damage to housing, medical issues, paying for weddings, and so on. Their responses to being excluded from the EMP and other types of formal credit ranged from dulled acceptance of the fact that they do not have enough social connections, to desperation and even resentment verging on anger framed in terms of class conflict. For instance, one group of local construction workers told us:

We can’t get loans. If we want to borrow, we need to find a guarantor and have collateral . . .

Without a guarantor we can’t borrow much money, maybe a few thousand yuan . . .

Guarantors have to have money, and only those in a good situation can make a contact with the rural credit cooperative—only 10% of people are in a good situation, while 90% are in a bad situation. (Group interview with non-borrowers in Migrant Work Township)

And the disabled head of one of the poorest households in the township told us: “It’s very difficult for us to borrow money because we have no social connections. The poorer we are, the harder it is to get a loan. The EMP is like the icing on the cake for the rich, but it doesn’t provide the poor with any temporary relief” (Interview with an excluded householder in Migrant Work Township).

Implementation in Diverse Economy Township

As in the other two townships, in Diverse Economy Township the EMP was initiated in 2008. However, the county Bureau of HRSS neither set any quota for the number of loans that should be provided nor did it designate a development agenda for the program—such as agriculture as in Agricultural Township and economic development as in Migrant Work Township. In Diverse Economy Township, Human Resources and Social Security is also controlled at the township level (as a “devolved work unit”); however, it has visibly more resources than its counterparts in Migrant Work Township or Agricultural Township, such as more staff, a nicer work space, and no evidence of “hanging sign work units.” Moreover, the officials responsible for the program have not been pressured by the country to ensure 100% repayment rates. Therefore, the township HRSS office initially provided loans to anyone who was eligible, and in 2012 it accepted 52 loan applications for RMB 100,000 each (Interviews with HRSS officials in Diverse Economy Township).

The loan application process was essentially the same as the other two townships, with one important difference—in addition to all the other requirements, borrowers also had to travel to the county seat to complete a week of training on how to effectively use their loans. This meant that the 52 loans represented a huge amount of administrative work, not only for the township officials, but also for county-level HRSS officials (who needed to organize these training sessions) and the financial institutions providing the loans (who needed to vet a higher number of township borrowers and guarantors). Therefore, in 2013 the EMP was “temporarily” halted by the township HRSS office in order to better allocate scarce time and resources to other more “useful” activities and services.²³ Ironically, the lack of a quota set by the county, which originally resulted in a larger number loans being provided than in the other two townships, ultimately became the justification for halting the EMP altogether, as the township HRSS office could exercise its discretion and simply decided that no one was eligible (Interviews with HRSS officials in Diverse Economy Township).

In Diverse Economy Township, the prevailing attitude toward the EMP was slight annoyance on the part of local officials (although not as pronounced as in Migrant Work Township) and indifference on the part of others in the community. Township officials felt that the program was unnecessary and a waste of time, so they reinterpreted the lack of a quota to mean that the provision of the program itself was optional. Most of the households we spoke to did not know about the existence of the EMP. Even those that did know and were considering applying for one of the microloans were not overly upset when they discovered that the program had been abandoned. This is mainly because Diverse Economy Township was one of the original pilot sites for the rural credit cooperative microcredit program, and approximately 80% of all households have received a rural credit cooperative microloan. Moreover, most households would much rather borrow interest-free informal credit from friends and family, due to the convenience and lower transaction costs—for example, no week-long training at the county seat. Nevertheless, as in the other two townships, during the period when the EMP was running there were excluded segments of the population. However, they were largely unaware of the existence of the program, and were more upset about being excluded from the more widespread rural credit cooperative microcredit program and/or from formal loans provided by the local branches of the Agricultural Bank of China and the Postal Savings Bank of China (Interviews with officials and households in Diverse Economy Township).²⁴

Embedded and Negotiated Variables Influencing Implementation

The examples above illustrate that EMP implementation is highly heterogeneous in different contexts. The actors involved conceived of microcredit in a variety of ways, leading implementers to target different individuals and groups, and to use very different methods to provide the loans (or not). The depictions of variation in implementation outlined in this section confirm the importance of some of the key variables that are often identified in the literature on local policy implementation in China. In particular, all three examples highlight that diverse socioeconomic contexts determine how interventions slot into the development terrain, how pressures and incentives influence decision making, and how the different personalities and backgrounds of implementers shape their perspectives on how implementation should occur.

However, in contrast to much of the mainstream research on development policy implementation, which often seeks to identify static variables for predictive purposes, the examples in this article show that the variables themselves are highly dynamic and manifest in diverse ways at the local level. This is because the entire implementation process is shaped by sets of interlocking developmental relationships. Thus, variables are not static entities that influence (or distort) standardized preexisting policies traveling from the center to the locality in a way that can be quantified, measured, and predicted. Rather, they are inherently relational in origin and nature, and therefore not only reproduce the wider developmental relationships that encourage the transfer of resources to winners, but also reflect particularistic relational dynamics at the meso and micro levels.

For instance, the three townships each represent distinct financial landscapes with different “relations, norms, actions and processes . . . [that] influence, directly or indirectly, the transactions of savers, borrowers and lenders” (Bouman and Hospes, 1994: 1). These diverse local contexts result in financial markets that are differently segmented in terms of what financial services are available, what financial services are used for, and, more broadly, how credit is conceptualized socioeconomically. This means that external financial interventions—such as microcredit—necessarily slot into different positions, and thus serve different functions in different places and for different actors (Tsai, 2004).

In Agricultural Township, the dominance of agriculture as the local livelihood strategy, combined with the fact that the county and township governments identified agricultural modernization as a developmental priority, meant that implementers and borrowers alike perceived the EMP as an opportunity to promote the improvement of agriculture. The remoteness of the township and relative lack of formal sources of credit compounded this perception, leading to negotiations between various government actors and a well-organized village-level farmer cooperative comprising the most prosperous local farmers. Emerging from this negotiation was a new model of EMP provision that reflected the wider imperative to funnel resources to “successful” actors in a way that was, nevertheless, unique to Agricultural Township.

In Migrant Work Township and Diverse Economy Township, access to credit was less constrained due to a higher number of financial institutions and more socioeconomic linkages with urban areas. This meant that the EMP was perceived as being a burden rather than an opportunity. In Migrant Work Township, the local official charged with providing the EMP loans had to work hard to find friends and family who had eligible guarantors and were willing to complete the time-consuming loan process, while in Diverse Economy Township the township government ultimately decided to halt the program. In both these cases, as with Agricultural Township, the only people who were eligible to borrow were relatively well-off local actors. However, unlike Agricultural Township, because financial landscapes were segmented in such a way that the EMP was basically a redundant source of credit for eligible borrowers, the loans were used as extra investment capital, for on-lending at a profit, or the program was cancelled. Nevertheless, even though the EMP was considered to be superfluous and unnecessary by the township-level officials in both Migrant Work Township and Diverse Economy Township, because of the different relational dynamics at play implementation manifested itself heterogeneously.

The case studies also highlight how pressures and incentives played an important role in shaping implementer behavior with regard to the EMP. However, this occurred in ways that challenge the usual depictions of local officials predictably prioritizing policy goals that are easily measurable and used to determine career advancement (Chung, 2000; Donaldson, 2009; O’Brien and Li, 1999; Zhao, 2007). While quantifiable pressures and incentives undoubtedly played a major role in shaping implementation in the three townships, it is also important to note that seemingly similar pressures and incentives resulted in different outcomes—meaning they were not perceived in the same way by different people.²⁵ For instance, local implementers formulated different strategies to ensure that county-imposed quotas were met in a way that limited time expenditure and risk. In Agricultural Township this meant supplying the entire quota to one borrower (the farmer cooperative) with one guarantor (the township government). In Migrant Work Township, the local implementer instead opted to lend to trustworthy friends and family. Implementers also demonstrated an ability to change their understanding of pressures and incentives over time. In Diverse Economy Township the lack of a quota was initially perceived as a top-down pressure to provide the service to anyone who met the eligibility criteria, but this understanding was later revised by local officials and the program was halted altogether.

There were also widely different understandings among officials with regard to what the EMP meant for their careers. Since the EMP does not feature prominently in the performance targets associated with the “cadre responsibility system” 干部责任制, it is unsurprising that the implementers in Migrant Work Township and Diverse Economy Township perceived the program as a potential career speed bump. They thus sought to mitigate the risks associated with the program by focusing on achieving the most quantifiable element (i.e., repayment rates) or by cancelling it altogether. However, although the EMP did not have an obvious or directly measurable benefit to their careers, county and township officials in Agricultural Township perceived it as a means of creating their own “model experience” 典型经验, which they hoped would improve their career prospects. This contradicts established theory on how and why implementers in China decide to devote time and energy to certain policies rather than others.

These examples demonstrate that the heterogeneous implementation of the EMP in the three townships does not fit into neat depictions of implementers responding to pressures by either “shirking” or “working” (O’Brien and Li, 1999: 181–82). It also cannot be characterized as solely the result of corrupt machinations or collusion between local governments and elites to capture subsidized funds for personal gain (OECD, 2003; Ong, 2011; Park and Ren, 2001). This does not mean that illegal and corrupt activities did not occur. Indeed, in Agricultural Township there was identity theft and in Migrant Work Township obvious nepotism. It does illustrate, however, that corruption and illegality were not the result of villainous rational actors coldly calculating how to best extract resources. Rather than considering these implementers as “principled agents” or “wily defiers” (O’Brien and Li, 1999: 168), they are better understood as attempting to adapt the provision of the EMP loans based on numerous dynamic factors, including their understanding of local developmental realities.

These varied understandings of the role of microcredit in inducing development reflected the wider imperatives calling for the consolidation of resources in the hands of the “successful” segments of the population. At the same time, the final outcomes emerged from local-level negotiations between people with diverse backgrounds and different understandings of what constitutes developmental progress, and were, therefore, distinct to the local contexts. For this reason, these examples illustrate that conventional methods of understanding variation in local implementation—by identifying important variables and assessing their respective influence over reified external interventions in local situations—are misleading. Instead of being an external and linear process, the implementation of the EMP was a local production that simultaneously reflected, emerged from, and contributed to relational dynamics at different levels.

Complexity and Emergence: Toward a Relational Understanding of Heterogeneous Implementation

Ultimately, the commonality (or constant) across the three localities is that implementation outcomes were forged by negotiations between a variety of actors and were, therefore, relational in nature. Instrumentality was a key feature of these relationships, as actors approached the negotiations over program implementation with their own interests in mind. However, these interests were diverse, and not always based solely on the types of material gains that would be expected by the classic rational actor, or Homo economicus. Instead, they were informed by the wider development landscape and shaped by the personalities and backgrounds of the different people involved. Moreover, these actors were not unchanging self-contained units, but were transformed through their encounters (Tsing, 2015), both with the EMP and each other. Thus, while we understand the basic socioeconomic frameworks—and sets of variables—structuring implementation behavior, it is still nearly impossible to accurately predict what form implementation will take in different situations because the relational dynamics that define the implementation events are in a constant state of flux.

This relational unpredictability underpinning EMP provision confirms David Mosse’s statement that conventional understandings of policy implementation processes and development impact assessments are “nonsense” because they are “based on a very mechanical notion of pulleys and levers” (Mosse and Kruckenberg, 2017: 199). This is due to the fact that top-down linear conceptions of development and impact do not take into account “messy partnerships and relationships” at different levels (Chambers, 2008: 174). Indeed, as outlined earlier in this article, the implementation of rural development policy/intervention in China is generally conceived of as a linear process that is inherently external rather than co-produced; is composed of atomized and measurable self-contained components; and is detached from wider developmental contexts and histories. We contend that in order to gain an in-depth understanding of heterogeneous development policy implementation outcomes at the local level, it is necessary to recognize that the “separation of ‘policy’, ‘implementation’ and ‘outcomes’ is a gross oversimplification of a much more complicated set of processes which involve the reinterpretation or transformation of policy during the implementation process” (Long, 2001: 25). This will allow for the deconstruction of “the concept of planned intervention so that it is seen for what it is—namely, an on-going, socially constructed and negotiated process, not simply the execution of an already-specified plan of action with expected outcomes” (Long, 1999: 4).

This type of relational orientation fundamentally challenges the assumption of linear causality that is implicit in much research on the local-level implementation of development projects both in China and elsewhere. In mainstream linear and substantialist (i.e., consisting of concrete things) research, the intervention itself is reified, and variation in its implementation is understood through the identification of a limited number of causal variables or determinants that are depicted as causing distortions. In this way, policy and implementation are perceived as scalable universal models achievable through best practices that eliminate these distortions (Chambers and Loubere, 2017; Escobar, 2004; Hobbes, 2014; Ramalingam, 2013). It is the imagined “ability to make one’s research framework apply to greater scales, without changing the research questions, [that] has become a hallmark of modern knowledge” (Tsing, 2015: 38).

Alternatively, by shifting the focus to relationships and negotiations it is possible to conceive of development interventions as being co-produced and reconstituted at the local level. They are, thus, revealed to be the result of (and contributing to) complex, diverse, and dynamic social processes. Indeed, the previous sections outline a number of factors that could be considered variables or determinants, each playing a role (to varying degrees) in the heterogeneous implementation outcomes of the EMP. However, they were not static and independent, but context specific—the products of particular times, places, and people that were in a state of continual and rapid change. Therefore, rather than perceiving these factors as reified entities, relational approaches prompt us to conceptualize development interventions as “self-organizing systems on the edge of chaos” that are being forged in zones “of diverse, self-organizing and emergent complexity which lies between top-down rigidity and random chaos” (Chambers, 2008: 174; Ramalingam, 2013). Moreover, from a relational perspective there are “no self-contained units, its units are encounter-based collaborations. Without self-contained units, it is impossible to compute costs and benefits, or functionality, to any ‘one’ involved” (Tsing, 2015: 33–34). In this way, development programs and policies are not really things that travel down from the center to the local level, getting distorted along the way, but are instead necessarily locally produced, while also reflective of macro dynamics. This means that implementation outcomes are not simply the sum of their parts, but are emergent in nature and can only be understood through the socioeconomic processes from which they originate and feed back into. These processes are not linear, durable, reproducible, or scalable, but rather take different forms each time.

The relational nature of the EMP policy implementation is evident in all three townships, but is particularly obvious in Agricultural Township. There, three key individuals—the head of the farmer cooperative, and the heads of the county and township branches of the HRSS—were the primary actors engaged in negotiations over how the program would be implemented at the local level. These actors effectively served as linchpins in the emergent social process of implementation, in that without their involvement the implementation and associated outcomes would have surely transpired much differently. However, it is not just their individual agency, or “development spirit” (Ahlers and Schubert, 2013), that was responsible for the creation of the Agricultural Township model of EMP provision. Rather, it was the relational bonds formed through transformative encounters between actors and others—creating new guanxi networks and arenas of negotiation both locally and non-locally—which ultimately resulted in a form of implementation that was unique to the township, but that nevertheless still reflected the wider imperatives implicit in China’s contemporary political economy of development requiring the transfer of resources to local “winners.” Therefore, only by focusing on these relational dynamics and recognizing the role of self-organizing complexity in producing emergent outcomes—while also acknowledging the importance of the wider development landscape in shaping these processes—is it possible to gain an in-depth understanding of how and why microcredit programs (or development programs more generally) are implemented at the local level in China and elsewhere.

Conclusion

Our examination of the implementation of one of the Chinese government’s largest microcredit programs—the EMP—in three townships illustrates that it was internalized by diverse actors and reconstituted at the local level through processes of negotiation. This resulted in implementation outcomes that were complex, emergent, and particularistic to different contexts; but which, nevertheless, reflected the fundamental features of the wider development landscape. For this reason, heterogeneous implementation at the local level must be understood from a relational perspective that adequately takes into account how the local implementation of external interventions and policies emerges from interlocking sets of developmental relationships at different levels.

These findings are in contrast to the majority of research analyzing local variation in the implementation of policy in China and elsewhere, which depicts external interventions as reified entities travelling from the central to the local level, and being influenced—or distorted—along the way by Homo economicus-esque implementers seeking to adapt implementation for their own personal (usually material) gain. Rather than depicting EMP implementation as being linear, external, and reified, this article illustrates how the local manifestations of the microcredit program emerged from self-organizing processes, negotiations, encounters, and wider developmental imperatives. It is important to point out, however, that by shifting the focus to the interface of interaction and developmental relations, making generalizations about (or predicting) variation in implementation becomes a much more difficult—perhaps even impossible—task. Moreover, discussions of best practice aimed at optimizing or standardizing implementation become meaningless as uniformity gives way to the reality of multiple and multidimensional understandings of what development actually entails. As Norman Long points out, at the local level there are “‘multiple realities’, which imply potentially conflicting social and normative interests, and diverse and discontinuous configurations of knowledge” (Long, 2001: 19).

These multiple realities and the relationships that they emerge from should be placed at the center of analysis. For this reason, future research on heterogeneous intervention would do well to draw on relational critiques of development process that challenge linear and substantialist perspectives (Chambers, 2008; Long, 1999, 2001; Mosse, 2010; Ramalingam, 2013), as well as critical development studies approaches that illuminate the inherently unequal relations of power that underpin and define the long-run relations from which the formulation, implementation, and outcomes of development planning and intervention emerge (Escobar, 1995; Tania Li, 2007; Yeh, 2013). By taking relational dynamics as the point of departure, it will become possible to more closely examine how developmental relationships are mediated at different levels and gain a better understanding of their nature.

However, shifting to an approach based on truly relational understandings of our world is no easy task. It comes with distinctly different analytical rewards that may seem to be of less value when subjected to the same criteria as research aiming to typify, categorize, and predict. We should, therefore, be careful not to inadvertently slip back into linear and substantialist modes of thinking that prompt us to precisely pin down inherently fluid and dynamic relations. In the words of Anna Lowenhaupt Tsing, “theories of heterogeneity are still in their infancy. To appreciate the patchy unpredictability associated with our current condition, we need to reopen our imaginations” (Tsing, 2015: 4–5). Ultimately, we call for a reimaging of how and why efforts to develop local areas manifest themselves in heterogeneous ways, and a rethinking of what this variation can tell us about local socioeconomic realities and wider development processes in contemporary China.

Footnotes

Acknowledgements

The authors would like to thank Ka Lin, Pingjun Yin, and Lijuan Pan for help in planning and conducting fieldwork. A debt of gratitude is also owed to Alison Hardie and Graeme Smith for helpful comments on earlier drafts of this article. We are also indebted to Kathryn Bernhardt, John Donaldson, and two anonymous referees for their invaluable comments and suggestions. The usual disclaimers apply.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was generously supported through grants and fellowships from the University of Leeds, the Worldwide Universities Network, the Universities’ China Committee in London, the Chiang Ching-kuo Foundation for International Scholarly Exchange, and the Australian Centre on China in the World at the Australian National University.

Notes

Author Biographies

Nicholas Loubere is an associate senior lecturer in the Study of Modern China at the Centre for East and South-East Asian Studies, Lund University. His research examines socioeconomic transformation in contemporary China with a focus on microcredit, and the impact of migration on local development and livelihoods.

Qiu Shen (PhD, Peking University) is a researcher in the China Center for International Knowledge on Development (CIKD) at the Development Research Center of the State Council, China. His research focuses on rural development social policy and the implementation of targeted poverty reduction programs in rural areas.

References

Ahlers

Anna L.

Schubert

Gunter

(2013) “Strategic modelling: ‘building a new socialist countryside’ in three Chinese counties.” China Q. 216: 831–49.

Ahlers

Anna L.

Schubert

Gunter

(2015) “Effective policy implementation in China’s local state.” Modern China 41, 4 (July): 372–405.

Atkinson

Paul

Hammersley

Martyn

(2007) Ethnography: Principles in Practice. London: Routledge.

Augustin-Jean

Louis

Ruining

Xue

(2011) “The nature of cooperatives in China: the implementation and paradoxes of the law on cooperatives in Shanxi province.” Pp. 187–201 in Alpermann

Björn

(ed.), Politics and Markets in Rural China. London: Routledge.

Bateman

Milford

(2010) Why Doesn’t Microfinance Work? The Destructive Rise of Local Neoliberalism. London: Zed Books.

Bislev

Ane

(2012) “Embedded microcredit—creating village cohesion on the basis of existing social networks.” Pp. 189–204 in Bislev

Ane

Thøgersen

Stig

(eds.), Organizing Rural China, Rural China Organizing. Plymouth, UK: Lexington Books.

Bouman

Frits J. A.

Hospes

Otto

(1994) “Financial landscapes reconstructed.” Pp. 1–7 in Bouman

Frits J. A.

Hospes

Otto

(eds.), Financial Landscapes Reconstructed: The Fine Art of Mapping Development. London: Westview.

Brown

Philip H.

De Brauw

Alan

Yang

(2009) “Understanding variation in the design of China’s new co-operative medical system.” China Q. 198: 304–29.

Chambers

Robert

(2008) Revolutions in Development Inquiry. London: Earthscan.

10.

Chambers

Robert

Loubere

Nicholas

(2017) “Liberating development inquiry: freedom, openness and participation in fieldwork.” Pp. 27–46 in Crawford

Gordon

Kruckenberg

Lena J.

Loubere

Nicholas

Morgan

Rosemary

(eds.), Understanding Global Development Research: Fieldwork Issues, Experiences and Reflections. Thousand Oaks, CA: SAGE.

11.

Charmaz

Kathy

(2006) Constructing Grounded Theory: A Practical Guide through Qualitative Analysis. Thousand Oaks, CA: SAGE.

12.

Chen

Jiandong

Dai

Ming

Wenxuan

Hou

Qiaobin

Feng

(2010) “The trend of the gini coefficient of China.” BWPI Working Paper 109. http://ideas.repec.org/p/bwp/bwppap/10910.html.

13.

Chung

Jae Ho

(2000) Central Control and Local Discretion in China: Leadership and Implementation during Post-Mao Decollectivization. Oxford: Oxford Univ. Press.

14.

Donaldson

John A.

(2009) “Why do similar areas adopt different developmental strategies? A study of two puzzling Chinese provinces.” J. of Contemporary China 18, 60: 421–44.

15.

Xiaoshan

(2008) “The current situation and future prospects for microfinance in China.” Pp. 2–9 in Microfinance in China. Geneva: World Microfinance Forum Geneva.

16.

Escobar

Arturo

(1995) Encountering Development: The Making and Unmaking of the Third World. Princeton, NJ: Princeton Univ. Press.

17.

Escobar

Arturo

(2004) “Other worlds are (already) possible: self-organisation, complexity, and post-capitalist cultures.” Pp. 349–58 in Sen

Jai

Anand

Anita

Escobar

Arturo

Waterman

Peter

(eds.), World Social Forum: Challenging Empires. New Delhi: Viveka Foundation.

18.

Gao

Yuan

(2011) “Rural development in Chongqing: the ‘every peasant household’s income to grow by 10,000 yuan’ project.” Modern China 37, 6 (Nov.): 623–45.

19.

Goodman

David S. G.

(2004) “The campaign to ‘open up the west’: national, provincial-level and local perspectives.” China Q. 178: 317–34.

20.

GÖbel

Christian

(2011) “Uneven policy implementation in rural China.” China J. 65: 53–76.

21.

Guangwen

Xiaoshan

Chengyu

Bai

Zhanwu

(2009) China Microfinance Industry Assessment Report. China Association of Microfinance. http://files.chwsda.webnode.com/200000073-aed4cafceb/Industry%20Assessment%20Report.pdf.

22.

Heilmann

Sebastian

(2008) “From local experiments to national policy: the origins of China’s distinctive policy process.” China J. 59: 1–30.

23.

Heilmann

Sebastian

Melton

Oliver

(2013) “The reinvention of development planning in China.” Modern China 39, 6 (Nov.): 580–628.

24.

Hickel

Jason

(2015) “The microfinance delusion: who really wins?” Guardian, June 14. http://www.theguardian.com/global-development-professionals-network/2015/jun/10/the-microfinance-delusion-who-really-wins.

25.

Hobbes

Michael

(2014) “Stop trying to save the world.” New Republic, Nov. 17. http://www.newrepublic.com/article/120178/problem-international-development-and-plan-fix-it.

26.

Hofmockel

Evan

(2005) “Microcredit in the rural areas of northwest Yunnan province, China: investigating the operation of, results of, and obstacles to microfinancial institutions.” Independent Study Project (ISP) Collection 419. http://digitalcollections.sit.edu/isp_collection/419/.

27.

Hsu

Becky Yang

(2014) “Alleviating poverty or reinforcing inequality? Interpreting micro-finance in practice, with illustrations from rural China.” British J. of Sociology 65, 2: 245–65.

28.

Jiangxi Provincial Department of HRSS (2009) “Xiao’e danbao daikuan zhengce wenjian huibian” (Compilation of employment microcredit policy documents).

29.

Jiangxi Provincial Department of HRSS (2012) “Guanyu wanshan xiao’e danbao daikuan zhengce jinyibu tuidong chuangye cujin jiuye de tongru” (Improving the employment microcredit policy through the promotion of employment through entrepreneurship), no. 48.

30.

Jiangxi Provincial People’s Government (2009) “Guanyu jinyibu zuohao cujin jiuye gongzuo de tongzhi” (Circular on further improving reemployment work), no. 5.

31.

Jiangxi Provincial Rural Credit Cooperative Union (2002) “Guanyu kaizhan zai-jiuye xiao’e daikuan danbao fuchi xiagang shiye renyuan zaijiuye de tongzhi” (Notification on the development of employment microcredit to support the reemployment of laid-off workers), no. 37.

32.

JXP [Jiangxi Provincial Department of HRSS and Jiangxi Province Rural Credit Cooperative Union] (2009) “Guanyu tuijin xiao’e danbao daikuan fuchi fanxiang nongmingong zizhu chuangye de shishi yijian” (Some suggestions on promoting employment microcredit to support migrant workers returning home to start their own businesses), no. 4.

33.

Lai

Hongyi Harry

(2002) “China’s western development program: its rationale, implementation, and prospects.” Modern China 28, 4 (Oct.): 432–66.

34.

Lampton

David M.

(1987) Policy Implementation in Post-Mao China. Berkeley: Univ. of California Press.

35.

Linda Chelan

(2007) “Working for the peasants? Strategic interactions and unintended consequences in the Chinese rural tax reform.” China J. 57: 89–106.

36.

Tania Murray

(2007) The Will to Improve: Governmentality, Development, and the Practice of Politics. Durham, NC: Duke Univ. Press.

37.

Lipsky

Michael

(1980) Street-Level Bureaucracy: Dilemmas of the Individual in Public Services. New York: Russell Sage Foundation.

38.

Liu

Mingxing

Juan

Wang

Ran

Tao

Murphy

Rachel

(2009) “The political economy of earmarked transfers in a state-designated poor county in western China: central policies and local responses.” China Q. 200: 973–94.

39.

Long

Norman

(1999) “The multiple optic of interface analysis.” UNESCO Background Paper on Interface (Oct.): 1–25.

40.

Long

Norman

(2001) Development Sociology: Actor Perspectives. London: Routledge.

41.

Loubere

Nicholas

(2016a) “Indebted to development: microcredit as (de)marginalisation in rural China.” J. of Peasant Studies. doi:10.1080/03066150.2016.1236025

42.

Loubere

Nicholas

(2016b) “Laying off responsibility: microcredit, entrepreneurship, and China’s industrial retrenchment.” Made in China: A Quarterly on Chinese Labour, Civil Society, and Rights 1, 2: 20–23.

43.

Loubere

Nicholas

(2017) “Questioning transcription: the case for the systematic and reflexive interviewing and reporting (SRIR) method.” Forum Qualitative Sozialforschung / Forum: Qualitative Social Research 18, 2. http://www.qualitative-research.net/index.php/fqs/article/view/2739.

44.

Manion

Melanie

(1991) “Policy implementation in the People’s Republic of China: authoritative decisions versus individual interests.” J. of Asian Studies 50, 2: 253–79.

45.

Mertha

Andrew

(2005) “China’s ‘soft’ centralization: shifting tiao/kuai authority relations.” China Q. 184: 791–810.

46.

Mertha

Andrew

(2009) “‘Fragmented authoritarianism 2.0’: political pluralization in the Chinese policy process.” China Q. 200: 995–1012.

47.

Ministry of Finance (2008) “Xiao’e danbao daikuan caizheng tiexi zijin guanli banfa” (Management approach for the employment microcredit financial subsidy fund), no. 100.

48.

Mood

Michelle S.

(2005) “Opportunists, predators and rogues: the role of local state relations in shaping Chinese rural development.” J. of Agrarian Change 5, 2: 217–50.

49.

Mosse

David

(2010) “A relational approach to durable poverty, inequality and power.” J. of Development Studies 46, 7: 1156–78.

50.

Mosse

David

Kruckenberg

Lena J.

(2017) “Beyond the ivory tower: fieldwork in development practice.” Pp. 193–212 in Crawford

Gordon

Kruckenberg

Lena J.

Loubere

Nicholas

Morgan

Rosemary

(eds.), Understanding Global Development Research: Fieldwork Issues, Experiences and Reflections. Thousand Oaks, CA: SAGE.

51.

O’brien

Kevin J.

Linjiang

(1999) “Selective policy implementation in rural China.” Comparative Politics 31, 2: 167–86.

52.

OECD (2003) Rural Finance and Credit Infrastructure in China. Paris: OECD Publishing.

53.

Ong

Lynette

(2011) “Greasing the wheels of development: rural credit in China.” Pp. 48–68 in Alpermann

Björn

(ed.), Politics and Markets in Rural China. London: Routledge.

54.

Park

Albert

Changqing

Ren

(2001) “Microfinance with Chinese characteristics.” World Development 29, 1: 39–62.

55.

PBC [People’s Bank of China, Ministry of Finance, State Economic and Trade Commission, and Ministry of Labor and Social Security] (2002) “Xiagang shiye renyuan xiao’e danbao daikuan guanli banfa” (Employment microcredit management approach), no. 394.

56.

Putterman

Louis

Xiao-Yuan

Dong

(2000) “China’s state-owned enterprises: their role, job creation, and efficiency in long-term perspective.” Modern China 26, 4 (Oct.): 403–47.

57.

Ramalingam

Ben

(2013) Aid on the Edge of Chaos: Rethinking International Cooperation in a Complex World. Oxford: Oxford Univ. Press.

58.

Robinson

Marguerite S.

(2008) “Supply and demand in microfinance: the case for a financial systems approach.” Pp. 45–64 in Hulme

David

Arun

Thankom

(eds.), Microfinance: A Reader. London: Routledge.

59.

Rutherford

Stuart

(2004) “The microfinance market: huge, diverse—and waiting for you.” Pp. 263–90 in Lont

Hotze

Hospes

Otto

(eds.), Livelihood and Microfinance: Anthropological and Sociological Perspectives on Savings and Debt. Delft: Eburon.

60.

Rutherford

Stuart

(2006) Grameen II: The First Five Years 2001–2005. Dhaka: MicroSave. http://www.cgdev.org/doc/blog/Rutherford,%20Grameen%20II–The%20First%20Five%20Years,%202001-2005.pdf.

61.

Smith

Graeme

(2010) “The hollow state: rural governance in China.” China Q. 203: 601–18.

62.

Smith

Graeme

(2013) “Measurement, promotions and patterns of behavior in Chinese local government.” J. of Peasant Studies 40, 6: 1027–50.

63.

State Council (2002) “Guanyu jinyibu zuohao xiagang shiye renyuan zaijiuye gongzuo de tongzhi” (Notice on further improving laid-off workers reemployment work), no. 12.

64.

State Council (2004) “Guanyu jinyibu jiaqiang nongcun gongzuo tigao nongye zonghe shengchan nengli ruogan zhengce de yijian” (Some suggestions on further strengthening rural work to improve agricultural production capacity), no. 1.

65.

State Council (2006) “Quanmian zhengque lijie shehuizhuyi xinnongcun jianshe” (Fully and correctly understand the building of a new socialist countryside).

66.

State Council (2008) “Guanyu zuohao cujin jiuye gongzuo de tongzhi” (Circular on the promotion of reemployment work), no. 5.

67.

State Council (2014) “Guanyu quanmian shenhua nongcun gaige jiakuai tuijin nong-ye xiandaihua de ruogan yijian” (Some suggestions on deepening rural reform through the modernization of agriculture), no. 1.

68.

Tan

Qiucheng

(2008) “Nongcun zhengce weishenme zai zhixing zhong rongyi zouyang” (Why rural policies are easily changed in implementation). Zhongguo nongcun guancha 4: 2–17.

69.

Teets

Jessica C.

Hurst

William

(2014) Local Governance Innovation in China: Experimentation, Diffusion, and Defiance. London: Routledge.

70.

ThØgersen

Stig

(2011) “Building a new socialist countryside: model villages in Hubei.” Pp. 172–186 in Alpermann

Björn

(ed.), Politics and Markets in Rural China. London: Routledge.

71.

Trappel

RenÉ

(2016) “New villages, old problems? Exploring policy implementation in a rapidly changing Chinese countryside.” J. of Current Chinese Affairs 45, 1: 5–10.

72.

Tsai

Kellee S.

(2004) “Imperfect substitutes: the local political economy of informal finance and microfinance in rural China and India.” World Development 32, 9: 1487–1507.

73.

Tsing

Anna Lowenhaupt

(2015) The Mushroom at the End of the World: On the Possibility of Life in Capitalist Ruins. Princeton, NJ: Princeton Univ. Press.

74.

Weber

Heloise

(2006) “The global political economy of microfinance and poverty reduction: locating local ‘livelihoods’ in political analysis.” Pp. 37–54 in Fernando

Jude L.

(ed.), Microfinance: Perils and Prospects. London: Routledge.

75.

White

Howard

(2010) “A contribution to current debates in impact.” Evaluation 16, 2: 153–64.

76.

World Bank (2014) Global Financial Development Report 2014: Financial Inclusion. Washington, DC: World Bank.

77.

Bin

ZHANG LIYAN (2013) “Farmer innovation diffusion via network building: a case of winter greenhouse diffusion in China.” Agriculture and Human Values 30, 4: 641–51.

78.

Yeh

Emily T.

(2013) Taming Tibet: Landscape Transformation and the Gift of Chinese Development. Ithaca, NY: Cornell Univ. Press.

79.

Yeh

Emily T.

O’brien

Kevin J.

Jingzhong

(2013) “Rural politics in contemporary China.” J. of Peasant Studies 40, 6: 915–28.

80.

Zang

Xiaowei

(2016) “Research on street-level discretion in the West: past, present, and the future.” Chinese Political Science Rev. doi:10.1007/s41111-016-0041-z.

81.

Zhang

Heather Xiaoquan

Loubere

Nicholas

(2015) “Rural finance and development in China: the state of the art and ways forward.” Pp. 151–74 in Zhang

Heather Xiaoquan

(ed.), Rural Livelihoods in China: Political Economy in Transition. London: Routledge.

82.

Zhao

Shukai

(2007) “Rural governance in the midst of underfunding, deception, and mistrust.” Chinese Sociology and Anthropology 39: 8–93.

83.

Zhou

Takeuchi

Hiroki

(2010) “Informal lenders and rural finance in China: a report from the field.” Modern China 36, 3 (May): 302–28.

84.

Zhou

Xueguang

(2010) “The institutional logic of collusion among local governments in China.” Modern China 36, 1 (Jan.): 47–78.