Business Friendliness: A Double-Edged Sword

Abstract

In this paper, we test the hypothesis that business-friendly local-government policies combined with weak legal institutions lead to lower economic welfare in the form of greater fraud activity. Using data of almost 3000 failed peer-to-peer (P2P) lending platforms in China, labeled as “runaways”, we find that they are more prevalent in provinces with business-friendly policies with weak law-enforcement regimes.

1. Introduction

In this paper, we addressed the problem of financial technology (FinTech) innovation and economic sustainability in emerging economies. Following Deng at al., we looked at the rise of the FinTech innovation of online peer-to-peer (P2P) lending platforms in China and the association with institutional conditions [1]. They reported a U-shaped relationship between the growth of P2P lending in a region and its sustainable development, defined as the sum of social, ecological, and economic welfare gained through economic activity without reducing the ability of future generations to meet their own needs. We delved deeper into this question by focusing on the constellation of institutional conditions that likely contribute to welfare destruction, which reduces the sustainability of a region’s developmental trajectory. The institutional conditions that we focus on comprise the policies supporting business friendliness and the associated legal-enforcement regime in a region.

In development economics, business friendliness refers to the ability of a region to attract and retain external investments of capital, which is a key factor in sustainable economic growth. It is usually defined as the ease of conducting business in a region, which is measured by the degree of economic freedom, the robustness of the legal and regulatory frameworks, and government support for business creation and growth. To promote business-friendly environments, governments use a variety of monetary and fiscal incentives such as tax holidays for new capital investments, research and development, employment, preferential treatment in business licensing, relocation assistance, and even subsidies for setup costs, simplified procedures for business registration and faster processing time, or reduced startup-capital requirements. Annually, the World Bank assesses business friendliness by collecting data on such factors as the procedures, time, costs, and minimum capital needed to start a business (http://www.doingbusiness.org/content/dam/doingBusiness/media/Annual-Reports/English/DB19-Chapters/DB19-Score-and-DBRankings.pdf). Countries/regions ranked higher are perceived to have more business-friendly regulatory regimes. Many have argued that regulation is a cost of doing business, and therefore the simpler and the fewer, the better [2,3,4]. There is a literature suggesting that business friendliness can have unintended consequences that negatively affect sustainable economic growth [5]. The purpose of this research is to explore the potential downsides of business-friendly policies in a fast-developing nation such as China and to understand the circumstances under which the adverse effects can happen.

Unintended consequences of business-friendly policies can occur for a variety of reasons: the misalignment between regulations and their enforcement can create undesirable incentives; the contradictions between different aspects of a policy package, for example rewarding short-term orientation to achieve long-term objectives can result in decision horizons that are mismatched with the time needed to realize the intended outcomes; and the circumstances (e.g., technological, cultural, and economic) under which a policy was first formulated may have changed, rendering it inappropriate for the objectives it was originally designed to achieve. For example, in countries with developing legal systems and weak law enforcement, business incentives aimed at economic development may morph into havens for money laundering. Unlike the laws they are supposed to represent, regulations (i.e., the rules and procedures) are less transparent and predictable, and thus may invite rent-seeking activities.

To examine the possible negative outcomes stemming from the ease of doing business in economies with weak legal institutions, we study the peer-to-peer lending (P2P thereafter) market in China. P2P lending is a form of informal consumer finance through which household lenders with spare cash looking for attractive returns meet household borrowers who need short-term cash for consumption, petty entrepreneurship, or private investment. While P2P lending has a long history around the world, the most notable of which are moneylenders or chettiars in India and the formal microfinance institutions (mFi) that followed [6]. The advent of the internet and the ease with which lending platforms can establish themselves have led to the triple-digit growth of informal financing, especially in emerging economies where formal credit markets are tight as there are no credit bureaus to standardize risk analysis. In 2016, the annual cumulative transaction volume of P2P-lending platforms in the world exceeded $400 billion with $72 billion in the US, UK, and EU, and $300 billion in China (https://www.bloomberg.com/quicktake/peer-peer-lending; http://m.sohu.com/n/483431935/). The introduction of internet-mediated lending, where there are few if any minimal loan amounts, no credit checks, and flexible repayment terms, has turbo charged the growth of P2P lending, often exceeding the speed at which financial regulators have been able to respond.

To test our hypotheses, we collect Chinese P2P data between 2007 and 2016 by scraping the web for P2P company information, news items, announcements, and other public web pages. We find that out of 2941 runaways, defined as P2P platforms that absconded, only 6% were incorporated in the least business-friendly provinces with the remaining 94% registered in more friendly locations. The regression results confirm that a region’s business friendliness in the year of a P2P’s registration is significantly and positively related to the probability of a runaway in the future, after controlling for a host of factors that might lead to corporate failure. Consistent with our hypothesis on the mitigation effect of legal institutions, we also find that the strength of law enforcement in a region reduces the positive relationship between business friendliness and the number of runaways.

Our study contributes to the studies on the development of institutions and sustainable economic growth in the following ways: While the literature suggests the importance of the institutional quality for healthy financial markets and economic growth, China’s economic growth has been an exception because it lacked strong legal institutions [7,8,9,10]. Some studies offer contradicting results on the importance of institutions in China and question the relevance and legitimacy of the existing theories on institutions and economic activities [11]. Our study sheds light on these discussions by examining how business friendliness, a proxy for government intervention, is related to risky investment decisions. We show that regulations promulgated by local governments in the hopes of promoting economic growth are only effective and efficient in those regions with strong formal institutions. Our study confirms the importance of legal institutions for a region’s sustainable economic development and demonstrates the importance of looking at disaggregated data.

In the remainder of this paper, we do the following: Section 2 provides background on the P2P phenomenon for those who may be less familiar with it; Section 3 discusses the literature, theory and hypotheses development; Section 4 describes methods; and we report the results in Section 5 and conclude by discussing the findings in Section 6.

2. Background

A recent OECD report says that China is in the midst of shifting from an export-oriented (exogenous) economic growth model to a domestic consumption driven (endogenous) model of growth [12]. The same report recognizes China’s increasing focus on sustainable development through its industrial policy of targeting investments in technology, for example, FinTech, electric vehicles, and renewable energy. To drive consumption, Chinese policymakers have endeavored to incentivize the growth of domestic credit markets, bank, and nonbank, formal and informal. In the area of nonbank informal credit markets, it has allowed the entry of FinTech startups companies to close the credit gap experienced by microborrowers and those in regions not well served by the big banks.

In spite of its support for technology innovation and startups and being the second fastest growing economy in the world (www.worldbank.org.en.country/china/overview), China is not known for its “business friendliness”. In fact, according to the World Bank’s 2018 report (http://www.doingbusiness.org/en/data/doing-business-score), China is ranked 46 among 190 economies on the ease of doing business index. Although the national ranking of business friendliness is not high, there is great variation across the provinces and cities in China. For example, a survey of the investment climate in China conducted by the World Bank in 2003 shows that the average score for government helpfulness is 0.35 (1.0 is perfect) with a standard deviation of 0.31, based on 731 firms from 18 cities responding to the survey [2]. Forbes 2018 annual list of China’s best 100 cities for doing business reports that Shanghai is first with a score of 0.723, which is comparable to such advanced economies as Belgium. The score for Hangzhou, ranked 10th in 2018 drops to 0.488, while that of Yichang, ranked 100th, is as low as 0.270. As well, the list captures less than 15% of the total number of cities in China, and so we believe there is a need for a framework which researchers can use to better understand why and when business-friendly policies may result in adverse outcomes.

Adverse outcomes from business-friendly policies may be amplified when local governments compete for investments and resources. A critical aspect of China’s economic reforms has been the process of gradual economic decentralization, which gave local governments fiscal authority over economic policies in their regions. Studies have cited examples of how local governments have chosen to promote economic growth by competing with regional rivals, especially since economic growth is a measure of state officials’ performance [13]. Some adverse outcomes have included local governments promulgating aggressive, and sometimes even illegal, practices such as leasing state-owned land and property at lower-than-market prices to related parties and providing extraordinary tax credits and financial subsidies to companies (http://cppcc.people.com.cn/n/2014/1225/c34948-26274811.html).

In countries where economic activity is largely market driven (i.e., prices are set by the forces of supply and demand for goods and services), lighter regulatory regimes can result in faster capital accumulation and economic development. In China, state-controlled enterprises (in which the controlling shareholder is the central government, its agencies, or other state-owned enterprises) account for approximately 70% of the firms that went public in China in 2006 [14], and the “big four” state-owned banks attract approximately 70% of total deposits. Evidence also shows that SOEs receive privileges compared to their nonstate counterparts in various aspects, such as in the initial public offering market [15], the product market [16], bank financing [17], and the M&A market [18]. Hence, where major industrial policy is still centralized, business-friendly policies by local governments to attract investment capital through “easier and faster establishment” regulation may lead to more harm than good. For example, to compete in the fast-growing internet economy, the local government supported Li Ke Gou, a Shandong online shopping platform established in July 2016, with free office space and official endorsements. By August of 2016, hundreds of investors were at the door of the company’s headquarters, demanding their money back. They were promised 8 times ROI on the $35 million investment the founders took for themselves (http://m.sohu.com/a/285442811_232843). A report by the National Audit Office of the PRC summarized the destructive consequences of such intragovernmental competition, such as inflated fiscal revenue, increased local government debt, and public and private rent-seeking (http://www.audit.gov.cn/n6/n41/c21449/content.html). We notice the same phenomenon in the P2P-lending market in China.

Since the first P2P platform was registered in 2006 in China, total lending ballooned to $300 billion by 2016 with the number of P2P platforms peaking at 3579 in 2015 (https://www.01caijing.com/article/12992.htm). Until 2016, P2P platforms in China were not regulated as other deposit-taking financial businesses such as banks. For example, the required registration procedures for P2P were the same as nonbank businesses, which were light. The consequence was a wave of P2P-platform abscondments or ‘runaways’ in which investors lost their money. P2P runaways run the gamut from those that simply stopped taking deposits and closed, after been in business for a while, to those in which the founders simply absconded with depositors’ money without having lent a dollar. According to Wind’s database and the data we collected, between 2007 and 2016 a total of 3410 platforms out of 6080, worth $16.78 billion, were shut down for various reasons. In addition, 4.66 million investors lost their money (exchange rate of 6.9 RMB to 1 USD). Since P2P are online platforms with lenders and borrowers from around the country, local authorities were limited in what they could do once abscondments occurred. Records show that P2P platforms were registered in all but two provinces (i.e., 32 provinces) during the sample period. This fact allows us to test the hypothesis that provinces with more business “friendly” policies are more likely to be associated with more runaway P2Ps.

The extent to which the law is enforced is a measure of the quality of an economy’s legal institutions, which have been shown to positively affect the development of its financial sector and growth [7,8,9]. In fact, regional commercial law enforcement varies greatly in China [2]. Their survey of managers from 18 cities shows that, on average, only 62%, with a deviation of 39%, believe that the legal system will protect their contracts and property rights in a business dispute. Thus, we conjecture that while local business-friendly policies work to attract P2P of all types, the strength of local law enforcement works in the opposite direction, to slow down and discourage those with nefarious intentions.

3. Literature Review and Theory

In this section, we briefly review the literature on business friendliness, peer-to-peer lending, and institutional theory. Institutional theory is the framework underpinning the theoretical model we propose to test in this paper. We then state the hypotheses and discuss them individually in detail.

3.1. Business Friendliness

Research in the role of regulatory policies on business performance and outcome has approached business friendliness from several related perspectives. For example, some describe the business environment as a government’s posture toward private enterprise. Generally, the business environment refers to a broad concept involving the positive and negative factors driving business establishment and growth [19,20]. We define business-friendly governments as those helpful to business startup and growth, whether it is in the sense of light government intervention or heavy government support for private enterprise.

The concept of ease of doing business describes the quantity of resources an enterprise has to expend in dealing with the government bureaucracy [2]. It refers to the lack of negative factors that hinder business establishment and growth so that scarce resources can be deployed to productive use [21]. It includes lower degrees of government intervention in business registration processes such as faster processing, less documentation, or less regulatory requirements [2,22]. Studies on business climate also focus on the negative aspects of the business environment such as bribery, corruption, and larceny by public officials [3,22,23,24], the corporate tax rate [25], and access to financial resources [26].

The above concepts are related because the light touch of government can also translate into less opportunity by public officials for rent-seeking behaviors [3]. Business friendliness also refers to positive factors in the business environment [27]. These factors include supportive infrastructure, availability of appropriately skilled labor, identifiable and accessible consumer markets, regional innovation capacity, and so on [28]. Business-friendly policies can manifest in affirmative ways, such as the active enforcement of property rights, and the provision of tax incentives, infrastructure, and financial support for business startups [25,26,27,28].

Many national and local governments have made business friendliness an important factor attracting capital and promoting growth [19,29]. For example, business friendliness is crucial in attracting foreign direct investment or FDI [16,30,31] and motivating outward foreign direct investment or OFDI [32,33]. It positively affects total factor productivity [22,27], enterprise performance [19,23] and exporting [34].

While most studies focus on the positive aspects of friendly business environments, scholars have recently begun to focus on the negative aspects of business friendliness. They find that when business friendliness abets public corruption and private rent-seeking, enterprise risk decisions [21] and innovation [20] are negatively affected. Although nascent, this research stream informs our formulation of the research question.

3.2. Peer-to-Peer Lending

The problem of access to capital for entrepreneurial activity has been an enduring problem in developing countries with thin financial markets. Policy solutions have ranged from the use of outright grants for start-up capital to uncollateralized loans by such institutions as the World Bank and Asian Development Bank. Yet, sustainable growth and economic success have been elusive. In one notable experiment, Nobel Prize-winning economist Muhammad Yunus, proposed the use of microloans as a way to break the consumption–investment conundrum (do you eat the seed corn to stave off hunger or plant it in hopes for a future harvest) faced by the world’s poor. The Grameen Bank, which he founded in the 1970s, relies on social networks to support a mutual insurance loan model that turned out to be transformative in the locations where it was practiced [35]. Similar types of lending systems have existed in Egypt, Brazil, Japan, and elsewhere in the world [36]. This model was introduced to China in the 1990s, although mutual private lending has been practiced in China as early as 300 AD [37]. The system of mutually insured microcredit protects lenders and borrowers. The former is protected because they have surety in the form of a group hedge, and the latter is protected because they have an alternative to moneylenders, with the associated costs and risks.

The first P2P-lending platform, Zopa, was launched in Europe in 2005 [38]. In the US, Prosper.com was the first in 2006. However, it was during the last global financial-market shock in 2007, in which formal financial institutions withdrew from the consumer microcredit market, and advances in internet payment technologies, that P2P-lending platforms experienced their greatest growth, especially in China. China’s first P2P-lending platform, Paipaidai, was founded in 2007 [39]. The rapid adoption of internet-based payment transactions and online credit scoring (https://globalfindex.worldbank.org/) allowed these platforms to proliferate at a rate that took authorities by surprise [37]. From the inception of the first platform, P2P firms grew by an average of 1000 platforms and 100 billion RMB transactions a month, making it the world’s leader in peer-to-peer lending. The P2P-lending market continued to grow around the world, perhaps because of the Great Recession of 2007 [37,40].

The P2P market varies around the world because of differences in economic environment, regulatory institutions, and policy orientation toward debt. In the US, the Securities and Exchange Commission (SEC) regulates P2P lending the same way they regulate nonbank financial institutions. These credit-giving institutions (PayPal, LoanMe, Freedom Financial Network, and so on) occupy the space that would have been served by P2P lenders in less financially developed markets (https://www.bis.org/publ/qtrpdf/r_qt1809e.htm). As the result of tight regulatory oversight and the lack of market demand, the P2P-lending market remains small and highly concentrated in the U.S., with Prosper and Lending Club controlling more than half the market share.

The market in China is diametrically opposite. The introduction and growth of the P2P market was unregulated in the beginning. Companies were simply registered as any other business entity, with no requirement to maintain a reserve, as with a bank. In 2013 the Chinese government adopted financial inclusion as part of its national economic development strategy. Within this framework, informal credit providers such as P2P platforms were viewed a channel to increase the coverage, availability, and satisfaction of credit for the general population (http://www.cbrc.gov.cn/chinese/files/2018/3D1972BB37BC43C2915E51BDEC4A47ED.pdf). It was not until the high-profile collapse of several large P2P platforms that the financial regulatory agency in China responded with consumer-protection measures.

The unorganized regulatory environment led to the rapid growth of P2P lending in China. In the early days, P2P companies assumed the role of credit intermediary, rather than information intermediary, the difference being the emphasis on the product in the former rather than the exchange in the latter. As credit intermediary, P2P companies had to assure the creditworthiness of borrowers on its platform, going so far as to guarantee borrowers’ principal and interest to the lenders [41]. Under this model, the platform obtains deposits from a large number of lenders by issuing notes. Different issues may carry different terms. The deposits are then packaged and loaned out to borrowers who apply through the platform. Hence, a borrower may get a loan from a larger number of lenders, and vice versa. The borrowers’ and lenders’ privacies are protected in this model since they do not know the identity of the counterparty. However, the default risk of these unsecured loans rises because of the anonymity. This is in contrast to a true P2P-lending model, in which the platform simply matches individual lenders with individual borrowers, taking a fee from the transaction. In such a model, lenders know the identity of their borrowers, and vice versa, which reduces moral hazard.

The rapid rise of the P2P platforms has attracted the interests of researchers [42,43,44,45,46,47,48,49]. The research in P2P markets have so far focused on the mechanisms to reduce default risk, appropriate interest rates, and other factors that drive success. For example, the growth rate, repayment rates, and size of the loan-loss provision affect P2P-lending platforms’ runaway behavior [50]. The regulation of information disclosure affects platforms’ runaway by reducing their credit guarantee [41]. Some use games to model the effect of regulatory schemes (e.g., improving the market entry and exit mechanism) on platform behavior [51]. In addition, researchers have examined borrowers’ personalities and objective financial information [52,53,54,55], as well as age, gender, marital status, education level, working years, income, company size, credit status, trust and trustworthiness [56,57], social networks [58,59], loan group characteristic [60,61], and loan terms in the textual information [62,63,64]. In these studies, the factor that seemed most important is the implied information asymmetry between borrowers and investors, with the latter not knowing the motivations of the former or the quality of the platforms into which they were depositing their monies.

3.3. Hypotheses

In China, business friendliness is simply a measure of the degree to which local governments competed for inbound investment capital and business establishment into their regions. As such, a more business-friendly province is said to have smaller government [2], fewer administrative barriers [21,22,23,24], broader access to financial capital [26], more developed factor market, and better infrastructure [65]. Factor markets and infrastructure development tended to vary more across the provinces than institutional development since local governments, with few exceptions, took their cue from the central government [65]. Provinces that were more business friendly were found to report higher levels of business-innovation activity and attracted concomitant human resources [28], faster-growing industries [29], and better enterprise performance [19,23].

Nevertheless, if business friendliness is associated with fewer government interventions, documentation for business startups, and regulatory barriers such as reserve capital requirements or business insurance, more liberal public resources in the form of tax incentives, infrastructure, and financial support, ex ante the variability in quality of business startups, will be higher in those provinces. More specifically, the lower the cost of business startups, the lower the cost of moral hazard to the bad actors, who can choose where to start their businesses. In the case of P2P platforms whose owners are investing other people’s money, the potential for abscondment is particularly high, since cash, unlike factory machines, is entirely liquid and fungible (i.e., redirectable at no cost). Hence,

Hypothesis 1a.

Provinces with more business-friendly policies will have more runaway P2P platforms.

According to the new institutional economics, institutions are both formal and informal [66]. Emerging economies often face many institutional voids, defined as the underdevelopment or absence of formal institutions that make the functioning of an organization efficient [67,68]. In the presence of institutional voids, informal institutions such as culture, norms, values, and most importantly, ideology, in a region play the role of shaping firm behaviors [69,70,71,72]. Ideology is the basis of an individual’s interpretation of the world around them [69]. When thinking about Chinese economic ideology, we should consider China’s economic structure at a particular point in time. Chinese ideology undergoes a tug-of-war between planned and free markets, with one or the other waning or waxing at different time periods. These competing ideologies alternately give rise to the general enthusiasm for private enterprise (as measured by the share of GDP by the private sector) or state-controlled means of production (state-owned enterprises). To the extent that the share of privately generated GDP increases, as a ratio of total GDP, the ideological stance of a region leans toward free markets, whereas the opposite is true when state intervention in economic activity is heavy handed. In the same vein, when private-sector GDP is high, trust in economic exchange is high because informal institutions are strong. The result of high trust is reduced moral-hazard risk.

Hypothesis 1b.

The positive relationship between business friendliness and P2P runaways is attenuated in provinces with more developed informal institutions.

Notwithstanding the importance of informal institutions, economies that are shown to develop sustainably also have strong formal institutions such as private-property rights and the robust enforcement of such rights. This is particularly true for large economies that become increasingly heterogeneous over time. Heterogeneous societies are less likely to develop strongly homogeneous ideologies because culture and creed is weaker by comparison to small homogeneous societies. In these situations, formal institutions (laws, rules, and regulations) become more important to the socioeconomic development of a region [69,73]. Research has suggested that formal institutions can effectively reduce the welfare-destroying effects of market failure. While countries such as China, India, and Russia have existing legal frameworks for governing commercial transactions, their enforcement has not been uniformly applied [74]. This is often the case when public officials are involved in private economic transactions (setting aside the legal construct of imminent domain in which the state is a counterparty in the transaction). Researchers have suggested that when business friendliness is associated with corruption, rent seeking, and crime, it is often the case that legal enforcement is weak [3,20,21]. Strong formal institutions are necessary for countering bad actors when informal institutions cannot function as the brakes to moral hazard.

However, legal enforcement is costly. Not only is the litigation itself costly but maintaining the agencies that enforce the law is costly. This cost is borne by the public, so that in regions that do not have sufficient public resources, enforcement tends to be weaker. In those instances, other institutions, such as the domain political party in the region, may fill the void. However, such measures are inefficient since political institutions are not designed to enforce regulations. As such, we can approximate a region’s commitment to legal enforcement by the relative share of public resources expended on those activities, which includes public security, prosecutor’s office, and the courts.

Hypothesis 2a.

The positive relationship between business friendliness and P2P runaways is attenuated in provinces with more effective law enforcement.

Finally, the new institutional theory argues that formal and informal institutions act in concert to shape firm behavior. One view of institutions is that they function as substitutes. A more prevalent view is that they are complementary [69,70]. When formal institutions are ineffective, they do not necessarily weaken the informal institutions. That is, actors’ culture and ideology are not affected by a poor understanding of the law. On the other hand, when informal and formal institutions are aligned, for example when free market ideology is supported by strong enforcement of private property, firms are more likely to act within the ‘iron cage’ of norms and avoid rent seeking or moral hazard. Hence,

Hypothesis 2b.

The positive relationship between business friendliness and P2P runaways is attenuated to a greater extent in provinces with both strong formal and informal institutions than when either is weak.

4. Methods

4.1. Data

We obtained information on 3244 runaway P2P-lending platforms registered in mainland China and Hong Kong between 2007 and 2016. The data was hand-collected by Think Tank Information Technology (Beijing) Co., Ltd. (Beijing, China). In the data, a P2P-lending platform is labeled as a runaway if it met at least one of the following conditions: stop business, website shutdown, change business, restructuring, withdraw difficulty, loss of contact, bankrupt, fraud, police recorded, vicious runaway, and court-confirmed abscondment (

Table 1. Runaway label and descriptions.

Label	Describe	Count
Stop Business	The platform has stopped business for more than three months, no news or announcements to the public about the continuance of the platform.	942
Website Shutdown	Unable to open the website, no news or announcements to the public about the continuance of the platform.	520
Change Business	The platform stops all the P2P business and transforms into other businesses.	3
Restructuring	The platform is merged and restructured by other companies	2
Withdraw Difficulty	Investors are difficult to withdraw. Or the platform default payment, and there is no announcement of future payment and withdraw.	359
Loss of Contact	The actual controller, other managers, and staffs of the platform could not be reached.	328
Bankrupt	The platform went bankrupt and lost contact.	46
Fraud	The platform is suspected of economic crimes. It has not been investigated.	208
Police Recorded	The platform is suspected of economic crimes. It has been investigated by police. The platform’s funds are frozen or suspended.	14
Vicious Runaway	The platform has been proven to be deliberately fraudulent and runaway.	488
Others	Other suspected runaways	31

).

According to another authoritative financial and economic database named Wind (https://www.wind.com.cn/NewSite/edb.html), as of 2016, there are 3410 runaway/problematic P2P-lending platforms. Of these, 1781 were confirmed already to be abscondments, while the status of the other 1629 were under further investigation. We reconciled the two sets of data, using each to cross-check the other, and estimated that our data account for 95% of runaway P2P-lending platforms in China. After deleting cases with indeterminate time of exit, we are left with 2941 cases for analysis, accounting for 86% of runways through 2016. The companies are located in 32 provincial-level regions in mainland China and Hong Kong. The data include their founding time, runaway or exit time, ownership, registered capital, total amount of capital involved, total borrowers involved, total lenders involved, and demographics, such as firm age (

Table 2. State level variables and descriptions.

Variables	Descriptions
Runaway year	The runaway year of the P2P-lending platforms
Founding year	The founding year of the P2P-lending platforms
Province	The province where the P2P platforms were registered in
Total amount involved	Sum of the total-amount outstanding balance of every P2P platform in the province and crisis in this year
Total P2P	The number of active P2Ps by 2016
Size	The funding size of the runaway P2Ps

).

4.2. Measures

Business Friendliness: The data for local government business friendliness come from Forbes’ China’s Best Cities for Business (http://china.forbeschina.com/review/list/002463.shtml). It is an annual ranking, from 2004, taking all the Chinese cities as units of analyses. The final rankings are published for the top 100 cities. The index is composed of eight items, including: city size, private-sector vitality, operation cost, innovation ability, freight capacity, passenger transport capacity, human resources, and consumption capacity. We calculate the provincial business friendliness index by taking an average score of the cities. The higher the value, the more business-friendly the province.

Law enforcement: We measure law enforcement by the proportion of expenditures on public security, the prosecutor’s office, and the court system, as a share of total local-public expenditures. The data come from the Annual Statistical Yearbook of China (http://www.stats.gov.cn/english/Statisticaldata/AnnualData/) and the Annual China Financial Statistics Yearbook (https://www.chinayearbooks.com/). The proportion of the expenditure on law enforcement also reflects the stage of development of the legal system [74]. The higher the ratio, the better the legal environment for doing business.

Private-sector GDP: We use the percentage of private-sector GDP as a proxy of informal institutions. The data for the percentage of private-sector GDP are a provincial-level measure is calculated from provincial statistical yearbooks and those published by local governments. For the provinces with missing information, we calculated their information from government website announcements and/or news reports. The measures are calculated as the ratio of local private-sector GDP to local total GDP in the P2P platform’s founding year (%).

Control variables: The control variables including provincial GDP and GDP growth. The data for provincial GDP and GDP growth come from the China Statistical Yearbook, which is published annually by the National Bureau of Statistics. We take natural logarithms to normalize the distribution for regression purposes. Finally, we include a control variable for firm size (in this case, funding amount). Firm size is a strong correlate with many individual-level firm characteristics such as structure, strategy, and resources that may impact performance success [75].

4.3. Estimation Models

According to our theoretical discussion, there are reasons to expect the relationship between business friendliness and the number or percentage of runaway P2P platforms to be positive and monotonic. In order to test the hypothesized functional form, we apply a model described as follows: We regress the number of runaway P2P platforms against business friendliness (Friendliness). Additional variables are included in the regression model to control for other determinants of runaways. Specifically, the variables are the P2P founding year’s private-sector GDP relative to local GDP (POE), founding year’s GDP (LnGDP), founding year and crisis year’s GDP growth (Growth₀ and Growth₁), and crisis year’s total number of P2P platforms (LnTotal). The last variable is a proxy for competitive intensity in the P2P industry, which is a standard I/O variable to explain firm failure. Therefore, to test Hypothesis 1, we estimated:

Runaways_it = β₀ + β₁ Friendliness_it0 + β₂ POE_it0 + β₃ LnGDP_it0 + β₄ Growth_it0 + β₅ Growth_it + β₆ LnTotal_it + β₇ LnSize_it + λ_t + ε_i,t

(1)

where Runaways_it is the natural logarithm of the number of P2P runaways in province i in year t. Friendliness_it0 is the business friendliness measure for province i in the year the runaway P2P was established. POE_it0 is the percent of the private sector’s GDP of the province GDP for province i in the year the runaway P2P was established. LnGDP_it0 is the natural logarithm of GDP for province i in the year the runaway P2P was established. Growth_it0 is province i’s GDP growth in the year the runaway P2P was established. Growth_it is province i’s GDP growth in the year the P2P failed. LnTotal_it is the natural logarithm of the total number of P2Ps in year t. LnSize_it is the natural logarithm of the total funding size of the runaway P2Ps in province i in year t.

To test Hypothesis 2, we estimate law enforcement’s influence:

Runaways_it = β₀ + β₁ Friendliness_it0 + β₂ POE_it0+ β₃ LnGDP_it0 + β₄ Growth_it0 + β₅ Growth_it + β₆ LnTotal_it + β₇ Law_it0 + β₈ LnSize_it +λ_t + ε_i,t

(2)

We estimate the interaction of private-sector GDP with the following model:

Runaways_it = β₀ + β₁ Friendliness_it0 + β₂ POE_it0 + β₃ LnGDP_it0 + β₄ Growth_it0 + β₅ Growth_it + β₆ LnTotal_it + β₇ Friendliness * POE_it0 + β₈ LnSize_it +λ_t + ε_i,t

(3)

The parameter λ_t is a time dummy variable that aims to capture the influence of economic factors that may also affect runaway platforms but are not controlled at the regional level, and ε_i,t is the i.i.d. residual.

We tested our hypothesis on the effect of runaway platforms using panel regression. The estimation technique allows us to control for unobservable heterogeneity and, therefore, eliminate the risk of obtaining biased results arising from heterogeneity (Hsiao, 1985). Second, panel data also allow us to avoid the problem of endogeneity, which might be present in our analyses and could also bias the estimates. ‘Friendliness’ is the target variable of interest. The results tell us that, for a particular year, whether provincial business friendliness explains the number of P2Ps established in that year that eventually ran away.

5. Results

5.1. Summary Statistics

We first report the number of total P2P platforms and runaways between 2010 and 2016 by province (Figure 1). It shows that the number of platforms and the runaways increased, especially in the period between 2012 and 2015 before slowing down in 2016.

Figure 2 shows the amount of money and the number of lenders affected in the runaways (Figure 2). The amount of money absconded surges in 2015 to nearly 90 billion RMB (approximately 14.52 billion USD using the 2015 average exchange rate), affecting more than 2 million lenders. The surge was driven by one runaway called “Ezubao”, which involved 65 billion RMB. Given how social media affects everything from political engagement [76] to stock price convergence [77], the Ezubao event had such profound social media coverage that the regulators finally stepped in to enact stricter regulations and enforcement in P2P lending, effectively wiping out meaningful differences between provinces in 2016 and beyond.

Table 3. Summary data on runaways in China by province.

P2P Incorporation Province	Active P2P in 2016	Total Runaways between 2011 and 2016	Amounts Involved in Runaways (in Millions Chinese Yuan)	Total Lenders
Anhui	468	138	¥2837	167,650
Beijing	1365	242	¥70,774	1,322,166
Chongqing	252	51	¥709	103,575
Fujian	287	70	¥603	105,148
Gansu	28	8	¥113	10,923
Guangdong	2239	481	¥14,960	671,591
Guangxi	146	42	¥197	52,333
Guizhou	97	23	¥251	42,272
Hainan	32	11	¥11	10,065
Hebei	283	85	¥174	111,977
Heilongjiang	38	12	¥146	15,525
Henan	223	62	¥841	188,898
Hong Kong	9	3	¥2	1345
Hubei	430	106	¥2052	149,960
Hunan	259	68	¥1092	72,607
Inner Mongolia	73	13	¥21	2435
Jiangsu	663	168	¥5483	197,395
Jiangxi	110	27	¥199	46,251
Jilin	77	20	¥45	39,725
Liaoning	34	9	¥38	9211
Ningxia	39	9	¥16	16,755
Qinghai	2	1	¥5	5670
Shaanxi	153	48	¥177	62,060
Shandong	1538	492	¥3567	381,937
Shanghai	1285	291	¥2694	343,112
Shanxi	68	17	¥12	12,520
Sichuan	357	78	¥955	95,043
Tianjin	125	26	¥48	28,146
Xinjiang	26	6	¥207	10,615
Xizang	1	0	0	0
Yunnan	85	35	¥506	37,788
Zhejiang	1364	299	¥7071	363,791
Grand Total	12,155	2941	¥115,805	4,678,489

reports the total number of runaway P2Ps by province, as well as the amount of investment and the number of lenders involved. As an internet platform, a firm can choose to incorporate in any province that may be different from its operating headquarters. Table 3 shows that P2Ps can be found in 32 provinces during the sample period, with 31 reporting runaways. There are 12,551 P2Ps established during the period, of which 2941 (23.4%) became runaways, affecting 4.6 million lenders and costing about 116 billion RMB of deposits.

To provide a preliminary view of the relationship between provincial business friendliness and the P2P runaways, we divide the provinces into four quartiles according to the Forbes’ list. Figure 3 shows how runaways were least likely to be found in the least business-friendly regions.

Table 4. Correlation Coefficient Matrix.

Variables	Mean	S.D.	1	2	3	4	5	6	7	8
1. LnRunaways	2.12	1.57	1
2. Business Friendlinessfounding year	0.64	0.12	0.360 **	1
3. Private Sector GDPfounding year	0.51	0.11	0.070	−0.067	1
4. LnGDPfounding year	17.70	0.95	0.400 **	0.865 **	−0.325 **	1
5. GDP Growthfounding year	0.12	0.05	0.076	0.112	−0.161	0.158	1
6. GDP Growthrunaway year	0.08	0.05	−0.140	0.181 *	−0.118	0.041	0.333 **	1
7. LntotalP2Pcrisis year	26.26	48.20	0.922 **	0.521 **	0.033	0.548 **	0.014	−0.075	1
8. LnSizecrisis year	17.33	17.27	0.851 ***	0.342 ***	−0.005	0.425 ***	0.286 ***	0.425 ***	0.815 ***	1

LnRunaways is the natural logarithm of the number of P2P runaways in a given province in a given year during the sample period of 2007 and 2016. Business Friendliness_{founding year} measures the provincial business friendliness measured by Forbes. This measure was taken in the year the runaway P2P was established. Private Sector GDP_{founding year} is the percentage of the local private-sector GDP of local GDP in the founding year of the runaway P2P. LnGDP_{founding year} is the natural logarithm of the provincial GDP in the founding year of the runaway P2P. GDP Growth_{founding year} is the provincial GDP growth in the founding year of the runaway P2P. GDP Growth_{runaway year} is the provincial GDP growth in the year of the P2Ps runaway. LntotalP2P_{crisis year} is the natural logarithm of the total number of existing P2P in the year P2Ps runaway. LnSize_{crisis year} is the natural logarithm of the total funding size of the runaway P2P. *, **, and *** denote statistical significance at the 10, 5, and 1% levels.

reports the means and standard deviations of the main variables used in the models and their correlations. The univariate result suggests that friendlier business environments are more attractive to runaway P2P platforms.

5.2. Regression Results

We then ran the base model to test hypothesis 1a. The results are shown in

Table 5. Business friendliness and P2P runaways.

	Number of Runaways
Business Friendlinessfounding year	1.731 * (0.069)
High Private Sector GDPfounding year	−0.067 (0.423)
LnGDPfounding year	−0.395 *** (0.001)
GDP growthfounding year	0.305 (0.804)
GDP growthrunaway year	−0.522 (0.519)
LntotalP2Pcrisis year	0.849 *** (0.000)
LnSizecrisis year	0.098 *** (0.010)
constant	2.701 (0.132)
Year effects	Yes
N	95
Adj. R2	0.926
Prob. > F	0.000

This table presents the empirical link between the provincial-level business friendliness and the number of runaways. The dependent variable is LnRunaways, the natural logarithm of P2P runaways in a given year. All the variables are defined earlier. We control year-fixed effects. * and *** denote statistical significance at the 10 and 1% levels. The reported p-values in the parentheses reflect White’s heteroskedasticity correction.

. The dependent variable for Table 5 is the natural logarithm of the number of runaway P2P in a particular province in a given year. After controlling for other factors that may contribute to the volume of runaways, we find that the provincial business environment still had significant and positive impact on the number of P2P runaways. The coefficient is 1.731 (p < 0.10), which supports Hypothesis 1a.

Note that the size of the regional GDP had a negative relationship to the number of P2P runaways. This has interesting implications. Provinces with established economic performance may be a result in more efficient capital allocation, effective government officials, better law enforcement, or combination of these factors. These factors point to an institutional environment that discourages welfare-destroying activities, including P2Ps that are bad actors.

We next tested Hypothesis 1b, which says that the positive relationship between business friendliness and runaways is attenuated by robust informal institutions, on the assumption that an economically efficient business-friendly environment is one that also discourages bad actors. We proxied the development of informal institutions by the size of the private sector to the provincial GDP. The results are shown in

Table 6. Business friendliness and P2P runaways: the attenuation effect of the private sector.

	Number of Runaways
Business Friendlinessfounding year	2.572 ** (0.019)
High Private Sector GDPfounding year	1.458 * (0.062)
Business Friendlinessfounding year * High Private Sector GDPfounding year	−2.320 ** (0.048)
LnGDPfounding year	−0.466 *** (0.000)
GDP growthfounding year	0.806 (0.524)
GDP growthrunaway year	−0.697 (0.395)
LntotalP2Pcrisis year	0.817 *** (0.000)
LnSizecrisis year	0.101 (0.009)
constant	3.319 * (0.072)
Year effects	Yes
N	95
Adj. R2	0.928
Prob. > F	0.000

This table presents the empirical link between the provincial-level business friendliness and the number of runaways by adding the effect of the private-sector. The dependent variable is LnRunaways, the natural logarithm of P2P runaways in a given year. To test the impact of whether informal institutions discourage inefficient decisions, we rank the provinces’ private-sector GDP as a percentage of its total provincial GDP each year and split the sample by the median value. The provinces with this variable higher than the annual median value are assigned with a value of 1 under the dummy variable named High Private Sector GDP_{founding year}, zero otherwise. All other variables are defined earlier. We control year-fixed effects. *, **, and *** denote statistical significance at the 10, 5, and 1% levels. The reported p-values in the parentheses reflect White’s heteroskedasticity correction.

.

We rank the provinces’ private-sector GDP as a percentage of its total provincial GDP each year and split the sample by the median value. The provinces with this variable higher than the annual median value is assigned a value of 1, labeled as High Private Sector GDP_{founding year}. We include this variable and its interaction with the business-friendliness variable in the baseline model. Consistent with our conjecture, the results in Table 6 show that while business friendliness increases the number of runaways, higher private-sector presence in the region attenuates the relationship. The coefficient of the interaction variable of the private-sector GDP and business friendliness is −2.320 and significant at 5% level.

In Hypothesis 2a, we stated that the positive link between business friendliness and the number of P2P runaways is weakened when the strength of the regional law enforcement regime is included. In

Table 7. Business friendliness and P2P runaways: the role of law enforcement.

	Number of Runaways
Law Enforcementfounding year	−10.644 ** (0.028)
Business Friendlinessfounding year	1.208 (0.198)
High Private Sector GDPfounding year	−0.033 (0.687)
LnGDPfounding year	−0.362 *** (0.003)
GDP growthfounding year	0.493 (0.688)
GDP growthrunaway year	−0.845 (0.310)
LntotalP2Pcrisis year	0.909 *** (0.000)
LnSizecrisis year	0.095 ** (0.013)
constant	3.070 * (0.090)
Year effects	Yes
N	95
Adj. R2	0.929
Prob. > F	0.000

This table presents the empirical link between the provincial-business friendliness and the number of runaways by adding the effect of the strength of the provincial law enforcement. The dependent variable is LnRunaways, the natural logarithm of P2P runaways in a given year. We measure law enforcement by the proportion of the public security, the procuratorate, and the court expenditure in local public expenditure of the total local public expenditure. All other variables are defined earlier. We control year-fixed effects. *, **, and *** denote statistical significance at the 10, 5, and 1% levels. The reported p-values in the parentheses reflect White’s heteroskedasticity correction.

, we test this hypothesis by including a law-enforcement measure in the base model. The results show that although there is a positive link between business friendliness and the number of runaways, it is not statistically significant anymore. In contrast, the variable Law Enforcement_{founding year} has significant negative effect on the number of the runaways, with a coefficient of −10.644 and significant at 5% level.

The result confirms the role of formal institutions in shaping economic activities and outcome. Unlike regulatory policies that are less predictable and transparent, rule of laws and the enforcement of it contribute to more sustainable economic decisions and discourage rent seeking and opportunistic activities.

Finally, we test Hypothesis 2b, in which we state that there is a complementary effect between the formal institutions (law enforcement) and the informal institutions (the size of the private sector). To test Hypothesis 2b, we split the sample into two groups by the strength of the provincial-level law enforcement. The cutoff point is the median value (0.05) of the relative share of public resources expended on those activities, which includes public security, procuratorate, and the courts. The results of

Table 8. Business friendliness and P2P runaways: The complementary effect of law enforcement and private sector.

	High Law Enforcement (≥0.05)	Low Law Enforcement (<0.05)
Business Friendlinessfounding year	4.657 *** (0.004)	−0.035 (0.985)
High Private Sector GDPfounding year	3.198 *** (0.003)	−0.503 (0.680)
Business Friendlinessfounding year * High Private Sector GDPfounding year	−5.124 *** (0.001)	0.832 (0.652)
LnGDPfounding year	−0.747 *** (0.001)	−0.237 * (0.077)
GDP growthfounding year	2.075 (0.197)	−2.321 (0.120)
GDP growthrunaway year	−1.197 (0.192)	−6.315 * (0.098)
LntotalP2Pcrisis year	0.820 *** (0.000)	1.121 *** (0.000)
LnSizecrisis year	0.081 (0.127)	0.082 (0.104)
constant	7.086 ** (0.019)	3.600 * (0.062)
Year effects	Yes	Yes
N	59	36
Adj. R2	0.943	0.956
Prob. > F	0.000	0.000

This table presents the empirical link between the provincial-level business friendliness and the number of runaways by examining the complementary effect of the strength of the provincial law enforcement and the development of the private sector. The dependent variable is LnRunaways, the natural logarithm of P2P runaways in a given year. We divide the sample to High and Low Law Enforcement groups by the median value of this variable, which is 0.05. All other variables are defined earlier. We control year fixed effects. *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels. The reported p-values in the parentheses reflect White’s heteroskedasticity correction.

show that the significant mitigation effect of the informal institutions on P2P runaways only exists in regions with strong law enforcement. The coefficient of the interaction variable of the private-sector GDP and business friendliness is −4.657 and significant at 1% level. These results show that only in places with strong law enforcement can private-sector prevalence play a significant role in shaping efficient economic activities that lead to positive outcome and more sustainable economic growth.

5.3. Robustness Tests

Our findings rely on the measurement of business friendliness, which may be subject to definitional and measurement bias. Hence, to test the robustness of our key results, we repeat the baseline regression model by using alternative measures of business friendliness. Specifically, we use two alternative measures of business friendliness. First, we compute the ranking of business friendliness by removing the items such as the freight and passenger transport capacity, which might be less related to the research question we investigated in this study on the internet platform registration. We name the first alternative measure of business friendliness: Robustness₁ (without freight and transport capacity). Second, we replace the business-friendliness measure by a so-called marketization index, which reflects the relationship between the local government and the local market. We collect the regional marketization data from the National Economic Research Institute (http://www.neri.org.cn/). The results are shown in

Table 9. Robustness tests.

	(1)	(2)
Robustness1	1.576 *
	(0.097)
Marketization		0.212 ***
		(0.005)
High Private Sector GDPfounding year	0.289 **	0.193 *
	(0.025)	(0.096)
LnGDPfounding year	−0.018	0.091
	(0.889)	(0.271)
GDP growthfounding year	−2.671 *	−1.560
	(0.083)	(0.278)
GDP growthrunaway year	1.403	−0.069
	(0.238)	(0.954)
LntotalP2Pcrisis year	0.010 ***	0.008 ***
	(0.000)	(0.000)
LnSizecrisis year	0.249 *** (0.000)	0.256 *** (0.000)
Constant	−4.463 **	−9.940 ***
	(0.019)	(1.715)
Year effects	Yes	Yes
N	95	103
Adj. R2	0.851	0.808
Prob. > F	0.000	0.000

This table presents the empirical link between the provincial-level business friendliness and the number of runaways using alternative measures of business friendliness. Robustness₁ refers to a friendliness measure without the factor of the passenger capacity. Marketization is a measure that reflects the relationship between the local government and the local market. We collected the regional marketization data from the National Economic Research Institute. All other variables are defined earlier. We control year-fixed effects. *, **, and *** denote statistical significance at the 10, 5, and 1% levels. The reported standard errors in the parentheses reflect White’s heteroskedasticity correction.

, which reports that our main findings remain qualitatively the same.

6. Discussions and Conclusions

This study examined the association between business friendliness and P2P runaways while exploring the moderating effect of formal and informal institutions. Why, given the high rate of failure, do depositors still put their money in these entities? We speculate that the P2P phenomenon arose during a period of economic growth in China’s development history, which fosters the tendency for investors’ herd behaviors and willingness to ignore negative information when anticipated returns are extraordinary or even guaranteed, as with many Ponzi schemes [78]. The high rate of failure, relative to other types of businesses, suggests that these FinTech companies were seen as technology startups rather than financial institutions subject to safety regulations. In fact, in the early days of the phenomenon, they were encouraged by the Chinese government precisely because they were seen as innovative technology solutions to the problem of credit availability. Therefore, investors’ risk tolerance was much higher, than say if these companies were regarded as savings institutions or even banks. Combined with information asymmetry it is no wonder that investors continued to pour money into these companies even in the face of the high rates of failure. With respect to the hypotheses, our results suggest the following implications.

6.1. Business Friendliness and P2P Runaways

The results support H1a, showing a positive relationship between business friendliness and P2P runaways. The results suggest that bad actors are more likely to choose regions with more friendly-business environments such as reduced procedures for setting up businesses, less processing time and/or reduced capital requirement to ply their trade.

6.2. Attenuation Effect of the Private Sector (Informal Institutions)

Our results support Hypothesis 1b, indicating that the association of business friendliness with P2P runaways is contingent on the development of the local private economy. A place’s informal institutions are a good constraint on market behavior. One possible reason is that other participants and managers in the market are more experienced at identifying those with undesirable motivations. Another explanation is that a region with a strong private economy can help corporations stay in business, avoid corruption, and reduce crime.

6.3. The Role of Law Enforcement (Formal Institutions)

Our results indicate that runaways will be significantly discouraged by the strength of law enforcement, which supports Hypothesis 2a and is consistent with prior findings [3,20,21]. When law enforcement is weak, markets located in regions with the so-called business friendly policies/loose regulation can easily grow to be a breeding ground for corruption, rent seeking, or even illegal activities, and ultimately, destroy the existing market order and damage the region’s potential for sustainable growth.

6.4. Complementary Effects of Formal and Informal Institutions

Existing studies have shown the importance of informal institutions in regulating the behavior of enterprises in the absence of formal institutions (Peng et al., 2009). We extended this literature by showing a complementary effect: when formal institutions are strong, informal institutions are also strong. Hypothesis 2b contributes to the new institutional economics literature by showing that institutions are not substitutes but are in fact complementary in their effects on economic welfare. In markets where law enforcement is strong, the private sector can play a more effective governing role on moral hazard. When enforcement is weak, however, no such effect was detected.

6.5. Conclusions

We are not questioning the value of business-friendly government policy in this paper. Indeed, the World Bank’s annual Doing Business survey reports that most business-friendly countries are strong on regulation and “focus their efforts more on protecting property rights” (http://www.doingbusiness.org/content/dam/doingBusiness/media/Annual-Reports/English/DB05-FullReport.pdf). It observes that thriving private sectors in business-friendly countries seems to be associated with higher economic growth. Using the size of the private sector as a proxy for the general business environment, we found that a thriving private sector in provinces with business-friendly policies actually reduces the probability of runaways. We showed that the positive welfare effect of business-friendly policies in provinces with thriving private sectors is further enhanced in those provinces with strong law-enforcement regimes.

We are concerned about economies with weak legal institutions. Regulations/policies promulgated by local governments in the hope of promoting economic growth do not necessarily lead to healthy economic activities. The rush to adopt regulations that encourage competitive inefficiencies may attract moral hazard. Economic policy alone is not sufficient to induce welfare-enhancing economic growth. The recent socioeconomic literature discusses the complementary effects of institutions in terms of how the effectiveness of an institution depends on the effectiveness of adjunct institutions [79]. Hence, business-friendly policies must be supported by strong business law enforcements.