Abstract
In this paper, we study regional discrimination in a peer-to-peer lending scenario and provide novel empirical evidence for theories of soft information collection and information cost. We find that the regional information matters for borrowers' funding probabilities and that discrimination is profit-oriented or taste-oriented depending on the specific region. Moreover, using borrowers' birthplace as an instrumental variable, we find no evidence of genuine discrimination based purely on region in the peer-to-peer lending market.
Acknowledgments
We are very grateful to the editor and two anonymous referees for their valuable comments and constructive suggestions. We would also like to thank Shulan Hu, Xiaoxi Li, Junmin Wan, Aluna Wang, Yi Xue as well as seminar and conference participants at Fukuoka University, University of Edinburgh, Shanghai University of Finance and Economics, Zhongnan University of Economics and Law, and Wuhan University of Technology for helpful comments. This research has received financial support from the Low and Middle Income Countries (LMIC) Travel and Partnerships Fund, which is a key component of Global Challenges Research Fund allocated by the University of Edinburgh. All remaining errors are our own.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
Tong Wang http://orcid.org/0000-0001-6416-6622
Correction Statement
This article has been republished with minor changes. These changes do not impact the academic content of the article.
Notes
1 as stated in Agarwal and Ben-David (Citation2018)
2 The lenders can also delegate the investment decisions to the platform. In this case, the investment choices are determined by algorithms and any algorithmic bias may also lead to a discriminatory result (Chander Citation2016; Hajian, Bonchi, and Castillo Citation2016).
3 There are three types of loan applications in total: unsecured, joint liability and field certification. Joint liability and field certification loans have a significant lower default risk and a higher likelihood of receiving loans compared with the unsecured loans.
5 The data portal of National Bureau of Statistics of China is http://data.stats.gov.cn
6 Actually, only 11 of 1767 applications failed after 31/12/2015.
7 Loan specification control variable include: ,
and
; and borrower profile control variables include: rating,
,
, gender, education,
,
,
,
,
and
.
8 We admit that, to some extent, the robustness of post-regulation results is limited due to an unbalanced sample size (there are 76,225 pre-regulation samples of unsecured loan applications, while only 1767 samples after the regulatory shock).
9 This is also the address shown in the application document.
10 Shanhaiguan is an important pass in ancient China, and all three northeastern provinces are outside Shanhaiguan. This saying de facto means ‘Do not invest in the north-eastern region’.
11 Data is from Statistics Bureau of China
12 Data are from the Statistics Bureau of China
13 We conducted a full analysis as we did with the eastern region. The results had no significant difference with the results in Table .
14 We rule out those provinces that are closer to the eastern region.
15 See the report published by Yingcan Consulting https://www.sohu.com/a/124127840_530780
16 See the region distribution map in Figure .
17 According to Table , the decision process is not always aligned with the profit maximizing objective.