Working Paper: NBER ID: w28724
Authors: Paul Gertler; Brett Green; Catherine Wolfram
Abstract: A new form of secured lending using “digital collateral” has recently emerged, most prominently in low- and middle-income countries. Digital collateral relies on lockout technology, which allows the lender to temporarily disable the flow value of the collateral to the borrower without physically repossessing it. We explore this new form of credit both in a model and in a field experiment using school-fee loans digitally secured with a solar home system. Securing a loan with digital collateral drastically reduces default rates (by 19 pp) and increases the lender’s rate of return (by 38 pp). Employing a variant of the Karlan and Zinman (2009) methodology, we decompose the total effect on repayment and find that roughly one-third is attributable to adverse selection, and two-thirds is attributable to moral hazard. In addition, access to digitally secured school-fee loans significantly increases school enrollment and school-related expenditures without detrimental effects to households’ balance sheet.
Keywords: Digital collateral; Secured lending; Credit markets; Moral hazard; Adverse selection; Education financing
JEL Codes: G20; I22; O16
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Digital collateral (D16) | Repayment rates (E43) |
| Digital collateral (D16) | Moral hazard (G52) |
| Digital collateral (D16) | Adverse selection (D82) |
| Digital collateral (D16) | Internal rate of return (IRR) (H43) |
| Digitally secured school-fee loans (I22) | School enrollment (I21) |
| Digitally secured school-fee loans (I22) | Expenditures (H59) |
| Digital collateral (D16) | Share of households for profitable loans (G51) |