Indirect Effects of Access to Finance

How to cite

Cai, Jing; Szeidl, Adam. 2022. "Indirect Effects of Access to Finance." AEA RCT Registry. AEARCTR-0009506. https://doi.org/10.1257/rct.9506-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Jing Cai* , Adam Szeidl

* Primary investigator

Registry ID
AEARCTR-0009506
DOI
https://doi.org/10.1257/rct.9506-1.0
First registered
Last updated
Status
Completed
Trial dates
2013-07-01 – 2020-09-01
Countries
China ()
Registry record
http://www.socialscienceregistry.org/trials/9506

Abstract

We created experimental variation across local markets in China in the share of firms having access to a new loan product, to measure the direct and indirect effects of access to finance. We find that: (1) Access to finance had a large positive direct effect on the performance of treated firms. (2) Access to finance had a similar-sized negative indirect effect on the performance of firms with treated competitors. Due to these opposing effects, we do not find detectable gains in aggregate producer surplus. (3) Access to finance had a positive direct effect on business practices, service quality, and consumer satisfaction, and a negative effect on price. None of these effects were offset by indirect effects, suggesting net gains in consumer surplus. (4) Two additional indirect effects were active: diffusion of borrowing to firms with treated peers, and diffusion of demand to firms with treated neighbors. (5) Combining several effects in a model- based evaluation, we estimate that the loan had a private return of 74%, most of which was offset by losses to competitors, and a social return of 60%, most of which was driven by gains to consumers.

Intervention

We created experimental variation across local markets in China in the share of firms having access to a new loan product, to measure the direct and indirect effects of access to finance.

Experimental design

We introduced the treatment using a combination of a market-level and a firm-level randomization. Out of our 78 markets, we randomized 37 to have high treatment intensity, 10 markets to have medium treatment intensity, and 31 markets to pure control.5 In high treatment intensity markets we treated a randomly selected 80% of firms; in medium intensity markets we treated a randomly selected 50% of firms; and in pure control markets we treated no firms. Every treated firm was visited every month for a year by a loan officer who provided information about the new loan product, and if the firm decided to apply, provided help with the application process, including filling in the relevant forms.

Design details

Randomization method
randomization done in office by a computer
Randomization unit
market and firms
Clusters
78 markets
Observations
3173 firms
Arms
Out of our 78 markets, we randomized 37 to have high treatment intensity, 10 markets to have medium treatment intensity, and 31 markets to pure control. Among the 3173 sample firms, 1436 were treated.
Primary outcomes
Firm revenue, profit, borrowing
Secondary outcomes
Firm management, employment, price

Cited by

No citation data available for this registration yet.

Keywords