Informational frictions and the persistence of cross-subsidies in equilibrium

How to cite

Pugliese, Joao Francisco; Braga, Otavio; Teixeira, Lucas Iten. 2026. "Informational frictions and the persistence of cross-subsidies in equilibrium." AEA RCT Registry. AEARCTR-0018745. https://doi.org/10.1257/rct.18745-1.1

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

Joao Francisco Pugliese* , Otavio Braga , Lucas Iten Teixeira

* Primary investigator

Registry ID
AEARCTR-0018745
DOI
https://doi.org/10.1257/rct.18745-1.1
First registered
Last updated
Status
In development
Trial dates
2026-05-29 – 2027-03-31
Countries
Brazil ()
JEL codes
D14, O16, G51, G41
Registry record
http://www.socialscienceregistry.org/trials/18745

Abstract

We study the role of informational frictions and biased consumer beliefs in sustaining market equilibria characterized by large, regressive cross-subsidies between different types of consumers. The context of this study is the Brazilian "consórcio" market, the largest formal instance of rotating finance associations ("savings groups"), representing about 20-30% of all yearly vehicle credit issued. A rotating finance group works as a closed-system of transfers where money is borrowed and lent across participants, and an allocation mechanism decides the ordering of who borrows first. Ex-ante cross-subsidies emerge in the group when the ordering is not fully randomized and those who borrow last are not compensated for their late position in the queue. Because the order of borrowing is partly random, there is uncertainty over credit issuance dates, and beliefs about when credit will be issued is central to determine the expected cost of financing through consórcios. We document the market dynamics leveraging administrative data from the Central Bank covering the universe of 60 million consórcio contracts across 75,000 groups that begun since 2010. As part of this trial, we recruit consumers online and conduct a survey with a "field" component where we: (i) elicit their beliefs about the probabilities of the timing of credit in this product; (ii) randomize a debiasing information intervention; (iii) measure willingness-to-pay for an alternative product with less cross-subsidization, where a larger share of credit is distributed through lotteries. We complement consumers' surveys by collecting survey responses from consórcio brokers, who are financial agents specialized in selling consórcios through banks, car dealerships, and specialized administrator firms.

Intervention

Consumers: our intervention consists of a short video illustrating the dynamics of credit allocation and timing in consórcios. The video combines conceptual points with empirical data obtained from the Central Bank data. Though 8 different scenes containing visual motions, our video conveys the mechanics of credit allocation and relevant statistics such as the proportion of credit issued through lottery or bidding, and the expected wait time for credit issuance through lottery.

Experimental design

The experimental design consists of an individual-level randomization into two treatment arms: (C) Control and (D) Debiasing/Disclosure and (DG) D + Group Sizes Explanation. Participants are stratified into six groups according to their (i) bidding capacity, and (ii) desired good to be acquired (called "segment"). Participants assigned to either (D) or (DG) are shown the video prepared by researchers. Participants assigned to (C) view a generic video about consórcios from the Central Bank YouTube channel, therefore consisting of an active control group.

Design details

Randomization method
Randomization of the main intervention is done through Qualtrics' survey flow "Randomizer" block with the "Evenly Present Elements" enabled.
Randomization unit
Individual
Clusters
1000 consumers 200 brokers
Observations
1000 consumers 200 brokers
Arms
Consumer survey: 500 consumers for (D) and (C). Brokers survey: 50 brokers for each arm.
Primary outcomes
Primary outcomes consist of: (i) quantitative, qualitative, and probabilistic beliefs about bidding share (as a proporiion of total allocations) and allocation timing; (ii) beliefs about own likelihood of bidding, best optimal bid, competitiveness of bid; (iii) opinions about the product; (iv) elicited choices between consórcio groups with different attributes; (v) elicited choices between consórcios and financing; (vi) elicited choices between consórcios and savings products; (vii) incentivized willingness-to-pay for an offer of an alternative consórcio group with a higher share of lottery allocations, which is rarely found in the market, through a MPL consisting of a choice between a gift card and the possibility of receiving the offer from a partner administrator. We also plan to obtain data on whether people buy this product.
Secondary outcomes
N/A.

Cited by

No citation data available for this registration yet.

Keywords