Why Don’t Small Firms Merge? Experimental Evidence on Information Barriers

How to cite

Hardy, Morgan; Witte, Marc; Menzel, Andreas; Kim, Seongyoon. 2021. "Why Don’t Small Firms Merge? Experimental Evidence on Information Barriers." AEA RCT Registry. AEARCTR-0007428. https://doi.org/10.1257/rct.7428-1.0

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

Morgan Hardy* , Marc Witte , Andreas Menzel , Seongyoon Kim

* Primary investigator

Registry ID
AEARCTR-0007428
DOI
https://doi.org/10.1257/rct.7428-1.0
First registered
Last updated
Status
On going
Trial dates
2019-08-01 – 2022-04-01
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/7428

Abstract

This research project aims to experimentally investigate information barriers to small firm consolidation. We exploit rich panel data on the universe of garment making firm owners in Hohoe, Ghana, including owner and owner household characteristics, firm characteristics and outcomes, and reported within-industry peer network interactions. We elicit firm owners’ willingness to pay for information about within-industry peer willingness to hire and be hired by the owner using Becker-Degroot-Marschak method and exploit the conditionally random variation to study the impacts of such information on firm-level growth and profitability as well as dyadic-level mergers.

Intervention

The intervention is the receipt of (a) list(s) of names of other firm owners the market that are willing to hire or be hired by the treated individual.

Experimental design

Following the Becker-Degroot-Marschak (BDM) method, each respondent receives 5
GHC (approximately 91 US cents at the time of the first wave) that can be used as offers
for the lists of other firm owners that are willing to work for or willing to hire the
respondent. Respondents can then bid for each list (in integer amounts), with a maximum
possible bid for each list of 3 GHC. Then, the price for the list is randomly drawn and takes
integer values between 0 and 4 GHC. If the bid is greater than or equal to the price, then
the list is successfully purchased, and any amount not spent is returned to the respondent
in cash. Each respondent was also placed into four different treatment groups, where the
information on the list varied. These groups are further described in the attached analysis plan.

Design details

Randomization method
Price randomization was completed by researchers using STATA prior to entering the field and preloaded on Survey CTO to be revealed to respondent (and enumerator) after willingness to pay exercise.
Randomization unit
Firm owner
Clusters
Information was randomized at the respondent (firm owner) level. There were 569 firm owners in the market identified during our census.
Observations
The number of observations varies by outcome (between 1 and 6 post-observations per firm owner).
Arms
This is non-trivial for this project as the random treatment is conditional on willingness to pay.
Primary outcomes
firm size (number of workers and wage expenses), firm performance (profits and sales), firm boundaries (outsourcing expenses and revenue, capital rental expenses and revenue)

Cited by

No citation data available for this registration yet.

Keywords