Government agents’ capacity and MSME performance

How to cite

Stegmann, Andreas; Szeidl, Adam; Yudhistira, Muhammad Halley. 2024. "Government agents’ capacity and MSME performance." AEA RCT Registry. AEARCTR-0013531. https://doi.org/10.1257/rct.13531-3.0

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

Andreas Stegmann* , Adam Szeidl , Muhammad Halley Yudhistira

* Primary investigator

Registry ID
AEARCTR-0013531
DOI
https://doi.org/10.1257/rct.13531-3.0
First registered
Last updated
Status
In development
Trial dates
2024-05-08 – 2025-12-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/13531

Abstract

MSME capabilities can be more effectively improved with customized programs such as mentoring or peer interactions. But we lack evidence on how the government can implement and scale such programs. To make progress, we study a government-run small-group mentoring program provided to thousands of MSMEs each year in Depok City, Indonesia. We introduce three randomized interventions to 2,200 firms and 63 mentors. (1) We randomize admittance to evaluate whether a government-run program can achieve similar gains as researcher-run programs. (2) We improve the government’s delivery capacity by training and monitoring mentors, to study the effect of state capacity on program effectiveness. (3) We enhance content on marketing practices, to evaluate the effect of providing content demanded specifically by the MSMEs. The results can help design scalable programs that meaningfully improve MSME performance.

Intervention

We administer two interventions to selected mentors who deliver weekly mentoring sessions to local MSMEs:

Intervention to improve mentoring delivery capacity.

This intervention will consist of the following elements: (i) two days of training in communication and organizational skills, and (ii) improved monitoring, in which a dedicated expert we hire regularly attends the meetings of these mentors, gives feedback about mentoring techniques, and is available for consultations.

Intervention about marketing practices.

This intervention will consist of the following elements: (i) two days of training about marketing and sales practices that are geared to the type of MSMEs in our sample. The training also covers the steps to obtain a halal certification, and (ii) monitoring specifically about marketing by dedicated experts who will provide additional feedback to mentors specifically about the marketing and halal certification process.

Experimental design

We conduct training and monitoring to selected mentors who are responsible to train and mentor local MSMEs. We randomize mentees and mentors into the following nested treatment arms, which feature an increasing set of features:

C: Control (350 firms). These applicant firms meet the eligibility criteria but will not be accepted to the program.

T1: Mentoring program (~360 firms, 21 mentors). These firms are accepted and receive the program as designed by the government.

T2: Mentoring with enhanced capacity (~360 firms, 21 mentors). In addition to T1, for these firms we improve delivery capacity.

T3: Mentoring with enhanced capacity and content (~360 firms, 21 mentors). In addition to T2, we improve content.

To create the control group, in our first intervention we randomize acceptance to the program. We plan to pre-select a pool of about 2,000 applicants who meet the minimum requirements of our government partner. Out of these, we randomize about 1,450 to be accepted to the program, and about 350 to be denied acceptance due to space constraints. The latter constitute our pure control group. This randomization will be stratified at the level of the 11 subdistricts.

Design details

Randomization method
Randomisation done in office by a computer
Randomization unit
We will randomize the acceptance into the program at the individual level, stratified at the level of subdistricts. We will randomize the treatment status at the mentor/village level.
Clusters
63 mentors/villages.
Observations
63 mentors at sub-district level 18 supervisory mentors 1100 mentees 350 control MSMEs
Arms
C: Control (350 firms). T1: Mentoring program (~360 firms, 21 mentors). T2: Mentoring with enhanced capacity (~360 firms, 21 mentors). T3: Mentoring with enhanced capacity and content (~360 firms, 21 mentors).
Primary outcomes
Firm performance, mentor practices, and mentee business practices.

Cited by

No citation data available for this registration yet.

Keywords