Fostering Social-Media Entrepreneurs: Evidence from A Field Experiment in China

How to cite

Lu, Fangzhou. 2024. "Fostering Social-Media Entrepreneurs: Evidence from A Field Experiment in China." AEA RCT Registry. AEARCTR-0015005. https://doi.org/10.1257/rct.15005-1.0

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

Fangzhou Lu*

* Primary investigator

Registry ID
AEARCTR-0015005
DOI
https://doi.org/10.1257/rct.15005-1.0
First registered
Last updated
Status
On going
Trial dates
2022-02-01 – 2025-12-31
Countries
China ()
JEL codes
C93, D1, G30
Registry record
http://www.socialscienceregistry.org/trials/15005

Abstract

We conduct a field experiment in China to incentivize small business owners to become social media content creators. Our primary subsidy is provided as a social media advertisement subsidy for content boosting fees, directly given to the social media platform. This subsidy is conditional upon business owners creating new video, live stream, or photo posts to promote their business. We intend to study the outcomes of revenue, profit and the number of employees after the intervention. We compare the effects with a direct cash transfer. Additionally, we also study whether increased sales enhance small businesses’ likelihood of opening online stores and subsequently accessing fintech credit.

Intervention

Our experiment provides the treatment group with a subsidy of 1,000 RMB as a social media advertisement fee, such as the Dou+ services on TikTok. Social media services like Dou+ are designed to boost the reach and visibility of user-generated content on the TikTok platform by allowing content creators to pay for increased expo- sure. Essentially, more people are likely to see the business’s video, streaming, or photo in their feed. This practice is also commonly observed across other social media platforms, where content creators leverage paid promotional tools to enhance user engagement and, consequently, their economic prospects. By providing households and small businesses with access to social media promotion services comparable to paid marketing services, we can eval- uate the potential economic benefits and behavioral shifts resulting from increased online presence and exposure. According to TikTok’s official website, one RMB of Dou+ reaches between 25 to 50 new viewers. Therefore, our 1,000 RMB subsidy allows small businesses to reach between 25,000 and 50,000 new views. Social media promoters on other platforms have a similar customer reach. This treatment is offered only on the condition that business owners create a new video, live stream, or photo post to promote their business. In the experiment, our control group receives nothing. We also design a direct cash transfer of the same amount, unconditional, to serve as an alternative control group.

In summary, The Treatment group received a social media content booster fee of 1000 RMB, the Control group received no intervention, and the Cash Transfer group received a direct cash transfer of 1000 RMB.

Experimental design

We randomize 5,000 small businesses in the provinces of Guangdong, Zhejiang, Jiangsu, and Liaoning, China between control and treatment groups. Randomization of treatment and control groups was done by stratifying businesses by sector and region.The re-randomization method was used to ensure balance on key baseline characteristics, including entrepreneur age, gender, years of schooling, business age, total asset value, and business revenue. The re-randomization method involves repeatedly randomizing until a desirable balance on covariates is achieved. Specifically, we first allocated small businesses to the treatment and control groups on the basis of a randomly generated number. Using this allocation, we then calculated the maximum and the average t-statistics on the differences in averages across the treatment and control groups for the following variables: Entrepreneur age, gender, years of schooling, business age, total asset value, business revenue/sales, If the maximum t-statistic for these variables was higher than 1.25 or the average t-statistic was higher than 0.35, we drew a new random number and allocated firms to the treatment and control groups on the basis of this new number. We repeated this process until the maximum t-statistic was 1.25 or lower and the average t-statistic was 0.35 or lower.

The randomization method adopted pairs businesses together based on the minimum distance between their baseline characteristics and randomly assigns them to treatment and control. Data is collected on a total of 5,000 small businesses – 1000 of whom have received the treatment intervention.

Baseline business surveys are administered to the treatment and control businesses before the intervention. Regular follow-ups are conducted via telephonic surveys to capture business performance metrics such as revenue, profit, and employee numbers. This is further cross-checked with the administrative data collected from social media platforms and fintech lenders. A midline survey is conducted one year post-intervention, and an endline survey is conducted two years and three years post-intervention.

The treatment group receives a social media advertisement subsidy of 1,000 RMB for content boosting fees, directly given to the social media platform, conditional upon business owners creating new video, live stream, or photo posts to promote their business. The control group receives no intervention.

Design details

Randomization method
randomization done by a computer in public. We use the Stata code automatically to re-randomized as follows. We first allocated small businesses to the treatment and control groups on the basis of a randomly generated number. Using this allocation, we then calculated the maximum and the average t-statistics on the differences in averages across the treatment and control groups for the following variables: Entrepreneur age, gender, years of schooling, business age, total asset value, business revenue/sales, If the maximum t-statistic for these variables was higher than 1.25 or the average t-statistic was higher than 0.35, we drew a new random number and allocated firms to the treatment and control groups on the basis of this new number. We repeated this process until the maximum t-statistic was 1.25 or lower and the average t-statistic was 0.35 or lower.
Randomization unit
individuals (entrepreneurs)
Clusters
5000 individuals (entrepreneurs)
Observations
5000 individuals (entrepreneurs)
Arms
1000 individuals (entrepreneurs) for social-media treatment, 1000 individuals (entrepreneurs) for cash treatment, 3000 for the control group
Primary outcomes
business revenue, profit, and employee growth. The entrepreneurs' social media behavior after intervention

Cited by

No citation data available for this registration yet.

Keywords