What Motivates Innovative Entrepreneurs? Evidence From a Global Experiment at MIT

How to cite

Sen, Ananya; Oh, Joohyun; Guzman, Jorge. 2019. "What Motivates Innovative Entrepreneurs? Evidence From a Global Experiment at MIT." AEA RCT Registry. AEARCTR-0004125. https://doi.org/10.1257/rct.4125-2.0

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

Ananya Sen* , Joohyun Oh , Jorge Guzman

* Primary investigator

Registry ID
AEARCTR-0004125
DOI
https://doi.org/10.1257/rct.4125-2.0
First registered
Last updated
Status
In development
Trial dates
2019-04-25 – 2019-11-25
Countries
Private
JEL codes
O 31, O 35, O 40, O 50
Registry record
http://www.socialscienceregistry.org/trials/4125

Abstract

Entrepreneurship plays a central role in translating new inventions into useful products and services, and it is often stated that income or wealth is critical in establishing the incentives to create these inventions in the first place. However, despite this strong emphasis of theory on income incentives for entrepreneurship, empirical research has found little evidence of them. Existing studies of comprehensive samples of business owners or self-employed individuals across the United States find that most entrepreneurs do not think they are innovating, do not think they are addressing a new market, have little intention of growth, and are often largely driven by non-pecuniary incentives such as the personal value of ‘being your own boss’. Critics of this lack of evidence often emphasize is that the empirical analysis has focused on the wrong sample. The entrepreneurs of economic theory are ‘innovative (or high-growth) entrepreneurs’, not small businesses and self-employed workers. They represent the small number of innovators bringing new ideas to the market, which are themselves only a small portion of new firms. Since self-employment can occur for many reasons, both positive and negative ones (e.g., lack of other employment options), insights from one group do not necessarily inform the other.

We perform the first causal study on what motivates innovative entrepreneurs. We carry out a series of field experiments which allows us to characterize the role of incentives on the commercialization of technology and the way it shapes the process through which new innovations enter society. We try to assess whether these motivations differ across gender and culture. We take advantage of a unique setting where we can observe innovative entrepreneurs all across the world considering the same task to help their startup: applying to various MIT entrepreneurship challenges.

The first experiment focuses on individuals likely to have interest in the MIT entrepreneurship challenges, this is a larger sample of broad range of entrepreneurship-oriented people. The second focuses on people who have already considered they are a good profile for these competitions by having registered but not yet completed the full application process. Our emails emphasize three different potential incentives that are thought to matter for entrepreneurs: income, social impact, and a control group merely highlighting the MIT brand. In our experiment we randomly send people emails with one of two treatments or the control.
Jointly with the emails we send to the mailing lists, we setup a web page for members of entrepreneurship communities to access on which we also randomize the incentive messages. We then track the behavior of these users on the website.

With these different treatments, we aim to analyze which priming condition engages the individuals more. Engagement will be measured by clicks and time spent. This will allow us a revealed preference way to measure the motivations of these entrepreneurs and why they might engage with these emails. We also aim to analyze these preferences across gender and regions.

Intervention

The type of entrepreneur the MIT challenge considers is precisely the type of entrepreneurs emphasized in economic theory—individuals with a new idea who are seeking to make it work in a new market. With the support of the Centers, we setup a few experiments that randomly emphasizes different messaging to individuals in the application process. The first experiment focuses on individuals likely to have interest in the MIT entrepreneurship challenge. In particular, we have access to different mailing lists which makes up our sample. We allocate different individuals in the mailing list into three categories: a control group and two treatment groups. This is a larger sample of broad range of entrepreneurship-oriented people. The second focuses on people who have already considered they are a good profile for the MIT challenges by registering for the competitions. Specifically, we take advantage of staggered deadlines where people can start the application by registering by a certain deadline but still have two weeks to complete the application. The existing data shows that only about 30 percent of the registrants finish their applications. This creates an opportunity for a second email to push them to apply, with randomization into one of three groups. Finally, we also partner with a few companies which promote the competition on their own website by embedding a link to the MIT entrepreneurship challenge website. We created this link such that if an individual clicks on it then he/she is randomly assigned to one of the three conditions as he/she lands on the website. We manipulate the content on the website such that it primes one of the three messages: control condition, money incentive or social incentive.

There has been a push to ensure greater participation of female entrepreneurs as well as entrepreneurs from different cultural backgrounds. In addition to looking at the main effects, we also want to see if there is any difference in behavior of entrepreneurs across gender and culture. This will help understand how preferences vary across these extremely important dimensions and how that might translate into eventual entrepreneurial behavior.

Experimental design

Our emails emphasize two different potential incentives that are thought to matter for entrepreneurs: income and social impact. In our experiment we randomly send people emails with one of our subject lines as follows.

1. Income Incentive: “Over $1 million in Prize Money! Apply to MIT’s Entrepreneurship Challenge!”
2. Social Impact Incentive: “Create Greater Shared Prosperity! Apply to MIT’s Entrepreneurship Challenge!”
3. Control Condition: "Apply to MIT’s Entrepreneurship Challenge!”

The email body is also amended to prime each of these messages in line with the conditions. We use similar messages when individuals visit the competition's website through the link provided to the partner companies.

Design details

Randomization method
Randomization was done in an office by a computer. Complete randomization was conducted using the "complete_ra" command in the R package "randomizr", with "set.seed" to make the randomization reproducible. In the case of the website link provided to our partners, the randomization will be done on the fly using a random number generator.
Randomization unit
The unit of randomization is the individual entrepreneur.
Clusters
30,000 individuals/entrepreneurs
Observations
30,000 individuals/entrepreneurs
Arms
10,000 individuals as control, 10,000 in the money treatment and 10,000 in the social incentives treatment.
Primary outcomes
The main outcomes of interest that we will focus on: (1) Open rate or clicks on emails (2) Time spent on website, number of pages visited (3) Registration or application rates

Cited by

No citation data available for this registration yet.

Keywords