Too Big to Innovate? Theory and Experiment

Last registered on August 31, 2025

View Trial History

Pre-Trial

Trial Information

General Information

Title

Too Big to Innovate? Theory and Experiment

RCT ID

AEARCTR-0014778

Initial registration date

November 07, 2024

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

November 15, 2024, 1:40 PM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated

August 31, 2025, 7:24 AM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Country

China

Region

Shanghai

Primary Investigator

Name

Mike Zhiren Wu

Affiliation

Monash University

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Xiangdong Qin

PI Affiliation

Shanghai Jiao Tong University

Contact Investigator

PI Name

Shicheng Jiang

PI Affiliation

East China University of Science and Technology

Contact Investigator

Additional Trial Information

Status

On going

Start date

2021-12-04

End date

2025-09-30

Keywords

Behavior, Firms & Productivity, Lab

Additional Keywords

Firm Innovation;

JEL code(s)

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

The evidence on whether small or large firms hold an advantage in innovation remains mixed. This paper revisits this question by combining lab experiments with theory to explain the "firm size effect" on innovation. We consider a setting where the principal chooses between a rigid, risk-free project, providing all agents with a uniform payoff, and a flexible, riskier project (with an option to revert to the risk-free project after observing outcomes), which offers potentially higher payoffs for each agent with a certain probability. Some agents (who fail the risky project or suffer a loss due to the principal) may perceive the principal as unfair, leading them to engage in costly punishment or "shading", directed at the principal and other agents. This act cannot be contractually prohibited ex-ante. We hypothesize that shading is more likely to occur in larger firms, generating more resistance to innovation. To reduce the potential shading, we test whether screening out agents with strong shading motives ex ante or team building activities can foster firm innovation.

External Link(s)

Registration Citation

Citation

Jiang, Shicheng, Xiangdong Qin and Mike Zhiren Wu. 2025. "Too Big to Innovate? Theory and Experiment." AEA RCT Registry. August 31. https://doi.org/10.1257/rct.14778-2.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

Intervention (Hidden)

Intervention Start Date

2021-12-04

Intervention End Date

2025-09-30

Primary Outcomes

Primary Outcomes (end points)

project carried out by the principal - initial choice (rigid, risk-free project / flexible, risky project)
adjustment made by the principal following the outcome of the risky project
ex-post punishment decision ("shading") of the agents

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

Before the main experiment, subjects complete a real-effort task to earn initial endowments (200 ECUs) to ensure they have non-negative payoffs.

The main experiment includes three different treatments, each with a different group size (N=4, 6, or 8). Each group consists of one principal (P) and multiple agents (A), with roles randomly assigned at the start and maintained throughout the experiment.

The timeline is as follows: At the beginning each round, the principal must choose between a risk-free project (project 1) and a risky project (project 2). If the project 1 is chosen, both the principal and the agents can secure a payoff of 100 experiment currency units (ECUs, henceforth). If the project 2 is chosen, with two-thirds of the probability, an agent may succeed in the project and gets a payoff of 200 ECUs. With one-third of the probability, the agent may failure in the project and gets nothing. The principal’s payoff will be the average payoff of the agents. After checking the performance of each agent, the principal can decide whether to revert to the risk-free option by switching back to the project 1. Regardless of the choice of the principal (whether it is project 1/project 2/ project 2 → project 1), each agent is given the opportunity to impose costly punishment on others (“shading”). An agent can spend 5 ECUs to reduce the payoff of the principal or all other agents (with positive payoff) by 50 ECUs, or spend 10 ECUs to reduce the payoff of both the principal and all other agents (with positive payoff) by 50 ECUs. The game is repeated for 10 rounds and only one round is randomly selected to calculate the final payoff.

In addition to examining whether larger firms may face greater challenges in pursuing risky investment or innovation, we also explore mechanisms to address these challenges. To achieve this, we run two additional treatments (group size = 6), each representing a different team management strategy. In the "screening" treatment, a simple game is used to differentiate agents by testing whether they are willing to bear costs to reduce income inequality when treated unfairly. Based on this game, groups are formed by pairing a randomly selected principal with five agents with weak shading motives. This allows us to test whether recruiting agents with weaker shading motives enhances firm innovation. In the "team building" treatment, we adapt a team building task from Jiang and Li (2019). In the randomly formed group with N=6, the principal and the five agents need to exchange their private information via group chat to solve a puzzle together, with the principal taking on the main responsibility in the solving the puzzle. Each group member receives a small reward if the puzzle is solved within the given time. We are interested in whether strengthened closeness between the principal and the agents promotes firm innovation.

Experimental Design Details

Randomization Method

Randomization done by the computer. We randomly assign students into different treatments.

Randomization Unit

individual level randomization

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

Around 220 groups in total

Sample size: planned number of observations

around 1320 subjects (university students)

Sample size (or number of clusters) by treatment arms

Around 40 groups for treatment with group size = 4, 6, 8
Around 60 groups for treatment with group size = 6 (with type classification)
Around 40 groups for treatment with group size = 6 (with team building)

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Supporting Documents and Materials

IRB