Thread by @KieranJGibson

I’m excited to share my job market paper for the 2025-2026 market!

In a real-effort experiment, participants working solely for money completed 23% more tasks than those offered the same monetary reward plus a charitable incentive (p < 0.02) 🤯

Thread below (1/)

The experiment design is as easy as 1, 2, 3!

Workers learn how to complete tasks, receive an incentive scheme, and then work for as long as they wish.

I recruited 3,412 participants on MTurk (via CloudResearch) across three pre-registered studies. Study 1 👇

(2/)

In the joint-incentive condition, participants earned the same money for themselves and for the charity as the other groups.

Despite having more incentives, the ECDFs show at every percentile, those working only for themselves completed the most tasks.

(3/)

Is this because people hate working for the charity?!

Nope. Even those who reported high levels of support for the charity exerted less effort.

In a follow-up study, I also show that the same charitable incentive increases effort relative to a no-incentive control 🤷🏻‍♂️

(4/)

This result supports the predictions of the framework I put forward in the paper (pre-registered 🤓)

It’s called “competing motivations”, where motivations interfere with each other instead of purely adding together.

Simplified model below.

(5/)

I extend the investigation of competing motivations by testing more incentives.

Study 2 replaces the charitable incentive with a lottery (a small chance of winning a gift card for every task), and also tests a higher monetary piece rate.

(6/)

Combing two extrinsic incentives did not increase effort! This is more evidence of interference and contrary to the additive approach to motivation.

In a follow up study, I also show that the same lottery incentive increases effort relative to a no-incentive control.

(7/)

Is this because participants reached a ceiling of effort?

Nope! More of the *same* incentive increased effort, but more of a different incentive did not.

This also rules out utility targeting as a possible mechanism for competing motivation…

(8/)

Participants in Self+Lottery who rated the gift card store relatively low exerted less effort compared to the average Self baseline.

This explores heterogenous preferences: the weaker the lottery incentive, the stronger the crowding out.

(9/)

Thanks for reading this far!

My JMP is about competing motivations, and I demonstrate crowding out in a real-effort experiment

A key policy takeaway: focus on the most compelling incentive instead of offering multiple reasons that can distract from what matters most.

(10/)

More information about me and my JMP on my academic site:


Some of my academic heroes who inspired this work @UriGneezy @Econ_4_Everyone @alexolegimas @sdellavi @Devin_G_Pope 🫡

(11/)kierangibson.net

@UriGneezy @Econ_4_Everyone @alexolegimas @sdellavi @Devin_G_Pope @threadreaderapp unroll