Does risk aversion prevent new product adoption?
How to cite
Killeen, Grady. 2024. "Does risk aversion prevent new product adoption?." AEA RCT Registry. AEARCTR-0014812. https://doi.org/10.1257/rct.14812-1.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
Does risk aversion prevent firms from adopting new products? This study offers firms new motorcycle helmets with or without an insurance contract and examines whether insurance increases uptake. Treated shops are offered the ability to choose between a guaranteed future cash payment or a larger payment that pays out only if the shops struggles to sell helmets, calibrated so that the value of the contracts is similar in expectation. The control group receives the guaranteed payment. I test for risk aversion by examine whether being offered the insurance contract increases adoption. In a secondary component, the markets entered at baseline are randomized to test for information externalities at a follow-up by examining whether shops in markets with an entrant are more willing to purchase helmet stock.
Intervention
To test for risk aversion, we allow treated shops to choose between the following interventions:
The shop may opt to receive a guaranteed payment of p*X + e at the follow-up, regardless of whether they stock helmets, where p is the probability of failing to sell out a batch of helmets by midline if they opt to stock them. Or an insurance offer that pays X if the shops stocks helmets and fails to sell out, and 0 otherwise. The insurance offer is slightly less valuable in expectation, but hedges risk. Control shops receive the payment of p*X + e.
Second, I randomize which markets are entered at baseline, approaching 5 shops in each. This is designed to induce entrants to test for information spillovers at the follow-up. I plan to survey 5 new shops in treatment and control markets during the follow-up and offer them a batch of helmets to see if shops in treated markets are more willing to stock helmets.
Experimental design
Markets will be assigned to treatment or control with equal probability, stratified by County. There will be about 7-8 treated markets per County, with 9 western Counties surveyed. 5 shops will be surveyed at baseline in each treated market. We aim to have 70 treatment markets.
The randomization of the insurance contract will be done at the firm level, with equal probability of being assigned to treatment or control stratified by market. I will randomize if markets have 2 or 3 treated firms, stratifying this design by county. I aim to sample 350 shops.
Design details
Cited by
No citation data available for this registration yet.