Establishing Properties of the Control Premium

How to cite

McGranaghan, Christina. 2019. "Establishing Properties of the Control Premium." AEA RCT Registry. AEARCTR-0004319. https://doi.org/10.1257/rct.4319-1.0

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

Christina McGranaghan*

* Primary investigator

Registry ID
AEARCTR-0004319
DOI
https://doi.org/10.1257/rct.4319-1.0
First registered
Last updated
Status
Completed
Trial dates
2019-06-19 – 2019-06-21
Countries
United States of America ()
JEL codes
C91, D12, Q40
Registry record
http://www.socialscienceregistry.org/trials/4319

Abstract

We propose a lab experiment to establish properties of the control premium in the context of a real effort task. Using a within-subject design, we test for the existence of a control premium as well as its sensitivity to the expected number of control events and the relevant stakes. This experiment also uses between-subject variation to determine whether participants exhibit an endowment effect with respect to control. In order to disentangle the control premium from probability weighting and the certainty effect, we compare participants' valuations of the control lottery to a lottery with similar stakes that does not feature a control component. Our framework has important implications for the design of Direct Load Control contracts in electricity.

Experimental design

Participants will be given the opportunity to pay in order to upgrade to having 60 instead of 30 seconds available to complete a real effort task. We first elicit participants' willingness to pay for this upgrade, and then offer them a series of lotteries which will determine the cost at which participants can choose to upgrade, but in some instances will result in participants not being able to upgrade at all. We will compare participants' stated indifference points for lotteries involving a control component to an estimated implied indifference point based on their reported willingness to pay for 60 seconds assuming zero value of control. The control premium will be defined as the difference between stated indifference points and their estimated control-neutral counterparts. We will then vary the stakes and odds of the offered lotteries to establish properties of the control premium.

Design details

Randomization method
Randomization done by the computer for the online survey
Randomization unit
Individual
Clusters
200 lab participants
Observations
200 lab participants
Arms
100 lab participants per treatment arm for the two out of ten questions in which subjects are endowed with an initial lottery
Primary outcomes
Estimated control premium for different control lotteries (base lottery, doubled probability of control event, doubled stakes)

Cited by

No citation data available for this registration yet.

Keywords