Interactions between Exponential Growth Bias and Procrastination: Evidence from Retirement Savings

How to cite

Goda, Gopi; Levy, Matthew; Tasoff, Joshua; Sojourner, Aaron; Manchester, Colleen. 2014. "Interactions between Exponential Growth Bias and Procrastination: Evidence from Retirement Savings." AEA RCT Registry. AEARCTR-0000314. https://doi.org/10.1257/rct.314-3.0

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

Gopi Goda* , Matthew Levy , Joshua Tasoff , Aaron Sojourner , Colleen Manchester

* Primary investigator

Registry ID
AEARCTR-0000314
DOI
https://doi.org/10.1257/rct.314-3.0
First registered
Last updated
Status
In development
Trial dates
2014-08-01 – 2015-01-15
Countries
Private
JEL codes
D14, J32, J26, J18, D03
Registry record
http://www.socialscienceregistry.org/trials/314

Abstract

Theory predicts that exponential-growth (EG) bias, the tendency to neglect compounding interest, and present-biased preferences, the tendency for people to underweight future consumption, will both lead to suboptimal saving decisions. However, there is a lack of empirical research on the incidence and interaction of these biases in predicting saving behavior. This study seeks to fill this gap in the following ways. First, we will measure the joint distribution of exponential-growth bias and present bias using a representative sample of the U.S. population, and correlate these measures with a variety of demographic characteristics and economic outcomes. Second, we will test a set of interventions designed to counteract present bias and EG bias on hypothetical saving decisions. We will give particular attention to the heterogeneity in treatment effects with respect to individuals' EG bias and present bias. Finally, by assuming more structure we can conduct a counterfactual analysis, estimating retirement savings if all bias were eliminated. These results will provide evidence on whether people make biased decisions regarding their retirement savings, the mechanism of the bias, and how best to design policy to improve decisions.

Intervention

Our interventions consist of two components: an informational intervention that relays current contributions in terms of future balances and future income, and an incentive to complete enrollment or contribution change paperwork in a specified time interval.

Experimental design

In this research project, we have four main goals:
1. Measure levels of present bias and exponential-growth bias in a nationally representative population.
2. Administer different kinds of interventions designed to counteract exponential-growth bias and present bias in a randomized control trial
3. Measure the effects of our interventions on hypothetical contributions into a tax-deferred retirement savings account
4. Determine whether the interventions targeted to present bias and exponential growth bias had different effects on individuals with different levels of these biases

Design details

Randomization method
Randomization done by ALP administrators.
Randomization unit
The unit of randomization is the individual survey respondent.
Clusters
N/A
Observations
3000, budget permitting
Arms
Approximately 300 subjects in each of the 9 following treatment arms: EGB Control/PB Control EGB Control/PB Limited EGB Control/PB Unlimited Balance/PB Control Balance/PB Limited Balance/PB Unlimited Income/PB Control Income/PB Limited Income/PB Unlimited
Primary outcomes
Our key outcome variable is hypothetical contributions into a tax-deferred retirement savings account

Cited by

No citation data available for this registration yet.

Keywords