The Behavioralist Goes Door-To-Door: Understanding Household Technological Diffusion Using a Theory-Driven Natural Field Experiment

How to cite

Price, Michael; Ponti, Giovanni; List, John; Jimenez-Gomez, David; Herberich, David; Giaccherini, Matilde. 2019. "The Behavioralist Goes Door-To-Door: Understanding Household Technological Diffusion Using a Theory-Driven Natural Field Experiment." AEA RCT Registry. AEARCTR-0004492. https://doi.org/10.1257/rct.4492-1.0

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

Michael Price* , Giovanni Ponti , John List , David Jimenez-Gomez , David Herberich , Matilde Giaccherini

* Primary investigator

Registry ID
AEARCTR-0004492
DOI
https://doi.org/10.1257/rct.4492-1.0
First registered
Last updated
Status
Completed
Trial dates
2010-01-01 – 2018-12-31
Countries
United States of America ()
Registry record
http://www.socialscienceregistry.org/trials/4492

Abstract

This paper uses a field experiment to estimate behavioral parameters in a structural model of residential adoption of technology. By embracing both economic and psychological factors, we are able to identify the role of prices, social norms, social pressure, and curiosity on the adoption decision. We find that prices and social norms influence the adoption decision along different margins, opening up the opportunity for economics and psychology to be strong complements in the diffusion process. In addition, welfare estimates from our structural model point to important household heterogeneities: whereas some consumers welcome the opportunity to purchase and learn about the new technology, for others the inconvenience and social pressure of the ask results in negative welfare. As a whole, our findings highlight that the design of optimal technological diffusion policies will require multiple instruments and a recognition of individual household heterogeneities.

Experimental design

Our field experiment was implemented by offering to households up to 2 packages of Compact Fluorescent Lightbulbs (CFLs) in a door-to-door design (each package contained 4 CFLs). The sample approached in 11 suburbs of Chicago (Arlington Heights, Elmwood Park, Evanston, Lemont, Libertyville, Oak Park, Glen Ellyn, River Forest, Roselle, Skokie, and Wheaton) had higher median incomes than in the U.S. average. Households were approached during a day during weekends by University of Chicago students employed through the Becker Center at the University of Chicago. Students were hired upon responding to job advertisements placed around the campus of the University of Chicago and on the University's main electronic help wanted web site (marketplace.uchicago.edu). Students spent Saturdays and Sundays to approach households over 4 one-hour blocks of time: 10 a.m. to 11 a.m., 11 a.m. to noon, 1 p.m. to 2 p.m. and 2 p.m. to 3 p.m. Households were grouped into blocks containing roughly 25 houses. Each block of houses was randomly assigned to a combination of treatments.
• In total, there are 18 treatment cells in the experimental design, such as warning treatments (no flyer, warning, opt-out), social norm treatments (Neutral Frame, social norm low, social norm high) and two different prices (high price and low price). A team of researchers and interns placed flyers on households in the Warning and Opt-Out treatments the day prior to the students visiting households. These flyers informed the household that it would be visited the following day by someone with an “offer for purchase and discuss energy saving light bulb options”. Subjects in the Opt-Out condition had the possibility of sorting out by checking the appropriate box in the flyer. The student (i.e. salesperson) used a script to tempt people to buy the CFLs. The scripts varied in two ways: stated social norm and sale price of CFL package. We set our baseline price for a package of 4 CFLs at $5.00 and we included one other price point, a low price of $1.00. In order to incorporate social norms, a script without a statement of social norms (Neutral Frame treatment), and two treatments involving a statement of social norms, “Social Norm Low” and ``Social Norm High”. Social Norm Low contained the phrase: “For instance, did you know that 70% of U.S. households own at least one CFL?” while the Social Norm High treatment containde the phrase: “For instance, did you know that 70% of the households that we surveyed in this area own at least one CFL?”

Design details

Randomization method
Randomization done in office by a computer
Randomization unit
Households were grouped into blocks containing roughly 25 houses. Each block of houses was randomly assigned to a treatment (18 treatments in total). For the entire experiment, a total of 8815 households were approached.
Clusters
164 household blocks
Observations
8815 households
Arms
Number of households approached for each treatment group: $1, No Social Norm, No Warning: 480 $1, No Social Norm, Warning: 474 $1, No Social Norm, Opt-Out: 473 $1, Low Social Norm, No Warning: 447 $1, Low Social Norm, Warning: 508 $1, Low Social Norm, Opt-Out: 535 $1, High Social Norm, No Warning: 454 $1, High Social Norm, Warning: 469 $1, High Social Norm, Opt-Out: 481 $5, No Social Norm, No Warning: 435 $5, No Social Norm, Warning: 546 $5, No Social Norm, Opt-Out: 501 $5, Low Social Norm, No Warning: 493 $5, Low Social Norm, Warning: 544 $5, Low Social Norm, Opt-Out: 491 $5, High Social Norm, No Warning: 431 $5, High Social Norm, Warning: 511 $5, High Social Norm, Opt-Out: 542
Primary outcomes
Decision to interact with Solicitor, Purchasing Quantity

Cited by

No citation data available for this registration yet.

Keywords