Collection of Delinquent Fines: Assessing the Effectiveness of Alternative Text Messages in the United Kingdom
How to cite
Green, Donald; Torgerson, David J.; John, Peter; Gallagher, Rory; Haynes, Laura. 2014. "Collection of Delinquent Fines: Assessing the Effectiveness of Alternative Text Messages in the United Kingdom." AEA RCT Registry. AEARCTR-0000253. https://doi.org/10.1257/rct.253-1.0
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
This paper reports the results of a large randomized controlled trial, led by the UK Cabinet Office’s Behavioural Insights Team, which was designed to test the effectiveness of mobile phone text messaging as an alternative method of inducing people to pay their outstanding fines. An adaptive trial design was used, first to test the effectiveness of text messaging against no treatment and then to test the relative effectiveness of alternative messages. Text messages, which are relatively inexpensive, are found to significantly increase average payment of delinquent fines. We found text messages to be especially effective when they address the recipient by name.
Intervention
On the Monday after their cases escalated to distress warrant status, HMCTS sent in- dividuals in the text message treatment groups a text (short message service [SMS]) to their mobile phones. These messages alert the owner that the message has arrived; the owner presses a button to see the text on the screen of the phone. Recipients of the text message viewed the sender as “HMCTS.” In the event that the message was undeliverable, the sender received a notification, which we used to classify text messages as “delivered” or “not delivered.”
As shown in Table 1, each message conveyed the same core information. Recipi- ents were reminded about their unpaid fines, warned that failure to pay would result in a warrant, instructed to call a payment hotline number, and given the reference identification number. The experimental variations on the standard treatment (see Table 1) were the personalized (PERSONAL) condition, in which the message was preceded by the recipient’s name; the personalized amount (AMOUNT) condition, in which the recipient was reminded of the total value of the outstanding fine; and the personalized name and amount (PERSONAL/AMOUNT) condition, which included both of these elements.
Experimental design
Our experimental design consisted of two phases: In Phase 1, we tested a series of alternative text treatments against a control condition in which no text was sent. The null hypothesis is that text messages fail to increase the payment of delinquent fines, which implies the use of one-tailed tests. In terms of point estimation, this phase of the experiment allowed us to gauge the overall effectiveness of text messaging and to assess tentatively the relative effectiveness of different types of messages. In Phase 2, the NO TEXT group was eliminated from the design, and the aim was to sharpen our estimates of the relative effectiveness of alternative messages. Two-sided tests are used to test the null hypothesis that subjects respond to alternative messages in the same way.
Design details
Cited by
No citation data available for this registration yet.