The Impact of Automated Reminders on Credit Outcomes: Results from an Experimental Pilot Program

• Published date: 01 December 2019
• Author: Stephen P. Roll,Stephanie Moulton
• Date: 01 December 2019
• DOI: http://doi.org/10.1111/joca.12252

WINTER 2019 VOLUME 53, NUMBER 4 1693

STEPHEN P. ROLL AND STEPHANIE MOULTON

The Impact of Automated Reminders on Credit

Outcomes: Results from an Experimental Pilot Program

This study evaluates the impact of an automated reminder program

designed to help credit counseling consumers manage their payment

obligations and nancial goals. Credit counseling consumers were ran-

domly assigned to receive reminders linked to their nancial goals and

payment obligations for one year after an initial credit counseling ses-

sion. We nd that consumers offered reminders were 21% less likely

than the control to experience severe (60+day) payment delinquen-

cies and were 12% less likely to experience a 30+day delinquency.

At the same time, these consumers saw a 10.5 point increase in credit

scores relative to the control group. There were no signicant impacts

on total credit card balances or installment debt levels. This analysis

provides promising evidence that automated reminders can providean

important complement to traditional credit counseling when it comes to

improving consumers’ credit proles and does so at a fairly low cost.

Credit is an important component of the nancial lives of consumers. For

U.S. households, revolving credit like that available through credit cards

comprises approximately two-thirds of the funds available for short-term

consumption in a given period (Fulford 2015). Research indicates that 40%

to 60% of U.S. consumers do not pay their credit card balances in full

each month, resulting in the accrual of interest on unpaid balances that can

place strain on household budgets (Consumer Financial Protection Bureau

2015; Laibson, Repetto, and Tobacman 2000). Missed or late payments

incur additional fees and charges that further increase revolving debt

balances and lower credit scores, reducing consumers’ ability to borrow

and increasing the cost of borrowing in future periods (Federal Trade

Commission 2013).

Stephen P. Roll (stephen.roll@wustl.edu) is a Research Assistant Professor at the Social Policy

Institute, Washington University in St. Louis. Stephanie Moulton (moulton.23@osu.edu) is an Asso-

ciate Professor at the John Glenn College of Public Affairs,The Ohio State University. This manuscript

has been prepared as part of an evaluation of the Sharpen YourFinancial Focus program, which was

jointly conducted by The Ohio State University and the National Foundation for Credit Counseling.

The Sharpen YourFinancial Focus program and this research were funded by Bank of America, Chase

Bank, Synchrony Financial, and Wells Fargo,in conjunction with the National Foundation for Credit

Counseling. The credit data in this evaluation were provided by Experian.

The Journal of Consumer Affairs, Winter 2019: 1693–1724

DOI: 10.1111/joca.12252

Copyright 2019 by The American Council on Consumer Interests

1694 THE JOURNAL OF CONSUMER AFFAIRS

Beginning in the 1950s, nonprot credit counseling agencies emerged

to help consumers manage their debt, improve their credit proles, and

serve as an alternative to bankruptcy (Hunt 2005). These agencies reach

a large number of consumers, providing services to between 1.5 and two

million consumers in 2013 and 2014; at the height of the recession they

serviced around four million consumers in a single year (Keating 2012;

National Foundation for Credit Counseling 2015). Consumers served

by these agencies tend to have blemished credit histories, with more

than $10,000 in revolving credit card debt and an average credit score

below 600 (Elliehausen, Christopher Lundquist, and Staten 2007; Roll and

Moulton 2016). Debt for these consumers is particularly costly. Analysis of

credit record data from the Consumer Financial Protection Bureau (2015)

indicates that, while consumers with subprime credit scores (below 620)

account for only 6% of new consumer spending on credit cards, their

outstanding balances comprise 19% of all outstanding revolving debt and

46% of fees assessed to consumers.

Is credit counseling an effective strategy to improve credit outcomes

for these distressed consumers? Studies tend to nd reductions in con-

sumer debt amounts but more limited evidence of improvement in credit

scores and payment delinquencies (Dobbie and Song 2016; Elliehausen,

Christopher Lundquist, and Staten 2007; Roll and Moulton 2016). Limita-

tions of prior studies include (1) lack of specicity regarding the mech-

anisms by which counseling affects credit outcomes and (2) difculty

accounting for consumer selection into counseling services. First, credit

counseling includes multiple interventions such as structural debt renego-

tiation through a debt management plan (DMP), as well as the provision

of more general nancial counseling and goal setting, which may or may

not include follow-up after the initial session. These interventions likely

affect credit outcomes in different ways. Second, consumers who select

to participate in counseling may be different from those who do not seek

counseling on unmeasured attributes that correlate with credit outcomes,

thereby biasing estimates.

This study overcomes these limitations by evaluating the impact of a

specic intervention—automated goal and payment reminders—through

a randomized study design. We expect reminders to reduce payment

delinquencies and improve overall credit quality by increasing consumer

attention to payment due dates and nancial goals set during the coun-

seling session. We leverage a eld experiment administered as part of

the National Foundation for Credit Counseling’s (NFCC®) Sharpen Your

Financial Focus®(Sharpen) initiative in 2014 and 2015. As part of the

credit counseling session, a random subset of consumers received an offer

WINTER 2019 VOLUME 53, NUMBER 4 1695

to participate in a reminder program. Those who enrolled received cus-

tomized, automated reminders for up to one year after the initial counseling

session. Consumers could select to receive reminders tied to nancial goals

or reminders tied to debt payments. Thirty-four percent of the consumers

offered reminders enrolled in the program, with over 80% specifying goal

reminders and almost 20% specifying payment reminders. On average,

enrolled consumers received 2.7 reminders over the course of a year.

Using administrative data collected by the NFCC merged with credit

attribute data provided by Experian, we compare credit outcomes for

consumers in the offer group to a control group of consumers not offered

enrollment (intent-to-treat, ITT). To estimate the impact of the reminders

on those who enrolled (treatment on the treated, TOT), we use a two-stage

regression model to rst predict the selection of a consumer into the

reminder program and then estimate the program’s effect on consumer

outcomes after accounting for this selection bias.

We nd that goal and payment reminders have signicant effects on

consumer credit scores and the propensity to have both minor (30+day)

and more severe (60+day) payment delinquencies over the evaluation

period. Consumers offered reminders were six percentage points less likely

to have a severe payment delinquency (p<.01) at any point over the

treatment period and were three percentage points less likely to have

a minor delinquency (p<.01). These impacts translate into reductions

in delinquency rates of 21% and 12%, respectively. At the same time,

these consumers saw a 10.5 point increase in credit scores relative to the

control group (p<.01). We nd no signicant effectsof reminders on total

revolving or installment debt balances one year after the offerof reminders.

The two-stage analysis shows an even greater effect of the TOT; those who

enrolled in the reminders were 14 percentage points—or 56%—less likely

to have any severe delinquency than the control group (p<.01), and their

credit scores increased by 26 points relative to the control (p<.01).

The contributions of our study are twofold. First, we contribute to

the literature on the effectiveness of nancial counseling interventions

for improving consumer nancial outcomes. While this study takes place

within the context of credit counseling, the ndings have implications for

nancial counseling and interventions intended to increase the nancial

capability of consumers more broadly. Researchers increasingly recognize

the need to disentangle the various mechanisms that underlie interven-

tions such as credit or nancial counseling (Collins and O’Rourke 2010;

Peeters et al. 2018). Studies tend to nd that more personalized approaches

with embedded follow-up have positive outcomes on longer term con-

sumer behaviors such as payment delinquencies or debt levels (Collins and