Assessing the Effectiveness of Financial Coaching: Evidence from the Boston Youth Credit Building Initiative

• Published date: 01 December 2019
• Author: Alicia S. Modestino,Rachel Sederberg,Liana Tuller
• Date: 01 December 2019
• DOI: http://doi.org/10.1111/joca.12265

WINTER 2019 VOLUME 53, NUMBER 4 1825

ALICIA S. MODESTINO , RACHEL SEDERBERG, AND

LIANA TULLER

Assessing the Effectiveness of Financial Coaching:

Evidence from the Boston Youth Credit

Building Initiative

Since the 2008 nancial crisis, there has been renewed interest in

providing nancial education to improve consumer nancial decision

making, especially among youth. Using a randomized controlled trial,

we estimate the causal effects of a nancial coaching program for young

adults from linked individual-level administrative credit reports and

self-reported survey responses. Withinsix months, the treatment group

was 10 percentage points more likely to have access to credit compared

to the control group. After 18months, the average credit score was 26

points higher for the treatment group versus the control group,raising

the likelihood of achievinga “good” credit rating by 8 percentage points.

Consequently, the treatment groupwas less likely to rely on alternative

nancial services and paid lower interest rates on car loans. These

impacts are largely driven by improvements in self-efcacy, offering

important insights for policymakers seeking to incorporate nancial

education into youth workforce developmentprograms.

Since the 2008 nancial crisis, there has been renewed interest in identi-

fying which individuals are at greatest risk for experiencing poor nancial

outcomes and in providing nancial education programs to improve their

nancial decision making. With the creation of the Consumer Financial

Protection Bureau (CFPB) in 2011, the federal government now spends

an additional $44 million per year on nancial education (CFPB 2017).

Moreover, special emphasis has been placed on developing nancial skills

Alicia S. Modestino (a.modestino@northeastern.edu) is an associate professor at Northeastern Uni-

versity in the School of Public Policyand Urban Affairs and the associate director of the Dukakis Center

for Urban and Regional Policy. Rachel Sederberg(sederberg.r@husky.neu.edu) is a PhD candidate in

the Economics Department at Northeastern University.Liana Tuller (tuller.l@husky.neu.edu) is a PhD

candidate in the Department of Sociology & Anthropology at Northeastern University. The authors

thank Constance Martin of the Boston Mayor’sOfce of Financial Empowerment at the City of Boston

for her guidance and support as well as Ricki Lowitz of WorkingCredit NFP (Not-for-Prot) for sup-

plying the data and providing insights regarding the collection methodology. We are also grateful to

Citi Community Development for their support of this work. The randomized control trial on which

this research was based has been registered with the American Economic Association RCT Registry

(#AEARCTR-0003011).

The Journal of Consumer Affairs, Winter 2019: 1825–1873

DOI: 10.1111/joca.12265

Copyright 2019 by The American Council on Consumer Interests

1826 THE JOURNAL OF CONSUMER AFFAIRS

among youth. At the federal level, nancial literacy became a required

element for youth workforce development programs under the 2014 Work-

force Innovation and Opportunity Act (WIOA). At the local level, the

number of states requiring a personal nancial education course in high

school increased from 7 to 17 over the past decade (Council for Economic

Education 2016).

However, prior studies evaluating the impact of nancial education

programs on outcomes have not produced consistent results, particularly

for programs aimed at youth, making their cost-effectiveness uncertain

at best (Brown et al. 2016; Cole, Paulson, and Shastry 2016; Hastings,

Madrian, and Skimmyhorn 2013; Lyons et al. 2006; McCormick 2009;

Walstad et al. 2017; Willis 2011). Approaches that focus on developing

nancial capability among youth by combining nancial literacy with

access to nancial products appear to hold greater promise (Johnson and

Sherraden 2007; Loke, Choi, and Libby 2015). Yetresearchers have found

that the effects of such programs are small (Fernandes, Lynch Jr., and

Netemeyer 2014) or affect only some outcomes (e.g., savings behavior)

but not others (e.g., reducing defaults) (Miller et al. 2014).

In response, policymakers have increasingly turned to nancial coach-

ing as a way to improve consumer nancial decision making by incor-

porating nancial knowledge, access to nancial products and services,

and one-on-one coach-client interactions (Collins, Baker, and Gorey2007).

One of the largest such initiatives, the Working Families Success Network,

currently includes 115 sites in over 30 cities offering a range of services to

low- and moderate-income adults. Financial coaching is an ongoing pro-

cess that involves setting goals, establishing a concrete plan of action, and

monitoring individual progress—all to change nancial habits and improve

long-term outcomes (Collins and O’Rourke 2012). Yetto date, only a hand-

ful of studies have assessed whether nancial coaching improves economic

knowledge, decision making, or well-being. Most studies fail to address

issues of positive selection among individuals who choose to enroll in such

programs, and few attempts to identify which program mechanisms may

be important for improving outcomes.

In this article, we estimate the causal effects of a one-year nancial

coaching program for young adults, the Boston Youth Credit Building

Initiative (BYCBI), on a range of nancial and behavioral outcomes.

Using linked individual-level data from administrative credit reports

and self-reported survey responses, we measure gains in credit scores

and ratings as well as improvements in nancial knowledge, behaviors,

self-efcacy, and well-being. A randomized controlled trial provides a

robust control group to assess program impacts, with stratication across

WINTER 2019 VOLUME 53, NUMBER 4 1827

demographic characteristics and recruitment sites to explore heterogeneity

in outcomes.

We nd that the BYCBI helps build credit by increasing access as well

as improving scores. Within the rst six months of the program, individuals

in the treatment group were 10 percentage points more likely to have

established a credit score compared to the control group. By the end of the

18-month observation window, the average credit score among individuals

with a credit le prior to the start of the program was 26 points higher

for the treatment group relative to the control group, raising the likelihood

of achieving a “good” credit rating by 8 percentage points. These effects

are even larger among those who complied with the program. We further

show that these improvements reect signicant changes in the underlying

factors that typically affect one’s credit score, such as having a history of

sustained on-time payments, and a lower likelihood of adverse events such

as delinquencies and collections. Finally, we demonstrate that the program

has meaningful impacts on individuals beyond improving their credit

scores, such as reducing their reliance on alternative nancial services

and lowering the interest rate they pay on car loans—outcomes that are

of interest to policymakers hoping to improve the nancial well-being of

low- and moderate-income groups.

Our ndings offer several policy-relevant lessons for future program

design. First, the subgroup analysis shows that the BYCBI has greater

impacts on younger individuals and African Americans, suggesting ways

that cities with limited resources may want to target similar programs. Sec-

ond, our mediation analysis demonstrates that much of the impact of the

program appears to be driven by improvementsin nancial self-efcacy—a

nding conrmed by focus group discussions—which may have been

the missing ingredient in prior nancial education programs. Finally,

the path toward better credit evolved over time and across different

dimensions—even after the program ended—suggesting that the program

may affect the behavior of the treatment group beyondthe short term. These

ndings are encouraging and offer important insights for cities and states

seeking effective ways to incorporate nancial education into youth work-

force development programs to comply with the new WIOA requirements.

RELEVANT LITERATURE

While the general consensus is that nancial education should have

a positive effect, the ndings from prior studies have been mixed,

including those aimed at youth, such that it has been difcult to assess

their cost-effectiveness (Brown et al. 2016; Hastings, Madrian, and