Supporting Statement A_30-day_10.14.2015_FINAL

BUREAU OF CONSUMER FINANCIAL PROTECTION
PAPERWORK REDUCTION ACT SUBMISSION
INFORMATION COLLECTION REQUEST

SUPPORTING STATEMENT PART A
OWNING A HOME STUDY
(OMB CONTROL NUMBER: 3170-XXXX)

TERMS OF CLEARANCE: Not applicable. This is a new information collection request.
ABSTRACT: The Dodd Frank Act directs the Consumer Financial Protection Bureau (CFPB)
to develop a program of consumer education and engagement. As part of that program, the
CFPB has developed a suite of online tools and resources, known as the Owning a Home project,
to help consumers make better, more informed decisions about mortgages. The purpose of this
information collection is to learn about the behavioral mechanisms and evaluate the hypotheses
underlying the Owning a Home project. This information collection is structured as a
randomized-controlled trial field study.

A. JUSTIFICATION
1. Circumstances Necessitating the Data Collection
The Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 (Pub. L. 111-203)
directs the Consumer Financial Protection Bureau (CFPB) to develop and implement “initiatives
intended to educate and empower consumers to make better informed decisions” (12 U.S.C. §
5493(d)). As part of fulfilling this mission, the CFPB has developed a suite of online tools and
resources to help consumers make better, more informed decisions about mortgages.
Collectively, these tools and resources are known as the Owning a Home project.
In order to ensure that the CFPB’s tools are as effective as possible, and to ensure that resources
are well-spent, the CFPB must study the impact of the Owning a Home project. The CFPB seeks
to gain insights as to (1) general consumer behavior, attitudes, and approaches to obtaining a
mortgage and mortgage shopping; (2) the validity of certain hypotheses underlying the
development of the Owning a Home project; (3) the mechanisms by which the Owning a Home
project impacts consumer behavior and attitudes; (4) which consumer segments or profiles
benefit the most from the Owning a Home project; and (5) the degree to which the Owning a
Home project impacts consumer behaviors, attitudes, and approaches to obtaining a mortgage.
To answer these questions, the CFPB intends to conduct a field study. We will recruit
prospective homebuyers and assign them to one of three study groups: those exposed to the
Page 1 of 17

Owning a Home tools and resources (the first treatment group), those who receive
encouragement to comparison shop for mortgage offers (the second treatment group), and those
not exposed to any additional resources (the control group). We will survey these groups of
homebuyers as they go through the process of shopping for and acquiring a mortgage, and will
compare the groups’ attitudes, behaviors, and outcomes.

2. Use of the Information

Use of the information
This information will be used to inform CFPB’s future work on the Owning a Home project. It
will also inform – at a high level – CFPB’s overall approach to consumer education and
engagement. As described in more detail in section 4, the Owning a Home project is premised
on two key hypotheses: (1) that comparison shopping in the mortgage market leads to better
consumer outcomes; and (2) that consumers often do not shop for a mortgage because shopping
in the mortgage market is difficult, confusing, and generally discouraged by the market structure.
The Owning a Home project seeks to remedy this situation by (1) encouraging consumers to
shop for mortgages and (2) supporting and empowering them with user-friendly tools and
resources designed to make shopping easier and give consumers greater confidence when
shopping. In order to explore the first hypothesis further, a second treatment group will be
encouraged to comparison shop for mortgage offers, but will not be exposed to the Owning a
Home suite.
This study will inform CFPB’s future work on Owning a Home as well as its overall approach to
consumer education and engagement in several ways. First, the study will help us understand to
what extent – and for what specific consumer segments or profiles – the above hypotheses are
true. There is some support in existing literature and much anecdotal wisdom to support these
hypotheses, but little solid research. This study will contribute significantly to the baseline
knowledge on which the CFPB and other governmental agencies and community intermediaries
operate.
Second, the study will help us understand more about how the market and behavioral forces
implicit in these hypotheses work in practice, which will help inform both the Owning a Home
project and related programs the CFPB might develop in the future. Third, the study will help us
understand to what extent, in what specific ways, and for what specific consumer segments or
profiles the Owning a Home suite is effective at encouraging consumers to shop and mitigating
the market conditions and behavioral forces that discourage shopping. This nuanced information
will help the CFPB refine and improve the Owning a Home suite over time.
The CFPB intends to publish aggregate and overall findings from the study so that other
researchers and agencies may benefit from the work, but it will not make raw data available to
other agencies or parties. This is a one-time research study. There are no plans to collect this
information on an on-going basis.
Page 2 of 17

Collection of the information
The target audience for the Owning a Home project is prospective homebuyers. Thus, the study
will attempt to replicate this target audience by recruiting respondents from among prospective
homebuyers. Potential respondents will be asked to complete a short screening questionnaire in
order to determine if they are eligible. Data obtained in the screening questionnaire does not
contain any directly identifying information and will be analyzed only in the aggregate to assess
which eligibility criteria have the most impact on disqualifying individuals. Upon completing
the screening questionnaire, respondents are presented with informed consent language and
asked to opt-in to the study.
Respondents will provide information in three primary ways. First, respondents are asked to
complete several short surveys, approximately two weeks apart, for approximately three months.
Second, the usage of the Owning a Home website is recorded for respondents in the Owning a
Home treatment group. Third, respondents are asked to submit some of their mortgage
documents.
The survey component will include questions on consumers’ current attitudes and behaviors
around shopping for and acquiring a mortgage. These questions will explore consumers’
attitudes to learn how consumers think about and perceive the mortgage shopping/acquisition
process, and consumers’ behaviors to learn more about what consumers actually do when
shopping for or acquiring a mortgage. The control group will help us understand more about the
status quo; that is, how consumers currently go about buying a home and getting a mortgage.
The treatment groups will help us understand how additional mortgage shopping and the Owning
a Home suite can alter that status quo.
Recording the Owning a Home treatment group’s use of the Owning a Home site will help us
learn more about what types of consumers are best served by the Owning a Home project, and
what pieces of the Owning a Home experience are most effective in improving consumer
attitudes and behaviors.
Finally, the collection of mortgage documents will enable us to learn whether changes in
attitudes and behaviors produced by either of the two treatments actually lead to improvements
in consumer outcomes, e.g.,. lower-priced mortgages. It is important to collect these documents
because data obtained from self-reporting of complex financial transactions is extremely
unreliable. Any personally identifiable information present on these documents will be redacted
prior to delivery to the CFPB.
Survey
The survey for this evaluation is organized into 10 modules. A brief description of the modules
is as follows, with additional justification below. Not all modules are administered in each
survey.
Module
A: Screening, consent &
contact information

Description of questions
Eligibility screening questions such as involvement in the mortgage industry and in
household’s financial decisions, and the timing of their plans to buy a home. After
Page 3 of 17

B: Homebuying stage
C: Search details
D: Empowerment
E: Knowledge
F: Beliefs about personal
home and mortgage terms
G: Correlates and
demographics
H: Lender summary &
disclosure
I: Request for documents

J: Wrap-up

participant is shown an informed consent and opts in, contact information is collected.
Participants’ stage of the home buying process (e.g. whether they have made an offer
on a home).
Identification of the sources used to learn about available homes and mortgage options,
and frequency of usage.
Participants’ confidence in searching for and evaluating mortgage information and
offers.
Factual questions on mortgages, including concepts like discount points and mortgage
insurance. Beliefs about real estate agents and lenders.
Pre- and post-purchase beliefs regarding a participants’ expectation of mortgage terms,
including purchase price, closing costs, and interest rates.
Correlates include time and risk preferences, financial literacy and numeracy, and need
for cognition. Demographics include race/ethnicity, income, employment status,
gender, age, and financial information relevant to mortgage search.
Summary of the contact respondent had with lenders throughout the mortgage process.
Respondent’s experience during closing, including how they used their closing
disclosures.
Request for documents. This is structured as a separate request that includes additional
privacy and consent language. If a participant opts not to submit documents, they are
instead asked questions about their mortgage terms.
Open-ended feedback from participants on the survey; treatment reinforcement.

Module A is required to identify eligible participants from the population of interest –
individuals who are interested in buying a home in the next few months, are involved in their
household’s financial decisions, who are planning on getting a mortgage to finance their home
purchase, and who are not real estate professionals. Module A will measure these concepts in
order to screen potential participants for eligibility. It will be administered one-time, and is
therefore not expected to represent an undue burden. At the conclusion of Module A,
participants are presented with informed consent language and asked to opt-in to the remainder
of the study.
Module B assesses participants’ stage of the home and mortgage search processes. It is used to
identify which subsequent questions and modules will be administered (or skipped), and
therefore is useful for minimizing participant burden. For example, participants who report that
they are still searching for a home will not be asked questions about when they put in an offer on
a home; conversely, those who say that they have put in an offer will not be asked how likely
they are to make an offer on a home in the next three months. Module B additionally provides
information on the dynamics of participants’ information search, for instance, at what point in
their home search they begin to think about mortgage terms.
Module C assesses participants’ information search for home and mortgage information,
including where they have looked for information and whether they have contacted lenders to
ask for background information, pre-approvals, or mortgage estimates. It is necessary for
measuring where participants are gathering information about mortgages.
Module D contains questions on participants’ empowerment, that is, their subjective ability to
navigate the mortgage process and confidence in doing so. These questions will allow
researchers to investigate whether the Owning a Home website increases empowerment, one of
Page 4 of 17

the outcomes of interest.
Module E contains factual questions that measure knowledge of the mortgage process and
products available including concepts like discount points and mortgage insurance. These
questions will allow researchers to investigate whether the Owning a Home website increases
knowledge, one of the outcomes of interest. It will also be used to assess whether knowledge
increases during the process of searching for a home or mortgage. By collecting this information
from both the treatment and control groups, researchers can compare the relative improvement in
knowledge (if any) between the groups.
Module F initially asks questions designed to assess participants’ beliefs about the mortgage
market, that is, the terms that they believe they will receive. After participants have purchased
their home, Module F is modified to ask parallel questions about the mortgage terms that they
actually received. By combining data on participants’ pre-purchase beliefs with information on
their demographic and financial characteristics, the research team will be able to assess whether
beliefs are reasonable, and whether they increase in accuracy over time. By comparing data on
participants’ pre- and post-purchase beliefs, the research team can look at whether participants
had accurate expectations about their mortgage outcomes, and whether expectations were more
accurate for participants in the treatment groups.
Module G collects background information on respondents’ characteristics, including risk and
time preferences, financial knowledge, numeracy, cognitive reflection, locus of control, and need
for cognition. These characteristics will be helpful in subsequent analyses for assessing whether
the Owning a Home tools are more effective for certain subsets of homebuyers. For instance, it
may be the case that those with a high level of knowledge of financial concepts are more easily
able to interpret the information that is provided on the website. In addition, Module G collects
basic demographic and socio-economic information including race/ethnicity, income,
employment status, gender, and age, as well as financial information relevant to mortgage
search. This data will be used for two purposes: (1) to assess whether randomization has been
conducted successfully and (2) to determine whether the Owning a Home tools are more
effective for certain home-buyers and whether participants are receiving mortgages that are
relatively expensive or inexpensive, given their background characteristics. Module G will be
administered only once to reduce burden on respondents.
Module H asks about the respondent’s overall contact with lenders and their experiences at
closing, including when they received their Closing Disclosure and whether they reviewed and
asked questions about it. Data from Module H can be combined with data from Module D
(measuring empowerment) to assess whether participants who were more confident about their
ability to navigate the mortgage process also asked more questions at closing.
Module I asks participants to send in their loan documents and provides privacy information and
instructions on how to submit the documents. Previous research has shown that self-reported

Page 5 of 17

mortgage information is often inaccurate. 1 Specifically, in a study in which borrowers were
asked to identify mortgage terms (using their loan documents), the percentage identifying a term
correctly was sometimes as low as 5%. 2 Given that mortgage outcomes are one of the primary
measures of interest, it is essential to collect accurate information on mortgage terms. To reduce
burden, participants will be given the opportunity to send in their documents directly through a
method of their choice: uploading electronic files, mobile capture, or sending a fax. They will
also be able to choose whether to send in their documents immediately or at a later time. Any
direct identifying information present on the documents will be redacted immediately, and the
source documents will be destroyed as soon as practicable. If participants decline to send in their
final mortgage documents, this module will instead ask questions about mortgage terms,
allowing for a coarse estimate of mortgage costs. In order to further reduce respondent burden,
Module I will be administered only when participants report having received a new loan estimate
or after closing.
Module J consists of a reinforcing treatment language for both treatment groups, as well as a
single open-ended, optional question, where participants can report any additional thoughts or
problems that they have with the survey. Data from Module J will be used to track any potential
problems with the survey collection that should be addressed by the research team. As it is an
optional question, it is expected to pose little burden to respondents. The final administration of
Module J includes additional questions for respondents in the Owning a Home treatment group
regarding their experience with the website.
3. Use of Information Technology
This collection will be administered almost entirely using electronic submission technology. The
survey component will be delivered electronically and responses will be received electronically.
The collection of Owning a Home site usage information will be completed entirely
electronically.
For the mortgage document submission component, we plan to give respondents several options
for submitting their documents. One option will be to scan (if necessary – many consumers may
receive mortgage documents electronically) and securely upload the documents. A second option
will be mobile capture technology, in which consumers take pictures of their paper documents
with their mobile phones. A third option will allow consumers to fax a copy of their paper
documents.
There are several reasons why electronic delivery and submission are appropriate for this
collection. The Owning a Home project is itself an online set of tools and resources in its current
iteration. Thus, our target audience for the Owning a Home project are prospective homebuyers
with at least moderate competence in using modern internet technology. Our recruitment pool for
Bucks, B., and Pence, K. (2008). Do borrowers know their mortgage terms? Journal of Urban Economics,
64(2): 218-33.
2 Lacko, J. M., and Pappalardo, J. K. (2010). The failure and promise of mandated consumer mortgage
disclosures: Evidence from qualitative interviews and a controlled experiment with mortgage borrowers.
American Economic Review: Papers and Proceedings, 100: 516-21.
Page 6 of 17
1

the study will be drawn from the existing customer base of a major online home shopping
website. Therefore, individuals in the study population are well-equipped to use an online survey
interface. Moreover, we anticipate that given the respondent pool, respondents will likely prefer
an online interface, and that conducting the study online will increase the response rate. Finally,
an online survey allows us to use dynamic programming to match the survey modules
administered to the participant’s individual home buying stage, thereby reducing burden.
We acknowledge that individuals who are comfortable with online tools and resources may be
different from the general population of prospective homebuyers. However, in order to recruit
consumers who are actively shopping for a home, we need access to a large number of people
who can easily be identified as prospective homebuyers. We do not intend to extrapolate our
findings to larger populations, including a general population of internet users or homebuyers.
4. Efforts to Identify Duplication
A key purpose of this research is to gain insight into: (1) general consumer behavior, attitudes,
and approaches to obtaining a mortgage and mortgage shopping; (2) the validity of certain
hypotheses underlying the development of the Owning a Home project. While some prior
research does exist on these topics, it is limited in scope.
Another key purpose of this collection is to gain insight into: (3) the mechanisms by which the
Owning a Home project impacts consumer behavior and attitudes; (4) which consumer segments
or profiles benefit the most from the Owning a Home project; and (5) the degree to which the
Owning a Home project impacts consumer behaviors, attitudes, and approaches to obtaining a
mortgage. Because the Owning a Home project has been specifically developed by the CFPB as
part of its consumer education and engagement mandate, no pre-existing sources of information
relevant to these objectives exists.
One of the hypotheses underlying the development of the Owning a Home project is that
additional mortgage shopping should allow consumers to obtain more favorable mortgage terms.
Previous research has explored the returns to shopping; however, there are two main limitations
of this work in addressing this hypothesis. First, the majority of existing research concentrates
on consumer goods other than mortgages. 3 Given that shopping for a home and mortgage is a
large, infrequently undertaken, and financially complex transaction, it is unclear whether results
from other markets will generalize to mortgage outcomes. A second limitation of existing
research is that methodologically, it has been limited to theoretical and correlational analysis. 4
3

See, for example, Griffith, R. et al. (2009). “Consumer Shopping Behavior: How Much Do Consumers Save?”
The Journal of Economic Perspectives, 23(2), pp. 99-120; Ratchford, B. and Srinivasan N. (1993). “An Empirical
Investigation of Returns to Search.” Marketing Science, 12(1), pp. 73-87; Brown, J. and Goolsbee, A. (2000).
“Does the Internet make Markets More Competitive? Evidence from the Life Insurance Industry.” NBER Working
Paper 7996.
4
See, for example, Cook M. et al. (2002). “Losing Interest: How Much Can Consumers Save by Shopping Around
for Financial Products?” Financial Services Authority Occasional Paper Series 19 and Woodward, S. and Hall, R.
(2012). “Diagnosing Consumer Confusion and Sub-Optimal Shopping Effort: Theory and Mortgage-Market
Evidence.” American Economic Review, 102(7), 3249-3279.
Page 7 of 17

Such research designs make it impossible to establish a causal relationship between shopping
behavior and mortgage outcomes, because homebuyers’ shopping behavior may be correlated
with other unobservable characteristics. Analyses that do not identify exogenous variation in
shopping behavior are therefore unable to cleanly identify returns to shopping. In contrast, the
current research will induce exogenous variation in shopping behavior through the use of a
randomized controlled trial.
We have relied on previous research in developing the approach to the Owning a Home project,
but it is not an adequate substitute for the planned collection. We expect that this study will
contribute two main sets of findings to existing research. First, the project will provide
information about the impact of Owning a Home and similar tools on mortgage outcomes. A
randomized controlled trial, such as the one described here, is the best method for cleanly
distinguishing the impact of financial education on mortgage outcomes from other individual
characteristics. Second, the project will provide estimates of the returns to search in the
mortgage market. As discussed, this work will broaden existing research on consumer search
using other consumer products and make methodological improvements in establishing the
causal effect of search on outcomes.

5. Efforts to Minimize Burdens on Small Entities
Not applicable. The information will be collected from private citizens, not business entities.
6. Consequences of Less Frequent Collection and Obstacles to Burden Reduction
This is a one-time collection of information. If the information is not collected, the CFPB will be
unable to make evidence-based decisions about how best to deliver on its consumer education
and engagement mandate. This study will enable a better understanding of consumers’ mortgage
shopping processes, which will enable the CFPB to make decisions about how to improve and
target its mortgage consumer education and engagement work.
7. Circumstances Requiring Special Information Collection
Participation in this study is entirely voluntary, and respondents can choose to cease their
involvement at any time. Respondents will be asked to report information every two weeks, for
a period of approximately three months. This frequency of survey was chosen so as to ensure
that we can capture the changes in consumers’ attitudes and behaviors over a relatively short
mortgage shopping/acquisition window. A less frequent survey schedule would require
respondents to recall how they felt or what actions they took over too long of a recall period. In
behavioral surveys, the best practice is to survey respondents frequently enough so that the
period of recall is not too long to confound their memories. To minimize burden, the individual
surveys will be kept short.
Respondents will be asked to respond to surveys within eight days of receipt. Again, the goal is
to capture consumers’ attitudes and behavioral reports while the information is still fresh in their
Page 8 of 17

mind. Moreover, given that respondents will be surveyed again two weeks later, the first survey
must be answered in a timely fashion so as not to conflate results with the subsequent survey.
With the exception of the reporting frequency noted above (every two weeks as opposed to
quarterly), this study is conducted in a manner consistent with the guidelines in 5 CFR 1320.
5(d)(2).
8. Consultation Outside the Agency
In accordance with 5 CFR 1320.8(d)(1), the Bureau published a notice Federal Register allowing
the public 60 days to comment on this proposed new collection of information. The Bureau
received three comments in response to this notice. The commenters suggested that the Bureau
should make the results of the study publicly available, that information about home ownership
is useful, and that more information about the home buying process should be added to the
Owning a Home website. The Bureau currently intends to follow all of these recommendations.
Further and in accordance with 5 CFR 1320.5(a)(1)(iv), the Bureau has published a notice in the
Federal Register allowing the public 30 days to comment on the submission of this information
collection request to the Office of Management and Budget.
The CFPB has also consulted on the design of this survey with Mick Couper, a professor and
survey design expert at the Population Studies Center at the University of Michigan. The CFPB
tested the survey instrument using two rounds of cognitive interviews. The first set of interviews
was conducted on September 25 and 26, 2014 while the second set was conducted from March 6,
2015 through March 10, 2015. During both sets of interviews, participants were asked openended questions regarding their understanding of study materials, revealing problems with
question comprehensibility and response options. Study materials were subsequently refined to
address these issues.
In addition to these interviews, the CFPB conducted a pilot test of the full study from October
27, 2014 through November 9, 2014. This pilot-phase study was approved by the Office of
Management and Budget under a separate generic PRA clearance. This study was conducted
using procedures that were designed to parallel those from the full study. Specifically,
participants were recruited through an email solicitation, answered screening questions, and
completed the baseline survey and up to two periodic surveys. Based on results of the pilot
study, the CFPB revised the survey instrument and recruitment language, incentive structure, and
implementation method. Examples of the changes are:
•

Survey instrument: Questions with low variance were eliminated in order to reduce the
length of the survey instrument. For example, 83.0% of respondents in the pilot study
could correctly identify the concept of “home equity”; therefore, we no longer ask this
Page 9 of 17

•

question. Eliminating questions reduces the length of the survey instrument and
corresponding respondent burden.
Implementation method: The survey invitation now contains unique links so that
participants do not have to enter a survey password. This change reduces respondent
burden and is expected to reduce attrition.

9. Payments or Gifts to Respondents
Prospective homebuyers are a hard-to-find population—only a small percentage of the
population is actively interested in purchasing a new home at any given time. Additionally,
searching for a home is a time-consuming activity. We believe incentives are necessary to ensure
an adequate response rate.
The current literature on survey incentives suggests that study response rates increase
dramatically when participants are compensated for their participation relative to when they
receive no incentive (Goritz et al. 2006). Our pilot study also provided estimates for the response
rates for the early surveys. Based on these estimates and the literature we believe monetary
incentives are necessary.
A few studies have tested the magnitude of payment on retention in an analogous situation to our
proposed study. Alexander et al. (2008) conducted a study quite similar to our intended design.
Participants in this study were recruited for a baseline survey and a three month follow-up
survey. Incentives were given out at the baseline and follow-up survey. Of 24 combinations of
baseline and follow-up incentives they found that a $5 incentive for the baseline and a $20
incentive for the follow-up survey resulted in the highest retention. Unfortunately, none of the
incentives tested were greater than $20 for a single stage of the study so we do not know if
paying individuals more than $20 for completing a single phase would dramatically increase
retention.
As part of our pilot study we experimentally manipulated the level of incentives homebuyers
were paid to test if the amounts impacted retention. Participants were offered incentives for
completing two different phases of the study. The first milestone was completion of the baseline
survey and the second was finishing the remainder of the pilot. The combinations of incentives
offered for completing the first and second phases were $5/$15, $5/$25, and $15/$15. Two
additional incentives were offered for submitting mortgage documents at the application and
closing stage, each $5 dollars. Participants could submit multiple sets of application documents
but they would only be paid for the first set. For the pilot study, participants had to complete at
least one of the two follow-up surveys after they completed the baseline in order to receive the
study completion incentive. In our full study, participants will need to complete five of six
follow-up surveys after completing the baseline survey to receive the study completion incentive.
We used statistical analyses to assess the causal impact of the different incentives on completion
rates of participants. With the incentive amounts tested we cannot claim that the marginal
Page 10 of 17

impacts of higher incentives produce statistically significant differences in retention. For later
stages in our pilot study the sample sizes became quite small and therefore our lack of statistical
significance may be due to insufficient power. Given the lack of clear direction from our pilot
testing we will defer to previous experimental results.
Our incentive scheme is multistage and dynamic. It is designed to maximize the response rate of
all necessary components of the study while minimizing the cost to the government.
Following the example of Alexander et al. (2008), participants are informed at recruitment and in
the informed consent stage that they will receive a $5 incentive for completing the baseline
survey, and an additional $20 incentive for completing the entire study.
During the course of the study, at a moment in time appropriate to the individual participant’s
stage in the homebuying process, we ask participants to submit their Loan Estimates (mortgage
offer disclosures). Participants are offered an additional incentive of $5 to submit their Loan
Estimate documents. Based on our pilot experience, we believe the additional incentive is
necessary to ensure a reasonable response rate. If a participant indicates in a later survey that
they have received additional Loan Estimates, they are invited to submit these documents as
well, but not offered an additional incentive.
After the participant has closed on their home, we ask participants to submit their Closing
Disclosures and offer an additional $10 incentive. Based on our pilot experience, we believe the
additional incentive is necessary to ensure a reasonable response rate. We offer a larger
incentive than at the Loan Estimate stage for two reasons: (1) Closing Disclosure data is
particularly important for our analysis; (2) participants are likely to be more fatigued, both with
respect to the home buying process and with respect to participating in the survey, at the closing
stage.
Finally, we anticipate that some number of study participants will have put an offer on a house,
but will not have closed on that home within the three month study timeframe. Our study
provides a novel contribution by tracking outcomes that occur at closing. Therefore we do not
want to miss out on collecting closing information from individuals who have completed several
surveys and are scheduled to close on a home, but will not complete that closing during the
timeframe of the study. Losing out on closing data from this population would be detrimental to
the study. Thus, we invite this limited population of participants to complete a follow-up survey
once they have closed.
Since these participants have already completed their obligation outlined in the consent form, we
feel that the follow-up survey request will require an additional incentive. We offer an additional
incentive of $5 to complete the follow-up survey. The following chart indicates the maximum
incentive offered to all participants:

Incentive

Completes the
baseline

Completes
the study,

Participant experience
Completes the study
Completes
and submits
the study

Completes the study
and submits
Page 11 of 17

Baseline ($5)
Study completion
($20)
Submits Loan
Estimate ($5)
Submits Closing
Disclosure ($10)
Completes followup survey ($5)
Total

survey, fails to
complete the
remainder of
the study

but
declines to
submit
documents

$5

documents, but still
searching for a home
(not scheduled to
close) at conclusion
of 3-month study
window
$5

and submits
documents,
closes
within 3
month study
window
$5

documents, home
purchase accepted
within 3-month study
window, but scheduled
to close after study
window is complete
$5

$5

$20

$20

$20

$20

$5

$5

$5

$10

$10
$5

$5

$25

$30

$40

$45

The incentives will be delivered through PayPal, Amazon e-giftcard, use-anywhere e-giftcard, or
reusable physical gift card. Incentives are delivered throughout the study – at the conclusion of
the baseline survey, at the conclusion of the study period, upon submission of documents, and at
the conclusion of the follow-up survey.
Participants will be able to complete all survey activities at home or in the location of their
choice, thus there are no expected travel costs. The surveys have been tested to minimize time
burden; additional details are provided below in Section 12.
Providing incentives for completion and minimizing costs for participants (in terms of time and
travel) will decrease missing data problems arising from survey non-response and participant
attrition. Reducing these problems is important for lowering the probability that there will not be
sufficient statistical power to derive conclusions from the study. Without statistical power, the
Bureau would have to determine whether it would want to repeat this data collection, thus
engendering additional costs to the government.
10. Assurances of Confidentiality
The Bureau will not disclose any personally identifiable information collected except to the
extent that it is required to do so by law and as provided in the Privacy Impact Assessment listed
below. Additionally, the Bureau will treat the information collected consistent with its
confidentiality regulations at 12 C.F.R. Part 1070, et seq.
The information contemplated here is covered under two separate Privacy Impact Assessments
(PIA). The Consumer Experience Research PIA is available at
http://files.consumerfinance.gov/f/201406_cfpb_consumer-experience-research_pia.pdf, and the
Consumer Education PIA that is available at
http://files.consumerfinance.gov/f/201409_cfpb_consumer-education_pia.pdf. Additionally,
information contemplated here is subject to the Privacy Act of 1974 and is covered under the
Bureau’s existing Consumer Education and Engagement Records System of Records Notice, 77
Page 12 of 17

FR60382.
Participants are informed that their responses will be kept confidential under 12 C.F.R. Part
1070, et seq. at several points during the survey process. This assurance is included in the initial
recruitment email, the informed consent, and at the document submission request. Participants
can view the full Privacy Act Statement from the initial study welcome page, the informed
consent page, and in every subsequent periodic survey welcome page.
11. Justification for Sensitive Questions

Respondents will be asked to provide basic demographic and socio-economic information
including their race/ethnicity, employment status, gender, and age. This information is necessary
to determine whether a socio-demographically diverse data set has been collected and to what
extent respondents are representative of the sampling frame or the broader population of U.S.
homebuyers. Additionally, it is necessary to determine that the randomization process was
correctly executed and that the treatment and control groups are similar with regard to known
factors.
Respondents will also be asked to provide financial information relevant to their home and
mortgage search, including an estimated purchase price, income level, liquid assets available,
and credit score. This information is necessary to determine whether an economically diverse
data set has been collected and to what extent respondents are representative of the sampling
frame or the broader population of U.S. home-buyers. In addition, this information will help to
determine whether respondents receiving relatively expensive or inexpensive mortgages, as
mortgage rates are partially determined by purchase price, a borrower’s creditworthiness (credit
score, income), and potential down payment (liquid assets).
Finally, respondents will be asked to provide mortgage loan documents. Previous research has
demonstrated that borrowers may not be aware of all of their mortgage terms; therefore, it is
necessary to see loan documents directly. Improvement in mortgage terms is one of the main
measures of interest.
Before providing any personally identifiable or otherwise sensitive information, participants will
be given informed consent documents. Such documents will give researchers permission to
collect and analyze demographic and financial information throughout the study. Additionally,
during survey collection, respondents may refuse to answer any question or participate in
additional survey activity, including providing mortgage documents.
12. Estimated Burden of Information Collection
The calculation of burden is complex, as attrition over the course of this longitudinal study
results in a diminishing number of participants at each stage. The below table includes estimates
of attrition and a detailed calculation of burden hours for each task. The per-task burden
estimates were derived from data collected during the pilot study. The total burden is estimated
Page 13 of 17

at 12,480 hours.
Line # Component

1 Arrive at study homepage
2 Screener (Module A) (start)
3 -- completed & eligible

Retention # Participants Time/
Burden Burden
person Hours
calculation
(mins)
notes
230,000
2
5673
74%
170,200
25%
42,550

4 Opt-in to study

59%

25,105

4 Finish Module A (contact details)

73%

18,327

5 Baseline Survey (start)

84%

15,395

6 -- complete

83%

12,778

7 Periodic Survey #1

57%

7,283

8 -- complete
9 Periodic Survey #2

94%
90%

6,846
6,161

5%

639

95%
90%
5%
95%
5%
95%
93%
95%
10%
90%
90%
95%
20%
80%
90%
95%
30%
70%
70%
30%

6,460
5,814
342
5,848
292
5,556
5,167
4,909
491
4,418
3,976
3,777
755
3,022
2,720
2,584
775
1,809
1,266
543

100%
90%
50%

1,266
1,139
570

10 + rejoiners
11 --> overall completion
12 Periodic Survey #3
13 + rejoiner rate (those who miss survey #2)
14 --> overall completion
15 -- Of completed, closed
16 -- Of completed, still searching
17 Periodic Survey #4
18 --> overall completion
19 -- Of completed, closed
20 -- Of completed, still searching
21 Periodic Survey #5
22 --> overall completion
23 -- Of completed, closed
24 -- Of completed, still searching
25 Periodic Survey #6
26 --> overall completion
27 -- Of completed, closed
28 -- Of completed, not closed
-- Of not closed, offer accepted
29
-- Of not closed, still searching
30
31
32 Follow up survey - eligible
33 -- close during follow up window
34 -- complete survey
Total

2

611

18

4226 avg start/fin

3

353 avg start/fin

3

332 avg start/fin

10

975 avg start/fin

3

123 avg start/fin

3

94 avg start/fin

3

65 avg start/fin

3

29

50

12,480

13. Estimated Total Annual Cost Burden to Respondents or Recordkeepers
Page 14 of 17

There are no costs to respondent for responding to this collection of information.
14. Estimated Cost to the Federal Government
The costs to the Government to conduct this collection consist of a firm-fixed price contract for
survey administration support and data matching, and the incentives paid to participants.
Category
Contractor support
Incentives to respondents
Total cost to the Government:

Method of computation
Firm fixed-price contract
See below

Cost
$207,133
$173,905
$381,038

Incentives to Respondents
The calculation of the total cost of incentives is complex, as attrition over the course of this
longitudinal study results in a diminishing number of participants at each stage. The below table
includes estimates of participation and incentive costs for each incentive payment. These
estimates were derived from data collected during the pilot study.
Component
Complete baseline survey
Complete study
Submit Loan Estimates
Submit Closing Disclosures
Complete follow-up survey
TOTAL

# Participants Incentive Total $
12,778
4,122
2,061
1,442
570

$5
$20
$5
$10
$5

$63,890
$82,440
$10,305
$14,420
$2,850
$173,905

15. Program Changes or Adjustments
Not applicable. This is a new information collection. There is no prior OMB approval.
16. Plans for Tabulation, Statistical Analysis, and Publication
Information from participants at baseline will be used to evaluate the quality of the
randomization process. Specifically, if assignment to treatment and control groups has occurred
randomly, then there should be no statistically distinguishable differences between the treatment
and control groups with regard to descriptive characteristics. The initial survey will include
information on basic demographic and socio-economic characteristics, information sources used
thus far in the mortgage search, and knowledge of and attitudes toward the mortgage process.
Page 15 of 17

The primary analytic technique for estimating the impact of the two treatment conditions is a
comparison of the outcomes for participants who were given treatment (the treatment groups) to
those who were not given access (the control group). The research team intends to use “intent to
treat” analysis, thereby estimating effects on those who had the opportunity to receive treatment,
including those who did and did not make use of the treatment. This analysis will identify the
overall causal impact of the two treatments -- access to the Owning a Home suite of tools, and
encouragement to comparison shop for mortgages -- on mortgage outcomes for those in the study
population.
Estimates of the differences between the two groups will be calculated using data from surveys,
Owning a Home website usage, and mortgage outcomes. For each difference, statistical
confidence intervals will be calculated using standard statistical assumptions.
For some outcomes, the research team is additionally interested in how outcomes changed during
the home shopping process. For example, it may be expected that respondents who have
completed their home purchase and have closed on a mortgage will have higher knowledge of
the mortgage process than those who are in the initial stages of their home search. As such,
changes in mortgage knowledge between baseline and the final survey will be compared for
those in treatment and control groups. Mortgage knowledge will be measured using scores from
Module E, as described above. Specifically, on an individual level, change in mortgage
knowledge will be calculated as:
Change in mortgage knowledge = (mortgage knowledge post-purchase) - (mortgage
knowledge at baseline)
A similar calculation can be performed for change in empowerment (using scores from Module
D).
In addition to variation that is imposed on the groups from random assignment, there will be
variation on an individual basis. The research team is interested in whether the Owning a Home
suite of tools (or encouragement to comparison shop) is more effective (in terms of improving
mortgage knowledge, empowerment, or mortgage outcomes) for individuals with particular
characteristics, such as higher financial literacy. Thus, within the treatment groups, outcomes
will be compared for respondents with differing baseline characteristics.
Finally, it is worthwhile to know whether shopping behavior itself improves mortgage outcomes,
and some academic research has investigated this question. 5 It is expected that participants in
the treatment groups may engage in more shopping behavior than those in the control group. In
order to isolate the effect of shopping, researchers will employ what is known as a “two stage
least squares” approach. This technique will statistically isolate the effect of shopping on
mortgage outcomes, allowing researchers to estimate the benefits from shopping.

Woodward, S. E., and Hall, R. E. (2010). Consumer confusion in the mortgage market: Evidence of less than a
perfectly transparent and competitive market. American Economic Review: Papers & Proceedings, 100: 511-15.
Page 16 of 17

5

The research team plans to analyze, integrate, and summarize data and findings from the research
in a report, which will be made publicly available.
Tentative Timetable
Information collected
Data analysis
Writing of report
Publication of aggregate report

February – September 2016
October 2016 – March 2017
April 2017 – September 2017
Approximately December 2017

17. Display of Expiration Date
The Bureau plans to display the expiration date for OMB approval of the information collection
on all instruments. Further, the OMB control number and expiration date will be displayed on the
Federal government’s electronic PRA docket at www.reginfo.gov
18. Exceptions to the Certification Requirement
The Bureau certifies the this collection of information is consistent with the requirements of 5
CFR 1320.9, and the related provisions of 5 CFR 1320.8(b)(3) and is not seeking an to these
certification requirements

Page 17 of 17