Washington, DC – ARC Meeting on 05/17/2017


My name is Ron Borzekowski. I’m the Assistant Director for the Office
of Research at the Consumer Financial Protection Bureau, or the CFPB. Today’s meeting is being held at the CFPB
Headquarters in Washington, D.C., and I’d like to welcome the ARC back for this important
discussion. Let me spend a few minutes telling you what
you can expect this morning. First, I will introduce the ARC members and
then CFPB’s acting Deputy Director, David Silberman, will provide some opening remarks. Following David’s remarks, we’ll engage with
the ARC in two short discussions. The first will focus on yesterday’s subcommittee
meetings where members of the staff and the ARC discussed the office’s research in the
areas of household balance sheets and disclosure. This will be followed by a presentation on
the Bureau’s recently announced work regarding the assessment of certain significant rules. The meeting should adjourn by 11:00 a.m. The Bureau established this Advisory Council
of leading academics from throughout the U.S., and the Council has served the Bureau well. Our Academic Research Council advises the
CFPB by sharing their unique experience on research strategy and methodology and today’s
public meeting and discussions in support of this important responsibility. As a reminder, the views of the ARC members
are their views and are generally appreciated and welcome. However, they do not necessarily represent
the views of the CFPB. So, let me now introduce the ARC members that
are here today. Justine Hastings. Justine is an Associate Professor at Department
of Economics in Watson’s Institute for International and Public Affairs at Brown University. Full professor, my apologies. I forgot to update. Before I go on, have you changed your title? Brigitte Madrian. Bridgette is the Aetna Professor of Public
Policy and Corporate Management at the Harvard Kennedy School. Melvin Stephens. Mel is a Professor of Economics at University
of Michigan. And John Lynch, our newest member of the ARC. John is the Ted Andersen Professor of Free
Enterprise at the Leeds School of Business at the University of Colorado, Boulder, and
Director of the Center for research on consumer financial decision making. Notably, John is also the host of the Boulder
Summer Conference on Financial Decision Making, an interdisciplinary conference now in its
seventh year that’s squarely focused on the issues that the Bureau wrestles with daily. Unable to be present today are two other ARC
members, John Campbell from Harvard University and Ian Ayers from Yale University. Now, unfortunately for us at the CFPB, John
Campbell, Justine Hastings, and Bostic will be finishing their service with the ARC this
year. Professors Campbell and Hastings are founding
members of the ARC, and we are very grateful for their valuable service and the contributions
over the last few years. Professor Bostic joined the ARC a bit later,
but is leaving shortly to pursue his new role as President of the Federal Reserve Bank of
Atlanta. So, we thank all of them. So, I’m now very pleased to introduce David
Silberman, acting Deputy Director of the Bureau. I thank you, Ron. I want to join Ron on behalf of Director Cordray,
who unfortunately can’t be here this morning, to welcome everyone to this 2017 meeting of
the Academic Research Council. As Ron indicated, we created the council nearly
six years ago as a technical advisory body of scholars to provide our office of research
with strong methodological and technical advice, along with guidance on data collection strategies
and methods. All in support of the Bureau’s underlying
commitment and unyielding commitment to bringing an evidence-based perspective to the oversight
of consumer financial markets. We are quite aware of how fortunate we are
to have such great leaders in their fields collaborate with us in our efforts to better
understand the behavior of consumers and the workings of these markets. I’ve had the privilege of being part of many
of the prior discussions of the annual ARC meetings, and can attest to the rich and fruitful
exchanges between these distinguished scholars and the talented researchers at the Bureau. I even have been able to understand parts
of what’s been said in each of these discussions. Today, I thought I would briefly recap some
of the work that our — sorry about that. Support of the Bureau’s mission. In doing so, I hope to illustrate the variety
of creative ways in which our researchers are going about the task of gathering and
analyzing evidence to inform our understanding of consumer financial markets and our policy
decisions with respect to those markets. In December of last year, we released a tool
we call Consumer Credit Trends. The tool is built on the Consumer Credit Panel,
a nationally representative sample of credit records from one of the top three nationwide
credit repositories. The credit records are deidentified before
being sent to the Bureau, meaning we do not receive names, addresses, social security
numbers, account numbers, or any other information that would reveal the identity of anyone in
our sample. This protects consumer’s privacy while enabling
us to generate valuable information on market trends. Using the data we receive at the individual
loan level, we could estimate originations for several different consumer credit products,
such as mortgages, credit cards, auto loans, and student loans. We also can disaggregate these data — did
I use the plural? For various subsegments of consumers, such
as older Americans, or those living in low-income communities, or those with subprime credit
scores. We update this information monthly as new
data arrives to supply timely ongoing information about these markets. This is not, however, simply an exercise in
counting, because different creditors report to the credit bureaus on different cycles
and because of lags in reporting, underlying consumer credit trends a sophisticated model
our researchers have built to enable them to construct reliable monthly estimates of
originations from incomplete or partial data. Highlights from the most recent release, which
we published just last week, show for instance, that while automobile lending overall continues
to increase mildly, up 3.4 percent year over year in March, lending to consumers with subprime
credit scores has been declining for 13 consecutive months. Conversely, while mortgage lending appears
to have cooled somewhat overall this year, likely in response to higher mortgage interest
rates that negatively affected the market for refinancing, lending activity to consumers
with a near prime or subprime score has remained stable and at levels that have hovered around
30 percent above the lending levels from one year ago. These are the kind of insights we’re able
to glean from the consumer credit trends. Now our consumer credit panel is not only
an important source of data, but also a valuable tool for conducting survey research to gain
insight into how consumers experience various elements of the consumer financial marketplace. For example, in January, we released the new
survey that spotlights the debt collection experience of a nationally representative
sample of consumers. To do this we mailed surveys to over 10,000
consumers selected from our consumer credit panel, and we obtained responses from over
2,000 of them, a very strong response rate. By using the panel, we were able to oversample
consumers with debts and collections, and then weight the responses so as to report
results that are nationally representative. Also, by linking responses back to deidentified
administrative records, we are able to compensate for a nonresponse bias, which is a challenge
in any survey research project, and we were able to use the data to analyze the results
of different segments of consumers while carefully preserving the confidentiality of each respondent. All this allows for robust analysis of consumer’s
debt collection experiences. For example, our survey enabled us, I think
for the first time, to provide a clear estimate of the size of the debt collection market. We found that one out of three consumers report
having been contacted over the prior year with respect to a past due debt, with the
majority of these consumers having been contacted about multiple debts. For consumers earning under 20,000, over half
were contacted by at least one collector during the prior year, and almost 80 percent of those
reported having been contacted about multiple debts. Our survey also provided insight into how
consumers experience debt collection, including information about the type of debt being collected,
when, how, and how often consumers were contacted by a collector, and the nature of their interactions
with collectors. And we uncovered important information about
the interrelationship of debt collection and collection lawsuits. For example, we found that about one out of
every seven consumers who said they were contacted about a debt during the past 12 months, also
reported having been sued during that period. Of those, only one out of four consumers reported
even attending the court hearing, meaning that three out of four consumers were subject
to default judgments. As our research shows, debt collection is
an important part of the way in which consumers interact with financial markets. Debt collection is also a market in which
consumers generally have no market power. Consumers do not choose their debt collector
and cannot switch collectors if they are unhappy. And it’s a market which collectors generally
have an incentive to extract as much as they can from each consumer. It is thus a market where regulation can play
an especially important role. In going forward, the results of our survey
will inform the consumer bureau’s approach to overseeing the debt collection industry. Let me give you one last example of cutting
edge research of a different type, informing a very different part of the Bureau’s agenda. As part of Project Catalyst, the Bureau’s
program to encourage consumer friendly innovations in consumer finance, we teamed up with American
Express to structure a randomized trial focused on studying the ethicacy of alternative strategies
to promote savings. American Express agreed to test various methods
of encouraging consumers with an American Express pre-paid card, to use a product feature
that offered them the opportunity to set money aside in a non-interest-bearing savings, quote,
wallet that was separate from the funds they used for regular transactions. The pilot lasted for three months and a follow
up survey was conducted nine months afterwards to let us look at consumer outcomes over a
longer period of time. This research yielded a number of key findings. We found that offering a small incentive,
just $10, to pre-paid card users to put some of their money into a savings wallet doubled
uptake of the wallet. Furthermore, consumers who used the savings
wallet tended to continue to use it even when they would no longer being encouraged or incentivized
to do so. And participants who were offered an incentive
to save reported significantly less use of pay day loans and paycheck advance products
over a period of 12 months, including the trial and the follow on period, as compared
to those who were not offered any incentive during the trial. So, these are just three examples of the work
the Office of Research has done in the past year. Our researchers also provide analytical support
to our fair lending and compliance missions, as well as to our consumer engagement team. They support project catalyst efforts to foster
innovation in consumer financial products and services, and they’re integrated into
our policymaking processes and are ultimately responsible for the close of evaluation of
our proposed and final rules through considerations of benefits, costs, and burdens as required
for governing statutes. Later this morning, you’ll be hearing about
the steps we are taking to research and assess the effectiveness of a number of significant
rules promulgated by the Bureau, and which will be approaching their five-year anniversary
in the coming year or two. And all of our research efforts we’ve benefited
from the wisdom and experience of the members are of our academic research Council. So, let me close by again, thanking the members
of the Council, welcoming John Lynch to the Council. We are grateful for your time and your insights,
and I look forward to the rest of the presentations and discussions this morning. Thank you. David, thank you very much. As I noted earlier, our next agenda item today
is to at least give readouts, to give some readouts of the two discussions that happened
yesterday. And with two of the ARC members not being
here, I will actually do the readout for economics of household balance sheets, and then Professor
Hastings, Justine, if you could follow on with the discussion of the disclosure discussion
that happened yesterday. The office has basically decided to focus
much of its work, although not all of its work, in two areas of research. We call them research agendas. One on what we call the dynamics of household
balance sheets, understanding how the financial lives of consumers evolve over time, the shocks
that they face, the persistence of those shocks, and the interrelation of particular products
in the markets with these consumers lives as they use them or don’t. So, that was part of this discussion yesterday. Staff at the Office of Research laid out a
number of those projects in high summary form, and then there was a deep dive in two particular
projects with the ARC members. The first was a discussion of early preliminary
work in a dynamic structural model that looked specifically at consumption and savings, focused
on low income consumers and extended periods where they may actually be in debt. We also had a complementary presentation using
very different methodologies, talking about pilot work to gather data and to understand
the shocks that such low-income consumers may face and may put them into a situation
where they’re looking for money, or need debt. So, following that conversation, the question
became, what direction should we move in going forward. And the discussion focused in two areas, new
data sources that the Bureau could potentially obtain, but also were we to actually use survey
methodologies, and as David mentioned we had the capacity to run surveys off of the consumer
credit panel using that as the administrative frame. Were we to do that, what would be the pluses
and minuses of taking that approach. And the pluses and minuses of taking that
approach, in particular if were trying to ascertain the assets that a given household
may have. That’s something where our current data is
a little bit weaker; an important area for us to develop in the future. So, we discussed what would happen if we used
the consumer credit panel for that work, any potential issues of selection, for example,
we know that there are about 11 percent of the US population that would not be in the
credit records and we could not sample or reach that way. We also discussed other issues, such as whether
this should be a one-shot survey of some sort, perhaps a panel, something more longitudinal,
whether there are newer technologies that we should be using rather than mail surveys
which is been our traditional, or Internet surveys, a traditional way of doing this. And the ARC brought forth, you know, various
ideas, both for other data sets that we could get, and the surveys that we could possibly
conduct. I think they were generally supportive of
this work and we were happy to hear that, and offered a lot of specifics. So I think, I don’t know if that’s a fair
summary and if any of you would like to join in or expound on that before Justine goes
on. Thank you, Ron. You know, listening to all the projects that
were discussed yesterday in the dynamics of household balance sheets, I thought it was
a very exciting research agenda. This idea of pulling together administrative
data along with household survey data to complement each other, is something that academic community
has been very excited about doing and pursuing very recently. And so, hearing that the Bureau is taking
the lead on this and actually creating some exciting products so we can learn a lot about
household balance sheets and better understand the situations households are finding themselves
in, I think is a quite exciting agenda and I look forward to seeing much more in the
future. Justine. Thanks. So, I also really enjoyed both of the conversations
yesterday. I’m going to summarize one of the other panels
in a moment, but I think one of the really commendable things both about the survey and
about generally much of the discussion, is the Bureau’s focus on measuring and understanding
and incorporating that understanding from data and research into decision-making. You know, a great thing we might think about,
for example, regulations for balloon mortgages. Well, if we don’t know anything about them
or what their impacts are, or even the characteristics of people who are taking them, how can we
actually design a regulation that we think is going to address a problem if problem exists? And so, taking this kind of deliberative approach
to actually measure things that we don’t see, and incorporating that into how we’re making
policy, it’s great to see that throughout all of the discussions and everything that
CFPB is doing. I think it really gives a lot of confidence
in the process, that people are really trying to understand where are there difficulties
and challenges, is affected by them, and what are the best directions to address those challenges? Thanks, much. This is a reasonable time if you want to summarize
the second conversation, then perhaps, you know, John and Brigitte, they have any reactions,
can chime in. Great. So, on that same theme of careful deliberation
and informed discussion, the Bureau yesterday, CFPB and ARC members engaged in a discussion
about the challenges of creating effective disclosure forms, and measuring the effectiveness
of disclosure forms once they are launched or put into the real world. And the discussion focused on a particular
example, which is the disclosure of prepaid cards. And with that example, we engaged in a discussion
around the challenges of effective disclosure policy. Heidi Johnson provided an overview of the
Office of Research projects that fall under disclosure research, and ARC member Eric Johnson
from Columbia Business School, led the discussion around an example, particularly highlighting
possible approaches for disclosure in the prepaid card industry. We used a visual of a hypothetical disclosure
requirement for prepaid cards. ARC members and CFPB staff brought their expertise
from successful research experiences to bear on this important question. The discussion included evidence about what
works for disclosure, given how busy people are and given their busyness and all the things
we all have to accomplish in a day, what is the right way to present information so that
busy people can use it and make the decisions they’d like to make if they had more than
24 hours in a day. And we discussed how people interpret average
fees versus their personal expectations over fees. We know a lot from a wealth of research about
how people think about, you know, project what they expect their usage is going to be
versus a typical person. We discussed the very important topic of how
firms might respond to consumer’s response, or changes in demand that happen when we implement
a disclosure, and how we might incorporate these responses into disclosure policy. So, thinking two steps ahead, you know, if
we want to enable consumers to make informed decisions in a transparent market, how do
we incorporate both consumer’s interpretation of information, and also how firms will respond
to regulations to ensure that at the end of the day we achieve our goal. For example, we brought evidence from real
policy changes in international settings where such responses of industry to regulation and
disclosure policy have hindered policy accomplishing its goals, and thought about basically how
we are going to design, how we might design disclosure that doesn’t just simply resolve
in new created fees that fall outside of just, the current regulations. So, how do we incorporate how firms are going
to interpret this disclosure regulation is that we in fact achieve the goals of transparent
markets. We discussed how to test disclosure in the
lab, and also in real-world settings, following rigorous research approaches that have been
tried and proven successful in the past. And there is robust discussion centered on
how to use disclosure to best protect the consumer and ensure transparent markets. If anyone would like to add to that summary,
please add to the summary. I thought it was a great discussion, and again,
a wonderful example of an agency really focused on using data and research to direct policy
so that we make sure policy is as effective as possible. Thanks very much. One thing that was, I think very notable of
that conversation when I was reflecting on it, and perhaps this will spur a little bit
of data, is that at least two of the ARC members have written research directly relevant to
that. Very, very relevant. And we’re able to bring her personal experience,
known experience from the field and from the lab to that conversation. I think the other thing that was quite interesting
was a, I don’t want to say it’s a difference of opinion, but various ARC members coming
to that discussion with various solutions. Some suggesting further research in a lab
setting, some suggesting further research and actual field settings, and then a long
discussion of whether there would be uses for administrative data to assess the effectiveness
of disclosure. So, it was a very rich discussion both because
of the backgrounds of the people involved, and the wide-ranging number of suggestions
or ideas on how we could continue that work. John, I don’t know if you want to — this
is your area. I guess, the only thing I would add is that
one of the really interesting dimensions of the discussion was if you’re trying to have
consumers understand the nature of fees that are associated with these products, but the
fees can be many different forms, how does one disclose that when any given individual
might incur different fees. And so, that issue of heterogeneity I thought
was a really interesting problem. Professor Madrian suggested as a criterion
for deciding what is an effective disclosure, one’s ability to correctly predict exactly
what fees one would experience personally. Which I thought was a good suggestion. Fantastic. Thanks, much. So, I think both conversations, to wrap up
quickly, and if these are conversations we hope to have with the ARC members as the year
goes on and to continue to engage them in a lot of our work, but both I think as Justine
carefully said, we are very directed at research and research methodology, and how the Bureau
can continue to build on the capacity it has to inform policy. You know, in that sense and very grateful
again for their time. Moving on to our last item of the morning,
let me introduce Paul Rothstein who is a section chief within the Office of Research and is
going to give a short presentation on two initiatives that the Bureau has just — particularly
an issue, the Bureau has just begun driven largely by the Office of Research involving
many people in the Bureau. So, Paul, the floor is yours, please. Hi. Yes? Good. All right. My pin goes down. All right, so we’ve been working on developing
this line of research for number of years. And I’ll say a little bit more about that
in a second. So, it’s very exciting to move into a more
public phase of discussing this work on assessment of significant rules. Okay, let me make sure this works, because
the last technical thing to make sure I’ve got. Good. All right. So, Dodd Frank, our governing statute, mandates
that we do these retrospective reviews that the statute calls assessment, and we’ve adopted
that term. So, the first, there are these five elements
that I’ve written up here on the slide, and I want to look at the first two together and
say, make a few comments, and then move on to the others. So, we are required to conduct an assessment
of each significant rule adopted under federal consumer financial law. And we are to publish report of the assessment
within five years of the effective date. So, there is this puzzle here. First of all, we have to define what is a
significant rule, and however we do that, there has to be a clear and well-defined effective
date so that we know when this report is due. So, in the real world, which the Bureau works
in, we produce rules and we amend them, sometimes prior to the effective date, sometimes afterward. We put a lot of thought in, into thinking
about what was the right way to think about what a rule was for purposes of these assessments. So, it was a reasonable exercise that we weren’t,
by some just frontalism, excluding requirements that should be part of thinking about what
a rule is doing, but also, you know, since there’s a five-year clock ticking we have
a well-defined effective date and we don’t get too far out, you know on the timeline
for, you know, we have to just cut off something like data collection because we have to write
a report. So, we thought real hard about this and obviously,
you know, as you might expect rules with a common effective date would be thought of
as together, but with perhaps a little bit of flexibility there for when just — in order
to produce a comprehensive or reasonable assessment we would go a little beyond that. And then significance, as we’ve put in some
of the Federal Register notices we’ve issued already, you know, is related to factors that
are just very intuitive and what you would expect, that we look at the compliance costs
associated with the package of the rules, with the common effective date. We look at the scope of the requirements that
are being changed. We look at the potential impacts on the market. But always keeping in mind that, you know,
this is for determining whether we will do an assessment so we can’t like, do the assessment
in order to determine if we are going to do the assessment. We have to make that decision based on information
that we have, largely information that we did in the rulemaking, so it’s forward-looking
information, what our expectations were, along with whatever information may have come along
to us up to the moment before we say, yes this is a significant rule, because we could
make that determination from the moment we issued the rule or sometime afterward. Obviously, as the public now knows, we’re
making that determination a little bit later, which is in many ways a reasonable thing to
do; kind of look around and see what’s happening for saying, oh yeah, this rule is a significant
rule, we should put the resources in to assessing its effectiveness. Okay. Third bullet point, third element, we’ll see
certain recommendations from the public regarding the rule. That’s required by the statute. And the assessment shall address effectiveness
of the rule in meeting the specific goals of the rule as stated by the Bureau and the
purposes and objectives of the Bureau, which is stated in Title 10 of the Dodd Frank Act. I won’t repeat the purposes and objectives
of the Bureau Title 10, there in our Federal Register notices that we’ve issued, but they’re
things that you would — common sense things that the Bureau, you know, is organic to its
mission that, if you know about the Bureau it’s preventing unfair deceptive practices,
it’s making sure consumers have information. That kind of thing. General goals of the Bureau. And very critically, we’ll reflect available
evidence and any data that the Bureau reasonably may collect. And so, it’s this last part that really puts
assessments into the realm of a research exercise, and moving away from simply relying on say,
comments about the rule, which are opinions about the effectiveness of the rule. There’s another thing out there, which would
be data about the effectiveness of the rule. Opinions are important, but it’s not the scope
of our assessments by any means. And when we look around at individual significant
rules, in some cases we’ll have lots of data in-house already because we have a supervision
and examination functions, we have complaints, we have enforcement actions that bring in
information and evidence to the Bureau. But in some cases we won’t have very much,
in which case we go out to stakeholders and to the entities that we regulate and ask for
information so that we can fulfill this requirement. And we’re in the early stages of doing that
on the ones for which we’ve already made the determinations, sort of the outreach part. So, one other technical point I want to make,
whenever you talk about an assessment, the first thing people will think about is well,
like relative to what? Effective relative to what? What’s the baseline against which you’re making
comparisons. Within the economics literature in this kind
of work there is a sense, there’s a general perspective of, well, that’s kind of obvious. It’s like, what the world would be if not
for the rule, as if that’s a very — as if that’s a clear, even in principle, that that’s
like the only answer. It is part of the answer. It’s certainly partly of interest, and we
sort of get at that, you know, in various ways. Formally, you think about before the rule
data and after the rule data, and then from the before rule data, well, in the current
context what does that tell us about the current context, because before the rule isn’t the
world. If not for the rule, it’s just before the
rule. But it’s a data that can inform that choice. Within the methodologies of economics there’s
a lot of interest in methodologies around carveouts, so that that gives you a much cleaner
sense of a group of guys, a group of entities that aren’t regulated currently, you can compare
them to similar guides after the rule, and then you get this clean sense against what
are they doing if not for the rule. But there’s other baselines of interest, in
that our assessments would be thinking of, you know, in some cases a rule has a safe
harbor. We might be interested in safe harbor. It makes no sense to think about the impact
of the safe harbor absent the requirements from which it is a safe harbor being in effect. So, that’s sort of a partial impact, so we
have part of the rule in effect, and you ask about if some other part of the rule was or
wasn’t in effect. That’s a very interesting question. Doesn’t fit in with the sort of plain vanilla
generic approach of thinking about this, and there’s also — it can be of interest to think
about against a baseline of what we thought the world might be like, what our expectations
were, relative to what actually occurred. There’s also a very interesting discussion. There are other kinds of baselines against
which we’ll be making comparisons, driven by what is substantively interesting to us,
to our stakeholders, to informing policy. So, that is the overview of our methodology,
and I’ll spend less time talking about other slides. All right. So, on the requirements front, this is just
to note that not only have we been thinking about this for a long time, but we actually
have performance goals around this. Performance Goal 1.1.2. Thank you. All right. So, that will complete all five-year regulation
assessments on schedule. We have a current goal and for Fiscal Year
17, but if this is all online, there were goals set in earlier years. And so, in fact, as a sort of public administration
matter, this work began within the Research Office and then kind of built out to within
the division of research and markets regulations, and then built out to Bureau stakeholders
in general, getting input across the Board, throughout the Bureau, over time really going
back a number of years, and ultimately leading to final decisions with the director on how
we were going to go about this. And in fact, at an earlier meeting where the
director was present, he got to see these slides. He said, oh, I now understand why you were
coming to me so much. I’ve got the full picture finally, of what
was required, and it was a lot. And he, I think he concluded that we weren’t
wasting his time and that — and it justified the fact we’re going to have to come back
to him a couple more times. So, looking forward to that. Also, within the government accountability
office has written a couple of reports on our progress so far. One in June 2015, one in July 2016 that recognized
our planning efforts. They just said, complete your plans. You know, you’re going along great. Just, but let’s tie some of this off. This was before we had issued any public facing
work. And they also said that we ought to get public
input on our plans for assessment, but that’s a best practice, and we had been intending
to do some of that but we really took that, you know, focused much more cleanly on doing
that and doing it early in the process. And that’s why I’m standing here today to
tell you a little bit about this. So, we issued our initial announcements of
plans to assess two rules for which we determined that they were significant. One, our remittance transfer rule, and the
other our servicing rule that we implemented under the — pursuant to the real estate service
— to amendments to RESP, which is the — right, which is the Real Estate Settlement Procedures
Act. It’s there. I just needed a moment there. Okay. So, we have blog posts up on our website,
which link the Federal Register notices, and there’s a public docket, and people can comment
on our plan. And I’ll tell you about each of those briefly
now. So, the remittance rule, so remittance consumers
sending money overseas. Those are consumer remittance transfers. Not businesses, just consumers. And the Dodd Frank Act created a new set of
statutory requirements around that for which we then created rules. And these new consumer protections required
providers to give accurate disclosures to consumers before they pay for remittance transfer,
with certain exceptions, to permit cancellations and provide refunds, and to investigate disputes
and remedy certain errors. So, when we think about this rule, there was
a large disclosure component and there is a large error resolution component. And the key objectives were to improve the
predictability of remittance transfers, predictability for the consumer sending money abroad. It really did specify originally that in many
circumstances, if you’re sending money overseas you should know exactly how much in the foreign
currency is going to be received on the back end. It could also be in dollars as well, dollars
to dollars. But on the back end you have your fees and
taxes and everything is taken out. To provide consumers with better information
for comparison shopping, which is more relevant in some contexts than in others. If you’re sitting in a bank, your bank, to
do a wire transfer, that may not be as relevant as if you’re walking down the street looking
at money transmitter rates. And then to provide these protections to consumers
while limiting market disruptions, and we amended this rule before it took effect in
order to limit market disruptions that we felt that we had good information on, might
occur if it took effect as written. So, what we’re — in terms of measures, we
are always taking the perspective that one thing we’ll do in these reports is you know,
provide of course, high level statistics on how the markets evolve. You know, apart from the nitty gritty of causality
and was the rule doing it or other things doing that, people are going to want to know
like, what has been happening in the market. So, we’ll always talk about that, trends,
and the number and types of providers, the number of remittances sent, the price of transfers,
and just a good data collection exercise and make a good place for people to see what happened
in that market. And then, we will, through outreach and other
data collection, we’ll look at provider activities undertaken to comply with the rule. And we are, from the provider side, would
be looking at evidence of cost savings that were related to the 100 Transfers of Safe
Harbor, meaning that if you’re providing 100 or fewer transfers in a year, you are not
subject to the rule because you’re only subject if you’re providing transfers in the normal
course of business. How valuable has that been to providers. We’re interested in essentially to exceptions
to the accuracy requirements for disclosures and requirements on error resolution that
we built into the rule to limit market disruption, limits that we put in on requirements to provide
foreign language disclosures, which are very important to consumers but could also be very
expensive for providers, so there’s a calibration that has to be done there, and which we did. And we’re also, from the consumer side — and
that really is the focus, largely the focus of the rules is — and the focus of the assessments,
is to really try to get into the benefits of these rules for consumers. What have we seen in terms of utilization
of error resolution, right. And to some extent, to the extent that we
can, within the time frame that we’ve got, some understanding of how — the value of
the disclosures. So, it’s very much like it — you know, going
back to the requirements, to use evidence and data that’s reasonably available, that
will depend on the rule and actually the development of data assets within itself, that we can
tap into. So, that’s a fairly straight forward set of
possible measures, but we are of course developing building out others, and we have a whole team
for each one of these assessments. Each project has a team and RMR, Research
Market Regulations, working on it. So, we’ve asked the public for information. We’ve asked members of the public to comment
on the assessment plan, to provide data and information that would be helpful to executing
the plan, to provide data and information on the effectiveness of the remittance rule,
and meeting the objectives of the rule as stated by the Bureau and the purposes and
objectives of the Bureau, and to provide certain recommendations regarding the rule, the closing
date on this one, because when we issued it, 60 days after the Federal Registry notice
of the closing date is May 23rd, and I’m thinking I didn’t tell you when the report is due. Although, I told you when the effective date
was, so it’s five years after that. The report is due at the end of October in
2018. So, we’ve got time to do this well, and we’re
really quite committed to doing that. On the servicing rule, that’s a very different
kind of rule. This was a rule that was very much — there
are some disclosure elements to it, but it was very much about operation. Timelines for doing certain things, how you
must engage with consumers, primarily consumers who are facing — in some kind of distress,
either by they’ve missed some payments or they’re heading into — they need loss mitigation
on their mortgage. And it’s one of the more operationally focused
rules that we’ve written. This rule took effect on January 10th, 2014,
so the report will be due on that date in 2019. Again, we’re assessing effectiveness relative
to the, you know, the stated purposes of the rule that the Bureau stated in the rulemaking
as well as the purpose and objectives of the Bureau, so RESPA itself really says to servicers,
you shall maintain and provide accurate information. You shall help borrowers avoid unwarranted
or unnecessary costs and fees. And you shall facilitate review for foreclosure
— well. And then we, building on that, added, said,
consistent with the purposes of RESPA will facilitate review for foreclosure avoidance
options. RESPA is a remedial consumer protection statute,
and our rules built off of those purposes of the statute in stating what we were trying
to achieve. So, the avoiding you know, avoidable fees
and avoidable foreclosure is a great concept to have in mind in thinking about the purposes
and things we might try to measure. And of course, what is an — we know what
a fee you paid, was it avoidable is the trickier concept. But we’ve put a lot of thinking into that,
into how we might be getting at that ultimate approximating that. So, the RESPA servicing rule requires, specifically
requires mortgage loan servicers in part to provide disclosures to borrowers related to
forced place insurance, which is the lender placed insurance, or when they don’t have
proof that you have in place, hazard insurance as you’re required to under your mortgage
contract. Respond to errors asserted by borrowers in
a timely manner. And follow certain procedures related to loss
mitigation applications and communication with borrowers. This is a large rule with many, many moving
parts. And one of the things we did over the period
in building up to this release, was to decide you know, what would our focus be to try to
give the investigation shape? We cannot do everything, and decide, to make
these decisions about what would our focus be. Of course, if great information, if great
information — if evidence of great — if we learn that there’s great information on
something important that we didn’t consider, you know, we’re interested in hearing about
that. But we’ve really tried to think hard about,
you know, what was this rule of statute trying to accomplish, what were we doing in that
environment, loss mitigation, error resolution, and avoidance of unnecessary forced placed
insurance, were really, really key at the time. So, those are the things that we’re trying
to look for. And this is not a very — this is not a very
— this is a single slide that tries to capture a lot of information, so it’s not a great
slide. But wanted to just give you a little flavor
of the kind of measures that we’ve been thinking about. So, you know, there is servicer compliance
activities, such as responding to loss mitigation application. Are they hitting the timelines that we’ve
— what are we finding about the timelines that services are hitting, and to bar notices
of error on these prescribed timelines. And, you know, to the extent that there is
variation there, even within the rules, even within the law, we can then kind of use that
variation to sort of then look at outcome. Sorry, to look at outcomes, various kinds
of outcomes related to that. And that’s one way of starting to get at this
— start to pick out a little bit around causality. On the consumer side, we’re looking for good
granular data on consumer activities, such as utilization of the rights provided by the
rule. Are they asserting errors and submitting loss
mitigation applications and actions prompted or enabled by the rules, such as are they
making additional payments in response to prompts, disclosure prompts. So, hey, we’re going to impose — we have
no evidence you have hazard insurance. Does that produce a spike in payment? You know, for people who — does that produce
a spike in payment off trend that would be sort of like, yeah, there was an avoidable
case here because we got a bump up on relative to a timeline. Of course, we can’t prove in that methodology
whether they might not have eventually paid. But we’re seeing a response to the disclosure. So again, it’s careful — in each case, specifying
the counter factual — look at the facts and circumstances, be real clear about that. You know, what will the evidence show? Additional payments or other consumer response
after early intervention by the servicer or consumer — after early intervention by the
servicer or consumer verification and hazard insurance. So, the early intervention by the servicer
is an actual — sorry, rule mandated requirement here, and once you’re in default for a certain
period that servicers will jump in and do a lot of outreach to you. Is that working? And in response to that, did we see cure? Did we see more orderly resolution of delinquency? You know, sometimes foreclosure or short sale
are the other options besides cure there. But having it take forever, relative to some
timelines that we take a long time, are those coming closer to what is the right timeline
so unnecessary fees don’t accumulate so that, you know, and there is actually some research
in the literature as to what is a good timeline for consumers who are in distress for final
resolution. Forever is certainly not the right answer. And so, we’re interested in these kinds of
changes that we might be observing in response to specific activities in the rule. And also, relative to before data that we
will have some — we will likely be able to obtain some of, from servicers. So, to the extent possible we will be relating
the above activities to consumer outcomes if the rule is sought to effect, including
fees and charges assessed and paid, the instance and severity of delinquency, and how delinquency
is resolved, including time to delinquency where resolution is cure, foreclosure, or
short sale. So, we’ve put a lot of thought into these
measures, but of course it’s a dynamic process because we go out we learn about what data
is reasonably available and some investigation, we’ll have to refine the investigations and
what we can do based on what we can actually gain information on. So, it’s this constant back and forth, which
is another real-world aspect of this kind of work, retrospective review, that doesn’t
— isn’t really in the text books. So, that’s going to be a very interesting
process, and that’s why we have teams working on this because not just — it’s not just
written down and then it rolls forward. It’s definitely a dynamic and something that
requires a lot of work. As I say internally, nobody’s full time job
is these assessments on these teams, but we’ve got an ongoing need for these resources. So, part of my job is to keep those resources
there and available for this work. So, we’ve asked the members of the public
the same thing for this remittance, as for the remittance rule, to comment on the assessment
plan, to provide data and information that would be helpful to executing the plan, to
provide data and information on the effectiveness of the, in this case, the RESPA servicing
rule, in meeting the objectives of the rule, and to provide certain recommendations regarding
the rule. The closing date on that is July 10th, so
a lot more time if anybody wants to write. All right. So, that’s the overview that I was asked to
provide of our assessment work and I’m happy to take questions, comments. So, thanks very much, Paul. As Paul mentioned, the Bureau is intentionally
committed to these retrospect reviews since the assessments of the rules, and we have
now asked through the two Federal Register notices, for information from data for the
public, which is a key ingredient to all of this. The reason that we decided and thought it
would be useful to discuss this this morning with the ARC members, is that the other input
to this will be any learnings that we can have from the literature from other places,
the cost benefit literature, the retrospective review literature, the evaluation literature. We have a lot of expertise in-house. We can always learn more. And if there are particularly good examples
that they have, or particular areas that they think we should look, or we should dig, or
particular methodologies we should use, we’re going to be interested in having those conversations
over the next 18 months as we’re engaged in these activities. And as Paul mentioned, October 28th of 2018
will be the first report, and five years later for the RESPA report is early in 2019. So, it’s a lot of work to do over the coming
months. I don’t know if there were any reactions from
the ARC members to some of this material and to sort of the, I would say, the large and
optimistic task that we’ve set in front of us. That’s a good characterization. Well, since you asked. So, first of all, I would just — I’m very
impressed at how much thought is going into this process. And particularly impressed that for reports
that aren’t going to be due for another 18 months, or you know, almost two years, that
the ducks are already in a row to get these done in a timely fashion. The one caution that I would give to you in
thinking about the impact of these regulations and in writing up the assessment is that there
are things that are easy to quantify, and there are things that are more difficult to
quantify. And then there are things that are fundamentally
impossible to quantify. And I’m speaking with my behavioral economist
hat on now, but it’s really easy to focus on the things that are easy to quantify, precisely
because they are easy to quantify. But they are not the only things that matter. So, I would just issue a caution to you know,
give due recognition to things that might matter, even if they’re difficult to measure. And I think in the domain of much of what
the CFPB does, you know, an important factor that’s difficult to quantify is the reduction
in stress, and the mental health issues that individuals feel around financial decision
making. So, there’s, you know, some interesting research
that has been done in Oregon on the impacts of providing health insurance to lower income
individuals who didn’t previously have it. And one of the interesting findings from that
research is that having health insurance, it turns out, you know, has big mental health
benefits. People are not as stressed out when they have
insurance and there’s lots of evidence that you know, from and others, this idea of scarcity
that when people have limited resources, that creates a lot of financial stress, and that
carries over to a lot of decisions that people make. And so, any rules that are reducing the number
of phone calls you have to make, or you know, reducing the number of errors and things that
are going on, or enabling you to stay in your home, anything that’s reducing stress is probably
going to have positive benefits, but things are that are going to be really hard to quantify. And I think there are other factors like that
and just make sure those get their due notice, even if you can’t measure them. Right. And one thing that’s — that’s great. It is useful to us in many ways that we’ve
got this language around assessing effectiveness relative to goals, you know, stated in the
rulemaking or in the purposes and objectives of the Euro. That’s both expansive, but also provides some
protection. Another way we could get, like you said, easy
to measure but aren’t the right things, is if we just focused on, for example, compliance
costs. That’s not what we’re required to do, but
that’s relatively easy to measure, and a lot of that information you know, can be collected. What you’ve described is a type of benefit,
which are typically the harder ones, but we are mandated to focus on that because, in
part, because the purpose and objectives includes that, and that’s a great framing. How would you think about the pre — reduction
of stress requires a baseline measurement of stress. How would one go about assessing what was
the — or assessing what was the level of stress pre-rule in order to think whether
we have impacted that level of stress? Yeah, I might have to think more about that. I mean, if you really had all your ducks in
a row you would have done that you know, three or four years ago. But, you know, there may be some — and there
are lots of national surveys that collect measures of stress, and you know, maybe possible
to identify population, a subpopulation most likely to be impacted by these rules, and
then do some tracking over time. But I’m just speaking off the top of my head. Yeah. I was going to put Justine on the spot given
her work in Rhode Island, if there’s any insights into policy evaluation or the data necessary
for such tasks, or generic advice as we embark on these projects. Sure. So, in Rhode Island we’ve started at Brown
University, the Rhode Island Innovative Policy Lab, which is a foundation funded partnership
with the Office of the Governor, to use data and science to improve policy and equity of
opportunity. And as such, one of the things that we’ve
been looking at are the impacts of policies and personal finance on measures of stress
in the household, which could manifest themselves in things such as crime rates. They could also manifest themselves with misbehavior
of children at school. An interesting project that we’re working
on with the Office of the Governor and USDA is examining the impact of the food stamp
consumption cycle on many measures of household’s wellbeing. And one of the topics that came up yesterday
in kind of survey measures, was measuring — using new technologies to measure outcome. So, for example, we’ve piloted a program that
uses text messages on phones to collect high-frequency data on things like stress. One of the things that we found in our pilot
is that kind of perceived misbehavior and stress of children appears to go up at the
end of the month among food stamp households in Rhode Island. This is consistent with some of the patterns
one might see with truancies and misbehavior at schools. And of course, measuring this leads to policy
recommendations, policy trials, and then the ability to measure the policy impact. And obviously, since these data are collected
and are being used for research driven policy in Rhode Island, to the extent that we can
harness that to inform national policies that also impact Rhode Island, we’re ready and
happy to do that because we will have better policy and we will have better policy impact
and increased equity of opportunity if we can actually use data to think about what
are the problems that policy can address, and then measure the effectiveness of policy
afterwards instead of just kind of implementing it and hoping that our aims actually became
reality. That’s very helpful. I’m thinking about these new technologies
and Brigitte is warning that we should have started but we — you know, we issued this
rule 18 months after the Bureau was born. This would have been a case where we would
have started our data collection. It’s 18 months before we were born, which
would have been particularly hard. John, I’m wondering if you have any insights
into this issue of intangible benefits and benefits to consumers. If anything comes off the top of your head,
or as a first cut, since I’m putting you on the spot from the marketing recyc literature,
we can think about some of the intangible benefits to consumers that may be a little
harder to quantify. Well, since you just brought up the issue
of consumer sense of financial stress, we have in our group, some research that’s building
on some CFPB research, or the CFPB has developed this scale for measuring consumer financial
wellbeing and we’ve been taking that in the direction of separating out indicators of
current stress from things that have to do with presumptions in more long run security. And so, I would say that if there are issues
in the future where current stress is a kind of an intangible you want to measure, we have
some help based on your own research. It is true. What John’s alluding to is the CFPB’s measures
to financial wellbeing, which we’re quite excited about and I’m glad to see are being
used more broadly and could become a baseline if we start a measure on a nationally representative
set of measures in that regard. I don’t know if there are any other final
comments before we wrap this up, but like I did say, we’ll be bothering you guys over
the coming years to help with this as we go through this effort. If there are any final words about this or
about anything that you heard in the last 24 to 36 hours that we’ve all been together
from any of the ARC members. All right. Then, fantastic. Thank you all very very much. I will call us adjourned for now until next
year’s annual meeting, and want to thank everybody for their participation yesterday, for today,
for the robust conversation, and for what I’m sure will be a lot of valuable input to
our work going forward. So, many thanks.

Leave a Reply

Your email address will not be published. Required fields are marked *