Popular in Social Sciences

Banks and Policy Organisations

Macroeconomic modelling of CBDC: a critical review

European Central Bank Ulrich Bindseil

Social Identities and Divisions

A new explanation of antisemitism: Jew hatred as a civilisational phenomenon

Boston University Liah Greenfeld

Banks and Policy Organisations

Financing Private Credit

Federal Reserve Bank of New York Nina Boyarchenko

Economics

Failing Banks

Massachusetts Institute of Technology Emil Verner

Culture and Society

The City as a Growth Machine: Toward a Political Economy of Place

New York University Harvey Moltoch

American Politics

Keeping the Republic

Boston College Marc Landy

British Politics

Entrepreneurial thinking: the politics and practice of policy impact

University of Sheffield Matthew Flinders

Consumer Credit Reporting Data

View more insights in Social Sciences - Economics - Financial Economics

View more figures from University of Illinois at Urbana-Champaign

View the page of Jialan Wang

Read the Study

Next in Economics

Corporate foreign bond issuance and interfirm loans in China

Ugo Panizza | Geneva Graduate Institute

Offshoring and non-monotonic employment effects across industries in general equilibrium

Daniel Baumgarten | Universität Paderborn

Are you curious about how consumer credit reporting data shape economic research and policy decisions? Discover the transformative role of credit files in understanding borrowing behaviors, economic measures, and the impact on financial markets.

Image courtesy of interviewee. November 15, 2024

Log-in or Sign-up to Faculti
Currently viewing this subject insight as a guest. You have insight(s) remaining for this month. Login to view 8000+ figures on the platform.

Generating Workspace...

Export as PDF

This video discusses the significance and development of consumer credit data, particularly within the context of the United States. It highlights the role of major credit reporting agencies—Equifax, TransUnion, and Experian—in collecting and selling data on individuals' credit habits, which lenders use to assess credit risk.

This video points out that the research into consumer credit data surged following the 2008 global financial crisis, which prominently involved the housing market. Important studies emerged in this field, leveraging credit data to analyze the crisis's impact and ongoing implications for the economy. The speaker emphasizes that while consumer credit data began mainly for lending purposes, its applications have expanded into various research domains, including health and economic mobility.

However, there are limitations and challenges associated with this data. This video notes that the credit reporting system is not comprehensive, missing significant populations like minors and undocumented individuals. Issues arise from the reliance on past credit behavior to assess current creditworthiness, which may overlook individuals who have never needed credit. Additionally, the quality of the data can be inconsistent due to varied reporting practices and changes in regulations, especially highlighted during the pandemic, which affected the accuracy of credit records.

The speaker advocates for better methodologies in using this data for research, noting the need to link it with other datasets for a more complete understanding of financial well-being. Overall, the discussion underscores the complexity and evolving nature of consumer credit data in both academic research and practical applications, pointing to the necessity of ongoing refinement and awareness of its limitations.

Language:

Export as PDF

X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:90000

1
00:00:00.145 --> 00:00:03.925
Is a data source that is increasingly become very widely

2
00:00:03.925 --> 00:00:07.485
used in financing and economics and what it comes from.

3
00:00:07.705 --> 00:00:09.685
And, I think a research paper

4
00:00:09.755 --> 00:00:12.845
that accompanies this interview on consumer credit reporting

5
00:00:12.845 --> 00:00:15.005
data, we primarily focus on the United States,

6
00:00:15.145 --> 00:00:18.245
but I wanna note that similar data sets are available in

7
00:00:18.475 --> 00:00:21.805
many other countries, including developed countries such

8
00:00:21.805 --> 00:00:24.245
as the UK or Europe as well as developing markets.

9
00:00:24.345 --> 00:00:25.445
But my expertise

10
00:00:25.465 --> 00:00:28.405
and our research paper are gonna focus primarily on the us

11
00:00:28.405 --> 00:00:30.405
So I will, I will focus on that for now.

12
00:00:30.745 --> 00:00:32.925
But essentially it's data collected

13
00:00:32.925 --> 00:00:35.685
by private firms in the US There's three major firms,

14
00:00:36.515 --> 00:00:39.525
Equifax, TransUnion and Experian.

15
00:00:39.985 --> 00:00:43.645
And they're collecting, information about our habits

16
00:00:44.185 --> 00:00:47.205
and, uses of credits such as taking out mortgages,

17
00:00:47.505 --> 00:00:49.205
credit cards, auto loans,

18
00:00:49.225 --> 00:00:51.485
and the such, such as any typical type of credit

19
00:00:51.485 --> 00:00:53.845
that we might think of, gathering that information together

20
00:00:54.385 --> 00:00:57.405
and then essentially selling it to lenders

21
00:00:57.465 --> 00:00:59.805
and other institutions that might want to use that data.

22
00:00:59.825 --> 00:01:03.405
In particular, thinking about if a new bank wants to lend

23
00:01:04.005 --> 00:01:05.005
a credit card to you

24
00:01:05.105 --> 00:01:06.965
or a mortgage, they want to see that data

25
00:01:06.965 --> 00:01:08.085
because they want to know whether

26
00:01:08.085 --> 00:01:09.365
you're a good credit risk or not.

27
00:01:09.385 --> 00:01:12.565
So essentially these firms collect this data,

28
00:01:12.745 --> 00:01:16.165
and then package it for essentially a huge fraction

29
00:01:16.165 --> 00:01:17.445
of the US population,

30
00:01:17.665 --> 00:01:20.725
and then resell it on to financial institutions

31
00:01:20.725 --> 00:01:23.005
and other companies that may be interested in using

32
00:01:23.005 --> 00:01:24.165
it for business purposes.

33
00:01:24.315 --> 00:01:27.205
Certainly the research community has researched,

34
00:01:27.205 --> 00:01:29.205
individuals, experiencing things such

35
00:01:29.205 --> 00:01:30.565
as the global financial crisis.

36
00:01:31.025 --> 00:01:34.605
And I would really trace back the origins of modern,

37
00:01:34.825 --> 00:01:38.205
or let's just say post-crisis household finance itself

38
00:01:38.225 --> 00:01:41.565
as a research field back to the use of consumer credit data.

39
00:01:41.745 --> 00:01:44.045
Um, thinking about a seminal papers such as those

40
00:01:44.065 --> 00:01:45.165
by Ian and Sufi.

41
00:01:45.385 --> 00:01:46.685
The earliest papers,

42
00:01:46.685 --> 00:01:50.205
in 2009 really dissecting this crisis is something that

43
00:01:50.745 --> 00:01:51.765
hit the global economy.

44
00:01:52.305 --> 00:01:54.765
In some ways, I think we're still suffering the

45
00:01:54.765 --> 00:01:57.205
repercussions of it politically and economically.

46
00:01:57.505 --> 00:01:59.525
Um, and, and relatively quick time.

47
00:01:59.525 --> 00:02:02.925
Their original paper was published in 2009 using credit

48
00:02:02.925 --> 00:02:04.485
reporting data because the heart

49
00:02:04.485 --> 00:02:06.965
of the global financial crisis, as we know,

50
00:02:07.065 --> 00:02:08.405
was the US housing market.

51
00:02:08.585 --> 00:02:12.165
And we realized that we didn't know much about this sector.

52
00:02:12.585 --> 00:02:15.045
The household sector in the United States is about two

53
00:02:15.045 --> 00:02:16.245
thirds of the US economy.

54
00:02:16.545 --> 00:02:20.005
Of course, the US is key in the global economy

55
00:02:20.065 --> 00:02:21.405
and the global financial system

56
00:02:21.585 --> 00:02:25.165
and fragility in the housing sector,

57
00:02:25.345 --> 00:02:27.645
in the consumer sector was what led

58
00:02:27.645 --> 00:02:29.885
to the cascading problems globally.

59
00:02:30.185 --> 00:02:33.125
And, me and Sufi were some of the first researchers put,

60
00:02:33.125 --> 00:02:34.565
of course, if you wanna study the mortgage

61
00:02:34.565 --> 00:02:35.845
market, you need credit data.

62
00:02:36.105 --> 00:02:39.485
And so they used this exact data source to shed some

63
00:02:39.485 --> 00:02:41.285
of the first light on the crisis.

64
00:02:41.425 --> 00:02:44.685
And I would say that we are still studying the financial

65
00:02:44.685 --> 00:02:46.685
crisis of 2008 and 2009.

66
00:02:47.105 --> 00:02:50.645
Uh, UAN keys is another example looking at refinances

67
00:02:50.645 --> 00:02:52.725
and equity extraction even into the 2010.

68
00:02:52.725 --> 00:02:54.805
So the Buddha Keys paper was from 2016

69
00:02:55.145 --> 00:02:56.285
and even the 2010s,

70
00:02:56.285 --> 00:02:59.165
I would say there's very important pathbreaking research.

71
00:02:59.585 --> 00:03:00.925
One example is, Ong

72
00:03:00.925 --> 00:03:04.765
and Noel from 2020 looking at mortgage restructuring back,

73
00:03:04.865 --> 00:03:06.325
during the financial crisis.

74
00:03:06.425 --> 00:03:08.805
So I think we are actually still studying,

75
00:03:08.875 --> 00:03:11.925
this phenomenon and still using credit data,

76
00:03:11.945 --> 00:03:13.565
to into this current decade.

77
00:03:13.885 --> 00:03:15.765
I would say that along with the development

78
00:03:15.765 --> 00:03:17.325
of consumer finance as a field,

79
00:03:17.445 --> 00:03:20.245
which again was a bit catalyzed by the financial crisis,

80
00:03:20.685 --> 00:03:23.805
I think that taking a step back a second,

81
00:03:24.125 --> 00:03:27.445
consumer credit data did pre-exist from the,

82
00:03:27.545 --> 00:03:28.725
financial crisis?

83
00:03:28.825 --> 00:03:30.645
And we have seen development

84
00:03:30.645 --> 00:03:32.925
of this data source in our research paper on consumer credit

85
00:03:32.925 --> 00:03:35.605
data this year does seek to advance like

86
00:03:35.605 --> 00:03:38.885
how we're using this data, help inform researchers about

87
00:03:38.885 --> 00:03:40.605
what are new ways to use this data

88
00:03:40.605 --> 00:03:42.165
that haven't really even been used before.

89
00:03:42.465 --> 00:03:45.045
But I think taking a step back, seminal work

90
00:03:45.145 --> 00:03:46.885
by the New York Federal Reserve, back

91
00:03:46.885 --> 00:03:49.205
before the financial crisis in the early two thousands,

92
00:03:49.205 --> 00:03:51.885
they really helped to broaden all the,

93
00:03:51.885 --> 00:03:53.565
economists within the Federal Reserve system.

94
00:03:53.565 --> 00:03:56.045
Hundreds of economists being able to do great work.

95
00:03:56.265 --> 00:03:58.685
So some of our co-authors, Wilbert Vander Klau

96
00:03:58.685 --> 00:04:02.045
and Don Lee really pioneered the New York Fed data set,

97
00:04:02.045 --> 00:04:04.405
which was the first widely available credit data

98
00:04:04.465 --> 00:04:05.605
set in the two thousands.

99
00:04:05.825 --> 00:04:08.925
But this type of database all the way back to the 1990s

100
00:04:08.925 --> 00:04:12.085
and even before with, the Federal Reserve system,

101
00:04:12.305 --> 00:04:15.405
the FTC and other primarily US government agencies.

102
00:04:15.425 --> 00:04:18.005
So I wanna make sure to credit these earlier waves

103
00:04:18.005 --> 00:04:20.885
of researchers, to not say kind of in finance.

104
00:04:21.065 --> 00:04:24.005
We sometimes have a short memory about economic history

105
00:04:24.005 --> 00:04:25.485
and even the history of our own field.

106
00:04:25.705 --> 00:04:27.645
But I think what we seek to do, me

107
00:04:27.645 --> 00:04:31.205
and my group of of co-authors, we seek to kind of push

108
00:04:31.365 --> 00:04:33.085
that advancement further,

109
00:04:33.315 --> 00:04:35.405
into the current era in, in the 2020s.

110
00:04:35.665 --> 00:04:37.725
And there's a lot of new developments. I would say.

111
00:04:37.725 --> 00:04:40.205
We are really only scratching the surface of our ability

112
00:04:40.225 --> 00:04:42.405
to use consumer credit data,

113
00:04:42.625 --> 00:04:44.445
and, to study a wide variety.

114
00:04:44.445 --> 00:04:45.925
And you think what we've seen in the past five

115
00:04:45.925 --> 00:04:48.285
to 10 years is not only using this data to study

116
00:04:48.825 --> 00:04:51.285
the obvious things such as the housing crisis

117
00:04:51.345 --> 00:04:54.085
and the financial crisis, but things not even that related

118
00:04:54.105 --> 00:04:56.125
to finance that you might think of such as health,

119
00:04:56.295 --> 00:04:57.845
healthcare, health insurance.

120
00:04:58.045 --> 00:05:00.965
I think one of the great research papers that I can imagine

121
00:05:01.225 --> 00:05:04.405
so important is linking, credit data

122
00:05:04.405 --> 00:05:06.685
with abortion records and looking at women

123
00:05:07.025 --> 00:05:09.285
who were either denied an abortion

124
00:05:09.505 --> 00:05:11.045
or we're allowed to have an abortion.

125
00:05:11.185 --> 00:05:13.725
How does that affect their general economic,

126
00:05:14.025 --> 00:05:16.405
of course health wellbeing as well as financial wellbeing?

127
00:05:16.425 --> 00:05:19.005
So I think these data have well advanced well

128
00:05:19.005 --> 00:05:21.565
beyond its origins in studying the financial crisis.

129
00:05:22.045 --> 00:05:26.525
I think that, the origin of what is consumer credit data

130
00:05:26.705 --> 00:05:29.405
and the reason this industry existed, it is a kind

131
00:05:29.405 --> 00:05:30.565
of a private data set.

132
00:05:30.565 --> 00:05:32.365
Um, these firms, it rose in the private

133
00:05:32.365 --> 00:05:33.685
market to serve a need.

134
00:05:33.865 --> 00:05:36.925
And that need was information asymmetry such

135
00:05:36.925 --> 00:05:38.365
that if you're a new bank

136
00:05:38.545 --> 00:05:40.645
and you are faced with the consumer, you've never seen

137
00:05:40.645 --> 00:05:42.485
that person before, it's very

138
00:05:42.485 --> 00:05:43.645
difficult to lend to that person.

139
00:05:43.745 --> 00:05:47.005
And we've seen an explosion the US credit markets,

140
00:05:47.065 --> 00:05:49.685
and some would say too much credit, which was what led

141
00:05:49.685 --> 00:05:50.925
to the, to the crisis.

142
00:05:51.305 --> 00:05:54.325
But certainly the development of this industry,

143
00:05:54.425 --> 00:05:56.965
was exactly to address information

144
00:05:57.045 --> 00:05:58.045
Asymmetry issues. And

145
00:05:58.045 --> 00:06:00.565
I think one aspect I would point to

146
00:06:00.565 --> 00:06:04.005
that's very interesting is thinking about linking

147
00:06:04.545 --> 00:06:05.565
per personal credit

148
00:06:05.565 --> 00:06:07.805
where we have this very well developed consumer credit

149
00:06:07.805 --> 00:06:10.045
reporting systems, or essentially any form

150
00:06:10.045 --> 00:06:11.325
of consumer debt in the US

151
00:06:11.325 --> 00:06:13.205
and many other countries is reported.

152
00:06:13.345 --> 00:06:16.485
So if I'm a new lender, whether I'm a large bank

153
00:06:16.665 --> 00:06:20.085
or a new FinTech firm, if there there's a customer who comes

154
00:06:20.085 --> 00:06:21.565
to me, I can pull their credit record

155
00:06:21.625 --> 00:06:25.045
and get very good visibility into their entire history,

156
00:06:25.045 --> 00:06:26.245
whether they're delinquent or not.

157
00:06:26.585 --> 00:06:29.725
So the current credit reporting system has solved a lot

158
00:06:29.725 --> 00:06:31.525
of the information is asymmetries, not

159
00:06:31.525 --> 00:06:33.045
that they're completely eliminated.

160
00:06:33.425 --> 00:06:35.405
Uh, so it's still the case that moral hazard,

161
00:06:35.745 --> 00:06:38.725
if I compare two people with the same credit score,

162
00:06:38.735 --> 00:06:40.165
there could be differences between them,

163
00:06:40.265 --> 00:06:41.805
but it's quite well developed.

164
00:06:42.025 --> 00:06:45.805
But in contrast to the small business le lending market,

165
00:06:45.865 --> 00:06:47.725
has very little information

166
00:06:47.825 --> 00:06:50.885
and we know that, the avail available information,

167
00:06:50.885 --> 00:06:53.445
if I'm a new entrepreneur opening, let's say a restaurant

168
00:06:53.785 --> 00:06:56.245
or a small manufacturing firm, it's much more difficult

169
00:06:56.245 --> 00:06:57.325
to get information on that.

170
00:06:57.345 --> 00:07:00.365
And some of my work, which links the consumer credit data

171
00:07:00.785 --> 00:07:04.045
to small business credit data, what I find is that in fact,

172
00:07:04.045 --> 00:07:06.445
small business lenders are using people's consumer records,

173
00:07:06.985 --> 00:07:09.245
to lend to their businesses

174
00:07:09.245 --> 00:07:12.605
because of the fact that they don't know, they don't know

175
00:07:12.705 --> 00:07:14.125
how to, underwrite

176
00:07:14.125 --> 00:07:15.725
or they don't know how to evaluate the risk

177
00:07:15.725 --> 00:07:19.325
of those businesses because, essentially that industry

178
00:07:19.625 --> 00:07:23.085
to manage small business credit risk has not developed

179
00:07:23.105 --> 00:07:25.005
as much as on the the consumer side.

180
00:07:25.005 --> 00:07:27.645
So there's actually borrowing a person's personal credit

181
00:07:27.665 --> 00:07:30.365
record in order to lend to their business.

182
00:07:30.425 --> 00:07:32.445
And there can be various issues with that,

183
00:07:32.465 --> 00:07:33.485
that you might imagine.

184
00:07:33.865 --> 00:07:37.325
My personal credit score may not dictate the quality

185
00:07:37.385 --> 00:07:40.805
of my new business that I, that I opened.

186
00:07:40.865 --> 00:07:42.725
But, and, and we know that minorities

187
00:07:42.825 --> 00:07:45.005
and women tend to have lower credit scores, so then

188
00:07:45.005 --> 00:07:47.685
that means there had gonna have a harder time starting up

189
00:07:47.685 --> 00:07:50.605
new businesses and that can impact the broader,

190
00:07:50.605 --> 00:07:53.205
macro economy and our business dynamism.

191
00:07:53.385 --> 00:07:54.445
Um, in many countries.

192
00:07:54.915 --> 00:07:58.405
This was really the core goal of our research paper,

193
00:07:58.615 --> 00:08:02.325
which is to shed light on ways of measuring and,

194
00:08:02.325 --> 00:08:03.885
and kind of advance best practices

195
00:08:04.225 --> 00:08:07.285
and hopefully generalize methodologies

196
00:08:07.365 --> 00:08:08.565
that we can all then share

197
00:08:08.625 --> 00:08:10.245
to advance the literature in that way.

198
00:08:10.245 --> 00:08:12.365
Because essentially what we've seen in the development

199
00:08:12.365 --> 00:08:14.165
of the literature is that, this

200
00:08:14.685 --> 00:08:17.205
consumer credit data became first became widespread in the

201
00:08:17.205 --> 00:08:19.605
two thousands and in the 2010s

202
00:08:19.605 --> 00:08:20.925
and 2020s, it kind of broadened out

203
00:08:20.925 --> 00:08:23.005
to many other economic fields, including researchers

204
00:08:23.305 --> 00:08:27.405
who their main expertise may be in public economics or labor

205
00:08:27.505 --> 00:08:29.365
or health and not be in credit per se,

206
00:08:29.365 --> 00:08:30.565
but they're using this type of data.

207
00:08:30.905 --> 00:08:34.805
And I think what we're hoping in the 2020s is to consolidate

208
00:08:34.805 --> 00:08:37.285
that and say, okay, we've had this mass proliferation

209
00:08:37.285 --> 00:08:39.685
of research, now let's, kind

210
00:08:39.685 --> 00:08:41.085
of compare different methodologies.

211
00:08:41.135 --> 00:08:43.845
Let's, let's create a frontier best practices

212
00:08:43.845 --> 00:08:46.885
so then we can all advance starting from the same basis

213
00:08:47.105 --> 00:08:49.525
of these are the best practices for measurement.

214
00:08:49.945 --> 00:08:51.925
And there's a few things to remember in that.

215
00:08:51.925 --> 00:08:53.245
As I mentioned at the very beginning,

216
00:08:53.375 --> 00:08:55.925
these data were not created for researchers.

217
00:08:56.315 --> 00:08:58.485
This was, created for business purposes,

218
00:08:58.905 --> 00:09:01.285
and those might not be exactly the same as

219
00:09:01.285 --> 00:09:04.605
what we might want from an economic measurement perspective.

220
00:09:05.105 --> 00:09:08.765
And there's very basic things such as even the definition of

221
00:09:08.915 --> 00:09:13.085
what is a credit card, they don't have, necessarily,

222
00:09:13.665 --> 00:09:15.445
not all of these data sets have an indicator

223
00:09:15.505 --> 00:09:17.405
for this is a credit card versus this is not.

224
00:09:17.545 --> 00:09:19.205
And a researcher has to do a lot

225
00:09:19.205 --> 00:09:21.765
of the legwork in creating these measures by themselves.

226
00:09:21.825 --> 00:09:24.885
So we hope to kind of, both for people

227
00:09:24.945 --> 00:09:26.285
who are not involved in credit data,

228
00:09:26.305 --> 00:09:28.005
but maybe reading these research papers,

229
00:09:28.005 --> 00:09:29.965
including the general public, they may not know

230
00:09:29.965 --> 00:09:31.925
that even the definition

231
00:09:31.925 --> 00:09:34.325
of a credit card can be subjective decision.

232
00:09:34.745 --> 00:09:36.285
And that researchers,

233
00:09:36.625 --> 00:09:39.645
can differ across two different research groups, in,

234
00:09:39.645 --> 00:09:42.045
in terms of how they make those subjective decisions.

235
00:09:42.265 --> 00:09:43.645
One thing to remember is

236
00:09:43.645 --> 00:09:46.045
that even though it has very broad population coverage in

237
00:09:46.045 --> 00:09:49.765
the United States, for example, over 200 million us

238
00:09:49.765 --> 00:09:51.365
adults have credit records,

239
00:09:51.835 --> 00:09:53.685
that also means that some people are missing.

240
00:09:53.825 --> 00:09:57.725
For example, obviously anyone under 18 is essentially not

241
00:09:57.725 --> 00:09:59.325
gonna be covered by this dataset.

242
00:09:59.345 --> 00:10:01.645
So even though they're very interesting questions about

243
00:10:01.745 --> 00:10:04.645
intergenerational, issues with economic

244
00:10:04.785 --> 00:10:06.925
and credit kind of access

245
00:10:07.345 --> 00:10:10.085
and mobility, we would not see the children

246
00:10:10.375 --> 00:10:11.525
until they turn 18.

247
00:10:11.585 --> 00:10:14.365
And of course, the age at which someone first gets a credit

248
00:10:14.365 --> 00:10:16.645
card varies by their parental income.

249
00:10:16.745 --> 00:10:20.685
And so we, even if we do see someone, we may see,

250
00:10:20.925 --> 00:10:22.925
a different selection of people who have credit

251
00:10:22.925 --> 00:10:25.765
by the time they're 18 versus do not enter the credit market

252
00:10:25.815 --> 00:10:27.325
until they're in their twenties.

253
00:10:28.105 --> 00:10:30.845
And so that is very important to take into account.

254
00:10:30.845 --> 00:10:33.485
And for example, this data is missing a lot

255
00:10:33.485 --> 00:10:36.765
of very important populations for policy purposes,

256
00:10:36.765 --> 00:10:38.405
including the most vulnerable populations.

257
00:10:38.405 --> 00:10:40.085
So people without social security numbers.

258
00:10:40.585 --> 00:10:44.045
Um, some data sets explicitly sample off of those

259
00:10:44.435 --> 00:10:45.725
with social security numbers.

260
00:10:45.825 --> 00:10:48.765
So obviously people who are have limited,

261
00:10:49.005 --> 00:10:52.325
documentation in terms of their, their employment status,

262
00:10:52.545 --> 00:10:54.685
or visa status in the United States,

263
00:10:54.685 --> 00:10:56.805
they might be very important for policy purposes,

264
00:10:56.875 --> 00:10:58.205
they would not be very covered.

265
00:10:58.465 --> 00:11:01.325
Or the other side, I think one of the contradictions

266
00:11:01.325 --> 00:11:04.605
that we do get into in this research paper is

267
00:11:04.605 --> 00:11:07.845
that the way the the industry has developed essentially over

268
00:11:07.845 --> 00:11:09.525
the course of the 20th century was

269
00:11:09.525 --> 00:11:11.445
that we measure your credit worthiness by

270
00:11:11.445 --> 00:11:12.845
how you've used credit in the past.

271
00:11:13.145 --> 00:11:15.365
But an inherent contradiction of that was that maybe someone

272
00:11:16.065 --> 00:11:18.205
is the most credit worthy if they've never used credit.

273
00:11:18.225 --> 00:11:19.365
If I've never needed to borrow,

274
00:11:19.625 --> 00:11:21.325
and this does apply to myself,

275
00:11:21.435 --> 00:11:22.845
I've never had to take a student loan.

276
00:11:22.845 --> 00:11:25.565
I've been so fortunate in my life, I've never had

277
00:11:25.565 --> 00:11:27.165
to take a mortgage, so I will never

278
00:11:27.165 --> 00:11:28.325
have a perfect credit score.

279
00:11:28.505 --> 00:11:30.165
So someone who purely lives off

280
00:11:30.165 --> 00:11:32.685
of their savings is extremely responsible.

281
00:11:32.945 --> 00:11:35.725
And I would say that, getting into the idea

282
00:11:35.905 --> 00:11:38.285
of what's missing, this is very good on credit,

283
00:11:38.705 --> 00:11:42.565
but credit is not all of our economic and financial lives.

284
00:11:42.665 --> 00:11:45.725
So the, the holy grail of research is really we,

285
00:11:45.825 --> 00:11:47.765
we would like, if we really wanna get a full picture

286
00:11:47.785 --> 00:11:50.365
of consumers' financial lives, we would want their credit,

287
00:11:50.365 --> 00:11:52.245
which this data source is very good at.

288
00:11:52.665 --> 00:11:54.565
But then we would also want their income

289
00:11:54.745 --> 00:11:57.085
and their assets, which we essentially

290
00:11:57.835 --> 00:11:59.245
have very little, almost zero.

291
00:11:59.425 --> 00:12:01.565
Um, we have some estimates of income,

292
00:12:01.945 --> 00:12:04.565
but this is not a labor economic data set.

293
00:12:04.585 --> 00:12:07.125
And so we would really not wanna use the measures of income

294
00:12:07.425 --> 00:12:10.725
for measuring someone's labor, mobility or wages.

295
00:12:11.065 --> 00:12:12.485
It could be used as an estimate

296
00:12:12.485 --> 00:12:15.005
of generally are they a six figure income person

297
00:12:15.185 --> 00:12:16.685
or a working class person,

298
00:12:17.225 --> 00:12:19.125
but we wouldn't really wanna use this data

299
00:12:19.185 --> 00:12:20.365
set to measure labor.

300
00:12:20.385 --> 00:12:23.725
And so you need to link those with more established

301
00:12:24.355 --> 00:12:26.645
data sets that that really measure people's wages

302
00:12:26.645 --> 00:12:28.245
and labor, that outcomes.

303
00:12:28.785 --> 00:12:32.245
Uh, so we are really missing huge, huge swaths

304
00:12:32.265 --> 00:12:34.245
of a person's financial life.

305
00:12:34.545 --> 00:12:37.285
Um, and through linking with other data sets, we are able

306
00:12:37.285 --> 00:12:39.245
to get closer to that holy grail.

307
00:12:39.505 --> 00:12:41.925
But I think, starting from the basis

308
00:12:41.945 --> 00:12:44.405
of consumer credit data, in order to use this data properly

309
00:12:44.405 --> 00:12:46.765
for research, you need to be aware of the limitations.

310
00:12:46.795 --> 00:12:48.325
Another thing we do want

311
00:12:48.505 --> 00:12:50.805
to promote in this research paper is

312
00:12:50.805 --> 00:12:53.565
that this is not perfect quality data.

313
00:12:53.595 --> 00:12:56.285
This is collected, from a hodgepodge.

314
00:12:56.285 --> 00:12:59.645
Essentially these three major credit reporting firms in the

315
00:12:59.645 --> 00:13:02.125
United States probably have contracts with dozens,

316
00:13:02.125 --> 00:13:03.885
if not hundreds of other firms

317
00:13:03.915 --> 00:13:06.445
that all focus on specific segments of the market.

318
00:13:06.505 --> 00:13:09.165
For example, they might contract with a company

319
00:13:09.195 --> 00:13:11.525
that essentially only goes to courts

320
00:13:11.785 --> 00:13:14.045
and collects court records including bankruptcies,

321
00:13:14.305 --> 00:13:15.405
tax liens and other things.

322
00:13:15.625 --> 00:13:18.445
And they might outsource to another firm to collect.

323
00:13:18.465 --> 00:13:20.325
Um, and some of these do have payroll data.

324
00:13:20.425 --> 00:13:22.125
So, there are some of these firms

325
00:13:22.125 --> 00:13:24.525
that have higher quality labor wage data,

326
00:13:24.525 --> 00:13:28.925
but it's kind of separate from the, credit side.

327
00:13:28.925 --> 00:13:31.845
But researchers may be able to get the linked version

328
00:13:31.945 --> 00:13:33.405
of both the credit and labor side.

329
00:13:33.405 --> 00:13:34.885
And as I mentioned, only a subset

330
00:13:34.945 --> 00:13:36.525
of these firms have small business data

331
00:13:36.545 --> 00:13:39.485
and we know that that is a very limited coverage compared

332
00:13:39.485 --> 00:13:40.965
with the consumer quality.

333
00:13:41.025 --> 00:13:44.365
So we do wanna emphasize some of these, missing pieces.

334
00:13:44.465 --> 00:13:47.445
And we don't wanna consider this kind of the end all be all

335
00:13:47.445 --> 00:13:50.765
of all data, but I think, over the 2010s

336
00:13:50.765 --> 00:13:53.965
and 2020s data quality has improved dramatically.

337
00:13:53.965 --> 00:13:56.525
So the reporting frequency in the past firms may have only

338
00:13:57.085 --> 00:13:58.485
reported spottily once a quarter,

339
00:13:58.915 --> 00:14:00.805
even longer than that to some extent.

340
00:14:00.825 --> 00:14:03.925
But as of 2024, most of these firms are,

341
00:14:03.945 --> 00:14:05.245
are reporting every single month.

342
00:14:05.245 --> 00:14:06.405
So we get up to date data.

343
00:14:06.655 --> 00:14:08.125
There have been some exceptions to that,

344
00:14:08.125 --> 00:14:09.285
such as during the pandemic.

345
00:14:09.625 --> 00:14:12.725
Uh, there's a little bit of a trade off between,

346
00:14:12.725 --> 00:14:15.565
things like consumer privacy versus the accuracy

347
00:14:15.565 --> 00:14:16.925
of the data consumer privacy as well

348
00:14:16.925 --> 00:14:17.965
as general consumer interest.

349
00:14:17.985 --> 00:14:19.885
So if we think back to 2020,

350
00:14:20.665 --> 00:14:22.485
the many people were becoming unemployed,

351
00:14:22.485 --> 00:14:25.485
really suffering a lot of hardships due to the pandemic.

352
00:14:25.585 --> 00:14:28.325
And so a lot of temporary rules were put in place such

353
00:14:28.325 --> 00:14:30.485
that you are delinquent on your student loan.

354
00:14:30.745 --> 00:14:33.485
It wouldn't necessarily get reported, wouldn't harm you that

355
00:14:33.485 --> 00:14:35.245
because this is essentially an act of God.

356
00:14:35.585 --> 00:14:37.365
Um, global pandemic is not

357
00:14:37.365 --> 00:14:39.845
that you personally were calling delinquent

358
00:14:39.845 --> 00:14:41.965
or not paying your bills on purpose.

359
00:14:42.545 --> 00:14:45.645
Uh, so we were became more lenient on credit reporting.

360
00:14:45.985 --> 00:14:47.925
Of course that affects the data that we see.

361
00:14:47.925 --> 00:14:49.325
So we know that data reported

362
00:14:49.325 --> 00:14:51.605
during the 2020s is not, was not not as accurate.

363
00:14:51.605 --> 00:14:53.925
It was not as up to date. And there were some specific

364
00:14:54.115 --> 00:14:56.645
instances in which mortgages, student loans

365
00:14:56.645 --> 00:14:59.365
and other things were put in a forbearance status so

366
00:14:59.365 --> 00:15:03.325
that everyone was marked as current, artificially so

367
00:15:03.325 --> 00:15:06.045
that we saw, whereas actual payment rates were going

368
00:15:06.045 --> 00:15:08.085
down for obvious reasons due

369
00:15:08.085 --> 00:15:11.485
to the distressed cause the delinquency rate was,

370
00:15:11.625 --> 00:15:12.765
was going down.

371
00:15:12.825 --> 00:15:15.645
So it made it seem like people's financial,

372
00:15:15.645 --> 00:15:17.165
health was getting better even though

373
00:15:17.165 --> 00:15:18.405
people are falling delinquent.

374
00:15:18.405 --> 00:15:20.165
And so with the exception of the pandemic,

375
00:15:20.395 --> 00:15:21.725
data quality has improved.

376
00:15:21.905 --> 00:15:24.765
But I would point out to two other examples of regulations

377
00:15:25.035 --> 00:15:28.005
that are meant to support data quality

378
00:15:28.185 --> 00:15:29.845
and also protect consumer privacy

379
00:15:30.025 --> 00:15:33.325
and protect consumers from things like having an incorrect

380
00:15:33.325 --> 00:15:35.805
credit report merged to your record.

381
00:15:35.915 --> 00:15:36.965
That can be devastating

382
00:15:36.965 --> 00:15:39.925
and could take years for somebody to correct a mistake

383
00:15:39.925 --> 00:15:42.765
that then you cannot get credit due to somebody else

384
00:15:42.765 --> 00:15:44.165
with the same name or something like that,

385
00:15:44.165 --> 00:15:47.245
which can be really, really challenging and devastating.

386
00:15:47.705 --> 00:15:50.725
But as a result in the 2010s, in the past five

387
00:15:50.725 --> 00:15:53.965
to 10 years or so, things like civil judgements,

388
00:15:53.975 --> 00:15:57.045
which are essentially being sued in a civil court

389
00:15:57.145 --> 00:15:59.685
by a creditor, for example, if you, fall

390
00:15:59.685 --> 00:16:01.645
behind on your credit card bill,

391
00:16:01.645 --> 00:16:03.325
that bank can eventually sue you.

392
00:16:03.545 --> 00:16:04.765
And there are records of this.

393
00:16:04.765 --> 00:16:06.965
This is a less commonly known phenomenon,

394
00:16:06.965 --> 00:16:09.085
but there's twice as many of them as bankruptcies.

395
00:16:09.105 --> 00:16:11.125
So, there were questions about,

396
00:16:11.185 --> 00:16:12.525
how accurate that data were.

397
00:16:12.585 --> 00:16:14.805
So essentially, today

398
00:16:15.315 --> 00:16:17.445
they're not really being reported in credit records.

399
00:16:17.545 --> 00:16:18.965
So they're still happening,

400
00:16:18.965 --> 00:16:20.325
people are still getting sued in court, but

401
00:16:20.325 --> 00:16:22.245
because there were questions about the quality of that data,

402
00:16:23.385 --> 00:16:25.885
the result was, due to regulatory

403
00:16:25.885 --> 00:16:26.925
and market equilibrium

404
00:16:26.925 --> 00:16:28.445
that now we're not reporting them at all

405
00:16:28.445 --> 00:16:30.685
because we don't know how accurate they are.

406
00:16:30.945 --> 00:16:33.485
But that means that as researchers we're not really able

407
00:16:33.485 --> 00:16:35.845
to observe them, we're not able to do research to say, well,

408
00:16:35.845 --> 00:16:38.405
what is the impact of a civil judgment on a consumer?

409
00:16:38.435 --> 00:16:42.645
It's probably not great to be sued by,

410
00:16:42.785 --> 00:16:45.965
by a creditor, but now we can't see what's happening to them

411
00:16:45.965 --> 00:16:47.125
because it's not showing up.

412
00:16:47.565 --> 00:16:49.805
A similar story with medical debt where there were concerns

413
00:16:49.805 --> 00:16:52.485
that, medical debt was not being,

414
00:16:52.625 --> 00:16:56.125
act was not really, an accurate predictor

415
00:16:56.125 --> 00:16:57.365
of your future credit worthiness and

416
00:16:57.365 --> 00:16:59.405
therefore it shouldn't really be part of your credit score.

417
00:16:59.705 --> 00:17:01.485
And there was a regulatory

418
00:17:01.705 --> 00:17:03.605
and industry debate on this such

419
00:17:03.605 --> 00:17:06.485
that medical debt is also now not being reported.

420
00:17:06.485 --> 00:17:07.525
Essentially very little

421
00:17:07.525 --> 00:17:09.725
of it is being reported in the data source.

422
00:17:09.865 --> 00:17:12.565
And then as can, as researchers,

423
00:17:12.625 --> 00:17:14.205
we cannot see what's going on.

424
00:17:14.205 --> 00:17:16.405
People are still having medical debt accumulated

425
00:17:16.405 --> 00:17:17.645
and still affecting their lives.

426
00:17:18.225 --> 00:17:20.485
Um, and so there can be a little bit of a trade off

427
00:17:20.485 --> 00:17:21.645
between consumer privacy

428
00:17:21.745 --> 00:17:25.205
and consumer welfare concerns with, just being able

429
00:17:25.205 --> 00:17:27.205
to observe, this information at all.

430
00:17:27.205 --> 00:17:28.525
And sometimes it just ends up that

431
00:17:28.525 --> 00:17:31.005
that information becomes excluded all together

432
00:17:31.425 --> 00:17:33.925
and we can think about is that good or bad?

433
00:17:33.925 --> 00:17:35.325
Would we rather have bad information

434
00:17:35.545 --> 00:17:37.485
or would we rather have no information?

435
00:17:37.755 --> 00:17:40.725
It's difficult, but I think we always found way, find ways

436
00:17:40.725 --> 00:17:42.605
around it and find other data sources.

437
00:17:43.025 --> 00:17:45.125
Uh, so I think we can be resourceful,

438
00:17:45.265 --> 00:17:47.685
but at the same time there are legitimate public policy

439
00:17:47.725 --> 00:17:50.445
concerns for, for the general public about this as well.

440
00:17:50.445 --> 00:17:51.445
So it's a tricky, it's a very tricky.

Faculti AI can make mistakes

Faculti

Popular in Social Sciences

Macroeconomic modelling of CBDC: a critical review

Financing Private Credit

Failing Banks

The City as a Growth Machine: Toward a Political Economy of Place

Keeping the Republic

Entrepreneurial thinking: the politics and practice of policy impact

Consumer Credit Reporting Data

Next in Economics

Guide

Platform and Category Pages

Contact Us

FAQs

What is Faculti?

How do you select interviews?

Interview Process

Page Structure

Popular in Social Sciences

Guide

Platform and Category Pages

Contact Us

FAQs

What is Faculti?

How do you select interviews?

Interview Process

Accessibility Options

Contrast Adjustment

Text Size Adjustment

Text Spacing Adjustment

Line Height Adjustment

Cursor Size Adjustment

Page Structure Options

Page Structure