1
00:00:00,090 --> 00:00:02,490
The following content is
provided under a Creative

2
00:00:02,490 --> 00:00:04,059
Commons license.

3
00:00:04,059 --> 00:00:06,330
Your support will help
MIT OpenCourseWare

4
00:00:06,330 --> 00:00:10,720
continue to offer high quality
educational resources for free.

5
00:00:10,720 --> 00:00:13,350
To make a donation or
view additional materials

6
00:00:13,350 --> 00:00:17,280
from hundreds of MIT courses,
visit MIT OpenCourseWare

7
00:00:17,280 --> 00:00:18,480
at ocw.mit.edu.

8
00:00:28,335 --> 00:00:30,210
ROBERT TOWNSEND: The
logistics for the class.

9
00:00:33,630 --> 00:00:35,770
Unfortunately, we seem
to have run out a bit.

10
00:00:35,770 --> 00:00:39,870
So if you could share the
syllabus that you have.

11
00:00:39,870 --> 00:00:43,230
I'm actually not going
to go over it in detail.

12
00:00:43,230 --> 00:00:46,500
I do want to give you sort
of the big picture of what

13
00:00:46,500 --> 00:00:47,040
we're doing.

14
00:00:47,040 --> 00:00:52,050
First of all, I'm
glad you're here.

15
00:00:52,050 --> 00:00:55,110
And I do want class
participation.

16
00:00:55,110 --> 00:01:03,400
There will be three problem
sets and a research proposal.

17
00:01:03,400 --> 00:01:06,060
The problem sets are
meant to be practice

18
00:01:06,060 --> 00:01:12,520
for you just to make sure you're
working through the material.

19
00:01:12,520 --> 00:01:15,580
Some of you take field
exams, for example.

20
00:01:15,580 --> 00:01:19,020
So these problem sets are
meant to be practices for that,

21
00:01:19,020 --> 00:01:20,190
as well.

22
00:01:20,190 --> 00:01:22,860
It's really meant to help you.

23
00:01:22,860 --> 00:01:27,390
One of the problem sets is going
to focus on actually getting

24
00:01:27,390 --> 00:01:30,930
your hands dirty with
the data, so to speak,

25
00:01:30,930 --> 00:01:32,580
and Whit's going
to help with that

26
00:01:32,580 --> 00:01:36,370
and give you access to
the Townsend tie data.

27
00:01:36,370 --> 00:01:40,380
And we'll have a specific
project geared around that.

28
00:01:40,380 --> 00:01:47,160
It's really a research
class so, I would like you

29
00:01:47,160 --> 00:01:49,680
to write a research proposal.

30
00:01:49,680 --> 00:01:52,180
Not exactly a paper--

31
00:01:52,180 --> 00:01:54,110
it can be short.

32
00:01:54,110 --> 00:01:56,130
But the idea is
that hopefully as we

33
00:01:56,130 --> 00:01:58,590
go through these
lectures, you will--

34
00:01:58,590 --> 00:02:00,390
and I will try to
help this process--

35
00:02:00,390 --> 00:02:03,120
you'll have ideas
about things you

36
00:02:03,120 --> 00:02:05,670
would like to work
on, things where

37
00:02:05,670 --> 00:02:08,940
you think the literature
is falling short.

38
00:02:08,940 --> 00:02:12,510
And you should begin to
sort of think about that.

39
00:02:12,510 --> 00:02:16,020
It's all too easy
in these courses

40
00:02:16,020 --> 00:02:19,890
to think that taking a course
means memorizing the lecture

41
00:02:19,890 --> 00:02:21,970
notes or learning
how to manipulate

42
00:02:21,970 --> 00:02:25,860
as in a certain model,
and not thinking broadly

43
00:02:25,860 --> 00:02:26,830
about next steps.

44
00:02:26,830 --> 00:02:31,260
You don't see faculty doing
the research, actually.

45
00:02:31,260 --> 00:02:35,730
I'm going to try to share with
you how literatures develop,

46
00:02:35,730 --> 00:02:39,930
compare and contrast different
styles, different attack

47
00:02:39,930 --> 00:02:41,838
strategies.

48
00:02:41,838 --> 00:02:43,380
In fact, the lectures
are going to be

49
00:02:43,380 --> 00:02:45,480
a blend between
overview material

50
00:02:45,480 --> 00:02:52,213
and the usual detailed
analysis of specific papers.

51
00:02:52,213 --> 00:02:53,880
Because that's a big
part of the course,

52
00:02:53,880 --> 00:02:56,500
I want you to get
some experience

53
00:02:56,500 --> 00:03:01,860
at trying to think about
things that way, as well.

54
00:03:01,860 --> 00:03:06,160
The syllabus does say
something about a midterm,

55
00:03:06,160 --> 00:03:09,480
which is basically the final
for my half of the course.

56
00:03:09,480 --> 00:03:12,180
This is joint with Abhijit.

57
00:03:12,180 --> 00:03:16,860
But my first gift to you is
that we're going to waive that.

58
00:03:16,860 --> 00:03:20,670
So there will be no final exam.

59
00:03:20,670 --> 00:03:22,440
And instead, the
emphasis is going

60
00:03:22,440 --> 00:03:27,210
to be on the problem sets,
and the research proposal,

61
00:03:27,210 --> 00:03:28,680
and class participation.

62
00:03:31,200 --> 00:03:34,410
Before I forget to say,
the recitation section

63
00:03:34,410 --> 00:03:37,020
is scheduled Thursdays.

64
00:03:37,020 --> 00:03:39,850
I had originally
thought it was Fridays.

65
00:03:39,850 --> 00:03:43,170
So we put To Be Determined in
all those recitation sections.

66
00:03:43,170 --> 00:03:46,990
And in fact, by this Thursday
noon-- just after this class,

67
00:03:46,990 --> 00:03:49,230
essentially--

68
00:03:49,230 --> 00:03:52,620
you would have already had
your first recitation section.

69
00:03:52,620 --> 00:03:55,080
But it's just too
early, so we're not

70
00:03:55,080 --> 00:03:57,540
going to meet on Thursday.

71
00:03:57,540 --> 00:04:00,600
But we will start in
earnest the following week.

72
00:04:00,600 --> 00:04:04,960
And if you start
looking at the syllabus,

73
00:04:04,960 --> 00:04:09,090
you will see that
there's a section

74
00:04:09,090 --> 00:04:12,480
at the beginning on data, which
I've kind of already mentioned.

75
00:04:12,480 --> 00:04:16,170
There's a section at the
beginning on computation,

76
00:04:16,170 --> 00:04:18,329
and also on the GIS.

77
00:04:18,329 --> 00:04:23,040
So if you want to think about
those as tools or skills,

78
00:04:23,040 --> 00:04:27,870
we're going to have
some TA sessions going

79
00:04:27,870 --> 00:04:30,690
over sort of some of the
computational aspects,

80
00:04:30,690 --> 00:04:34,840
for example, after the
lecture on Thursday.

81
00:04:34,840 --> 00:04:40,770
Though it will be obvious how
a computational recitation

82
00:04:40,770 --> 00:04:42,930
section fits in.

83
00:04:42,930 --> 00:04:46,710
Now I recognize I don't
want to overdo this,

84
00:04:46,710 --> 00:04:50,850
but the more I've talked
to many of you, the more--

85
00:04:50,850 --> 00:04:53,100
including yesterday,
actually, in some sense,

86
00:04:53,100 --> 00:04:55,440
and advising of second years--

87
00:04:55,440 --> 00:04:58,050
the more I realized
that students like

88
00:04:58,050 --> 00:05:01,710
you want to get trained
in computational skills

89
00:05:01,710 --> 00:05:03,730
and methods, and so on.

90
00:05:03,730 --> 00:05:05,910
So we can have some
extra sessions.

91
00:05:05,910 --> 00:05:07,320
I'm in the busy--

92
00:05:07,320 --> 00:05:11,640
in the process of setting up
extra computational sections

93
00:05:11,640 --> 00:05:14,550
on how to solve
certain problems.

94
00:05:14,550 --> 00:05:18,235
We'll probably make up for the
lost TA session on Thursday.

95
00:05:18,235 --> 00:05:19,860
And Whit, in fact,
is going to send out

96
00:05:19,860 --> 00:05:23,700
an email to each of
you to try to work

97
00:05:23,700 --> 00:05:26,280
on scheduling a backup time.

98
00:05:26,280 --> 00:05:28,740
So if we have some
extra sessions,

99
00:05:28,740 --> 00:05:33,040
we'll try to have as few
conflicts as possible

100
00:05:33,040 --> 00:05:34,110
so you can come.

101
00:05:38,850 --> 00:05:43,560
So I guess the only thing I want
to say about the syllabus that

102
00:05:43,560 --> 00:05:48,350
won't be obvious
from the lecture is

103
00:05:48,350 --> 00:05:53,670
today is the introductory
class, and there's all

104
00:05:53,670 --> 00:05:56,100
this supplementary material.

105
00:05:56,100 --> 00:05:59,580
And then on Thursday,
we'll start with Lecture 2.

106
00:05:59,580 --> 00:06:04,530
You'll get a dose of these sort
of micro-founded macro models.

107
00:06:04,530 --> 00:06:08,550
We're going to go
on to another sort

108
00:06:08,550 --> 00:06:12,720
of new wave of those models
featuring limited commitment

109
00:06:12,720 --> 00:06:20,460
and collateral constraints on
the growth TSP type issues.

110
00:06:20,460 --> 00:06:28,380
And then another section
focuses on basically versions

111
00:06:28,380 --> 00:06:31,890
where there's costly state
verification, or moral hazard,

112
00:06:31,890 --> 00:06:33,850
or adverse selection.

113
00:06:33,850 --> 00:06:37,530
So you can compare and
contrast the implications

114
00:06:37,530 --> 00:06:41,130
of having these different
financial underpinnings.

115
00:06:41,130 --> 00:06:46,160
Which you know is
complementary classes in macro

116
00:06:46,160 --> 00:06:51,150
are also getting at
those same issues.

117
00:06:51,150 --> 00:06:55,680
Then we'll have sort of the
intermediate class on microdata

118
00:06:55,680 --> 00:06:59,370
and building all the way
from household accounts

119
00:06:59,370 --> 00:07:01,890
to national income
accounts, and then

120
00:07:01,890 --> 00:07:06,130
move to actually using the
microdata for various tests

121
00:07:06,130 --> 00:07:09,660
so that these models, full
insurance, in various ways,

122
00:07:09,660 --> 00:07:11,880
then with these
obstacles to trade

123
00:07:11,880 --> 00:07:14,970
and so on throughout the class.

124
00:07:14,970 --> 00:07:19,020
So that's kind of the big
picture of the syllabus.

125
00:07:22,800 --> 00:07:26,490
Each lecture is pretty--

126
00:07:29,000 --> 00:07:29,715
a lot of fun.

127
00:07:29,715 --> 00:07:31,340
And there's a lot of
material in there.

128
00:07:31,340 --> 00:07:33,170
The lecture notes are
meant, as I've said,

129
00:07:33,170 --> 00:07:37,040
to tie the material
together both in general,

130
00:07:37,040 --> 00:07:38,828
and in its specifics.

131
00:07:38,828 --> 00:07:40,370
Today is a little
bit of an exception

132
00:07:40,370 --> 00:07:43,820
because it's just an
introductory lecture.

133
00:07:43,820 --> 00:07:47,330
I will try to have
all the slides posted.

134
00:07:47,330 --> 00:07:49,610
I'm not sure we managed
to get this thing posted,

135
00:07:49,610 --> 00:07:50,700
but I'll do that.

136
00:07:50,700 --> 00:07:53,060
I'll try to get that
done right after class.

137
00:07:53,060 --> 00:07:56,190
Normally the slides will be
posted in advance of the class.

138
00:07:56,190 --> 00:07:59,930
So if you feel more comfortable
looking at material,

139
00:07:59,930 --> 00:08:02,427
might facilitate asking
questions during class,

140
00:08:02,427 --> 00:08:03,260
which would be good.

141
00:08:03,260 --> 00:08:04,957
AUDIENCE: These
slides are posted.

142
00:08:04,957 --> 00:08:06,040
ROBERT TOWNSEND: Oh, good.

143
00:08:06,040 --> 00:08:08,720
Excellent.

144
00:08:08,720 --> 00:08:11,090
And most of the readings
are linked, not entirely

145
00:08:11,090 --> 00:08:11,720
all of them.

146
00:08:14,530 --> 00:08:20,120
So it should be easy to
explore your special topic.

147
00:08:20,120 --> 00:08:24,640
Are there questions about
the logistics of the class?

148
00:08:24,640 --> 00:08:27,190
And I'll hang around
a bit afterwards

149
00:08:27,190 --> 00:08:31,150
to answer any specific
questions you might have.

150
00:08:31,150 --> 00:08:36,679
So finance, growth,
and volatility--

151
00:08:36,679 --> 00:08:39,809
theory, data, and the
formulation of policy.

152
00:08:39,809 --> 00:08:44,550
You know, the course is
organized around these topics.

153
00:08:44,550 --> 00:08:48,791
There's so many ways
to think about this.

154
00:08:48,791 --> 00:08:55,365
Finance, financial access is a
very huge issue in development.

155
00:08:55,365 --> 00:08:57,470
You may be more used
to thinking about it

156
00:08:57,470 --> 00:09:00,380
from a micro perspective.

157
00:09:00,380 --> 00:09:04,100
But we not only do the
micro, we do the macro.

158
00:09:04,100 --> 00:09:07,850
The idea is that somehow
finance helps people

159
00:09:07,850 --> 00:09:09,140
with financial services.

160
00:09:09,140 --> 00:09:13,460
It may be alleviating
poverty, and in the context

161
00:09:13,460 --> 00:09:16,200
of these general
equilibrium models,

162
00:09:16,200 --> 00:09:19,010
it has additional effects.

163
00:09:19,010 --> 00:09:21,210
So many countries
around the world,

164
00:09:21,210 --> 00:09:25,850
not just the World Bank
and other aid agencies,

165
00:09:25,850 --> 00:09:31,920
are pushing, promoting,
advocating, financial access.

166
00:09:31,920 --> 00:09:34,790
There's new Gallup surveys
measuring this stuff

167
00:09:34,790 --> 00:09:38,210
to try to get at inequalities
across countries,

168
00:09:38,210 --> 00:09:46,700
to sort of push a
little bit the agenda.

169
00:09:46,700 --> 00:09:48,080
So that's finance and growth.

170
00:09:48,080 --> 00:09:51,290
I'm going to say
more in a second.

171
00:09:51,290 --> 00:09:55,230
Finance and volatility
sort of should

172
00:09:55,230 --> 00:10:02,540
bring to mind the recent US
and worldwide financial crisis.

173
00:10:02,540 --> 00:10:07,230
And it pushes you seemingly in
exactly the opposite direction,

174
00:10:07,230 --> 00:10:11,870
which is maybe the
housing market in the US.

175
00:10:11,870 --> 00:10:15,140
Maybe there was a
little too much effort

176
00:10:15,140 --> 00:10:20,090
for middle or even lower
income households to own homes,

177
00:10:20,090 --> 00:10:27,430
and that's led to certain
government agency financing

178
00:10:27,430 --> 00:10:28,480
arrangements.

179
00:10:28,480 --> 00:10:36,720
And in any event, it clearly
was a financial crisis.

180
00:10:36,720 --> 00:10:41,590
And most of the regulations
deal with seemingly trying

181
00:10:41,590 --> 00:10:45,460
to push us back toward more
restricted intermediation,

182
00:10:45,460 --> 00:10:48,490
as if financial
access, and credit,

183
00:10:48,490 --> 00:10:52,750
and financial
liberalization might,

184
00:10:52,750 --> 00:10:55,990
if not be a bad thing,
maybe go too far.

185
00:10:58,800 --> 00:11:02,510
Well, we're going to try
to be writing down models

186
00:11:02,510 --> 00:11:06,680
that think about not only
each of those separately,

187
00:11:06,680 --> 00:11:09,375
but to some extent,
having it on the same page

188
00:11:09,375 --> 00:11:10,625
to try to get at that tension.

189
00:11:13,220 --> 00:11:16,310
And actually growth is in
there with finance and growth.

190
00:11:16,310 --> 00:11:19,210
Growth and volatility is
another quote bilateral

191
00:11:19,210 --> 00:11:22,770
pairing if you wish.

192
00:11:22,770 --> 00:11:25,540
There's some idea that
finance causes growth,

193
00:11:25,540 --> 00:11:28,730
but growth does not
come without volatility,

194
00:11:28,730 --> 00:11:30,200
leaving the question
I just raised

195
00:11:30,200 --> 00:11:34,260
about finance and volatility.

196
00:11:34,260 --> 00:11:35,990
So let me sort of
give you a sense

197
00:11:35,990 --> 00:11:38,510
of each of these sub-pieces.

198
00:11:41,290 --> 00:11:43,610
Finance and growth,
well, here's the story,

199
00:11:43,610 --> 00:11:51,420
and I think it's pretty well
accepted by these three papers,

200
00:11:51,420 --> 00:11:53,900
for example, that
are very much cited,

201
00:11:53,900 --> 00:12:01,490
King and Levine, Rajan and
Zingales, and Ross Levine.

202
00:12:01,490 --> 00:12:05,090
You can read the
abstract, but basically,

203
00:12:05,090 --> 00:12:06,560
let's pick out a
few things here.

204
00:12:06,560 --> 00:12:10,020
Finance and Growth,
Schumpeter Might be Right.

205
00:12:10,020 --> 00:12:14,690
Cross country evidence that
financial systems promote

206
00:12:14,690 --> 00:12:20,710
growth, that the levels
of financial development

207
00:12:20,710 --> 00:12:24,250
are strongly associated
with per capita GDP growth,

208
00:12:24,250 --> 00:12:27,910
and capital accumulations and
improvements in efficiency

209
00:12:27,910 --> 00:12:29,230
like TFP.

210
00:12:29,230 --> 00:12:33,070
But furthermore, there seems
to be causality in the data,

211
00:12:33,070 --> 00:12:40,060
that predetermined level
of financial deepening

212
00:12:40,060 --> 00:12:43,540
or development are associated
with subsequent higher

213
00:12:43,540 --> 00:12:48,130
rates of all these good things.

214
00:12:48,130 --> 00:12:52,540
Rajan and Zingales come
at it in a related way.

215
00:12:52,540 --> 00:12:57,460
They're basically saying
that industrial sectors that

216
00:12:57,460 --> 00:12:59,920
are more in need
of external finance

217
00:12:59,920 --> 00:13:02,560
due to the nature
of the business

218
00:13:02,560 --> 00:13:06,310
develop disproportionately
faster starting

219
00:13:06,310 --> 00:13:10,900
from a lower level in countries
where the financial systems are

220
00:13:10,900 --> 00:13:14,470
liberalized or more liberal
than in other countries.

221
00:13:17,580 --> 00:13:20,050
And they're worried about
causality and so forth.

222
00:13:20,050 --> 00:13:26,620
And Levine's review, Finance
and Growth, Theory and Evidence,

223
00:13:26,620 --> 00:13:30,900
the summary in the abstract
with qualifications

224
00:13:30,900 --> 00:13:33,030
to follow from him.

225
00:13:33,030 --> 00:13:36,000
Countries with better
functioning banks grow faster,

226
00:13:36,000 --> 00:13:37,590
but however, it
doesn't seem to matter

227
00:13:37,590 --> 00:13:43,560
whether it's bank-led finance
or stocks or equity markets.

228
00:13:43,560 --> 00:13:50,520
He tries to cover the
literature on simultaneity bias,

229
00:13:50,520 --> 00:13:51,240
and so on.

230
00:13:54,410 --> 00:13:58,940
So Levine also points
out the importance

231
00:13:58,940 --> 00:14:02,750
of this question
and the importance,

232
00:14:02,750 --> 00:14:05,360
therefore, to research.

233
00:14:05,360 --> 00:14:07,910
And I must say,
in these slides I

234
00:14:07,910 --> 00:14:11,900
tried to indicate where there
were quotes, but at some point

235
00:14:11,900 --> 00:14:13,700
you may find it
hard to tell what

236
00:14:13,700 --> 00:14:17,960
was the author's statement,
which is quite clear here,

237
00:14:17,960 --> 00:14:20,990
versus my summary.

238
00:14:20,990 --> 00:14:23,840
But I wanted to see--

239
00:14:23,840 --> 00:14:25,820
I didn't want you have
to take my word for it.

240
00:14:25,820 --> 00:14:28,370
I actually wanted you to
see what these guys are

241
00:14:28,370 --> 00:14:32,450
saying in their papers.

242
00:14:32,450 --> 00:14:34,460
"Research that clarifies
our understanding

243
00:14:34,460 --> 00:14:37,250
of the role of finance
and economic growth

244
00:14:37,250 --> 00:14:40,370
have policy
implications and shape

245
00:14:40,370 --> 00:14:44,470
policy-oriented research."

246
00:14:44,470 --> 00:14:47,500
This is going to influence
the priority that policymakers

247
00:14:47,500 --> 00:14:52,390
and advisors attach to reforming
the financial service sector.

248
00:14:56,620 --> 00:14:58,720
And so when this
empirical work shows

249
00:14:58,720 --> 00:15:01,300
there is the connection between
finance and long-run growth,

250
00:15:01,300 --> 00:15:04,840
that basically advertises
the urgent need

251
00:15:04,840 --> 00:15:09,530
for research on the political,
legal, regulatory, or policy

252
00:15:09,530 --> 00:15:12,670
determinants of
financial development.

253
00:15:12,670 --> 00:15:15,790
Note he is not saying
that it's a done deal.

254
00:15:15,790 --> 00:15:18,230
And he has been one of
the lead contributors

255
00:15:18,230 --> 00:15:20,230
to this empirical work.

256
00:15:20,230 --> 00:15:22,460
But he's saying, in
fact, there is much more

257
00:15:22,460 --> 00:15:24,710
to do on the research side.

258
00:15:24,710 --> 00:15:27,860
Specific policy implications
are not immediate.

259
00:15:27,860 --> 00:15:34,910
It's not obvious how to go
from the general idea of policy

260
00:15:34,910 --> 00:15:40,100
to the way that financial
systems actually work,

261
00:15:40,100 --> 00:15:44,000
and in particular, the
way they are modeled

262
00:15:44,000 --> 00:15:46,980
at the theoretical level.

263
00:15:46,980 --> 00:15:50,690
So you'll see this over and over
again in these lectures, which

264
00:15:50,690 --> 00:15:51,380
is this--

265
00:15:51,380 --> 00:15:53,100
and I try to bring
out both sides--

266
00:15:53,100 --> 00:15:58,640
which is empirical work seems
to mesmerize, if not reassure

267
00:15:58,640 --> 00:16:02,900
theoretical work with almost
no mention to the data

268
00:16:02,900 --> 00:16:06,440
and often this gap in between.

269
00:16:06,440 --> 00:16:10,340
I'm going to try
to bridge the gap,

270
00:16:10,340 --> 00:16:13,470
show you how the literature
is trying to bridge the gap.

271
00:16:13,470 --> 00:16:16,370
But it's kind of reassuring
and refreshing to see

272
00:16:16,370 --> 00:16:22,730
Levine being so candid in this
rather famous review article.

273
00:16:22,730 --> 00:16:26,270
He says, "What do financial
intermediaries really do?"

274
00:16:26,270 --> 00:16:28,910
Well, you can think about
producing information,

275
00:16:28,910 --> 00:16:33,040
about investment possibilities,
or monitoring investments,

276
00:16:33,040 --> 00:16:36,020
maybe corporate
governance is better

277
00:16:36,020 --> 00:16:40,130
if they're somebody looking
at them, facilitate trading.

278
00:16:42,680 --> 00:16:44,960
Facilitate diversification
is important

279
00:16:44,960 --> 00:16:51,300
when there's risk for sure and
the management of that risk.

280
00:16:51,300 --> 00:16:53,840
Financial systems at
a theoretical level

281
00:16:53,840 --> 00:16:58,140
mobilize and pool savings and
generally ease the exchange

282
00:16:58,140 --> 00:16:59,070
of goods and services.

283
00:17:01,720 --> 00:17:05,130
So changes in these functions
that are brought about

284
00:17:05,130 --> 00:17:10,470
by, say, policy
changes could influence

285
00:17:10,470 --> 00:17:12,240
savings and investment,
and therefore,

286
00:17:12,240 --> 00:17:17,109
have influence on
economic growth.

287
00:17:17,109 --> 00:17:19,680
Now, then you start to think
about what is he really

288
00:17:19,680 --> 00:17:22,500
saying or not saying?

289
00:17:22,500 --> 00:17:26,380
One thing that sort of
caught my attention--

290
00:17:26,380 --> 00:17:29,130
and I was predisposed
to look for it--

291
00:17:29,130 --> 00:17:33,310
is this distinction about
levels versus growth rates.

292
00:17:33,310 --> 00:17:36,960
So is his vision one
of steadily improving

293
00:17:36,960 --> 00:17:41,460
financial intermediation on one
or several of these dimensions,

294
00:17:41,460 --> 00:17:45,940
pushing levels of income
up, so that an expanding

295
00:17:45,940 --> 00:17:48,190
improved financial
sector over time

296
00:17:48,190 --> 00:17:50,860
is causing growth in
the sense that you

297
00:17:50,860 --> 00:17:53,800
have ever-increasing levels?

298
00:17:53,800 --> 00:17:59,740
Or is he saying that
a once-and-for-all fix

299
00:17:59,740 --> 00:18:03,070
to the financial system, making
it go from a lower or higher

300
00:18:03,070 --> 00:18:08,110
level, and no changes after
that have an implication

301
00:18:08,110 --> 00:18:13,830
for sustained growth as
in quote, steady state?

302
00:18:13,830 --> 00:18:16,830
Not clear from what he said.

303
00:18:16,830 --> 00:18:20,385
And we'll see models that take
both these tacts actually.

304
00:18:23,970 --> 00:18:26,670
There are other sort of
caveats when you think

305
00:18:26,670 --> 00:18:30,630
about theoretical modeling.

306
00:18:30,630 --> 00:18:33,480
Are we talking about
technological improvement

307
00:18:33,480 --> 00:18:34,890
in the sense of TFP?

308
00:18:34,890 --> 00:18:37,710
Is that really what
improved intermediation

309
00:18:37,710 --> 00:18:39,030
is supposed to be giving us?

310
00:18:41,710 --> 00:18:44,170
But again, we should
maybe think also

311
00:18:44,170 --> 00:18:48,400
about transitions or even
potentially poverty traps

312
00:18:48,400 --> 00:18:52,240
if systems get stuck.

313
00:18:52,240 --> 00:18:57,100
Another caveat,
savings rates are not

314
00:18:57,100 --> 00:18:59,855
monotone with
increases in return.

315
00:18:59,855 --> 00:19:01,480
I mean, for one thing,
in price theory,

316
00:19:01,480 --> 00:19:05,410
just think about income
and substitution effects.

317
00:19:05,410 --> 00:19:08,780
One cuts against the other one.

318
00:19:08,780 --> 00:19:13,010
Or what about the
relationships between savings

319
00:19:13,010 --> 00:19:14,810
and improved allocation of risk?

320
00:19:17,870 --> 00:19:20,150
If you improve the
allocation of risk bearing,

321
00:19:20,150 --> 00:19:22,240
then you don't need to
save as a buffer stock.

322
00:19:22,240 --> 00:19:25,460
So savings may actually
go down and not up

323
00:19:25,460 --> 00:19:27,320
as a consequence
of improving one

324
00:19:27,320 --> 00:19:29,585
piece of the financial system.

325
00:19:34,470 --> 00:19:38,980
And then he makes his
pitch for thinking about--

326
00:19:38,980 --> 00:19:42,670
don't think broadly about
savings or especially credit

327
00:19:42,670 --> 00:19:43,390
as an input.

328
00:19:43,390 --> 00:19:45,700
Think about actually
how the financial system

329
00:19:45,700 --> 00:19:47,342
is put together.

330
00:19:47,342 --> 00:19:49,300
And we'll come back to
that, especially when we

331
00:19:49,300 --> 00:19:51,250
get to the measurement part.

332
00:19:54,280 --> 00:19:57,580
Does finance cause
reduced inequality?

333
00:20:00,160 --> 00:20:04,080
He said in this review,
not necessarily.

334
00:20:04,080 --> 00:20:06,040
And he reviews
empirical evidence.

335
00:20:06,040 --> 00:20:10,070
There are theoretical models
that claim this to be true,

336
00:20:10,070 --> 00:20:16,120
that the relationship is, say,
favorable to reduce poverty,

337
00:20:16,120 --> 00:20:18,357
reduce inequality, and so on.

338
00:20:18,357 --> 00:20:20,440
But certainly, in some of
the theoretical models--

339
00:20:20,440 --> 00:20:23,290
in some of them that
we're going to cover--

340
00:20:23,290 --> 00:20:25,960
improved financial
systems, at least early on,

341
00:20:25,960 --> 00:20:31,330
cause increases in inequality,
not decreases in inequality.

342
00:20:31,330 --> 00:20:33,130
Something along
the lines of what

343
00:20:33,130 --> 00:20:34,915
Kuznets had in mind, actually.

344
00:20:40,160 --> 00:20:43,340
And then he just goes
through some empirical work

345
00:20:43,340 --> 00:20:45,140
and doesn't come back.

346
00:20:45,140 --> 00:20:48,500
Not to belittle it, but
he does not come back

347
00:20:48,500 --> 00:20:50,675
to this underlying tension.

348
00:20:54,530 --> 00:20:55,910
Volatility and growth.

349
00:20:58,740 --> 00:20:59,240
Let's see.

350
00:21:02,210 --> 00:21:05,000
Before we start thinking that
all the volatility we see

351
00:21:05,000 --> 00:21:09,500
has something to do with finance
or unstable financial systems,

352
00:21:09,500 --> 00:21:14,440
maybe we should try to
document better those sources.

353
00:21:14,440 --> 00:21:21,330
So this paper in the
QJE is a decomposition

354
00:21:21,330 --> 00:21:24,150
of the sources of inequality,
looking in particular

355
00:21:24,150 --> 00:21:25,500
at low income countries.

356
00:21:28,470 --> 00:21:29,650
Take it as a fact--

357
00:21:29,650 --> 00:21:32,940
I'll show you a graph
momentarily from them--

358
00:21:32,940 --> 00:21:36,960
that there is more volatility
in growth rates among low income

359
00:21:36,960 --> 00:21:39,210
countries.

360
00:21:39,210 --> 00:21:41,910
And why is that happening?

361
00:21:41,910 --> 00:21:45,870
Could be that poor countries
somehow or other end up

362
00:21:45,870 --> 00:21:50,740
specialized in volatile sectors.

363
00:21:50,740 --> 00:21:52,990
Or it could be a
version of that,

364
00:21:52,990 --> 00:21:57,400
they just have a lot more
quote, specialization.

365
00:21:57,400 --> 00:22:00,785
They're not terribly
diversified over a multitude

366
00:22:00,785 --> 00:22:01,660
of different sectors.

367
00:22:04,270 --> 00:22:06,220
Could be that poor
countries experience

368
00:22:06,220 --> 00:22:09,640
more frequent or more
severe aggregate shocks

369
00:22:09,640 --> 00:22:13,520
sort of from a macro
policy perspective.

370
00:22:13,520 --> 00:22:17,720
And it's even possible that
poor countries experienced macro

371
00:22:17,720 --> 00:22:21,170
fluctuations that are somehow
correlated with the shocks

372
00:22:21,170 --> 00:22:22,040
in which they're--

373
00:22:22,040 --> 00:22:23,930
the sectors in which
they're specialized.

374
00:22:30,160 --> 00:22:32,890
So here's the picture.

375
00:22:32,890 --> 00:22:40,460
This is levels of per
capita GDP and the variance.

376
00:22:40,460 --> 00:22:42,730
Standard deviation
calculation, this morning,

377
00:22:42,730 --> 00:22:45,340
sort of thinking
about this lecture,

378
00:22:45,340 --> 00:22:48,100
I spooked myself a bit worrying
that it should have been

379
00:22:48,100 --> 00:22:49,560
growth rates and not levels.

380
00:22:49,560 --> 00:22:54,370
But actually when
they do the analysis,

381
00:22:54,370 --> 00:23:00,250
they do a composition in
the growth of value added.

382
00:23:00,250 --> 00:23:03,820
Again, I can't do justice
to this paper other

383
00:23:03,820 --> 00:23:06,040
than give you an overview.

384
00:23:06,040 --> 00:23:12,600
But what they're saying is that
the growth of GDP per worker

385
00:23:12,600 --> 00:23:16,500
in country J, this object
here, is a weighted average

386
00:23:16,500 --> 00:23:19,650
of the growth of GDP
in the various sectors

387
00:23:19,650 --> 00:23:23,310
in which country
J is specialized.

388
00:23:23,310 --> 00:23:28,800
And these AJs are the
weights or the proportions.

389
00:23:28,800 --> 00:23:33,510
And then they write down a quite
plausible statistical model

390
00:23:33,510 --> 00:23:37,200
that the growth of value
added GDP per worker

391
00:23:37,200 --> 00:23:43,200
in sector J of country S has a
country S component, a sector J

392
00:23:43,200 --> 00:23:47,580
component, and some residual.

393
00:23:47,580 --> 00:23:49,890
So that's kind of
the basic model.

394
00:23:49,890 --> 00:23:53,160
And what they do then is
a very clever, basically,

395
00:23:53,160 --> 00:23:55,890
variance co-variance
decomposition

396
00:23:55,890 --> 00:24:01,140
quantifying and answering
the questions that were

397
00:24:01,140 --> 00:24:04,890
raised on the previous slide.

398
00:24:04,890 --> 00:24:11,408
The answer is 60% of this
volatility is country specific.

399
00:24:11,408 --> 00:24:13,450
So I don't know how you
want to think about that.

400
00:24:13,450 --> 00:24:15,780
It's not even 50-50.

401
00:24:15,780 --> 00:24:19,260
More than half of it is somehow
a country specific phenomenon.

402
00:24:19,260 --> 00:24:24,690
On the other hand, there's a
very non-trivial residual 40%--

403
00:24:24,690 --> 00:24:29,460
almost 50, say-- which
is due to the following,

404
00:24:29,460 --> 00:24:33,960
namely diversification
is limited

405
00:24:33,960 --> 00:24:35,400
in low income countries.

406
00:24:35,400 --> 00:24:39,150
And low income countries do
tend to specialize in sectors

407
00:24:39,150 --> 00:24:41,370
that are relatively volatile.

408
00:24:46,140 --> 00:24:47,880
That sets aside
the issue of why,

409
00:24:47,880 --> 00:24:50,380
and whether there are any
possible remedies, and so on.

410
00:24:56,070 --> 00:24:59,170
A Model in which
Volatility is Inevitable.

411
00:24:59,170 --> 00:25:01,660
This is one of Daron's
models with Zilibotti.

412
00:25:01,660 --> 00:25:04,830
It's a great paper.

413
00:25:04,830 --> 00:25:09,270
And the idea is simple,
but very clever,

414
00:25:09,270 --> 00:25:12,090
which is basically there is
a fixed cost to opening up

415
00:25:12,090 --> 00:25:14,190
a new sector.

416
00:25:14,190 --> 00:25:16,140
And at any moment
in time, of course,

417
00:25:16,140 --> 00:25:19,000
you have a limited
amount of resources.

418
00:25:19,000 --> 00:25:20,500
So you've got to choose--

419
00:25:20,500 --> 00:25:23,640
you can't literally
afford to do everything.

420
00:25:23,640 --> 00:25:25,320
You can only do a
subset of things.

421
00:25:25,320 --> 00:25:31,020
So you're naturally vulnerable
to shocks in technologies

422
00:25:31,020 --> 00:25:33,120
that can't cover your--

423
00:25:33,120 --> 00:25:35,970
that are not able
to cover your risk.

424
00:25:35,970 --> 00:25:39,000
Now within that there
are some policy issues

425
00:25:39,000 --> 00:25:40,690
we might come back
to at some point.

426
00:25:44,030 --> 00:25:46,450
But for a while,
clearly, these countries

427
00:25:46,450 --> 00:25:49,060
are going to just be more
volatile, because they're

428
00:25:49,060 --> 00:25:52,450
less able to diversify.

429
00:25:56,810 --> 00:26:00,190
Then it gets a little more--

430
00:26:00,190 --> 00:26:02,380
what should I say--

431
00:26:02,380 --> 00:26:04,600
provocative.

432
00:26:04,600 --> 00:26:08,350
Which is, maybe
volatility is actually

433
00:26:08,350 --> 00:26:10,510
not something you'd
like to get rid of

434
00:26:10,510 --> 00:26:12,040
and eventually can
afford to do it.

435
00:26:12,040 --> 00:26:17,990
Maybe it's actually a necessary
ingredient for growth.

436
00:26:17,990 --> 00:26:22,260
In other words, if we come back
to the big motivation, the idea

437
00:26:22,260 --> 00:26:26,970
that finance causes
growth, OK, let's do it.

438
00:26:26,970 --> 00:26:29,850
And finance and macro
stability, well, maybe you're

439
00:26:29,850 --> 00:26:31,590
going to have the
downside of that,

440
00:26:31,590 --> 00:26:33,840
that there has to be some
instability if you're

441
00:26:33,840 --> 00:26:35,640
going to get the growth.

442
00:26:35,640 --> 00:26:37,650
If you think this is
entirely provocative,

443
00:26:37,650 --> 00:26:40,260
just think about
the mean variance

444
00:26:40,260 --> 00:26:43,500
idea of holding assets.

445
00:26:43,500 --> 00:26:46,440
You don't get a
higher mean for free.

446
00:26:46,440 --> 00:26:48,050
Higher means come
with more risk.

447
00:26:52,910 --> 00:26:55,670
So we should maybe
take this seriously,

448
00:26:55,670 --> 00:26:57,770
as did these authors.

449
00:26:57,770 --> 00:26:58,940
I think this is in the QJE.

450
00:27:03,080 --> 00:27:06,080
And they document
that countries that

451
00:27:06,080 --> 00:27:08,810
have had occasional
financial crises

452
00:27:08,810 --> 00:27:11,440
do grow faster on average.

453
00:27:14,080 --> 00:27:21,760
How did they measure
that basically risk idea?

454
00:27:21,760 --> 00:27:26,680
Well, they don't want to take
a simple variance measure,

455
00:27:26,680 --> 00:27:29,680
because that's like small
or big, left or right tail.

456
00:27:29,680 --> 00:27:30,430
It doesn't matter.

457
00:27:30,430 --> 00:27:31,847
They're trying to
get at a crisis.

458
00:27:31,847 --> 00:27:34,450
So their ideas look at
the growth of credit,

459
00:27:34,450 --> 00:27:39,400
or actually quite negative
growth, when you have a bust.

460
00:27:39,400 --> 00:27:42,230
First, the boom, maybe
inching along, maybe a trend.

461
00:27:42,230 --> 00:27:43,150
Who knows how fast?

462
00:27:43,150 --> 00:27:47,710
But occasionally, wham, you
get one of these episodes where

463
00:27:47,710 --> 00:27:50,200
the bottom falls out.

464
00:27:50,200 --> 00:27:53,100
And I can't help but
say, it's a bit odd

465
00:27:53,100 --> 00:27:55,780
that since the US had a
recent financial crisis,

466
00:27:55,780 --> 00:27:58,090
everyone's kind of tearing
up all the old work

467
00:27:58,090 --> 00:28:02,050
and starting over as if we
need a whole new agenda.

468
00:28:02,050 --> 00:28:03,980
Whereas, countries
throughout the world

469
00:28:03,980 --> 00:28:06,970
for, Rogoff would
say, hundreds of years

470
00:28:06,970 --> 00:28:09,460
have been experiencing
financial crises.

471
00:28:09,460 --> 00:28:12,160
So certainly we
should pay attention.

472
00:28:12,160 --> 00:28:13,750
Not to say something's
good or bad,

473
00:28:13,750 --> 00:28:17,440
but there's lots of
historical experience.

474
00:28:17,440 --> 00:28:23,550
So they look at the
skewness of credit growth.

475
00:28:23,550 --> 00:28:26,620
I always get the sign wrong,
but positive basically

476
00:28:26,620 --> 00:28:28,570
puts the mass in the left tail.

477
00:28:32,027 --> 00:28:33,860
And they actually have
a little model, which

478
00:28:33,860 --> 00:28:35,600
I'll come back to in a second.

479
00:28:41,520 --> 00:28:43,580
So here are some pictures.

480
00:28:43,580 --> 00:28:48,450
India versus Thailand, real
credit growth and GDP growth.

481
00:28:48,450 --> 00:28:59,420
So the dotted line is Thailand,
which way back to 1980,

482
00:28:59,420 --> 00:29:02,000
was expanding its
financial system

483
00:29:02,000 --> 00:29:05,600
as measured by real credit.

484
00:29:05,600 --> 00:29:09,050
And then this is the
famous, if not infamous,

485
00:29:09,050 --> 00:29:14,570
infamous financial crisis
that started in July 2, 1997,

486
00:29:14,570 --> 00:29:17,090
when the Bank of Thailand
realized it couldn't continue

487
00:29:17,090 --> 00:29:19,940
to hold the exchange
rate because they

488
00:29:19,940 --> 00:29:25,490
had forward commitments they
seem to have forgot about.

489
00:29:25,490 --> 00:29:28,800
And that spread, as you
know, to many, not all,

490
00:29:28,800 --> 00:29:32,810
but many of the Asian countries,
not even nearby countries.

491
00:29:32,810 --> 00:29:38,540
I mean, Korea is not near,
but they too suffered.

492
00:29:38,540 --> 00:29:42,380
However, if you look
at the-- and India

493
00:29:42,380 --> 00:29:43,850
did not have this crisis.

494
00:29:43,850 --> 00:29:46,430
And India has been
bragging a lot lately

495
00:29:46,430 --> 00:29:50,390
how clever they were to be
conservative that they did not

496
00:29:50,390 --> 00:29:55,520
suffer from the US
financial episode

497
00:29:55,520 --> 00:29:58,670
the way some other
countries have had.

498
00:29:58,670 --> 00:30:01,040
But if you look at the big
picture, here at least,

499
00:30:01,040 --> 00:30:05,360
up through 02, you can see
Thailand has kind of this

500
00:30:05,360 --> 00:30:09,710
V-shaped drop, and then
it's right back on.

501
00:30:09,710 --> 00:30:12,380
My measurements show
that the growth rate was

502
00:30:12,380 --> 00:30:16,070
only 4% after as opposed
to 5% or 6% before,

503
00:30:16,070 --> 00:30:18,470
so maybe there was
a structural shift.

504
00:30:18,470 --> 00:30:23,270
But this little drop doesn't
affect the overall level,

505
00:30:23,270 --> 00:30:26,480
India's level remains lower.

506
00:30:26,480 --> 00:30:29,310
So during this period,
country of Thailand

507
00:30:29,310 --> 00:30:31,560
was one of the fastest growing
countries in the world.

508
00:30:31,560 --> 00:30:34,520
You hear a lot about
China now and, rightly so,

509
00:30:34,520 --> 00:30:38,250
India actually, also.

510
00:30:38,250 --> 00:30:41,770
But this is a big--

511
00:30:41,770 --> 00:30:44,600
and I didn't realize that these
guys had focused on Thailand.

512
00:30:44,600 --> 00:30:47,180
We have a lot more to say
about Thailand in the class,

513
00:30:47,180 --> 00:30:50,750
but obviously we're not alone
in thinking about Thailand

514
00:30:50,750 --> 00:30:54,380
relative to other countries.

515
00:30:54,380 --> 00:30:58,320
Now whether I agree with them
about the financial crisis is--

516
00:30:58,320 --> 00:30:59,880
or that dip is-- another story.

517
00:31:03,650 --> 00:31:06,410
Now how does their model work?

518
00:31:06,410 --> 00:31:08,990
Actually, the paper
uses this as motivation

519
00:31:08,990 --> 00:31:11,480
and then goes on
to try to work out

520
00:31:11,480 --> 00:31:15,500
how you would find
countries experiencing

521
00:31:15,500 --> 00:31:18,950
these periodic financial crises.

522
00:31:18,950 --> 00:31:22,250
So the idea is it's a
credit constraint model.

523
00:31:22,250 --> 00:31:24,860
So there are
borrowing constraints

524
00:31:24,860 --> 00:31:26,360
like collateral constraints.

525
00:31:26,360 --> 00:31:32,360
And if they were locked in
place without any leniency,

526
00:31:32,360 --> 00:31:39,330
you'd have high return
businesses starved for credit.

527
00:31:39,330 --> 00:31:41,960
But what happens is people
begin to anticipate that

528
00:31:41,960 --> 00:31:45,440
actually they can be more
generous in their lending

529
00:31:45,440 --> 00:31:46,880
because of the moral--

530
00:31:46,880 --> 00:31:50,180
so-called moral hazard problem,
which is the government's going

531
00:31:50,180 --> 00:31:52,460
to bail them out--

532
00:31:52,460 --> 00:31:54,890
so bail out the private sector.

533
00:31:54,890 --> 00:31:57,830
So this is sort of a
political economy story.

534
00:32:00,850 --> 00:32:08,640
There's kind of like a political
economy-driven boom-bust cycle.

535
00:32:11,490 --> 00:32:17,520
Now, again, I don't
have the equations here,

536
00:32:17,520 --> 00:32:23,310
but the paper's online and you
can read through their model.

537
00:32:26,640 --> 00:32:30,290
So they're saying
crises are costly.

538
00:32:30,290 --> 00:32:32,220
They're associated
with bankruptcies

539
00:32:32,220 --> 00:32:33,800
and a dead-weight loss.

540
00:32:37,052 --> 00:32:38,760
And it's true that
when a crisis happens,

541
00:32:38,760 --> 00:32:41,730
you get this depressing
effect on new credit,

542
00:32:41,730 --> 00:32:43,810
and on investment,
and hampering growth,

543
00:32:43,810 --> 00:32:48,270
which is a big
conversation in the US,

544
00:32:48,270 --> 00:32:49,930
and actually, to some
extent in Europe.

545
00:32:57,400 --> 00:32:59,560
And they're saying if
their model is right--

546
00:32:59,560 --> 00:33:03,610
or they believe they're right--
that this sort of effect

547
00:33:03,610 --> 00:33:07,120
is going to be larger in
countries where the contract

548
00:33:07,120 --> 00:33:12,350
enforceability issues are
more severe like borrowing

549
00:33:12,350 --> 00:33:13,280
constraints.

550
00:33:13,280 --> 00:33:16,610
On the other hand,
not too severe,

551
00:33:16,610 --> 00:33:18,980
otherwise they won't get
this leverage effect.

552
00:33:22,750 --> 00:33:25,870
So then we come to
the trilogy which

553
00:33:25,870 --> 00:33:27,490
is finance and volatility.

554
00:33:27,490 --> 00:33:30,740
We did finance and growth,
growth and volatility,

555
00:33:30,740 --> 00:33:33,070
now finance and volatility.

556
00:33:33,070 --> 00:33:36,790
And I don't have too much
more to say about this other

557
00:33:36,790 --> 00:33:40,600
than to offer the reminder
that that Acemoglu

558
00:33:40,600 --> 00:33:43,440
paper and the
Ranciere et al paper

559
00:33:43,440 --> 00:33:48,010
suggest that volatility might
diminish over time as countries

560
00:33:48,010 --> 00:33:49,720
have higher and higher
level of output.

561
00:33:54,510 --> 00:33:57,060
But as I said at the
beginning, in fact,

562
00:33:57,060 --> 00:33:59,810
not all these sort of--

563
00:33:59,810 --> 00:34:02,250
it must be more than a
bunch-- it must be more

564
00:34:02,250 --> 00:34:06,330
than the transitivity.

565
00:34:06,330 --> 00:34:10,050
If A is associated with B, and
B is associated with C, than A

566
00:34:10,050 --> 00:34:12,060
should be associated
with C. And in the data,

567
00:34:12,060 --> 00:34:13,929
it doesn't go that way.

568
00:34:13,929 --> 00:34:17,159
So clearly there's something
being left out here.

569
00:34:17,159 --> 00:34:22,290
This paper, which I became
aware of more recently--

570
00:34:30,350 --> 00:34:36,080
and they seem to find that, in
fact, deeper financial systems

571
00:34:36,080 --> 00:34:39,270
are associated with
lower volatility.

572
00:34:39,270 --> 00:34:42,590
But I checked this
again this morning,

573
00:34:42,590 --> 00:34:45,150
they also have in
the regression,

574
00:34:45,150 --> 00:34:51,080
the levels of income and
the growth rates of income.

575
00:34:51,080 --> 00:34:53,540
So in other words,
they have to control

576
00:34:53,540 --> 00:34:55,219
for the other parts
of the trilogy,

577
00:34:55,219 --> 00:34:59,670
basically, and then
eek out this effect.

578
00:34:59,670 --> 00:35:03,260
Countries that are poor are
often dependent on trade.

579
00:35:03,260 --> 00:35:07,850
There are problems when
exchange rates move.

580
00:35:07,850 --> 00:35:10,860
There are issues about
devaluations, and so on.

581
00:35:10,860 --> 00:35:15,940
So they kind of have to
control for all of that.

582
00:35:15,940 --> 00:35:20,380
And then offer the
suggestion that finance

583
00:35:20,380 --> 00:35:23,480
might be a good thing.

584
00:35:23,480 --> 00:35:25,780
But believe me,
countries don't know.

585
00:35:25,780 --> 00:35:28,630
The IMF really
doesn't know either,

586
00:35:28,630 --> 00:35:33,190
and they're very interested
in the state of the literature

587
00:35:33,190 --> 00:35:36,040
and in further research
to try to find out.

588
00:35:36,040 --> 00:35:38,590
Because they have
client countries

589
00:35:38,590 --> 00:35:40,480
who want to know what to do.

590
00:35:40,480 --> 00:35:43,870
They go to the IMF, for
example, to get advice.

591
00:35:47,900 --> 00:35:55,070
So then we come to this
sort of policy discussion

592
00:35:55,070 --> 00:35:57,050
that's going on today.

593
00:35:57,050 --> 00:36:03,840
And to be candid, it
seems to basically

594
00:36:03,840 --> 00:36:06,900
depart from this
literature, from both

595
00:36:06,900 --> 00:36:11,550
the theoretical and the
empirical literature.

596
00:36:11,550 --> 00:36:15,330
Levine writes a new
paper now called

597
00:36:15,330 --> 00:36:20,710
Regulating Finance and
Regulators to Promote Growth.

598
00:36:20,710 --> 00:36:23,020
It's like, well, we
got to fix the problem.

599
00:36:23,020 --> 00:36:26,630
We know the financial crisis
was caused by bad regulation,

600
00:36:26,630 --> 00:36:31,870
so let's jump in and
regulate somehow or provide

601
00:36:31,870 --> 00:36:33,160
the right incentives.

602
00:36:42,050 --> 00:36:47,390
So how important is it that the
operation of-- how important

603
00:36:47,390 --> 00:36:49,250
is the operation of
the financial system

604
00:36:49,250 --> 00:36:51,200
for an economic growth?

605
00:36:51,200 --> 00:36:55,760
And which financial regulatory
reforms will improve

606
00:36:55,760 --> 00:36:57,320
financial sector operations?

607
00:36:57,320 --> 00:37:01,730
He's sort of assuming, in
some sense, an the answer

608
00:37:01,730 --> 00:37:06,800
that the other literature is,
and his own review was sort of

609
00:37:06,800 --> 00:37:10,280
much more equivocal about.

610
00:37:21,140 --> 00:37:25,880
This is another example,
this Thorsten Beck's work,

611
00:37:25,880 --> 00:37:31,810
which is highly regarded,
I think, in policy circles.

612
00:37:31,810 --> 00:37:35,420
And the idea is to come up
with a financial possibilities

613
00:37:35,420 --> 00:37:37,530
frontier.

614
00:37:37,530 --> 00:37:39,320
And again, this is
driven by the idea

615
00:37:39,320 --> 00:37:41,690
that we're going
to give countries

616
00:37:41,690 --> 00:37:43,940
specific recommendations.

617
00:37:43,940 --> 00:37:47,570
They want to know where
they stand relative

618
00:37:47,570 --> 00:37:52,290
to other countries, and what, if
anything, they should be doing.

619
00:37:52,290 --> 00:37:53,090
So here's the idea.

620
00:37:57,360 --> 00:38:03,160
This is the, say, the financial
possibilities frontier.

621
00:38:03,160 --> 00:38:05,500
This country is in on it.

622
00:38:05,500 --> 00:38:07,980
So somehow they
should move up there.

623
00:38:11,640 --> 00:38:16,020
This country B actually
has a higher level

624
00:38:16,020 --> 00:38:19,630
of financial depth than A.

625
00:38:19,630 --> 00:38:22,680
But rather than rest
on their levels,

626
00:38:22,680 --> 00:38:25,170
the idea is that if we knew
what this frontier was,

627
00:38:25,170 --> 00:38:26,880
it would tell us
they're actually further

628
00:38:26,880 --> 00:38:29,400
below the frontier than A was.

629
00:38:29,400 --> 00:38:35,070
They have more ground to go
to get up to that frontier.

630
00:38:38,580 --> 00:38:45,840
And then this will get your
eye here beyond the frontier.

631
00:38:45,840 --> 00:38:47,820
So the idea is that
some countries are just

632
00:38:47,820 --> 00:38:52,040
basically too deep, and
they should pull back.

633
00:38:56,342 --> 00:38:57,300
AUDIENCE: Question.

634
00:38:57,300 --> 00:38:58,437
ROBERT TOWNSEND: Yeah.

635
00:38:58,437 --> 00:39:00,270
AUDIENCE: In this paper,
what sort of things

636
00:39:00,270 --> 00:39:02,640
go on the x-axis?

637
00:39:02,640 --> 00:39:06,150
ROBERT TOWNSEND: Yeah,
so this is basically

638
00:39:06,150 --> 00:39:09,660
a huge sort of cross-country
regression with tons

639
00:39:09,660 --> 00:39:12,600
of x-variables.

640
00:39:12,600 --> 00:39:13,650
The next slide has--

641
00:39:16,730 --> 00:39:17,810
gives you some sense.

642
00:39:17,810 --> 00:39:21,080
There are variables in there
to try to control somehow

643
00:39:21,080 --> 00:39:23,560
for institutions.

644
00:39:23,560 --> 00:39:30,290
There are surveys of how easy
it is to conduct business

645
00:39:30,290 --> 00:39:32,540
in these countries.

646
00:39:32,540 --> 00:39:34,670
There are other
things associated

647
00:39:34,670 --> 00:39:36,020
with structural problems.

648
00:39:36,020 --> 00:39:39,595
There is even policy
stuff, although honestly, I

649
00:39:39,595 --> 00:39:40,220
don't remember.

650
00:39:40,220 --> 00:39:44,120
We can look it up and
see, sort of categorize

651
00:39:44,120 --> 00:39:45,260
all the x-variables.

652
00:39:49,030 --> 00:39:58,180
Now, I'm somewhat sympathetic
to looking at data.

653
00:39:58,180 --> 00:40:06,263
So we don't want to say this
is completely uninformative.

654
00:40:11,060 --> 00:40:13,160
And I am sympathetic
with the idea

655
00:40:13,160 --> 00:40:16,970
that knowing what has happened
to different countries

656
00:40:16,970 --> 00:40:20,240
at different points in
times might kind of give us

657
00:40:20,240 --> 00:40:22,290
pause, like leading indicators.

658
00:40:22,290 --> 00:40:24,920
There's a whole group.

659
00:40:24,920 --> 00:40:30,780
Actually, I was talking to
Ken Singleton the other day,

660
00:40:30,780 --> 00:40:34,450
and I don't think this
work is quite public yet,

661
00:40:34,450 --> 00:40:39,860
but there the idea is he's in
finance and a very good person.

662
00:40:39,860 --> 00:40:42,170
And he's trying to
measure the risk premium

663
00:40:42,170 --> 00:40:45,910
based on observed asset prices.

664
00:40:45,910 --> 00:40:51,110
And they seem to be showing that
when the risk premium is low,

665
00:40:51,110 --> 00:40:53,670
that's when you
run into trouble.

666
00:40:53,670 --> 00:40:56,300
Now that might sound like
counterintuitive at first,

667
00:40:56,300 --> 00:41:00,260
but the point is, if the markets
are saying, or understating,

668
00:41:00,260 --> 00:41:03,770
the amount of risk, then you
can imagine the financial system

669
00:41:03,770 --> 00:41:06,840
is just pumping too much money
into otherwise risky things.

670
00:41:06,840 --> 00:41:10,880
And that's when countries
run into trouble eventually.

671
00:41:10,880 --> 00:41:14,530
So it's good to have early
warnings, that's good.

672
00:41:14,530 --> 00:41:17,920
But that seems to me
the challenge here

673
00:41:17,920 --> 00:41:20,800
is to get beneath the frontier.

674
00:41:23,950 --> 00:41:28,480
In fact, I find the whole
jargon a bit misleading,

675
00:41:28,480 --> 00:41:30,880
because there is a utility
possibilities frontier which

676
00:41:30,880 --> 00:41:34,175
is like the bread and butter of
general equilibrium economics.

677
00:41:34,175 --> 00:41:37,705
One point is you can't
get beyond the frontier.

678
00:41:37,705 --> 00:41:38,830
Anyway, there's a question.

679
00:41:38,830 --> 00:41:41,620
AUDIENCE: Oh, I was going
to ask, in what sense

680
00:41:41,620 --> 00:41:43,990
does he talk about frontier?

681
00:41:43,990 --> 00:41:45,850
Because it sort of
looks from his graph

682
00:41:45,850 --> 00:41:47,920
as if he was saying this
is the optimal level

683
00:41:47,920 --> 00:41:49,420
of financial deepening for you.

684
00:41:49,420 --> 00:41:50,990
That's not what
a frontier means.

685
00:41:50,990 --> 00:41:53,888
So then he's talking
about pareto optimality.

686
00:41:53,888 --> 00:41:55,180
ROBERT TOWNSEND: No, that's me.

687
00:41:55,180 --> 00:41:58,780
So this is one instance
where my thoughts kind of

688
00:41:58,780 --> 00:42:03,130
merged into his, so
I'm the one worried

689
00:42:03,130 --> 00:42:05,270
about pareto optimality.

690
00:42:05,270 --> 00:42:08,980
AUDIENCE: So are there issues
with aggregation at that point,

691
00:42:08,980 --> 00:42:12,325
if you're worrying about
utility possibilities

692
00:42:12,325 --> 00:42:17,440
frontiers and pareto
optimality and stuff?

693
00:42:17,440 --> 00:42:21,250
ROBERT TOWNSEND: To construct a
utility possibilities frontier,

694
00:42:21,250 --> 00:42:23,460
we need to specify the model.

695
00:42:23,460 --> 00:42:26,080
We're going to have to layout
preferences, endowments,

696
00:42:26,080 --> 00:42:26,812
and technology.

697
00:42:26,812 --> 00:42:28,270
We're going to have
to take a stand

698
00:42:28,270 --> 00:42:35,530
on possible imperfections in
the credit and financial system.

699
00:42:35,530 --> 00:42:39,400
And just say that
sometimes it will turn out

700
00:42:39,400 --> 00:42:45,430
that despite limited commitment
or some information problems,

701
00:42:45,430 --> 00:42:48,250
it's a well-defined notion,
and markets should allow

702
00:42:48,250 --> 00:42:50,170
you to end up on the frontier.

703
00:42:50,170 --> 00:42:56,080
But there are other things like
pecuniary externalities, where

704
00:42:56,080 --> 00:42:58,840
if there is no remedy,
then you could end up

705
00:42:58,840 --> 00:43:01,610
off the frontier due
to an externality.

706
00:43:01,610 --> 00:43:04,570
So something close
to the same language

707
00:43:04,570 --> 00:43:09,340
works, but if I'm
understanding you correctly,

708
00:43:09,340 --> 00:43:13,760
they're not representative
consumer models, typically.

709
00:43:13,760 --> 00:43:16,180
They have all the
heterogeneity in them,

710
00:43:16,180 --> 00:43:18,550
and there are issues
about whether there

711
00:43:18,550 --> 00:43:22,390
is a pseudo representative
consumer that somehow

712
00:43:22,390 --> 00:43:24,250
approximates the behavior.

713
00:43:24,250 --> 00:43:27,520
So those are key
issues in these models.

714
00:43:27,520 --> 00:43:32,703
And most of them don't
aggregate up that way.

715
00:43:32,703 --> 00:43:34,120
But on the other
hand, we can look

716
00:43:34,120 --> 00:43:36,340
at the welfare implications.

717
00:43:42,670 --> 00:43:47,470
And finally,
there's this paper--

718
00:43:47,470 --> 00:43:51,040
an IMF working paper--

719
00:43:51,040 --> 00:43:52,660
on how to deal
with credit booms.

720
00:43:52,660 --> 00:43:57,760
And this seems to be a summary
of the current consensus.

721
00:43:57,760 --> 00:44:00,730
I hesitate to call it that.

722
00:44:00,730 --> 00:44:04,900
That credit booms, buttress,
investment and consumption,

723
00:44:04,900 --> 00:44:06,880
they can contribute
to long run deepening,

724
00:44:06,880 --> 00:44:10,930
but they often end up in costly
balance sheet disallocations,

725
00:44:10,930 --> 00:44:17,880
and dislocations, and more
often than acceptable,

726
00:44:17,880 --> 00:44:21,240
in devastating financial
crises whose cost greatly

727
00:44:21,240 --> 00:44:24,360
exceeds the benefit
associated with boom.

728
00:44:24,360 --> 00:44:33,340
So this is the view somehow
that is pretty typical.

729
00:44:33,340 --> 00:44:37,120
And so you end up
with this sort of goal

730
00:44:37,120 --> 00:44:40,660
to keep an eye on
financial liberalizations

731
00:44:40,660 --> 00:44:43,090
to limit credit
expansion somehow,

732
00:44:43,090 --> 00:44:45,280
although they don't
quite know how.

733
00:44:45,280 --> 00:44:49,000
And the menu of things
that they're looking at

734
00:44:49,000 --> 00:44:50,920
is very Basile like.

735
00:44:50,920 --> 00:44:54,850
They're basically ad hoc
international best practice

736
00:44:54,850 --> 00:44:56,500
statements.

737
00:44:56,500 --> 00:45:00,670
As if it were obvious how to
regulate the financial system.

738
00:45:00,670 --> 00:45:05,650
Those regulations are not coming
from fundamental considerations

739
00:45:05,650 --> 00:45:08,830
of how credit markets
work, and how they interact

740
00:45:08,830 --> 00:45:10,571
with the macro economy.

741
00:45:13,940 --> 00:45:21,110
So is there another way,
an alternative approach?

742
00:45:21,110 --> 00:45:23,360
I guess it's
self-evident from what

743
00:45:23,360 --> 00:45:28,640
I've been saying in this
introduction is to understand

744
00:45:28,640 --> 00:45:31,820
something I'm going to refer
to short-handedly as Applied

745
00:45:31,820 --> 00:45:36,920
General Equilibrium Development
Economics to understand

746
00:45:36,920 --> 00:45:41,300
the unit of analysis at
which this can be applied,

747
00:45:41,300 --> 00:45:46,490
and to clarify what that is
relative to other general

748
00:45:46,490 --> 00:45:48,652
equilibrium models
in the literature.

749
00:45:51,680 --> 00:45:56,600
So this may be a familiar,
if shocking, picture,

750
00:45:56,600 --> 00:45:59,060
so to speak, in the context.

751
00:45:59,060 --> 00:46:00,556
If you haven't seen
it before, this

752
00:46:00,556 --> 00:46:06,250
is a picture of a
village in India.

753
00:46:06,250 --> 00:46:11,730
We can enumerate the
household by their wealth,

754
00:46:11,730 --> 00:46:16,060
order them by their wealth,
and look at these years

755
00:46:16,060 --> 00:46:19,690
where panel data were gathered
in these so-called ICRISAT

756
00:46:19,690 --> 00:46:20,680
villages.

757
00:46:20,680 --> 00:46:22,720
This is [INAUDIBLE].

758
00:46:22,720 --> 00:46:26,250
And here you can see the
ups and downs of income.

759
00:46:26,250 --> 00:46:29,740
Idiosyncratic
aggregate risk, there's

760
00:46:29,740 --> 00:46:32,200
not a lot of co-movement here.

761
00:46:32,200 --> 00:46:34,660
The peaks are covering
up the valleys behind it.

762
00:46:34,660 --> 00:46:38,680
Households are not having up
and ups and downs together.

763
00:46:38,680 --> 00:46:43,330
There's plenty of
risk in the system.

764
00:46:43,330 --> 00:46:47,620
And this is how the
consumption picture turns out.

765
00:46:47,620 --> 00:46:50,290
I often refer to this
in a shorthand way

766
00:46:50,290 --> 00:46:53,370
as the Rocky Mountains
versus Kansas.

767
00:46:53,370 --> 00:46:57,010
And Kansas is pretty boring.

768
00:46:57,010 --> 00:46:59,800
It's very flat.

769
00:46:59,800 --> 00:47:03,860
But what it shows you very
dramatically is that somehow,

770
00:47:03,860 --> 00:47:07,510
some way, in these
village economies,

771
00:47:07,510 --> 00:47:11,360
they have figured out a way
to smooth a lot of the risk.

772
00:47:11,360 --> 00:47:12,970
In fact, they come very close--

773
00:47:12,970 --> 00:47:16,300
not totally there--
but very close

774
00:47:16,300 --> 00:47:18,790
to achieving the
optimal allocation

775
00:47:18,790 --> 00:47:25,090
of risk-bearing that
Arrow-Debreu and so on

776
00:47:25,090 --> 00:47:28,990
have been talking about.

777
00:47:28,990 --> 00:47:35,520
Now don't dismiss this
as OK, OK, we're back.

778
00:47:35,520 --> 00:47:37,127
And thank God it's development.

779
00:47:37,127 --> 00:47:38,085
We're back in villages.

780
00:47:41,910 --> 00:47:44,940
The point is that you
can use this framework

781
00:47:44,940 --> 00:47:48,510
and the data are available to
do this not just at the village

782
00:47:48,510 --> 00:47:52,170
level, collections of villages,
regions in the country,

783
00:47:52,170 --> 00:47:53,370
or even across countries.

784
00:47:53,370 --> 00:47:55,590
And every single one of
those things is done,

785
00:47:55,590 --> 00:47:57,520
and there's a huge literature.

786
00:47:57,520 --> 00:48:01,380
So we know quite a bit about
the allocation of risk-bearing,

787
00:48:01,380 --> 00:48:06,090
and in fact, where systems fail.

788
00:48:06,090 --> 00:48:11,575
Here's another look
at sort of an economy.

789
00:48:11,575 --> 00:48:15,210
This turns out to
be another village.

790
00:48:15,210 --> 00:48:17,370
And it's about the
diversification issue.

791
00:48:19,980 --> 00:48:22,470
So if you could do all this
great ex post smoothing

792
00:48:22,470 --> 00:48:26,010
somehow, you should
just specialize

793
00:48:26,010 --> 00:48:29,880
in what you're good at,
and your asset holdings

794
00:48:29,880 --> 00:48:33,570
might appear pretty specialized.

795
00:48:33,570 --> 00:48:37,170
And maybe the bulk
of your consumption

796
00:48:37,170 --> 00:48:39,660
is determined ex post by
borrowing, and lending,

797
00:48:39,660 --> 00:48:41,910
and gift-giving, and
whatever else they're doing.

798
00:48:41,910 --> 00:48:44,730
But no, not in medieval England.

799
00:48:44,730 --> 00:48:49,040
They divided up their land
in an extraordinary way.

800
00:48:49,040 --> 00:48:54,990
A typical farmer would have
60 or 70 plots fragmented

801
00:48:54,990 --> 00:48:56,970
throughout the village.

802
00:48:56,970 --> 00:49:02,700
This is also true in Bolivia,
up in Altiplano and so on,

803
00:49:02,700 --> 00:49:06,350
running down
different elevations.

804
00:49:06,350 --> 00:49:12,010
So the point here is not
all economies are alike,

805
00:49:12,010 --> 00:49:13,750
not even all villages are alike.

806
00:49:13,750 --> 00:49:16,660
And looking at
data in this case,

807
00:49:16,660 --> 00:49:18,940
partly through this
institution, you

808
00:49:18,940 --> 00:49:23,320
can see that the ex post
allocation of risk-bearing

809
00:49:23,320 --> 00:49:26,410
must evidently be more limited.

810
00:49:26,410 --> 00:49:29,470
In this case, it took a
private information model

811
00:49:29,470 --> 00:49:30,925
to try to explain that outcome.

812
00:49:35,440 --> 00:49:40,000
Again, to reassure you that
this is not just macro,

813
00:49:40,000 --> 00:49:45,393
it's macro development,
macro devo, devoted to macro,

814
00:49:45,393 --> 00:49:47,560
or I don't know, you can
turn that any way you want.

815
00:49:48,060 --> 00:49:49,040
[LAUGHTER]

816
00:49:49,540 --> 00:49:54,190
Someone made a joke
once, if in the slides

817
00:49:54,190 --> 00:49:56,540
you don't have a picture
of a villager or something,

818
00:49:56,540 --> 00:49:58,530
then you're not really
a development economist.

819
00:49:58,530 --> 00:50:02,000
I actually took this picture.

820
00:50:02,000 --> 00:50:05,810
But so you got land,
labor, and capital.

821
00:50:06,310 --> 00:50:08,286
[LAUGHTER]

822
00:50:10,760 --> 00:50:12,490
OK.

823
00:50:12,490 --> 00:50:14,050
So I love villages.

824
00:50:14,050 --> 00:50:17,290
And the reason is, you really
don't have to make things up.

825
00:50:17,290 --> 00:50:21,040
You can actually measure things
and get an approximate sense

826
00:50:21,040 --> 00:50:27,700
of what the preference, the
endowments, and technology are.

827
00:50:27,700 --> 00:50:30,970
Preferences are a bit trickier.

828
00:50:30,970 --> 00:50:35,540
So that's the way the language
we use as general equilibrium

829
00:50:35,540 --> 00:50:37,120
theorists--

830
00:50:37,120 --> 00:50:39,070
preferences,
endowments, technology,

831
00:50:39,070 --> 00:50:41,095
as I said in answer to
the earlier question.

832
00:50:44,220 --> 00:50:46,180
And then you can do
this kind of adding up.

833
00:50:46,180 --> 00:50:50,470
So this is like taking these
villages, the Thai villages,

834
00:50:50,470 --> 00:50:55,060
and adding up all the
households in the village

835
00:50:55,060 --> 00:50:56,248
with the financial accounts.

836
00:50:56,248 --> 00:50:57,790
We'll get to that
in a minute, that's

837
00:50:57,790 --> 00:50:59,920
the measurement part, the
midpoint of the course.

838
00:51:03,130 --> 00:51:07,030
And you can see in one region
near Bangkok, village output--

839
00:51:07,030 --> 00:51:09,680
GDP, except it's not national--

840
00:51:09,680 --> 00:51:10,600
is going down.

841
00:51:10,600 --> 00:51:12,500
In the other regions,
it's kind of going up.

842
00:51:12,500 --> 00:51:15,260
This is a second
panel, so to speak,

843
00:51:15,260 --> 00:51:18,910
is where their savings is
put into real or financial

844
00:51:18,910 --> 00:51:19,455
investments.

845
00:51:19,455 --> 00:51:21,080
And here you have
the balance payments.

846
00:51:23,850 --> 00:51:25,950
So it shouldn't be
too much of a stretch

847
00:51:25,950 --> 00:51:28,740
to realize that if you
have the measurement,

848
00:51:28,740 --> 00:51:31,680
you can indeed zoom
in and zoom out.

849
00:51:31,680 --> 00:51:34,870
You can go from households
out to the whole country,

850
00:51:34,870 --> 00:51:36,660
and intermediate
steps in between,

851
00:51:36,660 --> 00:51:38,230
and use these standard tools.

852
00:51:38,230 --> 00:51:40,710
In this case, we've
ended up with the tools

853
00:51:40,710 --> 00:51:43,425
that international
economists use,

854
00:51:43,425 --> 00:51:45,300
I'm using them internally
within the country.

855
00:51:48,645 --> 00:51:50,520
But at least if the
measurement is in common,

856
00:51:50,520 --> 00:51:52,080
we are allowed to do that.

857
00:51:52,080 --> 00:51:54,870
I'll just pause briefly
to talk about the data.

858
00:51:54,870 --> 00:51:57,360
I've already
mentioned we have it,

859
00:51:57,360 --> 00:52:01,240
and you should feel
free to use it.

860
00:52:01,240 --> 00:52:02,530
It's quite extensive.

861
00:52:02,530 --> 00:52:05,123
It's 15 years worth of data.

862
00:52:05,123 --> 00:52:06,165
And some of it's monthly.

863
00:52:08,840 --> 00:52:10,940
It's both rural and urban.

864
00:52:10,940 --> 00:52:16,730
Yeah, I'm in city
neighborhoods and spread out

865
00:52:16,730 --> 00:52:18,650
throughout the kingdom.

866
00:52:21,990 --> 00:52:25,340
This is zooming in and
zooming out, literally.

867
00:52:25,340 --> 00:52:30,170
So here you can see almost
like an aerial photo

868
00:52:30,170 --> 00:52:34,010
with the village locations
where we collect data.

869
00:52:37,310 --> 00:52:41,390
All the way over here you
can see zooming way out,

870
00:52:41,390 --> 00:52:44,480
where it is basically.

871
00:52:44,480 --> 00:52:49,640
It's somewhat near Bangkok,
this particular sample site,

872
00:52:49,640 --> 00:52:51,980
and this is sort of
the intermediate level.

873
00:52:51,980 --> 00:52:57,410
Here we're plotting at
the level of provinces

874
00:52:57,410 --> 00:53:03,770
the wealth level, which is in
common with those two diagrams.

875
00:53:03,770 --> 00:53:06,470
Well, we have this
extensive GIS system.

876
00:53:06,470 --> 00:53:11,120
I've made a campaign of putting
all available secondary data

877
00:53:11,120 --> 00:53:15,740
that I could get a hold of
over the years, a lot of it's

878
00:53:15,740 --> 00:53:17,480
available through
Dataverse, Harvard MIT.

879
00:53:20,060 --> 00:53:23,150
And we are in the
process of putting this

880
00:53:23,150 --> 00:53:26,540
onto this common
GIS platform where

881
00:53:26,540 --> 00:53:31,640
you can look at bank locations,
and factory locations, and road

882
00:53:31,640 --> 00:53:34,500
networks, and compute
travel time, and so on.

883
00:53:34,500 --> 00:53:37,670
So one of the--

884
00:53:37,670 --> 00:53:44,060
if you're interested, I can
have my GIS person maybe

885
00:53:44,060 --> 00:53:49,790
have a special session and
show you a little bit about how

886
00:53:49,790 --> 00:53:53,660
to use this system and
the kind of work that's

887
00:53:53,660 --> 00:53:56,120
possible to do with it.

888
00:53:56,120 --> 00:53:56,820
It's coming.

889
00:53:56,820 --> 00:53:59,000
There isn't that much
work in economics

890
00:53:59,000 --> 00:54:02,310
that has explicit general
equilibrium spatial modeling,

891
00:54:02,310 --> 00:54:03,080
but it's coming.

892
00:54:05,910 --> 00:54:09,410
So it's a wide open and
interesting research topic.

893
00:54:12,042 --> 00:54:14,000
So what is this general
equilibrium development

894
00:54:14,000 --> 00:54:14,500
approach?

895
00:54:14,500 --> 00:54:18,680
Well, it's not just
individual maximization.

896
00:54:18,680 --> 00:54:20,630
It's not just
partial equilibrium.

897
00:54:20,630 --> 00:54:22,430
It's general equilibrium.

898
00:54:22,430 --> 00:54:27,620
So things like the economy-wide
interest rate and wage rates

899
00:54:27,620 --> 00:54:32,150
are very much on our minds.

900
00:54:32,150 --> 00:54:34,670
From a development
perspective, it's not just

901
00:54:34,670 --> 00:54:38,900
the impact of some savings
account or improved credit

902
00:54:38,900 --> 00:54:45,210
product on individuals, it's
about efficiency of the system.

903
00:54:45,210 --> 00:54:48,050
In other words, it's not
just on the credit side,

904
00:54:48,050 --> 00:54:50,542
it's kind of on
the lending side.

905
00:54:50,542 --> 00:54:52,250
The money for lending
comes from savings.

906
00:54:52,250 --> 00:54:54,950
You have to think
about intermediation

907
00:54:54,950 --> 00:54:58,760
and the flow of funds,
and judge efficiency

908
00:54:58,760 --> 00:55:03,920
using this pareto criterion.

909
00:55:03,920 --> 00:55:09,980
It's not a foregone conclusion
what the best fitting market

910
00:55:09,980 --> 00:55:11,660
or contract structure is.

911
00:55:11,660 --> 00:55:14,030
That's something that
comes out of the analysis.

912
00:55:14,030 --> 00:55:17,630
That's why it's so good
to have the micro data,

913
00:55:17,630 --> 00:55:20,320
because you can actually test.

914
00:55:20,320 --> 00:55:21,640
You could have village--

915
00:55:21,640 --> 00:55:25,150
one village doing quite
well and a whole region

916
00:55:25,150 --> 00:55:29,850
in another part of the
country doing terribly.

917
00:55:29,850 --> 00:55:32,650
But we don't have
to guess or assume,

918
00:55:32,650 --> 00:55:38,020
because we'll in the
course acquire some tools

919
00:55:38,020 --> 00:55:39,460
for doing that analysis.

920
00:55:39,460 --> 00:55:45,130
So clearly it's not the ever
varying Washington consensus,

921
00:55:45,130 --> 00:55:48,040
which is for liberalization
and now for regulation.

922
00:55:48,040 --> 00:55:50,860
Look at Europe, just going
back and forth not knowing

923
00:55:50,860 --> 00:55:51,490
what to do.

924
00:55:51,490 --> 00:55:52,670
The debt's too high.

925
00:55:52,670 --> 00:55:54,040
We need to restrain it.

926
00:55:54,040 --> 00:55:54,680
Oh, my God!

927
00:55:54,680 --> 00:55:55,555
We need the stimulus.

928
00:55:59,430 --> 00:56:01,885
So we think through those
things at the same time

929
00:56:01,885 --> 00:56:02,635
with these models.

930
00:56:08,440 --> 00:56:12,180
And I don't mean to be
saying it's all about the US,

931
00:56:12,180 --> 00:56:12,780
obviously.

932
00:56:12,780 --> 00:56:13,700
It's about India.

933
00:56:13,700 --> 00:56:15,160
It's about the Philippines.

934
00:56:15,160 --> 00:56:17,520
It's Kenya where they are
allowing cell phones as

935
00:56:17,520 --> 00:56:21,630
opposed to Thailand where
they're effectively ruling it

936
00:56:21,630 --> 00:56:22,710
out.

937
00:56:22,710 --> 00:56:25,920
So those are the debates
that are happening

938
00:56:25,920 --> 00:56:28,620
in developing countries.

939
00:56:28,620 --> 00:56:31,620
You've got some stuff on the
website if you want to see.

940
00:56:31,620 --> 00:56:33,540
I will be tracking
Thailand a lot,

941
00:56:33,540 --> 00:56:35,850
but I've also been
writing a book on Mexico.

942
00:56:35,850 --> 00:56:39,840
And you'll see me
throw in some slides.

943
00:56:39,840 --> 00:56:43,200
I think the danger here is
you see too much Thailand,

944
00:56:43,200 --> 00:56:45,630
and then you begin to think
it's all about Thailand.

945
00:56:45,630 --> 00:56:48,510
Well, A, I'm not the only
one worrying about Thailand.

946
00:56:48,510 --> 00:56:53,760
But B, seeing other countries
and seeing the same analysis

947
00:56:53,760 --> 00:56:56,580
done in two different
countries really brings

948
00:56:56,580 --> 00:56:57,540
this stuff to light.

949
00:56:57,540 --> 00:57:00,420
So that's why
Mexico is sometimes

950
00:57:00,420 --> 00:57:01,440
paired in these slides.

951
00:57:05,070 --> 00:57:07,230
So the goal here
is to explain, yes,

952
00:57:07,230 --> 00:57:09,540
to understand what's
out there, but it's also

953
00:57:09,540 --> 00:57:16,530
normative to try to come
up with policies that

954
00:57:16,530 --> 00:57:18,780
allow more inclusive
financial systems

955
00:57:18,780 --> 00:57:20,520
if the data indicate that.

956
00:57:20,520 --> 00:57:24,360
But also to think about
the overall efficiency,

957
00:57:24,360 --> 00:57:28,500
and to think about market
design and optimal regulation.

958
00:57:31,140 --> 00:57:33,870
There is some logic in
all of this, basically.

959
00:57:33,870 --> 00:57:37,140
You assume something
or several things,

960
00:57:37,140 --> 00:57:39,330
you go out there and test.

961
00:57:39,330 --> 00:57:40,260
It may fit.

962
00:57:40,260 --> 00:57:42,150
If it does fit,
you've understood

963
00:57:42,150 --> 00:57:45,180
the financial contracts
and the imperfection.

964
00:57:45,180 --> 00:57:47,340
If it's full information
for risk sharing,

965
00:57:47,340 --> 00:57:49,230
then we have this
wonderful standard

966
00:57:49,230 --> 00:57:52,335
and it's clear, at least at
the village level in India,

967
00:57:52,335 --> 00:57:56,100
you shouldn't be intervening
with particular households.

968
00:57:56,100 --> 00:57:57,840
On the other hand,
if you reject,

969
00:57:57,840 --> 00:58:03,000
maybe it's due to a moral
hazard problem, the problem

970
00:58:03,000 --> 00:58:05,040
that people don't repay
and they walk away

971
00:58:05,040 --> 00:58:06,930
from their investments.

972
00:58:06,930 --> 00:58:08,570
The data can help indicate that.

973
00:58:08,570 --> 00:58:11,870
Now then the question is,
is it still constrained?

974
00:58:11,870 --> 00:58:13,860
Maybe it's second best optimal.

975
00:58:13,860 --> 00:58:19,960
Maybe this first best
standard is an illusion.

976
00:58:19,960 --> 00:58:22,450
But maybe when it's
second best, things

977
00:58:22,450 --> 00:58:26,290
like allowing more goods to
play the role as collateral,

978
00:58:26,290 --> 00:58:30,370
like changing the legal
system, those kinds of reforms

979
00:58:30,370 --> 00:58:33,550
might be indicated.

980
00:58:38,450 --> 00:58:44,480
Now there is a lot of general
equilibrium literature

981
00:58:44,480 --> 00:58:48,350
to sort of contrast with this
general equilibrium development

982
00:58:48,350 --> 00:58:50,360
point of view.

983
00:58:50,360 --> 00:58:52,880
This goes way back
to the foundations

984
00:58:52,880 --> 00:58:54,290
of general equilibrium.

985
00:58:54,290 --> 00:58:58,820
Herb Scarf, and his
students, Chauvin and Wally,

986
00:58:58,820 --> 00:59:01,760
that's called Applied
General Equilibrium

987
00:59:01,760 --> 00:59:04,010
and does not allow uncertainty.

988
00:59:04,010 --> 00:59:06,230
This type of modeling
is still used,

989
00:59:06,230 --> 00:59:08,160
although it's in
kind of pockets.

990
00:59:08,160 --> 00:59:11,210
It's used within the World Bank.

991
00:59:11,210 --> 00:59:15,090
All our climate guys over
here at MIT use this stuff.

992
00:59:15,090 --> 00:59:18,740
That's the basic conceptual
framework, the CGE model,

993
00:59:18,740 --> 00:59:23,060
that lies behind these
conclusions about how

994
00:59:23,060 --> 00:59:26,870
climate change is going
to influence the economy.

995
00:59:26,870 --> 00:59:31,670
So it is quite widely used.

996
00:59:31,670 --> 00:59:34,190
And there's some
great review articles.

997
00:59:34,190 --> 00:59:36,470
And it does have
some key strengths.

998
00:59:42,120 --> 00:59:49,590
Computable General
Equilibrium has this strength

999
00:59:49,590 --> 00:59:53,520
of drawing on the national
income and product accounts

1000
00:59:53,520 --> 00:59:56,330
and input-output matrices.

1001
00:59:56,330 --> 00:59:59,520
That's basically the successor
to the Supply General

1002
00:59:59,520 --> 01:00:00,630
Equilibrium.

1003
01:00:00,630 --> 01:00:03,570
And then we have the more
macro-like, if I could say

1004
01:00:03,570 --> 01:00:08,460
so, Dynamic Stochastic
General Equilibrium models.

1005
01:00:08,460 --> 01:00:12,750
So you associate this with
Prescott, for example,

1006
01:00:12,750 --> 01:00:15,070
and real business cycles.

1007
01:00:15,070 --> 01:00:17,130
So it's great on dynamics.

1008
01:00:17,130 --> 01:00:20,470
It's great in being
very clear about shocks.

1009
01:00:20,470 --> 01:00:22,830
However, we anticipated
this, it largely

1010
01:00:22,830 --> 01:00:27,390
assumes a
representative consumer.

1011
01:00:27,390 --> 01:00:31,440
As if you had complete markets
that allow this kind of Gorman

1012
01:00:31,440 --> 01:00:34,540
aggregation with
certain preferences.

1013
01:00:34,540 --> 01:00:36,240
So in that world,
you're not going

1014
01:00:36,240 --> 01:00:38,400
to have redistributive
wealth effects.

1015
01:00:44,190 --> 01:00:46,320
And it is kind of
interesting in sort

1016
01:00:46,320 --> 01:00:48,630
of a history of
thought point of view,

1017
01:00:48,630 --> 01:00:51,570
that despite the early
battles in macro,

1018
01:00:51,570 --> 01:00:58,380
this DSGE stuff basically
took over, and largely

1019
01:00:58,380 --> 01:01:02,130
without modeling the
financial sector.

1020
01:01:02,130 --> 01:01:06,290
And so I think that's why people
felt rather flat-footed when

1021
01:01:06,290 --> 01:01:08,750
the financial crisis came along.

1022
01:01:12,120 --> 01:01:23,580
So there is stuff, sort of
dynamic general equilibrium

1023
01:01:23,580 --> 01:01:29,526
models that started with
Bernanke and Gertler.

1024
01:01:29,526 --> 01:01:31,680
I keep looking at Alp
over there because I

1025
01:01:31,680 --> 01:01:36,240
know that we feature that
just now in the macro lecture.

1026
01:01:39,860 --> 01:01:43,400
It is kind of built
on micro underpinnings

1027
01:01:43,400 --> 01:01:47,570
and a certain view of a
costly state verification

1028
01:01:47,570 --> 01:01:49,790
as an underlying imperfection.

1029
01:01:49,790 --> 01:01:52,820
It does feature a lot
the role of credit

1030
01:01:52,820 --> 01:01:54,470
and a financial accelerator.

1031
01:01:57,850 --> 01:02:00,130
But largely, with
some exceptions

1032
01:02:00,130 --> 01:02:07,520
I'll say momentarily, this
stuff is an aggregate of model.

1033
01:02:07,520 --> 01:02:10,610
And it's quite
elaborate as a model.

1034
01:02:10,610 --> 01:02:11,930
You have the Central Bank.

1035
01:02:11,930 --> 01:02:13,700
You have households,
entrepreneurs,

1036
01:02:13,700 --> 01:02:16,160
but entrepreneurs are somehow
different from retailers.

1037
01:02:16,160 --> 01:02:19,010
And retailers are somehow
different from capital goods

1038
01:02:19,010 --> 01:02:20,690
producers.

1039
01:02:20,690 --> 01:02:25,970
And certainly in early versions
of Christiano and Eichenbaum,

1040
01:02:25,970 --> 01:02:31,280
and so on, they actually
only look at the aggregates

1041
01:02:31,280 --> 01:02:32,540
generated by that model.

1042
01:02:32,540 --> 01:02:37,590
They never they never assign
these sectors to data.

1043
01:02:37,590 --> 01:02:38,550
It's not tested.

1044
01:02:38,550 --> 01:02:42,730
It's not part of the way
they're doing business.

1045
01:02:42,730 --> 01:02:45,300
So maybe they don't need to if
they get the aggregates right,

1046
01:02:45,300 --> 01:02:50,160
but this is a picture of India.

1047
01:02:50,160 --> 01:02:54,720
This stuff is used
by central banks.

1048
01:02:54,720 --> 01:02:57,090
If you want to
see one in Brazil,

1049
01:02:57,090 --> 01:02:58,625
it's called samba or something.

1050
01:02:58,625 --> 01:03:00,000
I'm not sure if
that's the music.

1051
01:03:00,500 --> 01:03:01,260
[LAUGHTER]

1052
01:03:02,100 --> 01:03:04,860
They have one.

1053
01:03:04,860 --> 01:03:06,900
It's on the reading list.

1054
01:03:06,900 --> 01:03:10,710
But I must say, in a bit
more optimistic vein,

1055
01:03:10,710 --> 01:03:14,520
things are starting
to come together.

1056
01:03:14,520 --> 01:03:19,320
Hsieh and Klenow wrote
papers on India and China

1057
01:03:19,320 --> 01:03:21,270
on the distribution
of firm size.

1058
01:03:21,270 --> 01:03:23,528
And now it's almost
becoming standard

1059
01:03:23,528 --> 01:03:25,320
that if you write down
one of these models,

1060
01:03:25,320 --> 01:03:28,560
you have to look at the
implication for firm size.

1061
01:03:28,560 --> 01:03:30,380
Or some key parameter
in the model

1062
01:03:30,380 --> 01:03:32,760
is going to be calibrated
against the distribution

1063
01:03:32,760 --> 01:03:36,720
of firm sizes in
the US, for example.

1064
01:03:36,720 --> 01:03:38,760
Just like Rajan and
Zingales kind of

1065
01:03:38,760 --> 01:03:40,850
started that research path.

1066
01:03:40,850 --> 01:03:47,790
So the latest Larry paper
definitely uses that,

1067
01:03:47,790 --> 01:03:50,310
and uses other
financial variables,

1068
01:03:50,310 --> 01:03:53,430
starting to look at
bankruptcy, and so on.

1069
01:03:53,430 --> 01:03:55,710
So one can be
somewhat optimistic

1070
01:03:55,710 --> 01:03:59,250
that these seemingly
similar, in the sense

1071
01:03:59,250 --> 01:04:02,070
of general equilibrium,
but obviously diverse

1072
01:04:02,070 --> 01:04:07,620
attacks on how the financial
system work, or isn't

1073
01:04:07,620 --> 01:04:09,070
even necessarily model.

1074
01:04:09,070 --> 01:04:11,190
They're kind of coming together.

1075
01:04:11,190 --> 01:04:15,390
And I think the place where
it arguably has come together

1076
01:04:15,390 --> 01:04:22,290
the fastest, and where we really
quite understand quite a lot is

1077
01:04:22,290 --> 01:04:25,590
in development economics.

1078
01:04:25,590 --> 01:04:32,220
And it features the measurement
at the household level

1079
01:04:32,220 --> 01:04:33,570
using corporate accounts.

1080
01:04:33,570 --> 01:04:37,630
This is again the midpoint
of the series of lectures.

1081
01:04:37,630 --> 01:04:41,070
So we can get at TFP,
productivity change, wealth,

1082
01:04:41,070 --> 01:04:43,980
rates of return,
treating households

1083
01:04:43,980 --> 01:04:46,740
just like the firms are
treated in corporate financial

1084
01:04:46,740 --> 01:04:50,220
accounts, income statement,
balance sheet, statement

1085
01:04:50,220 --> 01:04:51,090
of cash flow.

1086
01:04:54,780 --> 01:04:57,810
Occupation choice is a key
ingredient in these models,

1087
01:04:57,810 --> 01:05:00,660
because typically in
developing countries,

1088
01:05:00,660 --> 01:05:02,630
you see households
making transitions.

1089
01:05:02,630 --> 01:05:05,040
They used to be in agriculture.

1090
01:05:05,040 --> 01:05:11,010
They set up a firm, maybe
become wage earners.

1091
01:05:11,010 --> 01:05:15,960
So that's a big part of
almost all of the models.

1092
01:05:15,960 --> 01:05:18,510
There's tests of these
micro underpinnings.

1093
01:05:18,510 --> 01:05:21,240
I listed a few papers here.

1094
01:05:21,240 --> 01:05:26,470
There's some warning that
underpinnings do matter,

1095
01:05:26,470 --> 01:05:28,450
that things are rare.

1096
01:05:28,450 --> 01:05:30,900
It's rarer that one
kind of underpinning

1097
01:05:30,900 --> 01:05:35,220
is a stand-in for
the other ones,

1098
01:05:35,220 --> 01:05:38,910
or at least it does not
always work out that way,

1099
01:05:38,910 --> 01:05:41,760
and the Matsuyama paper
is a pretty good reminder.

1100
01:05:45,930 --> 01:05:48,660
Well, actually he more says
things aren't monotonic.

1101
01:05:48,660 --> 01:05:51,810
And Boyd and Smith
and Gertler and Rogoff

1102
01:05:51,810 --> 01:05:55,110
have examples of how
these underpinnings really

1103
01:05:55,110 --> 01:05:59,885
matter to the macro or
regional phenomenon.

1104
01:06:02,850 --> 01:06:08,050
So part of-- part of the
agenda here is measurement,

1105
01:06:08,050 --> 01:06:11,080
and it's not just measurement
at the household level.

1106
01:06:11,080 --> 01:06:17,380
It moves from the national
income accounts and balance

1107
01:06:17,380 --> 01:06:20,660
sheets, and so on, to
differences of balance sheets,

1108
01:06:20,660 --> 01:06:24,080
so it's essentially
the flow of funds.

1109
01:06:24,080 --> 01:06:27,640
So if there's one thing that
seems obvious, but actually

1110
01:06:27,640 --> 01:06:30,490
remains challenging,
is that if you

1111
01:06:30,490 --> 01:06:33,350
wanted to model the financial
system in a country,

1112
01:06:33,350 --> 01:06:35,440
you should look at the
flow of funds data.

1113
01:06:35,440 --> 01:06:38,230
Because that really tells
you, at least at the level

1114
01:06:38,230 --> 01:06:42,160
of aggregation that is
presented in the public data,

1115
01:06:42,160 --> 01:06:44,920
what financial instruments
are actually used,

1116
01:06:44,920 --> 01:06:51,280
who's taking the debtor-lender
position in those instruments.

1117
01:06:51,280 --> 01:06:55,210
You can certainly see
very well with flow

1118
01:06:55,210 --> 01:06:56,950
of funds, the
movement in and out

1119
01:06:56,950 --> 01:06:59,620
of the formal financial system.

1120
01:06:59,620 --> 01:07:02,530
Because that's where the
measurement is the best,

1121
01:07:02,530 --> 01:07:05,740
actually, because banks
have to report stuff.

1122
01:07:05,740 --> 01:07:07,690
They try desperately
to hide a lot,

1123
01:07:07,690 --> 01:07:13,420
but at least flow of funds in
and out of the banking system

1124
01:07:13,420 --> 01:07:19,240
is available even for
low income countries.

1125
01:07:19,240 --> 01:07:21,640
Turns out the IMF
is loaded with data.

1126
01:07:21,640 --> 01:07:25,460
People at the fund
haven't been using it.

1127
01:07:25,460 --> 01:07:27,460
But now it's sort
of an awakening

1128
01:07:27,460 --> 01:07:30,040
that there's a lot of
things that could be

1129
01:07:30,040 --> 01:07:32,110
done with flow of funds data.

1130
01:07:32,110 --> 01:07:33,650
And I'm in touch
with those guys,

1131
01:07:33,650 --> 01:07:38,770
so potentially if
you're interested

1132
01:07:38,770 --> 01:07:43,780
in a particular country, get you
the latest flow of funds data.

1133
01:07:43,780 --> 01:07:45,850
You have to keep
an eye on things,

1134
01:07:45,850 --> 01:07:48,370
though, in the sense
of the informal sector

1135
01:07:48,370 --> 01:07:52,150
is typically not measured,
but in these countries

1136
01:07:52,150 --> 01:07:57,410
can be really key, especially
in low income countries.

1137
01:07:57,410 --> 01:07:58,910
And unfortunately,
you don't usually

1138
01:07:58,910 --> 01:08:02,140
get the geography,
although we've

1139
01:08:02,140 --> 01:08:05,600
had some projects we've
encouraged with Mexico

1140
01:08:05,600 --> 01:08:08,140
and are kind of
working with Brazil

1141
01:08:08,140 --> 01:08:09,690
to get the geography in there.

1142
01:08:17,840 --> 01:08:20,720
We'll come back to this
flow of funds diagram,

1143
01:08:20,720 --> 01:08:23,300
but it basically makes the
point that you can talk about--

1144
01:08:32,970 --> 01:08:38,290
Here is actually a village
again with the transactions

1145
01:08:38,290 --> 01:08:44,920
of the village with a
not-for-profit corporation,

1146
01:08:44,920 --> 01:08:47,859
financial corporations,
non-financial corporations,

1147
01:08:47,859 --> 01:08:50,270
and government, and so on.

1148
01:08:50,270 --> 01:08:54,200
So just to belabor
the point again,

1149
01:08:54,200 --> 01:08:57,970
the way that you measure
in principle is common.

1150
01:08:57,970 --> 01:09:01,210
And then you can zoom
in and out depending

1151
01:09:01,210 --> 01:09:07,060
on what you think might be
key, and take a look at orders

1152
01:09:07,060 --> 01:09:08,380
of magnitudes of things.

1153
01:09:11,180 --> 01:09:13,399
You may not realize it,
but I'm basically following

1154
01:09:13,399 --> 01:09:14,600
the order of the syllabus.

1155
01:09:14,600 --> 01:09:16,590
So after we get through
the measurement part,

1156
01:09:16,590 --> 01:09:19,490
including flow of funds,
we'll talk specifically

1157
01:09:19,490 --> 01:09:26,300
about micro level tests
of insurance, of credit.

1158
01:09:26,300 --> 01:09:28,250
This is too
complicated for today,

1159
01:09:28,250 --> 01:09:31,490
but we'll get into these
obstacles to trade,

1160
01:09:31,490 --> 01:09:34,700
and how to test whether
it's a moral hazard problem

1161
01:09:34,700 --> 01:09:38,090
or adverse selection problem.

1162
01:09:38,090 --> 01:09:41,569
Test incomplete markets
against information

1163
01:09:41,569 --> 01:09:44,720
constrained markets.

1164
01:09:44,720 --> 01:09:48,189
This is a lot of material,
and I'll be mindful of that.

1165
01:09:48,189 --> 01:09:51,770
I'll at least try to give
you some examples of how

1166
01:09:51,770 --> 01:09:57,170
this analysis is
being conducted.

1167
01:09:57,170 --> 01:10:01,910
And we end up back with Applied
General Equilibrium Development

1168
01:10:01,910 --> 01:10:06,890
Economics, which is going to be
you know what we're going to do

1169
01:10:06,890 --> 01:10:09,980
in class starting on Thursday.

1170
01:10:09,980 --> 01:10:13,060
We're going to start with--

1171
01:10:13,060 --> 01:10:14,360
we've already got the reduced--

1172
01:10:18,500 --> 01:10:22,740
the patient choice models.

1173
01:10:22,740 --> 01:10:26,330
There are several Banerjee
Newman, [INAUDIBLE],,

1174
01:10:26,330 --> 01:10:30,210
Lloyd-Ellis and Dan
Bernhardt, Aghion and Bolton.

1175
01:10:43,980 --> 01:10:49,424
The key parameters of the
financial underpinning.

1176
01:11:06,720 --> 01:11:09,910
We'll for the second
lecture of the class,

1177
01:11:09,910 --> 01:11:13,540
we will be sort of having all
the ingredients, the micro,

1178
01:11:13,540 --> 01:11:16,690
the estimation,
and the simulation.

1179
01:11:16,690 --> 01:11:20,590
This is another sort of
genre in the literature.

1180
01:11:20,590 --> 01:11:23,080
This is financial--
this one takes

1181
01:11:23,080 --> 01:11:27,280
the expansion of the
financial system as exogenous.

1182
01:11:27,280 --> 01:11:30,700
This one tries to
make it endogenous.

1183
01:11:30,700 --> 01:11:33,550
Key contributors
are, say for example,

1184
01:11:33,550 --> 01:11:36,970
Greenwood and Jovanovic,
[INAUDIBLE] and Smith.

1185
01:11:36,970 --> 01:11:41,020
Entirely theoretical models
basically in the style

1186
01:11:41,020 --> 01:11:45,830
of Ross Levine, taking a stand
on how intermediation works,

1187
01:11:45,830 --> 01:11:48,440
showing, for example, that
inequality might go up

1188
01:11:48,440 --> 01:11:51,590
before it goes down,
but not taken seriously

1189
01:11:51,590 --> 01:11:56,510
enough in the sense of going to
the data, which is what we'll

1190
01:11:56,510 --> 01:11:58,190
talk about next time.

1191
01:12:01,010 --> 01:12:05,420
I said quite a bit during
the lecture about policy

1192
01:12:05,420 --> 01:12:07,580
and assessing policy.

1193
01:12:07,580 --> 01:12:10,940
So Thailand, for
example, as you saw,

1194
01:12:10,940 --> 01:12:15,320
had financial deepening
going on, and at times,

1195
01:12:15,320 --> 01:12:18,890
rather substantial
financial deepening.

1196
01:12:18,890 --> 01:12:22,670
Once we have these models
and we've kind of estimated

1197
01:12:22,670 --> 01:12:24,500
or calibrated the
key parameter, we

1198
01:12:24,500 --> 01:12:26,090
can do these
welfare experiments.

1199
01:12:26,090 --> 01:12:32,270
We can actually say, who
benefited, and for that matter,

1200
01:12:32,270 --> 01:12:37,660
who lost from the expansion
of the financial system?

1201
01:12:37,660 --> 01:12:41,800
The beneficiaries are like
these poor credit constrained

1202
01:12:41,800 --> 01:12:45,580
households that don't have
access to the financial system.

1203
01:12:45,580 --> 01:12:47,920
But over time, as
they exogenously

1204
01:12:47,920 --> 01:12:51,580
are allowed to
enter that system,

1205
01:12:51,580 --> 01:12:55,030
they have huge increases in
their welfare and well-being.

1206
01:12:57,610 --> 01:13:00,070
Why does somebody lose?

1207
01:13:00,070 --> 01:13:05,410
Well, that growth of enterprise
means increased employment.

1208
01:13:05,410 --> 01:13:08,360
It's pressure on wages.

1209
01:13:08,360 --> 01:13:11,050
In the model and in Thailand,
the real wage eventually

1210
01:13:11,050 --> 01:13:13,090
shoots up rather dramatically.

1211
01:13:13,090 --> 01:13:15,550
That increases the
cost of doing business

1212
01:13:15,550 --> 01:13:22,160
and the profits of
entrepreneurs can drop.

1213
01:13:22,160 --> 01:13:26,330
So this is the wonderful part.

1214
01:13:26,330 --> 01:13:29,220
Now, I don't mean to say
this model is wonderful.

1215
01:13:29,220 --> 01:13:31,010
It has a lot of
limitations, which I will

1216
01:13:31,010 --> 01:13:32,870
share with you on Thursday.

1217
01:13:32,870 --> 01:13:35,990
But what's wonderful about
it is that you can actually

1218
01:13:35,990 --> 01:13:40,490
go through these distributive
welfare gains and losses

1219
01:13:40,490 --> 01:13:43,250
types of calculations.

1220
01:13:43,250 --> 01:13:48,210
This one with endogenous
financial deepening

1221
01:13:48,210 --> 01:13:50,610
comes back to my comment
on the slide which

1222
01:13:50,610 --> 01:13:54,570
had India versus Thailand,
and I said something obscure

1223
01:13:54,570 --> 01:13:58,320
about I had another view of
what happened in Thailand.

1224
01:13:58,320 --> 01:14:04,110
Well, part of that growth
was preceded by a real flat.

1225
01:14:04,110 --> 01:14:07,560
GDP went basically flat-lined.

1226
01:14:07,560 --> 01:14:08,700
And what was going on?

1227
01:14:08,700 --> 01:14:14,820
Well, they had certain interest
rate restrictions on deposit

1228
01:14:14,820 --> 01:14:16,740
rates and lending rates.

1229
01:14:16,740 --> 01:14:19,250
There was an oil shock.

1230
01:14:19,250 --> 01:14:23,170
Prices were going up,
and money was flowing out

1231
01:14:23,170 --> 01:14:25,920
of the financial system.

1232
01:14:25,920 --> 01:14:28,080
The banks ran into trouble.

1233
01:14:28,080 --> 01:14:30,840
The government essentially
took them over,

1234
01:14:30,840 --> 01:14:34,450
and the economy goes flat-line.

1235
01:14:34,450 --> 01:14:36,770
You should worry about this.

1236
01:14:36,770 --> 01:14:40,720
It's not necessarily the
case that quote, regulation,

1237
01:14:40,720 --> 01:14:45,250
or in this case,
national-- almost close

1238
01:14:45,250 --> 01:14:47,200
to nationalization
of the banking system

1239
01:14:47,200 --> 01:14:52,550
is necessarily the way to deal
with the financial crisis.

1240
01:14:52,550 --> 01:14:54,430
Now how on earth are
we able to model this?

1241
01:14:54,430 --> 01:14:56,330
It's really quite simple.

1242
01:14:56,330 --> 01:14:59,680
We just create a wedge
in the original Greenwood

1243
01:14:59,680 --> 01:15:02,585
and Jovanovic model,
which is the spread

1244
01:15:02,585 --> 01:15:04,210
between the borrowing
and lending rate.

1245
01:15:04,210 --> 01:15:06,160
And we say when the more the--

1246
01:15:06,160 --> 01:15:09,340
sorry, this sounds
very market-oriented--

1247
01:15:09,340 --> 01:15:13,210
the more the government
does, the more resources

1248
01:15:13,210 --> 01:15:17,500
gets squandered, and
that inefficiency wedge

1249
01:15:17,500 --> 01:15:20,920
kind of increases over and above
the cost of intermediation.

1250
01:15:20,920 --> 01:15:22,570
And at the calibrated
parameters,

1251
01:15:22,570 --> 01:15:27,190
that's actually enough for the
model simulation to go flat.

1252
01:15:27,190 --> 01:15:29,860
But likewise, then we can
liberalize the system, which

1253
01:15:29,860 --> 01:15:32,140
Thailand did, and back out.

1254
01:15:46,410 --> 01:15:50,120
And then we'll go way
beyond these two the lecture

1255
01:15:50,120 --> 01:15:51,920
next Tuesday.