1
00:00:00,000 --> 00:00:01,470
[SQUEAKING]

2
00:00:01,470 --> 00:00:02,450
[CLICKING]

3
00:00:02,450 --> 00:00:04,410
[RUSTLING]

4
00:00:04,410 --> 00:00:09,545
[CLICKING]

5
00:00:09,545 --> 00:00:10,670
FRANK SCHILBACH: All right.

6
00:00:10,670 --> 00:00:11,980
I'm going to get started.

7
00:00:11,980 --> 00:00:14,100
This is lecture 19, of 14.13.

8
00:00:14,100 --> 00:00:17,600
I'm going to finish up
some of what we discussed

9
00:00:17,600 --> 00:00:21,230
last time on gender
discrimination and identity

10
00:00:21,230 --> 00:00:26,010
And then we're going to talk
about nudges and default

11
00:00:26,010 --> 00:00:27,720
effects.

12
00:00:27,720 --> 00:00:31,370
So let me sort of fairly
quickly go through the remaining

13
00:00:31,370 --> 00:00:36,300
slides of lecture 18, and then
move to the lecture slides.

14
00:00:36,300 --> 00:00:38,990
So remember, we talked
last time about the study

15
00:00:38,990 --> 00:00:42,230
by Bertrand et al. about
gender identity norms,

16
00:00:42,230 --> 00:00:44,750
where we talked about because--

17
00:00:44,750 --> 00:00:49,820
so the authors argue there's
this norm that some men

18
00:00:49,820 --> 00:00:52,550
perceive-- and perhaps
even some women--

19
00:00:52,550 --> 00:00:55,430
that men should be earning
more than their wives.

20
00:00:55,430 --> 00:00:58,670
There's this cliff-- this
missing mass of households

21
00:00:58,670 --> 00:01:00,950
or couples in the US.

22
00:01:00,950 --> 00:01:04,220
We talked about different
reasons for that.

23
00:01:04,220 --> 00:01:06,540
Missing couples that don't
form in the first place.

24
00:01:06,540 --> 00:01:10,310
Such marriages and couples
are less happy and stable,

25
00:01:10,310 --> 00:01:12,470
so they end up in divorce.

26
00:01:12,470 --> 00:01:15,470
And there's also,
even for existing

27
00:01:15,470 --> 00:01:18,050
couples that form, whether
the wife or the woman

28
00:01:18,050 --> 00:01:20,990
has more earnings potential.

29
00:01:20,990 --> 00:01:23,750
In some cases, at
least, women tend

30
00:01:23,750 --> 00:01:28,070
to either work less few hours,
and therefore perhaps that

31
00:01:28,070 --> 00:01:29,690
was promoted and the like.

32
00:01:29,690 --> 00:01:33,110
And in addition, they
might be involved in more

33
00:01:33,110 --> 00:01:34,830
nonmarket and childcare work.

34
00:01:34,830 --> 00:01:36,200
It's called the second shift.

35
00:01:36,200 --> 00:01:39,320
They have to do more work
because the husband is very

36
00:01:39,320 --> 00:01:41,600
unhappy about her earning more.

37
00:01:41,600 --> 00:01:45,770
Which then, of course,
holds potentially

38
00:01:45,770 --> 00:01:48,680
the wife back at her work.

39
00:01:48,680 --> 00:01:51,200
Now, somebody was
asking last time--

40
00:01:51,200 --> 00:01:55,250
I believe Justin or
others-- about the arrival

41
00:01:55,250 --> 00:01:58,715
of children and the gender gap.

42
00:01:58,715 --> 00:02:00,590
So there's quite a few
studies like this one.

43
00:02:00,590 --> 00:02:02,240
So out of this here,
which is the study

44
00:02:02,240 --> 00:02:04,250
by Kleven et al. in Denmark.

45
00:02:04,250 --> 00:02:06,890
And these studies
essentially show

46
00:02:06,890 --> 00:02:12,020
there is a substantial what they
call child penalty for women,

47
00:02:12,020 --> 00:02:15,380
which is when you look at
men's and women's earnings.

48
00:02:15,380 --> 00:02:18,250
On the left-hand side, there's
a figure shows earnings.

49
00:02:18,250 --> 00:02:20,450
This is yearly data over time.

50
00:02:20,450 --> 00:02:23,000
On the right-hand side,
you see hours worked.

51
00:02:23,000 --> 00:02:25,040
Similarly, you see productivity.

52
00:02:25,040 --> 00:02:27,380
But essentially, you tend
to see is that for men--

53
00:02:27,380 --> 00:02:30,730
this is the gray
bar here on top--

54
00:02:30,730 --> 00:02:33,120
you see essentially
before the child is born,

55
00:02:33,120 --> 00:02:37,980
you match men and women to
have sort of similar earnings,

56
00:02:37,980 --> 00:02:40,640
the trajectory would
look very similar.

57
00:02:40,640 --> 00:02:42,260
And then, for men,
perhaps there's

58
00:02:42,260 --> 00:02:46,160
a slight decline or no further
for the increase over time.

59
00:02:46,160 --> 00:02:52,130
But for women, there's a
clear drop in earnings.

60
00:02:52,130 --> 00:02:54,340
And that drop persists
for many, many years.

61
00:02:54,340 --> 00:02:56,450
So it's not just like
there's a year after birth,

62
00:02:56,450 --> 00:02:59,150
there's lower earnings
for, on average, women.

63
00:02:59,150 --> 00:03:01,550
But there's a
persistent gap overall

64
00:03:01,550 --> 00:03:03,980
that has to be presumably
with some women

65
00:03:03,980 --> 00:03:08,450
never returning back to work,
some women remaining part time,

66
00:03:08,450 --> 00:03:11,390
just some women working a few
hours when they have children.

67
00:03:11,390 --> 00:03:14,060
Notice that that's not
just about hours worked.

68
00:03:14,060 --> 00:03:15,980
So when you look at
the earnings gap,

69
00:03:15,980 --> 00:03:17,500
it's about something like 20%.

70
00:03:17,500 --> 00:03:21,090
The long-run child
penalty here is like 20%.

71
00:03:21,090 --> 00:03:22,530
It's also, for
hours worked, it's

72
00:03:22,530 --> 00:03:25,610
something like 10%, which
must mean that there's

73
00:03:25,610 --> 00:03:30,200
an additional 10% coming
from productivity or wages

74
00:03:30,200 --> 00:03:31,940
conditional on hours worked.

75
00:03:31,940 --> 00:03:35,180
So it's not only that women
are working fewer hours,

76
00:03:35,180 --> 00:03:38,000
but they're also paid less
per hour, which presumably

77
00:03:38,000 --> 00:03:43,790
is once people start working
few hours, take more breaks,

78
00:03:43,790 --> 00:03:46,070
or just take a break
for a few years,

79
00:03:46,070 --> 00:03:49,490
that they're getting less
promotions and so on.

80
00:03:49,490 --> 00:03:51,980
And that sort of carries
through for a long time.

81
00:03:51,980 --> 00:03:54,410
Now, interestingly, some of
these child penalties are

82
00:03:54,410 --> 00:03:56,180
transmitted through
generations--

83
00:03:56,180 --> 00:03:58,422
meaning from parents
to daughters--

84
00:03:58,422 --> 00:04:00,380
suggesting that essentially,
in the same family

85
00:04:00,380 --> 00:04:03,650
where there is child
penalties from the parents--

86
00:04:03,650 --> 00:04:07,520
as in the woman that
is a childcare penalty

87
00:04:07,520 --> 00:04:10,640
for the mother, the
daughter will also

88
00:04:10,640 --> 00:04:13,730
be likely to have a
child penalty later on.

89
00:04:13,730 --> 00:04:17,630
Like, 20, 30 years once
she's been actually working.

90
00:04:17,630 --> 00:04:20,660
So there seems to be
something to do with the child

91
00:04:20,660 --> 00:04:24,080
environment and gender
identity, which is the norm

92
00:04:24,080 --> 00:04:26,960
that the woman is supposed
to take care of the children,

93
00:04:26,960 --> 00:04:28,850
versus the husband is not.

94
00:04:28,850 --> 00:04:33,830
And that seems to be sort of
permitted through generations

95
00:04:33,830 --> 00:04:36,680
in the child environment.

96
00:04:36,680 --> 00:04:38,960
Now, you might sort of say,
well, then, couldn't we

97
00:04:38,960 --> 00:04:43,340
just do some policies that
support women or support

98
00:04:43,340 --> 00:04:46,910
parents to help them deal
with children better?

99
00:04:46,910 --> 00:04:49,520
And wouldn't that help?

100
00:04:49,520 --> 00:04:51,590
And there's some policies
that people have done.

101
00:04:51,590 --> 00:04:54,090
In particular, in academia,
you might say, well,

102
00:04:54,090 --> 00:05:00,020
if a woman has a child or a
couple as a child in general,

103
00:05:00,020 --> 00:05:03,620
that's surely a hint
to productivity.

104
00:05:03,620 --> 00:05:06,078
And so therefore, one could
say, well, that's [INAUDIBLE]..

105
00:05:06,078 --> 00:05:07,870
That's in fact what
people have been doing,

106
00:05:07,870 --> 00:05:10,640
is sort of, let's stop people's
tenure clock which essentially

107
00:05:10,640 --> 00:05:12,140
is an assistant professor.

108
00:05:12,140 --> 00:05:14,600
People have something
like 7, 8, 9 years

109
00:05:14,600 --> 00:05:16,610
to do this research
until tenure.

110
00:05:16,610 --> 00:05:18,980
You might say, let's
give people another year

111
00:05:18,980 --> 00:05:20,710
if they have a child.

112
00:05:20,710 --> 00:05:25,210
Now, one proposal is to have
gender neutral clock stopping

113
00:05:25,210 --> 00:05:26,290
policies.

114
00:05:26,290 --> 00:05:30,460
Now, these policies are intended
to involve these gender neutral

115
00:05:30,460 --> 00:05:37,090
policies, meaning that both
parents, regardless of who's

116
00:05:37,090 --> 00:05:39,935
giving-- or not just
the person who's

117
00:05:39,935 --> 00:05:41,810
giving birth, but also
the other person who's

118
00:05:41,810 --> 00:05:43,820
taking care of the
child potentially

119
00:05:43,820 --> 00:05:47,460
is getting an additional
year on the tenure clock.

120
00:05:47,460 --> 00:05:50,510
So the intention is here in
particular to involve men more

121
00:05:50,510 --> 00:05:52,340
in child care.

122
00:05:52,340 --> 00:05:54,470
But you know, those
kinds of policies

123
00:05:54,470 --> 00:05:57,260
tend to be not enforced,
and potentially

124
00:05:57,260 --> 00:05:59,960
they might even enhance
gender inequality

125
00:05:59,960 --> 00:06:03,020
if it's the case that both men
and women get another year.

126
00:06:03,020 --> 00:06:05,810
But the man essentially
takes that as a year

127
00:06:05,810 --> 00:06:09,710
off where you can do additional
work while the woman is back

128
00:06:09,710 --> 00:06:12,920
taking care of the
household and the child.

129
00:06:12,920 --> 00:06:16,670
Then, such policies in some
ways might make things worse.

130
00:06:16,670 --> 00:06:18,800
And this is exactly
what some studies find,

131
00:06:18,800 --> 00:06:22,490
is that these gender neutral
policies, in particular,

132
00:06:22,490 --> 00:06:23,570
in economics.

133
00:06:23,570 --> 00:06:26,310
There's also some examples
from other settings.

134
00:06:26,310 --> 00:06:30,050
But essentially, introducing
such policies substantially

135
00:06:30,050 --> 00:06:33,930
reduce female tenure rates while
increasing male tenure rates.

136
00:06:33,930 --> 00:06:37,760
So here's a policy that
looks actually a good policy.

137
00:06:37,760 --> 00:06:39,410
It's gender neutral and so on.

138
00:06:39,410 --> 00:06:42,110
But in fact, it might not
be, depending on who takes

139
00:06:42,110 --> 00:06:44,310
advantage of that
and who doesn't.

140
00:06:44,310 --> 00:06:47,720
So it's a very tricky question
because some of these things

141
00:06:47,720 --> 00:06:49,580
are not enforceable.

142
00:06:49,580 --> 00:06:52,920
If instead you gave only these
tenure exemptions to women,

143
00:06:52,920 --> 00:06:56,120
then you would get issues in
households, that then women

144
00:06:56,120 --> 00:06:58,640
are sort of pushed out
of work, which in part I

145
00:06:58,640 --> 00:07:00,920
think some of these
neutral tenure

146
00:07:00,920 --> 00:07:02,773
clocks try to avoid
in the first place.

147
00:07:02,773 --> 00:07:04,190
So it's not clear
what to do here.

148
00:07:04,190 --> 00:07:05,930
But I think it's
important to understand

149
00:07:05,930 --> 00:07:08,330
that some policies
that look like they

150
00:07:08,330 --> 00:07:12,740
are in fact neutral or gender
neutral, they might in fact not

151
00:07:12,740 --> 00:07:13,810
be.

152
00:07:13,810 --> 00:07:17,440
Now, let me tell you what I [got
about one more paper related

153
00:07:17,440 --> 00:07:19,850
to gender and the labor
market or identity,

154
00:07:19,850 --> 00:07:23,920
which is very nice paper
by Bursztyn et al., which

155
00:07:23,920 --> 00:07:28,120
asks the question whether women
avoid career enhancing actions

156
00:07:28,120 --> 00:07:31,250
because these actions
signal undesirable traits.

157
00:07:31,250 --> 00:07:34,570
So in particular,
women might be--

158
00:07:34,570 --> 00:07:36,550
ambitious women
might be perceived

159
00:07:36,550 --> 00:07:42,250
as not attractive partners
in the labor market,

160
00:07:42,250 --> 00:07:46,460
in particular, in some settings
such as among MBA students.

161
00:07:46,460 --> 00:07:50,450
And so, in particular, as
you might be familiar with,

162
00:07:50,450 --> 00:07:54,250
there's lots of dating happening
among MBA students, where

163
00:07:54,250 --> 00:07:58,550
people try to essentially find
partners or somebody to get

164
00:07:58,550 --> 00:07:59,570
married to.

165
00:07:59,570 --> 00:08:03,800
And it appears to be the case
that being ambitious or being

166
00:08:03,800 --> 00:08:06,590
perhaps overly
ambitious is perceived

167
00:08:06,590 --> 00:08:10,340
as negatively as an
undesirable trait

168
00:08:10,340 --> 00:08:12,980
on the dating market for MBAs.

169
00:08:12,980 --> 00:08:14,160
Now, what Bursztyn et al.

170
00:08:14,160 --> 00:08:19,311
find-- and this is joint work
with Amanda Pallais and Thomas

171
00:08:19,311 --> 00:08:22,130
Fujiwara they find that
unmarried and unmarried

172
00:08:22,130 --> 00:08:24,920
female MBA students
perform similarly

173
00:08:24,920 --> 00:08:27,440
when the performance is
unobserved by classmates.

174
00:08:27,440 --> 00:08:29,600
So these are things like
exams, problem sets.

175
00:08:29,600 --> 00:08:31,517
These are things that
only the professor sees,

176
00:08:31,517 --> 00:08:33,440
but not necessarily
their classmates.

177
00:08:33,440 --> 00:08:34,830
To here, male and--

178
00:08:34,830 --> 00:08:35,330
sorry.

179
00:08:35,330 --> 00:08:38,220
Unmarried and married female
MBA students do similarly

180
00:08:38,220 --> 00:08:40,330
well when we compare them.

181
00:08:40,330 --> 00:08:44,640
But unmarried women have
lower participation rates.

182
00:08:44,640 --> 00:08:46,360
So these are
essentially things that,

183
00:08:46,360 --> 00:08:48,310
do you say something in class?

184
00:08:48,310 --> 00:08:52,288
And as you might know, in MBA
classes, in some, at least,

185
00:08:52,288 --> 00:08:54,580
there's somebody sitting in
the back actually recording

186
00:08:54,580 --> 00:08:58,330
each interaction, each
participation very carefully

187
00:08:58,330 --> 00:09:03,550
so you know exactly who says
how many things and how often,

188
00:09:03,550 --> 00:09:05,060
and so on.

189
00:09:05,060 --> 00:09:07,120
So it seems to me
that unmarried women

190
00:09:07,120 --> 00:09:10,390
tend to not want to look too
ambitious and too hard-working

191
00:09:10,390 --> 00:09:11,740
in class.

192
00:09:11,740 --> 00:09:13,960
Now, that's of course
only suggestive.

193
00:09:13,960 --> 00:09:16,570
So then, they also
have a field experiment

194
00:09:16,570 --> 00:09:19,270
where they ask people about
their desired salaries

195
00:09:19,270 --> 00:09:23,200
and willingness to travel and
work long hours in real stakes

196
00:09:23,200 --> 00:09:24,310
placement questioning.

197
00:09:24,310 --> 00:09:26,875
So this essentially, at the
end, there's this placement,

198
00:09:26,875 --> 00:09:28,390
there's these career fairs.

199
00:09:28,390 --> 00:09:31,810
But people ask these
questions about,

200
00:09:31,810 --> 00:09:33,060
what salary are you expecting?

201
00:09:33,060 --> 00:09:34,602
Are you willing to
travel really far?

202
00:09:34,602 --> 00:09:36,280
Are you willing to
work long hours?

203
00:09:36,280 --> 00:09:37,150
And so on.

204
00:09:37,150 --> 00:09:41,630
And of course, these
questions all signal ambition.

205
00:09:41,630 --> 00:09:44,330
And then, the really interesting
part of the experiment

206
00:09:44,330 --> 00:09:47,000
is, then, they varied
whether this information is

207
00:09:47,000 --> 00:09:48,660
perceived to be private.

208
00:09:48,660 --> 00:09:50,333
So you just saying
it to a recruiter

209
00:09:50,333 --> 00:09:52,250
but nobody else in your
class, or anybody else

210
00:09:52,250 --> 00:09:55,970
might find out, versus whether
there some expectation of some

211
00:09:55,970 --> 00:09:57,740
of their classmates
actually learning

212
00:09:57,740 --> 00:10:00,200
about their preferences.

213
00:10:00,200 --> 00:10:04,190
And what they find, then, is
that the single female students

214
00:10:04,190 --> 00:10:07,310
report lower desired
salary, and willingness

215
00:10:07,310 --> 00:10:09,380
to travel, and work hours--

216
00:10:09,380 --> 00:10:11,510
and work long hours
when they think

217
00:10:11,510 --> 00:10:14,870
that there's an expectation that
their classmates might find out

218
00:10:14,870 --> 00:10:15,415
about it.

219
00:10:15,415 --> 00:10:16,790
Which essentially
is to say, they

220
00:10:16,790 --> 00:10:19,400
perceive it to be
costly for their class,

221
00:10:19,400 --> 00:10:22,550
or their classmates find
out about their ambition.

222
00:10:22,550 --> 00:10:25,460
Now, importantly, two
other things are also true.

223
00:10:25,460 --> 00:10:27,330
One is for other groups.

224
00:10:27,330 --> 00:10:31,460
So, other women that are
not single or for men.

225
00:10:31,460 --> 00:10:35,610
Essentially, these responses
are unaffected by peer

226
00:10:35,610 --> 00:10:37,140
observability.

227
00:10:37,140 --> 00:10:39,500
And second, it seems
to be that what's

228
00:10:39,500 --> 00:10:44,270
driving the observability
result is the observability

229
00:10:44,270 --> 00:10:46,560
by single male peers.

230
00:10:46,560 --> 00:10:50,600
So when other female students
or other non-single male peers

231
00:10:50,600 --> 00:10:55,700
are potentially seeing these
answers, there's no effect.

232
00:10:55,700 --> 00:10:59,570
But the effect's really coming
from other single male peers

233
00:10:59,570 --> 00:11:02,300
potentially seeing it, which
is precisely what you expect

234
00:11:02,300 --> 00:11:04,970
if you think having
ambition is potentially

235
00:11:04,970 --> 00:11:09,590
a negative signal on
the dating market.

236
00:11:09,590 --> 00:11:11,810
So that's another
sort of study showing

237
00:11:11,810 --> 00:11:15,320
that these norms
or perceived norms

238
00:11:15,320 --> 00:11:19,040
are really important, not just
in terms of punishing women

239
00:11:19,040 --> 00:11:24,410
and in some ways
undeservedly for mistakes

240
00:11:24,410 --> 00:11:26,150
as we saw in Heather
Sarson's paper,

241
00:11:26,150 --> 00:11:31,640
but also by holding
women back by themselves.

242
00:11:31,640 --> 00:11:35,720
These are single women who are
highly qualified MBA students

243
00:11:35,720 --> 00:11:39,950
who are deliberately not as
ambitious and hard-working

244
00:11:39,950 --> 00:11:42,150
as perhaps you might think
they would like to be--

245
00:11:42,150 --> 00:11:45,950
which is what the private
responses suggest--

246
00:11:45,950 --> 00:11:49,910
because there's a
penalty waiting for them,

247
00:11:49,910 --> 00:11:54,720
or receive penalty coming
on the dating market.

248
00:11:54,720 --> 00:11:55,890
Any questions about this?

249
00:12:01,420 --> 00:12:05,200
So here's a question about,
can it be the opposite?

250
00:12:05,200 --> 00:12:06,650
Better ask this in person.

251
00:12:06,650 --> 00:12:09,090
It's a little bit hard
for me to follow the chat.

252
00:12:09,090 --> 00:12:10,930
But can it be that
more confident women

253
00:12:10,930 --> 00:12:12,230
tend to be married?

254
00:12:12,230 --> 00:12:12,730
Yes.

255
00:12:12,730 --> 00:12:13,690
In principle, yes.

256
00:12:13,690 --> 00:12:15,940
But I think what we see
here in the experiment

257
00:12:15,940 --> 00:12:19,330
is, notice that
what I was saying,

258
00:12:19,330 --> 00:12:23,920
I was careful to say these
are about there's a perceived

259
00:12:23,920 --> 00:12:26,260
penalty of being
overly ambitious.

260
00:12:26,260 --> 00:12:29,360
There may well be for some
women, being very ambitious

261
00:12:29,360 --> 00:12:29,860
and so on.

262
00:12:29,860 --> 00:12:32,620
That's a very highly
desirable trait.

263
00:12:32,620 --> 00:12:34,780
But it seems to be that
the perception, at least,

264
00:12:34,780 --> 00:12:44,130
is that being highly ambitious
and so on is not desired.

265
00:12:53,000 --> 00:12:55,340
So I think there's
another question

266
00:12:55,340 --> 00:12:57,980
by Lucy about married
women compensating

267
00:12:57,980 --> 00:12:59,090
for discrimination.

268
00:12:59,090 --> 00:13:00,360
I think that's all.

269
00:13:00,360 --> 00:13:01,860
There might be other
stuff going on.

270
00:13:01,860 --> 00:13:04,520
But I think the key part
here is the variation

271
00:13:04,520 --> 00:13:10,230
between private and public
among single female students.

272
00:13:10,230 --> 00:13:12,290
So when they are
asking in private,

273
00:13:12,290 --> 00:13:15,020
they are saying essentially
they want higher salaries.

274
00:13:15,020 --> 00:13:17,870
When they're asking in public
for higher salaries and they're

275
00:13:17,870 --> 00:13:19,850
willing to work long
hours, and so on.

276
00:13:19,850 --> 00:13:21,650
When they're
answering in public,

277
00:13:21,650 --> 00:13:23,802
they're essentially
holding back.

278
00:13:23,802 --> 00:13:25,760
Which seems to suggest--
and that's consistent,

279
00:13:25,760 --> 00:13:28,750
I guess, also with the
participation rates.

280
00:13:28,750 --> 00:13:32,660
And that's consistent with why
there is at least a perception

281
00:13:32,660 --> 00:13:38,490
of lower desirability, in
particular among their single

282
00:13:38,490 --> 00:13:42,720
male peers, which is what
this results is driven by.

283
00:13:42,720 --> 00:13:44,730
It's not clear, by the
way, that the study shows

284
00:13:44,730 --> 00:13:46,380
that that's actually true.

285
00:13:46,380 --> 00:13:49,620
That's about these
women's perceptions.

286
00:13:49,620 --> 00:13:51,120
The study also
doesn't necessarily

287
00:13:51,120 --> 00:13:53,760
show that this translates at
the end of the day in lower

288
00:13:53,760 --> 00:13:55,470
earnings, and so
on, and so forth.

289
00:13:55,470 --> 00:13:58,020
The authors argue
that that's the case.

290
00:13:58,020 --> 00:14:01,530
But these are only
sort of questions--

291
00:14:01,530 --> 00:14:04,423
real stakes questions and
placement questionnaires.

292
00:14:04,423 --> 00:14:06,840
It's just sort of essentially
more like a proof of concept

293
00:14:06,840 --> 00:14:09,210
arguing that this is
an important force that

294
00:14:09,210 --> 00:14:11,560
might be at play, that
women hold themselves back

295
00:14:11,560 --> 00:14:16,050
in some ways here and in
the case of a placement

296
00:14:16,050 --> 00:14:17,030
questionnaire.

297
00:14:26,310 --> 00:14:27,990
I had started this
already last time,

298
00:14:27,990 --> 00:14:31,680
only briefly going through
the Vesterlund study, which

299
00:14:31,680 --> 00:14:36,390
essentially is saying that
there's two forces going on.

300
00:14:36,390 --> 00:14:41,880
There's demand [INAUDIBLE]
in terms of tasks

301
00:14:41,880 --> 00:14:44,700
that women and men are
perhaps asked differently.

302
00:14:44,700 --> 00:14:48,210
And beliefs could be quite
important here, and mediating

303
00:14:48,210 --> 00:14:50,150
who has which tasks and when.

304
00:14:50,150 --> 00:14:52,650
So let me show you a little bit
of what that study is about.

305
00:14:52,650 --> 00:14:55,440
The study is essentially
asking the question

306
00:14:55,440 --> 00:14:59,280
whether women are not saying
no often enough In particular

307
00:14:59,280 --> 00:15:01,410
when it comes to work.

308
00:15:01,410 --> 00:15:03,900
And it's motivated by the
fact those female faculty

309
00:15:03,900 --> 00:15:06,240
members tend to
do a lot more what

310
00:15:06,240 --> 00:15:08,850
they call non-promotable tasks.

311
00:15:08,850 --> 00:15:11,790
These are tasks that essentially
everybody wants somebody

312
00:15:11,790 --> 00:15:13,320
to do the task,
but they don't want

313
00:15:13,320 --> 00:15:14,870
to do the tasks themselves.

314
00:15:14,870 --> 00:15:17,070
So that's like essentially
University committees,

315
00:15:17,070 --> 00:15:21,690
and here, also
undergraduate teaching,

316
00:15:21,690 --> 00:15:23,670
or undergraduate
student advising.

317
00:15:23,670 --> 00:15:25,320
Of course, everybody
loves undergrads

318
00:15:25,320 --> 00:15:28,180
and wants to advise them.

319
00:15:28,180 --> 00:15:30,690
However, it's not
something that's

320
00:15:30,690 --> 00:15:32,970
good for peoples' career.

321
00:15:32,970 --> 00:15:34,800
So at the end of the
day, faculty are often

322
00:15:34,800 --> 00:15:37,060
evaluated by research output.

323
00:15:37,060 --> 00:15:41,450
So any time that's spent
on committees, et cetera,

324
00:15:41,450 --> 00:15:43,830
these are not going to
be helping for promotions

325
00:15:43,830 --> 00:15:44,370
eventually.

326
00:15:44,370 --> 00:15:46,740
And then, one first
observation is that [? women ?]

327
00:15:46,740 --> 00:15:49,890
do that much more than men.

328
00:15:49,890 --> 00:15:52,230
Now you might sort of
ask, why is it that women

329
00:15:52,230 --> 00:15:53,790
spend their time differently?

330
00:15:53,790 --> 00:15:55,660
There's potentially
issues of demand.

331
00:15:55,660 --> 00:15:58,140
Women are just asked a lot
more to do these things.

332
00:15:58,140 --> 00:16:00,060
And there's potentially
supply issues

333
00:16:00,060 --> 00:16:01,290
conditional on being asked.

334
00:16:01,290 --> 00:16:04,590
Women might be just much
more likely to say yes.

335
00:16:04,590 --> 00:16:07,650
Now, what this paper argues
is that demand and supply

336
00:16:07,650 --> 00:16:08,550
interact.

337
00:16:08,550 --> 00:16:11,520
Because women are more
likely to say yes,

338
00:16:11,520 --> 00:16:16,080
they're also more likely to
be asked in the first place.

339
00:16:16,080 --> 00:16:16,710
OK.

340
00:16:16,710 --> 00:16:17,765
So why do we care?

341
00:16:17,765 --> 00:16:19,140
I think I already
mentioned this.

342
00:16:19,140 --> 00:16:20,900
There's sort of a few
broad perspectives.

343
00:16:20,900 --> 00:16:22,650
One is about, we kind
of try to understand

344
00:16:22,650 --> 00:16:24,552
individual
decision-making. and we

345
00:16:24,552 --> 00:16:26,760
want to understand how people
make optimal decisions.

346
00:16:26,760 --> 00:16:28,620
Here's another reason
where people report.

347
00:16:28,620 --> 00:16:31,950
Here is an example where people
potentially mis-optimize.

348
00:16:31,950 --> 00:16:34,530
There's a managerial
social planner perspective

349
00:16:34,530 --> 00:16:36,900
that if you want
to sort of organize

350
00:16:36,900 --> 00:16:41,070
or you want organizations
to be efficient,

351
00:16:41,070 --> 00:16:42,570
you want to make
sure that everybody

352
00:16:42,570 --> 00:16:44,970
does equal work in some way.

353
00:16:44,970 --> 00:16:46,900
And so you don't
want women who are

354
00:16:46,900 --> 00:16:51,570
very productive to be stuck
in a bunch of other tasks

355
00:16:51,570 --> 00:16:53,710
that hold them back.

356
00:16:53,710 --> 00:16:56,500
And then, there's a
public policy perspective.

357
00:16:56,500 --> 00:17:00,130
It might essentially
be that sex differences

358
00:17:00,130 --> 00:17:01,617
and the allocation of time.

359
00:17:01,617 --> 00:17:03,700
So women spending more
time on these kind of tasks

360
00:17:03,700 --> 00:17:07,960
might explain why they're less
likely to be promoted and so

361
00:17:07,960 --> 00:17:08,530
on.

362
00:17:08,530 --> 00:17:11,200
And then, understanding that
and understanding these issues

363
00:17:11,200 --> 00:17:13,240
might sort of help
with interventions

364
00:17:13,240 --> 00:17:15,480
to improve equity.

365
00:17:15,480 --> 00:17:16,069
Now, OK.

366
00:17:16,069 --> 00:17:17,069
So what are these tasks?

367
00:17:17,069 --> 00:17:18,361
Again there's promotable tasks.

368
00:17:18,361 --> 00:17:19,349
That's doing research.

369
00:17:19,349 --> 00:17:21,252
And there's
non-promotable tasks.

370
00:17:21,252 --> 00:17:22,960
So promotable tasks,
meaning essentially,

371
00:17:22,960 --> 00:17:25,619
if you do these tasks,
you're likely to be promoted

372
00:17:25,619 --> 00:17:29,790
or essentially get
tenure eventually.

373
00:17:29,790 --> 00:17:31,500
A non-promotable
task essentially

374
00:17:31,500 --> 00:17:33,760
are tasks that many
people could do the task,

375
00:17:33,760 --> 00:17:35,610
but everybody wants
somebody to do it.

376
00:17:35,610 --> 00:17:39,840
But often, nobody actually
wants to do it themselves.

377
00:17:39,840 --> 00:17:42,270
Now, the authors that first
[INAUDIBLE] field study

378
00:17:42,270 --> 00:17:45,360
where they have faculty at
a large public University

379
00:17:45,360 --> 00:17:47,490
where you send essentially
emails from a chair,

380
00:17:47,490 --> 00:17:50,190
and they're requested
to volunteer

381
00:17:50,190 --> 00:17:53,250
to join one of several
university-wide faculty senate

382
00:17:53,250 --> 00:17:54,330
committees.

383
00:17:54,330 --> 00:17:58,740
And women are essentially,
in these tasks,

384
00:17:58,740 --> 00:18:01,818
are way more likely to volunteer
conditional on being asked.

385
00:18:01,818 --> 00:18:03,360
So that's the first
piece of evidence

386
00:18:03,360 --> 00:18:05,430
that conditional on
being asked women

387
00:18:05,430 --> 00:18:07,530
are more likely to be asked.

388
00:18:07,530 --> 00:18:09,780
Then you might say, well,
perhaps there's some

389
00:18:09,780 --> 00:18:11,460
personality characteristic
[? it's ?] [? on. ?] Maybe

390
00:18:11,460 --> 00:18:12,127
women are nicer.

391
00:18:12,127 --> 00:18:15,970
Maybe women care more
about the public good.

392
00:18:15,970 --> 00:18:19,110
Now, what the authors
show is that that may well

393
00:18:19,110 --> 00:18:20,250
be true in some cases.

394
00:18:20,250 --> 00:18:23,040
But in other cases, it's also
about expectation about who

395
00:18:23,040 --> 00:18:25,140
does it and who doesn't.

396
00:18:25,140 --> 00:18:28,170
Moreover, the ask
about who is asked

397
00:18:28,170 --> 00:18:30,930
is quite different depending on
whether you're male or female.

398
00:18:30,930 --> 00:18:33,130
And that's where the
experiment comes in.

399
00:18:33,130 --> 00:18:34,983
So what they do in
an experiment is

400
00:18:34,983 --> 00:18:36,900
what's called the threshold
public goods game.

401
00:18:36,900 --> 00:18:40,080
It's very simple, which
is, a small group needs

402
00:18:40,080 --> 00:18:42,272
to find a volunteer for a task.

403
00:18:42,272 --> 00:18:43,980
These are groups of
3 people, and there's

404
00:18:43,980 --> 00:18:45,605
like 10 rounds where
people are matched

405
00:18:45,605 --> 00:18:48,820
over time with each other.

406
00:18:48,820 --> 00:18:53,070
And so it's essentially set up
to mimic real world situations

407
00:18:53,070 --> 00:18:56,130
where everyone prefers the task
to be undertaken by someone,

408
00:18:56,130 --> 00:18:57,370
but not themselves.

409
00:18:57,370 --> 00:18:59,100
As you don't want
to do it, you have

410
00:18:59,100 --> 00:19:00,450
to have somebody else to do it.

411
00:19:00,450 --> 00:19:03,180
You want that somebody
does it, but you just

412
00:19:03,180 --> 00:19:05,050
don't want to do it yourself.

413
00:19:05,050 --> 00:19:07,750
And so people get--

414
00:19:07,750 --> 00:19:11,020
the group of 3 gets two minutes
to decide whether to invest.

415
00:19:11,020 --> 00:19:12,486
And investing means
essentially you

416
00:19:12,486 --> 00:19:14,153
have to [? truly ?]
press a button here.

417
00:19:14,153 --> 00:19:16,000
Only one person can
invest, and the round

418
00:19:16,000 --> 00:19:17,950
ends when someone invests.

419
00:19:17,950 --> 00:19:21,130
If no one invests, all
the members get $1.

420
00:19:21,130 --> 00:19:26,140
If you are investing you get
$1.25 and the remaining members

421
00:19:26,140 --> 00:19:27,017
get $2.

422
00:19:27,017 --> 00:19:28,600
So if you're thinking
about this task,

423
00:19:28,600 --> 00:19:32,000
what are you thinking
about is now, A, is

424
00:19:32,000 --> 00:19:33,470
somebody else going to invest?

425
00:19:33,470 --> 00:19:37,980
And B, like conditional
on your beliefs about,

426
00:19:37,980 --> 00:19:40,190
is somebody else going
to invest any time soon?

427
00:19:40,190 --> 00:19:42,890
Do you want to invest or not?

428
00:19:42,890 --> 00:19:45,920
Clearly, it's the efficient
thing that somebody invests.

429
00:19:45,920 --> 00:19:48,740
But you don't want to be
yourself because then you get

430
00:19:48,740 --> 00:19:51,697
only $1.25 rather than the $2.

431
00:19:51,697 --> 00:19:53,780
And the clock ticks down
until one person invests,

432
00:19:53,780 --> 00:19:57,330
or no investment
is made 2 minutes.

433
00:19:57,330 --> 00:19:57,830
OK.

434
00:19:57,830 --> 00:19:59,480
So now what do we find?

435
00:19:59,480 --> 00:20:01,880
Or what did they find
in this experiment?

436
00:20:01,880 --> 00:20:05,420
Women are way more likely to
invest in each of these rounds.

437
00:20:05,420 --> 00:20:07,400
This is sort of
the red line here.

438
00:20:07,400 --> 00:20:11,030
The blue line here is for men.

439
00:20:11,030 --> 00:20:16,350
And so women are much more
likely to invest than men.

440
00:20:16,350 --> 00:20:19,130
And that's very clear when
you look at these figures.

441
00:20:21,800 --> 00:20:23,638
And that's also very
persistent over time.

442
00:20:23,638 --> 00:20:26,180
So it's not just like, perhaps
in the first round, and so on.

443
00:20:26,180 --> 00:20:27,500
This is even after 10 rounds.

444
00:20:27,500 --> 00:20:30,480
When people keep being
matched with others over time,

445
00:20:30,480 --> 00:20:33,510
women are much more
likely to invest.

446
00:20:33,510 --> 00:20:34,760
Now, how do we interpret this?

447
00:20:34,760 --> 00:20:36,010
So how do we think about this?

448
00:20:36,010 --> 00:20:37,243
What can we learn about this?

449
00:20:37,243 --> 00:20:38,910
One explanation that
I already mentioned

450
00:20:38,910 --> 00:20:40,890
is, women are much nicer.

451
00:20:40,890 --> 00:20:44,130
But can you infer
this from this task?

452
00:20:44,130 --> 00:20:46,180
You observe if women are
a lot nicer than men,

453
00:20:46,180 --> 00:20:48,638
well, then you should find that
women are more cooperative,

454
00:20:48,638 --> 00:20:50,790
as Lucy suggests.

455
00:20:50,790 --> 00:20:53,280
If women are a lot
more cooperative,

456
00:20:53,280 --> 00:20:56,250
you should see that
in women-only groups,

457
00:20:56,250 --> 00:20:58,590
women are much more
likely to invest

458
00:20:58,590 --> 00:21:00,720
than in male-only groups.

459
00:21:00,720 --> 00:21:04,210
And that's exactly what they
tried to do in the experiment.

460
00:21:04,210 --> 00:21:07,320
So they have essentially
single sex sessions

461
00:21:07,320 --> 00:21:10,870
would have all-female
and all-male groups.

462
00:21:10,870 --> 00:21:13,830
And what you find is, while
the probability of investing

463
00:21:13,830 --> 00:21:16,810
is actually exactly the
same for men and women,

464
00:21:16,810 --> 00:21:18,840
they also find that
when trying to predict

465
00:21:18,840 --> 00:21:25,043
who's investing who is not, any
measures of social preferences,

466
00:21:25,043 --> 00:21:27,210
risk preferences and so on
that they've [? listed ?]

467
00:21:27,210 --> 00:21:30,930
previously actually not
predictive of this behavior.

468
00:21:30,930 --> 00:21:33,600
That sort of suggests it's
really not about altruism

469
00:21:33,600 --> 00:21:37,320
in a sense of being
nicer or wanting.

470
00:21:37,320 --> 00:21:39,840
So in some sense, investing
is doing something nice

471
00:21:39,840 --> 00:21:41,670
for the others in your group.

472
00:21:41,670 --> 00:21:44,830
It doesn't seem to be
that it's about that.

473
00:21:44,830 --> 00:21:47,830
It's also not so much
about risk aversion.

474
00:21:47,830 --> 00:21:51,000
So you might say, well, I'm
investing because I really

475
00:21:51,000 --> 00:21:51,630
like certainty.

476
00:21:51,630 --> 00:21:53,820
If I invest, I know
exactly what I'm getting.

477
00:21:53,820 --> 00:21:55,860
If I'm not investing,
I have a chance

478
00:21:55,860 --> 00:21:58,200
of like getting the
$2 or essentially

479
00:21:58,200 --> 00:22:00,930
a lottery between $2 and
$1, depending on, then,

480
00:22:00,930 --> 00:22:03,150
the probability is
essentially about what

481
00:22:03,150 --> 00:22:05,310
you think is the probability
of others investing.

482
00:22:05,310 --> 00:22:07,390
So your risk aversion
might matter,

483
00:22:07,390 --> 00:22:09,720
but they also show that
risk aversion is really not

484
00:22:09,720 --> 00:22:10,900
important.

485
00:22:10,900 --> 00:22:13,480
So now they have
essentially two experiments.

486
00:22:13,480 --> 00:22:16,230
One is the experiment,
1 that I showed you.

487
00:22:16,230 --> 00:22:19,410
Women are more likely to
invest in mixed sex groups,

488
00:22:19,410 --> 00:22:23,700
and women and men are equally
likely to invest in single sex

489
00:22:23,700 --> 00:22:24,360
groups.

490
00:22:24,360 --> 00:22:25,860
And so now, then,
what they're doing

491
00:22:25,860 --> 00:22:27,443
is, then, they're
trying to say, well,

492
00:22:27,443 --> 00:22:29,700
could it be that beliefs
that women will invest

493
00:22:29,700 --> 00:22:31,228
are important here?

494
00:22:31,228 --> 00:22:32,520
So that's essentially the idea.

495
00:22:32,520 --> 00:22:35,370
If you think women are
more likely to invest

496
00:22:35,370 --> 00:22:38,670
than men, if women think
this and men think this,

497
00:22:38,670 --> 00:22:41,580
then essentially you're
going to get in a situation

498
00:22:41,580 --> 00:22:46,020
where if you have
a mixed sex group,

499
00:22:46,020 --> 00:22:48,360
then you will have one
woman saying to men,

500
00:22:48,360 --> 00:22:51,000
the woman thinks, oh, the men
are not going to invest anyway.

501
00:22:51,000 --> 00:22:53,850
So either it's like we're
investing or getting $1.

502
00:22:53,850 --> 00:22:56,308
So then, for the women,
it's optimal to invest.

503
00:22:56,308 --> 00:22:57,850
And the men are
going to think, well,

504
00:22:57,850 --> 00:22:59,058
the woman is going to invest.

505
00:22:59,058 --> 00:23:01,560
So why should they invest if
the woman is investing anyway?

506
00:23:01,560 --> 00:23:06,360
For them, it's then
optimal not to invest.

507
00:23:06,360 --> 00:23:10,240
Now, experiment 3,
then, get set this

508
00:23:10,240 --> 00:23:15,220
in some ways, which essentially
is about now getting

509
00:23:15,220 --> 00:23:18,005
women and men to ask to invest.

510
00:23:18,005 --> 00:23:20,380
What they do is, essentially,
they have what's called a--

511
00:23:20,380 --> 00:23:24,850
they call it a photo ask, which
they have 4 people per group.

512
00:23:24,850 --> 00:23:26,477
3 people as 4 can invest.

513
00:23:26,477 --> 00:23:28,060
That's essentially
the same as before.

514
00:23:28,060 --> 00:23:29,240
They call them green players.

515
00:23:29,240 --> 00:23:30,565
Not sure why, but they
essentially [? do ?]

516
00:23:30,565 --> 00:23:31,750
[? their ?] green players.

517
00:23:31,750 --> 00:23:34,720
And one person is
unable to invest,

518
00:23:34,720 --> 00:23:39,260
but he can ask one of the
three others to invest.

519
00:23:39,260 --> 00:23:40,570
It's not a binding request.

520
00:23:40,570 --> 00:23:42,400
Essentially, it's
sort of a nudge.

521
00:23:42,400 --> 00:23:44,317
So if you want to, instead
of essentially just

522
00:23:44,317 --> 00:23:46,750
suggest that they may
exert some social pressure,

523
00:23:46,750 --> 00:23:47,710
that's the red player.

524
00:23:47,710 --> 00:23:49,918
And then, the red player is
essentially incentivized.

525
00:23:49,918 --> 00:23:52,833
The red player wants the
investment to happen.

526
00:23:52,833 --> 00:23:54,250
So this is kind
of what this looks

527
00:23:54,250 --> 00:23:55,910
like, is essentially
three people

528
00:23:55,910 --> 00:23:57,220
that you see as the red player.

529
00:23:57,220 --> 00:23:58,720
And you can say,
whom would you ask?

530
00:23:58,720 --> 00:23:59,920
Well, any of these three.

531
00:23:59,920 --> 00:24:01,420
You see the photos,
so obviously you

532
00:24:01,420 --> 00:24:05,450
see the gender of the person.

533
00:24:05,450 --> 00:24:09,110
And so, what they then find
is that women are much more

534
00:24:09,110 --> 00:24:12,110
likely to be asked.

535
00:24:12,110 --> 00:24:15,920
You can see here, the red
lines are more to the right,

536
00:24:15,920 --> 00:24:18,500
or there's more red
to the right, which

537
00:24:18,500 --> 00:24:20,750
is the total times
asked to invest.

538
00:24:20,750 --> 00:24:24,590
So there's more asked to the
right for women than for men.

539
00:24:24,590 --> 00:24:26,510
Now, what did we
learn from that?

540
00:24:26,510 --> 00:24:28,520
Well, absent a request,
investment rates

541
00:24:28,520 --> 00:24:30,350
do not differ by gender.

542
00:24:30,350 --> 00:24:33,630
But when asked to invest,
women are more likely to do so.

543
00:24:33,630 --> 00:24:35,130
So there's two
things going on here.

544
00:24:35,130 --> 00:24:38,450
One is, when asked, women
are more responsive.

545
00:24:38,450 --> 00:24:41,210
And [INAUDIBLE] says that
the marginal increase

546
00:24:41,210 --> 00:24:44,450
of being asked is higher
for women than for men.

547
00:24:44,450 --> 00:24:48,830
And then, in addition, women are
also more likely to be asked.

548
00:24:48,830 --> 00:24:50,420
Which is to say,
when people have

549
00:24:50,420 --> 00:24:52,460
the choice about
whom should you ask,

550
00:24:52,460 --> 00:24:54,440
well, they kind of know
who is the person who

551
00:24:54,440 --> 00:25:03,530
is more likely to react, which
is the woman in the group.

552
00:25:03,530 --> 00:25:06,080
So they correctly
understand that if they

553
00:25:06,080 --> 00:25:08,660
ask the woman, that
it's going to have more

554
00:25:08,660 --> 00:25:10,520
of an impact than asking men.

555
00:25:10,520 --> 00:25:11,660
And that's true.

556
00:25:11,660 --> 00:25:16,170
So when women are asked, they're
also more likely to comply.

557
00:25:16,170 --> 00:25:19,820
So now, that essentially sort
of says these gender differences

558
00:25:19,820 --> 00:25:24,770
get amplified by increases
for women to contribute.

559
00:25:24,770 --> 00:25:29,000
So not only is it the case that
they're more likely to say yes,

560
00:25:29,000 --> 00:25:32,690
but on top of that, they
also will get more requests.

561
00:25:32,690 --> 00:25:37,017
Which of course, then, is like,
they do so much work overall.

562
00:25:37,017 --> 00:25:38,600
And so, what's key
here is about this.

563
00:25:38,600 --> 00:25:40,250
it's not about preferences.

564
00:25:40,250 --> 00:25:41,930
It's essentially
just an equilibrium

565
00:25:41,930 --> 00:25:45,920
that at the end of the
day will be sustained.

566
00:25:45,920 --> 00:25:48,110
Because again, if men think
women are going to do it

567
00:25:48,110 --> 00:25:50,180
anyway, and if
women think, well,

568
00:25:50,180 --> 00:25:51,862
nobody's going to do
it unless I myself,

569
00:25:51,862 --> 00:25:54,320
[? when ?] [? I ?] [? am ?] in
situations where it might be

570
00:25:54,320 --> 00:25:56,960
optimal for both women
and men to do it,

571
00:25:56,960 --> 00:25:59,690
conditional on
their [INAUDIBLE]..

572
00:25:59,690 --> 00:26:00,920
Any questions on this study?

573
00:26:14,595 --> 00:26:18,900
OK, so then, let me
just summarize a bit.

574
00:26:18,900 --> 00:26:22,950
So we saw large gender wage
and earnings gaps that even

575
00:26:22,950 --> 00:26:25,140
persisted even until now.

576
00:26:25,140 --> 00:26:26,700
Substantial progress
has been made

577
00:26:26,700 --> 00:26:28,980
due to the
technological advances

578
00:26:28,980 --> 00:26:31,620
and other improvements, but
sort of those persistent gender

579
00:26:31,620 --> 00:26:36,210
differences in many
societies, including the US.

580
00:26:36,210 --> 00:26:38,490
Biased beliefs and
identity concerns

581
00:26:38,490 --> 00:26:43,080
appear to play a particularly
important role, in particular,

582
00:26:43,080 --> 00:26:46,050
in some issues such as
educational differences

583
00:26:46,050 --> 00:26:49,080
and so on.

584
00:26:49,080 --> 00:26:51,750
And given the
technological advances,

585
00:26:51,750 --> 00:26:54,662
some of these technological
issues are less important now.

586
00:26:54,662 --> 00:26:57,120
But really, it seems to be that
biased beliefs and identity

587
00:26:57,120 --> 00:27:01,110
concerns are at the center of
explaining things in addition

588
00:27:01,110 --> 00:27:02,457
to the child penalty.

589
00:27:02,457 --> 00:27:04,290
But even the child
penalty, you might argue,

590
00:27:04,290 --> 00:27:08,037
is driven by beliefs,
and identity, and so on.

591
00:27:08,037 --> 00:27:09,870
And then there are these
feedback mechanisms

592
00:27:09,870 --> 00:27:12,930
between demand and
supplies of tasks

593
00:27:12,930 --> 00:27:14,730
that women and men are doing.

594
00:27:14,730 --> 00:27:16,980
And in particular, it seems
to be that, again, beliefs

595
00:27:16,980 --> 00:27:19,147
are important in the sense
of, like, it matters what

596
00:27:19,147 --> 00:27:20,550
people think other people do.

597
00:27:24,760 --> 00:27:28,570
And then that seems to create,
essentially, these inequalities

598
00:27:28,570 --> 00:27:29,878
across gender.

599
00:27:29,878 --> 00:27:31,420
Now, if you want to
sort of mitigate,

600
00:27:31,420 --> 00:27:33,400
or reduce, or eliminate
the gender gap,

601
00:27:33,400 --> 00:27:35,560
it's really important to
understand these issues.

602
00:27:35,560 --> 00:27:38,110
I think we have made
quite a bit of progress

603
00:27:38,110 --> 00:27:42,310
in understanding what's going
on, and identifying the issue.

604
00:27:42,310 --> 00:27:45,490
Now, I think the next steps
are to try and understand

605
00:27:45,490 --> 00:27:49,330
what are policies that are
in fact successful in helping

606
00:27:49,330 --> 00:27:51,400
us do that.

607
00:27:51,400 --> 00:27:55,120
So this lecture is about
defaults, nudges, and frames.

608
00:27:55,120 --> 00:27:57,640
And originally, I had planned
to also do mental accounting.

609
00:27:57,640 --> 00:28:00,220
But given that we
didn't finish last time,

610
00:28:00,220 --> 00:28:03,750
I moved this to a later
lecture, at least [INAUDIBLE]..

611
00:28:03,750 --> 00:28:05,507
OK, so where are
we in the course?

612
00:28:05,507 --> 00:28:06,840
We talk a lot about preferences.

613
00:28:06,840 --> 00:28:09,550
We talked a lot about beliefs
and the interaction of those,

614
00:28:09,550 --> 00:28:13,790
and sort of potential deviations
from the classical model.

615
00:28:13,790 --> 00:28:16,290
Now, we're going to talk about
non-standard decision-making.

616
00:28:16,290 --> 00:28:18,640
And one way to think
about this is to say,

617
00:28:18,640 --> 00:28:21,335
even if you understand
people's preferences very well,

618
00:28:21,335 --> 00:28:22,710
then you can
understand these are

619
00:28:22,710 --> 00:28:24,510
non-standard preferences
in some ways,

620
00:28:24,510 --> 00:28:26,490
and potentially biased beliefs.

621
00:28:26,490 --> 00:28:30,450
Even conditional on that,
people might behave in some ways

622
00:28:30,450 --> 00:28:32,880
where they deviate from
the neoclassical model.

623
00:28:32,880 --> 00:28:36,605
That is hard to explain just
by deviations and preferences

624
00:28:36,605 --> 00:28:37,313
[INAUDIBLE].

625
00:28:37,313 --> 00:28:38,730
So there are
additional deviations

626
00:28:38,730 --> 00:28:41,440
from the neoclassical model
that are beyond changes

627
00:28:41,440 --> 00:28:43,490
in preferences and beliefs.

628
00:28:43,490 --> 00:28:46,770
In particular, I'm going to
talk now about frames, defaults,

629
00:28:46,770 --> 00:28:47,340
and nudges.

630
00:28:47,340 --> 00:28:49,530
Mental accounting is a
similar case of that.

631
00:28:49,530 --> 00:28:52,530
We'll talk about this later.

632
00:28:52,530 --> 00:28:57,098
OK, and so the most famous
paper in this space,

633
00:28:57,098 --> 00:28:58,890
or one of the most
famous paper [INAUDIBLE]

634
00:28:58,890 --> 00:29:01,530
is about 401(k) savings.

635
00:29:01,530 --> 00:29:04,108
So companies offer
these 401(k) savings.

636
00:29:04,108 --> 00:29:04,650
That's right.

637
00:29:04,650 --> 00:29:07,680
So it's in fact the most common
voluntary savings vehicle

638
00:29:07,680 --> 00:29:08,880
in the US.

639
00:29:08,880 --> 00:29:12,660
What you do is you set aside
some money for retirement,

640
00:29:12,660 --> 00:29:15,300
and you choose what's called
the contribution rate.

641
00:29:15,300 --> 00:29:18,180
It's like what fraction of the
income you want to contribute,

642
00:29:18,180 --> 00:29:22,200
and you choose the
asset allocation.

643
00:29:22,200 --> 00:29:26,000
That is offered by
companies-- by employers.

644
00:29:26,000 --> 00:29:28,010
Now, why are
employers doing this?

645
00:29:28,010 --> 00:29:30,625
Because employers want to--

646
00:29:30,625 --> 00:29:32,000
it's seen as sort
of a nice thing

647
00:29:32,000 --> 00:29:37,110
that employers can do
for their employees.

648
00:29:37,110 --> 00:29:39,260
It's potentially
a way in which you

649
00:29:39,260 --> 00:29:44,570
can create some form of
solidarity with your employees.

650
00:29:44,570 --> 00:29:47,960
In a way, you want
your workers to do

651
00:29:47,960 --> 00:29:53,100
well in retirement beyond their
working for you, and so on.

652
00:29:53,100 --> 00:29:56,340
And then, sort of overall,
what companies often do

653
00:29:56,340 --> 00:29:58,560
is they offer matching
contributions up

654
00:29:58,560 --> 00:29:59,450
to a threshold.

655
00:29:59,450 --> 00:30:03,660
This often incentivizes that,
in fact, fairly strongly.

656
00:30:03,660 --> 00:30:05,760
There's also a penalty
for early withdrawal.

657
00:30:05,760 --> 00:30:09,180
Why is there a penalty
for early withdrawal?

658
00:30:09,180 --> 00:30:09,870
Yeah, exactly.

659
00:30:09,870 --> 00:30:12,415
It's like a soft commitment
device in some ways,

660
00:30:12,415 --> 00:30:15,040
in the sense of-- or it's a hard
commitment device in the sense

661
00:30:15,040 --> 00:30:17,100
that it's dominated.

662
00:30:17,100 --> 00:30:20,575
But essentially, it's like you
want-- it's sort of a mixture.

663
00:30:20,575 --> 00:30:22,950
You could have if like a hard
commitment, or a fully hard

664
00:30:22,950 --> 00:30:24,750
commitment, or very
hard commitment device.

665
00:30:24,750 --> 00:30:29,310
But you could say, you cannot
withdraw this until age 60,

666
00:30:29,310 --> 00:30:33,090
or something, or 55, or
65, whenever you retire.

667
00:30:33,090 --> 00:30:34,050
That's not what it is.

668
00:30:34,050 --> 00:30:36,870
Instead, there's a penalty
that's often like 10%.

669
00:30:36,870 --> 00:30:41,150
And the penalty is there
because you want to get people--

670
00:30:41,150 --> 00:30:42,660
avoid that people--

671
00:30:42,660 --> 00:30:44,700
what's called leakage,
that people essentially

672
00:30:44,700 --> 00:30:47,790
withdraw money early,
presumably because

673
00:30:47,790 --> 00:30:51,000
of self-control problems,
present bias, and so on.

674
00:30:51,000 --> 00:30:52,690
There's lots of
leakage, in fact,

675
00:30:52,690 --> 00:30:55,440
even though there's
a 10% withdrawal.

676
00:30:55,440 --> 00:30:56,970
There's a 10%
withdrawal as opposed

677
00:30:56,970 --> 00:30:59,650
to shutting it down entirely
because sometimes people have

678
00:30:59,650 --> 00:31:00,150
shocks.

679
00:31:00,150 --> 00:31:01,650
They have health
problems and so on.

680
00:31:01,650 --> 00:31:02,880
They need a bunch of money.

681
00:31:02,880 --> 00:31:05,400
And in those cases, of course,
you want to allow this,

682
00:31:05,400 --> 00:31:07,320
or the company wants to
allow that to people.

683
00:31:07,320 --> 00:31:09,240
So a 10% penalty
tends to be sort

684
00:31:09,240 --> 00:31:11,760
of compromise between that.

685
00:31:11,760 --> 00:31:13,650
There's tax deferral,
which is you pay,

686
00:31:13,650 --> 00:31:17,100
usually, a lower marginal tax
rate during retirement, right?

687
00:31:17,100 --> 00:31:20,100
When people are 30, 40
they earn a lot of money

688
00:31:20,100 --> 00:31:23,310
compared to when they're
70 during retirement.

689
00:31:23,310 --> 00:31:26,520
So people can pay
taxes when they receive

690
00:31:26,520 --> 00:31:29,100
the money at retirement,
so to the extent

691
00:31:29,100 --> 00:31:32,700
that your marginal tax
rate is lower then,

692
00:31:32,700 --> 00:31:34,630
you're going to save
some money from that.

693
00:31:34,630 --> 00:31:36,422
So that's another reason
why you might want

694
00:31:36,422 --> 00:31:39,200
to use your 401(k) savings.

695
00:31:39,200 --> 00:31:43,370
Now, what are some patterns
401(k) investments?

696
00:31:43,370 --> 00:31:46,110
2/3 of employees believe that
they're saving too little

697
00:31:46,110 --> 00:31:47,510
when you just ask them.

698
00:31:47,510 --> 00:31:51,680
And 1/4 of these intends
to raise their savings

699
00:31:51,680 --> 00:31:52,800
in the next two months.

700
00:31:52,800 --> 00:31:54,650
So that's sort of the typical
thing that sort of smells

701
00:31:54,650 --> 00:31:56,692
like self-control problems,
where you ask people,

702
00:31:56,692 --> 00:32:00,163
like, are you happy
with your savings?

703
00:32:00,163 --> 00:32:01,580
Should you be
saving more or less?

704
00:32:01,580 --> 00:32:04,730
Then people tend to say, oh, no,
no, I want to really save more.

705
00:32:04,730 --> 00:32:06,290
Next two months,
I'm going to do it.

706
00:32:06,290 --> 00:32:09,590
This is sort of the same as when
you ask people about the gym.

707
00:32:09,590 --> 00:32:13,500
I [INAUDIBLE] exercise
more every month.

708
00:32:13,500 --> 00:32:15,860
But it happens.

709
00:32:15,860 --> 00:32:17,690
Almost nobody follows through.

710
00:32:17,690 --> 00:32:20,780
In fact, the
reported undersavers

711
00:32:20,780 --> 00:32:22,550
have low savings rates.

712
00:32:22,550 --> 00:32:25,220
You find similar patterns
also in other surveys.

713
00:32:25,220 --> 00:32:27,770
If you're interested in more
of the Choi et al survey,

714
00:32:27,770 --> 00:32:32,050
it does a very nice job at
summarizing these patterns.

715
00:32:32,050 --> 00:32:36,070
Now, given these patterns,
what are standard economics

716
00:32:36,070 --> 00:32:37,492
tools to increase savings?

717
00:32:37,492 --> 00:32:38,950
Suppose you are
the social planner,

718
00:32:38,950 --> 00:32:42,130
or a company want to
increase people's savings.

719
00:32:42,130 --> 00:32:45,340
What would standard economics
1401, 14-A, or whatever

720
00:32:45,340 --> 00:32:49,630
tell you what should we do
to increase people's savings?

721
00:32:49,630 --> 00:32:51,160
How would you do that?

722
00:32:51,160 --> 00:32:54,070
First, the matching contribution
is very much like an incentive,

723
00:32:54,070 --> 00:32:57,520
so, like, a very standard,
straightforward incentive, Econ

724
00:32:57,520 --> 00:32:59,667
101 would tell you people
react to incentives.

725
00:32:59,667 --> 00:33:01,000
So let's give people incentives.

726
00:33:01,000 --> 00:33:03,160
We provide matching
contributions, and therefore

727
00:33:03,160 --> 00:33:03,740
people--

728
00:33:03,740 --> 00:33:07,868
essentially increases the
rate of return on savings.

729
00:33:07,868 --> 00:33:09,285
And these matching
contributions--

730
00:33:09,285 --> 00:33:12,030
I'll show you in a second--
tend to be very high, or fairly

731
00:33:12,030 --> 00:33:12,530
high.

732
00:33:12,530 --> 00:33:14,450
And so people should
be doing it more

733
00:33:14,450 --> 00:33:17,110
if you make it more
attractive [INAUDIBLE]

734
00:33:17,110 --> 00:33:19,353
under usual circumstances.

735
00:33:19,353 --> 00:33:20,270
What else could we do?

736
00:33:27,360 --> 00:33:30,900
And maybe that's not
necessarily standard economics.

737
00:33:30,900 --> 00:33:33,360
Of course, one option will
be, like, the default.

738
00:33:33,360 --> 00:33:35,890
But what other options
could have or we do?

739
00:33:35,890 --> 00:33:37,880
What else could we do
[INAUDIBLE] people?

740
00:33:37,880 --> 00:33:38,380
Right?

741
00:33:38,380 --> 00:33:40,200
So you could-- so the
matching contribution

742
00:33:40,200 --> 00:33:42,492
is a way of doing that,
essentially because I'm telling

743
00:33:42,492 --> 00:33:47,790
you I'm matching however much
you're contributing by adding

744
00:33:47,790 --> 00:33:50,760
so much money in your
account, and you're not

745
00:33:50,760 --> 00:33:52,050
supposed to withdraw it.

746
00:33:52,050 --> 00:33:55,210
So that's implicitly
giving you a return.

747
00:33:55,210 --> 00:33:57,210
Another thing I think
that you're already saying

748
00:33:57,210 --> 00:33:58,890
is what I have
here as well, which

749
00:33:58,890 --> 00:34:03,065
is provide additional choices
or better choices in some ways,

750
00:34:03,065 --> 00:34:05,190
or improving what's called
our financial education.

751
00:34:05,190 --> 00:34:06,960
That's to say,
maybe people are not

752
00:34:06,960 --> 00:34:08,230
investing in the right thing.

753
00:34:08,230 --> 00:34:10,360
And so maybe I could
either give them

754
00:34:10,360 --> 00:34:12,300
better investment
funds or better options

755
00:34:12,300 --> 00:34:13,960
that have higher returns.

756
00:34:13,960 --> 00:34:17,980
And by doing so, people will
be inclined to save more.

757
00:34:17,980 --> 00:34:20,280
Or maybe I can explain to
them better, for example,

758
00:34:20,280 --> 00:34:22,650
compound interest and so
on, and sort of [INAUDIBLE]

759
00:34:22,650 --> 00:34:24,900
that people don't quite
understand, perhaps

760
00:34:24,900 --> 00:34:26,250
because of lower education.

761
00:34:26,250 --> 00:34:28,167
Now, that's not necessarily
standard economics

762
00:34:28,167 --> 00:34:29,958
in the sense of, like,
people are confused.

763
00:34:29,958 --> 00:34:32,219
That's sort of not what
standard economics would say.

764
00:34:32,219 --> 00:34:34,170
But it might well
be that people are--

765
00:34:34,170 --> 00:34:39,239
if you explain to them better
what is compound interest,

766
00:34:39,239 --> 00:34:42,672
and if you can save now 5% of
your income, how much will that

767
00:34:42,672 --> 00:34:44,130
be in 30 years from
now, people are

768
00:34:44,130 --> 00:34:46,532
often surprised because
that's very high.

769
00:34:46,532 --> 00:34:47,699
You could sort of get them--

770
00:34:47,699 --> 00:34:49,440
provide them with
additional choices,

771
00:34:49,440 --> 00:34:51,199
maybe make it more
attractive for them.

772
00:34:51,199 --> 00:34:53,909
You could even sort of subsidize
it and say, here's a fund,

773
00:34:53,909 --> 00:34:55,159
and I'm giving you more money.

774
00:34:55,159 --> 00:34:57,534
Notice that that would be
quite expensive for the company

775
00:34:57,534 --> 00:34:58,450
to do over time.

776
00:34:58,450 --> 00:35:00,690
So companies tend
to not do that.

777
00:35:00,690 --> 00:35:02,370
Now, it turns out,
none of these tools

778
00:35:02,370 --> 00:35:03,760
tend to be very effective.

779
00:35:03,760 --> 00:35:07,230
So people have tried this,
despite the fact that--

780
00:35:07,230 --> 00:35:09,510
they have been
matching contributions

781
00:35:09,510 --> 00:35:13,590
despite the effect of
financial education, sort

782
00:35:13,590 --> 00:35:16,320
of information sessions,
providing more choices,

783
00:35:16,320 --> 00:35:16,960
and so on.

784
00:35:16,960 --> 00:35:18,793
Notice that it actually
sometimes backfires,

785
00:35:18,793 --> 00:35:20,040
because people get confused.

786
00:35:20,040 --> 00:35:22,930
There's too many choices, or
what's called choice overload.

787
00:35:22,930 --> 00:35:26,250
So these things tend to
not work particularly well.

788
00:35:26,250 --> 00:35:30,990
Now, one question again, why
are companies actually trying

789
00:35:30,990 --> 00:35:33,067
to increase people's savings?

790
00:35:33,067 --> 00:35:34,650
So in some sense,
you might say, well,

791
00:35:34,650 --> 00:35:37,095
we're offering-- here's
your 401(k) savings.

792
00:35:37,095 --> 00:35:43,470
We'll offer it to you, but
whoever wants to actually do it

793
00:35:43,470 --> 00:35:44,130
can use it.

794
00:35:44,130 --> 00:35:46,830
And otherwise, we really
don't care that much.

795
00:35:46,830 --> 00:35:49,920
So why are companies still
interested in doing this?

796
00:35:49,920 --> 00:35:51,323
Why are people
paying this match?

797
00:35:51,323 --> 00:35:52,740
We had some answer
already before,

798
00:35:52,740 --> 00:35:55,450
but that's actually not
quite the answer why

799
00:35:55,450 --> 00:35:58,330
this is [? going. ?]

800
00:35:58,330 --> 00:36:00,290
So to be clear,
there's two questions.

801
00:36:00,290 --> 00:36:02,665
One question that we talked
about a little bit previously

802
00:36:02,665 --> 00:36:06,100
was, like, why might the company
offer savings options anyway?

803
00:36:06,100 --> 00:36:08,320
And the second one is
conditional on offering that--

804
00:36:08,320 --> 00:36:09,820
why is the company
really interested

805
00:36:09,820 --> 00:36:11,175
in increasing take-up?

806
00:36:11,175 --> 00:36:12,550
Because they say,
well, whatever.

807
00:36:12,550 --> 00:36:13,110
We offer it.

808
00:36:13,110 --> 00:36:14,620
Some people like it,
some people don't.

809
00:36:14,620 --> 00:36:16,203
If they don't like
it, why do we care?

810
00:36:18,480 --> 00:36:20,960
So I always thought, you know--

811
00:36:20,960 --> 00:36:24,013
when I first read about this,
I thought companies are nice

812
00:36:24,013 --> 00:36:26,180
and they want to sort of
be good to their employers,

813
00:36:26,180 --> 00:36:28,160
and they want the best
for their workers.

814
00:36:28,160 --> 00:36:30,950
And you're already a lot
more cynical in a sense of,

815
00:36:30,950 --> 00:36:33,350
like, you're kind
of saying, how do we

816
00:36:33,350 --> 00:36:36,800
get workers to work a lot more?

817
00:36:36,800 --> 00:36:37,320
It could be.

818
00:36:37,320 --> 00:36:39,320
I mean, I haven't heard
of what you were saying.

819
00:36:39,320 --> 00:36:42,470
But I think in principle,
that could be a reason where--

820
00:36:42,470 --> 00:36:45,260
in a way, I think companies
like workers to be happy.

821
00:36:45,260 --> 00:36:49,310
And to the extent that if
they consume more right now,

822
00:36:49,310 --> 00:36:52,142
that probably makes them happier
and perhaps more productive,

823
00:36:52,142 --> 00:36:53,600
so then, in some
ways, probably you

824
00:36:53,600 --> 00:36:56,000
have the incentives to have
people not save too much.

825
00:36:56,000 --> 00:36:57,708
Because in some ways,
you don't care that

826
00:36:57,708 --> 00:37:02,900
much about whether somebody
who is 60 is particularly happy

827
00:37:02,900 --> 00:37:03,530
or has money.

828
00:37:03,530 --> 00:37:07,200
Because essentially, you
can't really engage or you--

829
00:37:07,200 --> 00:37:08,510
age discrimination is illegal.

830
00:37:08,510 --> 00:37:11,660
So you can't fire, like,
a 57-year-old person who

831
00:37:11,660 --> 00:37:13,160
is doing their job
reasonably well,

832
00:37:13,160 --> 00:37:16,890
but maybe is not as productive
as you'd like them to be.

833
00:37:16,890 --> 00:37:22,070
And so to them, if that person
has saved a lot of money,

834
00:37:22,070 --> 00:37:24,770
they probably are
more likely to retire.

835
00:37:24,770 --> 00:37:27,530
You guys are more
cynical than I am.

836
00:37:27,530 --> 00:37:32,730
So the reported reason for
a lot of these policies

837
00:37:32,730 --> 00:37:35,790
is, in fact, the reason that
there is IRS nondiscrimination

838
00:37:35,790 --> 00:37:36,820
rules.

839
00:37:36,820 --> 00:37:38,020
What does that mean?

840
00:37:38,020 --> 00:37:44,910
Essentially, by the IRS rules,
it cannot be that people who

841
00:37:44,910 --> 00:37:50,430
are paid a lot can benefit from
certain retirement and other

842
00:37:50,430 --> 00:37:53,280
schemes way more than people
who are paid very little.

843
00:37:53,280 --> 00:37:56,190
Now it turns out that
low income or low wage

844
00:37:56,190 --> 00:38:00,150
workers are much
less likely to save

845
00:38:00,150 --> 00:38:04,640
in 401(k) and other
savings schemes.

846
00:38:04,640 --> 00:38:07,310
Therefore, they're much
less likely to receive

847
00:38:07,310 --> 00:38:10,460
any of these 401(k) matching
and other contributions.

848
00:38:10,460 --> 00:38:12,440
And therefore, essentially,
their companies

849
00:38:12,440 --> 00:38:14,990
tend to have way
more money spent

850
00:38:14,990 --> 00:38:18,310
on sort of high paid employees
than low paid employees.

851
00:38:18,310 --> 00:38:20,060
And there are some
nondiscrimination rules

852
00:38:20,060 --> 00:38:23,360
by the IRS that
says that's illegal.

853
00:38:23,360 --> 00:38:24,740
You can't do that.

854
00:38:24,740 --> 00:38:29,180
And therefore, companies are
very keen on getting lower--

855
00:38:29,180 --> 00:38:30,740
if you are like a
CEO or something,

856
00:38:30,740 --> 00:38:33,330
you want your
workers to earn very

857
00:38:33,330 --> 00:38:38,570
little to invest in 401(k)
and other types of schemes

858
00:38:38,570 --> 00:38:40,970
that they get rewarded
for, because that allows

859
00:38:40,970 --> 00:38:43,670
them, essentially, to pay
the high paid employees

860
00:38:43,670 --> 00:38:46,958
higher bonuses, in these
kinds of schemes as well.

861
00:38:46,958 --> 00:38:48,500
That's essentially
one of the reasons

862
00:38:48,500 --> 00:38:50,780
why companies are
very, very interested

863
00:38:50,780 --> 00:38:54,410
in increasing their 401(k)
savings, in particular

864
00:38:54,410 --> 00:38:58,040
among workers with low wages.

865
00:38:58,040 --> 00:39:03,560
And that's exactly what the
default to default policy's

866
00:39:03,560 --> 00:39:04,100
tend to do.

867
00:39:04,100 --> 00:39:05,600
Chad, did you have
another question,

868
00:39:05,600 --> 00:39:07,854
or was that just your
hand that's still up?

869
00:39:11,020 --> 00:39:14,330
OK, so let me move on.

870
00:39:14,330 --> 00:39:17,642
So why would you participate
in 401(k) savings schemes?

871
00:39:17,642 --> 00:39:18,850
We already talked about this.

872
00:39:18,850 --> 00:39:20,770
What are the costs
of nonparticipation?

873
00:39:20,770 --> 00:39:23,440
There are essentially
three main reasons

874
00:39:23,440 --> 00:39:25,120
from the worker's perspective.

875
00:39:25,120 --> 00:39:26,800
There's foregone tax benefits.

876
00:39:26,800 --> 00:39:29,960
And tax benefits are, again,
your marginal tax rate,

877
00:39:29,960 --> 00:39:31,600
usually while
people are working,

878
00:39:31,600 --> 00:39:33,730
tends to be higher than
the marginal tax rate

879
00:39:33,730 --> 00:39:35,240
when people are retiring.

880
00:39:35,240 --> 00:39:40,370
So essentially, by
investing in 401(k) savings,

881
00:39:40,370 --> 00:39:44,060
they can defer taxes
and save some taxes.

882
00:39:44,060 --> 00:39:49,530
Second, people forgo
the employer match.

883
00:39:49,530 --> 00:39:51,290
I'll tell you about
this in the second.

884
00:39:51,290 --> 00:39:52,730
And number three is--

885
00:39:52,730 --> 00:39:54,320
that's a little
more subtle, which

886
00:39:54,320 --> 00:39:57,470
is foregone consumption
smoothing, which, essentially,

887
00:39:57,470 --> 00:40:00,440
is your marginal--

888
00:40:00,440 --> 00:40:02,990
the marginal value
of a dollar for you,

889
00:40:02,990 --> 00:40:06,220
while you're working
at age 30 or 40,

890
00:40:06,220 --> 00:40:09,760
tends to be lower than the
marginal value of a dollar

891
00:40:09,760 --> 00:40:11,367
once you're 70 or 80.

892
00:40:11,367 --> 00:40:13,450
The reason being that,
essentially, you're richer.

893
00:40:13,450 --> 00:40:15,970
You have more money available
when you're 30 or 40,

894
00:40:15,970 --> 00:40:18,790
so you actually want to smooth
consumption to later in time.

895
00:40:18,790 --> 00:40:22,480
Of course, there's a discounting
motive going the other way.

896
00:40:22,480 --> 00:40:26,320
But it shouldn't be that,
essentially, at age 59 or 64,

897
00:40:26,320 --> 00:40:28,630
you earn a lot of money
and consume a lot, and then

898
00:40:28,630 --> 00:40:33,370
the year after, once you retire,
your consumption drops a lot.

899
00:40:33,370 --> 00:40:36,100
Economists would say you
should smooth this out by,

900
00:40:36,100 --> 00:40:38,200
essentially, saving some
money to the future.

901
00:40:38,200 --> 00:40:41,410
And 401(k) savings [INAUDIBLE]
might help you with that.

902
00:40:41,410 --> 00:40:43,900
Now, why do companies care?

903
00:40:43,900 --> 00:40:44,950
I already said that.

904
00:40:44,950 --> 00:40:49,120
Essentially-- whoops,
sorry, going back.

905
00:40:49,120 --> 00:40:51,430
Non-highly compensated
employees don't save enough,

906
00:40:51,430 --> 00:40:53,950
or don't save very much, in
particular in new schemes.

907
00:40:53,950 --> 00:40:56,020
And the IRS
nondiscrimination tests--

908
00:40:56,020 --> 00:40:57,730
has essentially
nondiscrimination tests

909
00:40:57,730 --> 00:41:02,685
in pension plans that
precisely try to prevent that.

910
00:41:02,685 --> 00:41:04,060
So really, what
you want to do is

911
00:41:04,060 --> 00:41:10,050
you want to get your low
income workers to save.

912
00:41:10,050 --> 00:41:12,960
OK, so now here comes the
paper by Madrian and Shea.

913
00:41:12,960 --> 00:41:16,620
That's a very successful and one
of the most successful papers

914
00:41:16,620 --> 00:41:18,360
in behavioral economics.

915
00:41:18,360 --> 00:41:24,117
The reason being that, A,
standard economics has sort of

916
00:41:24,117 --> 00:41:25,950
failed in some ways,
in the sense of saying,

917
00:41:25,950 --> 00:41:29,010
like, we tried
matching contributions,

918
00:41:29,010 --> 00:41:31,590
we tried giving people
education or information,

919
00:41:31,590 --> 00:41:33,690
and so on, and so forth,
and none of these things

920
00:41:33,690 --> 00:41:36,130
have worked very well.

921
00:41:36,130 --> 00:41:39,112
And then instead, something
that shouldn't matter--

922
00:41:39,112 --> 00:41:41,320
which is what I'm going to
show you in a second-- the

923
00:41:41,320 --> 00:41:44,500
defaults mattered a lot,
and had huge effects,

924
00:41:44,500 --> 00:41:48,320
which is very hard for
standard economics to explain.

925
00:41:48,320 --> 00:41:50,287
And then on top of that,
this is in a domain

926
00:41:50,287 --> 00:41:51,620
that we really care about a lot.

927
00:41:51,620 --> 00:41:53,680
Savings is one of the
most important choices

928
00:41:53,680 --> 00:41:56,150
that economists have considered.

929
00:41:56,150 --> 00:41:57,700
And so that's a
very good example

930
00:41:57,700 --> 00:42:02,500
of an early example of an
aspect of the economics,

931
00:42:02,500 --> 00:42:04,270
in this case
default, that really

932
00:42:04,270 --> 00:42:06,670
matters in some ways
that's hard to explain

933
00:42:06,670 --> 00:42:08,020
for standard economics.

934
00:42:08,020 --> 00:42:10,270
And the standard
tools of economics

935
00:42:10,270 --> 00:42:13,350
seem to not have been
working very well.

936
00:42:13,350 --> 00:42:15,570
OK, so what does the
paper actually do?

937
00:42:15,570 --> 00:42:19,620
It's a large publicly-traded,
fortune 500 healthcare company.

938
00:42:19,620 --> 00:42:22,350
You can enroll in 401(k)
savings plans any day

939
00:42:22,350 --> 00:42:24,660
by essentially filling
out an enrollment form,

940
00:42:24,660 --> 00:42:27,330
or calling the
401(k) record keeper.

941
00:42:27,330 --> 00:42:28,855
What does that mean?

942
00:42:28,855 --> 00:42:30,480
Essentially, you call
somebody and say,

943
00:42:30,480 --> 00:42:32,280
I want you to invest
in 401(k) savings

944
00:42:32,280 --> 00:42:33,570
when you're in this company.

945
00:42:33,570 --> 00:42:37,430
And you're going to
say, I'm going to--

946
00:42:37,430 --> 00:42:40,130
I want to save, say,
5% of my earnings.

947
00:42:40,130 --> 00:42:42,260
I want to save every month.

948
00:42:42,260 --> 00:42:43,850
Or you could fill out a form.

949
00:42:43,850 --> 00:42:46,560
Now, these days, a
lot of this is online.

950
00:42:46,560 --> 00:42:49,310
So if at MIT, you want to
change your 401(k) savings,

951
00:42:49,310 --> 00:42:51,800
you essentially have to
fill out an online form,

952
00:42:51,800 --> 00:42:55,350
and change your online
allocation [? potential. ?]

953
00:42:55,350 --> 00:42:59,210
So there's potentially small
transaction costs involved.

954
00:42:59,210 --> 00:43:02,090
But notice that the transaction
cost tends to be very minor,

955
00:43:02,090 --> 00:43:04,580
in a sense, like, it's going
to take you maybe 10 minutes,

956
00:43:04,580 --> 00:43:07,110
maybe 20 minutes-- maybe if
you think about it for a while,

957
00:43:07,110 --> 00:43:08,120
it takes you an hour--

958
00:43:08,120 --> 00:43:12,220
but really, it's a
very minor cost of time

959
00:43:12,220 --> 00:43:15,910
overall that this
costs you compared

960
00:43:15,910 --> 00:43:20,780
to the potentially huge benefits
of employer matches and so on.

961
00:43:20,780 --> 00:43:23,170
Now, what are these
employer matches?

962
00:43:23,170 --> 00:43:26,290
In this particular company,
it's a 50% matching contribution

963
00:43:26,290 --> 00:43:28,210
for the first 6%.

964
00:43:28,210 --> 00:43:29,920
So again, what are
these percentages?

965
00:43:29,920 --> 00:43:33,820
The percentages are essentially
the fraction of your income

966
00:43:33,820 --> 00:43:37,620
that you'd like to
contribute to these savings.

967
00:43:37,620 --> 00:43:42,510
For example, 5% or 3%, or
7% is what you're saving.

968
00:43:42,510 --> 00:43:44,130
And now the company
says they're going

969
00:43:44,130 --> 00:43:48,360
to match 50% for the first 6%.

970
00:43:48,360 --> 00:43:51,180
For example, if an
employee chooses 4%,

971
00:43:51,180 --> 00:43:53,520
the company pays
an additional 2%

972
00:43:53,520 --> 00:43:57,100
into that savings
account right away.

973
00:43:57,100 --> 00:43:58,830
So that's actually
quite a bit of money.

974
00:43:58,830 --> 00:44:01,980
If an employee chooses
10%, the company

975
00:44:01,980 --> 00:44:05,460
pays an additional 3%,
because that employee

976
00:44:05,460 --> 00:44:07,380
is above the cap of 6%.

977
00:44:07,380 --> 00:44:11,580
So you cannot get more
than 3% from the company.

978
00:44:11,580 --> 00:44:13,140
Other companies
these days have even,

979
00:44:13,140 --> 00:44:17,940
like, 100% matching
contribution for 5% or the like.

980
00:44:17,940 --> 00:44:20,020
That tends to be fairly common.

981
00:44:20,020 --> 00:44:23,790
So for example, MIT has
100% matching contribution

982
00:44:23,790 --> 00:44:28,610
for the first, I think, 5%.

983
00:44:28,610 --> 00:44:30,920
In this specific
company, the employees

984
00:44:30,920 --> 00:44:34,370
are first eligible after
one year of employment

985
00:44:34,370 --> 00:44:36,355
before the change
happened of the policy

986
00:44:36,355 --> 00:44:38,330
that I'm going to talk
about in a second.

987
00:44:38,330 --> 00:44:41,090
Now, notice that this
is a terrible policy

988
00:44:41,090 --> 00:44:43,130
from a behavioral
economics perspective,

989
00:44:43,130 --> 00:44:45,680
in terms of if you wanted
people to sign up and go

990
00:44:45,680 --> 00:44:47,543
do this, that's not
what you would do.

991
00:44:47,543 --> 00:44:50,210
If I'm telling you, like, here's
a thing that I want you to sign

992
00:44:50,210 --> 00:44:53,540
up for, in a year from now
please come back and do it,

993
00:44:53,540 --> 00:44:56,270
well, in a year from
now, people might forget.

994
00:44:56,270 --> 00:44:57,620
People might procrastinate.

995
00:44:57,620 --> 00:45:01,590
People might be busy
with other things.

996
00:45:01,590 --> 00:45:06,742
So it's not a policy that's
great to do to start with

997
00:45:06,742 --> 00:45:07,700
to get people involved.

998
00:45:07,700 --> 00:45:09,200
And in some sense,
there's no wonder

999
00:45:09,200 --> 00:45:12,950
that participation early
on was kind of low.

1000
00:45:12,950 --> 00:45:15,230
Now, the company does
that because turnover

1001
00:45:15,230 --> 00:45:16,130
tends to be high.

1002
00:45:16,130 --> 00:45:18,050
And you try to
avoid paying people

1003
00:45:18,050 --> 00:45:20,013
early on that
might leave anyway.

1004
00:45:20,013 --> 00:45:21,680
And you have to deal
with the paperwork,

1005
00:45:21,680 --> 00:45:23,127
and so on and so forth.

1006
00:45:23,127 --> 00:45:25,460
Notice that that might be
exactly the wrong thing to do,

1007
00:45:25,460 --> 00:45:28,610
because once people have paid
into their 401(k) savings,

1008
00:45:28,610 --> 00:45:31,280
maybe that creates some loyalty
and makes workers to stay

1009
00:45:31,280 --> 00:45:32,750
more at the company anyway.

1010
00:45:32,750 --> 00:45:35,240
But for what it's worth,
in this particular case,

1011
00:45:35,240 --> 00:45:40,160
employees were only eligible
and one year after employment.

1012
00:45:40,160 --> 00:45:43,000
Now, then there's
a discontinuity

1013
00:45:43,000 --> 00:45:47,830
of the 401(k) plan default
based on the date of hire.

1014
00:45:47,830 --> 00:45:49,540
So here's the old cohort--

1015
00:45:49,540 --> 00:45:53,050
the window cohort, as I call
it-- and the new cohort.

1016
00:45:53,050 --> 00:45:57,910
So the old cohort are people who
are hired between 1st of April,

1017
00:45:57,910 --> 00:46:00,480
'96 and 31st of March, '97.

1018
00:46:00,480 --> 00:46:03,110
So they are essentially people
who have been previously hired.

1019
00:46:03,110 --> 00:46:04,960
This is what I just told you.

1020
00:46:04,960 --> 00:46:08,140
You can, after one year--

1021
00:46:08,140 --> 00:46:11,580
after being hired after
one year, you can--

1022
00:46:11,580 --> 00:46:13,060
you're not
automatically enrolled,

1023
00:46:13,060 --> 00:46:15,380
but you can essentially
just call somebody and say,

1024
00:46:15,380 --> 00:46:18,430
I would like to start
saving in my 401(k) savings.

1025
00:46:18,430 --> 00:46:20,890
Or you could fill out this form.

1026
00:46:20,890 --> 00:46:22,390
Then, there is the
new cohort, which

1027
00:46:22,390 --> 00:46:24,940
essentially is automatic
enrollment, which

1028
00:46:24,940 --> 00:46:28,690
is everything is exactly the
same, but by the date of hire.

1029
00:46:28,690 --> 00:46:30,760
So when you're
hired immediately,

1030
00:46:30,760 --> 00:46:32,170
you're automatically enrolled.

1031
00:46:32,170 --> 00:46:35,150
So they're not asking you.

1032
00:46:35,150 --> 00:46:36,650
They're giving you
some information,

1033
00:46:36,650 --> 00:46:38,470
but they say, unless
you change, unless you

1034
00:46:38,470 --> 00:46:40,810
want to make a
change, you're going

1035
00:46:40,810 --> 00:46:43,810
to be automatically
enrolled immediately.

1036
00:46:43,810 --> 00:46:47,020
And when you automatically
enrolled somebody,

1037
00:46:47,020 --> 00:46:49,750
then you need to also specify
where the money goes to.

1038
00:46:49,750 --> 00:46:51,740
So they have an
automatic enrollment.

1039
00:46:51,740 --> 00:46:54,100
You need to choose the
contribution rate, which is

1040
00:46:54,100 --> 00:46:55,970
the default contribution rate.

1041
00:46:55,970 --> 00:46:58,690
And they need to have a
default fund allocation, which

1042
00:46:58,690 --> 00:47:01,870
is-- so they have a default
contribution rate which is 3%.

1043
00:47:01,870 --> 00:47:03,940
And you need to put the
money somewhere, right?

1044
00:47:03,940 --> 00:47:06,340
So if I'm automatically
doing something for you

1045
00:47:06,340 --> 00:47:08,590
with your money, I need to
put the money somewhere.

1046
00:47:08,590 --> 00:47:10,690
What the company is
doing is the company

1047
00:47:10,690 --> 00:47:15,190
puts the money into
the money market fund.

1048
00:47:15,190 --> 00:47:17,350
Is that a good idea, a bad idea?

1049
00:47:17,350 --> 00:47:18,850
Why is the company
putting the money

1050
00:47:18,850 --> 00:47:19,990
into the money market fund?

1051
00:47:19,990 --> 00:47:22,448
Is that what you're supposed
to be doing with your savings?

1052
00:47:22,448 --> 00:47:34,340
[INAUDIBLE] Or
maybe first explain

1053
00:47:34,340 --> 00:47:36,410
what is a money market
fund, and then tell

1054
00:47:36,410 --> 00:47:37,610
is it a good or bad idea.

1055
00:47:37,610 --> 00:47:40,770
Somebody says they saw dot come
bubble coming, potentially.

1056
00:47:40,770 --> 00:47:44,340
But I think if you save for
retirement, what should you do?

1057
00:47:44,340 --> 00:47:48,060
Here's free investment
advice for everybody.

1058
00:47:48,060 --> 00:47:48,923
What should you do?

1059
00:47:48,923 --> 00:47:50,340
What are people
telling you, like,

1060
00:47:50,340 --> 00:47:52,063
if you go to somebody
who does finance?

1061
00:47:52,063 --> 00:47:53,730
[INAUDIBLE] but
essentially, what people

1062
00:47:53,730 --> 00:47:55,770
would argue you
should be doing is

1063
00:47:55,770 --> 00:47:57,960
you should invest in
stocks or some forms

1064
00:47:57,960 --> 00:48:03,480
of diversified
portfolio of stocks.

1065
00:48:03,480 --> 00:48:05,970
You should buy, essentially,
the stock market, S&P 500,

1066
00:48:05,970 --> 00:48:08,820
or the world stock
market, or the like,

1067
00:48:08,820 --> 00:48:12,750
which essentially is risky, but
also has a much higher return.

1068
00:48:12,750 --> 00:48:17,910
Now, given that if you started
investing at age 25 or 30,

1069
00:48:17,910 --> 00:48:20,160
if you do that for
30, 40 years, that

1070
00:48:20,160 --> 00:48:24,160
has a very likely very high
long-run return, and a much

1071
00:48:24,160 --> 00:48:26,160
higher long-run return
than you will have if you

1072
00:48:26,160 --> 00:48:28,160
invest in the money market.

1073
00:48:28,160 --> 00:48:32,170
Now, there's a question of what
is the optimal risk and so on.

1074
00:48:32,170 --> 00:48:36,870
But investing, nobody--
no investment advisor

1075
00:48:36,870 --> 00:48:38,740
in their right mind,
maybe perhaps partially

1076
00:48:38,740 --> 00:48:41,480
because they want to make money,
but I think including people

1077
00:48:41,480 --> 00:48:43,570
who want the best
for you will tell you

1078
00:48:43,570 --> 00:48:46,430
you should invest everything
and the money market.

1079
00:48:46,430 --> 00:48:48,843
Now why are the
companies doing that?

1080
00:48:48,843 --> 00:48:51,010
Essentially, the company
is worried about liability,

1081
00:48:51,010 --> 00:48:53,500
and about getting sued, or
getting, at least, in trouble

1082
00:48:53,500 --> 00:48:55,510
with their employees.

1083
00:48:55,510 --> 00:48:57,160
If they default--
so essentially,

1084
00:48:57,160 --> 00:49:01,090
if I'm not offering
any retirement savings,

1085
00:49:01,090 --> 00:49:04,390
then I'm like, of course,
that's not a big risk to take.

1086
00:49:04,390 --> 00:49:06,160
If I'm in a sense
saying, I'm hiring you,

1087
00:49:06,160 --> 00:49:08,230
and I'm putting your
money without really

1088
00:49:08,230 --> 00:49:09,760
getting explicit
consent for you,

1089
00:49:09,760 --> 00:49:12,040
without you explicitly
saying you want it,

1090
00:49:12,040 --> 00:49:16,120
and putting your money into some
investment fund, potentially

1091
00:49:16,120 --> 00:49:19,210
a risky investment fund, and
that thing essentially crashes,

1092
00:49:19,210 --> 00:49:20,950
and people lose a
lot of money, they're

1093
00:49:20,950 --> 00:49:23,080
going to get really upset
with you, potentially.

1094
00:49:23,080 --> 00:49:25,240
And that's, from the
company's perspective,

1095
00:49:25,240 --> 00:49:26,653
just not worth doing.

1096
00:49:26,653 --> 00:49:28,570
Essentially, the company
wants to help people.

1097
00:49:28,570 --> 00:49:33,270
Again, remember a part of
the motive of the company

1098
00:49:33,270 --> 00:49:35,970
was just to get people to
contribute in the first place,

1099
00:49:35,970 --> 00:49:40,130
to essentially get around the
IRS nondiscrimination rules.

1100
00:49:40,130 --> 00:49:43,840
And so the company is really
not necessarily that interested

1101
00:49:43,840 --> 00:49:47,810
in high returns
for their workers.

1102
00:49:47,810 --> 00:49:50,130
And in particular, companies
worry potentially about,

1103
00:49:50,130 --> 00:49:52,130
essentially, the stock
market crashing, and then

1104
00:49:52,130 --> 00:49:54,297
sort of people getting
really upset with them, which

1105
00:49:54,297 --> 00:49:55,589
would be quite costly for them.

1106
00:49:55,589 --> 00:49:57,297
And then they would
be, like, why did you

1107
00:49:57,297 --> 00:49:58,700
put the money in
this risky thing

1108
00:49:58,700 --> 00:50:01,880
while in fact, you could have
put it in the money market,

1109
00:50:01,880 --> 00:50:02,612
and so on?

1110
00:50:02,612 --> 00:50:04,070
And the reason
being is essentially

1111
00:50:04,070 --> 00:50:05,720
because there's no
explicit consent,

1112
00:50:05,720 --> 00:50:09,230
because it's essentially
a default option where

1113
00:50:09,230 --> 00:50:12,420
the default is such that
you're automatically enrolled.

1114
00:50:12,420 --> 00:50:14,360
So unless you tell
me otherwise, you're

1115
00:50:14,360 --> 00:50:16,760
going to be-- the money's
going to put there.

1116
00:50:16,760 --> 00:50:20,300
And so there's never
something where

1117
00:50:20,300 --> 00:50:25,800
the person says, yes, I want the
money to be invested in x or y.

1118
00:50:25,800 --> 00:50:28,830
So then, this here
is the main result

1119
00:50:28,830 --> 00:50:31,080
of the paper, which
is the participation

1120
00:50:31,080 --> 00:50:35,100
rates in 401(k) savings
accounts are dramatically

1121
00:50:35,100 --> 00:50:37,440
higher with automatic
enrollment compared

1122
00:50:37,440 --> 00:50:39,640
to the previous regime.

1123
00:50:39,640 --> 00:50:41,910
So you see here, this
is the old cohort.

1124
00:50:41,910 --> 00:50:43,350
This is the window cohort.

1125
00:50:43,350 --> 00:50:44,850
And these are previous cohorts.

1126
00:50:44,850 --> 00:50:47,910
So you see that prior
to automatic enrollment,

1127
00:50:47,910 --> 00:50:49,950
the participation increases--

1128
00:50:49,950 --> 00:50:51,188
or increased with the tenure.

1129
00:50:51,188 --> 00:50:53,230
That is to say, the longer
you're at the company,

1130
00:50:53,230 --> 00:50:54,870
the higher the
fraction of people

1131
00:50:54,870 --> 00:50:57,900
who participate
in 401(k) savings.

1132
00:50:57,900 --> 00:51:01,380
But even people who have been
in the company for 20 plus years

1133
00:51:01,380 --> 00:51:06,030
have only, like, a
contribution rate of about 83%.

1134
00:51:06,030 --> 00:51:07,830
Now, if you do automatic
enrollment, which

1135
00:51:07,830 --> 00:51:14,267
is getting people
enrolled-- but people

1136
00:51:14,267 --> 00:51:15,850
are free to choose
whatever they want.

1137
00:51:15,850 --> 00:51:18,880
If they don't want to invest
money into the 401(k) savings

1138
00:51:18,880 --> 00:51:19,630
account.

1139
00:51:19,630 --> 00:51:21,950
They can opt out at
any point in time.

1140
00:51:21,950 --> 00:51:24,790
But when you do that,
people get essentially way

1141
00:51:24,790 --> 00:51:28,150
more likely to be enrolled.

1142
00:51:28,150 --> 00:51:31,720
The participation rate
is as high as 86%.

1143
00:51:31,720 --> 00:51:33,940
So there are essentially
dramatic differences

1144
00:51:33,940 --> 00:51:36,520
in what you would argue is a
very important choice, which

1145
00:51:36,520 --> 00:51:39,550
is, if you think about what are
important choices in your life,

1146
00:51:39,550 --> 00:51:42,027
savings choice is one of
the most important ones.

1147
00:51:42,027 --> 00:51:43,360
So there's dramatic differences.

1148
00:51:43,360 --> 00:51:45,220
And when you do
automatic enrollment,

1149
00:51:45,220 --> 00:51:48,910
people are way more likely
to enroll than if not.

1150
00:51:48,910 --> 00:51:51,790
Now, in addition-- sorry.

1151
00:51:51,790 --> 00:51:55,430
And then this is here
is, I guess, the part

1152
00:51:55,430 --> 00:51:57,140
about the low income workers.

1153
00:51:57,140 --> 00:51:58,850
When you look at--

1154
00:51:58,850 --> 00:52:00,560
this is-- when you
look at the first two

1155
00:52:00,560 --> 00:52:05,250
columns of this table,
here you see these effects.

1156
00:52:05,250 --> 00:52:07,605
So first you see, this is
the window cohort here.

1157
00:52:07,605 --> 00:52:09,230
This is the cohort
that was essentially

1158
00:52:09,230 --> 00:52:11,720
before the change was done.

1159
00:52:11,720 --> 00:52:14,600
And here is the
new cohort, which

1160
00:52:14,600 --> 00:52:18,470
is the cohort that received
automatic enrollment.

1161
00:52:18,470 --> 00:52:23,240
The difference is dramatic
from like 37% to 86%.

1162
00:52:23,240 --> 00:52:26,630
And this is now by
compensation levels of workers.

1163
00:52:26,630 --> 00:52:28,880
And what you see here
is, in the window cohort,

1164
00:52:28,880 --> 00:52:33,800
you see this huge gradient
here by compensation.

1165
00:52:33,800 --> 00:52:35,870
So workers who are--
and this is before

1166
00:52:35,870 --> 00:52:37,310
the automatic enrollment.

1167
00:52:37,310 --> 00:52:41,330
Workers who received
fewer than $20,000

1168
00:52:41,330 --> 00:52:45,350
per year, only 12.5%
of them actually

1169
00:52:45,350 --> 00:52:48,110
were participating in 401(k)
savings to start with.

1170
00:52:48,110 --> 00:52:51,230
And among people who earned
a lot already anyway,

1171
00:52:51,230 --> 00:52:53,047
this thing was, like, 68%.

1172
00:52:53,047 --> 00:52:54,380
There's a huge discrepancy here.

1173
00:52:54,380 --> 00:53:00,300
And this is precisely where
the IRS discrimination

1174
00:53:00,300 --> 00:53:01,490
rules were [? fighting. ?]

1175
00:53:01,490 --> 00:53:04,310
Now, but if you look at
after the enrollment,

1176
00:53:04,310 --> 00:53:06,740
there's still-- notice that
there's still some gradient.

1177
00:53:06,740 --> 00:53:09,770
There's still, essentially,
like, a higher participation

1178
00:53:09,770 --> 00:53:11,930
rate among the high
paid employees.

1179
00:53:11,930 --> 00:53:14,480
But the grading is much
flatter, because essentially,

1180
00:53:14,480 --> 00:53:19,670
so many among the low income
workers or low wage workers

1181
00:53:19,670 --> 00:53:21,780
are now enrolled as well.

1182
00:53:21,780 --> 00:53:24,620
So there's huge effect, and the
effective is disproportionately

1183
00:53:24,620 --> 00:53:28,070
large among low income workers.

1184
00:53:28,070 --> 00:53:31,140
What are potential
reasons for that?

1185
00:53:31,140 --> 00:53:32,528
So they have more likely--

1186
00:53:32,528 --> 00:53:34,070
essentially, their
credit constraints

1187
00:53:34,070 --> 00:53:35,240
are more likely to bind.

1188
00:53:35,240 --> 00:53:37,490
Or they have a high
marginal utility of money,

1189
00:53:37,490 --> 00:53:40,340
and therefore, they might
be less inclined to save.

1190
00:53:40,340 --> 00:53:42,140
Notice that low
income workers also

1191
00:53:42,140 --> 00:53:45,227
have a high marginal utility
of income in the future.

1192
00:53:45,227 --> 00:53:47,060
So in some ways, it's
not obvious, actually,

1193
00:53:47,060 --> 00:53:48,920
that when you think about--
[? because when you ?] write

1194
00:53:48,920 --> 00:53:51,650
down a model, you think about
should you save right now

1195
00:53:51,650 --> 00:53:54,835
versus consume money--

1196
00:53:54,835 --> 00:53:58,030
so consume versus save.

1197
00:53:58,030 --> 00:54:00,740
You will get, essentially, what
this-- what this depends on

1198
00:54:00,740 --> 00:54:03,350
is the ratio of your
marginal utility right now

1199
00:54:03,350 --> 00:54:04,970
versus in the future.

1200
00:54:04,970 --> 00:54:07,430
And the marginal utility of
poor workers in the future

1201
00:54:07,430 --> 00:54:08,347
will also be high.

1202
00:54:08,347 --> 00:54:10,430
So it's not obvious that
they shouldn't be saving.

1203
00:54:10,430 --> 00:54:11,330
But I think it's
right that we're

1204
00:54:11,330 --> 00:54:12,747
saying they have
more emergencies,

1205
00:54:12,747 --> 00:54:16,290
and more reasons to essentially
not want the money right now.

1206
00:54:16,290 --> 00:54:19,770
So one thing could be just it's
mechanical in the sense of,

1207
00:54:19,770 --> 00:54:22,190
in some ways, it could just
be that for whatever reason,

1208
00:54:22,190 --> 00:54:25,010
the poor are not
saving to start with.

1209
00:54:25,010 --> 00:54:26,690
Maybe that's in
part due to tenure.

1210
00:54:26,690 --> 00:54:28,550
Maybe it's due to
some other reasons

1211
00:54:28,550 --> 00:54:33,380
that we don't quite understand.

1212
00:54:33,380 --> 00:54:37,370
But it might also be
that the poor might

1213
00:54:37,370 --> 00:54:41,650
be more prone to default
effects, which could be in part

1214
00:54:41,650 --> 00:54:42,680
what Carmen was saying.

1215
00:54:42,680 --> 00:54:45,620
They're sort of--
in a way, perhaps

1216
00:54:45,620 --> 00:54:48,170
financial sophistication,
or education,

1217
00:54:48,170 --> 00:54:50,420
or information-- just what's
the right thing to do--

1218
00:54:50,420 --> 00:54:51,410
might be lower.

1219
00:54:51,410 --> 00:54:52,430
And you might, as
in, like, I don't

1220
00:54:52,430 --> 00:54:53,460
understand what's going on.

1221
00:54:53,460 --> 00:54:54,980
So I'm just doing what
the company does for me

1222
00:54:54,980 --> 00:54:56,240
and just hope for the best.

1223
00:54:56,240 --> 00:54:58,340
And then you might
just go with whatever

1224
00:54:58,340 --> 00:55:01,370
that is, and you're less
confident in making different--

1225
00:55:01,370 --> 00:55:04,022
or overriding such choices.

1226
00:55:04,022 --> 00:55:05,480
Another explanation--
and I'm going

1227
00:55:05,480 --> 00:55:07,710
to talk about this when we
talk about poverty a bit--

1228
00:55:07,710 --> 00:55:11,660
which is the idea that poverty
taxes people's bandwidth

1229
00:55:11,660 --> 00:55:12,360
and attention.

1230
00:55:12,360 --> 00:55:13,735
That's to say,
like, essentially,

1231
00:55:13,735 --> 00:55:15,727
being poor is really
hard in various ways.

1232
00:55:15,727 --> 00:55:17,810
You have to struggle with
so many different things

1233
00:55:17,810 --> 00:55:19,670
about paying your bills,
and so on and so forth,

1234
00:55:19,670 --> 00:55:21,670
which is precisely kind
of what Jose [INAUDIBLE]

1235
00:55:21,670 --> 00:55:23,800
was saying in part.

1236
00:55:23,800 --> 00:55:25,807
Your marginal utility
of money might be high,

1237
00:55:25,807 --> 00:55:27,390
and you're really
concerned with that.

1238
00:55:27,390 --> 00:55:29,840
And you worried about
feeding your children,

1239
00:55:29,840 --> 00:55:31,203
and so on and so forth.

1240
00:55:31,203 --> 00:55:32,870
And it's just really
hard to find a time

1241
00:55:32,870 --> 00:55:35,045
to think about stuff carefully.

1242
00:55:35,045 --> 00:55:36,920
And therefore, you might,
again, sort of just

1243
00:55:36,920 --> 00:55:38,682
go with the default.
Because you just

1244
00:55:38,682 --> 00:55:40,640
don't have the bandwidth--
the mental bandwidth

1245
00:55:40,640 --> 00:55:46,250
to think about what's best
for you and override that.

1246
00:55:46,250 --> 00:55:47,750
So that's an open
question, in fact,

1247
00:55:47,750 --> 00:55:52,250
whether the poor are more likely
to be prone to default effects.

1248
00:55:52,250 --> 00:55:56,570
But I think what's almost
surely the case, or what's

1249
00:55:56,570 --> 00:55:59,510
been shown in various studies,
is that these effects tend

1250
00:55:59,510 --> 00:56:02,630
to have larger effects
among poor, low compensation

1251
00:56:02,630 --> 00:56:04,820
workers, which is precisely
one of the reasons

1252
00:56:04,820 --> 00:56:08,840
why they were so
popular overall.

1253
00:56:08,840 --> 00:56:12,230
OK, so then in
addition, not only

1254
00:56:12,230 --> 00:56:15,770
are there large effects on the
contribution decision overall.

1255
00:56:15,770 --> 00:56:18,890
It's like, do you want your the
participation decision overall?

1256
00:56:18,890 --> 00:56:21,170
Do you want to
participate versus not?

1257
00:56:21,170 --> 00:56:24,720
Like, do you want to contribute
any money versus not?

1258
00:56:24,720 --> 00:56:27,650
But also there seem to be large
effects on the contribution

1259
00:56:27,650 --> 00:56:28,740
rate.

1260
00:56:28,740 --> 00:56:31,250
That is to say, I
think I told you--

1261
00:56:31,250 --> 00:56:35,037
remember, I told you the default
contribution rate was 3%.

1262
00:56:35,037 --> 00:56:36,620
So if you were
automatically enrolled,

1263
00:56:36,620 --> 00:56:42,200
the company would just give
you a 3% contribution standard.

1264
00:56:42,200 --> 00:56:45,810
You could change that to 4%,
2%, 7%, whatever you want to do.

1265
00:56:45,810 --> 00:56:48,230
But the default would be 3%.

1266
00:56:48,230 --> 00:56:51,220
And you see, essentially,
the majority of people--

1267
00:56:51,220 --> 00:56:53,780
the fraction of employees
in the new cohort,

1268
00:56:53,780 --> 00:56:58,253
which is the automatic
enrollment, chooses 3%.

1269
00:56:58,253 --> 00:57:00,420
There's another thing that's
quite interesting here.

1270
00:57:00,420 --> 00:57:03,860
What you see is that
among people who

1271
00:57:03,860 --> 00:57:05,462
arguably would have chosen--

1272
00:57:05,462 --> 00:57:07,045
if you just look at
the window cohort,

1273
00:57:07,045 --> 00:57:09,530
there's some people, for
example, who chose 6%.

1274
00:57:09,530 --> 00:57:12,800
11% of people in
the window cohort--

1275
00:57:12,800 --> 00:57:15,980
that's the cohort that was
just before the automatic

1276
00:57:15,980 --> 00:57:17,030
enrollment--

1277
00:57:17,030 --> 00:57:21,470
11% of people chose 6%.

1278
00:57:21,470 --> 00:57:24,590
That fraction actually
goes down to 7%.

1279
00:57:24,590 --> 00:57:28,640
That is to say, some people
would have chosen more than 3%

1280
00:57:28,640 --> 00:57:30,560
in the absence of
automatic enrollment,

1281
00:57:30,560 --> 00:57:33,055
now they are moving down to 3%.

1282
00:57:33,055 --> 00:57:34,430
So there is some
effect-- there's

1283
00:57:34,430 --> 00:57:37,855
a bunch of mass coming
from 0% moving to 3%.

1284
00:57:37,855 --> 00:57:39,230
But there's also
some mass coming

1285
00:57:39,230 --> 00:57:41,000
from here that moves
from, essentially,

1286
00:57:41,000 --> 00:57:43,910
positive amounts down to 3%.

1287
00:57:43,910 --> 00:57:46,250
So there's sort of very
heterogeneous effect.

1288
00:57:46,250 --> 00:57:47,840
Some people, or
quite a few people,

1289
00:57:47,840 --> 00:57:50,840
increased their savings from
very low rates, or 0 rates,

1290
00:57:50,840 --> 00:57:51,393
to 3%.

1291
00:57:51,393 --> 00:57:53,810
But there's always another
other effect that, essentially,

1292
00:57:53,810 --> 00:57:56,630
some people who would
have sat down and done

1293
00:57:56,630 --> 00:57:58,730
the choice themselves,
they would have probably

1294
00:57:58,730 --> 00:58:01,410
chosen something like
6%, 7%, whatever,

1295
00:58:01,410 --> 00:58:05,570
even 15%, and some of
them have moved down to 3%

1296
00:58:05,570 --> 00:58:07,680
by the automatic enrollment.

1297
00:58:07,680 --> 00:58:09,960
So there's some negative
effect on savings

1298
00:58:09,960 --> 00:58:13,350
coming from that, potentially.

1299
00:58:13,350 --> 00:58:18,860
In addition, the asset
allocation, here,

1300
00:58:18,860 --> 00:58:23,410
people essentially predominantly
just choose the default.

1301
00:58:23,410 --> 00:58:27,450
Here you see, essentially,
the assets or stocks, bonds,

1302
00:58:27,450 --> 00:58:28,420
and money market.

1303
00:58:28,420 --> 00:58:31,190
Think about money market
essentially as like a 0%--

1304
00:58:31,190 --> 00:58:32,720
or a close to 0%--

1305
00:58:32,720 --> 00:58:36,370
very safe option.

1306
00:58:36,370 --> 00:58:39,740
And in the new cohort,
80% of people--

1307
00:58:39,740 --> 00:58:42,680
I'm not sure how well you can
see this-- but 80% of people

1308
00:58:42,680 --> 00:58:44,660
choose the money market,
which essentially

1309
00:58:44,660 --> 00:58:47,960
is exactly what the company
is choosing for them.

1310
00:58:47,960 --> 00:58:50,090
Well, if you have--

1311
00:58:50,090 --> 00:58:51,710
in the other
cohorts, this is all

1312
00:58:51,710 --> 00:58:53,390
conditional on participating.

1313
00:58:53,390 --> 00:58:59,170
The fraction who chooses stocks
and bonds is way, way higher.

1314
00:58:59,170 --> 00:59:02,580
And as we discussed,
the long-run return

1315
00:59:02,580 --> 00:59:05,730
to investing in the money
market is way, way lower.

1316
00:59:05,730 --> 00:59:10,350
So that's potentially,
again, another reason why--

1317
00:59:10,350 --> 00:59:12,660
in fact, why defaults
aren't increasing

1318
00:59:12,660 --> 00:59:15,420
the faction of people that
are choosing to save overall,

1319
00:59:15,420 --> 00:59:17,710
or to participate on
the savings overall.

1320
00:59:17,710 --> 00:59:22,330
The long-run savings might
not necessarily be high.

1321
00:59:22,330 --> 00:59:23,920
Now, let me summarize this.

1322
00:59:23,920 --> 00:59:26,290
I think I said,
already, most of this.

1323
00:59:26,290 --> 00:59:29,858
So 40% to 50% of individuals
are passive savers,

1324
00:59:29,858 --> 00:59:31,150
as you might want to call them.

1325
00:59:31,150 --> 00:59:33,170
They follow the default plan.

1326
00:59:33,170 --> 00:59:35,680
They essentially-- both in
terms of the participation,

1327
00:59:35,680 --> 00:59:37,872
but also on the contribution
rate and the asset

1328
00:59:37,872 --> 00:59:39,580
allocation-- they
essentially do whatever

1329
00:59:39,580 --> 00:59:43,596
the company chooses for them.

1330
00:59:43,596 --> 00:59:46,880
The suggestive choice
is not very attractive

1331
00:59:46,880 --> 00:59:49,580
unless it's the default.
Notice-- remember,

1332
00:59:49,580 --> 00:59:50,510
the window cohort.

1333
00:59:50,510 --> 00:59:51,927
The window cohort
were essentially

1334
00:59:51,927 --> 00:59:55,660
the cohort that now could just
do whatever they wanted to do.

1335
00:59:55,660 --> 01:00:00,760
They were told on the 1st of
April, I think, '97 or '98,

1336
01:00:00,760 --> 01:00:03,552
now you can invest immediately.

1337
01:00:03,552 --> 01:00:04,760
You can do whatever you want.

1338
01:00:04,760 --> 01:00:07,163
We're not choosing for you,
but you are now eligible,

1339
01:00:07,163 --> 01:00:09,580
even for people who have been
at the company for something

1340
01:00:09,580 --> 01:00:11,538
like, say, six months,
three months, and so on.

1341
01:00:11,538 --> 01:00:14,230
The company just made
the participation

1342
01:00:14,230 --> 01:00:15,550
available to anybody.

1343
01:00:15,550 --> 01:00:17,140
Since they had been
hired previously,

1344
01:00:17,140 --> 01:00:18,658
they didn't default people in.

1345
01:00:18,658 --> 01:00:20,950
They just made it, like, you
can change your choice now

1346
01:00:20,950 --> 01:00:21,920
if you wanted.

1347
01:00:21,920 --> 01:00:23,830
But among those people,
the window cohort

1348
01:00:23,830 --> 01:00:25,870
is very much looking
like the old cohort.

1349
01:00:25,870 --> 01:00:28,478
They're not sort of following
the company [INAUDIBLE]..

1350
01:00:28,478 --> 01:00:30,770
You might say, well, the
company has a new default now.

1351
01:00:30,770 --> 01:00:33,073
And the new default,
that's a recommendation,

1352
01:00:33,073 --> 01:00:34,990
so I should just follow
that default because I

1353
01:00:34,990 --> 01:00:36,290
don't know what to do.

1354
01:00:36,290 --> 01:00:37,840
That's not what
people seem to do.

1355
01:00:37,840 --> 01:00:39,520
So it's not like
that-- so there doesn't

1356
01:00:39,520 --> 01:00:44,340
seem to be that the default is a
perceived choice that's suggest

1357
01:00:44,340 --> 01:00:46,990
[INAUDIBLE] company, because the
window cohort really does not

1358
01:00:46,990 --> 01:00:50,530
follow this default or
the suggested default.

1359
01:00:50,530 --> 01:00:54,023
The window cohort really
looks like the old cohort.

1360
01:00:54,023 --> 01:00:55,440
Now, I only showed
you the results

1361
01:00:55,440 --> 01:00:57,080
for one company for
Madrian and Shea.

1362
01:00:57,080 --> 01:01:00,030
But it's a very robust result
that, if you look at the paper

1363
01:01:00,030 --> 01:01:01,140
by Choi et al.

1364
01:01:01,140 --> 01:01:04,380
that looks at this in more
detail, that finds this

1365
01:01:04,380 --> 01:01:06,560
over and over again in
different companies as well,

1366
01:01:06,560 --> 01:01:10,523
default effects tend to be
very strong in those companies.

1367
01:01:10,523 --> 01:01:11,940
Now one question
you might ask is,

1368
01:01:11,940 --> 01:01:14,400
what's explaining-- what's
the underlying reason why

1369
01:01:14,400 --> 01:01:17,730
these default effects
are so strong?

1370
01:01:17,730 --> 01:01:19,230
There are potential
candidates which

1371
01:01:19,230 --> 01:01:22,110
are awareness, implicit
endorsement by the company,

1372
01:01:22,110 --> 01:01:26,310
inattention/memory, or
potentially present bias.

1373
01:01:26,310 --> 01:01:29,370
And what researchers
found, it's sort of not

1374
01:01:29,370 --> 01:01:30,630
necessarily conclusive.

1375
01:01:30,630 --> 01:01:33,090
But I think the suggestive
evidence is very much that--

1376
01:01:33,090 --> 01:01:34,920
and particularly from
Blumenstock et al.,

1377
01:01:34,920 --> 01:01:37,470
that we think that
present bias might

1378
01:01:37,470 --> 01:01:41,430
be a quite important
reason, in addition

1379
01:01:41,430 --> 01:01:43,710
to the cognitive
costs of thinking

1380
01:01:43,710 --> 01:01:45,760
through different
savings scenarios.

1381
01:01:45,760 --> 01:01:47,820
So people essentially
tend to push this off

1382
01:01:47,820 --> 01:01:50,250
and say, oh, I'm going to
change it in the future.

1383
01:01:50,250 --> 01:01:51,970
At some point, I'll get to it.

1384
01:01:51,970 --> 01:01:54,000
I'll get to it when
it's important.

1385
01:01:54,000 --> 01:01:56,520
Notice that it
requires some naiveté,

1386
01:01:56,520 --> 01:01:59,460
because as we discussed,
if you were sophisticated,

1387
01:01:59,460 --> 01:02:02,400
you might of push it off
maybe a period or two.

1388
01:02:02,400 --> 01:02:05,070
But you would never push it
off by 10 years, or five years,

1389
01:02:05,070 --> 01:02:08,490
because you know you're
going to lose a lot of money.

1390
01:02:08,490 --> 01:02:11,130
But if you have some present
bias plus some naiveté,

1391
01:02:11,130 --> 01:02:14,940
you might say, well, I'm going
to do it in the future and not

1392
01:02:14,940 --> 01:02:16,380
right now.

1393
01:02:16,380 --> 01:02:21,240
And that seems to fit the
data the best, in part

1394
01:02:21,240 --> 01:02:24,360
because papers such as the
one by Blumenstock et al.

1395
01:02:24,360 --> 01:02:26,192
have ruled out some
other explanation,

1396
01:02:26,192 --> 01:02:28,025
for example, the
inattention/memory stories.

1397
01:02:28,025 --> 01:02:29,280
They set reminders.

1398
01:02:29,280 --> 01:02:31,020
That didn't seem
to have an effect.

1399
01:02:31,020 --> 01:02:34,488
They also could rule out
the implicit endorsement,

1400
01:02:34,488 --> 01:02:36,405
because they essentially
explicitly randomized

1401
01:02:36,405 --> 01:02:39,660
the people who were defaulted
in versus defaulted out.

1402
01:02:39,660 --> 01:02:41,070
That also had effects.

1403
01:02:41,070 --> 01:02:42,630
So it doesn't seem to be that.

1404
01:02:42,630 --> 01:02:44,940
They also made
people very clearly

1405
01:02:44,940 --> 01:02:48,540
aware of the different
retirement savings options.

1406
01:02:48,540 --> 01:02:51,710
And again, that didn't
seem to make a big effect.

1407
01:02:51,710 --> 01:02:54,300
So while we don't have
conclusive evidence that

1408
01:02:54,300 --> 01:02:58,020
present bias is sort of
causing these default defects,

1409
01:02:58,020 --> 01:03:00,057
in the retirement
savings choice scheme,

1410
01:03:00,057 --> 01:03:02,640
perhaps if I had to guess, or
if I had to say which I think is

1411
01:03:02,640 --> 01:03:05,020
the most likely explanation,
we think that present bias

1412
01:03:05,020 --> 01:03:11,720
and naiveté seem to be the
best explanations that we have

1413
01:03:11,720 --> 01:03:14,240
so far.

1414
01:03:14,240 --> 01:03:18,140
So now, one question you
might sort of say is, well,

1415
01:03:18,140 --> 01:03:20,565
is automatic enrollment optimal?

1416
01:03:20,565 --> 01:03:22,190
I showed you,
essentially, that once we

1417
01:03:22,190 --> 01:03:25,910
do automatic enrollment, people
are more likely to participate.

1418
01:03:25,910 --> 01:03:29,650
However, while they're
more likely to participate,

1419
01:03:29,650 --> 01:03:34,780
also these defaults seem to make
some people save less, right?

1420
01:03:34,780 --> 01:03:36,280
We saw some people
going essentially

1421
01:03:36,280 --> 01:03:38,370
from 6% participation--

1422
01:03:38,370 --> 01:03:41,150
contribution rate to
3% contribution rate.

1423
01:03:41,150 --> 01:03:45,200
They also have a much more
conservative asset allocation.

1424
01:03:45,200 --> 01:03:46,870
So in the long run,
it's not obvious,

1425
01:03:46,870 --> 01:03:49,930
in the case of this company,
whether people are actually

1426
01:03:49,930 --> 01:03:52,330
saving more.

1427
01:03:52,330 --> 01:03:55,220
Now, so how would we
answer this question?

1428
01:03:55,220 --> 01:03:57,740
We're going to get back to
this when we talk about policy.

1429
01:03:57,740 --> 01:04:00,520
But how should we
set the default?

1430
01:04:00,520 --> 01:04:01,990
What should we do?

1431
01:04:01,990 --> 01:04:03,850
What are the considerations
that we have?

1432
01:04:03,850 --> 01:04:06,175
And how can we decide
what's good for people?

1433
01:04:06,175 --> 01:04:10,353
You can essentially force
people to make a choice.

1434
01:04:10,353 --> 01:04:12,520
There were some companies
where, essentially, people

1435
01:04:12,520 --> 01:04:15,220
were very, very strongly
encouraged to-- essentially,

1436
01:04:15,220 --> 01:04:15,980
here's a form.

1437
01:04:15,980 --> 01:04:16,480
Fill it out.

1438
01:04:16,480 --> 01:04:17,590
Tell me what you want.

1439
01:04:17,590 --> 01:04:19,510
And was sort of part
of the hiring package,

1440
01:04:19,510 --> 01:04:21,970
and essentially was
sort of arguing--

1441
01:04:21,970 --> 01:04:23,470
I don't think that's
actually legal.

1442
01:04:23,470 --> 01:04:26,550
But it was essentially sort of
implying that one [INAUDIBLE]

1443
01:04:26,550 --> 01:04:28,580
to doing this form before
you can start working,

1444
01:04:28,580 --> 01:04:31,033
and otherwise, you can't
start working at this company.

1445
01:04:31,033 --> 01:04:32,950
And now, we can look at,
like, what are people

1446
01:04:32,950 --> 01:04:35,620
actually choosing when
everybody chooses actively?

1447
01:04:35,620 --> 01:04:38,360
And that's presumably when
people are paying attention,

1448
01:04:38,360 --> 01:04:41,080
or maybe people do the best
they can for their choices,

1449
01:04:41,080 --> 01:04:43,150
and they're not
automatically enrolled

1450
01:04:43,150 --> 01:04:46,250
either in 0%, or 3%,
or other percentages.

1451
01:04:46,250 --> 01:04:48,590
And so this paper
for by Carroll et al.

1452
01:04:48,590 --> 01:04:50,330
looks attractive choices.

1453
01:04:50,330 --> 01:04:52,190
And when people
choose actively, they

1454
01:04:52,190 --> 01:04:55,310
look a lot like the new
cohort in Madrian and Shea.

1455
01:04:55,310 --> 01:04:56,810
So essentially,
that's sort of seems

1456
01:04:56,810 --> 01:05:00,920
to say that the default
did in fact-- the default

1457
01:05:00,920 --> 01:05:05,210
into automatic enrollment
seemed to have alleviated

1458
01:05:05,210 --> 01:05:07,460
undersavings.

1459
01:05:07,460 --> 01:05:08,960
Because when people
choose actively,

1460
01:05:08,960 --> 01:05:11,240
they choose a lot-- look
a lot like the new cohort

1461
01:05:11,240 --> 01:05:12,530
compared to the old cohort.

1462
01:05:12,530 --> 01:05:14,480
It seems to be that
people, when they're

1463
01:05:14,480 --> 01:05:17,290
defaulted in the no savings
option, they kind of maybe

1464
01:05:17,290 --> 01:05:20,670
think, oh, I'll do it in the
future, and so on and so forth.

1465
01:05:20,670 --> 01:05:23,138
But then they never
sort of get to it.

1466
01:05:23,138 --> 01:05:25,430
Now, when you actually ask
them actually what you want,

1467
01:05:25,430 --> 01:05:29,820
it seems like people are
mostly wanting to enroll.

1468
01:05:29,820 --> 01:05:31,650
Now, one thing that
is quite interesting

1469
01:05:31,650 --> 01:05:35,430
here is that it seems like the
default effects seem to mostly

1470
01:05:35,430 --> 01:05:36,870
disappear after three years.

1471
01:05:36,870 --> 01:05:41,450
That's true for
quite a few studies.

1472
01:05:41,450 --> 01:05:43,008
However, even if
that's the case,

1473
01:05:43,008 --> 01:05:44,800
well, there's no catch-up
in levels, right?

1474
01:05:44,800 --> 01:05:49,800
So if you start contributing
after three years, while you

1475
01:05:49,800 --> 01:05:51,390
miss out on three
years of savings,

1476
01:05:51,390 --> 01:05:53,190
you're never going
to catch up overall.

1477
01:05:53,190 --> 01:05:57,330
Moreover, people tend to
change their employers

1478
01:05:57,330 --> 01:05:59,730
very frequently,
particularly in the US.

1479
01:05:59,730 --> 01:06:01,590
So if it takes you only--

1480
01:06:01,590 --> 01:06:04,620
if it takes you three years to
sign up for retirement savings

1481
01:06:04,620 --> 01:06:06,870
every time you
switch your employer,

1482
01:06:06,870 --> 01:06:09,940
you're going to
not save very much.

1483
01:06:09,940 --> 01:06:12,210
Chetty et al had an example
where they find, in fact,

1484
01:06:12,210 --> 01:06:16,080
very clear effects on long-run
savings of these types

1485
01:06:16,080 --> 01:06:18,400
of default [INAUDIBLE].

1486
01:06:18,400 --> 01:06:21,962
Now, let me tell you a
little bit now about--

1487
01:06:21,962 --> 01:06:23,920
I'm going to sort of move
the frames and nudges

1488
01:06:23,920 --> 01:06:26,462
until next time, which I think
should be fine because there's

1489
01:06:26,462 --> 01:06:28,150
not that much left anyway.

1490
01:06:28,150 --> 01:06:30,640
But let me tell you a little
bit about a cautionary tale

1491
01:06:30,640 --> 01:06:32,650
about defaults.

1492
01:06:32,650 --> 01:06:35,250
This is a paper by-- a very
short paper by Cronqvist

1493
01:06:35,250 --> 01:06:38,080
and Thaler that looks
at the privatization

1494
01:06:38,080 --> 01:06:40,150
of social security in Sweden.

1495
01:06:40,150 --> 01:06:42,170
So they had 456 funds.

1496
01:06:42,170 --> 01:06:45,580
And there was one default
fund chosen by the government.

1497
01:06:45,580 --> 01:06:48,220
And in 2000, this
default fund was

1498
01:06:48,220 --> 01:06:50,410
discouraged with a massive
marketing campaign.

1499
01:06:50,410 --> 01:06:53,020
Essentially people said, like,
don't choose the default fund.

1500
01:06:53,020 --> 01:06:54,490
Just pick for yourself.

1501
01:06:54,490 --> 01:06:56,810
Make the choice that's
the best for you.

1502
01:06:56,810 --> 01:06:59,230
And then in 2003, this
marketing campaign ended.

1503
01:06:59,230 --> 01:07:01,330
And now essentially,
the majority, again,

1504
01:07:01,330 --> 01:07:08,130
sort of chose the
default account or fund.

1505
01:07:08,130 --> 01:07:12,600
Now, as it happens, the
portfolios actively chosen

1506
01:07:12,600 --> 01:07:16,750
in 2000 did a lot
worse than the default.

1507
01:07:16,750 --> 01:07:21,347
So people, essentially,
were choosing on their own.

1508
01:07:21,347 --> 01:07:23,430
And in part, I think because
they're just unlucky,

1509
01:07:23,430 --> 01:07:26,790
because it happened to be at the
wrong time and the wrong place,

1510
01:07:26,790 --> 01:07:28,770
chose sort of very risky assets.

1511
01:07:28,770 --> 01:07:31,390
And the default tended to be
somewhat more conservative.

1512
01:07:31,390 --> 01:07:34,630
So people lost a lot of money.

1513
01:07:34,630 --> 01:07:36,270
So now, that's the
more general issue

1514
01:07:36,270 --> 01:07:39,600
is that active choice is less
attractive if consumers are

1515
01:07:39,600 --> 01:07:41,370
less financially sophisticated.

1516
01:07:41,370 --> 01:07:43,170
Again, in the Cronqvist
and Thaler case,

1517
01:07:43,170 --> 01:07:44,378
maybe they were just unlucky.

1518
01:07:44,378 --> 01:07:46,710
But in general, if you want
people to choose actively

1519
01:07:46,710 --> 01:07:48,335
and they don't know
what they're doing,

1520
01:07:48,335 --> 01:07:51,235
well, then they're going
to be potentially worse.

1521
01:07:51,235 --> 01:07:52,860
There is similar
evidence, for example,

1522
01:07:52,860 --> 01:07:55,027
for health care insurance
choices, where people just

1523
01:07:55,027 --> 01:07:57,730
don't know what health insurance
choice is best for them.

1524
01:07:57,730 --> 01:08:00,240
So you can give
them more options,

1525
01:08:00,240 --> 01:08:04,183
they might be potentially
doing worse, because they just

1526
01:08:04,183 --> 01:08:05,100
don't know what to do.

1527
01:08:05,100 --> 01:08:06,933
And then they choose
options that are worse,

1528
01:08:06,933 --> 01:08:09,610
and they're more likely, for
example, to be exploited.

1529
01:08:09,610 --> 01:08:12,840
And then, having a default
choice that is perhaps not

1530
01:08:12,840 --> 01:08:15,450
optimal for everybody, but it's
pretty conservative and pretty

1531
01:08:15,450 --> 01:08:19,104
good for most people, might
be, in fact, quite valuable.

1532
01:08:19,104 --> 01:08:20,479
And so now, when
you think about,

1533
01:08:20,479 --> 01:08:25,720
like, optimal
decision regimes, when

1534
01:08:25,720 --> 01:08:27,910
you think of an active
choice versus default,

1535
01:08:27,910 --> 01:08:32,270
consumer heterogeneity makes
active choice more attractive.

1536
01:08:32,270 --> 01:08:35,770
If you think people want
very different things,

1537
01:08:35,770 --> 01:08:38,040
and they know about
what's best for them,

1538
01:08:38,040 --> 01:08:39,580
and we don't have
that information--

1539
01:08:39,580 --> 01:08:42,130
we don't know what's good for
some people versus others--

1540
01:08:42,130 --> 01:08:44,290
well, it's best to let
people choose on their own.

1541
01:08:44,290 --> 01:08:45,998
Let's let them pick
what's good for them,

1542
01:08:45,998 --> 01:08:49,410
and then they'll be happy.

1543
01:08:49,410 --> 01:08:52,335
But active choice only
improves the outcome

1544
01:08:52,335 --> 01:08:54,210
if people actually know
what's good for them.

1545
01:08:54,210 --> 01:08:55,710
If everybody is
confused and doesn't

1546
01:08:55,710 --> 01:08:57,359
know what they're
doing, well then,

1547
01:08:57,359 --> 01:09:00,180
letting people choose
actively will potentially

1548
01:09:00,180 --> 01:09:03,380
make things worse for
quite a few people.

1549
01:09:03,380 --> 01:09:05,880
And so you want to make sure
that defaults don't make people

1550
01:09:05,880 --> 01:09:06,450
worse off.

1551
01:09:06,450 --> 01:09:07,658
This is what Jose was saying.

1552
01:09:07,658 --> 01:09:09,600
In particular, people
might sort of oversave

1553
01:09:09,600 --> 01:09:11,758
if you default them
into high savings.

1554
01:09:11,758 --> 01:09:13,800
And then they have credit
card debt or other sort

1555
01:09:13,800 --> 01:09:15,680
of problems arising from that.

1556
01:09:15,680 --> 01:09:17,790
So we want to be quite careful.

1557
01:09:17,790 --> 01:09:21,240
One option one can try and do
is try to provide information

1558
01:09:21,240 --> 01:09:23,330
sessions and make sure
people know, here's--

1559
01:09:23,330 --> 01:09:26,609
or trying to sort of provide
a curated set of three, four

1560
01:09:26,609 --> 01:09:28,350
options that we think
are good options

1561
01:09:28,350 --> 01:09:31,439
for different types of people,
provide information for them,

1562
01:09:31,439 --> 01:09:34,560
make sure they
understand these options,

1563
01:09:34,560 --> 01:09:38,013
and that then people
choose on their own.

1564
01:09:38,013 --> 01:09:40,430
Another popular alternative--
I'm going to talk about this

1565
01:09:40,430 --> 01:09:42,170
a little bit when we
talk about policy--

1566
01:09:42,170 --> 01:09:43,878
which is called the
smart plan, or, like,

1567
01:09:43,878 --> 01:09:47,330
auto-escalation plans, which are
essentially more like automatic

1568
01:09:47,330 --> 01:09:55,310
raises in the future,
where you essentially

1569
01:09:55,310 --> 01:09:57,200
can commit to
automatically raising

1570
01:09:57,200 --> 01:10:00,500
your savings in the future.

1571
01:10:00,500 --> 01:10:03,590
You say, essentially, every
time I get a pay raise, 1%,

1572
01:10:03,590 --> 01:10:05,240
or some fraction
of that pay raise,

1573
01:10:05,240 --> 01:10:07,880
or one percentage point of my
earnings of that pay raise,

1574
01:10:07,880 --> 01:10:11,070
will go through my savings.

1575
01:10:11,070 --> 01:10:14,690
So then, you don't
have to reduce

1576
01:10:14,690 --> 01:10:18,770
your consumption or your salary
right now that you receive,

1577
01:10:18,770 --> 01:10:19,970
but in the future.

1578
01:10:19,970 --> 01:10:20,860
That's good.

1579
01:10:20,860 --> 01:10:23,870
In particular, it addresses
present bias and loss aversion.

1580
01:10:23,870 --> 01:10:26,810
Because people are really averse
to reducing their paycheck

1581
01:10:26,810 --> 01:10:30,485
right now, and for present
bias and loss aversion reasons.

1582
01:10:30,485 --> 01:10:32,180
When I was asking
about information

1583
01:10:32,180 --> 01:10:34,760
versus financial education,
I mean kind of both of those.

1584
01:10:34,760 --> 01:10:38,150
You can find information about
what are these options, what's

1585
01:10:38,150 --> 01:10:41,370
good about this investment
option versus another,

1586
01:10:41,370 --> 01:10:43,610
and so providing
financial education,

1587
01:10:43,610 --> 01:10:48,470
but in particular, financial
education plus active choice.

1588
01:10:48,470 --> 01:10:50,540
Essentially, by
active choice, I mean

1589
01:10:50,540 --> 01:10:52,580
pushing people really
to choose, and saying,

1590
01:10:52,580 --> 01:10:53,805
look, what do you want?

1591
01:10:53,805 --> 01:10:57,260
Do you want option A,
which is no retirement

1592
01:10:57,260 --> 01:11:01,520
savings, or option B, which is
like 3%, or 6%, or whatever.

1593
01:11:01,520 --> 01:11:03,650
The current format is more--

1594
01:11:03,650 --> 01:11:06,110
essentially, is,
like, no information,

1595
01:11:06,110 --> 01:11:09,030
and the default is
essentially 0 savings.

1596
01:11:09,030 --> 01:11:11,270
And then, you have to
actively choose on your own

1597
01:11:11,270 --> 01:11:11,910
as an employer.

1598
01:11:11,910 --> 01:11:14,660
And many employees--
sorry, as an employee.

1599
01:11:14,660 --> 01:11:18,410
Many employees just don't know
what to do, and get confused.

1600
01:11:18,410 --> 01:11:20,660
And then they get frazzled,
and then procrastinate it,

1601
01:11:20,660 --> 01:11:24,240
and then never do it, perhaps
in part because of confusion.

1602
01:11:24,240 --> 01:11:28,130
OK, and then other settings
have enormous before the fact.

1603
01:11:28,130 --> 01:11:30,200
This is, for example,
true for organ donations.

1604
01:11:30,200 --> 01:11:33,920
I talked about this already in
the first or second lecture.

1605
01:11:33,920 --> 01:11:39,080
There's some other settings
such as voter registration,

1606
01:11:39,080 --> 01:11:41,100
green energy options, and so on.

1607
01:11:41,100 --> 01:11:44,840
So default effects can be
really, really powerful.

1608
01:11:44,840 --> 01:11:47,840
But one wants to be quite
careful in using them.

1609
01:11:47,840 --> 01:11:50,870
I'm going to talk
about this a little bit

1610
01:11:50,870 --> 01:11:52,900
at the end of this
lecture next time.

1611
01:11:52,900 --> 01:11:54,090
Let me stop here.

1612
01:11:54,090 --> 01:12:04,366
Let me just tell you what to
read, which is-- one second--

1613
01:12:04,366 --> 01:12:09,610
which is the paper by Ariely et
al., sections one through four.

1614
01:12:09,610 --> 01:12:13,390
You can read more if you want,
but section one through four

1615
01:12:13,390 --> 01:12:14,740
already gets you pretty far.

1616
01:12:14,740 --> 01:12:17,770
This is really a fun
paper that questions

1617
01:12:17,770 --> 01:12:21,610
a lot of assumptions of
neoclassical economics,

1618
01:12:21,610 --> 01:12:24,350
makes you wonder where
preferences come from

1619
01:12:24,350 --> 01:12:26,750
and what they really mean.

1620
01:12:26,750 --> 01:12:28,390
So it's a fun read.

1621
01:12:28,390 --> 01:12:31,540
I'll talk about this in
the remaining part of this

1622
01:12:31,540 --> 01:12:34,330
lecture-- which is a bit
on frames and nudges--

1623
01:12:34,330 --> 01:12:37,810
more generally next
time on Wednesday.

1624
01:12:37,810 --> 01:12:39,860
Let me know if you
have any questions.

1625
01:12:39,860 --> 01:12:41,730
Thank you so much.