1
00:00:00,040 --> 00:00:02,460
The following content is
provided under a Creative

2
00:00:02,460 --> 00:00:03,870
Commons license.

3
00:00:03,870 --> 00:00:06,910
Your support will help MIT
OpenCourseWare continue to

4
00:00:06,910 --> 00:00:10,560
offer high quality educational
resources for free.

5
00:00:10,560 --> 00:00:13,460
To make a donation or view
additional materials from

6
00:00:13,460 --> 00:00:16,180
hundreds of MIT courses, visit
mitopencourseware@ocw.mit.edu.

7
00:00:22,340 --> 00:00:22,780
PROFESSOR: All right.

8
00:00:22,780 --> 00:00:27,130
So today we are going to start
by reviewing income and

9
00:00:27,130 --> 00:00:28,690
substitution effects.

10
00:00:28,690 --> 00:00:31,110
Because that's a pretty hard
concept and pretty central to

11
00:00:31,110 --> 00:00:33,160
a lot of what we'll do for
the rest of the semester.

12
00:00:33,160 --> 00:00:35,750
And then we're going to dive
in and talk about an

13
00:00:35,750 --> 00:00:37,630
application, a more interesting
application, of

14
00:00:37,630 --> 00:00:40,750
income and substitution effects
which is the effects

15
00:00:40,750 --> 00:00:43,100
of wages on labor supply.

16
00:00:43,100 --> 00:00:44,050
So let's review.

17
00:00:44,050 --> 00:00:46,850
If you take the handout, grab
the handout and look at the

18
00:00:46,850 --> 00:00:48,720
first figure, it's the same
as the last figure of the

19
00:00:48,720 --> 00:00:51,200
previous lecture.

20
00:00:51,200 --> 00:00:54,520
To review, remember, whenever
the price changes, a price

21
00:00:54,520 --> 00:00:59,170
change can be decomposed into
two effects, the substitution

22
00:00:59,170 --> 00:01:01,980
effect and the income effect.

23
00:01:01,980 --> 00:01:05,690
The substitution effect is the
change in the quantity

24
00:01:05,690 --> 00:01:08,430
demanded when the
price changes,

25
00:01:08,430 --> 00:01:11,300
holding utility constant.

26
00:01:11,300 --> 00:01:13,370
And as we proved last
time, that is always

27
00:01:13,370 --> 00:01:16,210
negative, 0 or negative.

28
00:01:16,210 --> 00:01:18,750
It is always non-positive.

29
00:01:18,750 --> 00:01:22,670
It's always true that when
a price goes up, the

30
00:01:22,670 --> 00:01:23,730
substitution effect
is negative.

31
00:01:23,730 --> 00:01:25,760
We proved that both
mathematically and graphically

32
00:01:25,760 --> 00:01:29,790
last time showing that if you're
going to hold utility

33
00:01:29,790 --> 00:01:32,120
constant, and the price of a
good is going to go up, you're

34
00:01:32,120 --> 00:01:34,040
going to shift away
from that good.

35
00:01:34,040 --> 00:01:34,680
OK.

36
00:01:34,680 --> 00:01:35,640
That's the substitution
effect.

37
00:01:35,640 --> 00:01:39,410
In our example, we showed
graphically how you measure a

38
00:01:39,410 --> 00:01:40,850
substitution effect.

39
00:01:40,850 --> 00:01:45,660
You draw a new imaginary budget
constraint, BC3, which

40
00:01:45,660 --> 00:01:50,130
is parallel to the new budget
constraint, BC2.

41
00:01:50,130 --> 00:01:53,710
So it's got the new price ratio
but tangent to the old

42
00:01:53,710 --> 00:01:55,270
indifference curve.

43
00:01:55,270 --> 00:01:57,295
So the key thing to understand
is the imaginary budget

44
00:01:57,295 --> 00:01:59,950
constraint, BC3, where
it comes from.

45
00:01:59,950 --> 00:02:01,390
It's parallel to the new
budget constraint.

46
00:02:01,390 --> 00:02:03,360
That is it's got the new
marginal rate of

47
00:02:03,360 --> 00:02:07,920
transformation, the new slope,
but it's tangent to the old

48
00:02:07,920 --> 00:02:09,135
indifference curve.

49
00:02:09,135 --> 00:02:11,840
That gets you to point B. And so
the movement from A to B is

50
00:02:11,840 --> 00:02:14,620
the substitution effect.

51
00:02:14,620 --> 00:02:18,930
Then we have an income effect
which is, in fact, utility

52
00:02:18,930 --> 00:02:21,410
isn't held constant when
prices change.

53
00:02:21,410 --> 00:02:22,930
In fact, utility falls, because

54
00:02:22,930 --> 00:02:25,980
you're effectively poorer.

55
00:02:25,980 --> 00:02:27,280
You're effectively poorer.

56
00:02:27,280 --> 00:02:28,500
Utility is falling.

57
00:02:28,500 --> 00:02:30,780
And since you're effectively
poorer, that further reduces

58
00:02:30,780 --> 00:02:33,560
demand if the good is normal.

59
00:02:33,560 --> 00:02:37,840
So if it's a normal good, if
it's a good where lower income

60
00:02:37,840 --> 00:02:40,260
causes less consumption of
it, the fact that you're

61
00:02:40,260 --> 00:02:42,590
effectively poorer further
lowers the consumption from

62
00:02:42,590 --> 00:02:44,440
point B to point C.

63
00:02:44,440 --> 00:02:46,810
So the total price effect is the
one we demonstrated at the

64
00:02:46,810 --> 00:02:48,270
beginning of the last lecture.

65
00:02:48,270 --> 00:02:54,490
We raised the price of movies
from $8 to $12.

66
00:02:54,490 --> 00:02:56,160
And we saw the number
of movies consumed

67
00:02:56,160 --> 00:02:58,040
fell from 6 to 4.

68
00:02:58,040 --> 00:03:00,870
But what we can see now to
understand what's underneath

69
00:03:00,870 --> 00:03:04,980
that is two things, an effect of
the fact that prices change

70
00:03:04,980 --> 00:03:08,200
holding utility constant,
and the fact that you're

71
00:03:08,200 --> 00:03:09,450
effectively poorer.

72
00:03:12,060 --> 00:03:13,820
And that's the key thing.

73
00:03:13,820 --> 00:03:15,790
No, your income hasn't
actually gone down.

74
00:03:15,790 --> 00:03:20,940
But that $96 your parents gave
you can buy you less.

75
00:03:20,940 --> 00:03:23,780
Your opportunity set has
been restricted.

76
00:03:23,780 --> 00:03:25,430
And that makes you effectively
poorer.

77
00:03:25,430 --> 00:03:27,250
And so you buy less
for that reason.

78
00:03:27,250 --> 00:03:31,970
And so you get the total
movement from A to C.

79
00:03:31,970 --> 00:03:35,950
Now, as we emphasized last time,
this will be the case if

80
00:03:35,950 --> 00:03:37,380
it's a normal good.

81
00:03:37,380 --> 00:03:38,900
So substitution effects
are done.

82
00:03:38,900 --> 00:03:40,400
Substitute effects are
always negative,

83
00:03:40,400 --> 00:03:41,880
nothing fun about that.

84
00:03:41,880 --> 00:03:44,100
Income effects are a little
more interesting, because

85
00:03:44,100 --> 00:03:47,770
goods can be not normal
but inferior.

86
00:03:47,770 --> 00:03:52,780
We have inferior goods which
are ones such that they're

87
00:03:52,780 --> 00:03:55,030
crummy stuff that as your
income goes up, you

88
00:03:55,030 --> 00:03:56,480
want less of it.

89
00:03:56,480 --> 00:03:58,170
And that can change
the analysis.

90
00:03:58,170 --> 00:04:01,400
So if we look at Figure 7-2,
now we're talking about the

91
00:04:01,400 --> 00:04:03,270
price change with an
inferior good.

92
00:04:03,270 --> 00:04:04,800
And now imagine someone
choosing

93
00:04:04,800 --> 00:04:07,510
between steak and potatoes.

94
00:04:07,510 --> 00:04:08,830
So now the choice is between
steak and potatoes.

95
00:04:12,590 --> 00:04:17,380
And steak costs $5 a pound,
initially, and potatoes cost

96
00:04:17,380 --> 00:04:19,029
$1 a pound.

97
00:04:19,029 --> 00:04:20,839
Initially, you have
an income of $25.

98
00:04:20,839 --> 00:04:29,030
So someone has an income
of y equals $25.

99
00:04:29,030 --> 00:04:34,635
The price of steak is $5, and
the price of potatoes is $1.

100
00:04:37,730 --> 00:04:41,950
So your budget constraint, your
original BC1, runs from

101
00:04:41,950 --> 00:04:44,310
you can either have 5 steak, or
25 potatoes, or something

102
00:04:44,310 --> 00:04:44,840
in between.

103
00:04:44,840 --> 00:04:47,990
And so individuals choose
point A where they're

104
00:04:47,990 --> 00:04:52,410
consuming 8.3 potatoes.

105
00:04:52,410 --> 00:05:00,580
They choose point A.

106
00:05:00,580 --> 00:05:01,860
I don't know what the
number of steaks is.

107
00:05:01,860 --> 00:05:04,290
We probably also ought to label
that, the number of

108
00:05:04,290 --> 00:05:05,110
steaks that comes from that.

109
00:05:05,110 --> 00:05:05,710
But whatever.

110
00:05:05,710 --> 00:05:08,020
It comes out of the
utility function.

111
00:05:08,020 --> 00:05:14,040
So then we say, now let's
imagine that the price of

112
00:05:14,040 --> 00:05:17,210
potatoes rises to $3 a pound.

113
00:05:17,210 --> 00:05:21,900
There's a blight on the potatoes
like there was in

114
00:05:21,900 --> 00:05:24,660
Ireland in the 1800s.

115
00:05:24,660 --> 00:05:28,670
There's a potato blight, and
that shifts in the supply

116
00:05:28,670 --> 00:05:31,870
curve for potatoes raising the
price of potatoes from $1 a

117
00:05:31,870 --> 00:05:35,150
pound to $3 a pound.

118
00:05:35,150 --> 00:05:40,410
Now, what we know is that that
will move consumers, given the

119
00:05:40,410 --> 00:05:42,810
utility function that we've
chosen here, that will move

120
00:05:42,810 --> 00:05:45,910
consumers from point A to point
C. Once again, that's

121
00:05:45,910 --> 00:05:49,780
not labeled, but some lower
amount of potatoes.

122
00:05:49,780 --> 00:05:52,840
That will move them from point
A to point C. So, ultimately,

123
00:05:52,840 --> 00:05:56,240
they'll choose fewer potatoes
and fewer steaks.

124
00:05:56,240 --> 00:05:58,780
But, in fact, what we can see is
that's the composition of a

125
00:05:58,780 --> 00:06:01,030
substitution effect which is
negative, and an income effect

126
00:06:01,030 --> 00:06:02,410
which is positive.

127
00:06:02,410 --> 00:06:04,460
So if we do our standard
decomposition, we draw a new

128
00:06:04,460 --> 00:06:07,530
monetary budget constraint
BC3.

129
00:06:07,530 --> 00:06:10,830
It's parallel to the new budget
constraint, BC2, so the

130
00:06:10,830 --> 00:06:12,480
same price ratio.

131
00:06:12,480 --> 00:06:13,970
It's parallel.

132
00:06:13,970 --> 00:06:18,560
But it's tangent to the old
indifference curve at point B

133
00:06:18,560 --> 00:06:21,810
which is actually to the left
of the ultimate choice at

134
00:06:21,810 --> 00:06:22,860
point C.

135
00:06:22,860 --> 00:06:26,590
So the substitution effect
takes us from A to B. The

136
00:06:26,590 --> 00:06:30,850
income effect actually takes us
back from B to C. That is

137
00:06:30,850 --> 00:06:35,090
as that budget constraint shifts
from BC3 to BC2, as you

138
00:06:35,090 --> 00:06:40,060
get poorer, you choose
more potatoes.

139
00:06:40,060 --> 00:06:42,400
So the substitution effect would
say that from the price

140
00:06:42,400 --> 00:06:44,860
change of potatoes alone, we go
all the way to point B. We

141
00:06:44,860 --> 00:06:47,620
massively reduce our consumption
of potatoes.

142
00:06:47,620 --> 00:06:51,760
But because we're poorer,
effectively, we now consume

143
00:06:51,760 --> 00:06:54,170
more potatoes.

144
00:06:54,170 --> 00:06:56,170
Because we're effectively
poorer, we now

145
00:06:56,170 --> 00:06:57,420
consume more potatoes.

146
00:07:02,270 --> 00:07:06,210
And so, on net, you get
a reduction in potato

147
00:07:06,210 --> 00:07:08,040
consumption.

148
00:07:08,040 --> 00:07:10,350
But it offsets the substitution
effect.

149
00:07:10,350 --> 00:07:11,710
So that's when income effects
can be a little more

150
00:07:11,710 --> 00:07:12,130
interesting.

151
00:07:12,130 --> 00:07:13,590
It's going to be a little more
interesting exercise.

152
00:07:13,590 --> 00:07:15,450
When you think about substitute
effects in the same

153
00:07:15,450 --> 00:07:16,350
way, it's not that
interesting.

154
00:07:16,350 --> 00:07:18,170
It's just look, quantity fell.

155
00:07:18,170 --> 00:07:20,290
It doesn't really matter why.

156
00:07:20,290 --> 00:07:21,390
You don't see, in
the real world,

157
00:07:21,390 --> 00:07:23,340
substitution income effects.

158
00:07:23,340 --> 00:07:25,670
What's interesting is when
they're opposed to each other.

159
00:07:25,670 --> 00:07:27,190
That's when it gets
more interesting.

160
00:07:27,190 --> 00:07:29,550
And so you see this small
reduction you get from the

161
00:07:29,550 --> 00:07:30,800
substitution effect alone.

162
00:07:35,530 --> 00:07:37,140
By the way, there's
two handouts.

163
00:07:37,140 --> 00:07:37,430
Right?

164
00:07:37,430 --> 00:07:38,430
Jessica, is there
two handouts?

165
00:07:38,430 --> 00:07:39,110
There should be.

166
00:07:39,110 --> 00:07:41,460
There's tables as well.

167
00:07:41,460 --> 00:07:42,450
I didn't actually get it.

168
00:07:42,450 --> 00:07:44,490
Jessica, grab me one of those.

169
00:07:44,490 --> 00:07:45,850
There's tables as
well as graphs.

170
00:07:45,850 --> 00:07:48,390
So make sure you have
both handouts.

171
00:07:48,390 --> 00:07:49,360
Anyone else need tables?

172
00:07:49,360 --> 00:07:50,390
Am I the only one who
didn't get it?

173
00:07:50,390 --> 00:07:51,820
OK, good.

174
00:07:51,820 --> 00:07:56,790
So, in principle, the income
effect could be so large it

175
00:07:56,790 --> 00:07:58,200
could offset the substitution
effect.

176
00:07:58,200 --> 00:07:59,170
There's no reason,
theoretically,

177
00:07:59,170 --> 00:08:00,320
that couldn't happen.

178
00:08:00,320 --> 00:08:04,070
That is, in principle, you could
derive preferences such

179
00:08:04,070 --> 00:08:06,650
that the income effect is
so large it offsets the

180
00:08:06,650 --> 00:08:08,540
substitution effect--
thank you--

181
00:08:08,540 --> 00:08:11,980
and the price increase actually
leads to more

182
00:08:11,980 --> 00:08:13,800
potatoes being consumed.

183
00:08:13,800 --> 00:08:15,640
That is what we'd call
a Giffen good as I

184
00:08:15,640 --> 00:08:16,890
talked about last time.

185
00:08:21,830 --> 00:08:25,730
So if you look at the table,
the top table, this sort of

186
00:08:25,730 --> 00:08:29,110
lays out our possibilities.

187
00:08:29,110 --> 00:08:29,840
So look at the top table.

188
00:08:29,840 --> 00:08:32,150
It sort of maps out the possible
sets of things that

189
00:08:32,150 --> 00:08:33,000
can happen.

190
00:08:33,000 --> 00:08:36,240
So if we have a normal good,
and the price of that good

191
00:08:36,240 --> 00:08:39,789
rises, then we know that the

192
00:08:39,789 --> 00:08:42,270
substitution effect is negative.

193
00:08:42,270 --> 00:08:44,250
The income effect is negative.

194
00:08:44,250 --> 00:08:45,430
So the total effect
is negative.

195
00:08:45,430 --> 00:08:46,210
Quantity falls.

196
00:08:46,210 --> 00:08:47,530
That's the law of demand.

197
00:08:47,530 --> 00:08:48,750
We talked about that
last time, downward

198
00:08:48,750 --> 00:08:50,430
sloping demand curves.

199
00:08:50,430 --> 00:08:51,870
Likewise, if the price
falls, the

200
00:08:51,870 --> 00:08:53,910
substitution effect is positive.

201
00:08:53,910 --> 00:08:55,090
The income effect is positive.

202
00:08:55,090 --> 00:08:58,050
You're now richer because the
price of the good fell.

203
00:08:58,050 --> 00:09:00,200
And so, therefore,
demand goes up.

204
00:09:00,200 --> 00:09:02,300
Quantity consumed goes up, once
again, downward sloping

205
00:09:02,300 --> 00:09:03,230
demand curve.

206
00:09:03,230 --> 00:09:04,870
Price rises, you consume
less of it.

207
00:09:04,870 --> 00:09:06,060
Price falls, you consume
more of it.

208
00:09:06,060 --> 00:09:08,460
That's what we learned about
in the first lecture.

209
00:09:08,460 --> 00:09:12,730
However, once goods are
inferior, all bets are off.

210
00:09:12,730 --> 00:09:14,910
Because now the income effect
is the opposite sign of the

211
00:09:14,910 --> 00:09:15,720
substitution effect.

212
00:09:15,720 --> 00:09:18,360
It's possible it could
be larger.

213
00:09:18,360 --> 00:09:20,770
So the total effect
is ambiguous.

214
00:09:20,770 --> 00:09:24,810
You could actually get an upward
sloping demand curve.

215
00:09:24,810 --> 00:09:27,610
You could actually get that a
price rise leads to more of a

216
00:09:27,610 --> 00:09:30,290
good, and a fall leads
to less of a good.

217
00:09:30,290 --> 00:09:31,390
Now will you?

218
00:09:31,390 --> 00:09:32,720
Only if it's a Giffen good.

219
00:09:32,720 --> 00:09:34,610
And, in fact, there's a lot of
controversy in economics about

220
00:09:34,610 --> 00:09:36,500
whether any good in the world
has ever been a Giffen good.

221
00:09:36,500 --> 00:09:39,600
At most, there's maybe one or
two examples people can find.

222
00:09:39,600 --> 00:09:42,760
Even then, it's controversial.

223
00:09:42,760 --> 00:09:45,890
So I think it's fine in life to
assume that demand curves

224
00:09:45,890 --> 00:09:46,850
slope down.

225
00:09:46,850 --> 00:09:49,000
I think, in fact, I don't see
convincing evidence that any

226
00:09:49,000 --> 00:09:51,250
subset or set of goods
are Giffen goods.

227
00:09:51,250 --> 00:09:53,480
I think it's just generally fine
it life to assume demand

228
00:09:53,480 --> 00:09:55,450
curves slope down.

229
00:09:55,450 --> 00:09:57,420
Nonetheless, it's important to
understand this theoretical

230
00:09:57,420 --> 00:10:00,000
possibility even if it's
just theoretical.

231
00:10:00,000 --> 00:10:01,570
Because it's important to
understand income and

232
00:10:01,570 --> 00:10:03,690
substitution effects.

233
00:10:03,690 --> 00:10:04,440
OK.

234
00:10:04,440 --> 00:10:08,710
Questions about that, either
on substitution effect or

235
00:10:08,710 --> 00:10:10,860
price changes?

236
00:10:10,860 --> 00:10:11,450
OK.

237
00:10:11,450 --> 00:10:15,330
So now, armed with that, let's
go onto the more interesting

238
00:10:15,330 --> 00:10:17,095
case which is labor supply.

239
00:10:17,095 --> 00:10:19,710
It's more interesting, because
as I'll come to in a few

240
00:10:19,710 --> 00:10:24,510
minutes, we talk about labor
supply, labor is typically

241
00:10:24,510 --> 00:10:26,090
going to be an inferior good.

242
00:10:26,090 --> 00:10:27,890
So things are going to get a
little more interesting.

243
00:10:27,890 --> 00:10:29,600
So let's talk about that.

244
00:10:29,600 --> 00:10:32,940
So the question you want to ask
here is how hard do folks

245
00:10:32,940 --> 00:10:33,690
decide to work?

246
00:10:33,690 --> 00:10:38,430
How many hours of labor do
folks decide to provide?

247
00:10:38,430 --> 00:10:40,970
As we talked about when we
talked about minimum wage,

248
00:10:40,970 --> 00:10:44,360
just as we all have to decide
between consuming pizza and

249
00:10:44,360 --> 00:10:47,670
consuming movies, or consuming
steak and consuming potatoes,

250
00:10:47,670 --> 00:10:50,380
we also have to decide between
how much labor we're going to

251
00:10:50,380 --> 00:10:52,690
provide and how much we're
going to consume.

252
00:10:52,690 --> 00:10:55,210
The more labor you provide to
the market, the more you

253
00:10:55,210 --> 00:10:58,170
consume, but the less
fun you get to have.

254
00:10:58,170 --> 00:11:00,040
Fun, we call leisure.

255
00:11:00,040 --> 00:11:02,840
The less labor you provide to
the market, the more fun you

256
00:11:02,840 --> 00:11:04,630
get to have, the more leisure
you get, but the less you get

257
00:11:04,630 --> 00:11:06,470
to consume, because you
have less income.

258
00:11:06,470 --> 00:11:08,200
And that's the trade-off we
talked about when we talked

259
00:11:08,200 --> 00:11:10,230
about the effect of
minimum wage.

260
00:11:10,230 --> 00:11:12,430
Now let's come back and
get underneath that

261
00:11:12,430 --> 00:11:13,840
labor supply curve.

262
00:11:13,840 --> 00:11:14,770
So we talked about
the minimum wage.

263
00:11:14,770 --> 00:11:18,600
We talked about the labor supply
curve which was how the

264
00:11:18,600 --> 00:11:21,750
hours you provide respond to
the wage and a labor demand

265
00:11:21,750 --> 00:11:24,030
curve, which was how the
hours that firms want

266
00:11:24,030 --> 00:11:25,140
respond to the wage.

267
00:11:25,140 --> 00:11:28,330
Now let's get underneath
the supply curve.

268
00:11:28,330 --> 00:11:30,240
A minute ago, we were talking
about the demand curves and

269
00:11:30,240 --> 00:11:31,900
getting underneath the demand
curve for consumers.

270
00:11:31,900 --> 00:11:35,940
Well, now let's get underneath
the supply curve for labor.

271
00:11:35,940 --> 00:11:42,410
Now, the key thing is that when
we talk about labor, it's

272
00:11:42,410 --> 00:11:44,480
not a good, it's a bad.

273
00:11:44,480 --> 00:11:46,480
The typical person doesn't
want to work.

274
00:11:46,480 --> 00:11:48,280
The typical person is
not in this room.

275
00:11:48,280 --> 00:11:49,770
You guys like to work.

276
00:11:49,770 --> 00:11:51,770
The typically person actually
doesn't like to work.

277
00:11:54,610 --> 00:11:59,750
Leisure is a normal good.

278
00:12:02,360 --> 00:12:05,250
For the typical person, leisure
is a normal good.

279
00:12:05,250 --> 00:12:06,600
They like time off.

280
00:12:13,160 --> 00:12:17,430
Leisure is a good, which
means labor is a bad.

281
00:12:17,430 --> 00:12:19,870
They don't like to work.

282
00:12:19,870 --> 00:12:21,000
The problem is we don't
know how to

283
00:12:21,000 --> 00:12:24,450
model bads in economics.

284
00:12:24,450 --> 00:12:26,020
It's just we're used to
trading off between

285
00:12:26,020 --> 00:12:26,900
two things you want.

286
00:12:26,900 --> 00:12:28,980
When I used to trade-off, we
know how to model something

287
00:12:28,980 --> 00:12:30,630
you want to get something
you don't want.

288
00:12:30,630 --> 00:12:32,600
Indifference curves wouldn't
work, because

289
00:12:32,600 --> 00:12:34,020
more wouldn't be better.

290
00:12:34,020 --> 00:12:36,130
If you drew an indifference
curve for labor, it would

291
00:12:36,130 --> 00:12:37,960
violate the more is
better assumption.

292
00:12:37,960 --> 00:12:38,730
Because you wouldn't
want more.

293
00:12:38,730 --> 00:12:40,480
You'd want less.

294
00:12:40,480 --> 00:12:43,320
So the modeling trick we're
going to use whenever we're

295
00:12:43,320 --> 00:12:48,880
modeling bads, is to model the
complementary good and then,

296
00:12:48,880 --> 00:12:51,930
in the end, solve for the bad.

297
00:12:51,930 --> 00:12:53,020
We're not going to
model labor.

298
00:12:53,020 --> 00:12:55,100
We're going to model leisure.

299
00:12:55,100 --> 00:12:58,790
And given the total amount of
hours you have to supply, the

300
00:12:58,790 --> 00:13:00,680
total hours minus the
amount of leisure is

301
00:13:00,680 --> 00:13:02,320
the amount of labor.

302
00:13:02,320 --> 00:13:03,880
So we're going to
model leisure.

303
00:13:03,880 --> 00:13:05,660
We're going to model the
good and then solve for

304
00:13:05,660 --> 00:13:08,080
labor at the end.

305
00:13:08,080 --> 00:13:10,210
So, in other words, if you have
24 hours a day you can

306
00:13:10,210 --> 00:13:16,230
work, then your amount of hours
of work is 24 minus the

307
00:13:16,230 --> 00:13:20,880
amount of leisure N. Call it N
or call it L. We'll call it N

308
00:13:20,880 --> 00:13:24,320
because L, typically, we
think of as labor.

309
00:13:24,320 --> 00:13:26,750
Let's call leisure N for
reasons I don't quite

310
00:13:26,750 --> 00:13:27,220
understand.

311
00:13:27,220 --> 00:13:29,430
Let's just use that.

312
00:13:29,430 --> 00:13:31,370
Basically the amount
of hours you can

313
00:13:31,370 --> 00:13:33,120
work is 24 minus leisure.

314
00:13:33,120 --> 00:13:37,440
So if we solve for the optimal
amount of leisure you want, we

315
00:13:37,440 --> 00:13:41,300
can obviously get the amount
of labor you supply.

316
00:13:41,300 --> 00:13:44,160
So the trick when modeling a bad
is not to model the bad.

317
00:13:44,160 --> 00:13:46,840
It's to model the complementary
good.

318
00:13:46,840 --> 00:13:49,720
In this case, the complementary
good is leisure.

319
00:13:49,720 --> 00:13:53,670
So we're going to model the
trade-off between leisure and

320
00:13:53,670 --> 00:13:59,030
consumption and use the result
of that to solve for the

321
00:13:59,030 --> 00:14:01,640
amount of labor you supply.

322
00:14:01,640 --> 00:14:04,190
So it's the general modeling
trick you need to understand,

323
00:14:04,190 --> 00:14:05,980
which is turn a bad
into a good.

324
00:14:05,980 --> 00:14:07,240
That's the modeling trick.

325
00:14:07,240 --> 00:14:08,230
Because we know how
to model the

326
00:14:08,230 --> 00:14:09,330
trade-off between two goods.

327
00:14:09,330 --> 00:14:12,110
We don't know how to model
the trade-off with a bad.

328
00:14:12,110 --> 00:14:16,710
So to think about that, let's
go to Figure 7-3, and let's

329
00:14:16,710 --> 00:14:20,740
talk about what's underneath
a labor supply curve.

330
00:14:20,740 --> 00:14:25,860
What's underneath a labor supply
curve is the trade-off

331
00:14:25,860 --> 00:14:30,130
between how much leisure you
want and how much consumption

332
00:14:30,130 --> 00:14:33,690
you can have. So you see here,
here's a trade-off.

333
00:14:33,690 --> 00:14:37,080
On the y-axis is the amount
of goods you can have. You

334
00:14:37,080 --> 00:14:38,680
earn a wage, w.

335
00:14:38,680 --> 00:14:40,610
The y-axis is the amount of
goods you can have from a

336
00:14:40,610 --> 00:14:43,080
day's work.

337
00:14:43,080 --> 00:14:44,820
So you earn w per hour.

338
00:14:44,820 --> 00:14:46,650
That means the most goods
you can have from a

339
00:14:46,650 --> 00:14:47,540
day's work is 24w.

340
00:14:47,540 --> 00:14:50,310
If you worked all 24 hours
at that wage, you

341
00:14:50,310 --> 00:14:53,170
can have 24w goods.

342
00:14:53,170 --> 00:14:56,760
On the other hand, if you work
not at all, then you take 24

343
00:14:56,760 --> 00:15:01,820
hours in leisure and have no
consumption from that day.

344
00:15:01,820 --> 00:15:04,030
So we see as you move
to the right on the

345
00:15:04,030 --> 00:15:05,730
x-axis, that's leisure.

346
00:15:05,730 --> 00:15:06,840
That's the good.

347
00:15:06,840 --> 00:15:08,290
As you move to the left,
that's labor.

348
00:15:08,290 --> 00:15:09,540
That's the bad.

349
00:15:09,540 --> 00:15:10,220
OK?

350
00:15:10,220 --> 00:15:10,930
That's just illustrating.

351
00:15:10,930 --> 00:15:12,090
But we're going to
model the good.

352
00:15:12,090 --> 00:15:13,340
We're going to model leisure.

353
00:15:16,330 --> 00:15:18,740
Your trade-off is between how
much you want to consume and

354
00:15:18,740 --> 00:15:20,810
how much leisure you
want to take.

355
00:15:20,810 --> 00:15:22,060
Now, here's what's
interesting.

356
00:15:25,940 --> 00:15:27,920
In general, what determines
the slope of a budget

357
00:15:27,920 --> 00:15:29,170
constraint?

358
00:15:31,010 --> 00:15:32,930
What determines the slope
of a budget constraint?

359
00:15:32,930 --> 00:15:34,640
AUDIENCE: Marginal rate
of transformation.

360
00:15:34,640 --> 00:15:35,556
PROFESSOR: Which is what?

361
00:15:35,556 --> 00:15:37,300
AUDIENCE: Ratio between
prices.

362
00:15:37,300 --> 00:15:38,190
PROFESSOR: Ratio
between prices.

363
00:15:38,190 --> 00:15:40,180
Prices determine the slope
of the budget constraint.

364
00:15:40,180 --> 00:15:41,260
But here's what's tricky.

365
00:15:41,260 --> 00:15:44,830
What's the price of leisure?

366
00:15:44,830 --> 00:15:45,700
AUDIENCE: Wage.

367
00:15:45,700 --> 00:15:46,180
PROFESSOR: The wage.

368
00:15:46,180 --> 00:15:46,985
Why?

369
00:15:46,985 --> 00:15:51,645
AUDIENCE: Because for every hour
you take having leisure,

370
00:15:51,645 --> 00:15:54,452
you are effectively using
money that you

371
00:15:54,452 --> 00:15:55,330
could gain at work.

372
00:15:55,330 --> 00:15:56,090
PROFESSOR: Exactly.

373
00:15:56,090 --> 00:16:03,390
The key is the economic concept
of opportunity cost,

374
00:16:03,390 --> 00:16:06,040
which we've talked about and
will continue to talk about

375
00:16:06,040 --> 00:16:10,780
this semester, opportunity cost.
By not working, you are

376
00:16:10,780 --> 00:16:13,120
forgoing earning a wage.

377
00:16:13,120 --> 00:16:15,500
So that is the price
of leisure.

378
00:16:15,500 --> 00:16:17,290
You may not think of it this
way, but, once again, that's

379
00:16:17,290 --> 00:16:19,530
why we're the dismal science.

380
00:16:19,530 --> 00:16:21,650
When you go home today, and you
sit on the couch, and you

381
00:16:21,650 --> 00:16:25,990
watch TV for an hour, you
have just paid a price.

382
00:16:25,990 --> 00:16:27,710
And that price is what you
could have earned by

383
00:16:27,710 --> 00:16:29,760
working that hour.

384
00:16:29,760 --> 00:16:31,510
Every action has a price.

385
00:16:31,510 --> 00:16:38,040
And the price of leisure is the
wage you forgo, The wage

386
00:16:38,040 --> 00:16:42,850
you forgo by sitting around
is the price of leisure.

387
00:16:50,010 --> 00:16:54,760
Let's assume here that the price
of goods is $1, that the

388
00:16:54,760 --> 00:16:57,200
goods you're going
to buy cost $1.

389
00:16:57,200 --> 00:16:58,560
Whatever your consumption,
it costs $1.

390
00:16:58,560 --> 00:16:59,730
That's the trick we always
use with modeling.

391
00:16:59,730 --> 00:17:01,600
Make as many things
$1 as you can.

392
00:17:01,600 --> 00:17:02,870
That makes the model easy.

393
00:17:02,870 --> 00:17:04,609
So let's assume that the price
of the goods you're going to

394
00:17:04,609 --> 00:17:05,990
buy are $1.

395
00:17:05,990 --> 00:17:09,760
So the slope of the budget
constraint is minus w over 1.

396
00:17:09,760 --> 00:17:11,569
The slope of the budget
constraint is just the price

397
00:17:11,569 --> 00:17:13,990
of leisure which is minus w .

398
00:17:13,990 --> 00:17:17,440
So the trade-off with the price
of goods of $1, the

399
00:17:17,440 --> 00:17:21,800
trade-off between taking leisure
and consuming is that

400
00:17:21,800 --> 00:17:25,410
if you take leisure,
an hour of leisure,

401
00:17:25,410 --> 00:17:26,789
you get w fewer goods.

402
00:17:29,470 --> 00:17:32,500
And if you work an hour, you get
w more goods, but you lose

403
00:17:32,500 --> 00:17:34,780
an hour of leisure.

404
00:17:34,780 --> 00:17:39,150
And that gives you the trade-off
between how much you

405
00:17:39,150 --> 00:17:41,190
consume and how much leisure you
take which determines how

406
00:17:41,190 --> 00:17:42,660
much you work.

407
00:17:42,660 --> 00:17:42,960
OK.

408
00:17:42,960 --> 00:17:45,530
Questions about that?

409
00:17:45,530 --> 00:17:49,290
Now let's take this framework
and ask, what happens when the

410
00:17:49,290 --> 00:17:51,810
wage changes, Figure 7-4.

411
00:17:54,630 --> 00:17:57,780
So we have an original
outcome with the

412
00:17:57,780 --> 00:17:59,030
budget constraint BC1.

413
00:18:06,780 --> 00:18:10,320
We have an original budget
constraint, BC1.

414
00:18:10,320 --> 00:18:15,700
Now imagine the wage goes
up, so we move to BC2.

415
00:18:15,700 --> 00:18:18,770
BC2 is a budget constraint
with a higher wage.

416
00:18:18,770 --> 00:18:20,290
The wage goes up.

417
00:18:20,290 --> 00:18:23,020
So what we're going to see is
you're going to move from

418
00:18:23,020 --> 00:18:27,720
point A where you work N1 hours
to point C where you

419
00:18:27,720 --> 00:18:30,030
work N3 hours.

420
00:18:30,030 --> 00:18:32,250
That's where your indifference
curves are tangent with the

421
00:18:32,250 --> 00:18:33,500
new budget constraint.

422
00:18:35,920 --> 00:18:36,730
Not work, take leisure.

423
00:18:36,730 --> 00:18:37,300
I'm sorry.

424
00:18:37,300 --> 00:18:40,470
We take leisure of N1 hours
to leisure of N3 hours.

425
00:18:40,470 --> 00:18:42,920
The wage going up
has reduced your

426
00:18:42,920 --> 00:18:45,790
leisure which makes sense.

427
00:18:45,790 --> 00:18:47,580
If the wage goes up,
you work harder.

428
00:18:47,580 --> 00:18:48,270
Right?

429
00:18:48,270 --> 00:18:50,750
So your wage going up, we always
first take if there's a

430
00:18:50,750 --> 00:18:52,850
leisure and then convert
to labor.

431
00:18:52,850 --> 00:18:59,010
Wage goes up, leisure falls
from N1 to N3, which means

432
00:18:59,010 --> 00:19:00,720
labor goes up.

433
00:19:00,720 --> 00:19:03,360
But actually two things are
happening here, the

434
00:19:03,360 --> 00:19:05,740
substitution and
income effect.

435
00:19:05,740 --> 00:19:09,830
The substitution effect, which
we see by drawing the

436
00:19:09,830 --> 00:19:15,040
imaginary budget constraint BC*
which is parallel to BC2

437
00:19:15,040 --> 00:19:18,100
but tangent to the original
difference curve, the

438
00:19:18,100 --> 00:19:22,240
substitution effect is a very
large reduction in leisure.

439
00:19:22,240 --> 00:19:25,895
It moves all the way
from N1 to N2.

440
00:19:25,895 --> 00:19:27,590
The substitution effect
is a very large

441
00:19:27,590 --> 00:19:28,930
reduction in leisure.

442
00:19:28,930 --> 00:19:33,760
The income effect is that
leisure is a normal good.

443
00:19:33,760 --> 00:19:36,830
I'm now richer, because
my wage has gone up.

444
00:19:36,830 --> 00:19:39,570
So I want to buy more of it.

445
00:19:39,570 --> 00:19:41,430
So I buy more leisure.

446
00:19:41,430 --> 00:19:45,620
And that moves me
from N2 to N3.

447
00:19:45,620 --> 00:19:50,700
So, basically, now the income
effect offsets the

448
00:19:50,700 --> 00:19:54,690
substitution effect even with
a normal good, or with a

449
00:19:54,690 --> 00:19:55,860
normal good.

450
00:19:55,860 --> 00:19:57,800
With a normal good, the income
effect offsets that

451
00:19:57,800 --> 00:19:59,950
substitution effect.

452
00:19:59,950 --> 00:20:03,220
And that's because the money
you're getting, you're using

453
00:20:03,220 --> 00:20:06,210
to buy leisure.

454
00:20:06,210 --> 00:20:10,940
So, in fact, if you flip to 7-5,
you can see a case where

455
00:20:10,940 --> 00:20:12,190
the income effect dominates.

456
00:20:19,460 --> 00:20:22,410
And you actually get that a
wage increase leads you to

457
00:20:22,410 --> 00:20:24,600
work less hard.

458
00:20:24,600 --> 00:20:25,160
Now, think about that.

459
00:20:25,160 --> 00:20:25,970
If I'd said to you--

460
00:20:25,970 --> 00:20:27,200
I probably should have
started with this--

461
00:20:27,200 --> 00:20:29,540
if you increase the wage, will
people work more or less hard?

462
00:20:29,540 --> 00:20:31,110
Your initial instinct would
have been more hard.

463
00:20:31,110 --> 00:20:32,730
You would have thought, well,
if your wage goes

464
00:20:32,730 --> 00:20:33,580
up, you work harder.

465
00:20:33,580 --> 00:20:34,990
But that's because your instinct
was focused on the

466
00:20:34,990 --> 00:20:36,660
substitution effect.

467
00:20:36,660 --> 00:20:37,920
You're thinking about
the income effect.

468
00:20:37,920 --> 00:20:41,200
Here's a case where
I started at N1.

469
00:20:41,200 --> 00:20:43,610
The substitution effect
leads me to N2.

470
00:20:43,610 --> 00:20:47,410
But I feel so much richer from
that higher wage that I

471
00:20:47,410 --> 00:20:48,870
actually move all
the way to N3.

472
00:20:48,870 --> 00:20:53,690
My leisure goes up, and
I work less hard.

473
00:20:53,690 --> 00:20:59,280
Now, unlike a Giffen good, this
is totally plausible.

474
00:20:59,280 --> 00:21:00,750
Why is it plausible?

475
00:21:00,750 --> 00:21:02,980
Well, let me do give you
a simple intuition

476
00:21:02,980 --> 00:21:04,470
for why it's plausible.

477
00:21:04,470 --> 00:21:08,280
Let's say that you're someone
who has a certain amount of

478
00:21:08,280 --> 00:21:09,980
things you want to
buy every week.

479
00:21:09,980 --> 00:21:12,250
You don't save. You have a
certain amount of things you

480
00:21:12,250 --> 00:21:13,320
want to buy every week.

481
00:21:13,320 --> 00:21:15,210
You have to pay your rent, you
have to buy your food, you

482
00:21:15,210 --> 00:21:17,740
have to buy your other goodies,
a certain budget.

483
00:21:17,740 --> 00:21:19,440
A lot of people live
on a budget.

484
00:21:19,440 --> 00:21:20,710
You have a certain budget.

485
00:21:20,710 --> 00:21:23,220
And the truth is you're happy
with that budget.

486
00:21:23,220 --> 00:21:25,420
That's kind of what
you want to do.

487
00:21:25,420 --> 00:21:28,070
Now let's say I doubled
your wage.

488
00:21:28,070 --> 00:21:31,550
Well, now to meet the budget you
can work half as hard and

489
00:21:31,550 --> 00:21:33,620
still meet the same budget.

490
00:21:33,620 --> 00:21:36,780
So you'll work less hard.

491
00:21:36,780 --> 00:21:39,370
You could say, look, I can get
more leisure and consume the

492
00:21:39,370 --> 00:21:42,070
same amount of goods
as I did before.

493
00:21:42,070 --> 00:21:44,410
So I'll work less hard.

494
00:21:44,410 --> 00:21:45,640
That's a totally
plausible case.

495
00:21:45,640 --> 00:21:48,340
That's a case of what we
call target income.

496
00:21:48,340 --> 00:21:51,370
If someone has a target income,
and their wage goes

497
00:21:51,370 --> 00:21:53,790
up, they'll work less.

498
00:21:53,790 --> 00:21:56,120
Now, that's not necessarily
the truth.

499
00:21:56,120 --> 00:21:59,240
But it's, at least to me, sort
of a plausible case of how

500
00:21:59,240 --> 00:22:01,380
people might behave. And that's
a case where income

501
00:22:01,380 --> 00:22:03,630
effects can dominate.

502
00:22:03,630 --> 00:22:12,480
So if we, once again, go to the
second chart on that page,

503
00:22:12,480 --> 00:22:13,500
now we see the income
and substitution

504
00:22:13,500 --> 00:22:14,750
effects for labor supply.

505
00:22:19,700 --> 00:22:21,380
Once again, we're assuming
leisure is a normal good.

506
00:22:21,380 --> 00:22:22,610
We're always going to assume
leisure is a normal good.

507
00:22:22,610 --> 00:22:25,050
We're never going to assume
people don't like leisure.

508
00:22:25,050 --> 00:22:29,530
Assuming leisure is a normal
good, then as the wage rises,

509
00:22:29,530 --> 00:22:33,260
the substitution effect is
you take less leisure.

510
00:22:33,260 --> 00:22:34,720
This table is a bit different
than the other table.

511
00:22:34,720 --> 00:22:36,370
Instead of the first panel being
normal and the second

512
00:22:36,370 --> 00:22:38,670
panel being inferior,
the first panel is

513
00:22:38,670 --> 00:22:39,930
what happens to leisure.

514
00:22:39,930 --> 00:22:42,750
The second panel converts it
to what happens to labor.

515
00:22:42,750 --> 00:22:45,530
So for instance, in the first
cell, when the wage rises, the

516
00:22:45,530 --> 00:22:48,300
substitution effect on leisure
is unambiguously negative.

517
00:22:48,300 --> 00:22:50,810
You clearly take less leisure
when the wage rises.

518
00:22:50,810 --> 00:22:53,100
So, likewise, you
have more labor.

519
00:22:53,100 --> 00:22:55,530
So on the bottom panel, labor
is clearly greater than or

520
00:22:55,530 --> 00:22:56,710
less than 0.

521
00:22:56,710 --> 00:23:00,240
But the income effect is
positive for leisure.

522
00:23:00,240 --> 00:23:01,450
You're rich, you take
more leisure.

523
00:23:01,450 --> 00:23:04,110
Or, likewise, negative for
labor, you're richer, so your

524
00:23:04,110 --> 00:23:05,480
work less hard.

525
00:23:05,480 --> 00:23:09,550
And, therefore, the
net is ambiguous.

526
00:23:09,550 --> 00:23:15,240
So with goods consumption, we
needed goods to be inferior

527
00:23:15,240 --> 00:23:17,630
for there to be a Giffen
good type phenomena.

528
00:23:17,630 --> 00:23:23,500
Here, even with leisure being
normal, you can have a Giffen

529
00:23:23,500 --> 00:23:24,530
good type phenomena.

530
00:23:24,530 --> 00:23:26,250
It's much less random.

531
00:23:26,250 --> 00:23:28,430
And, in some sense, this
is why we learn income

532
00:23:28,430 --> 00:23:29,540
substitution effects.

533
00:23:29,540 --> 00:23:31,750
To be honest, they're just
not that interesting for

534
00:23:31,750 --> 00:23:32,970
consumption.

535
00:23:32,970 --> 00:23:34,800
The book makes a big deal out
of them and talks about

536
00:23:34,800 --> 00:23:36,420
consumer price indices
and all that.

537
00:23:36,420 --> 00:23:38,920
It's just not that important
for consumption.

538
00:23:38,920 --> 00:23:40,240
Because we know in consumption
if prices goes

539
00:23:40,240 --> 00:23:41,670
up, you consume less.

540
00:23:41,670 --> 00:23:43,010
It's just not that
interesting.

541
00:23:43,010 --> 00:23:45,570
It's much more interesting for
things like labor supply.

542
00:23:45,570 --> 00:23:47,420
And we talk about savings
in a number of lectures.

543
00:23:47,420 --> 00:23:48,700
It's the same thing.

544
00:23:48,700 --> 00:23:49,550
There, it's more interesting.

545
00:23:49,550 --> 00:23:51,560
Because now they can often
offset each other in

546
00:23:51,560 --> 00:23:53,100
meaningful ways.

547
00:23:53,100 --> 00:23:55,210
And so now this is why the
tools of income and

548
00:23:55,210 --> 00:23:58,220
substitution effects become
much more important.

549
00:23:58,220 --> 00:23:59,890
OK?

550
00:23:59,890 --> 00:24:09,230
So if we put this together, if
we go to Figure 7-6, we can

551
00:24:09,230 --> 00:24:13,640
now think about deriving where
labor supply comes from.

552
00:24:13,640 --> 00:24:15,770
Where does labor supply
come from?

553
00:24:15,770 --> 00:24:17,960
Well, first, you've got the
consumer's decision of how

554
00:24:17,960 --> 00:24:18,600
hard to work.

555
00:24:18,600 --> 00:24:19,270
So here's a case.

556
00:24:19,270 --> 00:24:23,440
It's sort of small, but
you can take a look.

557
00:24:23,440 --> 00:24:26,920
Here's a case where you've got
someone initially working,

558
00:24:26,920 --> 00:24:30,110
taking 16 hours of leisure and,
therefore, working eight

559
00:24:30,110 --> 00:24:32,750
hours, at a wage of W1.

560
00:24:38,780 --> 00:24:41,250
Now their wage goes up to W2.

561
00:24:41,250 --> 00:24:44,100
They choose to take 12 hours
of leisure and, therefore,

562
00:24:44,100 --> 00:24:48,890
work 12 hours.

563
00:24:48,890 --> 00:24:50,920
This is someone who works harder
when the wage goes up.

564
00:24:50,920 --> 00:24:52,600
That is, the income effect
does not offset the

565
00:24:52,600 --> 00:24:55,600
substitution effect.

566
00:24:55,600 --> 00:24:59,670
Now, we can take that to draw
a demand for leisure curve

567
00:24:59,670 --> 00:25:02,050
just like we drew any
other demand curve.

568
00:25:02,050 --> 00:25:04,080
It's the same technique
as last time.

569
00:25:04,080 --> 00:25:07,700
Just bring those point and say,
look, at a wage of W1,

570
00:25:07,700 --> 00:25:08,780
leisure is 16.

571
00:25:08,780 --> 00:25:10,250
At a wage of W2,
leisure is 12.

572
00:25:10,250 --> 00:25:13,500
We have a downward sloping
demand for leisure, standard

573
00:25:13,500 --> 00:25:15,180
downward sloping demand
for leisure.

574
00:25:15,180 --> 00:25:19,063
But we can convert that to a
supply of labor, which is what

575
00:25:19,063 --> 00:25:19,340
we care about.

576
00:25:19,340 --> 00:25:21,300
Nobody cares about the demand
for leisure curve.

577
00:25:21,300 --> 00:25:24,270
We care about the supply
of labor curve.

578
00:25:24,270 --> 00:25:26,210
You just subtract
these from 24.

579
00:25:26,210 --> 00:25:29,880
You use the supply of labor
curve which is upward sloping.

580
00:25:29,880 --> 00:25:32,780
So as long as substitution
effects dominate income

581
00:25:32,780 --> 00:25:37,610
effects, we'll get an upward
sloping labor supply curve.

582
00:25:37,610 --> 00:25:42,900
But it's certainly possible
that if income effects

583
00:25:42,900 --> 00:25:45,900
dominates substitution effects,
you could get a

584
00:25:45,900 --> 00:25:50,180
downward sloping supply curve,
if you will, what we call in

585
00:25:50,180 --> 00:25:53,720
labor economics, a
backward-bending supply curve,

586
00:25:53,720 --> 00:25:55,260
a supply curve that goes
the wrong way.

587
00:25:55,260 --> 00:25:57,600
Instead of sloping up like
supply curves are supposed to,

588
00:25:57,600 --> 00:26:00,280
it goes the wrong way
and slopes down.

589
00:26:00,280 --> 00:26:01,890
And we can see that's
plausible.

590
00:26:01,890 --> 00:26:03,900
The target income case I
just described to you

591
00:26:03,900 --> 00:26:05,500
would deliver that.

592
00:26:05,500 --> 00:26:07,660
The target income case I just
described to you would deliver

593
00:26:07,660 --> 00:26:09,450
a downward sloping
supply of labor.

594
00:26:09,450 --> 00:26:11,740
As the wage rose, people would
work less and less.

595
00:26:14,620 --> 00:26:17,240
That's a totally
plausible case.

596
00:26:17,240 --> 00:26:18,770
And that's why income and
substitution effects are

597
00:26:18,770 --> 00:26:19,370
interesting.

598
00:26:19,370 --> 00:26:21,540
Because they can deliver
this weird result.

599
00:26:21,540 --> 00:26:25,050
They can get the wrong
signed supply curve.

600
00:26:25,050 --> 00:26:27,790
Questions about income
and substitution

601
00:26:27,790 --> 00:26:31,620
effects or labor supply?

602
00:26:31,620 --> 00:26:34,330
So what I want to spend the
rest of the lecture on is

603
00:26:34,330 --> 00:26:38,570
talking about well,
what is the case?

604
00:26:38,570 --> 00:26:40,750
Do labor supply curves
slope up or down?

605
00:26:40,750 --> 00:26:41,970
And what do we know
about that?

606
00:26:41,970 --> 00:26:46,600
Well, this is probably the major
focus of a field we call

607
00:26:46,600 --> 00:26:47,850
labor economics.

608
00:26:49,990 --> 00:26:52,570
And there's an excellent course
on labor economics,

609
00:26:52,570 --> 00:26:56,045
14.64 taught by Josh Angrist,
which goes into much detail in

610
00:26:56,045 --> 00:26:56,730
the entire field.

611
00:26:56,730 --> 00:26:59,810
But one of the main focuses of
the field is understanding the

612
00:26:59,810 --> 00:27:02,560
elasticity of labor supply,
and is it positive or

613
00:27:02,560 --> 00:27:04,270
negative, and how big is it.

614
00:27:04,270 --> 00:27:06,250
So, basically, measuring the
slope of the labor supply

615
00:27:06,250 --> 00:27:08,320
curve is the focus of this
literature, the elasticity of

616
00:27:08,320 --> 00:27:09,940
the labor supply.

617
00:27:09,940 --> 00:27:12,730
Now, what I want to do is start
with a historical fact,

618
00:27:12,730 --> 00:27:13,960
and then I'll come to
the modern age.

619
00:27:13,960 --> 00:27:16,110
Let's think about
30 years ago.

620
00:27:16,110 --> 00:27:20,250
30 years ago, all men
worked and less than

621
00:27:20,250 --> 00:27:21,550
half of women worked.

622
00:27:21,550 --> 00:27:24,610
It was more normal for women
not to work than to work,

623
00:27:24,610 --> 00:27:25,250
married women.

624
00:27:25,250 --> 00:27:25,590
I'm sorry.

625
00:27:25,590 --> 00:27:27,060
Less than half of married
women worked.

626
00:27:30,790 --> 00:27:32,630
Now, married women could work.

627
00:27:32,630 --> 00:27:36,230
I'm not talking 60 years ago
or 80 years ago when there

628
00:27:36,230 --> 00:27:37,340
were marriage bars.

629
00:27:37,340 --> 00:27:39,730
Literally, firms wouldn't hire
you if you were married.

630
00:27:39,730 --> 00:27:40,420
It's true.

631
00:27:40,420 --> 00:27:41,500
If you're interested in
that, you can actually

632
00:27:41,500 --> 00:27:42,700
read Claudia Goldin.

633
00:27:42,700 --> 00:27:45,530
She's a labor historian who's
written about the early 20th

634
00:27:45,530 --> 00:27:46,760
century when, literally,
women could be

635
00:27:46,760 --> 00:27:48,670
fired for being married.

636
00:27:48,670 --> 00:27:49,790
We're not talking
about that era.

637
00:27:49,790 --> 00:27:51,710
I'm talking about 30 years ago
when you could work if you

638
00:27:51,710 --> 00:27:52,090
were married.

639
00:27:52,090 --> 00:27:53,030
It's no problem.

640
00:27:53,030 --> 00:27:59,400
But most women chose not to,
maybe 40 years ago now.

641
00:27:59,400 --> 00:28:03,750
So, in that case, let's think
about two groups.

642
00:28:03,750 --> 00:28:10,260
Let's think about married
men, and let's think

643
00:28:10,260 --> 00:28:12,520
about married women.

644
00:28:12,520 --> 00:28:17,730
And let's just posit,
hypothetically, how big we

645
00:28:17,730 --> 00:28:21,190
think their substitution and
income effects would be.

646
00:28:26,600 --> 00:28:28,560
Let's start with substitution
effect.

647
00:28:28,560 --> 00:28:33,050
Do we think the substitution
effect would be bigger?

648
00:28:33,050 --> 00:28:36,740
This is the change in the wage
holding utility constant.

649
00:28:36,740 --> 00:28:39,180
Do we think that would have a
bigger effect on leisure and,

650
00:28:39,180 --> 00:28:42,280
therefore, labor for men
or for women and why?

651
00:28:42,280 --> 00:28:42,900
Don't yell it out.

652
00:28:42,900 --> 00:28:44,570
Somebody, raise their
hand and tell me.

653
00:28:44,570 --> 00:28:46,726
Do we think that the
substitution effect would be

654
00:28:46,726 --> 00:28:51,080
bigger for men or
women and why?

655
00:28:51,080 --> 00:28:51,830
Remember the name.

656
00:28:51,830 --> 00:28:52,950
It's the substitution effect.

657
00:28:52,950 --> 00:28:54,200
That's the key to the answer.

658
00:28:58,520 --> 00:28:58,785
Yeah.

659
00:28:58,785 --> 00:29:02,000
AUDIENCE: I think it would be
the same, because they each

660
00:29:02,000 --> 00:29:05,444
have equal use for the goods.

661
00:29:05,444 --> 00:29:08,396
Maybe their income effect
would be different.

662
00:29:08,396 --> 00:29:08,888
PROFESSOR: OK.

663
00:29:08,888 --> 00:29:10,190
[UNINTELLIGIBLE PHRASE].

664
00:29:10,190 --> 00:29:11,565
They each have equal
use for the goods.

665
00:29:11,565 --> 00:29:12,840
Well let's deal with where
the substitution

666
00:29:12,840 --> 00:29:13,410
effect comes from.

667
00:29:13,410 --> 00:29:16,600
Let's break it down.

668
00:29:16,600 --> 00:29:19,460
So you're someone
who's deciding.

669
00:29:19,460 --> 00:29:23,720
You've got you and your wife,
and you're each deciding how

670
00:29:23,720 --> 00:29:27,050
to respond to a change
in the wage.

671
00:29:27,050 --> 00:29:30,100
Now, you both value the
goods the same.

672
00:29:30,100 --> 00:29:32,490
But it's goods versus leisure.

673
00:29:32,490 --> 00:29:34,085
What's the other feature
that you're going

674
00:29:34,085 --> 00:29:35,280
to be thinking about?

675
00:29:35,280 --> 00:29:37,780
Think about a married man
40 years ago, and

676
00:29:37,780 --> 00:29:39,783
the wage goes down.

677
00:29:39,783 --> 00:29:41,900
AUDIENCE: They have
to work more.

678
00:29:41,900 --> 00:29:42,660
PROFESSOR: No.

679
00:29:42,660 --> 00:29:43,880
We're just doing substitution
effects.

680
00:29:43,880 --> 00:29:44,240
That's unambiguous.

681
00:29:44,240 --> 00:29:45,980
If the goes down,
they work less.

682
00:29:45,980 --> 00:29:47,120
We're just doing substitution
effects.

683
00:29:47,120 --> 00:29:48,220
That's ambiguous.

684
00:29:48,220 --> 00:29:52,540
The question is, if they work
less, what do they do?

685
00:29:52,540 --> 00:29:55,360
Whereas think about a married
women 40 years ago.

686
00:29:55,360 --> 00:29:56,680
If she works less,
what does she do?

687
00:29:56,680 --> 00:29:58,670
What does a married man do?

688
00:29:58,670 --> 00:29:59,340
Nothing.

689
00:29:59,340 --> 00:30:00,060
There's nothing to do.

690
00:30:00,060 --> 00:30:01,050
Your friends are all at work.

691
00:30:01,050 --> 00:30:03,250
You can't go play golf.

692
00:30:03,250 --> 00:30:04,360
You can't do anything.

693
00:30:04,360 --> 00:30:06,370
You don't take care of kids,
because men didn't take care

694
00:30:06,370 --> 00:30:08,910
of kids 40 years ago.

695
00:30:08,910 --> 00:30:09,670
What do you do?

696
00:30:09,670 --> 00:30:11,100
There's nothing to do.

697
00:30:11,100 --> 00:30:13,860
Whereas a woman, married woman,
if the wage goes down

698
00:30:13,860 --> 00:30:16,730
40 years ago, you can take
care of kids instead.

699
00:30:16,730 --> 00:30:18,840
You can hang out with other
women who aren't working.

700
00:30:18,840 --> 00:30:20,670
There's plenty to do.

701
00:30:20,670 --> 00:30:22,030
Based on that, now change
your answer.

702
00:30:22,030 --> 00:30:22,840
Where do you think
the substitution

703
00:30:22,840 --> 00:30:24,082
effect would be bigger?

704
00:30:24,082 --> 00:30:26,830
AUDIENCE: [INAUDIBLE PHRASE].

705
00:30:26,830 --> 00:30:28,160
PROFESSOR: In women,
it would be bigger.

706
00:30:28,160 --> 00:30:30,860
Because men, there's less of
a substitution effect.

707
00:30:30,860 --> 00:30:33,940
Because it's all about
substitutability of options.

708
00:30:33,940 --> 00:30:35,760
There's no good alternative
option to work for

709
00:30:35,760 --> 00:30:37,480
men 40 years ago.

710
00:30:37,480 --> 00:30:39,130
It's either work or nothing.

711
00:30:39,130 --> 00:30:41,870
Basically, everybody worked.

712
00:30:41,870 --> 00:30:44,620
So, basically, there's
no good substitution

713
00:30:44,620 --> 00:30:45,960
effect option for men.

714
00:30:45,960 --> 00:30:50,030
For women, there's lots
of outside options.

715
00:30:50,030 --> 00:30:54,300
There's sociability, there's
child rearing, et cetera.

716
00:30:54,300 --> 00:30:57,590
The substitution effect will
be larger the more things

717
00:30:57,590 --> 00:31:00,680
there are you can
substitute to.

718
00:31:00,680 --> 00:31:03,180
Men don't have anything to
substitute to from work.

719
00:31:03,180 --> 00:31:07,300
Women have options to substitute
to from work.

720
00:31:07,300 --> 00:31:09,870
For men, this is going
to be very small.

721
00:31:09,870 --> 00:31:13,750
For women, this will be big.

722
00:31:13,750 --> 00:31:15,820
We know the sign.

723
00:31:15,820 --> 00:31:18,090
The smallest this can be is 0.

724
00:31:18,090 --> 00:31:20,020
We know the sign.

725
00:31:20,020 --> 00:31:21,530
But it's going to be a very
small substitution effect,

726
00:31:21,530 --> 00:31:23,930
because I don't have
a lot else to do if

727
00:31:23,930 --> 00:31:24,560
my wage goes down.

728
00:31:24,560 --> 00:31:27,060
Women, if it's a low
wage, why work?

729
00:31:27,060 --> 00:31:29,060
You can be much more effective
taking care of the kids or

730
00:31:29,060 --> 00:31:30,010
hanging out with your friends.

731
00:31:30,010 --> 00:31:31,150
Why work for a low wage?

732
00:31:31,150 --> 00:31:34,100
Men, there's nothing
else to do.

733
00:31:34,100 --> 00:31:38,410
So that's the relative size to
the substitution effects.

734
00:31:38,410 --> 00:31:42,770
Now, the income effect, I think,
is a little bit harder.

735
00:31:42,770 --> 00:31:45,200
And let's come back and
think about what

736
00:31:45,200 --> 00:31:46,925
drives an income effect.

737
00:31:54,320 --> 00:31:57,810
We talked about the income
effect as being delta q over

738
00:31:57,810 --> 00:31:59,870
delta y, how much a
quantity changes

739
00:31:59,870 --> 00:32:01,750
when your income changes.

740
00:32:01,750 --> 00:32:03,670
But, in reality, what's going
to matter for your income

741
00:32:03,670 --> 00:32:08,660
effect given when you start
today, is going to be not only

742
00:32:08,660 --> 00:32:13,140
delta q over delta y, how your
taste for work changed or

743
00:32:13,140 --> 00:32:19,010
income changes, but also how
hard you worked to start.

744
00:32:19,010 --> 00:32:21,090
Think of it this way.

745
00:32:21,090 --> 00:32:25,550
The income effect is how much
richer you feel if your wage

746
00:32:25,550 --> 00:32:27,370
goes up, or how much
poorer you feel if

747
00:32:27,370 --> 00:32:28,900
your wage goes down.

748
00:32:28,900 --> 00:32:35,270
If you are working 0 hours,
the income effect is 0.

749
00:32:35,270 --> 00:32:37,110
You don't feel any richer if the
wage goes up, because you

750
00:32:37,110 --> 00:32:40,390
don't earn any money.

751
00:32:40,390 --> 00:32:43,800
The more hours you work the
bigger the income effect is,

752
00:32:43,800 --> 00:32:46,490
because the bigger that
shock is to you.

753
00:32:46,490 --> 00:32:48,620
So we can think of the income
effect, a shorthand for the

754
00:32:48,620 --> 00:32:56,140
income effect, is going to be
h times dh/dy, the hours you

755
00:32:56,140 --> 00:33:00,820
work times how your hours
change with your income.

756
00:33:00,820 --> 00:33:03,600
OK?

757
00:33:03,600 --> 00:33:06,450
Now, to prove this it involves
using complicated algebra.

758
00:33:06,450 --> 00:33:07,710
We're not going to get into
it in this course.

759
00:33:07,710 --> 00:33:09,960
I worked hard last night to
see if I could make the

760
00:33:09,960 --> 00:33:11,710
algebra less complicated,
and I can't.

761
00:33:11,710 --> 00:33:16,000
I just have to try to work
intuitively on this.

762
00:33:16,000 --> 00:33:19,050
The notion of the income effect
is bigger the more

763
00:33:19,050 --> 00:33:20,080
you're in the market.

764
00:33:20,080 --> 00:33:23,220
You can think about
it for goods too.

765
00:33:23,220 --> 00:33:28,260
Think about the income effect
of a change in the price of

766
00:33:28,260 --> 00:33:30,760
something you buy a lot
of for something you

767
00:33:30,760 --> 00:33:33,090
buy very little of.

768
00:33:33,090 --> 00:33:37,210
So let's say you're someone
who's buying two Starbucks a

769
00:33:37,210 --> 00:33:43,830
day, and you very rarely
go see a movie.

770
00:33:43,830 --> 00:33:47,570
Well, if the price of a movie
goes up 10%, or the price of

771
00:33:47,570 --> 00:33:49,500
Starbucks goes up 10%,
which is going to

772
00:33:49,500 --> 00:33:51,400
make you feel poorer?

773
00:33:51,400 --> 00:33:52,940
The price of Starbucks
going up, because you

774
00:33:52,940 --> 00:33:54,540
buy a lot of Starbucks.

775
00:33:54,540 --> 00:33:57,710
So how much poorer you'll feel,
or the income effect,

776
00:33:57,710 --> 00:34:01,840
will depend on your
starting point.

777
00:34:01,840 --> 00:34:04,160
The more you're in a market,
the more you'll feel the

778
00:34:04,160 --> 00:34:05,330
income effect.

779
00:34:05,330 --> 00:34:07,250
Now, based on that, who's going
to have a bigger income

780
00:34:07,250 --> 00:34:09,050
effect, men or women?

781
00:34:09,050 --> 00:34:10,800
Same person, what
do you think?

782
00:34:10,800 --> 00:34:13,100
The income effect is going to be
stronger the more you're in

783
00:34:13,100 --> 00:34:13,920
the market.

784
00:34:13,920 --> 00:34:15,062
So who's going to have a
bigger income effect?

785
00:34:15,062 --> 00:34:15,909
AUDIENCE: Married men would have
a bigger income effect.

786
00:34:15,909 --> 00:34:16,403
PROFESSOR: Exactly.

787
00:34:16,403 --> 00:34:18,920
Married men would have a bigger
income effect, because

788
00:34:18,920 --> 00:34:19,650
they're in the market.

789
00:34:19,650 --> 00:34:21,360
Married women, most of
them don't work.

790
00:34:21,360 --> 00:34:23,480
So there's no income effect.

791
00:34:23,480 --> 00:34:30,800
So this is going to be big for
men and small for women.

792
00:34:30,800 --> 00:34:32,650
There's another issue,
which is does dh/dy

793
00:34:32,650 --> 00:34:33,440
differ for men and women?

794
00:34:33,440 --> 00:34:35,000
I'll leave that alone.

795
00:34:35,000 --> 00:34:37,489
Let's assume they both have
the same underlying income

796
00:34:37,489 --> 00:34:39,230
elasticity.

797
00:34:39,230 --> 00:34:41,380
But, certainly, the initial
hours are much bigger from men

798
00:34:41,380 --> 00:34:42,630
than for women.

799
00:34:44,560 --> 00:34:49,020
So what does this mean in terms
of the labor supply

800
00:34:49,020 --> 00:34:52,239
curves you would see for married
men and married women

801
00:34:52,239 --> 00:34:55,460
40 years ago?

802
00:34:55,460 --> 00:34:57,830
Based on these facts, what
would you think?

803
00:34:57,830 --> 00:34:58,680
Yeah?

804
00:34:58,680 --> 00:34:59,770
AUDIENCE: They would have
opposite slopes.

805
00:34:59,770 --> 00:35:00,100
PROFESSOR: Yeah.

806
00:35:00,100 --> 00:35:02,590
So, in particular, the female
labor supply curve

807
00:35:02,590 --> 00:35:03,450
would look like what?

808
00:35:03,450 --> 00:35:04,848
It would slope up or down?

809
00:35:04,848 --> 00:35:06,720
AUDIENCE: It would slope up.

810
00:35:06,720 --> 00:35:07,880
PROFESSOR: It would slope up.

811
00:35:07,880 --> 00:35:12,500
You'd have an upward sloping
curve, because you'd have

812
00:35:12,500 --> 00:35:14,820
these big substitution effects
and small income effects.

813
00:35:14,820 --> 00:35:17,530
So it would look much more
like Figure 7-4.

814
00:35:21,680 --> 00:35:23,220
You'd have the big substitution
effect when the

815
00:35:23,220 --> 00:35:25,650
wage goes up and a small
offsetting income effect.

816
00:35:25,650 --> 00:35:27,750
Think about the woman who
is not working at all.

817
00:35:27,750 --> 00:35:29,340
She's now working at
all at $8 an hour.

818
00:35:29,340 --> 00:35:31,530
You raise her wage
to $12 an hour.

819
00:35:31,530 --> 00:35:32,650
She's like, hey, I wasn't
working at all.

820
00:35:32,650 --> 00:35:33,650
So there's no income effect.

821
00:35:33,650 --> 00:35:36,390
But now I'm going to go to
work and make some money.

822
00:35:36,390 --> 00:35:37,200
So it's upward sloping.

823
00:35:37,200 --> 00:35:40,690
But for men, it's going to look
more potentially like

824
00:35:40,690 --> 00:35:42,330
Figure 7-5.

825
00:35:42,330 --> 00:35:45,050
There's a small substitution
effect but a potentially big

826
00:35:45,050 --> 00:35:49,050
income effect or bigger
than women.

827
00:35:49,050 --> 00:35:51,160
Now, how big it is,
that's not clear.

828
00:35:51,160 --> 00:35:52,460
Because, once again, men
have nothing to do

829
00:35:52,460 --> 00:35:54,010
if they don't work.

830
00:35:54,010 --> 00:35:59,660
So it could be this ends up
being bigger and smaller.

831
00:35:59,660 --> 00:36:03,230
It's not clear how big
this ends up being.

832
00:36:03,230 --> 00:36:06,540
But it's at least possible that
you could have men having

833
00:36:06,540 --> 00:36:11,020
a backward-bending or downward
sloping labor supply curve.

834
00:36:11,020 --> 00:36:12,910
Because the income effect could
even more than offset

835
00:36:12,910 --> 00:36:14,950
the substitution effect.

836
00:36:14,950 --> 00:36:17,410
But, in reality, given the way
I set up the example, you'd

837
00:36:17,410 --> 00:36:18,660
think men would basically
have a pretty

838
00:36:18,660 --> 00:36:20,650
inelastic labor supply.

839
00:36:20,650 --> 00:36:23,090
You'd think, 40 years ago, these
things would basically

840
00:36:23,090 --> 00:36:25,950
both be 0, both offset.

841
00:36:25,950 --> 00:36:28,735
And, basically, you'd have a
situation where the change in

842
00:36:28,735 --> 00:36:30,000
the wages didn't matter
much for men.

843
00:36:32,940 --> 00:36:36,540
And, in fact, that's
what people found.

844
00:36:36,540 --> 00:36:38,570
This is a wonderful case of the
convergence of truth with

845
00:36:38,570 --> 00:36:41,020
theory and a wonderful chance
to see the power of some

846
00:36:41,020 --> 00:36:43,620
pretty simplistic theory.

847
00:36:43,620 --> 00:36:46,110
The intuition is exactly borne
out in the data, which is

848
00:36:46,110 --> 00:36:47,860
males, 40 years ago,
would have a very

849
00:36:47,860 --> 00:36:49,070
inelastic labor supply.

850
00:36:49,070 --> 00:36:51,660
Their labor supply curves were
virtually vertical and maybe

851
00:36:51,660 --> 00:36:52,410
backward-bending.

852
00:36:52,410 --> 00:36:54,920
There's some controversy
on that.

853
00:36:54,920 --> 00:36:56,190
Some estimates got
backward-bending.

854
00:36:56,190 --> 00:36:56,900
Some didn't.

855
00:36:56,900 --> 00:37:00,300
But there were certainly
not upward sloping.

856
00:37:00,300 --> 00:37:02,060
It was basically vertical.

857
00:37:02,060 --> 00:37:04,460
Women had a very few elastic
labor supply.

858
00:37:04,460 --> 00:37:07,380
The elasticity is estimated
to be around 1.

859
00:37:07,380 --> 00:37:10,490
That is every 1% change
in the wage lead to

860
00:37:10,490 --> 00:37:12,360
1% more labor supply.

861
00:37:12,360 --> 00:37:15,550
So that's a fairly elastic
labor supply for women.

862
00:37:15,550 --> 00:37:19,990
Where, for men, the estimate
was basically 0.

863
00:37:19,990 --> 00:37:21,920
And that's kind of neat, because
we're actually getting

864
00:37:21,920 --> 00:37:23,230
confirmation in the data
of what the theory

865
00:37:23,230 --> 00:37:25,960
would have told us.

866
00:37:25,960 --> 00:37:30,480
Now, someone else tell me what
do you think has happened over

867
00:37:30,480 --> 00:37:36,190
the last 40 years relative to
these elasticities of married

868
00:37:36,190 --> 00:37:39,305
men and married women.

869
00:37:39,305 --> 00:37:40,736
Yeah.

870
00:37:40,736 --> 00:37:43,836
AUDIENCE: The elasticity of
married women has gone down in

871
00:37:43,836 --> 00:37:45,510
the last 40 years--

872
00:37:45,510 --> 00:37:45,915
PROFESSOR: Why?

873
00:37:45,915 --> 00:37:47,390
Speak up so the class
can hear you.

874
00:37:47,390 --> 00:37:48,868
Why is that?

875
00:37:48,868 --> 00:37:51,832
AUDIENCE: Because women
work more often now

876
00:37:51,832 --> 00:37:52,326
than they did before.

877
00:37:52,326 --> 00:37:54,230
PROFESSOR: Women work
more often now.

878
00:37:54,230 --> 00:37:58,450
So the income effect is going to
be getting bigger for them.

879
00:37:58,450 --> 00:38:00,700
So the income effect is going
up, because their

880
00:38:00,700 --> 00:38:03,100
initial h is bigger.

881
00:38:03,100 --> 00:38:06,900
Plus there's actually now,
in some sense, less good

882
00:38:06,900 --> 00:38:08,420
opportunities if you're
not working.

883
00:38:08,420 --> 00:38:10,420
So when we had our first kid,
and we lived in Brookline,

884
00:38:10,420 --> 00:38:12,660
which is sort of an urban city,
and my wife decided to

885
00:38:12,660 --> 00:38:15,660
stay at home, she didn't have
moms to hang out with.

886
00:38:15,660 --> 00:38:17,680
It was just nannies
at the park.

887
00:38:17,680 --> 00:38:19,670
And it wasn't that much fun.

888
00:38:19,670 --> 00:38:23,130
And so, basically, the
substitution effect is

889
00:38:23,130 --> 00:38:25,550
shrinking, because the outside
options aren't quite as good

890
00:38:25,550 --> 00:38:29,760
as they were, as the norms
shift towards work.

891
00:38:29,760 --> 00:38:33,580
Whereas for men, actually it's
becoming more normal for men

892
00:38:33,580 --> 00:38:36,110
to be engaged in child care.

893
00:38:36,110 --> 00:38:38,840
My best friend is a
stay-at-home dad.

894
00:38:38,840 --> 00:38:42,770
It's becoming more normal
for that to happen.

895
00:38:42,770 --> 00:38:45,360
And so the substitution
effect is rising.

896
00:38:45,360 --> 00:38:47,720
It's not implausible that if
you cut a man's wage down,

897
00:38:47,720 --> 00:38:48,440
he'll just say forget it.

898
00:38:48,440 --> 00:38:49,050
My wife's going to work.

899
00:38:49,050 --> 00:38:50,340
I'm taking care of the kids.

900
00:38:50,340 --> 00:38:53,880
That would be socially
ostracizing 40 years ago.

901
00:38:53,880 --> 00:38:56,180
But it's not that odd now.

902
00:38:56,180 --> 00:38:58,850
And, likewise, as men are less
engaged in the labor force and

903
00:38:58,850 --> 00:39:00,930
spending more time at home,
their income effects are

904
00:39:00,930 --> 00:39:04,280
falling, because their
initial h is smaller.

905
00:39:04,280 --> 00:39:06,950
So you're getting a convergence
in these labor

906
00:39:06,950 --> 00:39:09,270
supply elasticities.

907
00:39:09,270 --> 00:39:11,050
What really seems to be
happening is mostly

908
00:39:11,050 --> 00:39:13,530
convergence down for women,
not much up for men.

909
00:39:13,530 --> 00:39:15,800
So men are maybe going
from 0 to 0.1.

910
00:39:15,800 --> 00:39:18,980
Women are coming from
like 1 to 1/2.

911
00:39:18,980 --> 00:39:21,880
So what you're seeing is that
men aren't actually working

912
00:39:21,880 --> 00:39:23,310
that much less.

913
00:39:23,310 --> 00:39:24,510
There's a few stay-at-home
dads.

914
00:39:24,510 --> 00:39:25,830
But they're still not
the majority.

915
00:39:25,830 --> 00:39:28,550
Women are working a lot
more, and kids are in

916
00:39:28,550 --> 00:39:30,520
child care a lot more.

917
00:39:30,520 --> 00:39:34,020
So what you're seeing over
time is you're seeing men

918
00:39:34,020 --> 00:39:35,860
being a little more responsive,
but not that much

919
00:39:35,860 --> 00:39:36,140
more responsive.

920
00:39:36,140 --> 00:39:39,380
They're still, basically,
working all the time.

921
00:39:39,380 --> 00:39:41,230
Women are working a lot
more and being more

922
00:39:41,230 --> 00:39:43,530
responsive to wages.

923
00:39:43,530 --> 00:39:49,690
And there's a reduction coming
in both women's leisure and

924
00:39:49,690 --> 00:39:52,840
production of child
care at home.

925
00:39:52,840 --> 00:39:57,790
Now, that raises a very
interesting question of is

926
00:39:57,790 --> 00:39:59,740
this is a good thing?

927
00:39:59,740 --> 00:40:01,430
Now this is a very deep
and hard topic.

928
00:40:01,430 --> 00:40:03,340
In economics, we think if people
do something it's good,

929
00:40:03,340 --> 00:40:06,200
or they wouldn't have done it.

930
00:40:06,200 --> 00:40:08,770
It is true that if you look
at data on self-reported

931
00:40:08,770 --> 00:40:14,020
well-being or happiness data,
married women report a general

932
00:40:14,020 --> 00:40:16,680
decline in happiness, over the
last 40 years, as they've

933
00:40:16,680 --> 00:40:20,100
entered the labor force
more and more.

934
00:40:20,100 --> 00:40:24,700
And the issue is, is this
something which is a good way

935
00:40:24,700 --> 00:40:27,560
for society to spend
its resources, to

936
00:40:27,560 --> 00:40:29,535
have everyone working?

937
00:40:29,535 --> 00:40:30,720
We're consuming more.

938
00:40:30,720 --> 00:40:32,370
Consumption has gone up.

939
00:40:32,370 --> 00:40:34,180
We're consuming more, but
we're getting less

940
00:40:34,180 --> 00:40:35,830
leisure as a family.

941
00:40:35,830 --> 00:40:37,370
Because the men aren't working
that much less.

942
00:40:37,370 --> 00:40:38,940
The women are working
a lot more.

943
00:40:38,940 --> 00:40:41,910
So we're getting less
leisure as a family.

944
00:40:41,910 --> 00:40:42,720
How do we feel about
that outcome.

945
00:40:42,720 --> 00:40:43,730
That's an interesting
question.

946
00:40:43,730 --> 00:40:47,000
And we'll talk about that some
more later on in the semester.

947
00:40:47,000 --> 00:40:48,040
OK.

948
00:40:48,040 --> 00:40:49,190
Questions about this?

949
00:40:49,190 --> 00:40:50,189
Yeah.

950
00:40:50,189 --> 00:40:51,439
AUDIENCE: [INAUDIBLE PHRASE].

951
00:41:08,540 --> 00:41:10,960
PROFESSOR: That's a really
good question.

952
00:41:10,960 --> 00:41:13,530
And let me talk about that
for a couple of minutes.

953
00:41:13,530 --> 00:41:23,320
The definition of unemployment
is those employed

954
00:41:23,320 --> 00:41:25,490
over looking for work.

955
00:41:29,480 --> 00:41:32,610
If the number of people employed
does not change, and

956
00:41:32,610 --> 00:41:35,770
women suddenly want to work, and
they report to surveyors

957
00:41:35,770 --> 00:41:38,680
that they're looking
for work--

958
00:41:38,680 --> 00:41:39,750
that's the employment rate.

959
00:41:39,750 --> 00:41:40,390
I'm sorry.

960
00:41:40,390 --> 00:41:41,660
The unemployment rate--

961
00:41:41,660 --> 00:41:42,370
I'm sorry--

962
00:41:42,370 --> 00:41:46,660
is going to be those looking
over employed.

963
00:41:46,660 --> 00:41:48,310
My bad.

964
00:41:48,310 --> 00:41:50,010
The unemployment rate
is going to be those

965
00:41:50,010 --> 00:41:52,060
looking over those employed.

966
00:41:52,060 --> 00:41:53,310
So the unemployment rate is how
many people are looking

967
00:41:53,310 --> 00:41:54,490
for work over how many
are employed.

968
00:41:54,490 --> 00:42:00,000
If women start suddenly looking
for work, and there's

969
00:42:00,000 --> 00:42:02,920
no jobs to be had, that will
raise the unemployment rate.

970
00:42:02,920 --> 00:42:06,670
So one thing that's been a focus
of a lot of research has

971
00:42:06,670 --> 00:42:09,700
been do increases in the
supply of labor lead to

972
00:42:09,700 --> 00:42:12,160
increases in unemployment?

973
00:42:12,160 --> 00:42:14,310
What you've expressed is what's
often called the lump

974
00:42:14,310 --> 00:42:15,820
of labour view.

975
00:42:15,820 --> 00:42:18,700
The lump of labour view is
basically the view that

976
00:42:18,700 --> 00:42:21,870
there's a fixed box of
production in the economy.

977
00:42:21,870 --> 00:42:24,450
And as more workers come in to
fill that box, there will be

978
00:42:24,450 --> 00:42:26,100
more unemployment.

979
00:42:26,100 --> 00:42:28,940
The alternative view is that
the economy is dynamic.

980
00:42:28,940 --> 00:42:31,940
And as more women are working,
and earning income, and buying

981
00:42:31,940 --> 00:42:34,490
stuff, that makes more jobs.

982
00:42:34,490 --> 00:42:37,100
So our standard of consumption
is way higher

983
00:42:37,100 --> 00:42:38,990
than 40 years ago.

984
00:42:38,990 --> 00:42:41,430
We all have much cooler stuff
than 40 years ago.

985
00:42:41,430 --> 00:42:46,170
You have no idea how bad life
sucked 40 years ago.

986
00:42:46,170 --> 00:42:48,200
We have way better stuff.

987
00:42:48,200 --> 00:42:50,680
We have that stuff, because
women are working and making

988
00:42:50,680 --> 00:42:52,810
the income to buy it, which
means people have to make it

989
00:42:52,810 --> 00:42:54,580
which makes jobs.

990
00:42:54,580 --> 00:42:56,910
So, in fact, the existing
evidence is labor supply

991
00:42:56,910 --> 00:43:00,390
shocks do not cause unemployment
increases.

992
00:43:00,390 --> 00:43:02,010
This is something I've
worked a lot on.

993
00:43:02,010 --> 00:43:07,210
What you see, a very interesting
case is in Europe.

994
00:43:07,210 --> 00:43:10,190
In the US and all over the
world, we have assistance of

995
00:43:10,190 --> 00:43:14,460
what we call Social Security,
a term you've

996
00:43:14,460 --> 00:43:15,780
all heard, I'm sure.

997
00:43:15,780 --> 00:43:21,400
The Social Security program is
a program which provides

998
00:43:21,400 --> 00:43:25,230
income when you're retired.

999
00:43:25,230 --> 00:43:28,280
So it provides income when
you're retired to help you

1000
00:43:28,280 --> 00:43:29,760
deal with the fact that you
don't have a source of labor

1001
00:43:29,760 --> 00:43:30,100
income anymore.

1002
00:43:30,100 --> 00:43:33,060
And that's a program that
virtually every country, and

1003
00:43:33,060 --> 00:43:34,570
all developed countries
have a very generous

1004
00:43:34,570 --> 00:43:35,990
social security program.

1005
00:43:35,990 --> 00:43:39,180
But they're different in the
US than in other countries.

1006
00:43:39,180 --> 00:43:42,950
In the US, the way the social
security program works is when

1007
00:43:42,950 --> 00:43:45,990
you hit 62, you get a choice.

1008
00:43:45,990 --> 00:43:49,570
You can stop working and get
your benefits from Social

1009
00:43:49,570 --> 00:43:53,350
Security, and then you get them
every year until you die.

1010
00:43:53,350 --> 00:43:57,880
Or you can keep working, delay
getting your benefits, but

1011
00:43:57,880 --> 00:44:00,700
they'll increase what you
get to offset the delay.

1012
00:44:00,700 --> 00:44:03,800
So, in other words, if I retire
at 63 rather than 62,

1013
00:44:03,800 --> 00:44:06,380
given that I'm going to die at
the same date, I'm going to

1014
00:44:06,380 --> 00:44:08,600
get one fewer year of
benefits in my life.

1015
00:44:08,600 --> 00:44:12,390
But they raise them by 6.7%
to compensate for that.

1016
00:44:12,390 --> 00:44:14,430
So I get one fewer year of
benefits, but every year it's

1017
00:44:14,430 --> 00:44:15,570
6.7% higher.

1018
00:44:15,570 --> 00:44:17,570
And it turns out, given life
expectancy, that works out to

1019
00:44:17,570 --> 00:44:19,440
be a roughly fair deal.

1020
00:44:19,440 --> 00:44:22,080
So, basically, at 62, your
choice is I can get one more

1021
00:44:22,080 --> 00:44:24,240
year of benefits or I
get higher benefits

1022
00:44:24,240 --> 00:44:25,580
for one fewer years.

1023
00:44:25,580 --> 00:44:27,940
And that's a choice that's
a roughly fair deal.

1024
00:44:27,940 --> 00:44:28,180
OK.

1025
00:44:28,180 --> 00:44:29,030
Questions about that?

1026
00:44:29,030 --> 00:44:31,490
Am I making sense of that?

1027
00:44:31,490 --> 00:44:34,310
In Europe, it's not
a fair deal.

1028
00:44:34,310 --> 00:44:37,605
In Europe the way it works is
they say, you can get one more

1029
00:44:37,605 --> 00:44:38,390
year of benefits.

1030
00:44:38,390 --> 00:44:40,810
But if you decide to work this
year, we're not given you any

1031
00:44:40,810 --> 00:44:44,050
more in the future.

1032
00:44:44,050 --> 00:44:46,690
So let me describe how it works
in the Netherlands.

1033
00:44:46,690 --> 00:44:50,430
At age 55, the Netherlands says,
if you decide to retire

1034
00:44:50,430 --> 00:44:55,130
this year, we will replace 90%
of your wages in social

1035
00:44:55,130 --> 00:44:56,800
security payments to make sure
your income doesn't suffer

1036
00:44:56,800 --> 00:44:58,600
when you retire.

1037
00:44:58,600 --> 00:45:02,160
If you don't retire and work,
you're going to give up

1038
00:45:02,160 --> 00:45:04,240
sitting at home earning
90% of your wage.

1039
00:45:04,240 --> 00:45:07,650
That is the opportunity
cost of working.

1040
00:45:07,650 --> 00:45:10,140
It's that you have forgone the
ability to sit at home and get

1041
00:45:10,140 --> 00:45:11,920
90% of your wage.

1042
00:45:11,920 --> 00:45:15,280
So what is your net
wage if you work?

1043
00:45:15,280 --> 00:45:17,200
10% of what you would
have earned.

1044
00:45:17,200 --> 00:45:20,150
So if you're earning $20 an
hour, then your choice is you

1045
00:45:20,150 --> 00:45:24,110
can sit at home for $18 or
work for $20 an hour.

1046
00:45:24,110 --> 00:45:27,470
So your net wage for working
is $2 an hour.

1047
00:45:27,470 --> 00:45:34,380
The return to work, the
opportunity cost of leisure is

1048
00:45:34,380 --> 00:45:35,830
only $2 an hour.

1049
00:45:35,830 --> 00:45:38,030
You're only forgoing $2 an
hour by sitting at home.

1050
00:45:38,030 --> 00:45:41,390
But wait, there's more.

1051
00:45:41,390 --> 00:45:44,340
If you sit at home, you don't
have to pay the payroll taxes

1052
00:45:44,340 --> 00:45:48,370
of financing the system
that are almost 50%.

1053
00:45:48,370 --> 00:45:52,060
If you work, you have to
pay the payroll taxes.

1054
00:45:52,060 --> 00:45:55,910
Which means that if you
work, you lose money.

1055
00:45:55,910 --> 00:45:58,600
Because if you work, you forgo
getting to sit at home at 90%

1056
00:45:58,600 --> 00:46:00,480
of your wage, and you
pay a tax that's

1057
00:46:00,480 --> 00:46:02,400
about 40% of your wages.

1058
00:46:02,400 --> 00:46:06,080
So, actually, you will lose 30%
of your salary by working

1059
00:46:06,080 --> 00:46:08,080
relative to sitting at home.

1060
00:46:08,080 --> 00:46:11,140
Guess what people do in
the Netherlands at 55?

1061
00:46:11,140 --> 00:46:12,050
They sit at home.

1062
00:46:12,050 --> 00:46:15,440
No one works after 55 in the
Netherlands on the books.

1063
00:46:15,440 --> 00:46:16,810
They work off the books
painting houses

1064
00:46:16,810 --> 00:46:18,160
and doing odd jobs.

1065
00:46:18,160 --> 00:46:20,280
No one works on the
books after 55.

1066
00:46:20,280 --> 00:46:21,610
Economics works, guys.

1067
00:46:21,610 --> 00:46:23,540
If you pay your guys to stay
at home, they stay at home.

1068
00:46:26,160 --> 00:46:30,040
Now, if you ask European
politicians, why do you have

1069
00:46:30,040 --> 00:46:31,200
this screwed up system?

1070
00:46:31,200 --> 00:46:32,780
They'll say, well, it's easy.

1071
00:46:32,780 --> 00:46:34,820
We want to get those old
guys out to make jobs

1072
00:46:34,820 --> 00:46:36,470
for the young guys.

1073
00:46:36,470 --> 00:46:38,980
We need to pay those old guys
to stay at home to make jobs

1074
00:46:38,980 --> 00:46:40,490
for the young guys.

1075
00:46:40,490 --> 00:46:42,330
And then you point out, have
you noticed that Europe has

1076
00:46:42,330 --> 00:46:45,060
higher unemployment than
American, even though we don't

1077
00:46:45,060 --> 00:46:46,150
do that and you do?

1078
00:46:46,150 --> 00:46:48,140
And that's because
you're wrong.

1079
00:46:48,140 --> 00:46:49,450
It doesn't work that way.

1080
00:46:49,450 --> 00:46:52,880
Because by paying the old guys
to sit at home, you have to

1081
00:46:52,880 --> 00:46:56,540
have such high taxes that no
one makes new businesses.

1082
00:46:56,540 --> 00:46:59,470
And so there's not jobs for
the young guys to have.

1083
00:46:59,470 --> 00:47:00,510
So it's true.

1084
00:47:00,510 --> 00:47:02,500
In theory, you've made jobs for
the young guys by leaving

1085
00:47:02,500 --> 00:47:03,690
the old guys at home.

1086
00:47:03,690 --> 00:47:07,480
But by imposing the 40% tax rate
that you've had to impose

1087
00:47:07,480 --> 00:47:10,380
to make it possible to pay the
old guys to sit at home,

1088
00:47:10,380 --> 00:47:12,410
you've killed job creation
in your country.

1089
00:47:12,410 --> 00:47:14,250
And, as a result, there's
not the jobs for

1090
00:47:14,250 --> 00:47:17,180
young guys to get.

1091
00:47:17,180 --> 00:47:20,250
That's a very long-winded way
of answering your question

1092
00:47:20,250 --> 00:47:26,300
that supply, in substance,
creates its own demand.

1093
00:47:26,300 --> 00:47:28,060
So more labor supply will not
necessarily cause more

1094
00:47:28,060 --> 00:47:29,710
unemployment.

1095
00:47:29,710 --> 00:47:33,700
And we're going to talk about
one more thing before we stop.

1096
00:47:33,700 --> 00:47:35,920
I've just talked about a vast
empirical literature in how

1097
00:47:35,920 --> 00:47:39,570
people understand the effects
of wages on labor supply.

1098
00:47:39,570 --> 00:47:41,680
Well, how do they do it?

1099
00:47:41,680 --> 00:47:45,180
Well, you could say, look, we
can just look at how you earn

1100
00:47:45,180 --> 00:47:47,180
a higher wage than you do.

1101
00:47:47,180 --> 00:47:50,290
And we'll ask, do you work
harder than you?

1102
00:47:50,290 --> 00:47:53,280
And we'll say, the guys who earn
higher wages work harder.

1103
00:47:53,280 --> 00:47:55,310
If guys who earn higher wages
work harder, that means labor

1104
00:47:55,310 --> 00:47:56,600
supply slopes up.

1105
00:47:56,600 --> 00:47:58,420
If guys who earn higher wages
don't work harder, that means

1106
00:47:58,420 --> 00:47:59,880
labor supply slopes down.

1107
00:47:59,880 --> 00:48:01,540
What's wrong with that?

1108
00:48:01,540 --> 00:48:01,970
Yeah.

1109
00:48:01,970 --> 00:48:04,434
AUDIENCE: Those who are getting
paid more probably are

1110
00:48:04,434 --> 00:48:06,130
getting paid because they
want to work harder.

1111
00:48:06,130 --> 00:48:07,260
PROFESSOR: Yeah.

1112
00:48:07,260 --> 00:48:08,940
Maybe you guys are different.

1113
00:48:08,940 --> 00:48:11,270
Maybe you're talented,
and you're not.

1114
00:48:11,270 --> 00:48:13,740
And maybe because you're
talented, maybe you're driven,

1115
00:48:13,740 --> 00:48:15,580
and you're not.

1116
00:48:15,580 --> 00:48:18,520
And because you're driven, you
work harder and get paid a

1117
00:48:18,520 --> 00:48:20,150
higher wage.

1118
00:48:20,150 --> 00:48:24,190
So I'm not learning anything
about the causal effect of the

1119
00:48:24,190 --> 00:48:25,400
wage on your labor supply.

1120
00:48:25,400 --> 00:48:30,030
I've just documented a
correlation between wage and

1121
00:48:30,030 --> 00:48:31,410
labor supply.

1122
00:48:31,410 --> 00:48:34,600
How can we get the causal effect
of your wage on your

1123
00:48:34,600 --> 00:48:35,270
labor supply?

1124
00:48:35,270 --> 00:48:38,090
Well, once again, ideally
we'd run an experiment.

1125
00:48:38,090 --> 00:48:40,940
We'd assign you a higher wage.

1126
00:48:40,940 --> 00:48:42,260
We'd find someone
just like you.

1127
00:48:42,260 --> 00:48:43,590
Not you, you're not driven.

1128
00:48:43,590 --> 00:48:45,470
We find someone just like you.

1129
00:48:45,470 --> 00:48:45,560
No offense.

1130
00:48:45,560 --> 00:48:47,170
You know I'm joking.

1131
00:48:47,170 --> 00:48:50,560
We'd find someone just like you
and, randomly, by a flip

1132
00:48:50,560 --> 00:48:52,260
of a coin, assign them
a lower wage.

1133
00:48:52,260 --> 00:48:56,070
And we'd see how your labor
supply differed.

1134
00:48:56,070 --> 00:48:57,560
Now, it seems like you
couldn't do that.

1135
00:48:57,560 --> 00:48:59,330
But, in fact, the US did that.

1136
00:48:59,330 --> 00:49:01,770
In the 1970s, we ran what was
called the negative income tax

1137
00:49:01,770 --> 00:49:04,740
experiment where we literally
assigned people different wage

1138
00:49:04,740 --> 00:49:08,250
rates through taxing them
by different amounts.

1139
00:49:08,250 --> 00:49:10,990
And that was part of what gave
us this very convincing

1140
00:49:10,990 --> 00:49:14,000
evidence from 40 years ago
of these responses.

1141
00:49:14,000 --> 00:49:15,990
So where we get this is from
a real experiment we

1142
00:49:15,990 --> 00:49:17,780
ran 40 years ago.

1143
00:49:17,780 --> 00:49:20,460
The problem is that's a pretty
hard experiment to run.

1144
00:49:20,460 --> 00:49:23,850
It's pretty expensive, and
there's some ethical issues.

1145
00:49:23,850 --> 00:49:26,560
So what do you do today
to estimate that?

1146
00:49:26,560 --> 00:49:28,996
What you can do today is say,
well, we can't run the

1147
00:49:28,996 --> 00:49:29,042
experiment.

1148
00:49:29,042 --> 00:49:31,780
But the government runs
it for us every time

1149
00:49:31,780 --> 00:49:34,060
they change tax rates.

1150
00:49:34,060 --> 00:49:36,490
Because if you take two people
that are identical--

1151
00:49:36,490 --> 00:49:39,630
so let's say you and you
were identical--

1152
00:49:39,630 --> 00:49:41,880
and I change your tax rate
because you live in

1153
00:49:41,880 --> 00:49:43,000
Massachusetts.

1154
00:49:43,000 --> 00:49:45,180
I don't change your tax rate
because you live in New York.

1155
00:49:45,180 --> 00:49:47,760
I can see what happens to
you relative to you.

1156
00:49:47,760 --> 00:49:51,060
Because I've now essentially
run this experiment by the

1157
00:49:51,060 --> 00:49:51,990
government changing
someone's tax rate

1158
00:49:51,990 --> 00:49:54,500
and not someone else's.

1159
00:49:54,500 --> 00:49:57,290
That's the way we do it if we
can't run a true, randomized

1160
00:49:57,290 --> 00:49:57,740
experiment.

1161
00:49:57,740 --> 00:50:00,690
And that gives very, very
similar answers.

1162
00:50:00,690 --> 00:50:02,470
Let me stop there.

1163
00:50:02,470 --> 00:50:04,880
And we will come back.

1164
00:50:04,880 --> 00:50:10,670
Next lecture we'll talk about
applying this model.

1165
00:50:10,670 --> 00:50:12,950
So I guess in section
on Friday, we

1166
00:50:12,950 --> 00:50:14,330
review for the exam.

1167
00:50:14,330 --> 00:50:16,030
In section on Friday, we
review for the exam.

1168
00:50:16,030 --> 00:50:17,040
So show up to that.

1169
00:50:17,040 --> 00:50:18,650
And the exam is next week.

1170
00:50:18,650 --> 00:50:24,560
The exam will cover through
my next lecture.