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
00:00:23,924 --> 00:00:29,230
PROFESSOR: OK, so what we're
going to do today is the last

8
00:00:29,230 --> 00:00:31,780
in what I'd say are the
core set of lectures.

9
00:00:31,780 --> 00:00:33,910
Our core set of lectures,
we started with

10
00:00:33,910 --> 00:00:35,170
talking about the market.

11
00:00:35,170 --> 00:00:37,466
We then moved on and talked
about consumer theory and did

12
00:00:37,466 --> 00:00:39,130
a series of lectures on that.

13
00:00:39,130 --> 00:00:40,340
Now we're doing producer
theory.

14
00:00:40,340 --> 00:00:43,070
This is the last in our series
of lectures on producer theory

15
00:00:43,070 --> 00:00:45,110
and then basically we
move on to topics.

16
00:00:45,110 --> 00:00:47,020
So the remainder of the section
we'll talk about

17
00:00:47,020 --> 00:00:50,480
things like international trade,
uncertainty, equity and

18
00:00:50,480 --> 00:00:53,920
efficiency, asymmetric
information

19
00:00:53,920 --> 00:00:54,610
in insurance markets.

20
00:00:54,610 --> 00:00:57,980
We'll move on to in the last
part of the course showing you

21
00:00:57,980 --> 00:00:59,855
how you can apply what we've
learned in the basics to

22
00:00:59,855 --> 00:01:02,250
answer a bunch of interesting,
real world questions.

23
00:01:02,250 --> 00:01:05,450
So this is the last of our
core basics lectures.

24
00:01:05,450 --> 00:01:08,040
What we're going to do here is
fit in something that's fallen

25
00:01:08,040 --> 00:01:13,150
through the cracks, which is
we've talked about firms and

26
00:01:13,150 --> 00:01:15,700
their decisions about
how much to produce.

27
00:01:15,700 --> 00:01:17,820
And we've talked about
the output side.

28
00:01:17,820 --> 00:01:20,250
But we haven't talked about
the input side at all.

29
00:01:20,250 --> 00:01:24,070
How do firms decide what kind
of the inputs to use and in

30
00:01:24,070 --> 00:01:25,230
what ratio to use et cetera.

31
00:01:25,230 --> 00:01:26,560
We talked a bit about it.

32
00:01:26,560 --> 00:01:30,350
We talked about isoquants and
isocosts and doing that

33
00:01:30,350 --> 00:01:31,800
tradeoff between inputs.

34
00:01:31,800 --> 00:01:35,160
We haven't really talked about
the input markets themselves.

35
00:01:35,160 --> 00:01:37,940
So firms go and they say, look
I've done my isoquants and

36
00:01:37,940 --> 00:01:40,200
isocosts and I want
63 workers.

37
00:01:40,200 --> 00:01:42,960
Well they've then got to go to
a market for labor and hire

38
00:01:42,960 --> 00:01:45,340
those workers and how does
that actually work.

39
00:01:45,340 --> 00:01:47,830
So today what we want to
focus on is the demand

40
00:01:47,830 --> 00:01:49,760
side of input markets.

41
00:01:49,760 --> 00:01:53,520
That is, what's the actual
market analysis by which a

42
00:01:53,520 --> 00:01:56,640
firm having maximized its
profits and deciding how many

43
00:01:56,640 --> 00:02:01,940
workers it wants goes and
actually finds those workers.

44
00:02:01,940 --> 00:02:04,470
So we're going to do is talk
about demand for factors.

45
00:02:04,470 --> 00:02:06,670
In particular today we'll focus
on the demand for labor.

46
00:02:06,670 --> 00:02:08,530
Although the demand for capital,
the analysis will be

47
00:02:08,530 --> 00:02:09,789
very similar.

48
00:02:09,789 --> 00:02:13,460
But today we're going to focus
on the demand for labor.

49
00:02:13,460 --> 00:02:16,490
And what we're going to do is
begin by focusing on the

50
00:02:16,490 --> 00:02:19,540
demand for labor in a
competitive factor market.

51
00:02:19,540 --> 00:02:24,160
So we're going to begin by
talking about competitive

52
00:02:24,160 --> 00:02:25,410
factor markets.

53
00:02:36,170 --> 00:02:40,890
What I mean by that is that a
perfectly competitive factor

54
00:02:40,890 --> 00:02:45,320
market is one where, just as
perfect competition and output

55
00:02:45,320 --> 00:02:48,260
markets means there's lots of
sellers selling the same good,

56
00:02:48,260 --> 00:02:51,710
a perfectly competitive factor
market means there's lots of

57
00:02:51,710 --> 00:02:54,790
sellers, in this case workers,
selling an identical good.

58
00:02:54,790 --> 00:02:56,000
That is their labor.

59
00:02:56,000 --> 00:02:58,230
So the notion is we're in a
market where there's many,

60
00:02:58,230 --> 00:03:00,770
many workers firms could hire,
all of whom are equally

61
00:03:00,770 --> 00:03:02,640
qualified for a job.

62
00:03:02,640 --> 00:03:05,310
So this is not, obviously,
a high-tech market.

63
00:03:05,310 --> 00:03:07,740
This is some low-tech,
construction, other sort of

64
00:03:07,740 --> 00:03:10,630
blue collar market, where
there's lots of workers out

65
00:03:10,630 --> 00:03:13,930
there who could equally well
be qualified for a job.

66
00:03:13,930 --> 00:03:16,280
In fact what we're going to
assume is that there's a

67
00:03:16,280 --> 00:03:19,100
perfectly flat labor
supply curve.

68
00:03:19,100 --> 00:03:22,040
Let's assume a perfectly flat
labor supply curve.

69
00:03:22,040 --> 00:03:24,720
Perfectly elastic labor supply
just to make life easy.

70
00:03:24,720 --> 00:03:26,140
Obviously it's not
true in reality.

71
00:03:26,140 --> 00:03:27,770
Let's assume we're looking at
some market with perfectly

72
00:03:27,770 --> 00:03:31,320
elastic labor supply.

73
00:03:31,320 --> 00:03:36,090
Now how do we think about what
happens in factor markets in

74
00:03:36,090 --> 00:03:36,520
that world?

75
00:03:36,520 --> 00:03:38,290
Well once again let's start
with the short run.

76
00:03:42,030 --> 00:03:46,040
So in the short run
capital's fixed.

77
00:03:46,040 --> 00:03:50,520
So a firm has said, look, I've
done my short run profit

78
00:03:50,520 --> 00:03:52,780
maximization, my isoquants
and isocosts.

79
00:03:52,780 --> 00:03:54,790
I've decided how many workers
I want given a

80
00:03:54,790 --> 00:03:57,310
fixed level of capital.

81
00:03:57,310 --> 00:04:02,210
And that gives me some
demand for labor.

82
00:04:02,210 --> 00:04:05,130
I can derive a demand for labor
curve by essentially

83
00:04:05,130 --> 00:04:09,270
saying, at different wage rates,
given a fixed capital

84
00:04:09,270 --> 00:04:14,620
price, that will shift my
isocost curve, going back to

85
00:04:14,620 --> 00:04:17,269
the producer theory, at
different wage rates that will

86
00:04:17,269 --> 00:04:19,769
shift my isocost curve, that
will cause me to demand

87
00:04:19,769 --> 00:04:20,670
different amounts of labor.

88
00:04:20,670 --> 00:04:26,830
So that traces out a demand
curve for labor.

89
00:04:26,830 --> 00:04:30,550
And we can see that graphically
in figure 18-1.

90
00:04:30,550 --> 00:04:34,920
So you've got a perfectly
elastic labor supply curve and

91
00:04:34,920 --> 00:04:38,870
then you've got a downward
sloping labor demand curve.

92
00:04:38,870 --> 00:04:43,060
And that labor demand curve
comes from the profit

93
00:04:43,060 --> 00:04:44,330
maximization.

94
00:04:44,330 --> 00:04:46,190
The other way to think about
how we get there though is

95
00:04:46,190 --> 00:04:46,590
interesting.

96
00:04:46,590 --> 00:04:52,100
You say, how do we think about
the marginal benefit versus

97
00:04:52,100 --> 00:04:54,595
the marginal cost of hiring
another worker.

98
00:04:54,595 --> 00:04:58,150
We know the marginal cost is
the wage, that's easy.

99
00:04:58,150 --> 00:05:02,590
What's the marginal benefit
of hiring another worker?

100
00:05:02,590 --> 00:05:08,150
Well recall that another unit of
labor raises output by the

101
00:05:08,150 --> 00:05:10,100
marginal product of labor.

102
00:05:10,100 --> 00:05:11,420
Remember the marginal
product of labor.

103
00:05:11,420 --> 00:05:13,120
We talked about this
a while ago.

104
00:05:13,120 --> 00:05:19,690
This is delta q delta l.

105
00:05:19,690 --> 00:05:24,310
So the next worker raises your
output by an amount marginal

106
00:05:24,310 --> 00:05:25,940
product of labor.

107
00:05:25,940 --> 00:05:29,190
That's what you get from
the next worker.

108
00:05:29,190 --> 00:05:31,820
But that's a quantity.

109
00:05:31,820 --> 00:05:33,380
The firm cares about profit
not quantity.

110
00:05:33,380 --> 00:05:35,480
So what it cares about
is revenues.

111
00:05:35,480 --> 00:05:40,190
So the revenues from the next
worker would be the marginal

112
00:05:40,190 --> 00:05:44,860
revenues that are made on that
next unit times the marginal

113
00:05:44,860 --> 00:05:47,430
product of labor.

114
00:05:47,430 --> 00:05:51,290
That's the marginal benefit to
the firm of the next unit of

115
00:05:51,290 --> 00:05:55,240
labor is the marginal product
that that worker produces

116
00:05:55,240 --> 00:05:57,630
times the marginal revenue
the firm raises from

117
00:05:57,630 --> 00:05:59,970
selling that next unit.

118
00:05:59,970 --> 00:06:03,370
So they have to consider two
things, two margins.

119
00:06:03,370 --> 00:06:07,910
How much is it worth them to
sell that next unit and how

120
00:06:07,910 --> 00:06:11,540
much will be produced by that
next unit of worker?

121
00:06:11,540 --> 00:06:26,100
So if we have a perfectly
competitive labor market, the

122
00:06:26,100 --> 00:06:28,740
marginal cost is going
to be the wage.

123
00:06:28,740 --> 00:06:31,640
So we're going to set this
equal to the wage.

124
00:06:31,640 --> 00:06:34,640
So in a perfectly competitive
labor market the marginal cost

125
00:06:34,640 --> 00:06:37,140
of another worker is the wage
and this is the marginal

126
00:06:37,140 --> 00:06:38,520
benefit of another worker.

127
00:06:38,520 --> 00:06:40,630
So this is going to be the
condition the firm's going to

128
00:06:40,630 --> 00:06:42,240
use to decide how many
workers to hire.

129
00:06:47,300 --> 00:06:49,570
We're going to call
this the marginal

130
00:06:49,570 --> 00:06:52,340
revenue product of labor.

131
00:06:52,340 --> 00:06:54,260
So you're going to set the
marginal revenue product of

132
00:06:54,260 --> 00:06:55,430
labor equal to the wage.

133
00:06:55,430 --> 00:06:57,400
That's going to be your profit
maximizing condition--

134
00:06:57,400 --> 00:06:58,760
the marginal revenue product.

135
00:06:58,760 --> 00:07:01,040
The marginal product is
about quantities.

136
00:07:01,040 --> 00:07:02,990
Marginal revenue product
is about dollars.

137
00:07:02,990 --> 00:07:05,900
What's the dollars that the next
unit of labor produces

138
00:07:05,900 --> 00:07:09,390
for you is your marginal
revenue product.

139
00:07:09,390 --> 00:07:15,310
Now if the output market is also
perfectly competitive.

140
00:07:15,310 --> 00:07:17,720
That is, imagine for a minute
now, take one further step.

141
00:07:17,720 --> 00:07:21,175
Not a perfectly competitive
labor supply but also a

142
00:07:21,175 --> 00:07:23,620
perfectly competitive
output market.

143
00:07:23,620 --> 00:07:25,740
So it's not a monopolist. It's
selling in a perfectly

144
00:07:25,740 --> 00:07:26,620
competitive output market.

145
00:07:26,620 --> 00:07:28,780
Then we know what the
marginal revenue is.

146
00:07:28,780 --> 00:07:31,310
We know the marginal revenue
is the price.

147
00:07:31,310 --> 00:07:37,860
So for a perfectly competitive
output market we could rewrite

148
00:07:37,860 --> 00:07:40,570
this as price times marginal
product of

149
00:07:40,570 --> 00:07:44,810
labor equals the wage.

150
00:07:44,810 --> 00:07:46,620
Because we know the marginal
revenue is the price in a

151
00:07:46,620 --> 00:07:47,920
perfectly competitive
output market.

152
00:07:51,020 --> 00:07:53,710
And basically this makes it even
easier to see which is to

153
00:07:53,710 --> 00:07:56,740
say, look, how many workers
do you hire?

154
00:07:56,740 --> 00:08:00,610
You hire until the wage you pay
that worker is equal to

155
00:08:00,610 --> 00:08:03,540
the price you sell your good for
times how many goods that

156
00:08:03,540 --> 00:08:04,790
worker produces.

157
00:08:06,975 --> 00:08:10,190
If one worker produces 100 goods
and each good sells for

158
00:08:10,190 --> 00:08:15,430
100, then you'll only pay
the worker $10,000.

159
00:08:15,430 --> 00:08:19,725
Basically that is going
to determine your

160
00:08:19,725 --> 00:08:20,870
labor demand curve.

161
00:08:20,870 --> 00:08:26,150
And so the labor demand curve
is also labeled the marginal

162
00:08:26,150 --> 00:08:27,540
revenue product of labor
curve, I'm sorry.

163
00:08:31,300 --> 00:08:32,355
Why is this diminishing?

164
00:08:32,355 --> 00:08:33,650
Why is it downward sloping?

165
00:08:33,650 --> 00:08:35,820
Because remember the marginal
product of labor's

166
00:08:35,820 --> 00:08:37,380
diminishing.

167
00:08:37,380 --> 00:08:39,830
The marginal product of
labor's diminishing.

168
00:08:39,830 --> 00:08:45,050
As a result this curve
is downward sloping.

169
00:08:45,050 --> 00:08:47,910
Now here price is fixed,
so it doesn't matter.

170
00:08:47,910 --> 00:08:49,980
Marginal revenue also is
diminishing so that's going to

171
00:08:49,980 --> 00:08:50,950
make it even more downward
sloping than the

172
00:08:50,950 --> 00:08:52,060
more general case.

173
00:08:52,060 --> 00:08:54,370
But in this specific case of
perfectly competitive output

174
00:08:54,370 --> 00:08:57,140
markets where this is just a
constant price you get this

175
00:08:57,140 --> 00:08:59,760
downward sloping marginal
revenue product of labor curve

176
00:08:59,760 --> 00:09:02,770
because the marginal product of
labor is diminishing, as we

177
00:09:02,770 --> 00:09:04,020
talked about.

178
00:09:09,180 --> 00:09:13,240
So that's the analysis of what
we see for a perfectly

179
00:09:13,240 --> 00:09:15,320
competitive factor market.

180
00:09:15,320 --> 00:09:19,610
So the equilibrium is where the
labor supply curve, which

181
00:09:19,610 --> 00:09:23,480
is perfectly elastic, intersects
this marginal

182
00:09:23,480 --> 00:09:26,870
revenue product curve, which
is determined by how

183
00:09:26,870 --> 00:09:28,880
productive the workers are and
how much money they're making

184
00:09:28,880 --> 00:09:30,520
for you with each unit
they produce.

185
00:09:30,520 --> 00:09:31,770
And that gets you the short
run equilibrium.

186
00:09:39,910 --> 00:09:41,290
Questions about that?

187
00:09:41,290 --> 00:09:42,460
Questions about what
we're doing here?

188
00:09:42,460 --> 00:09:46,500
So that just says the underlying
analysis of where

189
00:09:46,500 --> 00:09:50,780
demand for labor comes from or
where the equilibrium level of

190
00:09:50,780 --> 00:09:51,480
labor is going to come.

191
00:09:51,480 --> 00:09:53,470
It's going to come from
intersection of this demand

192
00:09:53,470 --> 00:09:56,080
with the supply,
which is flat.

193
00:09:56,080 --> 00:09:59,620
Now how is this going to
differ in the long run?

194
00:09:59,620 --> 00:10:01,730
Well let me ask the question
this way, forget the math,

195
00:10:01,730 --> 00:10:03,810
forget the graphs, I'm just
going to ask you intuitively.

196
00:10:03,810 --> 00:10:09,100
In the long run, do we think
the long run demand for

197
00:10:09,100 --> 00:10:12,790
workers will be more elastic or
less elastic than the short

198
00:10:12,790 --> 00:10:17,020
run demand, in general.

199
00:10:17,020 --> 00:10:19,080
Will the long run demand curve,
this is short run

200
00:10:19,080 --> 00:10:21,500
demand curve, will the long run
demand curve typically be

201
00:10:21,500 --> 00:10:23,690
more elastic or less elastic
than the short run demand

202
00:10:23,690 --> 00:10:26,240
curve for workers?

203
00:10:26,240 --> 00:10:26,830
Somebody take a guess.

204
00:10:26,830 --> 00:10:27,524
Yeah.

205
00:10:27,524 --> 00:10:28,736
AUDIENCE: More elastic.

206
00:10:28,736 --> 00:10:29,690
PROFESSOR: More elastic.

207
00:10:29,690 --> 00:10:30,940
Why?

208
00:10:33,142 --> 00:10:34,380
It's more elastic.

209
00:10:34,380 --> 00:10:35,767
But intuitively, don't worry
about the graph, I'm just

210
00:10:35,767 --> 00:10:36,320
looking for intuition.

211
00:10:36,320 --> 00:10:36,615
Yeah.

212
00:10:36,615 --> 00:10:38,006
AUDIENCE: Because in
the long run you

213
00:10:38,006 --> 00:10:38,790
can substitute capital.

214
00:10:38,790 --> 00:10:39,960
PROFESSOR: Exactly.

215
00:10:39,960 --> 00:10:42,960
More substitutability equals
more elasticity.

216
00:10:42,960 --> 00:10:45,380
General intuition you want to
remember for this course.

217
00:10:45,380 --> 00:10:48,820
In the long run if I can
substitute towards capital,

218
00:10:48,820 --> 00:10:51,360
then that long run demand curve
for labor will be even

219
00:10:51,360 --> 00:10:53,140
more elastic than the short
run demand curve.

220
00:10:53,140 --> 00:10:54,470
I'm not going to work through
the math or anything.

221
00:10:54,470 --> 00:10:59,960
I just want you to remember that
intuition that when you

222
00:10:59,960 --> 00:11:03,790
have more margins you can use,
that's more substitutability,

223
00:11:03,790 --> 00:11:05,040
that means more elasticity.

224
00:11:08,390 --> 00:11:13,810
So the idea is in the short run,
if the wage increases for

225
00:11:13,810 --> 00:11:17,850
workers all you can do is if
you hire fewer workers you

226
00:11:17,850 --> 00:11:20,270
just produce less, so you're
sort of stuck.

227
00:11:20,270 --> 00:11:22,140
But in the long run if the wage
decreases you just say,

228
00:11:22,140 --> 00:11:24,640
fine, I'll just use
machines instead.

229
00:11:24,640 --> 00:11:26,590
So in the long run you can
substitute away from workers

230
00:11:26,590 --> 00:11:28,010
towards machines.

231
00:11:28,010 --> 00:11:30,910
So in the long run your
demand curve is

232
00:11:30,910 --> 00:11:32,160
going to be more elastic.

233
00:11:34,920 --> 00:11:37,710
Questions about that?

234
00:11:37,710 --> 00:11:40,590
Now this is all relatively
straightforward, just follows

235
00:11:40,590 --> 00:11:42,540
from producer theory.

236
00:11:42,540 --> 00:11:46,470
The sort of stuff you had to
do in the exam last night.

237
00:11:46,470 --> 00:11:50,540
Let's now talk about a little
more interesting case which is

238
00:11:50,540 --> 00:11:53,310
one of my other favorite words
in economics which is the case

239
00:11:53,310 --> 00:11:53,970
of monopsony.

240
00:11:53,970 --> 00:11:56,410
Not monopoly, but monopsony.

241
00:12:02,240 --> 00:12:05,270
We've been talking in lectures
about monopoly which is the

242
00:12:05,270 --> 00:12:08,700
case where one firm
is the only seller

243
00:12:08,700 --> 00:12:11,230
in the output market.

244
00:12:11,230 --> 00:12:13,490
One firm is the only seller
in the output market.

245
00:12:13,490 --> 00:12:16,780
A parallel case in input
markets is the case of

246
00:12:16,780 --> 00:12:22,230
monopsony which is when one firm
is the only demander in

247
00:12:22,230 --> 00:12:25,040
the input market.

248
00:12:25,040 --> 00:12:28,050
So monopoly is when one firm
is the only seller in the

249
00:12:28,050 --> 00:12:29,160
output market.

250
00:12:29,160 --> 00:12:34,120
The parallel in input markets
is monopsony which is where

251
00:12:34,120 --> 00:12:38,330
one firm is the only buyer
in the input market.

252
00:12:38,330 --> 00:12:46,010
And the key thing that's going
to drive monopsony is that

253
00:12:46,010 --> 00:12:52,930
when there are barriers to exit
from a factor market it's

254
00:12:52,930 --> 00:12:54,400
going to create a monopsony.

255
00:12:54,400 --> 00:12:56,440
And any time there are barriers
to exit, any time

256
00:12:56,440 --> 00:12:59,960
workers are stuck and cannot
leave a market, workers are

257
00:12:59,960 --> 00:13:02,750
stuck working one place, that
will give the employer market

258
00:13:02,750 --> 00:13:05,560
power over those workers.

259
00:13:05,560 --> 00:13:09,490
The classic example is
the company town.

260
00:13:09,490 --> 00:13:14,510
In the 1800s when there were
mining operations and they'd

261
00:13:14,510 --> 00:13:16,740
come in and they'd
hire people.

262
00:13:16,740 --> 00:13:17,940
And basically there
was nowhere else

263
00:13:17,940 --> 00:13:18,700
to work in the area.

264
00:13:18,700 --> 00:13:20,695
These were areas which were
dying agricultural areas there

265
00:13:20,695 --> 00:13:21,910
was nowhere else to work.

266
00:13:21,910 --> 00:13:23,460
You'd go work for the
mining company.

267
00:13:23,460 --> 00:13:25,940
That mining company had
tremendous monopsony power

268
00:13:25,940 --> 00:13:28,900
over you because basically there
was nowhere else to work

269
00:13:28,900 --> 00:13:29,740
within a decent area.

270
00:13:29,740 --> 00:13:30,490
There weren't cars yet.

271
00:13:30,490 --> 00:13:33,200
You couldn't just commute
to somewhere else.

272
00:13:33,200 --> 00:13:37,920
A modern example is MIT's
monopsony power over me.

273
00:13:37,920 --> 00:13:40,930
MIT has monopsony
power over me.

274
00:13:40,930 --> 00:13:41,940
Why is that?

275
00:13:41,940 --> 00:13:45,420
Well that's because my life's
pretty comfortable now.

276
00:13:45,420 --> 00:13:47,050
I'm in a house I've lived
in a long time.

277
00:13:47,050 --> 00:13:48,260
My kids are in schools I like.

278
00:13:48,260 --> 00:13:49,880
I've got a lot of friends.

279
00:13:49,880 --> 00:13:52,720
It'd be a real pain in the
ass for me to move.

280
00:13:52,720 --> 00:13:57,840
And as a result MIT has some,
not unlimited, don't tell

281
00:13:57,840 --> 00:14:00,410
them, but they have some
monopsony power over me

282
00:14:00,410 --> 00:14:03,970
because they know that that's
a barrier to my exit.

283
00:14:03,970 --> 00:14:07,480
To my exiting MIT a barrier
is that I'm very

284
00:14:07,480 --> 00:14:09,830
comfortable and satisfied.

285
00:14:09,830 --> 00:14:15,030
And given the nature tenets of
psychology, it is harder to

286
00:14:15,030 --> 00:14:18,180
move someone with a pull
than with the push.

287
00:14:18,180 --> 00:14:21,200
So someone could come and try
to pull me away, but they're

288
00:14:21,200 --> 00:14:22,910
going to have to blow me
away with an offer--

289
00:14:22,910 --> 00:14:24,490
once again, don't
tell MIT this--

290
00:14:24,490 --> 00:14:25,490
they'll have to blow
me away an offer

291
00:14:25,490 --> 00:14:27,490
because I'm pretty satisfied.

292
00:14:27,490 --> 00:14:30,760
And that satisfaction inherently
gives MIT some

293
00:14:30,760 --> 00:14:32,220
monopsony power over me.

294
00:14:35,420 --> 00:14:38,720
Now the question is what
implications does this have.

295
00:14:38,720 --> 00:14:41,140
And the implications are
quite interesting.

296
00:14:41,140 --> 00:14:43,630
And what they are is a
complicated flip of the

297
00:14:43,630 --> 00:14:44,330
monopoly case.

298
00:14:44,330 --> 00:14:46,760
Let's look at figure 18-2.

299
00:14:46,760 --> 00:14:48,460
Here's an example of
a company town.

300
00:14:51,220 --> 00:14:54,000
Now let's imagine a labor market
where you've got some

301
00:14:54,000 --> 00:14:56,260
labor demand curve, the marginal
revenue product of

302
00:14:56,260 --> 00:15:00,530
labor, MRPL, and some
labor supply curve.

303
00:15:00,530 --> 00:15:02,480
And now labor supply
is upward sloping.

304
00:15:02,480 --> 00:15:07,930
This is a not a perfectly
competitive labor market now.

305
00:15:07,930 --> 00:15:10,560
This labor supply is
now upward sloping.

306
00:15:10,560 --> 00:15:14,800
So the demand curve for
labor we're going to

307
00:15:14,800 --> 00:15:16,790
say is 60 minus l.

308
00:15:16,790 --> 00:15:20,740
So in our example the marginal
revenue product of labor is

309
00:15:20,740 --> 00:15:22,690
going to be 60 minus l.

310
00:15:22,690 --> 00:15:25,140
The notion is you have some
downward sloping demand curve

311
00:15:25,140 --> 00:15:28,105
for labor because of diminishing
marginal product.

312
00:15:38,220 --> 00:15:41,270
The wage you're willing to pay
is diminishing in the number

313
00:15:41,270 --> 00:15:41,800
of workers.

314
00:15:41,800 --> 00:15:44,050
To get that first worker
you'll pay a high wage,

315
00:15:44,050 --> 00:15:45,310
because you got to produce
something.

316
00:15:45,310 --> 00:15:47,030
But as you hire more workers
the wage you're

317
00:15:47,030 --> 00:15:48,140
willing to pay falls.

318
00:15:48,140 --> 00:15:50,640
So the demand curve is
downward sloping.

319
00:15:50,640 --> 00:15:53,490
And let's say that our labor
supply curve is that the

320
00:15:53,490 --> 00:15:56,230
amount of labor workers are
willing to supply is

321
00:15:56,230 --> 00:15:57,680
the wage over 2.

322
00:15:57,680 --> 00:15:58,950
I'm just making this up.

323
00:15:58,950 --> 00:16:02,080
These are just made up numbers
just to make the math work.

324
00:16:02,080 --> 00:16:04,590
So this is it just an upward
sloping labor supply curve.

325
00:16:04,590 --> 00:16:06,460
A higher wage calls
forth more labor.

326
00:16:08,960 --> 00:16:16,110
So let's ask what happens
in the competitive case.

327
00:16:16,110 --> 00:16:18,820
Well in the competitive case
you set the supply equal to

328
00:16:18,820 --> 00:16:19,590
the demand.

329
00:16:19,590 --> 00:16:24,860
Well supply is that firms are
going to set the wage equals

330
00:16:24,860 --> 00:16:27,170
60 minus l.

331
00:16:27,170 --> 00:16:29,850
Labor supply is l
equals w over 2.

332
00:16:33,020 --> 00:16:35,790
So we could just have two
equations and two unknowns we

333
00:16:35,790 --> 00:16:39,050
can solve and get that the
amount of labor supplied in

334
00:16:39,050 --> 00:16:41,520
the competitive case
is 20 units.

335
00:16:41,520 --> 00:16:44,320
20 hours, 20 days,
weeks, whatever.

336
00:16:44,320 --> 00:16:47,600
20 units.

337
00:16:47,600 --> 00:16:49,280
That's the competitive
outcome.

338
00:16:49,280 --> 00:16:54,060
And the wage we can read off
the labor supply curve.

339
00:16:54,060 --> 00:16:55,980
If they're going to supply 20
units, they're going to need a

340
00:16:55,980 --> 00:16:58,030
wage of 40.

341
00:16:58,030 --> 00:17:00,520
So once again we can read that
off the labor supply or the

342
00:17:00,520 --> 00:17:04,440
labor demand curve if there's
going to be 20 workers the

343
00:17:04,440 --> 00:17:05,700
wage is going to be 40.

344
00:17:05,700 --> 00:17:09,089
So the wage, labor competitive
and the wage

345
00:17:09,089 --> 00:17:11,260
competitive is 40.

346
00:17:11,260 --> 00:17:13,740
So in a competitive market with
this demand and supply

347
00:17:13,740 --> 00:17:16,170
curve you should know by now you
just set them equal, you

348
00:17:16,170 --> 00:17:21,630
solve, you get an outcome
of 20 workers

349
00:17:21,630 --> 00:17:23,930
working a wage of 40.

350
00:17:23,930 --> 00:17:25,869
Now let's imagine this isn't
the competitive case.

351
00:17:25,869 --> 00:17:27,359
Let's imagine it's the
monopsony case.

352
00:17:27,359 --> 00:17:29,140
Let's imagine this firm
has monopsony power.

353
00:17:29,140 --> 00:17:32,020
Workers can't exit.

354
00:17:32,020 --> 00:17:35,830
And let's further assume that
the firm cannot wage

355
00:17:35,830 --> 00:17:37,110
discriminate.

356
00:17:37,110 --> 00:17:38,760
Just as we talked about
monopolists that couldn't

357
00:17:38,760 --> 00:17:41,900
price discriminate, we're going
to talk about a firm

358
00:17:41,900 --> 00:17:43,280
they can't wage discriminate.

359
00:17:43,280 --> 00:17:45,870
It has to pay one wage to
all of its workers.

360
00:17:45,870 --> 00:17:47,400
And we'll come back once again
with this assumption.

361
00:17:47,400 --> 00:17:49,800
Just as we talked about price
discrimination we'll come back

362
00:17:49,800 --> 00:17:51,220
and talk about wage
discrimination.

363
00:17:51,220 --> 00:17:54,740
But for now assume a non-wage
discriminating monopsonist.

364
00:17:54,740 --> 00:17:58,500
They have to pay one wage
to all their workers.

365
00:17:58,500 --> 00:18:04,470
Well what that means is just as
when a monopolist wanted to

366
00:18:04,470 --> 00:18:08,020
sell more goods it had to
lower the price, if a

367
00:18:08,020 --> 00:18:11,270
monopsonist wants to hire
more workers it has

368
00:18:11,270 --> 00:18:14,150
to raise the wage.

369
00:18:14,150 --> 00:18:14,810
Parallel thing.

370
00:18:14,810 --> 00:18:16,950
Just as the monopolist had to
lower the price to sell more

371
00:18:16,950 --> 00:18:19,730
units because it had to respect
the demand curve, a

372
00:18:19,730 --> 00:18:22,460
monopsonist if it wants to hire
more workers has to raise

373
00:18:22,460 --> 00:18:26,440
the wage because it has to
respect the supply curve--

374
00:18:26,440 --> 00:18:27,270
parallel.

375
00:18:27,270 --> 00:18:32,000
And what this will do is that
will lead them to under hire

376
00:18:32,000 --> 00:18:34,920
workers ct too low a wage.

377
00:18:34,920 --> 00:18:38,460
Just as monopoly led firms to
under produce at too high a

378
00:18:38,460 --> 00:18:41,550
price, monopsony will
lead to under hiring

379
00:18:41,550 --> 00:18:43,660
at too high a wage.

380
00:18:43,660 --> 00:18:46,220
So once again, to think about
this, let's think about the

381
00:18:46,220 --> 00:18:50,460
firm's decision to hire
an extra worker.

382
00:18:50,460 --> 00:18:53,020
What is the firm's total
expenditure on labor?

383
00:18:53,020 --> 00:18:57,170
Its expenditure on labor is the
wage, which is a function

384
00:18:57,170 --> 00:19:00,640
of the amount of labor, times
the amount of labor.

385
00:19:00,640 --> 00:19:03,430
That's the expenditure
on labor.

386
00:19:03,430 --> 00:19:08,230
So its marginal expenditure, if
you take the derivative, is

387
00:19:08,230 --> 00:19:16,050
going to be w plus
dw/dl times l.

388
00:19:16,050 --> 00:19:17,300
That's its marginal
expenditure.

389
00:19:19,630 --> 00:19:22,040
If you want to hire an
additional worker what's the

390
00:19:22,040 --> 00:19:23,270
marginal cost.

391
00:19:23,270 --> 00:19:24,510
Well I want to hire
an additional

392
00:19:24,510 --> 00:19:25,000
worker, what's the cost?

393
00:19:25,000 --> 00:19:30,050
I've got a pay him w and to hire
him I have to raise the

394
00:19:30,050 --> 00:19:36,040
wage, so I have to pay all my
previous workers more as well.

395
00:19:36,040 --> 00:19:38,770
So to hire one more worker I've
got to pay that worker a

396
00:19:38,770 --> 00:19:42,120
wage and in order to entice
him I've got to raise the

397
00:19:42,120 --> 00:19:43,850
wage, which means I've got to
pay a higher wage to all my

398
00:19:43,850 --> 00:19:45,040
previous workers too.

399
00:19:45,040 --> 00:19:48,680
Once again, remember, I
have to pay one wage.

400
00:19:48,680 --> 00:19:52,610
So if I'm going to hire that
worker there is the same

401
00:19:52,610 --> 00:19:56,240
poisoning effect that we
saw with monopolists.

402
00:19:56,240 --> 00:19:59,390
With monopolists the poisoning
effect was if I want to sell

403
00:19:59,390 --> 00:20:01,350
one more unit I'm going to have
to undercut my price on

404
00:20:01,350 --> 00:20:03,150
all previous units.

405
00:20:03,150 --> 00:20:06,622
For a monopsonist, if I want to
hire one more worker I have

406
00:20:06,622 --> 00:20:09,160
to pay all my previous
workers more.

407
00:20:09,160 --> 00:20:11,330
And that's going to mean that
there's a poisoning effective

408
00:20:11,330 --> 00:20:12,220
in reverse.

409
00:20:12,220 --> 00:20:14,350
That's going to cost me
a lot of money to

410
00:20:14,350 --> 00:20:17,140
hire that extra worker.

411
00:20:17,140 --> 00:20:20,870
So we can actually derive now
a marginal expenditure curve

412
00:20:20,870 --> 00:20:23,100
just as we derived a marginal
revenue curve for the

413
00:20:23,100 --> 00:20:24,310
monopolist.

414
00:20:24,310 --> 00:20:28,040
The marginal expenditure
curve.

415
00:20:28,040 --> 00:20:33,010
So we know expenditure
is w of l times l.

416
00:20:33,010 --> 00:20:42,090
And we know from the supply
curve, we can rewrite this as

417
00:20:42,090 --> 00:20:44,960
w equals 2l.

418
00:20:44,960 --> 00:20:46,980
So that says that the
expenditure on

419
00:20:46,980 --> 00:20:51,170
labor is 2l times l.

420
00:20:51,170 --> 00:20:55,240
So plugging in from the supply
curve the expenditure on labor

421
00:20:55,240 --> 00:20:56,750
is 2l times l.

422
00:20:56,750 --> 00:20:58,880
Its wage is a function
of labor times labor.

423
00:20:58,880 --> 00:21:00,930
So 2l times l.

424
00:21:00,930 --> 00:21:06,030
So that means that marginal
expenditure is 4l.

425
00:21:06,030 --> 00:21:10,240
The marginal expenditure
is 4l.

426
00:21:10,240 --> 00:21:13,430
So we can now draw a marginal
expenditure curve that's

427
00:21:13,430 --> 00:21:14,980
steeper than the labor
supply curve.

428
00:21:14,980 --> 00:21:17,280
Once again it's confusing
but it's all parallel.

429
00:21:17,280 --> 00:21:18,570
It's just we're flipping
everything around from

430
00:21:18,570 --> 00:21:22,620
monopolist. Instead of drawing
that marginal revenue curve

431
00:21:22,620 --> 00:21:26,160
that was steeper than the
product demand curve, now

432
00:21:26,160 --> 00:21:28,600
we're drawing a marginal
expenditure curve which is

433
00:21:28,600 --> 00:21:31,920
steeper than the labor
supply curve.

434
00:21:31,920 --> 00:21:36,150
And once again to find the
outcome we'll find the

435
00:21:36,150 --> 00:21:39,010
intersection of that marginal
expenditure curve with

436
00:21:39,010 --> 00:21:41,910
marginal cost.

437
00:21:41,910 --> 00:21:43,760
Well here we'll find the
intersection of marginal

438
00:21:43,760 --> 00:21:45,700
expenditure curve, I'm sorry,
with labor demand.

439
00:21:45,700 --> 00:21:47,510
I'm sorry, so the parallel was
we found the intersection of

440
00:21:47,510 --> 00:21:50,070
marginal revenue with marginal
cost. Now find the

441
00:21:50,070 --> 00:21:53,030
intersection of marginal
expenditure with labor demand.

442
00:21:53,030 --> 00:21:57,280
That intersection is going
to happen at 12 workers.

443
00:21:57,280 --> 00:22:01,360
So the firm is going
to hire 12 workers.

444
00:22:01,360 --> 00:22:03,960
But what ways you're going to
pay-- once again our first

445
00:22:03,960 --> 00:22:06,390
temptation is to look at that
high intersection and say,

446
00:22:06,390 --> 00:22:08,260
well at that intersection
what's the wage.

447
00:22:08,260 --> 00:22:10,140
We didn't put that on the
diagram for a reason.

448
00:22:10,140 --> 00:22:12,160
Maybe we should have
to throw you off.

449
00:22:12,160 --> 00:22:14,570
Remember, to figure out the wage
you've got to respect the

450
00:22:14,570 --> 00:22:15,820
labor supply curve.

451
00:22:15,820 --> 00:22:18,180
Just like you have to respect
the product demand curve to

452
00:22:18,180 --> 00:22:19,480
figure out the price.

453
00:22:19,480 --> 00:22:24,050
So what's the wage when I hire
12 workers they pay 24.

454
00:22:24,050 --> 00:22:28,180
So the monopsonist is going to
hire 12 workers and pay 24.

455
00:22:28,180 --> 00:22:29,860
It's going to hire the number
of workers where marginal

456
00:22:29,860 --> 00:22:33,620
expenditure equals
the labor demand.

457
00:22:33,620 --> 00:22:34,970
That's going to determine
the quantity.

458
00:22:34,970 --> 00:22:36,480
And the wage they're going
to read off the

459
00:22:36,480 --> 00:22:37,730
labor supply curve.

460
00:22:40,090 --> 00:22:43,710
And as a result this firm is
going to under hire at too low

461
00:22:43,710 --> 00:22:47,110
a wage relative to the
perfect competition.

462
00:22:47,110 --> 00:22:48,540
Relative to the competition
they're going to hire fewer

463
00:22:48,540 --> 00:22:50,215
workers at a lower wage.

464
00:22:53,090 --> 00:22:56,380
So if you think about this,
let's come back to

465
00:22:56,380 --> 00:22:59,680
the example of MIT.

466
00:22:59,680 --> 00:23:04,990
MIT, say, would like to expand
the economics department.

467
00:23:04,990 --> 00:23:07,780
But to do so it's got to poach
an economics professor away

468
00:23:07,780 --> 00:23:09,740
from another university.

469
00:23:09,740 --> 00:23:11,390
Well, it poaches another
professor away from another

470
00:23:11,390 --> 00:23:13,190
university.

471
00:23:13,190 --> 00:23:17,590
And let's say that at MIT pays
all its professors the same.

472
00:23:17,590 --> 00:23:18,880
If they're going to poach
a professor from another

473
00:23:18,880 --> 00:23:21,030
university they're going
to have to pay

474
00:23:21,030 --> 00:23:23,680
the rest of us more.

475
00:23:23,680 --> 00:23:25,160
And that's going to cost
them a ton of money.

476
00:23:25,160 --> 00:23:26,900
So they think think we'd rather
have a little bit more

477
00:23:26,900 --> 00:23:28,630
crowded undergrad class, we
don't really see it here

478
00:23:28,630 --> 00:23:32,750
today, but maybe in general,
more crowded undergrad class

479
00:23:32,750 --> 00:23:34,540
and not get that extra professor
to avoid having to

480
00:23:34,540 --> 00:23:38,220
pay a higher wage to all our
existing professors.

481
00:23:38,220 --> 00:23:44,270
So as a result MIT will under
hire professors and they'll

482
00:23:44,270 --> 00:23:49,020
under hire professors
at too low a wage.

483
00:23:49,020 --> 00:23:54,220
And once you can determine how
big monopsony power is, what

484
00:23:54,220 --> 00:23:57,090
determined how big monopoly
power was?

485
00:23:57,090 --> 00:24:00,768
What was the key thing I don't
want to say what it is because

486
00:24:00,768 --> 00:24:02,460
it will give it away, what's the
key thing that determined

487
00:24:02,460 --> 00:24:04,600
the size of monopoly power.

488
00:24:04,600 --> 00:24:04,800
Yeah.

489
00:24:04,800 --> 00:24:05,420
AUDIENCE: Elasticity demand.

490
00:24:05,420 --> 00:24:06,880
PROFESSOR: Elasticity
of demand.

491
00:24:06,880 --> 00:24:10,490
So just like that, the key
factors going to determine the

492
00:24:10,490 --> 00:24:12,970
market power of monopsonist is
going to be the elasticity of

493
00:24:12,970 --> 00:24:15,140
supply of labor.

494
00:24:15,140 --> 00:24:18,660
The elasticity of demand for a
good is what determined the

495
00:24:18,660 --> 00:24:21,236
market power of monopolist
because as goods were more

496
00:24:21,236 --> 00:24:23,110
elasticity demanded, they
had less ability

497
00:24:23,110 --> 00:24:25,240
to jack up the price.

498
00:24:25,240 --> 00:24:27,600
The elasticity of supply of
labor is what's going to

499
00:24:27,600 --> 00:24:29,500
determine the market power
of monopsonists.

500
00:24:29,500 --> 00:24:31,690
Because if I have
more options.

501
00:24:31,690 --> 00:24:34,220
they can't underpay me.

502
00:24:34,220 --> 00:24:38,580
So if I'm very willing to move
to another university.

503
00:24:38,580 --> 00:24:42,040
That is as my labor supply
curve gets flatter, then

504
00:24:42,040 --> 00:24:45,750
there's less market power that
they have. In particular, in a

505
00:24:45,750 --> 00:24:47,460
perfectly competitive labor
market there's no monopsony

506
00:24:47,460 --> 00:24:49,180
power at all.

507
00:24:49,180 --> 00:24:52,250
So monopsony power is a function
of the options facing

508
00:24:52,250 --> 00:24:53,500
their workers.

509
00:24:56,340 --> 00:24:57,590
Questions about that?

510
00:25:00,440 --> 00:25:03,970
Now the key question this all
raises, at least when I first

511
00:25:03,970 --> 00:25:08,130
learned about it and think
about it, is it maybe was

512
00:25:08,130 --> 00:25:10,080
plausible to think that
monopolists can only charge

513
00:25:10,080 --> 00:25:11,710
one price for their good.

514
00:25:11,710 --> 00:25:15,210
That the iPod is what
the iPod costs.

515
00:25:15,210 --> 00:25:17,250
And you couldn't start charging
different amounts for

516
00:25:17,250 --> 00:25:18,380
iPods to different people.

517
00:25:18,380 --> 00:25:22,200
That would get bad
press and stuff.

518
00:25:22,200 --> 00:25:24,120
But it's seems a little bit
stranger to think that

519
00:25:24,120 --> 00:25:27,450
employers have to pay one wage
to all their workers.

520
00:25:27,450 --> 00:25:29,604
As a matter of fact we know that
MIT doesn't pay the same

521
00:25:29,604 --> 00:25:32,370
to all its professors.

522
00:25:32,370 --> 00:25:35,470
There is some wage
discrimination.

523
00:25:35,470 --> 00:25:38,450
In particular, it's a well known
fact that the way to get

524
00:25:38,450 --> 00:25:40,870
a raise as a professor is to
get an offer from another

525
00:25:40,870 --> 00:25:42,130
university.

526
00:25:42,130 --> 00:25:45,740
Because MIT, the pay structure
professors at competitive

527
00:25:45,740 --> 00:25:49,030
departments like economics is
typically they will underpay

528
00:25:49,030 --> 00:25:51,470
you until the university comes
and says you're worth more,

529
00:25:51,470 --> 00:25:54,900
and then they'll ratchet up
to try to match them.

530
00:25:54,900 --> 00:25:57,120
So there is wage discrimination
in practice at

531
00:25:57,120 --> 00:25:58,880
universities as there is
in most workplaces.

532
00:25:58,880 --> 00:26:00,420
There's very few workplaces
where all

533
00:26:00,420 --> 00:26:01,300
workers make the same.

534
00:26:01,300 --> 00:26:03,200
There's wage discrimination.

535
00:26:03,200 --> 00:26:06,160
So does that mean this
model is irrelevant?

536
00:26:06,160 --> 00:26:07,550
And the answer is,
no, it doesn't.

537
00:26:07,550 --> 00:26:11,750
Because there are still major
barriers to perfect wage

538
00:26:11,750 --> 00:26:12,330
discrimination.

539
00:26:12,330 --> 00:26:14,420
There's some wage
discrimination, but there's a

540
00:26:14,420 --> 00:26:17,750
lot of barriers to perfect
wage discrimination.

541
00:26:17,750 --> 00:26:22,410
The most important one is
workplace norms or fairness.

542
00:26:22,410 --> 00:26:26,310
So what MIT should do, here's
the MIT optimal strategy.

543
00:26:26,310 --> 00:26:29,100
The older the professor, the
less they should pay them.

544
00:26:29,100 --> 00:26:30,460
Not because they're
less productive.

545
00:26:30,460 --> 00:26:32,430
Marginal productivity is
constant, it's not, we get

546
00:26:32,430 --> 00:26:35,400
less productive as we get older,
but put that aside.

547
00:26:35,400 --> 00:26:38,770
It's because the older you are,
the less likely you are

548
00:26:38,770 --> 00:26:40,820
to get up and move.

549
00:26:40,820 --> 00:26:42,640
And the less likely, quite
frankly, other universities

550
00:26:42,640 --> 00:26:44,120
are going to want to hire
you and take you away.

551
00:26:44,120 --> 00:26:45,030
Because no one gets
that excited about

552
00:26:45,030 --> 00:26:47,270
hiring a 60 year old.

553
00:26:47,270 --> 00:26:49,607
So what you should do is take
all the 60 over professors and

554
00:26:49,607 --> 00:26:53,220
say we're cutting your wage
in half or by a third.

555
00:26:53,220 --> 00:26:57,910
Because the truth, we've written
our lectures already,

556
00:26:57,910 --> 00:27:01,450
the wage is already well above
our marginal product.

557
00:27:01,450 --> 00:27:02,950
We're all doing pretty
well anyway.

558
00:27:02,950 --> 00:27:04,980
And the truth is people
wouldn't leave.

559
00:27:04,980 --> 00:27:07,100
So why doesn't MIT do that?

560
00:27:07,100 --> 00:27:09,790
MIT doesn't do that because I'm
going to someday be one of

561
00:27:09,790 --> 00:27:12,200
those old guys.

562
00:27:12,200 --> 00:27:13,710
I'm getting there rapidly.

563
00:27:13,710 --> 00:27:15,910
I say, wait a second, if they're
going to do to me when

564
00:27:15,910 --> 00:27:19,790
I'm 60, I'm going to get out of
here while I'm 45 because I

565
00:27:19,790 --> 00:27:20,950
don't want to be in
that situation.

566
00:27:20,950 --> 00:27:22,740
Because that's unfair.

567
00:27:22,740 --> 00:27:25,610
Workplace norms matter.

568
00:27:25,610 --> 00:27:28,210
Employers really do not like
to discriminate within the

569
00:27:28,210 --> 00:27:31,830
workplace because it breeds bad
blood and ultimately can

570
00:27:31,830 --> 00:27:34,120
lower productivity.

571
00:27:34,120 --> 00:27:36,370
And this is something that we
miss in our basic models.

572
00:27:36,370 --> 00:27:39,760
We don't have fairness and
workplace norms in our models.

573
00:27:39,760 --> 00:27:42,220
So wage is just about setting
the wage that maximizes the

574
00:27:42,220 --> 00:27:44,290
profit, which means screwing
the 60 year olds.

575
00:27:44,290 --> 00:27:47,195
But in fact, in reality,
that's not the

576
00:27:47,195 --> 00:27:48,040
way workplaces work.

577
00:27:48,040 --> 00:27:49,830
And that's what the point
of labor economics is.

578
00:27:49,830 --> 00:27:52,640
If you're interested in this we
have an excellent course in

579
00:27:52,640 --> 00:27:56,080
labor economics that follows
up on these issues.

580
00:27:56,080 --> 00:27:58,400
But a key issue is how much
wage discrimination can be

581
00:27:58,400 --> 00:28:01,600
done given workplace norms,
given the notions of fairness

582
00:28:01,600 --> 00:28:04,710
we have. And the answer is it
might be kind of tough.

583
00:28:04,710 --> 00:28:10,100
Because basically MIT doesn't
want to worry about upsetting

584
00:28:10,100 --> 00:28:14,420
all its younger faculty by
mistreating its older faculty.

585
00:28:14,420 --> 00:28:16,220
And part of that could
be solidarity.

586
00:28:16,220 --> 00:28:17,560
Some of those guys are
my friends and I

587
00:28:17,560 --> 00:28:18,930
feel bad for them.

588
00:28:18,930 --> 00:28:22,090
But partly it could be just more
selfish which is, I don't

589
00:28:22,090 --> 00:28:25,670
want to be at a place that's
going to discriminate in that

590
00:28:25,670 --> 00:28:28,450
way against me when
I get older.

591
00:28:28,450 --> 00:28:30,770
And that's just something that's
missed by the basic

592
00:28:30,770 --> 00:28:32,870
14.01 models.

593
00:28:32,870 --> 00:28:34,280
Questions about that?

594
00:28:34,280 --> 00:28:34,565
Yeah.

595
00:28:34,565 --> 00:28:36,678
AUDIENCE: Couldn't you
model the person's

596
00:28:36,678 --> 00:28:38,506
life wage or something?

597
00:28:38,506 --> 00:28:40,450
PROFESSOR: So yeah.

598
00:28:40,450 --> 00:28:46,260
So what you can do is you could
say to people, look,

599
00:28:46,260 --> 00:28:49,640
when we're hiring you we're
going to overpay you relative

600
00:28:49,640 --> 00:28:52,080
to other universities when
you're young and we'll

601
00:28:52,080 --> 00:28:54,970
underpay you when
you're older.

602
00:28:54,970 --> 00:28:57,960
And so on a lifetime basis
you'll be fine.

603
00:28:57,960 --> 00:29:00,760
What would the problem
with that be?

604
00:29:00,760 --> 00:29:01,880
So let's say the say-- yeah.

605
00:29:01,880 --> 00:29:02,680
AUDIENCE: You'd leave.

606
00:29:02,680 --> 00:29:05,900
PROFESSOR: I'd leave as soon as
I reached that point where

607
00:29:05,900 --> 00:29:08,380
I was being paid more elsewhere
that didn't pay this

608
00:29:08,380 --> 00:29:09,420
downward slope.

609
00:29:09,420 --> 00:29:11,530
And, in fact, there's a lot of
interesting labor economics

610
00:29:11,530 --> 00:29:13,570
theory which says the optimal
formal labor contract is

611
00:29:13,570 --> 00:29:15,010
actually the opposite.

612
00:29:15,010 --> 00:29:16,720
It's to overpay when
you're old and

613
00:29:16,720 --> 00:29:17,800
underpay when you're young.

614
00:29:17,800 --> 00:29:19,070
And the notion is
to get people to

615
00:29:19,070 --> 00:29:20,990
want to stick around.

616
00:29:20,990 --> 00:29:23,240
And so in some sense, common
labor theory says exactly the

617
00:29:23,240 --> 00:29:24,450
opposite of what
you suggested.

618
00:29:24,450 --> 00:29:25,960
You want to overpay older
people to get

619
00:29:25,960 --> 00:29:27,010
them to stick around.

620
00:29:27,010 --> 00:29:28,850
And for many years in America
that's often the way labor

621
00:29:28,850 --> 00:29:29,390
markets worked.

622
00:29:29,390 --> 00:29:31,440
We had very generous pensions
and health benefits and high

623
00:29:31,440 --> 00:29:32,650
wages for older workers.

624
00:29:32,650 --> 00:29:35,510
That equilibrium is now breaking
down because in this

625
00:29:35,510 --> 00:29:37,620
more competitive labor market
you can't afford to overpay

626
00:29:37,620 --> 00:29:40,060
those older workers because
then you can't attract the

627
00:29:40,060 --> 00:29:41,130
younger workers in
the first place.

628
00:29:41,130 --> 00:29:43,650
And we're moving towards a
flatter profile by age.

629
00:29:43,650 --> 00:29:45,750
But that's exactly the set of
interesting issues we have to

630
00:29:45,750 --> 00:29:48,920
deal with in labor economics in
setting pay that we don't

631
00:29:48,920 --> 00:29:50,240
really get into here.

632
00:29:50,240 --> 00:29:53,250
Other questions or
comments on that?

633
00:29:53,250 --> 00:29:55,530
Of course there's another every
reason why MIT couldn't

634
00:29:55,530 --> 00:29:58,600
do this, which is it's
against the law.

635
00:29:58,600 --> 00:30:01,890
We have age discrimination laws
in our country which say

636
00:30:01,890 --> 00:30:03,740
you cannot discriminate
against people based

637
00:30:03,740 --> 00:30:05,340
purely on their age.

638
00:30:05,340 --> 00:30:07,920
MIT could say, well look,
I could demonstrate a

639
00:30:07,920 --> 00:30:08,930
productivity difference.

640
00:30:08,930 --> 00:30:10,800
This guy is publishing fewer
articles than he did 20 years

641
00:30:10,800 --> 00:30:14,690
ago et cetera, but they do
have a legal hurdle to

642
00:30:14,690 --> 00:30:16,550
overcome as well as an
administrative hurdle.

643
00:30:16,550 --> 00:30:18,540
It would be a pain in the ass
for MIT to have to figure out

644
00:30:18,540 --> 00:30:22,070
exactly how to shift their wage
schedule to do all this.

645
00:30:22,070 --> 00:30:24,240
And that causes administrative
costs and extra Deans and

646
00:30:24,240 --> 00:30:27,440
extra things and they just don't
want to deal with it.

647
00:30:27,440 --> 00:30:30,460
So there are other barriers as
well which is administrative

648
00:30:30,460 --> 00:30:31,610
costs and legal costs.

649
00:30:31,610 --> 00:30:34,160
But I think probably the main
barrier is just the difficult

650
00:30:34,160 --> 00:30:38,930
issue of workplace norms
and how it affects the

651
00:30:38,930 --> 00:30:41,556
productivity of the workers who
are behind while you get

652
00:30:41,556 --> 00:30:43,360
rid of these older workers or
underpay these older workers

653
00:30:43,360 --> 00:30:44,610
who aren't as productive.

654
00:30:47,120 --> 00:30:50,040
Questions or thoughts on that?

655
00:30:50,040 --> 00:30:53,840
So now with this monopsony model
in mind I want to go

656
00:30:53,840 --> 00:31:00,020
back and revisit a major topic
that we talked about early in

657
00:31:00,020 --> 00:31:01,350
the course.

658
00:31:01,350 --> 00:31:03,460
And a couple times in this
course we talked about as an

659
00:31:03,460 --> 00:31:06,820
example of how governments can
screw up markets we talked

660
00:31:06,820 --> 00:31:08,880
about the minimum wage.

661
00:31:08,880 --> 00:31:12,360
We talked about if you take a
competitive labor market and

662
00:31:12,360 --> 00:31:17,350
impose a minimum wage above the
competitive level, that

663
00:31:17,350 --> 00:31:19,610
could lead to deadweight loss.

664
00:31:19,610 --> 00:31:22,270
Because what will happen is at
that higher wage firms will

665
00:31:22,270 --> 00:31:24,080
want fewer workers.

666
00:31:24,080 --> 00:31:26,340
Workers who would be happy to
work at the competitive wage

667
00:31:26,340 --> 00:31:28,130
will not be able to work.

668
00:31:28,130 --> 00:31:31,050
Trades which would make social
welfare higher won't be made.

669
00:31:31,050 --> 00:31:33,800
And there will be
deadweight loss.

670
00:31:33,800 --> 00:31:37,850
The monopsony model says that
may not be the case.

671
00:31:37,850 --> 00:31:41,860
Because in the monopsony model,
a minimum wage can play

672
00:31:41,860 --> 00:31:44,770
the same role that optimal price
regulation paid with

673
00:31:44,770 --> 00:31:45,960
monopolists.

674
00:31:45,960 --> 00:31:49,705
Remember with monopolists we
said if you regulated a

675
00:31:49,705 --> 00:31:52,460
monopolist and forced them to
charge a competitive price,

676
00:31:52,460 --> 00:31:54,130
that you could actually
force them into

677
00:31:54,130 --> 00:31:56,090
the competitive outcome.

678
00:31:56,090 --> 00:32:02,260
Well if a minimum wage is set
above the prevailing wage but

679
00:32:02,260 --> 00:32:05,140
that's because a monopsony
prevailing wage is too low,

680
00:32:05,140 --> 00:32:07,560
and the minimum wage is set at
the competitive level, then

681
00:32:07,560 --> 00:32:09,610
you could actually increase
employment and improve

682
00:32:09,610 --> 00:32:11,710
outcomes with a minimum wage.

683
00:32:11,710 --> 00:32:12,560
Sort of counter intuitive.

684
00:32:12,560 --> 00:32:14,180
So let's look at it,
pretty confusing.

685
00:32:14,180 --> 00:32:16,010
Let's look at figure 18-3.

686
00:32:16,010 --> 00:32:20,660
Figure 18-3 once again parallels
a figure you saw on

687
00:32:20,660 --> 00:32:24,320
the entire flip side for price
regulation of a monopolist,

688
00:32:24,320 --> 00:32:27,380
this is the parallel which
is wage regulation of a

689
00:32:27,380 --> 00:32:29,020
monopsonist. Let's walk
through this.

690
00:32:29,020 --> 00:32:32,460
It's pretty confusing so let's
walk through this slowly.

691
00:32:32,460 --> 00:32:37,610
Initially you have a monopsonist
who is hiring at

692
00:32:37,610 --> 00:32:42,210
the point where their marginal
expenditure curve, which is

693
00:32:42,210 --> 00:32:43,880
the dashed line and then
the solid line.

694
00:32:43,880 --> 00:32:49,190
So the line me1 is original
marginal expenditure curve.

695
00:32:49,190 --> 00:32:52,990
It's the me1 plus the me2,
it's that segment.

696
00:32:52,990 --> 00:32:55,080
That line which is, once again,
more elastic than the

697
00:32:55,080 --> 00:32:56,480
supply curve.

698
00:32:56,480 --> 00:33:01,560
That intersects the demand curve
at a labor supply l1.

699
00:33:01,560 --> 00:33:06,770
So they hire l1 workers and
they pay a wage w1.

700
00:33:06,770 --> 00:33:10,150
That's the initial monopsony
equilibrium.

701
00:33:10,150 --> 00:33:13,680
The competitive equilibrium is
where supply equals demand.

702
00:33:13,680 --> 00:33:16,180
That would be hiring l2
workers and paying

703
00:33:16,180 --> 00:33:18,930
a higher wage w2.

704
00:33:18,930 --> 00:33:25,320
So the monopsonist is under
hiring relative to the

705
00:33:25,320 --> 00:33:27,030
competitive firm.

706
00:33:27,030 --> 00:33:29,280
Now let's say the government
rolls in and says we're going

707
00:33:29,280 --> 00:33:32,210
to set a minimum wage and we're
going to happen to get

708
00:33:32,210 --> 00:33:34,530
it right and set it at the
competitive wage level, w2.

709
00:33:37,080 --> 00:33:39,460
Well now let's think about
the monopsonist calculus.

710
00:33:39,460 --> 00:33:42,970
The monopsonist new marginal
expenditure curve is the old

711
00:33:42,970 --> 00:33:44,140
one for the solid segment.

712
00:33:44,140 --> 00:33:47,220
So where it says me2, that solid
segment is still the

713
00:33:47,220 --> 00:33:48,260
marginal expenditure curve.

714
00:33:48,260 --> 00:33:50,780
So as they think about hiring
additional workers, they're

715
00:33:50,780 --> 00:33:52,170
working down that curve.

716
00:33:52,170 --> 00:33:56,420
But once they get to l2 workers
they can't lower the

717
00:33:56,420 --> 00:33:57,900
wage anymore.

718
00:33:57,900 --> 00:34:01,560
So that marginal expenditure
curve, they can no longer

719
00:34:01,560 --> 00:34:03,720
lower the wage below w2.

720
00:34:03,720 --> 00:34:05,990
So their new marginal
expenditure curve hops down

721
00:34:05,990 --> 00:34:08,139
and becomes the minimum wage.

722
00:34:08,139 --> 00:34:09,790
So the new marginal expenditure
curve is the two

723
00:34:09,790 --> 00:34:10,860
segments labeled me2.

724
00:34:10,860 --> 00:34:14,530
It's the horizontal segment
from the y-axis to e2.

725
00:34:14,530 --> 00:34:18,600
And then it jumps up to that
upward sloping segment to the

726
00:34:18,600 --> 00:34:20,199
right of l2.

727
00:34:20,199 --> 00:34:23,750
So the new marginal expenditure
curve is basically

728
00:34:23,750 --> 00:34:26,040
at a wage that's above minimum
wage they continue to behave

729
00:34:26,040 --> 00:34:29,040
like a monopsonist. But once you
hit the minimum wage and

730
00:34:29,040 --> 00:34:32,400
they can't lower the wage
anymore, what happens?

731
00:34:32,400 --> 00:34:35,350
The poisoning effect goes away,
just like we talked

732
00:34:35,350 --> 00:34:41,570
about with the monopolist.
Essentially what this has done

733
00:34:41,570 --> 00:34:43,190
has killed the poisoning
effect.

734
00:34:43,190 --> 00:34:45,659
Because it said as you're
thinking about hiring that

735
00:34:45,659 --> 00:34:52,790
next worker to the right of l1,
typically say I want to

736
00:34:52,790 --> 00:34:55,460
hire them because I'm going to
have to raise my wage to

737
00:34:55,460 --> 00:34:56,170
everybody else.

738
00:34:56,170 --> 00:34:58,250
But you're already paying
everybody else a higher wage.

739
00:34:58,250 --> 00:34:59,100
You're already paying everybody

740
00:34:59,100 --> 00:35:00,320
else the minimum wage.

741
00:35:00,320 --> 00:35:01,290
So there's no poisoning
effect.

742
00:35:01,290 --> 00:35:02,770
You're not going to have to pay
them a higher wage to hire

743
00:35:02,770 --> 00:35:04,060
that next worker because you're
already paying them

744
00:35:04,060 --> 00:35:05,550
that higher wage.

745
00:35:05,550 --> 00:35:08,720
So just as optimal price
regulation undoes the

746
00:35:08,720 --> 00:35:12,890
poisoning effect on the demand
side, optimal wage regulation

747
00:35:12,890 --> 00:35:15,610
undoes that poisoning effect
on the supply side and can

748
00:35:15,610 --> 00:35:17,100
lead you to the optimal
outcome.

749
00:35:17,100 --> 00:35:20,520
So minimum wage can actually
increase employment.

750
00:35:20,520 --> 00:35:21,500
Pretty bizarre.

751
00:35:21,500 --> 00:35:24,095
Minimum wage is our whipping boy
for this course about how

752
00:35:24,095 --> 00:35:28,560
it causes dead weight loss and
leads to lower employment.

753
00:35:28,560 --> 00:35:31,390
Here we're saying, actually,
if we start a monopsony

754
00:35:31,390 --> 00:35:34,980
equilibrium, a minimum wage
could increase employment.

755
00:35:34,980 --> 00:35:38,440
Of course, as you should know,
the minimum wage could reduce

756
00:35:38,440 --> 00:35:40,455
employment even in a monopsony
setting if it

757
00:35:40,455 --> 00:35:42,080
gets set too high.

758
00:35:42,080 --> 00:35:45,900
So if the minimum wage
got set very high--

759
00:35:45,900 --> 00:35:47,010
Jessica maybe this is something
we should actually

760
00:35:47,010 --> 00:35:50,020
add for next year-- so if you
can imagine a minimum wage

761
00:35:50,020 --> 00:35:56,050
that's set very high, that could
lead to a level of labor

762
00:35:56,050 --> 00:36:00,550
supply that's actually below
the monopsony level.

763
00:36:00,550 --> 00:36:10,020
So just as with optimal price
regulation, we talked about

764
00:36:10,020 --> 00:36:12,470
how setting a price too high can
make things worse, setting

765
00:36:12,470 --> 00:36:14,380
a minimum wage too high
can make things worse.

766
00:36:14,380 --> 00:36:17,290
So it's ultimately an empirical
question of does the

767
00:36:17,290 --> 00:36:20,320
minimum wage raise employment,
which would require two

768
00:36:20,320 --> 00:36:23,540
conditions, a monopsony market
and a well-set minimum wage.

769
00:36:23,540 --> 00:36:25,120
Or does it lower employment?

770
00:36:25,120 --> 00:36:27,850
Which can happen either with a
competitive labor market or it

771
00:36:27,850 --> 00:36:31,220
could happen with a poorly-set
minimum wage.

772
00:36:31,220 --> 00:36:33,130
This pretty confusing so you'll
probably have to go

773
00:36:33,130 --> 00:36:33,470
home and think more
about this.

774
00:36:33,470 --> 00:36:35,240
But are there questions now
about anything that's not

775
00:36:35,240 --> 00:36:36,490
apparent from the diagram?

776
00:36:38,980 --> 00:36:46,210
So now this is why we have
empirical economics.

777
00:36:46,210 --> 00:36:47,710
We have empirical
economics, well

778
00:36:47,710 --> 00:36:48,740
really we have two reasons.

779
00:36:48,740 --> 00:36:50,550
One, is sometimes we know the
direction of what we're

780
00:36:50,550 --> 00:36:52,780
looking for, we want measure
its magnitude.

781
00:36:52,780 --> 00:36:54,700
Sometimes we don't even know the
direction, not to mention

782
00:36:54,700 --> 00:36:55,650
the magnitude.

783
00:36:55,650 --> 00:36:57,420
Here's a case we don't even
know the direction.

784
00:36:57,420 --> 00:36:59,490
Will increasing the minimum wage
raise or lower employment

785
00:36:59,490 --> 00:37:01,870
and by how much?

786
00:37:01,870 --> 00:37:03,630
Well how do we test this?

787
00:37:03,630 --> 00:37:06,050
Well the traditional way to
look at it to say, OK, the

788
00:37:06,050 --> 00:37:08,430
minimum wage changes over
time, let's look at what

789
00:37:08,430 --> 00:37:14,020
happens to employment when
the minimum wage goes up.

790
00:37:14,020 --> 00:37:17,700
And what people found was an
increase in minimum wage

791
00:37:17,700 --> 00:37:20,690
tended to be associated
with lower employment.

792
00:37:20,690 --> 00:37:23,510
When the minimum wage went
up, employment fell.

793
00:37:23,510 --> 00:37:28,160
So people took that to mean
that there was a situation

794
00:37:28,160 --> 00:37:30,750
where we're either in a
competitive market or we're

795
00:37:30,750 --> 00:37:33,000
screwing up the monopsony market
by setting too high a

796
00:37:33,000 --> 00:37:33,650
minimum wage.

797
00:37:33,650 --> 00:37:36,860
Because we raised the minimum
wage, employment fell.

798
00:37:36,860 --> 00:37:39,400
And what is wrong with
drawing that

799
00:37:39,400 --> 00:37:40,420
conclusion from that evidence?

800
00:37:40,420 --> 00:37:42,280
Or could someone tell me a story
about why that might not

801
00:37:42,280 --> 00:37:44,500
be a convincing piece
of evidence?

802
00:37:44,500 --> 00:37:47,180
Why the fact that when we
raise the minimum wage

803
00:37:47,180 --> 00:37:50,190
employment falls, why that
may not be by itself be

804
00:37:50,190 --> 00:37:51,010
compelling.

805
00:37:51,010 --> 00:37:53,110
What problem you might have with
that piece of evidence.

806
00:37:53,110 --> 00:37:57,360
If you read that in the New York
Times tomorrow, look at

807
00:37:57,360 --> 00:37:59,820
this graph, we raised the wage
and employment falls.

808
00:37:59,820 --> 00:38:01,630
Clearly minimum wage is bad.

809
00:38:01,630 --> 00:38:04,211
What should your first thought
be upon reading that article?

810
00:38:09,400 --> 00:38:10,660
Well what do we care about?

811
00:38:10,660 --> 00:38:13,680
We care about causation
not correlation.

812
00:38:13,680 --> 00:38:16,960
So what could be causing this
to be correlation but not a

813
00:38:16,960 --> 00:38:18,210
causation effect?

814
00:38:25,490 --> 00:38:26,020
Someone want to try?

815
00:38:26,020 --> 00:38:29,613
AUDIENCE: It would depend
on when you

816
00:38:29,613 --> 00:38:30,920
raise the minimum wage.

817
00:38:30,920 --> 00:38:33,245
PROFESSOR: Right, in particular
what story might

818
00:38:33,245 --> 00:38:34,474
cause this effect?

819
00:38:34,474 --> 00:38:38,538
AUDIENCE: If they're in a
depression and they decide to

820
00:38:38,538 --> 00:38:39,090
raise the minimum wage.

821
00:38:39,090 --> 00:38:42,050
PROFESSOR: Right, what if
governments, worried about

822
00:38:42,050 --> 00:38:44,790
workers in bad economies, that's
exactly when they raise

823
00:38:44,790 --> 00:38:45,450
the minimum wage.

824
00:38:45,450 --> 00:38:46,950
What if the government raised
the minimum wage in bad

825
00:38:46,950 --> 00:38:48,780
economy, because that's exactly
when they're worried

826
00:38:48,780 --> 00:38:50,380
about workers suffering.

827
00:38:50,380 --> 00:38:53,010
Then you would see that a
higher minimum wage is

828
00:38:53,010 --> 00:38:53,800
associated with lower
employment.

829
00:38:53,800 --> 00:38:55,830
But it's got nothing to do with
the higher minimum wage.

830
00:38:55,830 --> 00:38:57,320
It's got to do with the fact
that you raised the minimum

831
00:38:57,320 --> 00:38:59,490
wage when unemployment
is falling anyway.

832
00:38:59,490 --> 00:39:02,090
It's causation versus
correlation.

833
00:39:02,090 --> 00:39:05,160
You have to be critical reader
of evidence like this.

834
00:39:05,160 --> 00:39:08,460
The minimum wage is not
handed down by God.

835
00:39:08,460 --> 00:39:11,960
It's determined by legislators
who are subject to political

836
00:39:11,960 --> 00:39:14,340
pressures which may depend
the state of the economy.

837
00:39:14,340 --> 00:39:19,030
If the minimum wage increases
when employment happens to be

838
00:39:19,030 --> 00:39:22,610
falling, then it will look
like the minimum wage has

839
00:39:22,610 --> 00:39:25,810
harmed employment when
it really hasn't.

840
00:39:25,810 --> 00:39:27,910
So what we do about this?

841
00:39:30,430 --> 00:39:36,190
What we do about this is to try
to think about a way we

842
00:39:36,190 --> 00:39:41,030
can find a causal relationship
between the minimum wage and

843
00:39:41,030 --> 00:39:42,350
employment.

844
00:39:42,350 --> 00:39:48,830
And one way to do that is to
try to find cases where a

845
00:39:48,830 --> 00:39:52,720
minimum wage increased and
yet we know there was no

846
00:39:52,720 --> 00:39:55,960
independent change in
economic activity.

847
00:39:55,960 --> 00:39:58,900
And the way economists have
done that is by looking at

848
00:39:58,900 --> 00:40:00,570
state minimum wages.

849
00:40:00,570 --> 00:40:03,140
Turns out a lot of states
actually set minimum wages

850
00:40:03,140 --> 00:40:06,040
higher than the national
minimum wage.

851
00:40:06,040 --> 00:40:08,220
Not really much anymore because
the national minimum

852
00:40:08,220 --> 00:40:10,140
wage has gone up a lot
the last decade.

853
00:40:10,140 --> 00:40:12,765
But about a decade ago the
national minimum wage, or

854
00:40:12,765 --> 00:40:14,570
about 15 years ago, the national
minimum wage was at a

855
00:40:14,570 --> 00:40:17,840
real historical minimum in
real terms. And a lot of

856
00:40:17,840 --> 00:40:20,820
states exceeded that in setting
their minimum wage.

857
00:40:20,820 --> 00:40:22,870
So take two states, for
example New Jersey and

858
00:40:22,870 --> 00:40:24,120
Pennsylvania.

859
00:40:25,760 --> 00:40:30,480
New Jersey raised its minimum
wage, Pennsylvania doesn't.

860
00:40:30,480 --> 00:40:32,730
But any economic shock is going
to hit New Jersey and

861
00:40:32,730 --> 00:40:33,880
Pennsylvania pretty similarly.

862
00:40:33,880 --> 00:40:35,570
They're both right next to each
other on the East coast.

863
00:40:35,570 --> 00:40:38,570
Any recession's going to hit
them pretty similarly.

864
00:40:38,570 --> 00:40:40,850
So what you could do is you
could ask what happens to

865
00:40:40,850 --> 00:40:45,720
employment in New Jersey when
they raise their minimum wage

866
00:40:45,720 --> 00:40:48,060
relative to Pennsylvania,
which suffers the same

867
00:40:48,060 --> 00:40:52,210
economic shocks but doesn't
raise its minimum wage.

868
00:40:52,210 --> 00:40:53,960
We try to achieve the
gold standard.

869
00:40:53,960 --> 00:40:54,910
What's the gold standard?

870
00:40:54,910 --> 00:40:56,500
The gold standard is
a randomized trial.

871
00:40:56,500 --> 00:41:00,465
What we'd like is literally a
trial where we change the

872
00:41:00,465 --> 00:41:01,750
minimum wage randomly
at different places

873
00:41:01,750 --> 00:41:02,880
and different times.

874
00:41:02,880 --> 00:41:04,980
That's ever going to happen.

875
00:41:04,980 --> 00:41:07,430
So we try to approximate in
what we call a natural

876
00:41:07,430 --> 00:41:10,770
experiment or quasi experiment
which say, are there

877
00:41:10,770 --> 00:41:14,680
experimental interventions that
nature gives us, even if

878
00:41:14,680 --> 00:41:16,420
they're not perfectly
randomized trials.

879
00:41:16,420 --> 00:41:17,180
And this is one.

880
00:41:17,180 --> 00:41:20,920
Here's a situation where we have
two states, very similar,

881
00:41:20,920 --> 00:41:26,710
one raises the minimum
wage and one doesn't.

882
00:41:26,710 --> 00:41:28,160
And they're set by
economic shocks.

883
00:41:28,160 --> 00:41:29,410
You can look at that.

884
00:41:32,290 --> 00:41:35,090
Another way people have taken
this approach is say, well

885
00:41:35,090 --> 00:41:38,030
when the minimum wage increases
it's going to have

886
00:41:38,030 --> 00:41:40,390
different effects in
different places.

887
00:41:40,390 --> 00:41:41,450
And why is that?

888
00:41:41,450 --> 00:41:44,320
That's because the minimum wage
is going to be higher

889
00:41:44,320 --> 00:41:47,670
relative to the market wage in
some areas than in others.

890
00:41:47,670 --> 00:41:49,210
So, for example, when they
raised the minimum wage a

891
00:41:49,210 --> 00:41:54,190
dollar nationally, most people
in Massachusetts already make

892
00:41:54,190 --> 00:41:55,730
more than they raised it to.

893
00:41:55,730 --> 00:41:57,000
So it's not going to affect
Massachusetts much.

894
00:41:57,000 --> 00:42:01,570
Whereas in Mississippi
that's a big hit.

895
00:42:01,570 --> 00:42:07,100
So given a national raise we can
compare states that were

896
00:42:07,100 --> 00:42:09,660
hit harder by that raise to
states that were hit less

897
00:42:09,660 --> 00:42:12,480
hard, another way to
try to do this.

898
00:42:12,480 --> 00:42:15,990
A number of studies have done
this and they've come up with

899
00:42:15,990 --> 00:42:19,180
the striking conclusion that the
minimum wage, if anything,

900
00:42:19,180 --> 00:42:23,570
raises employment, at least at
the levels that we've seen

901
00:42:23,570 --> 00:42:25,450
over the last 15,20 years.

902
00:42:25,450 --> 00:42:27,990
Changes in the minimum wage are
actually associated with

903
00:42:27,990 --> 00:42:30,450
modest, but increases
in employment.

904
00:42:30,450 --> 00:42:33,820
And certainly no decreases.

905
00:42:33,820 --> 00:42:37,270
They've been associated with
modest increases in employment

906
00:42:37,270 --> 00:42:38,680
and not decreases.

907
00:42:38,680 --> 00:42:43,530
And this is really, really
striking because this is

908
00:42:43,530 --> 00:42:45,880
something that economists
just never

909
00:42:45,880 --> 00:42:47,940
even really took seriously.

910
00:42:47,940 --> 00:42:49,750
We always taught that minimum
wage was the bad boy of

911
00:42:49,750 --> 00:42:53,470
economics, and this is saying,
no, in fact, if you take it

912
00:42:53,470 --> 00:42:59,060
seriously and look at the
evidence carefully you can

913
00:42:59,060 --> 00:43:02,380
actually see that a minimum
wage can actually increase

914
00:43:02,380 --> 00:43:06,080
employment as this
model shows.

915
00:43:06,080 --> 00:43:09,190
Now, this is still the subject
of some controversy, there's

916
00:43:09,190 --> 00:43:11,640
still a lot of work on it.

917
00:43:11,640 --> 00:43:14,825
But it then raises the
question of, well how

918
00:43:14,825 --> 00:43:16,450
can this really be?

919
00:43:16,450 --> 00:43:18,780
Don't we have a pretty
competitive labor market?

920
00:43:18,780 --> 00:43:22,280
How do we really have a
monopsony labor market?

921
00:43:22,280 --> 00:43:24,640
And the answer is that if you
look at who's getting the

922
00:43:24,640 --> 00:43:27,840
minimum wage it's largely
younger and low-skilled

923
00:43:27,840 --> 00:43:31,790
workers who don't have a lot
of employment options.

924
00:43:31,790 --> 00:43:37,730
So basically McDonald's in a
given area in a city could

925
00:43:37,730 --> 00:43:39,880
have some monopsony power.

926
00:43:39,880 --> 00:43:41,670
Because people don't
have cars.

927
00:43:41,670 --> 00:43:43,620
These low-income urban youths
don't have cars.

928
00:43:43,620 --> 00:43:45,390
They can't go somewhere
else to work.

929
00:43:45,390 --> 00:43:48,090
They don't have skills so they
can even work retail because

930
00:43:48,090 --> 00:43:50,260
they don't have good enough
skills to work retail.

931
00:43:50,260 --> 00:43:52,290
So McDonald's has them.

932
00:43:52,290 --> 00:43:54,550
McDonald's has some monopsony
power over them because

933
00:43:54,550 --> 00:43:56,250
McDonald's is a job
they can get.

934
00:43:56,250 --> 00:43:58,730
Or McDonald's and Burger King
together maybe are the jobs

935
00:43:58,730 --> 00:43:59,690
they can get.

936
00:43:59,690 --> 00:44:01,620
That gives them some
monopsony power.

937
00:44:01,620 --> 00:44:03,500
So given that's where the
minimum wage is going to bite

938
00:44:03,500 --> 00:44:06,110
the most it's maybe not
implausible that the minimum

939
00:44:06,110 --> 00:44:09,590
wage could actually increase
employment which is what we

940
00:44:09,590 --> 00:44:12,130
see in the studies.

941
00:44:12,130 --> 00:44:13,160
Questions about that?

942
00:44:13,160 --> 00:44:13,360
Yeah.

943
00:44:13,360 --> 00:44:20,108
AUDIENCE: [INAUDIBLE] anti-trust
would it be

944
00:44:20,108 --> 00:44:22,036
possible [INAUDIBLE]

945
00:44:22,036 --> 00:44:24,260
PROFESSOR: Yeah exactly.

946
00:44:24,260 --> 00:44:25,730
So let's take my inner
city example.

947
00:44:25,730 --> 00:44:26,860
McDonald's isn't the
only employer.

948
00:44:26,860 --> 00:44:28,820
There's Burger King and Wendy's
and lots of other

949
00:44:28,820 --> 00:44:30,400
low-skilled employers.

950
00:44:30,400 --> 00:44:33,800
The notion is if there's a few
of them that should be enough

951
00:44:33,800 --> 00:44:37,530
to break any monopsony power
unless they collude.

952
00:44:37,530 --> 00:44:39,770
So likewise just as there's not
supposed to be collusion

953
00:44:39,770 --> 00:44:41,880
on the output side, there are
laws against collusion on the

954
00:44:41,880 --> 00:44:44,360
input side in the same way.

955
00:44:44,360 --> 00:44:46,154
But once again, just as those
laws are hard to enforce on

956
00:44:46,154 --> 00:44:48,050
the output side they're
hard to enforce.

957
00:44:48,050 --> 00:44:52,135
Because what you can do is you
can get together in the back

958
00:44:52,135 --> 00:44:55,120
room, or they can just say,
Wendy's and Burger King can

959
00:44:55,120 --> 00:44:56,910
wait and see what McDonald's
does and just

960
00:44:56,910 --> 00:44:58,360
follow in lock step.

961
00:44:58,360 --> 00:45:01,290
So there's lots of ways to
get around those rules.

962
00:45:01,290 --> 00:45:05,390
But yes, just as there's
antitrust laws on the output

963
00:45:05,390 --> 00:45:09,710
side, there are the labor market
laws on the input side

964
00:45:09,710 --> 00:45:11,610
which get in the way
of collusion.

965
00:45:11,610 --> 00:45:13,870
The difference is those
are more on a

966
00:45:13,870 --> 00:45:16,100
sector by sector basis.

967
00:45:16,100 --> 00:45:24,470
So, for example, the
unionization of the workers

968
00:45:24,470 --> 00:45:27,735
affects the collusion ability
of the employers.

969
00:45:27,735 --> 00:45:31,350
So if the workers are unionized
it's more lax in

970
00:45:31,350 --> 00:45:33,470
terms of allowing employers to
collude as well because the

971
00:45:33,470 --> 00:45:34,710
workers are colluding.

972
00:45:34,710 --> 00:45:36,190
If workers aren't unionized
it's less lax.

973
00:45:36,190 --> 00:45:39,110
And there's a complicated body
of labor law about that.

974
00:45:39,110 --> 00:45:40,940
Other questions?

975
00:45:40,940 --> 00:45:41,245
Yeah.

976
00:45:41,245 --> 00:45:45,640
AUDIENCE: So it would be bad for
unions to try and get an

977
00:45:45,640 --> 00:45:49,000
industry-wide standard
wage or--

978
00:45:49,000 --> 00:45:51,630
PROFESSOR: Well that's about
efficiency versus equity.

979
00:45:51,630 --> 00:45:53,380
If a union got an industry-wide
standardized

980
00:45:53,380 --> 00:45:56,200
wage then that's going to
penalize the talented workers

981
00:45:56,200 --> 00:45:59,530
who would want to go work
elsewhere and help the less

982
00:45:59,530 --> 00:46:00,710
talented workers.

983
00:46:00,710 --> 00:46:03,180
And basically a lot of the
complaints employers have

984
00:46:03,180 --> 00:46:05,710
about unions is that they lose
their talented workers because

985
00:46:05,710 --> 00:46:08,700
the union doesn't allow them
to pay those differentials.

986
00:46:08,700 --> 00:46:11,080
Teachers is a great example
that's very controversial

987
00:46:11,080 --> 00:46:12,150
right now which is,
should there be

988
00:46:12,150 --> 00:46:13,710
merit pay for teachers.

989
00:46:13,710 --> 00:46:16,240
A lot of teachers say, no, there
shouldn't be merit pay

990
00:46:16,240 --> 00:46:19,360
because that's going to violate
workplace norms and

991
00:46:19,360 --> 00:46:20,380
it's unfair.

992
00:46:20,380 --> 00:46:22,870
But we might not be getting the
most talented people into

993
00:46:22,870 --> 00:46:24,150
teaching as a result.

994
00:46:27,060 --> 00:46:30,510
So we'll come back next time
and we'll start our topics

995
00:46:30,510 --> 00:46:32,790
part of the course by talking
about international trade and

996
00:46:32,790 --> 00:46:34,640
whether it's good or bad.

997
00:46:34,640 --> 00:46:35,550
Answer, it's good.

998
00:46:35,550 --> 00:46:36,800
But we'll talk about why.