1
00:00:00,500 --> 00:00:02,405
[SQUEAKING]

2
00:00:02,405 --> 00:00:03,367
[RUSTLING]

3
00:00:03,367 --> 00:00:07,215
[CLICKING]

4
00:00:11,070 --> 00:00:19,990
JONATHAN GRUBER: Externalities,
so, so far in the class,

5
00:00:19,990 --> 00:00:21,970
we once again remember
the big picture.

6
00:00:21,970 --> 00:00:23,950
We started with the
first fundamental theorem

7
00:00:23,950 --> 00:00:27,190
overall for economics, which
is that the competitive market

8
00:00:27,190 --> 00:00:29,950
will maximize total
social welfare.

9
00:00:29,950 --> 00:00:32,470
Then we said that will not
be true under conditions

10
00:00:32,470 --> 00:00:33,430
of market failure.

11
00:00:33,430 --> 00:00:35,830
Remember, market failure
doesn't mean market collapse.

12
00:00:35,830 --> 00:00:39,610
It means when there are
barriers to the market achieving

13
00:00:39,610 --> 00:00:41,860
this first best outcome, OK?

14
00:00:41,860 --> 00:00:45,310
One barrier is
imperfect competition.

15
00:00:45,310 --> 00:00:48,910
One barrier was
imperfect information.

16
00:00:48,910 --> 00:00:53,560
A third barrier to
welfare maximization

17
00:00:53,560 --> 00:00:57,300
or a third source of market
failure is externalities, OK?

18
00:00:57,300 --> 00:00:59,050
That's the third source
of market failure,

19
00:00:59,050 --> 00:01:00,820
and we're going to
talk about that today.

20
00:01:00,820 --> 00:01:01,950
What is an externality?

21
00:01:01,950 --> 00:01:04,629
Let me be very clear
about the definition.

22
00:01:04,629 --> 00:01:10,720
An externality occurs
whenever one party's actions

23
00:01:10,720 --> 00:01:15,178
makes another party
better or worse off, OK?

24
00:01:15,178 --> 00:01:16,720
Let's do it-- let's
make it a person,

25
00:01:16,720 --> 00:01:20,230
whether my actions make
you better or worse off,

26
00:01:20,230 --> 00:01:23,540
but I don't bear the
consequences of that.

27
00:01:23,540 --> 00:01:27,040
So, when my actions make
you better or worse off,

28
00:01:27,040 --> 00:01:29,560
but I don't bear
the consequences,

29
00:01:29,560 --> 00:01:31,000
then there's an externality.

30
00:01:34,610 --> 00:01:37,460
So let's talk about
that in the context.

31
00:01:37,460 --> 00:01:40,460
Let's start with the classic
example of externalities, which

32
00:01:40,460 --> 00:01:47,560
is a negative
production externality,

33
00:01:47,560 --> 00:01:52,630
a negative production
externality, OK?

34
00:01:52,630 --> 00:01:56,290
The classic example
is you've got a river.

35
00:01:56,290 --> 00:01:58,760
On that river is a steel plant.

36
00:01:58,760 --> 00:02:01,870
That steel plant
produces steel, OK?

37
00:02:01,870 --> 00:02:03,880
But, as a byproduct
of that production,

38
00:02:03,880 --> 00:02:05,543
it dumps sludge
into the river, OK?

39
00:02:05,543 --> 00:02:06,710
You guys remember The Lorax?

40
00:02:06,710 --> 00:02:08,740
So it's basically The Lorax, OK?

41
00:02:08,740 --> 00:02:12,280
So, basically, the steel plant
dumps sludge into the river,

42
00:02:12,280 --> 00:02:13,030
OK?

43
00:02:13,030 --> 00:02:17,600
That sludge floats down the
river and kills the fish.

44
00:02:17,600 --> 00:02:21,320
Those killed fish mean that
fishermen cannot make as much

45
00:02:21,320 --> 00:02:25,730
money fishing on the river, OK?

46
00:02:25,730 --> 00:02:33,230
So, basically, there's sludge
coming out of the factory,

47
00:02:33,230 --> 00:02:35,870
and that sludge we're going
to make the assumption

48
00:02:35,870 --> 00:02:38,677
is directly proportional
to the production of steel.

49
00:02:38,677 --> 00:02:41,010
So we're going to say, for
every unit of steel produced,

50
00:02:41,010 --> 00:02:43,820
there's one unit of sludge
that emits on the factory just

51
00:02:43,820 --> 00:02:45,200
to make this model easy.

52
00:02:45,200 --> 00:02:47,740
So, every unit of steel,
there's one unit of sludge.

53
00:02:47,740 --> 00:02:50,390
That sludge flows down the
river and kills the fish.

54
00:02:50,390 --> 00:02:52,280
Unfortunately, there
are fishermen down

55
00:02:52,280 --> 00:02:54,560
this river who are
trying to catch fish,

56
00:02:54,560 --> 00:02:58,100
and that hurts their
livelihood, OK?

57
00:02:58,100 --> 00:03:00,380
So this is a classic
example of living

58
00:03:00,380 --> 00:03:04,550
by an externality, because
the steel plant's behavior is

59
00:03:04,550 --> 00:03:07,008
imposing a cost
on the fishermen.

60
00:03:07,008 --> 00:03:09,050
Their sludge is imposing
a cost on the fishermen,

61
00:03:09,050 --> 00:03:11,425
but the steel plant doesn't
bear any consequence of that.

62
00:03:11,425 --> 00:03:13,972
They just dump the sludge
and forget about it, OK?

63
00:03:13,972 --> 00:03:15,430
So that's what we
mean by negative.

64
00:03:15,430 --> 00:03:18,050
It's a negative externality
because my actions

65
00:03:18,050 --> 00:03:18,913
are hurting you.

66
00:03:18,913 --> 00:03:21,080
The steel plant's actions
are hurting the fishermen.

67
00:03:21,080 --> 00:03:23,163
It's a production externality
because it comes out

68
00:03:23,163 --> 00:03:26,375
of the production process,
in this case, for steel.

69
00:03:26,375 --> 00:03:28,250
So that's what we mean
by negative production

70
00:03:28,250 --> 00:03:29,420
externality.

71
00:03:29,420 --> 00:03:33,200
It means that, when one
party's production adversely

72
00:03:33,200 --> 00:03:36,410
affects another party, but
the party doing the production

73
00:03:36,410 --> 00:03:38,840
doesn't bear any
consequences of that, then

74
00:03:38,840 --> 00:03:41,110
that's a negative
production externality.

75
00:03:41,110 --> 00:03:42,360
So what effect does that have?

76
00:03:42,360 --> 00:03:45,230
Let's go to figure 22-1
and talk graphically

77
00:03:45,230 --> 00:03:50,270
about how we think about
production externalities, OK?

78
00:03:50,270 --> 00:03:51,945
This is the market for steel.

79
00:03:51,945 --> 00:03:53,570
In figure 22-1 is
the market for steel,

80
00:03:53,570 --> 00:03:55,862
the quantity of steel on the
x-axis, the price of steel

81
00:03:55,862 --> 00:03:58,220
on the y-axis, OK?

82
00:03:58,220 --> 00:04:01,250
The market is initially
in equilibrium at point A.

83
00:04:01,250 --> 00:04:05,060
That is where demand, which is
the downward-sloping blue line,

84
00:04:05,060 --> 00:04:08,290
equals supply, which is
upward-sloping blue line.

85
00:04:08,290 --> 00:04:13,020
Now, as we said, in a
perfectly competitive market,

86
00:04:13,020 --> 00:04:17,220
demand represents the
marginal willingness

87
00:04:17,220 --> 00:04:23,010
to pay for the good, which is
equal to the marginal benefit

88
00:04:23,010 --> 00:04:25,670
that consumers get from
consuming the good, OK?

89
00:04:25,670 --> 00:04:27,420
The marginal benefit
of consuming the good

90
00:04:27,420 --> 00:04:29,280
is the marginal
willingness to pay.

91
00:04:29,280 --> 00:04:32,760
And that's what's represented
by the demand curve, OK?

92
00:04:32,760 --> 00:04:34,260
We're not going to
touch that here.

93
00:04:34,260 --> 00:04:37,000
We're going to leave
that alone, OK?

94
00:04:37,000 --> 00:04:41,690
The supply curve is the firm's
marginal willingness to supply,

95
00:04:41,690 --> 00:04:44,870
which is their
marginal cost, OK?

96
00:04:44,870 --> 00:04:46,640
So, in the perfectly
competitive market

97
00:04:46,640 --> 00:04:50,340
where marginal benefit
equals marginal cost,

98
00:04:50,340 --> 00:04:51,590
we get equilibrium.

99
00:04:51,590 --> 00:04:54,590
And that yields the
welfare-maximizing outcome.

100
00:04:54,590 --> 00:04:55,520
The market succeeds.

101
00:04:55,520 --> 00:04:58,390
It does not fail, OK?

102
00:04:58,390 --> 00:05:00,800
The difference now
is we're now going

103
00:05:00,800 --> 00:05:04,780
to drive a wedge between
privately perceived

104
00:05:04,780 --> 00:05:08,230
benefits and social benefits.

105
00:05:08,230 --> 00:05:11,020
So, for the consumption of
steel with the demand curve,

106
00:05:11,020 --> 00:05:13,050
we're going to see the
benefit to individuals

107
00:05:13,050 --> 00:05:14,378
is the benefit to society.

108
00:05:14,378 --> 00:05:15,420
They're one and the same.

109
00:05:15,420 --> 00:05:18,010
That's what we've
assumed all course.

110
00:05:18,010 --> 00:05:19,600
But, for the supply
of steel, we're

111
00:05:19,600 --> 00:05:20,860
going to say wait a second.

112
00:05:20,860 --> 00:05:25,852
The benefit to society is
different than the benefit

113
00:05:25,852 --> 00:05:26,810
to the steel producers.

114
00:05:26,810 --> 00:05:27,352
Or I'm sorry.

115
00:05:27,352 --> 00:05:30,330
The cost to society is different
than the cost to the steel

116
00:05:30,330 --> 00:05:31,177
producers.

117
00:05:31,177 --> 00:05:32,760
The cost to the steel
producers, which

118
00:05:32,760 --> 00:05:36,540
is their private marginal
cost, is the supply curve,

119
00:05:36,540 --> 00:05:42,470
but the social marginal
cost adds the damage

120
00:05:42,470 --> 00:05:44,180
they're doing to the fishermen.

121
00:05:44,180 --> 00:05:48,340
That is society encompasses
all the actors in society.

122
00:05:48,340 --> 00:05:51,650
It encompasses both the steel
plant and the fishermen.

123
00:05:51,650 --> 00:05:54,290
So the marginal cost
to society is the cost

124
00:05:54,290 --> 00:05:59,180
of producing the steel plus
the marginal damage being

125
00:05:59,180 --> 00:06:00,290
done to the fishermen.

126
00:06:00,290 --> 00:06:04,970
So social marginal cost
equals private marginal cost

127
00:06:04,970 --> 00:06:07,738
plus marginal damage.

128
00:06:07,738 --> 00:06:09,780
Social marginal cost equals
private marginal cost

129
00:06:09,780 --> 00:06:11,940
plus the marginal damage.

130
00:06:11,940 --> 00:06:13,320
And we see that as the red line.

131
00:06:16,750 --> 00:06:22,510
What that means, from a welfare
perspective, what we care about

132
00:06:22,510 --> 00:06:25,030
is social marginal
benefits and costs, not

133
00:06:25,030 --> 00:06:27,220
private marginal
benefits and costs.

134
00:06:27,220 --> 00:06:29,800
So what that means is
the social optimum,

135
00:06:29,800 --> 00:06:34,602
the welfare-maximizing optimum,
is actually at point C. Point

136
00:06:34,602 --> 00:06:36,310
C is the welfare-maximizing
optimum where

137
00:06:36,310 --> 00:06:41,994
the social marginal cost equals
the social marginal benefit,

138
00:06:41,994 --> 00:06:42,950
OK?

139
00:06:42,950 --> 00:06:47,780
And, therefore, we overproduce.

140
00:06:47,780 --> 00:06:50,840
What's happening is
the steel company,

141
00:06:50,840 --> 00:06:52,790
not considering
the damage they're

142
00:06:52,790 --> 00:06:56,590
doing through production to the
fisherman, produces too much.

143
00:06:56,590 --> 00:07:00,050
The steel company
produces at the point

144
00:07:00,050 --> 00:07:03,800
where private
marginal cost equals

145
00:07:03,800 --> 00:07:06,140
private marginal benefit,
which, in our case,

146
00:07:06,140 --> 00:07:09,420
equals social
marginal benefit, OK?

147
00:07:09,420 --> 00:07:12,360
That's the private
market decision.

148
00:07:12,360 --> 00:07:15,320
But, in fact, it should
be producing at the point

149
00:07:15,320 --> 00:07:18,373
where social
marginal cost equals

150
00:07:18,373 --> 00:07:20,540
private marginal benefit,
private marginal benefit--

151
00:07:20,540 --> 00:07:24,440
equals social
marginal benefit, OK?

152
00:07:24,440 --> 00:07:26,330
And the point where
social marginal cost

153
00:07:26,330 --> 00:07:29,840
equals social marginal
benefit is lower production.

154
00:07:29,840 --> 00:07:30,770
Why?

155
00:07:30,770 --> 00:07:33,170
Because lower
production avoids--

156
00:07:33,170 --> 00:07:37,490
reduces the damage being done
to the fishermen down the river,

157
00:07:37,490 --> 00:07:38,460
OK?

158
00:07:38,460 --> 00:07:42,120
So the optimum, from
society's perspective,

159
00:07:42,120 --> 00:07:45,250
is point C. In other words,
there is a market failure.

160
00:07:45,250 --> 00:07:48,690
The private market
is not delivering

161
00:07:48,690 --> 00:07:51,990
the welfare-maximizing outcome.

162
00:07:51,990 --> 00:07:55,960
And we can see that
creates a deadweight loss.

163
00:07:55,960 --> 00:07:59,210
The deadweight loss
is the units that

164
00:07:59,210 --> 00:08:02,123
are traded that are
socially inefficient.

165
00:08:02,123 --> 00:08:03,540
Why are they
socially inefficient?

166
00:08:03,540 --> 00:08:04,910
They're privately efficient.

167
00:08:04,910 --> 00:08:08,255
If you see Q2 and Q1, before
we introduced externalities,

168
00:08:08,255 --> 00:08:10,880
we'd say, well, it's a shame if
they don't get produced, right?

169
00:08:10,880 --> 00:08:12,422
If we ignore
externalities, we'd say,

170
00:08:12,422 --> 00:08:15,730
Q2 and Q1, well, they have a
benefit higher than their cost.

171
00:08:15,730 --> 00:08:17,120
So they should get produced.

172
00:08:17,120 --> 00:08:18,912
But, actually, in a
world of externalities,

173
00:08:18,912 --> 00:08:21,200
their benefit is
lower than their cost

174
00:08:21,200 --> 00:08:24,620
because their cost incorporates
the damage done to the fish.

175
00:08:24,620 --> 00:08:26,240
So there's a deadweight loss.

176
00:08:26,240 --> 00:08:27,349
Critically, remember,
you've got to know

177
00:08:27,349 --> 00:08:29,182
how to draw these
deadweight loss triangles.

178
00:08:29,182 --> 00:08:31,130
Remember, deadweight
loss triangles

179
00:08:31,130 --> 00:08:34,010
always point to the optimum.

180
00:08:34,010 --> 00:08:38,059
The deadweight lost triangle
is the area ABC, OK?

181
00:08:38,059 --> 00:08:40,880
It's drawn-- it's
units that are sold

182
00:08:40,880 --> 00:08:43,850
where the social
marginal cost exceeds

183
00:08:43,850 --> 00:08:45,800
the social marginal benefit.

184
00:08:45,800 --> 00:08:47,790
And that's that
deadweight loss triangle.

185
00:08:47,790 --> 00:08:50,600
So there's an
inefficiency arising

186
00:08:50,600 --> 00:08:53,060
from the fact the
private actors do not

187
00:08:53,060 --> 00:08:56,660
account for the social
implications of their actions,

188
00:08:56,660 --> 00:08:57,820
OK?

189
00:08:57,820 --> 00:09:00,640
Questions about that?

190
00:09:00,640 --> 00:09:02,810
So we have here a
classic example,

191
00:09:02,810 --> 00:09:05,810
perhaps the classic example
in all of economics,

192
00:09:05,810 --> 00:09:08,230
of a market failure.

193
00:09:08,230 --> 00:09:09,990
The classic example
is a market failure

194
00:09:09,990 --> 00:09:15,030
happens when the social
implications of your actions

195
00:09:15,030 --> 00:09:17,255
are different than the
private implications

196
00:09:17,255 --> 00:09:19,380
since people maximize their
own private well-being.

197
00:09:19,380 --> 00:09:21,162
That's what Adam Smith
sort of taught us.

198
00:09:21,162 --> 00:09:22,620
The notion of the
invisible hand is

199
00:09:22,620 --> 00:09:24,660
that the market acting
in its own interests

200
00:09:24,660 --> 00:09:26,820
will deliver the best
outcomes for society.

201
00:09:26,820 --> 00:09:28,740
We're saying, no,
that's not true

202
00:09:28,740 --> 00:09:30,420
if the market's
own interests has

203
00:09:30,420 --> 00:09:36,990
implications for other parties
that are not accounted for, OK?

204
00:09:36,990 --> 00:09:44,420
Now externalities don't have to
just be on the production side.

205
00:09:44,420 --> 00:09:51,197
We can also have negative
consumption externalities.

206
00:09:54,400 --> 00:09:59,170
That would be a case where
my literally consuming a good

207
00:09:59,170 --> 00:10:02,240
makes you worse off.

208
00:10:02,240 --> 00:10:07,960
My consuming a good
makes you worse off, OK?

209
00:10:07,960 --> 00:10:10,930
So let me start with
a simple question.

210
00:10:10,930 --> 00:10:14,380
Let's start with a perfectly
competitive market.

211
00:10:14,380 --> 00:10:19,680
If I consume a good, I
raise demand for that good.

212
00:10:19,680 --> 00:10:22,630
That raises the price.

213
00:10:22,630 --> 00:10:25,110
Is that an externality?

214
00:10:25,110 --> 00:10:27,390
In a perfectly competitive
market, it's not.

215
00:10:27,390 --> 00:10:28,930
Why isn't that an externality?

216
00:10:28,930 --> 00:10:29,730
If I consume the--

217
00:10:29,730 --> 00:10:31,772
if I want to consume the
good, the price goes up.

218
00:10:31,772 --> 00:10:33,200
Everyone pays a higher price.

219
00:10:33,200 --> 00:10:34,570
Why is that not an externality?

220
00:10:34,570 --> 00:10:34,780
Yeah?

221
00:10:34,780 --> 00:10:36,420
STUDENT: Because you
also bear the cost.

222
00:10:36,420 --> 00:10:38,462
JONATHAN GRUBER: Because
you also bear the costs.

223
00:10:38,462 --> 00:10:40,180
An externality only
occurs when you don't

224
00:10:40,180 --> 00:10:41,640
bear the costs of your actions.

225
00:10:41,640 --> 00:10:44,670
When you want a car, and,
therefore, the price of cars

226
00:10:44,670 --> 00:10:47,560
goes up, you pay
that higher price.

227
00:10:47,560 --> 00:10:49,440
Externalities only
occur when you

228
00:10:49,440 --> 00:10:51,572
don't bear the consequences.

229
00:10:51,572 --> 00:10:53,280
So, in general,
consumption externalities

230
00:10:53,280 --> 00:10:55,820
don't happen through
causing higher prices.

231
00:10:55,820 --> 00:10:58,800
Consumption externalities
happen more directly when

232
00:10:58,800 --> 00:11:00,960
my consumption affects you.

233
00:11:00,960 --> 00:11:05,130
So the best example of
this would be smoking, OK?

234
00:11:05,130 --> 00:11:08,632
When I smoke, it affects you.

235
00:11:08,632 --> 00:11:10,090
It affects you in
a number of ways.

236
00:11:10,090 --> 00:11:12,400
Most directly, if I
smoke in this classroom,

237
00:11:12,400 --> 00:11:16,150
you get secondhand smoke,
and you get ill as a result.

238
00:11:16,150 --> 00:11:17,860
But that's not all.

239
00:11:17,860 --> 00:11:23,290
It affects you because, if
I smoke, and I get sick,

240
00:11:23,290 --> 00:11:28,020
and my health care costs go
up, then, well, I work at MIT.

241
00:11:28,020 --> 00:11:30,630
All my fellow MIT employees
bear those costs because we all

242
00:11:30,630 --> 00:11:32,130
share health insurance.

243
00:11:32,130 --> 00:11:34,152
And, when I retire,
all society bears

244
00:11:34,152 --> 00:11:35,610
those costs because
those costs are

245
00:11:35,610 --> 00:11:37,470
paid for by the
Medicare program, which

246
00:11:37,470 --> 00:11:40,470
is financed by taxation.

247
00:11:40,470 --> 00:11:42,620
So my health care costs
are an externality.

248
00:11:42,620 --> 00:11:44,690
Secondhand smoke
is an externality.

249
00:11:44,690 --> 00:11:47,060
What are some other
externalities from smoking?

250
00:11:47,060 --> 00:11:51,010
What are other externalities
that can occur from smoking?

251
00:11:51,010 --> 00:11:51,530
Yeah?

252
00:11:51,530 --> 00:11:53,840
STUDENT: Environmental
damage from the production.

253
00:11:53,840 --> 00:11:55,882
JONATHAN GRUBER: Well,
that would be a production

254
00:11:55,882 --> 00:11:56,900
externality, OK?

255
00:11:56,900 --> 00:11:58,567
We're going to leave
that alone for now.

256
00:11:58,567 --> 00:12:00,890
I'm just talking about
from consuming cigarettes.

257
00:12:00,890 --> 00:12:03,830
From consuming cigarettes, what
else-- what other damage comes?

258
00:12:03,830 --> 00:12:04,475
So I make you--

259
00:12:04,475 --> 00:12:06,350
I may make you sick
through secondhand smoke.

260
00:12:06,350 --> 00:12:08,660
I might raise health care costs.

261
00:12:08,660 --> 00:12:11,815
Well, I might raise
health care costs, OK?

262
00:12:11,815 --> 00:12:13,190
What else is
another externality?

263
00:12:13,190 --> 00:12:14,570
What else does smoking do?

264
00:12:17,820 --> 00:12:20,330
Well, it turns out
there are 100,000.

265
00:12:20,330 --> 00:12:22,620
This is a number which I
triple check because I still

266
00:12:22,620 --> 00:12:23,640
can't believe it.

267
00:12:23,640 --> 00:12:26,610
100,000 people every
single year die

268
00:12:26,610 --> 00:12:29,340
in fires caused by
smokers, not in the US,

269
00:12:29,340 --> 00:12:32,470
worldwide, which
is a crazy number.

270
00:12:32,470 --> 00:12:35,280
But, if you think about how
tightly packed slums are

271
00:12:35,280 --> 00:12:36,960
in developing
countries, one person

272
00:12:36,960 --> 00:12:38,752
falling asleep with
their cigarette burning

273
00:12:38,752 --> 00:12:41,220
can kill thousands
of people, OK?

274
00:12:41,220 --> 00:12:44,850
That is an externality
because my action to smoke

275
00:12:44,850 --> 00:12:46,522
has killed you, OK?

276
00:12:46,522 --> 00:12:48,730
And I'm clearly not going
to compensate you for that.

277
00:12:48,730 --> 00:12:52,430
So that's another
externality, OK?

278
00:12:52,430 --> 00:12:56,840
What about the fact that smokers
are less productive at work.

279
00:12:56,840 --> 00:12:59,150
They have to take
more smoke breaks.

280
00:12:59,150 --> 00:13:01,070
They might get sick more often.

281
00:13:01,070 --> 00:13:04,217
Is that an externality or not?

282
00:13:04,217 --> 00:13:06,300
The fact that smokers are
less productive at work,

283
00:13:06,300 --> 00:13:07,540
is that an externality or not?

284
00:13:07,540 --> 00:13:08,200
And why or why not?

285
00:13:08,200 --> 00:13:08,710
Yeah?

286
00:13:08,710 --> 00:13:10,110
STUDENT: Not necessarily
because, if they're

287
00:13:10,110 --> 00:13:11,500
less productive, they're
going to do less work

288
00:13:11,500 --> 00:13:12,250
and get paid less.

289
00:13:12,250 --> 00:13:14,530
JONATHAN GRUBER: Exactly,
it's not an externality

290
00:13:14,530 --> 00:13:16,910
if they're paid less.

291
00:13:16,910 --> 00:13:19,933
This is the key thing, which is
it's only an externality if you

292
00:13:19,933 --> 00:13:21,100
don't bear the consequences.

293
00:13:21,100 --> 00:13:22,930
If smokers are less
productive at work,

294
00:13:22,930 --> 00:13:24,700
and they get paid
less as a result

295
00:13:24,700 --> 00:13:26,950
by exactly the same amount
they're less productive,

296
00:13:26,950 --> 00:13:28,960
there's no externality.

297
00:13:28,960 --> 00:13:31,043
But, if their wage
doesn't fully adjust,

298
00:13:31,043 --> 00:13:32,710
and, therefore, their
lower productivity

299
00:13:32,710 --> 00:13:35,085
affects everybody else in the
firm or the firm's profits,

300
00:13:35,085 --> 00:13:37,270
that is an externality, OK?

301
00:13:37,270 --> 00:13:39,610
So this is the deep
aspect of externalities.

302
00:13:39,610 --> 00:13:41,500
You have to think
about whether people

303
00:13:41,500 --> 00:13:44,800
are compensating for it, OK?

304
00:13:44,800 --> 00:13:46,910
You have to think about that.

305
00:13:46,910 --> 00:13:54,260
OK, most importantly, the
fact I kill myself by smoking

306
00:13:54,260 --> 00:13:56,840
is not an externality, OK?

307
00:13:56,840 --> 00:13:59,818
Smokers die seven years
earlier on average.

308
00:13:59,818 --> 00:14:01,610
Roughly speaking, every
cigarette you smoke

309
00:14:01,610 --> 00:14:03,590
lowers your life
by seven minutes.

310
00:14:03,590 --> 00:14:07,220
It's pretty linear, OK?

311
00:14:07,220 --> 00:14:08,390
But you know what?

312
00:14:08,390 --> 00:14:10,820
If I sit by myself on a rock
in the middle of nowhere

313
00:14:10,820 --> 00:14:18,090
and smoke until I die, no
problem because, in that case,

314
00:14:18,090 --> 00:14:21,440
the social implications are
the private implications.

315
00:14:21,440 --> 00:14:23,300
I've made my privately
optimal decision,

316
00:14:23,300 --> 00:14:24,883
and there's no effect
on anybody else.

317
00:14:24,883 --> 00:14:26,550
So it's also socially optimal.

318
00:14:26,550 --> 00:14:28,610
There's only an externality if
we have one these mechanisms,

319
00:14:28,610 --> 00:14:30,380
like if I'm smoking, I'm in
the woods, and I start a fire,

320
00:14:30,380 --> 00:14:31,760
and the fire
company has to come.

321
00:14:31,760 --> 00:14:33,068
That's an externality.

322
00:14:33,068 --> 00:14:34,610
But, as long as I
just sit by myself,

323
00:14:34,610 --> 00:14:35,870
and I don't bother anybody--

324
00:14:35,870 --> 00:14:37,160
I just smoke until I die--

325
00:14:37,160 --> 00:14:39,160
there's no externalities, OK?

326
00:14:39,160 --> 00:14:41,780
So externalities come through
the effects on others.

327
00:14:41,780 --> 00:14:45,860
So let's think about the
externalities of smoking.

328
00:14:45,860 --> 00:14:49,070
Let's think about a negative
consumption externality, OK?

329
00:14:49,070 --> 00:14:51,980
Here we have the
market for cigarettes.

330
00:14:51,980 --> 00:14:53,530
On the x-axis, we
have the number--

331
00:14:53,530 --> 00:14:56,030
I'm sorry, figure 22-2.

332
00:14:56,030 --> 00:14:57,800
On the x-axis, we
have the number

333
00:14:57,800 --> 00:15:00,680
of cigarettes, quantity
of cigarettes consumed.

334
00:15:00,680 --> 00:15:03,470
On the y-axis, we have the
price of cigarettes per pack.

335
00:15:03,470 --> 00:15:06,170
We have an initial
equilibrium at point A, which

336
00:15:06,170 --> 00:15:10,480
is where the private
marginal benefit equals

337
00:15:10,480 --> 00:15:12,223
the private marginal cost.

338
00:15:12,223 --> 00:15:13,640
Here we're going
to assume there's

339
00:15:13,640 --> 00:15:15,350
no externalities from
producing tobacco.

340
00:15:15,350 --> 00:15:17,433
Let's assume there's no
sludge produced, whatever.

341
00:15:17,433 --> 00:15:18,620
That's a separate issue, OK?

342
00:15:18,620 --> 00:15:20,370
There may be production
externalities too.

343
00:15:20,370 --> 00:15:21,078
We covered those.

344
00:15:21,078 --> 00:15:22,910
You already know how
to think about those.

345
00:15:22,910 --> 00:15:24,535
But here let's assume
there aren't any.

346
00:15:24,535 --> 00:15:26,095
Let's assume the
social marginal cost

347
00:15:26,095 --> 00:15:27,470
equals the private
marginal cost,

348
00:15:27,470 --> 00:15:28,940
no production externalities.

349
00:15:28,940 --> 00:15:30,950
But there is a
consumption externality.

350
00:15:30,950 --> 00:15:33,620
Every cigarette I smoke
is bad for society.

351
00:15:33,620 --> 00:15:36,230
What that means is the
social marginal benefit

352
00:15:36,230 --> 00:15:39,502
is below the private
marginal benefit.

353
00:15:39,502 --> 00:15:41,960
The social marginal benefit is
the private marginal benefit

354
00:15:41,960 --> 00:15:44,420
minus the marginal
damage I'm doing.

355
00:15:44,420 --> 00:15:47,690
MD is the Marginal Damage,
the marginal damage I'm doing.

356
00:15:47,690 --> 00:15:50,190
That's estimated to be about--

357
00:15:50,190 --> 00:15:52,850
absent secondhand smoke, the
damage of smoking is about

358
00:15:52,850 --> 00:15:54,770
$0.50 a pack.

359
00:15:54,770 --> 00:15:59,210
The secondhand smoke
part is really hard,

360
00:15:59,210 --> 00:16:02,290
and the estimates are anywhere
from $0.01 to $2 per pack.

361
00:16:02,290 --> 00:16:04,910
So that's hard to know
how big that is, OK?

362
00:16:04,910 --> 00:16:09,680
But, certainly, we have
this negative consumption

363
00:16:09,680 --> 00:16:14,060
externality, which
is that, basically,

364
00:16:14,060 --> 00:16:18,260
every pack of cigarettes I smoke
is worth at least $0.50 less

365
00:16:18,260 --> 00:16:21,320
to society than it's worth to
me because I have these external

366
00:16:21,320 --> 00:16:24,830
effects on society and
perhaps a lot more than that.

367
00:16:24,830 --> 00:16:27,740
As a result, I should smoke--

368
00:16:27,740 --> 00:16:31,400
I choose to smoke at point
A, but the social optimum

369
00:16:31,400 --> 00:16:32,652
is point C.

370
00:16:32,652 --> 00:16:34,610
So, once again, I've
created a deadweight loss.

371
00:16:34,610 --> 00:16:37,040
Once again, I'm over consuming.

372
00:16:37,040 --> 00:16:38,420
There's overconsumption here.

373
00:16:38,420 --> 00:16:40,253
Just like there was
overproduction of steel.

374
00:16:40,253 --> 00:16:42,290
There's overconsumption
and a deadweight loss

375
00:16:42,290 --> 00:16:45,230
because there are units that are
privately optimal to consume,

376
00:16:45,230 --> 00:16:46,550
but not socially optimal.

377
00:16:51,330 --> 00:16:57,017
Now, externalities, this
is kind of a fun topic

378
00:16:57,017 --> 00:16:58,100
because it is interesting.

379
00:16:58,100 --> 00:17:02,120
I talk-- so let's talk for a
second about secondhand smoke

380
00:17:02,120 --> 00:17:04,720
and whether that's
actually an externality.

381
00:17:04,720 --> 00:17:07,079
So almost all the damage
of secondhand smoke

382
00:17:07,079 --> 00:17:08,849
is not done by
smoking in a crowd.

383
00:17:08,849 --> 00:17:10,442
It's done to family members.

384
00:17:10,442 --> 00:17:12,150
Almost all the damage
of secondhand smoke

385
00:17:12,150 --> 00:17:13,170
is done to family members.

386
00:17:13,170 --> 00:17:15,839
Mostly, it's that you make your
family members sick by smoking.

387
00:17:15,839 --> 00:17:18,290
Is that an externality?

388
00:17:18,290 --> 00:17:19,630
When or when is it not an--

389
00:17:19,630 --> 00:17:21,869
under what conditions might
it not be an externality?

390
00:17:21,869 --> 00:17:22,060
Yeah?

391
00:17:22,060 --> 00:17:23,880
STUDENT: Well, I guess it
wouldn't be an externality

392
00:17:23,880 --> 00:17:25,280
if like let's say you die.

393
00:17:25,280 --> 00:17:27,200
And then the consequences
of them getting sick

394
00:17:27,200 --> 00:17:28,453
doesn't affect you at all.

395
00:17:28,453 --> 00:17:29,370
Or, well, I guess it--

396
00:17:29,370 --> 00:17:31,703
JONATHAN GRUBER: No, no, then
it-- no, then it would be.

397
00:17:31,703 --> 00:17:33,770
OK, so what-- under what
conditions-- under what

398
00:17:33,770 --> 00:17:35,520
conditions would it
not be an externality?

399
00:17:35,520 --> 00:17:37,082
When-- yeah?

400
00:17:37,082 --> 00:17:40,260
STUDENT: If your-- like,
if your family gets sick,

401
00:17:40,260 --> 00:17:41,508
and you are impacted by that.

402
00:17:41,508 --> 00:17:43,550
JONATHAN GRUBER: Yeah, if
I care about my family,

403
00:17:43,550 --> 00:17:46,790
in particular, if I
maximize family utility,

404
00:17:46,790 --> 00:17:48,995
then it's not an externality.

405
00:17:48,995 --> 00:17:50,600
If I maximize not
my own utility,

406
00:17:50,600 --> 00:17:58,360
but my family's utility, then
I will essentially internalize,

407
00:17:58,360 --> 00:18:04,230
internalize, the externality.

408
00:18:04,230 --> 00:18:08,100
Just like a lower wage means
that I bear the consequences

409
00:18:08,100 --> 00:18:12,080
for being a less productive
worker, if I care

410
00:18:12,080 --> 00:18:14,150
about my whole
family's happiness,

411
00:18:14,150 --> 00:18:17,540
and I make my kids
sick by smoking,

412
00:18:17,540 --> 00:18:19,280
then my smoking
decision will actually

413
00:18:19,280 --> 00:18:22,040
reflect the total
consequences for my family.

414
00:18:22,040 --> 00:18:24,290
I will smoke only if it's
optimal for my family for me

415
00:18:24,290 --> 00:18:24,790
to do so.

416
00:18:24,790 --> 00:18:26,150
It doesn't mean I won't smoke.

417
00:18:26,150 --> 00:18:27,170
It just means I must--

418
00:18:27,170 --> 00:18:29,230
I'll have to enjoy it
enough that it's worth

419
00:18:29,230 --> 00:18:31,277
making my kids sick, OK?

420
00:18:31,277 --> 00:18:33,110
It doesn't mean that's
an incorrect decision

421
00:18:33,110 --> 00:18:35,240
because, after all, the
odds are you don't make

422
00:18:35,240 --> 00:18:37,100
your kids that
sick, and, you know,

423
00:18:37,100 --> 00:18:39,590
you might like
smoking a lot, OK?

424
00:18:39,590 --> 00:18:42,020
It's just that you will
internalize the externality

425
00:18:42,020 --> 00:18:44,120
because you only
smoke to the extent

426
00:18:44,120 --> 00:18:46,320
that it is optimal
for the whole family.

427
00:18:46,320 --> 00:18:50,140
So it's not necessarily
an externality, OK?

428
00:18:50,140 --> 00:18:53,050
So we'd like, actually, to test
whether it's an externality.

429
00:18:53,050 --> 00:18:55,790
There's actually a
clever test of this,

430
00:18:55,790 --> 00:18:58,960
which is there's a test of
whether people maximize family

431
00:18:58,960 --> 00:19:04,690
utility, which is, if people
maximize family utility,

432
00:19:04,690 --> 00:19:08,500
what would be the
implications of giving

433
00:19:08,500 --> 00:19:11,370
a father $1 versus
giving a mother $1?

434
00:19:11,370 --> 00:19:12,745
If both the father
and the mother

435
00:19:12,745 --> 00:19:15,050
are maximizing a family
utility function,

436
00:19:15,050 --> 00:19:17,050
the same family
utility function, then

437
00:19:17,050 --> 00:19:19,340
should it matter whether I
give $1 to a father or $1

438
00:19:19,340 --> 00:19:20,530
to a mother?

439
00:19:20,530 --> 00:19:22,853
No, it shouldn't because
that's $1 to the family.

440
00:19:22,853 --> 00:19:25,270
We're maximizing family utility
subject to a family budget

441
00:19:25,270 --> 00:19:26,200
constraint.

442
00:19:26,200 --> 00:19:28,460
It shouldn't matter
who gets the dollar.

443
00:19:28,460 --> 00:19:30,340
So one test of family
utility maximization

444
00:19:30,340 --> 00:19:33,650
is does it matter
who gets the money.

445
00:19:33,650 --> 00:19:36,960
And it turns out
it does a lot, OK?

446
00:19:36,960 --> 00:19:40,110
So there was a
great test of this.

447
00:19:40,110 --> 00:19:46,530
In the UK, they used to have a
tax system where, essentially,

448
00:19:46,530 --> 00:19:48,520
there was a credit they
gave for every kid.

449
00:19:48,520 --> 00:19:50,103
And the way the
credit worked is there

450
00:19:50,103 --> 00:19:52,850
would be a check sent home, OK?

451
00:19:52,850 --> 00:19:53,870
Then they changed it.

452
00:19:53,870 --> 00:19:54,860
So, instead, they
said, well, instead

453
00:19:54,860 --> 00:19:56,318
of sending a check
home, we're just

454
00:19:56,318 --> 00:19:58,072
going to add it into
pay, into wages.

455
00:19:58,072 --> 00:19:59,780
So, instead of getting
a check sent home,

456
00:19:59,780 --> 00:20:01,450
we're just going to
raise your wages.

457
00:20:01,450 --> 00:20:03,260
Well, it had no effect
on family budgets.

458
00:20:03,260 --> 00:20:04,552
They literally just changed it.

459
00:20:04,552 --> 00:20:07,220
The difference was, back
in-- this was in the '70s.

460
00:20:07,220 --> 00:20:08,990
Men worked and women didn't.

461
00:20:08,990 --> 00:20:11,980
So, when the check came
home, women controlled it.

462
00:20:11,980 --> 00:20:14,558
But, when it was in
wages, men controlled it.

463
00:20:14,558 --> 00:20:16,350
So, if there's family
utility maximization,

464
00:20:16,350 --> 00:20:17,430
it shouldn't matter.

465
00:20:17,430 --> 00:20:19,180
But it turned out, as
soon as they changed

466
00:20:19,180 --> 00:20:21,345
the way they paid it,
spending on kids went down,

467
00:20:21,345 --> 00:20:22,720
and spending on
drugs and alcohol

468
00:20:22,720 --> 00:20:24,820
went up because,
basically, guys don't care

469
00:20:24,820 --> 00:20:26,410
about kids as much as women do.

470
00:20:26,410 --> 00:20:27,842
Sorry, guys.

471
00:20:27,842 --> 00:20:30,570
It's just-- I don't know what
the evolutionary biology is.

472
00:20:30,570 --> 00:20:32,580
At least, in the
'70s, they didn't.

473
00:20:32,580 --> 00:20:35,027
So that's a rejection of
family utility maximization.

474
00:20:35,027 --> 00:20:36,610
Who had the dollars
actually mattered.

475
00:20:36,610 --> 00:20:37,985
It's kind of a
neat study for how

476
00:20:37,985 --> 00:20:39,718
you think about these theories.

477
00:20:39,718 --> 00:20:42,010
So it suggests that, secondhand
smoke, probably, people

478
00:20:42,010 --> 00:20:43,677
don't perfectly
maximize family utility.

479
00:20:43,677 --> 00:20:48,270
So there probably are
some externalities, OK?

480
00:20:48,270 --> 00:20:50,250
So that's a negative
consumption externality.

481
00:20:50,250 --> 00:20:53,350
There can also, of course,
be positive externalities.

482
00:20:53,350 --> 00:20:59,020
So let's talk about a positive
consumption externality.

483
00:20:59,020 --> 00:21:01,120
Let's talk about my neighbor.

484
00:21:01,120 --> 00:21:05,030
I don't get along
with my neighbor, OK?

485
00:21:05,030 --> 00:21:07,580
And, partly, it's
because my neighbor

486
00:21:07,580 --> 00:21:09,710
has a habit of
starting big projects

487
00:21:09,710 --> 00:21:11,450
and leaving them half done.

488
00:21:11,450 --> 00:21:14,390
And, about 25 years
ago, 20 years ago,

489
00:21:14,390 --> 00:21:17,140
he started a big project
to landscape his yard,

490
00:21:17,140 --> 00:21:19,130
created these huge
mounds of dirt

491
00:21:19,130 --> 00:21:23,170
that I look at directly from
my kitchen, and then stopped.

492
00:21:23,170 --> 00:21:27,650
So, for 20 years, I've had to
stare at huge mounds of dirt,

493
00:21:27,650 --> 00:21:28,640
OK?

494
00:21:28,640 --> 00:21:33,170
Now let's think about
my neighbor's decision

495
00:21:33,170 --> 00:21:35,850
to go ahead and get rid
of those piles of dirt.

496
00:21:35,850 --> 00:21:40,290
Let's say that that
would cost $1,000.

497
00:21:40,290 --> 00:21:44,310
So the cost is $1,000, OK?

498
00:21:44,310 --> 00:21:47,130
And let's say the
benefit to my neighbor

499
00:21:47,130 --> 00:21:48,990
from doing so is clearly
less than $1,000,

500
00:21:48,990 --> 00:21:50,320
or he would have done it.

501
00:21:50,320 --> 00:21:52,590
Let's say it's $800.

502
00:21:52,590 --> 00:21:54,900
So that's why he leaves
those piles of dirt

503
00:21:54,900 --> 00:21:56,840
because the cost is
$1,000 to remove them,

504
00:21:56,840 --> 00:21:58,860
and it's only worth $800 to him.

505
00:21:58,860 --> 00:22:02,080
But what he's not accounting
for is, if he removed them,

506
00:22:02,080 --> 00:22:09,020
there'd be a positive benefit
to me of another $300.

507
00:22:09,020 --> 00:22:11,180
So, actually, the
total social benefit

508
00:22:11,180 --> 00:22:15,540
of removing the dirt piles is
higher than the social cost.

509
00:22:15,540 --> 00:22:19,160
So, from a social
perspective, he should do it,

510
00:22:19,160 --> 00:22:21,410
but he doesn't because,
privately, it's

511
00:22:21,410 --> 00:22:24,770
not optimal to do so.

512
00:22:24,770 --> 00:22:29,060
So that is a positive
consumption externality.

513
00:22:29,060 --> 00:22:29,660
Yeah?

514
00:22:29,660 --> 00:22:32,230
STUDENT: Does this mean you'd
be willing to pay him like $200

515
00:22:32,230 --> 00:22:32,730
to do it?

516
00:22:32,730 --> 00:22:34,105
JONATHAN GRUBER:
Well, this leads

517
00:22:34,105 --> 00:22:38,270
to a very deep question, which
is can't all externalities

518
00:22:38,270 --> 00:22:40,070
simply be internalized.

519
00:22:40,070 --> 00:22:42,330
Let's take this example.

520
00:22:42,330 --> 00:22:47,230
Why can't I just go over
and offer to pay him, OK?

521
00:22:47,230 --> 00:22:49,480
Well, and, in fact, with
any example, we can do that.

522
00:22:49,480 --> 00:22:53,820
Why can't the fishermen just
go pay the steel plant, OK?

523
00:22:53,820 --> 00:22:58,523
Why can't you pay me not
to smoke in class, OK?

524
00:22:58,523 --> 00:23:00,190
Indeed, with any of
these externalities,

525
00:23:00,190 --> 00:23:02,942
there's a question why can't
they all be internalized.

526
00:23:02,942 --> 00:23:04,650
And, indeed, there's
a school of thought,

527
00:23:04,650 --> 00:23:06,050
which suggests that
externalities aren't really

528
00:23:06,050 --> 00:23:06,550
a problem.

529
00:23:06,550 --> 00:23:08,130
They can all just
be internalized.

530
00:23:08,130 --> 00:23:10,860
But, of course, that's
totally wrong, OK?

531
00:23:10,860 --> 00:23:12,340
Let's start with a hard example.

532
00:23:12,340 --> 00:23:16,590
Let's talk about the biggest
environmental externality,

533
00:23:16,590 --> 00:23:19,960
which is global warming, OK?

534
00:23:19,960 --> 00:23:24,980
With global warming, every
single time you drive,

535
00:23:24,980 --> 00:23:27,740
you are bringing
people of Bangladesh

536
00:23:27,740 --> 00:23:30,170
that much closer to
being under water.

537
00:23:30,170 --> 00:23:31,985
How could you possibly
negotiate that?

538
00:23:31,985 --> 00:23:34,610
How could you possibly negotiate
where the people of Bangladesh

539
00:23:34,610 --> 00:23:35,970
would come and say, well,
drive a little bit less

540
00:23:35,970 --> 00:23:37,530
so I don't go under water?

541
00:23:37,530 --> 00:23:40,010
OK, that's not happening, OK?

542
00:23:40,010 --> 00:23:42,590
But, even with these simple
cases, think about this.

543
00:23:42,590 --> 00:23:46,580
Why can't I just go to my
neighbor and offer him $200?

544
00:23:46,580 --> 00:23:48,490
There's three problems.

545
00:23:48,490 --> 00:23:54,700
First problem is there's
the fact that I don't really

546
00:23:54,700 --> 00:23:59,700
know what his costs are
and what his benefits are

547
00:23:59,700 --> 00:24:01,620
and that I might--

548
00:24:01,620 --> 00:24:05,450
I don't want to offer more than
I have to to get him to do it,

549
00:24:05,450 --> 00:24:06,040
OK?

550
00:24:06,040 --> 00:24:08,410
The second problem is he
doesn't know how I feel.

551
00:24:08,410 --> 00:24:10,120
So there's an
information asymmetry,

552
00:24:10,120 --> 00:24:11,703
which makes negotiations hard.

553
00:24:11,703 --> 00:24:13,120
There's a third
problem too, which

554
00:24:13,120 --> 00:24:17,093
is it'd just be deeply
weird to do that, right?

555
00:24:17,093 --> 00:24:18,760
I mean, it's just not
how society works.

556
00:24:18,760 --> 00:24:20,800
You think about a classic
case of an externality you've

557
00:24:20,800 --> 00:24:23,008
probably all run into, which
is your neighbor playing

558
00:24:23,008 --> 00:24:25,710
their music too loud, OK?

559
00:24:25,710 --> 00:24:27,960
If your neighbor plays
their music too loud,

560
00:24:27,960 --> 00:24:31,390
that is an externality
on you, OK?

561
00:24:31,390 --> 00:24:33,390
Now, in principle, you
could go to your neighbor

562
00:24:33,390 --> 00:24:36,420
and say, well, look,
I'm studying for a test.

563
00:24:36,420 --> 00:24:39,100
This test will raise
my grade by 10 points.

564
00:24:39,100 --> 00:24:41,640
A higher grade in this
class will raise my earnings

565
00:24:41,640 --> 00:24:43,620
by $1,000.

566
00:24:43,620 --> 00:24:47,028
So I'm willing to pay you $83 to
stop playing music because I've

567
00:24:47,028 --> 00:24:48,570
calculated my lifetime
earning effect

568
00:24:48,570 --> 00:24:50,210
of your playing the music.

569
00:24:50,210 --> 00:24:54,490
OK, even at MIT, that would
be sort of deeply weird to do.

570
00:24:54,490 --> 00:24:55,240
So what do you do?

571
00:24:55,240 --> 00:24:58,590
You either shut up about
it, or you go yell at them.

572
00:24:58,590 --> 00:24:59,888
OK, but yelling induces--

573
00:24:59,888 --> 00:25:01,430
it's not necessarily
an efficient way

574
00:25:01,430 --> 00:25:03,690
to resolve it because maybe they
really want to play the music.

575
00:25:03,690 --> 00:25:05,420
The efficient thing would be,
if it's worth more than $83

576
00:25:05,420 --> 00:25:07,045
for them to play the
music, they should

577
00:25:07,045 --> 00:25:08,930
get to play it and
just pay you $83.

578
00:25:08,930 --> 00:25:10,470
But, in fact, that doesn't work.

579
00:25:10,470 --> 00:25:12,470
So, in fact, private
solutions to these problems

580
00:25:12,470 --> 00:25:14,673
simply do not work, OK?

581
00:25:14,673 --> 00:25:17,090
It's just hard to figure out
how you can really get people

582
00:25:17,090 --> 00:25:23,945
privately to internalize
these externalities

583
00:25:23,945 --> 00:25:26,570
because negotiation is difficult
and because it's just socially

584
00:25:26,570 --> 00:25:27,320
awkward.

585
00:25:27,320 --> 00:25:28,940
There is a famous
apocryphal story

586
00:25:28,940 --> 00:25:31,743
told of a famous economist
who was on a flight

587
00:25:31,743 --> 00:25:33,410
and wanted to get
work done and couldn't

588
00:25:33,410 --> 00:25:35,618
because the person next to
him wouldn't stop talking.

589
00:25:35,618 --> 00:25:37,515
So they actually offered
them $10 to shut up.

590
00:25:37,515 --> 00:25:39,140
I don't believe that
actually happened,

591
00:25:39,140 --> 00:25:42,240
but, you know, it makes
for a good story, OK?

592
00:25:42,240 --> 00:25:44,120
So that's a positive
consumption externality.

593
00:25:44,120 --> 00:25:48,620
Finally, we have positive
production externalities.

594
00:25:52,080 --> 00:25:53,940
The classic example of
a positive production

595
00:25:53,940 --> 00:26:00,420
externality is R&D
by private firms, OK?

596
00:26:00,420 --> 00:26:03,530
When a firm does
research and development,

597
00:26:03,530 --> 00:26:06,370
they don't just create
learning for themselves.

598
00:26:06,370 --> 00:26:09,750
They create learning that might
benefit other firms as well,

599
00:26:09,750 --> 00:26:10,630
OK?

600
00:26:10,630 --> 00:26:12,610
And, indeed, the best
economic estimates

601
00:26:12,610 --> 00:26:16,080
suggest that the social
returns to $1 of R&D

602
00:26:16,080 --> 00:26:18,840
are 2 and 1/2 times
the private returns,

603
00:26:18,840 --> 00:26:21,000
that every dollar
of R&D a firm does

604
00:26:21,000 --> 00:26:23,170
benefits society
by 2 and 1/2 times

605
00:26:23,170 --> 00:26:25,550
how much it benefits the firm.

606
00:26:25,550 --> 00:26:31,130
And, as a result, firms
under invest in R&D, OK?

607
00:26:31,130 --> 00:26:35,570
As a result, firms
under invest in R&D, OK?

608
00:26:35,570 --> 00:26:39,318
And they are-- and
that's leading them to--

609
00:26:39,318 --> 00:26:41,610
that's leading to too little
R&D being done in society,

610
00:26:41,610 --> 00:26:43,790
and that affects all of us
because that affects growth.

611
00:26:43,790 --> 00:26:45,623
That's what my new book
Jumpstarting America

612
00:26:45,623 --> 00:26:46,400
is all about.

613
00:26:46,400 --> 00:26:47,942
It's about why we
need the government

614
00:26:47,942 --> 00:26:50,790
to come in and invest more in
R&D because firms under invest,

615
00:26:50,790 --> 00:26:51,350
OK?

616
00:26:51,350 --> 00:26:54,710
Essentially, firms don't
account for the spillovers

617
00:26:54,710 --> 00:26:56,910
that their investments
have on others.

618
00:26:56,910 --> 00:26:58,880
So a great example
used in the book

619
00:26:58,880 --> 00:27:01,400
is the example of when
two drug companies were

620
00:27:01,400 --> 00:27:02,705
racing to invent statins.

621
00:27:02,705 --> 00:27:04,580
You guys are too young
to know about statins,

622
00:27:04,580 --> 00:27:07,472
but statins are basically
a cholesterol-lowering drug

623
00:27:07,472 --> 00:27:08,180
that's a miracle.

624
00:27:08,180 --> 00:27:10,305
It saves hundreds of
thousands of lives every year.

625
00:27:10,305 --> 00:27:11,750
Best guess is
about 200,000 lives

626
00:27:11,750 --> 00:27:14,030
a year are saved by people
lowering their cholesterol

627
00:27:14,030 --> 00:27:15,380
through being on statins.

628
00:27:15,380 --> 00:27:17,630
Statins were being
invented in the early 1980s

629
00:27:17,630 --> 00:27:20,780
by two rival drug companies,
Merck in the US and Sankyo

630
00:27:20,780 --> 00:27:21,750
in Japan.

631
00:27:21,750 --> 00:27:23,900
And they were racing to
develop these statins.

632
00:27:23,900 --> 00:27:25,632
And then Sankyo
suddenly stopped.

633
00:27:25,632 --> 00:27:27,340
And Merck found out
through the grapevine

634
00:27:27,340 --> 00:27:29,450
it was because some
dogs got sick in drug--

635
00:27:29,450 --> 00:27:31,128
in the animal trials.

636
00:27:31,128 --> 00:27:32,670
And so Merck went
to Sankyo and said,

637
00:27:32,670 --> 00:27:34,040
hey, we heard some
dogs are sick.

638
00:27:34,040 --> 00:27:34,707
What's going on?

639
00:27:34,707 --> 00:27:36,380
Sankyo said we're not
going to tell you.

640
00:27:36,380 --> 00:27:37,712
You're our competitor.

641
00:27:37,712 --> 00:27:39,170
So Merck offered
to pay them money.

642
00:27:39,170 --> 00:27:40,170
They offered to partner.

643
00:27:40,170 --> 00:27:41,545
Sankyo said no way.

644
00:27:41,545 --> 00:27:42,920
This is private
R&D, and we don't

645
00:27:42,920 --> 00:27:44,003
want to share it with you.

646
00:27:44,003 --> 00:27:45,990
So Merck stopped too.

647
00:27:45,990 --> 00:27:48,420
Five years later, some
academics got permission

648
00:27:48,420 --> 00:27:49,590
to run trials on statins.

649
00:27:49,590 --> 00:27:51,210
It turns out that what happened
to the dogs had nothing

650
00:27:51,210 --> 00:27:52,265
to do with the drug.

651
00:27:52,265 --> 00:27:52,890
They were fine.

652
00:27:52,890 --> 00:27:54,060
It's totally safe.

653
00:27:54,060 --> 00:27:57,300
And statins were invented and
save 200,000 lives a year,

654
00:27:57,300 --> 00:27:59,160
but five years after
they should have.

655
00:27:59,160 --> 00:28:02,550
Literally, one
million people died

656
00:28:02,550 --> 00:28:04,080
because there was--
they could not

657
00:28:04,080 --> 00:28:07,170
benefit from the spillovers
of R&D knowledge, OK?

658
00:28:07,170 --> 00:28:09,540
This is an example of what
we mean by under-investment

659
00:28:09,540 --> 00:28:11,682
in R&D, OK?

660
00:28:11,682 --> 00:28:13,890
So we have externalities
can be negative or positive.

661
00:28:13,890 --> 00:28:16,290
They can be production
side or consumption side.

662
00:28:16,290 --> 00:28:18,320
Questions about that?

663
00:28:18,320 --> 00:28:18,860
Yeah?

664
00:28:18,860 --> 00:28:21,470
STUDENT: So, regardless if
they're negative or positive,

665
00:28:21,470 --> 00:28:23,500
externalities still
create deadweight loss.

666
00:28:23,500 --> 00:28:24,190
JONATHAN GRUBER: Yes,
regardless if they're

667
00:28:24,190 --> 00:28:26,030
negative or positive,
externalities still create

668
00:28:26,030 --> 00:28:27,947
deadweight loss because,
if they're positive--

669
00:28:27,947 --> 00:28:31,690
I should-- Jason,
just sent me a note.

670
00:28:31,690 --> 00:28:34,315
Next year, we should have in
the handout a positive graph.

671
00:28:34,315 --> 00:28:35,440
Well, I'll just do it here.

672
00:28:35,440 --> 00:28:38,030
I can-- I'm capable
of drawing a graph.

673
00:28:38,030 --> 00:28:42,400
So let's think about R&D. OK,
here's the quantity of R&D, OK?

674
00:28:42,400 --> 00:28:45,226
Here's the cost of
R&D, the price of R&D.

675
00:28:45,226 --> 00:28:47,090
So basically-- and here's the--

676
00:28:51,490 --> 00:28:54,070
so here's going to be
the demand for R&D, OK?

677
00:28:54,070 --> 00:28:55,570
Here's going to be
the demand, which

678
00:28:55,570 --> 00:28:57,640
is private marginal benefit.

679
00:28:57,640 --> 00:29:03,753
And here's the supply, which is
the private marginal cost, OK?

680
00:29:03,753 --> 00:29:05,170
And let's just say
that there's no

681
00:29:05,170 --> 00:29:07,340
externalities from actually--

682
00:29:07,340 --> 00:29:09,065
this is a consumption
externality.

683
00:29:09,065 --> 00:29:11,360
This is a production
externality.

684
00:29:11,360 --> 00:29:14,410
So, basically, the point is
that, when a firm does R&D,

685
00:29:14,410 --> 00:29:17,830
they do it until the
benefits to the firm

686
00:29:17,830 --> 00:29:19,340
equal the costs to the firm.

687
00:29:19,340 --> 00:29:24,210
So they do an optimum-- they do
an amount of R&D of, you know,

688
00:29:24,210 --> 00:29:29,180
Q1 at a price of P1.

689
00:29:29,180 --> 00:29:32,390
But what they're missing is
that, in fact, what they're

690
00:29:32,390 --> 00:29:36,980
missing is that
the costs to them

691
00:29:36,980 --> 00:29:41,250
are actually well
below, well below,

692
00:29:41,250 --> 00:29:43,710
what it truly costs because
they are benefiting others

693
00:29:43,710 --> 00:29:45,810
by doing it, OK?

694
00:29:45,810 --> 00:29:51,118
So the supply curve, the
true social supply curve,

695
00:29:51,118 --> 00:29:51,660
is down here.

696
00:29:51,660 --> 00:29:55,080
The social marginal cost is
the private marginal cost

697
00:29:55,080 --> 00:30:04,470
minus the social benefits that
we get from doing that R&D.

698
00:30:04,470 --> 00:30:08,920
So, as a result, they should be
doing Q2 R&D, but they're not.

699
00:30:08,920 --> 00:30:10,710
They're doing too
little R&D. And that's

700
00:30:10,710 --> 00:30:12,330
making a deadweight loss.

701
00:30:12,330 --> 00:30:14,640
The deadweight loss,
remember, is with reference

702
00:30:14,640 --> 00:30:15,870
to the optimal point.

703
00:30:15,870 --> 00:30:17,770
This is the deadweight loss.

704
00:30:17,770 --> 00:30:20,640
This is deadweight loss
from under producing

705
00:30:20,640 --> 00:30:23,447
R&D. The difference between--

706
00:30:23,447 --> 00:30:24,030
no, I'm sorry.

707
00:30:24,030 --> 00:30:24,660
I got that wrong.

708
00:30:24,660 --> 00:30:26,180
These triangles are
always confusing.

709
00:30:26,180 --> 00:30:27,555
In the drawing,
I got that wrong.

710
00:30:27,555 --> 00:30:30,210
OK, it's the difference being
the social marginal cost,

711
00:30:30,210 --> 00:30:32,160
between the social
marginal cost,

712
00:30:32,160 --> 00:30:34,210
and the private
marginal benefit.

713
00:30:34,210 --> 00:30:37,845
So, basically, let me think
about this for one second.

714
00:30:37,845 --> 00:30:39,470
This is always a
little bit hard to do.

715
00:30:39,470 --> 00:30:41,340
So, basically, there are units--

716
00:30:41,340 --> 00:30:43,220
there are units
they under produce.

717
00:30:43,220 --> 00:30:46,460
So, essentially, what their--
they should be producing this

718
00:30:46,460 --> 00:30:49,127
many units, yeah, between the--

719
00:30:49,127 --> 00:30:51,210
it's where the private
marginal-- yeah, it's this.

720
00:30:51,210 --> 00:30:53,130
I had it right.

721
00:30:53,130 --> 00:30:55,130
That's the deadweight loss, OK?

722
00:30:55,130 --> 00:30:57,420
Yeah, so, basically,
what you have is

723
00:30:57,420 --> 00:30:59,670
you're going to have
underproduction of R&D,

724
00:30:59,670 --> 00:31:02,490
just like you had
overproduction of steel, OK?

725
00:31:02,490 --> 00:31:03,574
Yeah?

726
00:31:03,574 --> 00:31:05,520
STUDENT: Why would the
social marginal cost

727
00:31:05,520 --> 00:31:08,545
go down when the social
marginal benefit goes up?

728
00:31:08,545 --> 00:31:10,420
JONATHAN GRUBER: Because
the marginal benefit

729
00:31:10,420 --> 00:31:14,860
is what's the marginal benefit
of another dollar of R&D.

730
00:31:14,860 --> 00:31:16,150
That's basically--

731
00:31:16,150 --> 00:31:19,540
that's, essentially, what's
the knowledge created, OK?

732
00:31:19,540 --> 00:31:21,840
You could view this either
way, but the idea here

733
00:31:21,840 --> 00:31:24,520
is I'm producing R&D.
I'm not consuming R&D.

734
00:31:24,520 --> 00:31:26,270
This is the benefit
of consuming R&D. This

735
00:31:26,270 --> 00:31:28,187
is sort of the benefit
of society of consuming

736
00:31:28,187 --> 00:31:31,030
that R&D. So you think of it as
lowering the cost of producing

737
00:31:31,030 --> 00:31:34,610
the R&D is sort of the
way we think about it, OK?

738
00:31:34,610 --> 00:31:36,590
But the main thing is not--

739
00:31:36,590 --> 00:31:37,176
yeah?

740
00:31:37,176 --> 00:31:39,765
STUDENT: So is the
marginal change

741
00:31:39,765 --> 00:31:45,200
between the social curve and the
private curve, is that linear?

742
00:31:45,200 --> 00:31:46,192
Or, as it grows--

743
00:31:46,192 --> 00:31:47,900
JONATHAN GRUBER: That's
a great question.

744
00:31:47,900 --> 00:31:48,817
I'm always making it--

745
00:31:48,817 --> 00:31:49,790
I'm making it constant.

746
00:31:49,790 --> 00:31:52,760
I'm assuming marginal damage or
marginal benefits are constant.

747
00:31:52,760 --> 00:31:55,390
STUDENT: So, mathematically,
whichever triangle

748
00:31:55,390 --> 00:31:56,805
you choose is the same?

749
00:31:56,805 --> 00:31:59,430
JONATHAN GRUBER: Yeah, exactly,
because I'm making it constant.

750
00:31:59,430 --> 00:32:01,097
And, in fact, you
could imagine it could

751
00:32:01,097 --> 00:32:03,290
be growing or shrinking, OK?

752
00:32:03,290 --> 00:32:05,803
Now, with this in
mind and realizing

753
00:32:05,803 --> 00:32:07,970
that private sector can't
solve this for the reasons

754
00:32:07,970 --> 00:32:11,180
we talked about, let's
talk about the role--

755
00:32:11,180 --> 00:32:14,531
let's talk about the
role of government,

756
00:32:14,531 --> 00:32:18,300
government solutions.

757
00:32:18,300 --> 00:32:20,120
So, once again,
remember the basic logic

758
00:32:20,120 --> 00:32:23,180
of this class, which is that--

759
00:32:23,180 --> 00:32:28,650
the basic logic of the
class, which is that if--

760
00:32:28,650 --> 00:32:31,670
the basic logic of the class
is that the market knows best

761
00:32:31,670 --> 00:32:33,072
unless there's a market failure.

762
00:32:33,072 --> 00:32:35,030
If there's a market
failure, the private market

763
00:32:35,030 --> 00:32:36,470
will deliver a deadweight loss.

764
00:32:36,470 --> 00:32:38,720
Now we have to ask can the
government actually make it

765
00:32:38,720 --> 00:32:39,260
better.

766
00:32:39,260 --> 00:32:41,177
Remember, we talked about
monopoly regulation.

767
00:32:41,177 --> 00:32:44,743
The government may or may
not make it better, OK?

768
00:32:44,743 --> 00:32:47,160
Information asymmetries, it
may or may not make it better.

769
00:32:47,160 --> 00:32:48,320
Same thing with
externalities, the government

770
00:32:48,320 --> 00:32:49,130
may not make it better.

771
00:32:49,130 --> 00:32:51,213
So let's talk about how
the government, in theory,

772
00:32:51,213 --> 00:32:52,658
could make it better, OK?

773
00:32:52,658 --> 00:32:54,200
Well, there's two
ways the government

774
00:32:54,200 --> 00:32:55,460
could make it better.

775
00:32:55,460 --> 00:32:58,100
One way is by regulation.

776
00:32:58,100 --> 00:33:03,350
So go back to figure 22-1, OK?

777
00:33:03,350 --> 00:33:05,480
The government could
literally regulate

778
00:33:05,480 --> 00:33:08,240
and could say, look, we know
the optimal level of steel

779
00:33:08,240 --> 00:33:10,290
to be produced is Q2.

780
00:33:10,290 --> 00:33:12,740
We're just going to
tell you to produce Q2.

781
00:33:12,740 --> 00:33:13,670
That's it.

782
00:33:13,670 --> 00:33:15,770
Problem solved, OK?

783
00:33:15,770 --> 00:33:18,290
We just say, hey, steel plant,
we know the optimum is Q2.

784
00:33:18,290 --> 00:33:19,010
You produce Q2.

785
00:33:19,010 --> 00:33:20,690
Problem solved.

786
00:33:20,690 --> 00:33:23,180
The problem with that is
that requires the government

787
00:33:23,180 --> 00:33:24,980
to know quite a lot.

788
00:33:24,980 --> 00:33:26,900
The government needs
to know both the supply

789
00:33:26,900 --> 00:33:31,310
and the demand curves to
figure out where Q2 is

790
00:33:31,310 --> 00:33:33,560
and what they should regulate.

791
00:33:33,560 --> 00:33:36,462
Let's say all the government
knows is the damage being done,

792
00:33:36,462 --> 00:33:37,920
and let's say the
damage is linear,

793
00:33:37,920 --> 00:33:40,550
or they can approximate
it as linear.

794
00:33:40,550 --> 00:33:43,390
Then there's a much
easier solution,

795
00:33:43,390 --> 00:33:46,213
which is a corrective tax.

796
00:33:46,213 --> 00:33:47,880
What if the government
came in and said,

797
00:33:47,880 --> 00:33:50,217
look, I don't where demand
and supply curves are?

798
00:33:50,217 --> 00:33:51,050
I don't really know.

799
00:33:51,050 --> 00:33:52,467
It's really hard
to figure it out.

800
00:33:52,467 --> 00:33:56,035
All I know is that, for every
unit of steel you produce,

801
00:33:56,035 --> 00:33:57,410
which is a unit
of sludge, you're

802
00:33:57,410 --> 00:34:00,348
killing $100 worth of fish.

803
00:34:00,348 --> 00:34:01,140
That's what I know.

804
00:34:01,140 --> 00:34:05,170
That I can study
environmentally, OK?

805
00:34:05,170 --> 00:34:08,040
What if I simply
tax the steel plant

806
00:34:08,040 --> 00:34:10,830
by $100 for every unit
of steel they produced?

807
00:34:10,830 --> 00:34:13,389
That is, if I imposed a
tax on the steel plant--

808
00:34:13,389 --> 00:34:14,763
STUDENT: Don't you mean sludge?

809
00:34:14,763 --> 00:34:16,180
JONATHAN GRUBER:
One unit of steel

810
00:34:16,180 --> 00:34:18,190
is one unit of sludge
in this example.

811
00:34:18,190 --> 00:34:21,520
So I'm going to tax every
unit of steel they produce,

812
00:34:21,520 --> 00:34:24,880
which is the same as producing
one unit of sludge, OK?

813
00:34:24,880 --> 00:34:26,170
What if I impose that tax?

814
00:34:26,170 --> 00:34:28,870
Well, let's look at figure 22-3.

815
00:34:28,870 --> 00:34:33,440
What does that do to
the firm's decision, OK?

816
00:34:33,440 --> 00:34:34,940
Well, before the
government came in,

817
00:34:34,940 --> 00:34:37,380
the firm was
producing at point A

818
00:34:37,380 --> 00:34:39,130
where their private
marginal costs equaled

819
00:34:39,130 --> 00:34:40,588
the private marginal
benefit, which

820
00:34:40,588 --> 00:34:42,170
is the social marginal benefit.

821
00:34:42,170 --> 00:34:45,860
Now the government comes
in and levies a tax.

822
00:34:45,860 --> 00:34:49,903
It levies a tax at exactly
MD, the marginal damage.

823
00:34:49,903 --> 00:34:50,570
What does it do?

824
00:34:50,570 --> 00:34:53,300
It shifts their private
marginal cost curve

825
00:34:53,300 --> 00:34:56,290
to the social
marginal cost curve.

826
00:34:56,290 --> 00:35:00,240
It has caused the firm to
internalize the externality

827
00:35:00,240 --> 00:35:03,510
because now the firm is
paying an amount exactly

828
00:35:03,510 --> 00:35:06,790
equal to the damage
they're doing to society.

829
00:35:06,790 --> 00:35:10,710
So a corrective tax
can cause the firm

830
00:35:10,710 --> 00:35:14,090
to internalize the externality.

831
00:35:14,090 --> 00:35:18,128
Corrective tax caused the firm
to internalize the externality,

832
00:35:18,128 --> 00:35:18,960
OK?

833
00:35:18,960 --> 00:35:21,270
Essentially, a corrective
tax by the government

834
00:35:21,270 --> 00:35:24,420
can get us to the right
answer because it gets firms

835
00:35:24,420 --> 00:35:26,720
to do the right thing, OK?

836
00:35:26,720 --> 00:35:29,990
It gets firms to pay attention
to the social costs, not just

837
00:35:29,990 --> 00:35:31,340
the--

838
00:35:31,340 --> 00:35:34,345
not just the private costs.

839
00:35:34,345 --> 00:35:35,720
Similarly, we
could do same thing

840
00:35:35,720 --> 00:35:37,250
with a positive externality.

841
00:35:37,250 --> 00:35:40,950
What could the government do
with a positive externality?

842
00:35:40,950 --> 00:35:41,450
Yeah?

843
00:35:41,450 --> 00:35:42,350
STUDENT: Subsidize production.

844
00:35:42,350 --> 00:35:44,017
JONATHAN GRUBER: It
could subsidize by--

845
00:35:44,017 --> 00:35:48,380
so imagine I knew
exactly how much social

846
00:35:48,380 --> 00:35:51,500
benefit there was per unit
of R&D. If I subsidized

847
00:35:51,500 --> 00:35:55,560
the firm doing R&D, then I
would lower their costs, right?

848
00:35:55,560 --> 00:35:58,100
If I offered them a
subsidy of this amount,

849
00:35:58,100 --> 00:36:00,262
their cost curve
would shift down.

850
00:36:00,262 --> 00:36:01,220
I would lower the cost.

851
00:36:01,220 --> 00:36:04,110
So R&D would get
to the right point.

852
00:36:04,110 --> 00:36:07,100
So a corrective tax of
the amount of damage

853
00:36:07,100 --> 00:36:09,590
gets firms to internalize
the externality.

854
00:36:09,590 --> 00:36:12,020
A corrective subsidy
of that amount

855
00:36:12,020 --> 00:36:15,530
gets firms to internalize the
externality or individuals

856
00:36:15,530 --> 00:36:17,480
to externalize the internality.

857
00:36:17,480 --> 00:36:19,790
And we can get to
the optimal outcome

858
00:36:19,790 --> 00:36:22,160
by the government
imposing a corrective tax

859
00:36:22,160 --> 00:36:23,820
or subsidy of the right amount.

860
00:36:23,820 --> 00:36:26,360
Now we could of course also
get there with regulation.

861
00:36:26,360 --> 00:36:27,470
It's just a lot harder.

862
00:36:30,200 --> 00:36:32,410
Questions about that?

863
00:36:32,410 --> 00:36:37,490
OK, this is our first
example of good taxes, OK?

864
00:36:37,490 --> 00:36:40,490
Taxes have been bad
throughout this course.

865
00:36:40,490 --> 00:36:43,375
The role of taxes has been
distortionary to the economy.

866
00:36:43,375 --> 00:36:44,750
We haven't talked
about it a lot.

867
00:36:44,750 --> 00:36:46,792
We'll talk about it more
in a couple of lectures.

868
00:36:46,792 --> 00:36:48,840
They've been distortionary
to the economy.

869
00:36:48,840 --> 00:36:51,110
This is saying, no,
a tax can actually

870
00:36:51,110 --> 00:36:55,250
play a positive role because
a tax can correct a market

871
00:36:55,250 --> 00:36:57,120
failure.

872
00:36:57,120 --> 00:36:59,180
Now, as always, if the
tax is set incorrectly,

873
00:36:59,180 --> 00:37:00,748
it could make things worse.

874
00:37:00,748 --> 00:37:03,290
OK, if you set a tax that was
five times the marginal damage,

875
00:37:03,290 --> 00:37:04,887
it would make things worse.

876
00:37:04,887 --> 00:37:07,220
But, if you set it correctly,
it can make things better.

877
00:37:07,220 --> 00:37:09,553
We're offering the potential
for government intervention

878
00:37:09,553 --> 00:37:11,070
to make things better here.

879
00:37:11,070 --> 00:37:13,400
Questions about that?

880
00:37:13,400 --> 00:37:15,135
OK, so what do we have?

881
00:37:15,135 --> 00:37:17,260
We have a situation where
the private market is not

882
00:37:17,260 --> 00:37:21,120
delivering the optimal outcome.

883
00:37:21,120 --> 00:37:23,460
The private market is not
delivering the optimal outcome

884
00:37:23,460 --> 00:37:26,713
where it seems hard to
think of private solutions,

885
00:37:26,713 --> 00:37:29,130
but where a government solution,
either through regulation

886
00:37:29,130 --> 00:37:31,920
or easier corrective
taxation, can get us

887
00:37:31,920 --> 00:37:34,000
to the optimal outcome.

888
00:37:34,000 --> 00:37:36,465
So now let's ask how does this
actually work in practice.

889
00:37:40,050 --> 00:37:41,760
And let's talk
about two examples.

890
00:37:41,760 --> 00:37:44,430
Let's talk about environmental
externalities and health

891
00:37:44,430 --> 00:37:45,700
externalities.

892
00:37:49,965 --> 00:37:51,590
Start with environmental
externalities.

893
00:37:51,590 --> 00:37:53,630
And, of course,
the most important

894
00:37:53,630 --> 00:37:56,685
is global warming, OK?

895
00:37:56,685 --> 00:37:59,060
Currently, the amount of carbon
dioxide in the atmosphere

896
00:37:59,060 --> 00:38:03,770
is at its highest
level in 400,000 years.

897
00:38:03,770 --> 00:38:06,380
Basically, every year becomes
the hottest year on record

898
00:38:06,380 --> 00:38:07,160
almost linearly.

899
00:38:07,160 --> 00:38:09,610
Almost monotonically, every year
is the hottest year on record.

900
00:38:09,610 --> 00:38:10,318
We're heating up.

901
00:38:10,318 --> 00:38:13,228
Scientists predict
that it's possible--

902
00:38:13,228 --> 00:38:15,020
the central prediction
is that temperatures

903
00:38:15,020 --> 00:38:19,072
will rise by more than 2 degrees
Fahrenheit by the end of the--

904
00:38:19,072 --> 00:38:20,780
I'm sorry, by more
than 2 degrees Celsius

905
00:38:20,780 --> 00:38:26,460
by the end of the century, but
it could be more than that.

906
00:38:26,460 --> 00:38:28,850
There's actually,
currently, the best estimate

907
00:38:28,850 --> 00:38:34,220
is that there's as
much as a 10% chance

908
00:38:34,220 --> 00:38:35,960
that temperatures
go up by 10 degrees

909
00:38:35,960 --> 00:38:37,543
by the end of the
century, which would

910
00:38:37,543 --> 00:38:40,400
end human life, basically,
in most of the world, OK?

911
00:38:40,400 --> 00:38:43,680
There is a non-trivial chance
we're all gone by 2100,

912
00:38:43,680 --> 00:38:46,190
not my problem, largely
not your problem,

913
00:38:46,190 --> 00:38:48,880
certainly your
kids' problem, OK?

914
00:38:48,880 --> 00:38:52,310
OK, by and large,
we are basically--

915
00:38:52,310 --> 00:38:55,375
we are basically-- we have a--
we know for sure there's going

916
00:38:55,375 --> 00:38:56,840
to be negative implications.

917
00:38:56,840 --> 00:38:59,630
Basically, we are essentially--

918
00:38:59,630 --> 00:39:02,590
unless there's a radical
new technology invented,

919
00:39:02,590 --> 00:39:03,920
Bangladesh is gone.

920
00:39:03,920 --> 00:39:05,180
It's over for Bangladesh.

921
00:39:05,180 --> 00:39:06,680
Cape Cod is gone.

922
00:39:06,680 --> 00:39:07,910
Much of Florida is gone.

923
00:39:07,910 --> 00:39:09,830
That's already happening, OK?

924
00:39:09,830 --> 00:39:12,080
At this point, the
question is can we actually

925
00:39:12,080 --> 00:39:15,200
stop the entire East and West
coasts and much of the South

926
00:39:15,200 --> 00:39:17,720
from disappearing as well
and many other countries

927
00:39:17,720 --> 00:39:20,360
in the world from
disappearing, OK?

928
00:39:20,360 --> 00:39:23,600
That's the sort of decision
we have to make now, OK?

929
00:39:23,600 --> 00:39:27,080
So, basically, this is a
classic negative externality

930
00:39:27,080 --> 00:39:28,825
because, that
negative situation,

931
00:39:28,825 --> 00:39:30,950
you were not thinking about
that when you filled up

932
00:39:30,950 --> 00:39:32,263
your car last time.

933
00:39:32,263 --> 00:39:33,680
You're not thinking
about the fact

934
00:39:33,680 --> 00:39:35,222
that the fossil
fuels you're emitting

935
00:39:35,222 --> 00:39:36,710
are contributing to that, OK?

936
00:39:36,710 --> 00:39:39,690
It's a classic
negative externality.

937
00:39:39,690 --> 00:39:41,060
So what can the government do?

938
00:39:41,060 --> 00:39:44,630
Well, the natural solution
would be corrective taxation.

939
00:39:44,630 --> 00:39:47,930
The natural solution would
be to have a carbon tax,

940
00:39:47,930 --> 00:39:50,570
to literally say this is
the amount-- we actually

941
00:39:50,570 --> 00:39:54,320
have a pretty good sense
from engineering models what

942
00:39:54,320 --> 00:39:59,030
the cost of carbon is, what the
marginal cost of carbon is, OK?

943
00:39:59,030 --> 00:40:01,190
And we could
literally impose a tax

944
00:40:01,190 --> 00:40:04,167
on the use of carbon
of that amount.

945
00:40:04,167 --> 00:40:05,750
I think it would
amount to something--

946
00:40:05,750 --> 00:40:06,860
I don't know the
numbers these days.

947
00:40:06,860 --> 00:40:09,230
It's something like between
$0.25 and $0.50 a gallon

948
00:40:09,230 --> 00:40:10,580
of gas.

949
00:40:10,580 --> 00:40:12,860
So it's a lot, but it's not--

950
00:40:12,860 --> 00:40:14,480
we've seen gas prices
in the last year

951
00:40:14,480 --> 00:40:16,200
go up and down by that much.

952
00:40:16,200 --> 00:40:19,168
OK, that's not an
outrageous amount, OK?

953
00:40:19,168 --> 00:40:20,960
In Europe, they already
have gasoline taxes

954
00:40:20,960 --> 00:40:23,510
well above that level, OK?

955
00:40:23,510 --> 00:40:27,923
So corrective taxation, in
principle, could be the answer.

956
00:40:27,923 --> 00:40:29,840
We could literally just
use engineering models

957
00:40:29,840 --> 00:40:32,760
to compute the costs,
social costs of carbon.

958
00:40:32,760 --> 00:40:34,350
We could put a tax on it.

959
00:40:34,350 --> 00:40:37,740
And then at least we would stop
global warming going forward.

960
00:40:37,740 --> 00:40:39,370
You know, Bangladesh
may be gone,

961
00:40:39,370 --> 00:40:43,170
but we can maybe save a lot
of the rest of the world, OK?

962
00:40:43,170 --> 00:40:45,810
So that's in theory.

963
00:40:45,810 --> 00:40:48,390
In practice, in
1994, Bill Clinton

964
00:40:48,390 --> 00:40:53,040
proposed a $0.03 gas tax
and lost Congress, OK?

965
00:40:53,040 --> 00:40:56,550
In practice, people
don't like gas taxes.

966
00:40:56,550 --> 00:40:59,890
It's very hard politically
in the US and other places.

967
00:40:59,890 --> 00:41:03,220
And that is why the world has
turned to a different approach,

968
00:41:03,220 --> 00:41:05,310
which is quantity
regulation, which

969
00:41:05,310 --> 00:41:08,490
is say, look, in practice, we
should have a global carbon

970
00:41:08,490 --> 00:41:09,033
tax.

971
00:41:09,033 --> 00:41:10,950
In theory, we should
have a global carbon tax.

972
00:41:10,950 --> 00:41:11,992
In practice, that's hard.

973
00:41:11,992 --> 00:41:14,700
That's why we have negotiations.

974
00:41:14,700 --> 00:41:17,040
That's why we try to
have a global negotiation

975
00:41:17,040 --> 00:41:19,770
to try to get a global
cap on carbon emissions,

976
00:41:19,770 --> 00:41:23,445
actually have a quantity
regulation, to actually have

977
00:41:23,445 --> 00:41:25,110
a quantity regulation.

978
00:41:25,110 --> 00:41:25,860
We started this.

979
00:41:25,860 --> 00:41:27,510
The first true
global negotiation

980
00:41:27,510 --> 00:41:32,857
was in Kyoto, Japan
in December 1997.

981
00:41:32,857 --> 00:41:35,190
I was fortunate enough to be
there for that negotiation.

982
00:41:35,190 --> 00:41:37,260
I was in the Clinton
administration at the time.

983
00:41:37,260 --> 00:41:41,320
And we got to go over
and do that negotiation.

984
00:41:41,320 --> 00:41:43,320
It was actually pretty
neat because they decided

985
00:41:43,320 --> 00:41:44,850
I was going at the last
minute, and the only plane left

986
00:41:44,850 --> 00:41:45,975
to go on was Air Force Two.

987
00:41:45,975 --> 00:41:48,350
So I got to fly over with the
vice president on Air Force

988
00:41:48,350 --> 00:41:50,240
Two, which was pretty
cool, super cool.

989
00:41:50,240 --> 00:41:51,520
They have really
nice seats and stuff.

990
00:41:51,520 --> 00:41:53,562
And so I sat down, and
the phone next to me rang.

991
00:41:53,562 --> 00:41:54,517
And I was like--

992
00:41:54,517 --> 00:41:55,100
I answered it.

993
00:41:55,100 --> 00:41:55,640
I was like hello.

994
00:41:55,640 --> 00:41:56,640
They're like, hey, John.

995
00:41:56,640 --> 00:41:57,510
I'm like oh my god.

996
00:41:57,510 --> 00:41:59,610
It was them calling from
Japan, but like getting

997
00:41:59,610 --> 00:42:02,190
a personal phone call on
a plane was super cool.

998
00:42:02,190 --> 00:42:03,930
So, anyway, so I
went over to Kyoto.

999
00:42:03,930 --> 00:42:07,890
I learned how these negotiations
work, which is, over five days,

1000
00:42:07,890 --> 00:42:10,440
I slept four hours.

1001
00:42:10,440 --> 00:42:12,090
In Japan, they sell--

1002
00:42:12,090 --> 00:42:14,020
they sold, at the time,
this coffee in cans.

1003
00:42:14,020 --> 00:42:16,325
So you just chug these
cans of coffee all the time

1004
00:42:16,325 --> 00:42:16,950
and stay awake.

1005
00:42:16,950 --> 00:42:19,325
And, basically, everyone is
so tired by the end they just

1006
00:42:19,325 --> 00:42:20,880
agree just to kind
of get it done.

1007
00:42:20,880 --> 00:42:23,550
And that's sort of way
that negotiations work.

1008
00:42:23,550 --> 00:42:25,590
So we agreed to the
Kyoto global warming

1009
00:42:25,590 --> 00:42:28,660
treaty, which would have
lowered emissions worldwide,

1010
00:42:28,660 --> 00:42:30,990
but the US did not sign on.

1011
00:42:30,990 --> 00:42:33,438
The US refused to sign on.

1012
00:42:33,438 --> 00:42:34,980
There's been continuing
negotiations.

1013
00:42:34,980 --> 00:42:37,770
Most recently, we know about
the Paris round of negotiations,

1014
00:42:37,770 --> 00:42:40,440
which the US did sign on to,
but the current administration

1015
00:42:40,440 --> 00:42:42,420
has pulled us back out of.

1016
00:42:42,420 --> 00:42:44,730
So we have a problem, which
is that, basically, we're

1017
00:42:44,730 --> 00:42:47,160
heading to this
environmental catastrophe,

1018
00:42:47,160 --> 00:42:52,380
and the world can't
agree on actions to take.

1019
00:42:52,380 --> 00:42:57,000
And it's not really a choice.

1020
00:42:57,000 --> 00:43:00,450
I mean, we have to do something,
or our grandkids will all

1021
00:43:00,450 --> 00:43:02,580
be under water, or
our great-grandkids

1022
00:43:02,580 --> 00:43:03,870
will all be under water.

1023
00:43:03,870 --> 00:43:05,850
We have to do something.

1024
00:43:05,850 --> 00:43:08,400
The optimistic case
that we'll do something

1025
00:43:08,400 --> 00:43:11,220
comes from the example of what's
called chlorofluorocarbons.

1026
00:43:11,220 --> 00:43:13,048
When I was a kid,
many, many products

1027
00:43:13,048 --> 00:43:15,090
were made with what's
called chlorofluorocarbons.

1028
00:43:15,090 --> 00:43:16,390
They were in refrigerators.

1029
00:43:16,390 --> 00:43:19,440
They were in aerosol
sprays, et cetera.

1030
00:43:19,440 --> 00:43:21,090
Scientists realized
they were actually

1031
00:43:21,090 --> 00:43:23,040
damaging the ozone
layer, which protects us

1032
00:43:23,040 --> 00:43:25,458
from ultraviolet
rays from the sun.

1033
00:43:25,458 --> 00:43:27,000
And people were
like, yeah, whatever,

1034
00:43:27,000 --> 00:43:29,458
much like they are with global
warming now, yeah, whatever.

1035
00:43:29,458 --> 00:43:32,550
But then a fucking hole
opened up in the ozone layer.

1036
00:43:32,550 --> 00:43:34,710
Like, literally, it was
like, oh my god, there's

1037
00:43:34,710 --> 00:43:36,090
a hole in the ozone layer.

1038
00:43:36,090 --> 00:43:38,160
And 180 countries got
together almost overnight

1039
00:43:38,160 --> 00:43:39,647
and banned chlorofluorocarbons.

1040
00:43:39,647 --> 00:43:41,730
Like, literally, almost
overnight, they were gone.

1041
00:43:41,730 --> 00:43:43,853
It was amazing,
international cooperation,

1042
00:43:43,853 --> 00:43:45,270
terrific international
cooperation

1043
00:43:45,270 --> 00:43:47,370
to take an environmental
catastrophe on and deal

1044
00:43:47,370 --> 00:43:47,870
with it.

1045
00:43:47,870 --> 00:43:49,662
The problem is global
warming doesn't quite

1046
00:43:49,662 --> 00:43:50,790
work that way, OK?

1047
00:43:50,790 --> 00:43:53,610
By the time we say, oh my god,
Bangladesh is under water,

1048
00:43:53,610 --> 00:43:55,265
it's too late.

1049
00:43:55,265 --> 00:43:56,640
So the question
is sort of how do

1050
00:43:56,640 --> 00:43:58,553
we get politicians
and the public

1051
00:43:58,553 --> 00:44:00,720
interested in taking something
on when we don't have

1052
00:44:00,720 --> 00:44:03,060
the symbol, like a hole
in the ozone layer,

1053
00:44:03,060 --> 00:44:06,600
to actually represent the
damage that's being done.

1054
00:44:06,600 --> 00:44:09,360
And that is the challenge facing
something like global warming,

1055
00:44:09,360 --> 00:44:12,450
but we have to take it on, OK?

1056
00:44:12,450 --> 00:44:14,142
So that's environmental
externalities.

1057
00:44:14,142 --> 00:44:15,600
The other big type
of externalities

1058
00:44:15,600 --> 00:44:22,190
are health externalities,
are health externalities.

1059
00:44:22,190 --> 00:44:26,660
I talked about smoking,
but, indeed, there

1060
00:44:26,660 --> 00:44:29,090
are huge externalities levied
by a bunch of activities

1061
00:44:29,090 --> 00:44:31,230
that we do that
impact our health.

1062
00:44:31,230 --> 00:44:35,680
So, for example,
drinking, drunk driving

1063
00:44:35,680 --> 00:44:40,730
causes 13,000 deaths per
year, 13,000 deaths per year,

1064
00:44:40,730 --> 00:44:45,160
OK, to put a sort of blunt face
on it, four 9/11s every year

1065
00:44:45,160 --> 00:44:49,750
from drunk driving plus
400,000 injuries every year

1066
00:44:49,750 --> 00:44:52,540
from people driving drunk.

1067
00:44:52,540 --> 00:44:57,910
Consuming gasoline, global
warming is a huge externality.

1068
00:44:57,910 --> 00:45:00,370
Perhaps one of the biggest
externalities facing us,

1069
00:45:00,370 --> 00:45:03,190
the biggest social
externality, is obesity, OK?

1070
00:45:03,190 --> 00:45:06,400
Obesity causes a
lot of illnesses

1071
00:45:06,400 --> 00:45:08,170
that cost a lot of money.

1072
00:45:08,170 --> 00:45:10,930
Projections are that children
born in the year 2000,

1073
00:45:10,930 --> 00:45:15,250
so about your kid, about your
generation, one third of them

1074
00:45:15,250 --> 00:45:18,400
will get diabetes before they
die based on current weight

1075
00:45:18,400 --> 00:45:19,260
projections.

1076
00:45:19,260 --> 00:45:21,400
Now, as you notice,
looking around this room,

1077
00:45:21,400 --> 00:45:24,900
that's not a problem of the
elite East Coast people, OK?

1078
00:45:24,900 --> 00:45:27,400
It's not a problem of-- it's a
problem of the less educated.

1079
00:45:27,400 --> 00:45:30,370
It's a problem concentrated
more in the South,

1080
00:45:30,370 --> 00:45:33,910
but, nonetheless, if you look
at a number of southern states,

1081
00:45:33,910 --> 00:45:36,970
the obesity rate is above 35%.

1082
00:45:36,970 --> 00:45:39,670
Literally, more than one in
three people in the state

1083
00:45:39,670 --> 00:45:41,530
are obese, OK?

1084
00:45:41,530 --> 00:45:44,590
This is a huge problem, and it's
going to cause huge social--

1085
00:45:44,590 --> 00:45:47,180
it's going to have huge social
consequences for our country.

1086
00:45:47,180 --> 00:45:48,710
The question is what
do we do about these.

1087
00:45:48,710 --> 00:45:50,877
What do we do about things
like smoking and drinking

1088
00:45:50,877 --> 00:45:52,000
and obesity?

1089
00:45:52,000 --> 00:45:55,560
And there's essentially--
there's, essentially,

1090
00:45:55,560 --> 00:45:57,160
four answers.

1091
00:45:57,160 --> 00:46:00,310
The first is information.

1092
00:46:00,310 --> 00:46:03,130
Can we just inform
people about the damages?

1093
00:46:03,130 --> 00:46:06,970
And, indeed, this has been
shown to work with smoking.

1094
00:46:06,970 --> 00:46:11,290
OK, we knew smoking was bad
for you in about 1954, OK?

1095
00:46:11,290 --> 00:46:13,450
But we only really
got through to people

1096
00:46:13,450 --> 00:46:16,610
starting really in
the 1970s and '80s,

1097
00:46:16,610 --> 00:46:18,010
but it's had an enormous effect.

1098
00:46:18,010 --> 00:46:19,940
Smoking has fallen
incredibly in the US

1099
00:46:19,940 --> 00:46:21,110
through that information.

1100
00:46:21,110 --> 00:46:23,230
But here's the
interesting thing, OK?

1101
00:46:23,230 --> 00:46:27,297
Smoking rates-- smoking used
to be 50% in the entire--

1102
00:46:27,297 --> 00:46:28,880
every adult, 50% of
all adults smoked.

1103
00:46:28,880 --> 00:46:31,360
It didn't matter race,
gender, class, whatever.

1104
00:46:31,360 --> 00:46:33,040
Now smoking is
essentially down to zero

1105
00:46:33,040 --> 00:46:35,050
among the well-educated
and still about 20%

1106
00:46:35,050 --> 00:46:36,760
against the less educated.

1107
00:46:36,760 --> 00:46:39,580
So information
works, but it works

1108
00:46:39,580 --> 00:46:42,950
in a very inequitable way.

1109
00:46:42,950 --> 00:46:45,200
OK, so information
is one solution.

1110
00:46:45,200 --> 00:46:47,830
The second solution is taxation.

1111
00:46:50,490 --> 00:46:53,310
And, indeed, this has been shown
to work for cigarettes once

1112
00:46:53,310 --> 00:46:55,190
again.

1113
00:46:55,190 --> 00:46:57,250
Smoking is price sensitive.

1114
00:46:57,250 --> 00:46:59,455
The elasticity of smoking
with respect to the price

1115
00:46:59,455 --> 00:47:01,600
is about minus 0.4.

1116
00:47:01,600 --> 00:47:04,300
About every 10% you raise
the price of cigarettes,

1117
00:47:04,300 --> 00:47:06,160
there's about 4% less smoking.

1118
00:47:06,160 --> 00:47:06,880
It works.

1119
00:47:06,880 --> 00:47:10,890
In particular, youth smoking
is very price sensitive.

1120
00:47:10,890 --> 00:47:13,610
Youths are very price sensitive
because youths have less money.

1121
00:47:13,610 --> 00:47:16,570
So they're very price sensitive.

1122
00:47:16,570 --> 00:47:18,940
So it actually works.

1123
00:47:18,940 --> 00:47:20,348
But that's sort
of the easy case.

1124
00:47:20,348 --> 00:47:22,390
Taxing cigarettes is easy
because every cigarette

1125
00:47:22,390 --> 00:47:23,320
is bad for you.

1126
00:47:23,320 --> 00:47:25,900
Taxing alcohol is trickier
because, after all,

1127
00:47:25,900 --> 00:47:28,780
most of the damage is done
by a tiny share of drinkers.

1128
00:47:28,780 --> 00:47:32,050
Most of us will consume alcohol
responsibly most of our lives

1129
00:47:32,050 --> 00:47:34,072
and not cause any
external damage, OK?

1130
00:47:34,072 --> 00:47:36,530
But most of the damage is done
by a tiny share of drinking.

1131
00:47:36,530 --> 00:47:37,650
So taxes is trickier.

1132
00:47:37,650 --> 00:47:39,400
If I proposed a big
rise in alcohol taxes,

1133
00:47:39,400 --> 00:47:40,690
people would say wait a second.

1134
00:47:40,690 --> 00:47:41,690
I'm responsible drinker.

1135
00:47:41,690 --> 00:47:42,653
Why are you taxing--

1136
00:47:42,653 --> 00:47:43,570
why are you taxing me?

1137
00:47:43,570 --> 00:47:44,903
So that's a little bit trickier.

1138
00:47:44,903 --> 00:47:46,480
Not to mention
obesity, taxing food

1139
00:47:46,480 --> 00:47:48,580
is maybe the
trickiest of all, OK?

1140
00:47:48,580 --> 00:47:50,850
So taxes are trickier.

1141
00:47:50,850 --> 00:47:52,910
You could maybe try--

1142
00:47:52,910 --> 00:47:55,320
an alternative thing you
could do is penalties.

1143
00:47:55,320 --> 00:47:57,750
So, instead of
taxing alcohol, we

1144
00:47:57,750 --> 00:48:00,327
could just steepen the
penalties for drunk driving.

1145
00:48:00,327 --> 00:48:02,160
You know, if you killed
a few drunk drivers,

1146
00:48:02,160 --> 00:48:04,620
there would be less
drunk driving, OK?

1147
00:48:04,620 --> 00:48:07,050
But the problem is that's
a pretty extreme penalty.

1148
00:48:07,050 --> 00:48:07,920
And what if you got it wrong?

1149
00:48:07,920 --> 00:48:09,360
You'd feel kind of bad
about killing someone

1150
00:48:09,360 --> 00:48:11,130
because the breathalyzer
didn't work.

1151
00:48:11,130 --> 00:48:12,180
There was a series of
articles, actually,

1152
00:48:12,180 --> 00:48:14,597
in The New York Times about
how terrible breathalyzers are

1153
00:48:14,597 --> 00:48:16,590
and how inaccurate they are, OK?

1154
00:48:16,590 --> 00:48:19,900
So problem with penalties is we
can't enforce them perfectly,

1155
00:48:19,900 --> 00:48:20,640
OK?

1156
00:48:20,640 --> 00:48:21,872
So that's the third solution.

1157
00:48:21,872 --> 00:48:24,330
The final solution and the one
that's really most discussed

1158
00:48:24,330 --> 00:48:28,050
right now is illegality.

1159
00:48:28,050 --> 00:48:29,800
What if we just made
these things illegal?

1160
00:48:29,800 --> 00:48:33,340
And this comes to the
discussion of marijuana.

1161
00:48:33,340 --> 00:48:36,380
Should marijuana be legal, OK?

1162
00:48:36,380 --> 00:48:42,350
Well, illegality is an extreme
form of lowering externalities.

1163
00:48:42,350 --> 00:48:43,850
Now, obviously,
when pot is illegal,

1164
00:48:43,850 --> 00:48:45,317
people still smoke pot.

1165
00:48:45,317 --> 00:48:47,150
But it's still true,
when you make it legal,

1166
00:48:47,150 --> 00:48:49,080
it's consumed at
much higher levels.

1167
00:48:49,080 --> 00:48:50,880
OK, legality does matter.

1168
00:48:50,880 --> 00:48:52,880
So, for example, people
have done studies.

1169
00:48:52,880 --> 00:48:55,197
Yes, people under 21
drink, but, literally,

1170
00:48:55,197 --> 00:48:57,530
if you look at people the day
after their 21st birthday,

1171
00:48:57,530 --> 00:48:59,450
they drink much more than
the day before their 21st.

1172
00:48:59,450 --> 00:49:01,242
Not on the 21st birthday,
that's the party.

1173
00:49:01,242 --> 00:49:01,867
We ignore that.

1174
00:49:01,867 --> 00:49:04,117
But, the day after their
21st birthday and thereafter,

1175
00:49:04,117 --> 00:49:06,350
they're drinking at much
higher levels than before.

1176
00:49:06,350 --> 00:49:08,303
Legality matters, OK?

1177
00:49:08,303 --> 00:49:10,220
It's also true, the day
after a 21st birthday,

1178
00:49:10,220 --> 00:49:12,428
people are much more likely
to die in a drunk driving

1179
00:49:12,428 --> 00:49:15,680
accident than the day before
their 21st birthday, OK?

1180
00:49:15,680 --> 00:49:17,280
So legality matters.

1181
00:49:17,280 --> 00:49:18,890
So the point is we
have a whole series

1182
00:49:18,890 --> 00:49:20,660
of tools to think about this.

1183
00:49:20,660 --> 00:49:23,490
And the question is how should
we combine and use them.

1184
00:49:23,490 --> 00:49:26,690
And the answer is
take 14.41, and I'll

1185
00:49:26,690 --> 00:49:27,950
teach you all about it.

1186
00:49:27,950 --> 00:49:29,510
But we don't any more time here.

1187
00:49:29,510 --> 00:49:32,390
This just raises the issues to
think about with externalities.

1188
00:49:32,390 --> 00:49:33,830
It's an important topic
to think about it.

1189
00:49:33,830 --> 00:49:35,760
I realize that's a lot
to cover in one lecture,

1190
00:49:35,760 --> 00:49:36,380
but I just wanted
to sort of give you

1191
00:49:36,380 --> 00:49:38,300
a taste for how
economists think about

1192
00:49:38,300 --> 00:49:41,050
and analyze this kind
of market failure.