WEBVTT

00:00:00.040 --> 00:00:02.460
The following content is
provided under a Creative

00:00:02.460 --> 00:00:03.870
Commons license.

00:00:03.870 --> 00:00:06.910
Your support will help MIT
OpenCourseWare continue to

00:00:06.910 --> 00:00:10.560
offer high quality educational
resources for free.

00:00:10.560 --> 00:00:13.460
To make a donation or view
additional materials from

00:00:13.460 --> 00:00:17.390
hundreds of MIT courses, visit
MIT OpenCourseWare at

00:00:17.390 --> 00:00:18.640
ocw.mit.edu.

00:00:22.460 --> 00:00:24.000
PROFESSOR: We're going to start
with an interesting

00:00:24.000 --> 00:00:32.040
application of demand curve
analysis, of the kind of

00:00:32.040 --> 00:00:34.320
indifference curve and
constrained choice analysis

00:00:34.320 --> 00:00:38.200
we've been doing, the
case of food stamps.

00:00:38.200 --> 00:00:40.100
And then we're going to move
on and talk about deriving

00:00:40.100 --> 00:00:41.510
demand curves.

00:00:41.510 --> 00:00:43.710
So let's talk about
food stamps.

00:00:43.710 --> 00:00:44.850
It's an interesting case.

00:00:44.850 --> 00:00:48.530
This is a policy that's been in
place in the US government

00:00:48.530 --> 00:00:53.440
for a very long time, which are
essentially coupons that

00:00:53.440 --> 00:00:55.430
individuals can use
to buy food.

00:00:55.430 --> 00:00:56.880
It used to literally
be coupons.

00:00:56.880 --> 00:00:59.800
It used to literally, you'd
get a stamp, a coupon, and

00:00:59.800 --> 00:01:01.510
you'd go to the supermarket and
hand this in instead of

00:01:01.510 --> 00:01:03.830
cash to buy your goods.

00:01:03.830 --> 00:01:05.660
Now it's actually
a debit card.

00:01:05.660 --> 00:01:08.830
And it's given to low-income
individuals as a way of

00:01:08.830 --> 00:01:10.020
redistributing income
in society.

00:01:10.020 --> 00:01:13.900
So individuals of income below
a certain level, typically,

00:01:13.900 --> 00:01:17.960
say, a family with income below
about $25,000 a year,

00:01:17.960 --> 00:01:24.760
below what we call the US
poverty line, will be eligible

00:01:24.760 --> 00:01:27.020
for food stamps, which is a
debit card they can use to

00:01:27.020 --> 00:01:29.670
charge their food purchases.

00:01:29.670 --> 00:01:33.160
Now, what I want to talk about
today is how we can use the

00:01:33.160 --> 00:01:36.470
kind of analysis we've done so
far to think about the effect

00:01:36.470 --> 00:01:38.010
of food stamps.

00:01:38.010 --> 00:01:40.830
So let's start with
figure 6-1.

00:01:40.830 --> 00:01:43.150
And let's think about someone
with an original budget line.

00:01:43.150 --> 00:01:46.280
If someone has a budget line,
they have income of $1,000.

00:01:46.280 --> 00:01:50.050
OK, a very poor person, they
have income of $1,000.

00:01:50.050 --> 00:01:52.450
And let's talk about them, their
choices between food and

00:01:52.450 --> 00:01:53.270
all other goods.

00:01:53.270 --> 00:01:54.390
Once again we have
this analysis.

00:01:54.390 --> 00:01:55.930
We put everything in
two dimensions.

00:01:55.930 --> 00:01:59.230
What we think about is via
mental accounting, as we

00:01:59.230 --> 00:02:00.790
talked about last time.

00:02:00.790 --> 00:02:03.010
However you want to justify it
to yourself, the way we model

00:02:03.010 --> 00:02:04.900
it is thinking about people
having this choice along two

00:02:04.900 --> 00:02:07.290
dimensions, food and other
goods, and they

00:02:07.290 --> 00:02:09.800
have income of $1000.

00:02:09.800 --> 00:02:10.949
Let's say we'll have
two individuals,

00:02:10.949 --> 00:02:12.460
individual x and y.

00:02:12.460 --> 00:02:16.570
Individual x doesn't care
that much about food.

00:02:16.570 --> 00:02:19.010
They really like consuming
other goods.

00:02:19.010 --> 00:02:21.470
So they spend $600 of their
income on other

00:02:21.470 --> 00:02:23.560
goods and $400 on food.

00:02:23.560 --> 00:02:26.015
Individual y cares
a lot about food.

00:02:26.015 --> 00:02:29.300
They end up spending $900
of their budget on food.

00:02:29.300 --> 00:02:31.690
And that should be 100.

00:02:31.690 --> 00:02:32.910
100 on other goods.

00:02:32.910 --> 00:02:37.590
So that y-intercept should
be 100 for person y.

00:02:37.590 --> 00:02:40.780
So they spend 900 on food
and 100 on other goods.

00:02:40.780 --> 00:02:43.820
Now let's say the government
comes in and wants to consider

00:02:43.820 --> 00:02:45.130
two options.

00:02:45.130 --> 00:02:47.800
The government decides, look,
we want to help poor people.

00:02:47.800 --> 00:02:50.560
And particularly we want to give
people like these people

00:02:50.560 --> 00:02:53.890
$500 in resources.

00:02:53.890 --> 00:02:55.180
Let's think of two different
ways the

00:02:55.180 --> 00:02:56.850
government could do that.

00:02:56.850 --> 00:02:58.060
One way the government
could do it is it

00:02:58.060 --> 00:02:59.530
could give them cash.

00:02:59.530 --> 00:03:00.960
It could say, look, we're
going to send

00:03:00.960 --> 00:03:03.500
you a check for $500.

00:03:03.500 --> 00:03:06.070
What would that do to their
budget constraint?

00:03:06.070 --> 00:03:08.450
Well that would shift it out, as
we talked about last time.

00:03:08.450 --> 00:03:11.440
It would be an outward shift
of $500 at every point.

00:03:11.440 --> 00:03:13.460
So their new budget constraint
would be the line that runs

00:03:13.460 --> 00:03:17.470
from $1,500 of all other goods
to $1,500 dollars in food.

00:03:17.470 --> 00:03:18.570
So include that.

00:03:18.570 --> 00:03:21.570
So both the solid and dashed
portions would be their new

00:03:21.570 --> 00:03:23.360
budget constraint.

00:03:23.360 --> 00:03:26.830
So the new budget constraint
would run from 1,500 to 1,500.

00:03:26.830 --> 00:03:32.450
Person x would choose to
move from x1 to x2.

00:03:32.450 --> 00:03:34.800
I'm sorry, they would choose
to move from x1.

00:03:34.800 --> 00:03:36.250
It's not labeled as a point.

00:03:36.250 --> 00:03:39.710
But they would actually choose
to move from indifference

00:03:39.710 --> 00:03:42.020
curve 2 to indifference
curve 3.

00:03:42.020 --> 00:03:44.200
OK, that's where they'd choose
to move if they could choose

00:03:44.200 --> 00:03:45.720
along this entire line.

00:03:45.720 --> 00:03:47.990
Let me sort of, well, I'm not
going to draw it, because I'll

00:03:47.990 --> 00:03:48.730
do a bad job.

00:03:48.730 --> 00:03:50.730
But basically, if you can think
about the new budget

00:03:50.730 --> 00:03:54.530
constraint running from 1,500 to
1,500, person x would move

00:03:54.530 --> 00:03:58.140
from indifference curve 2
to indifference curve 3.

00:03:58.140 --> 00:03:59.390
They would choose--

00:04:01.640 --> 00:04:02.390
Actually, you know what?

00:04:02.390 --> 00:04:03.700
Let's put this graph aside.

00:04:03.700 --> 00:04:04.970
It's not quite right along
the number of dimensions.

00:04:04.970 --> 00:04:07.170
I'm going to draw this, because
there's a number of

00:04:07.170 --> 00:04:08.420
problems with that.

00:04:11.260 --> 00:04:19.029
So you've got an original budget
line that runs from

00:04:19.029 --> 00:04:23.730
1,000 to 1,000.

00:04:23.730 --> 00:04:28.110
And we've got person x up here,
and they're choosing to

00:04:28.110 --> 00:04:31.725
spend 400 on food and
600 on other goods.

00:04:31.725 --> 00:04:33.590
And we've got person
y down here.

00:04:33.590 --> 00:04:35.890
This is person x, and they're
in section x1.

00:04:42.000 --> 00:04:47.150
And person y intercepts at y1
where they choose to spend 900

00:04:47.150 --> 00:04:50.600
on food and 100 on
other goods.

00:04:50.600 --> 00:04:52.350
Now, the government first says,
look, we're going to

00:04:52.350 --> 00:04:54.440
give people $500 in cash.

00:04:54.440 --> 00:04:58.050
That just shifts the budget
constraint out parallel, but

00:04:58.050 --> 00:05:02.360
now runs from 1,500 to 1,500.

00:05:02.360 --> 00:05:04.920
Let's say that the choices
people make--

00:05:04.920 --> 00:05:06.853
Person x would say, great, I'm
going to take that money, I'm

00:05:06.853 --> 00:05:09.920
going to spend almost all
of it on other goods.

00:05:09.920 --> 00:05:14.880
So I'm going to move to a point
like x2, where I'm going

00:05:14.880 --> 00:05:20.480
to consume $1,200 on--

00:05:20.480 --> 00:05:22.730
I'm going to consume--

00:05:22.730 --> 00:05:23.210
They were consuming--

00:05:23.210 --> 00:05:23.910
No, I'm sorry.

00:05:23.910 --> 00:05:28.440
15, right.

00:05:28.440 --> 00:05:32.360
They're going to move from
spending $600 on other goods

00:05:32.360 --> 00:05:35.940
and $400 on food to spending
$1,100 of their dollars on

00:05:35.940 --> 00:05:38.224
other goods, and--

00:05:38.224 --> 00:05:39.940
let me think for a second.

00:05:39.940 --> 00:05:41.380
Let's say they would
go vertically.

00:05:41.380 --> 00:05:44.380
So let's say they'd choose
to spend all

00:05:44.380 --> 00:05:46.180
of it on other goods.

00:05:46.180 --> 00:05:51.990
So they'd take the whole 500,
and they'd go from spending--

00:05:51.990 --> 00:05:54.460
they'd continue to spend 400 on
food, but now they'd spend

00:05:54.460 --> 00:05:56.220
1,100 on other goods.

00:05:56.220 --> 00:05:58.270
So person x, they would
continue to

00:05:58.270 --> 00:05:59.800
spend $400 on food.

00:05:59.800 --> 00:06:02.420
They'd say I'm going to take
that entire $500, instead of

00:06:02.420 --> 00:06:04.860
spending $600 on other goods,
I'm going to spend $1,100 on

00:06:04.860 --> 00:06:06.400
other goods.

00:06:06.400 --> 00:06:09.384
Let's say person y, they would
say, well I'm going to

00:06:09.384 --> 00:06:10.030
sort of split it.

00:06:10.030 --> 00:06:11.430
I'm going to spend
some on food.

00:06:11.430 --> 00:06:14.200
So this is their new
intercept x2.

00:06:14.200 --> 00:06:16.180
I'm going to spend some on food,
and I'm going to spend

00:06:16.180 --> 00:06:16.870
some on other goods.

00:06:16.870 --> 00:06:19.000
So I'm going to go out here.

00:06:19.000 --> 00:06:20.240
I'm going to spend now--

00:06:20.240 --> 00:06:22.030
instead of spending
$900 on food, I'll

00:06:22.030 --> 00:06:25.590
spend $1,200 on food.

00:06:25.590 --> 00:06:27.050
I'll take $300 and
spend it on food.

00:06:27.050 --> 00:06:29.935
Instead of spending $100 on
other goods, I'll spend $300

00:06:29.935 --> 00:06:31.460
on other goods.

00:06:31.460 --> 00:06:33.660
OK, so that's person y2.

00:06:33.660 --> 00:06:37.780
OK, so that's what they'd do if
we gave them $500 in cash.

00:06:37.780 --> 00:06:39.360
Now say the government came in
and said, you know, we're

00:06:39.360 --> 00:06:42.056
going to give you $500 but in
the form of a coupon that you

00:06:42.056 --> 00:06:43.880
can spend on food.

00:06:43.880 --> 00:06:45.710
So the first question is, what
does that do to the budget

00:06:45.710 --> 00:06:47.130
constraint?

00:06:47.130 --> 00:06:47.880
This is a bit tricky.

00:06:47.880 --> 00:06:49.280
We've got to think about this.

00:06:49.280 --> 00:06:50.300
Think about their budget
constraint.

00:06:50.300 --> 00:06:57.820
What that says is for anyone who
wants to continue to, at

00:06:57.820 --> 00:07:01.390
least spend $1,000 on--

00:07:01.390 --> 00:07:06.460
anyone who wants to at least
spend $500 on food, it does

00:07:06.460 --> 00:07:08.810
not change their opportunities
at all.

00:07:08.810 --> 00:07:11.610
So the new budget constraint
looks like this.

00:07:11.610 --> 00:07:14.270
It's a solid line to here,
and then it goes down.

00:07:14.270 --> 00:07:16.710
This intersects at 500.

00:07:16.710 --> 00:07:21.100
It's a solid line to the point
at $500, and then it goes down

00:07:21.100 --> 00:07:22.390
and follows the old
budget constraint.

00:07:22.390 --> 00:07:23.640
Why does it do that?

00:07:23.640 --> 00:07:28.750
Because for anyone to this side,
they are $500 richer

00:07:28.750 --> 00:07:31.280
regardless of whether you
give them cash or food.

00:07:31.280 --> 00:07:32.290
Either way they're
$500 richer.

00:07:32.290 --> 00:07:32.950
Why?

00:07:32.950 --> 00:07:35.720
Because as long as you intended
to spend $500 on food

00:07:35.720 --> 00:07:38.370
anyway, it doesn't matter the
form in which the government

00:07:38.370 --> 00:07:39.420
gives you the money.

00:07:39.420 --> 00:07:40.215
Think about that for a second.

00:07:40.215 --> 00:07:41.890
It's very important.

00:07:41.890 --> 00:07:44.950
If you were going to spend $500
on food anyway, it does

00:07:44.950 --> 00:07:47.840
not matter if the government
gives you a check for 500, or

00:07:47.840 --> 00:07:50.350
a food card for 500.

00:07:50.350 --> 00:07:51.930
Why is that?

00:07:51.930 --> 00:07:55.170
That's because your budget
is what we call fungible.

00:07:55.170 --> 00:07:58.120
You can always move money around
within your budget.

00:07:58.120 --> 00:08:01.700
So let's say you're spending
500 on cash.

00:08:01.700 --> 00:08:02.950
Let's take a person like y.

00:08:02.950 --> 00:08:04.030
Let's take a person like y.

00:08:04.030 --> 00:08:08.890
They were spending 900 on food
and 100 on other goods.

00:08:08.890 --> 00:08:10.090
The government comes
in and gives them a

00:08:10.090 --> 00:08:12.640
food card for 500.

00:08:12.640 --> 00:08:15.045
Well to them that's the same
as $500 in cash because

00:08:15.045 --> 00:08:17.150
they're already spending
$900 on food.

00:08:17.150 --> 00:08:19.670
They can just take some of the
cash on food, spend it on

00:08:19.670 --> 00:08:22.460
other things, and use the card
for the food instead.

00:08:22.460 --> 00:08:25.060
So for them, there's no
difference in giving them cash

00:08:25.060 --> 00:08:27.120
or giving them the food card.

00:08:27.120 --> 00:08:28.180
It has the same effect.

00:08:28.180 --> 00:08:30.130
They move to y2 either way.

00:08:30.130 --> 00:08:31.610
Now let's take a
person like x.

00:08:31.610 --> 00:08:34.826
Well, if you gave them cash,
$500 cash, they'd spend none

00:08:34.826 --> 00:08:35.890
of it on food.

00:08:35.890 --> 00:08:39.200
So they've been consuming
$1,100 of other

00:08:39.200 --> 00:08:40.400
goods and $400 of food.

00:08:40.400 --> 00:08:42.760
But they can't do that if you
give them a food card, right,

00:08:42.760 --> 00:08:44.250
if you give them food stamps.

00:08:44.250 --> 00:08:45.710
That's not a choice.

00:08:45.710 --> 00:08:52.630
They are now constrained to
move to a point like x3.

00:08:52.630 --> 00:08:55.580
They're now constrained to move
to a point like x3 where

00:08:55.580 --> 00:08:58.770
they have to go to this
intersection.

00:08:58.770 --> 00:09:01.630
Because this point is
not attainable.

00:09:01.630 --> 00:09:02.710
This point is not attainable.

00:09:02.710 --> 00:09:03.910
If you give them food stamps,
the new budget

00:09:03.910 --> 00:09:04.940
constraint is this.

00:09:04.940 --> 00:09:06.770
So x2 is no longer attainable.

00:09:06.770 --> 00:09:10.960
They have to move to
a point like x3.

00:09:10.960 --> 00:09:15.330
So what do we know about their
level of happiness

00:09:15.330 --> 00:09:16.800
at x3 versus x2?

00:09:19.940 --> 00:09:21.801
Someone raise their
hand and tell me.

00:09:21.801 --> 00:09:22.643
Yeah.

00:09:22.643 --> 00:09:23.910
AUDIENCE: It's lower.

00:09:23.910 --> 00:09:24.475
PROFESSOR: It's what?

00:09:24.475 --> 00:09:25.320
AUDIENCE: Lower.

00:09:25.320 --> 00:09:25.910
PROFESSOR: It's lower.

00:09:25.910 --> 00:09:26.900
And how do you know that?

00:09:26.900 --> 00:09:30.161
AUDIENCE: Because it's kind of
like, if it was on the full

00:09:30.161 --> 00:09:32.420
curve, they would
be elsewhere.

00:09:32.420 --> 00:09:33.900
PROFESSOR: Well OK, so there's
two different ways to see it.

00:09:33.900 --> 00:09:35.090
One is you could say, look
they're at a lower

00:09:35.090 --> 00:09:35.900
indifference curve.

00:09:35.900 --> 00:09:37.360
You can see what's wrong
with this graph.

00:09:37.360 --> 00:09:38.540
The indifference curves cross.

00:09:38.540 --> 00:09:39.570
The indifference curves
can never cross.

00:09:39.570 --> 00:09:41.360
So that's wrong there.

00:09:41.360 --> 00:09:43.640
They're on a lower indifference
curve, OK.

00:09:43.640 --> 00:09:45.763
But what's the other way
to think about it?

00:09:45.763 --> 00:09:47.535
AUDIENCE: The marginal rate
of substitution and

00:09:47.535 --> 00:09:48.870
transformation aren't
the same.

00:09:48.870 --> 00:09:50.130
PROFESSOR: The marginal rate
of substitution and

00:09:50.130 --> 00:09:51.260
transformation aren't
the same.

00:09:51.260 --> 00:09:52.200
That's another way
to think about.

00:09:52.200 --> 00:09:54.320
And that's always true
at the optimum.

00:09:54.320 --> 00:09:57.240
But what else do we know about
a point like this?

00:09:57.240 --> 00:10:00.680
They could have chosen that
point before and didn't.

00:10:00.680 --> 00:10:01.030
Right?

00:10:01.030 --> 00:10:04.920
When they had the cash, they
had the option of choosing

00:10:04.920 --> 00:10:07.210
this point, but they didn't,
they chose a different point.

00:10:07.210 --> 00:10:15.450
So we know by something called
revealed preference that

00:10:15.450 --> 00:10:16.130
they're worse off.

00:10:16.130 --> 00:10:18.000
This is a very important
concept.

00:10:18.000 --> 00:10:25.280
If someone makes a choice that
they turned down before, then

00:10:25.280 --> 00:10:27.900
by revealed preference they're
less well off.

00:10:27.900 --> 00:10:31.400
We've revealed that they're
worse, because they could have

00:10:31.400 --> 00:10:32.890
chosen this point before,
but they didn't.

00:10:32.890 --> 00:10:35.300
When you gave them the cash,
they chose this point.

00:10:35.300 --> 00:10:38.310
So by revealed preference we
revealed they're worse off.

00:10:38.310 --> 00:10:39.370
So it's the same
as saying their

00:10:39.370 --> 00:10:41.210
indifference curve is lower.

00:10:41.210 --> 00:10:44.120
We've revealed they're
worse off.

00:10:44.120 --> 00:10:48.810
So what we've learned is for
person y, they don't care if

00:10:48.810 --> 00:10:52.390
you give them cash or a food
card, food stamps.

00:10:52.390 --> 00:10:55.090
It's called food stamps, but
it's now a debit card.

00:10:55.090 --> 00:11:00.140
Person x is made worse off if
you give them the food stamps

00:11:00.140 --> 00:11:02.980
instead of the cash.

00:11:02.980 --> 00:11:04.690
Why do it?

00:11:04.690 --> 00:11:05.950
You're the government.

00:11:05.950 --> 00:11:10.760
The US government spends $35
billion every year giving

00:11:10.760 --> 00:11:12.970
people food stamps
instead of cash.

00:11:12.970 --> 00:11:15.810
Why don't we just take that
$35 billion and give it to

00:11:15.810 --> 00:11:18.530
them in cash?

00:11:18.530 --> 00:11:18.960
Yeah?

00:11:18.960 --> 00:11:20.864
AUDIENCE: Probably because the
government doesn't trust

00:11:20.864 --> 00:11:22.768
people to spend it on what
they actually need.

00:11:22.768 --> 00:11:26.100
And that will just lead to
more poverty and people

00:11:26.100 --> 00:11:28.020
wasting it on things
they don't need.

00:11:28.020 --> 00:11:30.050
PROFESSOR: Or to put it more
succinctly, what if this axis

00:11:30.050 --> 00:11:31.990
was not labeled other goods
but labeled cocaine.

00:11:34.560 --> 00:11:37.080
Then we might be sad that you
took the whole $500 and spent

00:11:37.080 --> 00:11:38.230
it on cocaine.

00:11:38.230 --> 00:11:42.400
We might want you to take that
$500 and spend it on food.

00:11:42.400 --> 00:11:43.830
So it's paternalism.

00:11:43.830 --> 00:11:46.610
The reason we give the guys food
stamps instead of cash is

00:11:46.610 --> 00:11:48.750
we don't trust them
with the cash.

00:11:48.750 --> 00:11:52.180
If we trusted people with the
cash, there'd be no reason not

00:11:52.180 --> 00:11:53.110
to give them the cash.

00:11:53.110 --> 00:11:57.740
We are unambiguously making them
worse off by forcing them

00:11:57.740 --> 00:12:00.370
to consume a bundle that's on
a lower utility curve, lower

00:12:00.370 --> 00:12:01.220
indifference curve.

00:12:01.220 --> 00:12:03.900
But since we don't trust them,
since we're paternalistic, we

00:12:03.900 --> 00:12:06.830
are willing to go ahead and
force them to do that.

00:12:06.830 --> 00:12:10.050
So then the interesting question
becomes, well, how

00:12:10.050 --> 00:12:11.570
much are we costing them?

00:12:11.570 --> 00:12:12.510
In fact, it's not obvious.

00:12:12.510 --> 00:12:15.880
If everyone in the world looks
like y, then there's no cost

00:12:15.880 --> 00:12:16.560
to food stamps.

00:12:16.560 --> 00:12:19.270
There's no good done either.

00:12:19.270 --> 00:12:21.570
Then it doesn't matter if we
give them cash or food stamps.

00:12:21.570 --> 00:12:25.490
But if a lot of people look
like x, then there is a

00:12:25.490 --> 00:12:27.360
welfare cost to people.

00:12:27.360 --> 00:12:30.430
They are worse off, from
their own perspective,

00:12:30.430 --> 00:12:31.360
getting food stamps.

00:12:31.360 --> 00:12:33.400
Society may think they're
better off.

00:12:33.400 --> 00:12:36.000
So how do we tell?

00:12:36.000 --> 00:12:36.890
How do we tell?

00:12:36.890 --> 00:12:38.330
Can anyone take a guess?

00:12:38.330 --> 00:12:40.320
If you're now an empirical
economist, and you want to

00:12:40.320 --> 00:12:45.000
test, how would you tell if
people are like x or like y?

00:12:45.000 --> 00:12:45.610
Any ideas?

00:12:45.610 --> 00:12:48.180
It's tricky, but let's see if we
have any budding empirical

00:12:48.180 --> 00:12:48.840
economists here.

00:12:48.840 --> 00:12:49.190
Yeah?

00:12:49.190 --> 00:12:51.965
AUDIENCE: You'd see if they're
spending any cash beyond the

00:12:51.965 --> 00:12:54.875
$500 that you gave them, because
then you'd basically

00:12:54.875 --> 00:12:55.850
[UNINTELLIGIBLE].

00:12:55.850 --> 00:12:57.665
PROFESSOR: See if they're buying
food beyond the $500

00:12:57.665 --> 00:12:58.350
you gave them.

00:12:58.350 --> 00:12:59.840
AUDIENCE: They might spend
$100 cash on food.

00:12:59.840 --> 00:13:00.320
PROFESSOR: Excellent.

00:13:00.320 --> 00:13:01.290
So that's one way you'd do it.

00:13:01.290 --> 00:13:03.640
You could look at people who
get food stamps and see if

00:13:03.640 --> 00:13:04.540
they're spending more.

00:13:04.540 --> 00:13:05.550
That's a great idea.

00:13:05.550 --> 00:13:09.300
The other thing you could
do is you could actually

00:13:09.300 --> 00:13:12.920
literally run an experiment
where you take people who are

00:13:12.920 --> 00:13:15.390
getting food stamps and replace
them with cash or vice

00:13:15.390 --> 00:13:17.910
versa and see what happens
to their behavior.

00:13:17.910 --> 00:13:21.540
When we do this, we find
that about 15% of

00:13:21.540 --> 00:13:22.980
people are like x.

00:13:22.980 --> 00:13:24.700
Or in other words, the
way to say it, is

00:13:24.700 --> 00:13:27.940
about $0.15 more precisely.

00:13:27.940 --> 00:13:30.820
When you give people food stamps
instead of cash, they

00:13:30.820 --> 00:13:33.940
spend 15% more on food than they
would if you just gave

00:13:33.940 --> 00:13:36.280
them the cash.

00:13:36.280 --> 00:13:43.020
So there's about 15% lower
utility compared to what

00:13:43.020 --> 00:13:47.500
they'd want for spending it on
the food instead of the cash.

00:13:47.500 --> 00:13:53.120
So the question is,
is it worth it?

00:13:53.120 --> 00:13:55.620
We're basically taking people
and making them spend $0.15

00:13:55.620 --> 00:13:57.800
more on food than
they'd want to.

00:13:57.800 --> 00:13:58.880
That's the right way
to think about it.

00:13:58.880 --> 00:14:00.870
If you give them food stamps
instead of cash, they spend

00:14:00.870 --> 00:14:02.680
$0.15 more on food than
they would if you

00:14:02.680 --> 00:14:04.240
just gave them cash.

00:14:04.240 --> 00:14:05.390
Is it worth it?

00:14:05.390 --> 00:14:06.400
That's a great question.

00:14:06.400 --> 00:14:08.730
It depends on how stupid we
think people are and how

00:14:08.730 --> 00:14:10.730
paternalistic we want to be.

00:14:10.730 --> 00:14:12.750
If we think people would really
waste the money, then

00:14:12.750 --> 00:14:15.210
$0.15 is not much to give up
to make sure they eat.

00:14:15.210 --> 00:14:17.160
If we think nobody would waste
the money, then we're just

00:14:17.160 --> 00:14:20.140
throwing $0.15 down the toilet
by making them buy food

00:14:20.140 --> 00:14:21.640
instead of goods they prefer.

00:14:21.640 --> 00:14:23.990
And that's the interesting kind
of public policy question

00:14:23.990 --> 00:14:25.260
we have to struggle with.

00:14:25.260 --> 00:14:28.320
We think about government policy
and redistribution.

00:14:28.320 --> 00:14:29.500
That's exactly the kind
of question we

00:14:29.500 --> 00:14:30.030
need to struggle with.

00:14:30.030 --> 00:14:31.940
And we'll come back to that
again later in the course when

00:14:31.940 --> 00:14:34.490
we talk about efficiency
versus equity.

00:14:34.490 --> 00:14:37.470
OK, questions about that?

00:14:37.470 --> 00:14:38.376
Yeah?

00:14:38.376 --> 00:14:39.626
AUDIENCE: [INAUDIBLE PHRASE].

00:14:47.610 --> 00:14:48.210
PROFESSOR: Sure.

00:14:48.210 --> 00:14:49.930
I mean, so basically you
make a good point.

00:14:49.930 --> 00:14:51.990
We sort of like to know.

00:14:51.990 --> 00:14:53.030
Actually that's a
very good point.

00:14:53.030 --> 00:14:55.240
You say, when we run these
experiments and replace the

00:14:55.240 --> 00:14:56.710
food stamps with cash,
we like to know what

00:14:56.710 --> 00:14:57.500
they spend the cash.

00:14:57.500 --> 00:15:00.330
We want to know not just what
happens to food consumption.

00:15:00.330 --> 00:15:02.830
So if you run the experiment,
and you say, I was giving you

00:15:02.830 --> 00:15:03.880
food stamps.

00:15:03.880 --> 00:15:05.800
I now cash you out and
give you cash.

00:15:05.800 --> 00:15:07.985
And I find you spend
15% less on food.

00:15:07.985 --> 00:15:10.010
Well what do you
spend more on?

00:15:10.010 --> 00:15:11.865
If it's clothes, maybe
we're not so worried.

00:15:11.865 --> 00:15:13.750
If it's cocaine, maybe we are.

00:15:13.750 --> 00:15:14.630
So that's a very good point.

00:15:14.630 --> 00:15:15.770
That's something we
could look at.

00:15:15.770 --> 00:15:18.460
Excellent.

00:15:18.460 --> 00:15:21.590
OK, so that's an example of
how we can use the kind of

00:15:21.590 --> 00:15:24.550
analysis we did last time to
think about policy making.

00:15:24.550 --> 00:15:25.510
Once again, this is
an incredibly

00:15:25.510 --> 00:15:27.270
simple framework, right?

00:15:27.270 --> 00:15:31.600
Yet I just described to you a
succinct way to think about

00:15:31.600 --> 00:15:32.920
the implications for society of

00:15:32.920 --> 00:15:33.960
different government policies.

00:15:33.960 --> 00:15:36.630
That's the power of this kind
of simplified framework.

00:15:36.630 --> 00:15:39.860
Now let's move on, and let's
get to the core of

00:15:39.860 --> 00:15:41.240
why we did all this.

00:15:41.240 --> 00:15:43.530
The reason we did all this is we
wanted to figure out how we

00:15:43.530 --> 00:15:47.560
come up with demand curves,
where demand curves come from.

00:15:47.560 --> 00:15:49.530
The stork doesn't bring them.

00:15:49.530 --> 00:15:51.690
Demand curves come from
underlying utility

00:15:51.690 --> 00:15:52.070
maximization.

00:15:52.070 --> 00:15:55.020
And we'll see that now.

00:15:55.020 --> 00:15:58.230
And basically the way to do
this is to return to our

00:15:58.230 --> 00:15:59.930
example from last time.

00:15:59.930 --> 00:16:05.880
Your parents gave you $96.

00:16:05.880 --> 00:16:10.860
You could buy movies at $8 a
pop or pizzas at $16 a pop.

00:16:10.860 --> 00:16:12.540
So we said last time,
if you turn to the

00:16:12.540 --> 00:16:15.320
next page of the handout.

00:16:15.320 --> 00:16:21.040
What we said last time is if
given your utility function, u

00:16:21.040 --> 00:16:24.560
equals square root of p times
m, you would choose

00:16:24.560 --> 00:16:26.960
a point like a.

00:16:26.960 --> 00:16:31.340
If the price of pizzas was $16,
the price of movies was

00:16:31.340 --> 00:16:37.070
$8, your income was $96, you
would choose a point like a,

00:16:37.070 --> 00:16:39.820
where you consumed--

00:16:39.820 --> 00:16:44.570
At point a, you're consuming six
movies and three pizzas.

00:16:44.570 --> 00:16:51.670
Once again that should
be p on the y-axis.

00:16:51.670 --> 00:16:58.510
You're consuming six movies and
three pizzas at point a.

00:16:58.510 --> 00:17:06.024
Now let's say the price
of pizzas rises.

00:17:09.869 --> 00:17:16.880
I'm sorry, now let's say the
price of movies rises to $12.

00:17:16.880 --> 00:17:21.874
So the price of movies
rises from $8 to $12.

00:17:21.874 --> 00:17:24.619
Well what does that do to
the budget constraint?

00:17:24.619 --> 00:17:28.290
That steepens the budget
constraint, moves it inward.

00:17:28.290 --> 00:17:31.530
Because now think about
your opportunity set.

00:17:31.530 --> 00:17:33.910
For the same income of $96 you
can buy the same number of

00:17:33.910 --> 00:17:38.100
pizzas you could have before,
but now you're

00:17:38.100 --> 00:17:41.260
buying fewer movies.

00:17:41.260 --> 00:17:42.690
Same number of pizzas you could
have bought before, but

00:17:42.690 --> 00:17:44.690
now you can buy fewer movies.

00:17:44.690 --> 00:17:46.890
So your new budget constraint,
you have a new constrained

00:17:46.890 --> 00:17:47.700
opportunity set.

00:17:47.700 --> 00:17:49.430
With a steeper budget
constraint, the slope, instead

00:17:49.430 --> 00:17:52.830
of being minus 1/2,
is minus 3/4.

00:17:52.830 --> 00:17:57.290
And given the preference I
wrote, u equals square root of

00:17:57.290 --> 00:18:01.500
c times m, you should be able
to show yourself that you'd

00:18:01.500 --> 00:18:07.870
now choose a point like b, where
you have three pizzas

00:18:07.870 --> 00:18:10.670
but now only four movies.

00:18:10.670 --> 00:18:12.310
So you reduce the number of
movies, you keep the number of

00:18:12.310 --> 00:18:14.030
pizzas constant.

00:18:14.030 --> 00:18:16.950
And you check that we still
spent our total budget.

00:18:16.950 --> 00:18:22.750
Well, 4 times 12 plus 3
times 16 is still $96.

00:18:22.750 --> 00:18:25.060
So we're still spending
our total budget.

00:18:25.060 --> 00:18:27.960
The marginal rate of
substitution you can compute

00:18:27.960 --> 00:18:30.130
if you write it down from that
utility function, will be

00:18:30.130 --> 00:18:32.600
minus 3/4, which is the same
as the marginal rate of

00:18:32.600 --> 00:18:34.890
transformation with
this new price.

00:18:34.890 --> 00:18:39.800
So you will choose a
point like point b.

00:18:39.800 --> 00:18:43.720
Now let's say instead
the price of movies

00:18:43.720 --> 00:18:47.230
fell from $8 to $6.

00:18:47.230 --> 00:18:50.470
Well in that case, your budget
constraint would flatten.

00:18:50.470 --> 00:18:53.210
It would move outwards.

00:18:53.210 --> 00:18:57.040
Your opportunity set would
expand in that case, because

00:18:57.040 --> 00:18:59.330
effectively you're richer.

00:18:59.330 --> 00:19:00.280
Your opportunity set expands.

00:19:00.280 --> 00:19:02.860
You move to bc little 3.

00:19:02.860 --> 00:19:04.740
You move to bc little 3.

00:19:04.740 --> 00:19:07.356
And given those preferences I
wrote down, u equals square

00:19:07.356 --> 00:19:10.870
root of p times m.

00:19:10.870 --> 00:19:12.710
u equals square root
of p times m.

00:19:12.710 --> 00:19:17.220
You end up choosing point c,
with the same three pizzas but

00:19:17.220 --> 00:19:20.170
now eight movies.

00:19:20.170 --> 00:19:21.420
Once again, how do we
know that's right?

00:19:21.420 --> 00:19:23.450
Well first of all the marginal
rate of substitution, you can

00:19:23.450 --> 00:19:25.350
compute, will equal the
new marginal rate of

00:19:25.350 --> 00:19:26.600
transformation.

00:19:28.610 --> 00:19:32.650
And also you can see you spend
your entire $96 income.

00:19:32.650 --> 00:19:34.925
You're still roughly splitting
it with $48 on movies and $48

00:19:34.925 --> 00:19:37.630
on pizza, exactly
splitting it.

00:19:37.630 --> 00:19:41.110
So all we've done here--

00:19:41.110 --> 00:19:42.250
Forget the bottom diagram
for a second.

00:19:42.250 --> 00:19:45.830
All we're doing in this top
diagram is saying, given your

00:19:45.830 --> 00:19:55.800
utility is u equals square root
of p times m and given

00:19:55.800 --> 00:19:59.380
your income and the prices,
these are the choices you

00:19:59.380 --> 00:20:01.290
would make as prices change.

00:20:01.290 --> 00:20:03.960
Are there questions
about that?

00:20:03.960 --> 00:20:06.665
Now armed with that, we can
draw a demand curve.

00:20:06.665 --> 00:20:07.910
Because what have we done?

00:20:07.910 --> 00:20:12.190
We've just given you three
different prices for movies

00:20:12.190 --> 00:20:14.740
and three different quantities
of movies you choose.

00:20:14.740 --> 00:20:18.830
We know when the price
of movies was $8,

00:20:18.830 --> 00:20:20.760
you chose six movies.

00:20:20.760 --> 00:20:22.010
That's point b.

00:20:26.195 --> 00:20:27.445
I'm sorry, that's point a.

00:20:31.460 --> 00:20:32.960
The points are mislabeled
too on this.

00:20:32.960 --> 00:20:34.540
I'm sorry.

00:20:34.540 --> 00:20:36.380
If you go to this bottom
graph, these points are

00:20:36.380 --> 00:20:37.450
mislabeled.

00:20:37.450 --> 00:20:40.845
So it should go b, a, c.

00:20:40.845 --> 00:20:42.095
It should go b, a, c.

00:20:46.680 --> 00:20:50.670
So when the price of movies is
$8, that's point a in the

00:20:50.670 --> 00:20:55.160
middle, you choose 6 movies.

00:20:55.160 --> 00:20:59.260
When the price of movies
rises to $12, your

00:20:59.260 --> 00:21:00.410
demand for movies falls.

00:21:00.410 --> 00:21:02.550
You only choose 4 movies.

00:21:02.550 --> 00:21:07.890
When the price of movies falls
to $6, you choose 8 movies.

00:21:07.890 --> 00:21:09.610
Thus the demand curve.

00:21:09.610 --> 00:21:10.310
And we're done.

00:21:10.310 --> 00:21:13.080
That's where demand
curves come from.

00:21:13.080 --> 00:21:15.040
They just come from utility,
constrained utility

00:21:15.040 --> 00:21:16.440
maximization.

00:21:16.440 --> 00:21:20.500
You just take your utility
function, you maximize it,

00:21:20.500 --> 00:21:22.970
given the constraint the budget
constraint places on

00:21:22.970 --> 00:21:24.970
you, and boom, you have
a demand curve.

00:21:27.740 --> 00:21:31.920
Now note that this a particular
case we did.

00:21:31.920 --> 00:21:34.000
And it's a particular case
that's interesting.

00:21:34.000 --> 00:21:37.260
In this case as we change the
price of movies, what happened

00:21:37.260 --> 00:21:39.890
to demand for pizzas?

00:21:39.890 --> 00:21:40.780
AUDIENCE: Stayed the same.

00:21:40.780 --> 00:21:41.800
PROFESSOR: Stayed the same.

00:21:41.800 --> 00:21:43.910
That is a particular case.

00:21:43.910 --> 00:21:45.510
It's basically the case that
will happen with utility

00:21:45.510 --> 00:21:46.690
function of this form.

00:21:46.690 --> 00:21:51.190
It's a case of what we call
no cross-price elasticity.

00:21:51.190 --> 00:21:58.040
This example has no cross-price
elasticity.

00:21:58.040 --> 00:22:01.600
What that means is that in this
particular case we chose,

00:22:01.600 --> 00:22:03.840
as the price of one good
changes, it does not change

00:22:03.840 --> 00:22:05.520
your demand for the
other good.

00:22:05.520 --> 00:22:08.040
That's a special case that will
not in general be true.

00:22:08.040 --> 00:22:12.190
You can imagine if your income
was only $96, and the price of

00:22:12.190 --> 00:22:14.720
movies was swinging around, that
might affect your taste

00:22:14.720 --> 00:22:15.710
for pizzas.

00:22:15.710 --> 00:22:17.880
That might affect your demand
for pizzas as well, because

00:22:17.880 --> 00:22:19.160
you're only have
a fixed budget.

00:22:19.160 --> 00:22:20.220
That's a more general case.

00:22:20.220 --> 00:22:23.070
We've chosen a particular case
here with no cross-price

00:22:23.070 --> 00:22:23.810
elasticity.

00:22:23.810 --> 00:22:24.910
But don't think that's
general.

00:22:24.910 --> 00:22:26.690
This is not a general lesson.

00:22:26.690 --> 00:22:28.390
There's the price of one good
changes the other goods

00:22:28.390 --> 00:22:29.240
unaffected.

00:22:29.240 --> 00:22:31.590
In fact, in general, both goods
will be affected when

00:22:31.590 --> 00:22:33.470
any one price changes.

00:22:33.470 --> 00:22:37.290
That's a more general result.

00:22:37.290 --> 00:22:39.740
You should be able to check at
home that you can do this

00:22:39.740 --> 00:22:43.530
exact same exercise for pizzas
and draw the demand curve for

00:22:43.530 --> 00:22:45.660
pizzas the exact same way.

00:22:45.660 --> 00:22:47.190
You'll still get a flat--
you'll still get no

00:22:47.190 --> 00:22:47.690
cross-price elasticity.

00:22:47.690 --> 00:22:52.320
You'll still get this flat
curve-- well now it will be

00:22:52.320 --> 00:22:57.190
vertical curve with respect
to movie purchases.

00:22:57.190 --> 00:22:59.890
But you can see as you change
the price of pizzas, you'll

00:22:59.890 --> 00:23:04.020
find a well-defined pizza
demand curve as well.

00:23:04.020 --> 00:23:05.350
OK?

00:23:05.350 --> 00:23:07.850
So that's where demand
curves come from.

00:23:07.850 --> 00:23:13.860
We basically maximize utility at
different prices given your

00:23:13.860 --> 00:23:18.650
income, and we end up with a
demand curve that shows us the

00:23:18.650 --> 00:23:21.170
relationship between how many
movies you choose and the

00:23:21.170 --> 00:23:22.420
price of movies.

00:23:24.930 --> 00:23:28.690
Demand curves themselves
can also shift.

00:23:28.690 --> 00:23:32.530
We talked about that in
the second lecture.

00:23:32.530 --> 00:23:34.370
We talked about demand
curve shifting.

00:23:34.370 --> 00:23:36.710
And one reason demand
curves can shift is

00:23:36.710 --> 00:23:39.120
because you get richer.

00:23:39.120 --> 00:23:41.290
So let's talk about how
that can happen.

00:23:41.290 --> 00:23:43.410
Let's now turn to figure 6-3,
which is really tiny.

00:23:46.690 --> 00:23:47.930
My bifocals are in
at the mall.

00:23:47.930 --> 00:23:48.960
I just haven't picked
them up yet.

00:23:48.960 --> 00:23:50.680
So let's take the glasses
off for this one.

00:23:54.400 --> 00:23:55.490
Now let's take a case.

00:23:55.490 --> 00:24:01.380
Once again, originally you're
at point a, where you're

00:24:01.380 --> 00:24:04.940
choosing six movies
and three pizzas.

00:24:04.940 --> 00:24:07.530
Now let's say your
income rises.

00:24:07.530 --> 00:24:09.350
Your parents are feeling
generous.

00:24:09.350 --> 00:24:11.740
And instead of giving
you $96, they're

00:24:11.740 --> 00:24:16.700
going to give you $128.

00:24:16.700 --> 00:24:19.890
Once again, on that y-axis it
should be labeled p, not c.

00:24:19.890 --> 00:24:26.230
You can now afford up to 8
movies and up to 16 pizzas

00:24:26.230 --> 00:24:28.070
with your $128 income.

00:24:28.070 --> 00:24:29.490
So your budget constraint
has shifted

00:24:29.490 --> 00:24:33.750
outwards from bc1 to bc2.

00:24:33.750 --> 00:24:36.170
At that new higher budget
constraint, you're going to

00:24:36.170 --> 00:24:40.200
choose, instead of choosing a,
which is six movies and three

00:24:40.200 --> 00:24:43.260
pizzas, you'll choose
b, which is eight

00:24:43.260 --> 00:24:44.770
movies and four pizzas.

00:24:44.770 --> 00:24:47.530
You're richer so you choose
more of both.

00:24:47.530 --> 00:24:53.360
Likewise if your parents cut
your income to $64, you're

00:24:53.360 --> 00:24:54.690
budget constraint will
shift inwards.

00:24:54.690 --> 00:24:56.270
Your opportunity set will
be constricted.

00:24:56.270 --> 00:24:58.060
You move to budget
constraint three.

00:24:58.060 --> 00:25:01.460
And you choose fewer of both
pizzas and movies.

00:25:01.460 --> 00:25:04.440
So you can see as your budget
constraint shifts, how you

00:25:04.440 --> 00:25:06.960
choose different amounts of
both pizza and movies.

00:25:06.960 --> 00:25:09.600
We can translate that
to shifting

00:25:09.600 --> 00:25:11.670
demand curves for movies.

00:25:11.670 --> 00:25:13.910
So if you draw that down to
the next diagram, you say,

00:25:13.910 --> 00:25:18.310
look, I can now graph that at
a given price of movies--

00:25:18.310 --> 00:25:19.650
prices have not changed
in the example.

00:25:19.650 --> 00:25:21.450
The slope of the budget
constraint is the same.

00:25:21.450 --> 00:25:24.180
Only your income has changed.

00:25:24.180 --> 00:25:29.645
At a given price of movies of
$8, as my income changes, I am

00:25:29.645 --> 00:25:31.090
on different demand curves.

00:25:31.090 --> 00:25:34.640
You can see the demand curve for
movies shifting out and in

00:25:34.640 --> 00:25:35.960
as my income changes.

00:25:35.960 --> 00:25:37.810
So as my income went up, the
demand curve for movies went

00:25:37.810 --> 00:25:40.010
out and moved from point
a to point b.

00:25:40.010 --> 00:25:43.820
As my income fell, the demand
curve for movies went up,

00:25:43.820 --> 00:25:46.480
shifted in, moved from
point a to point c.

00:25:46.480 --> 00:25:49.530
We can then drop that down one
more level, just to make life

00:25:49.530 --> 00:25:54.330
especially interesting, and draw
the relationship between

00:25:54.330 --> 00:25:56.940
your income and your
demand for movies.

00:25:56.940 --> 00:25:58.570
And that's the third figure.

00:25:58.570 --> 00:26:01.455
Here we graph the relationship
of your income and your demand

00:26:01.455 --> 00:26:01.740
for movies.

00:26:01.740 --> 00:26:03.770
This is not a demand curve.

00:26:03.770 --> 00:26:06.300
Demand curves only relate
price to quantity.

00:26:06.300 --> 00:26:10.360
This is what we call
an Engel curve.

00:26:10.360 --> 00:26:13.150
Those of you who studied your
socialism theory will remember

00:26:13.150 --> 00:26:15.290
Engels worked with Marx.

00:26:15.290 --> 00:26:16.240
It's not him.

00:26:16.240 --> 00:26:17.450
It's Engel not Engels.

00:26:17.450 --> 00:26:19.100
Different guy.

00:26:19.100 --> 00:26:24.140
So basically this is
an Engel curve.

00:26:24.140 --> 00:26:26.760
And basically it shows the
relationship between your

00:26:26.760 --> 00:26:30.180
income and your demand
for a good.

00:26:30.180 --> 00:26:33.810
And this turns out to be a
very important concept.

00:26:33.810 --> 00:26:36.870
Because an important thing that
we'll focus on now is

00:26:36.870 --> 00:26:45.500
what we call the income
elasticity of demand.

00:26:45.500 --> 00:26:48.700
We've talked about price
elasticities.

00:26:48.700 --> 00:26:51.030
What's the price elasticity
of demand?

00:26:51.030 --> 00:26:52.795
Someone quickly, someone raise
their hand and tell me, what's

00:26:52.795 --> 00:26:55.392
the price elasticity
of demand?

00:26:55.392 --> 00:26:57.080
Get some other folks
involved here.

00:26:57.080 --> 00:26:57.560
Yeah?

00:26:57.560 --> 00:26:59.960
AUDIENCE: How demand changes
with the price of the item?

00:26:59.960 --> 00:27:02.070
PROFESSOR: Right, so as price
changes, how demand changes.

00:27:02.070 --> 00:27:03.880
The income elasticity
is the same concept.

00:27:03.880 --> 00:27:08.530
It's a change in demand as
your income changes.

00:27:08.530 --> 00:27:09.990
In the book it's this
fancy letter.

00:27:09.990 --> 00:27:12.670
I can't write, so I'm going
to call it gamma.

00:27:12.670 --> 00:27:15.730
But it's some c thing that
I can't draw in the book.

00:27:15.730 --> 00:27:23.470
Which is delta Q over Q
over delta y over y.

00:27:23.470 --> 00:27:26.000
So just like the price
elasticity is the percentage

00:27:26.000 --> 00:27:28.590
change of quantity, percentage
change in price, the income

00:27:28.590 --> 00:27:32.660
elasticity is the percent
change in quantity with

00:27:32.660 --> 00:27:33.910
percent change in income.

00:27:37.270 --> 00:27:41.820
Once again, just like price
elasticities are locals.

00:27:41.820 --> 00:27:42.870
You talked about
it in section.

00:27:42.870 --> 00:27:45.140
You talked about, sort of,
local versus global price

00:27:45.140 --> 00:27:47.030
elasticities and how it's really
local to that segment

00:27:47.030 --> 00:27:47.750
of the curve.

00:27:47.750 --> 00:27:49.690
Income elasticities
are local too.

00:27:49.690 --> 00:27:52.190
Your income elasticity will
obviously change along an

00:27:52.190 --> 00:27:56.520
Engel, could change along
an Engel curve.

00:27:56.520 --> 00:28:00.660
But the key point is that for
most goods the Engel curve is

00:28:00.660 --> 00:28:02.300
upward sloping.

00:28:02.300 --> 00:28:06.240
That is for most goods, this
is greater than zero.

00:28:06.240 --> 00:28:08.550
Just like we talked about the
price elasticity being less

00:28:08.550 --> 00:28:13.530
than zero in general, this is
less general, but for most

00:28:13.530 --> 00:28:15.450
goods, the income elasticity
is greater than zero.

00:28:15.450 --> 00:28:17.830
We call these normal goods.

00:28:21.290 --> 00:28:23.800
Normal goods are goods for which
the income elasticity is

00:28:23.800 --> 00:28:24.250
greater than zero.

00:28:24.250 --> 00:28:27.560
As you have more income,
you buy more of them.

00:28:27.560 --> 00:28:32.250
On the other hand, if the income
elasticity was less

00:28:32.250 --> 00:28:36.340
than zero, we would call
those inferior goods.

00:28:40.750 --> 00:28:46.550
Inferior goods, goods where as
your income goes up, you buy

00:28:46.550 --> 00:28:47.800
less of them.

00:28:49.800 --> 00:28:50.080
Yeah.

00:28:50.080 --> 00:28:53.300
AUDIENCE: Is there any term
for when income elasticity

00:28:53.300 --> 00:28:55.085
equals zero?

00:28:55.085 --> 00:28:57.840
PROFESSOR: If it equals zero,
you're just income inelastic.

00:28:57.840 --> 00:28:59.140
It's in between normal
and inferior.

00:28:59.140 --> 00:29:00.210
I don't think there's
a precise term.

00:29:00.210 --> 00:29:02.570
You're just income inelastic.

00:29:02.570 --> 00:29:05.870
So can someone tell me how you
could get an inferior good?

00:29:05.870 --> 00:29:07.236
Does anyone have a good
idea of an example

00:29:07.236 --> 00:29:09.130
of an inferior good?

00:29:09.130 --> 00:29:10.556
How could a good be inferior?

00:29:10.556 --> 00:29:11.042
Yeah.

00:29:11.042 --> 00:29:15.416
AUDIENCE: Canned food, because
if you have a low income, you

00:29:15.416 --> 00:29:17.360
buy canned food because
it's cheaper.

00:29:17.360 --> 00:29:19.320
But [INAUDIBLE].

00:29:19.320 --> 00:29:21.710
PROFESSOR: Exactly, so canned
food versus fresh food.

00:29:21.710 --> 00:29:25.790
As your income goes up, you'll
substitute away from canned

00:29:25.790 --> 00:29:26.580
food to fresh food.

00:29:26.580 --> 00:29:27.920
So actually as you get
richer, you'll

00:29:27.920 --> 00:29:29.490
consume less canned food.

00:29:29.490 --> 00:29:31.690
So canned food is an
inferior good.

00:29:31.690 --> 00:29:32.200
Excellent.

00:29:32.200 --> 00:29:34.140
The classic example
uses potatoes.

00:29:34.140 --> 00:29:37.190
Potatoes is a good
cost-effective, cheap source

00:29:37.190 --> 00:29:39.190
of nutrition.

00:29:39.190 --> 00:29:40.920
But, you know, no one wants to
eat potatoes all the time if

00:29:40.920 --> 00:29:41.460
they don't have to.

00:29:41.460 --> 00:29:43.710
So when the income goes up, we
substitute away from potatoes

00:29:43.710 --> 00:29:46.820
towards arugula or whatever.

00:29:46.820 --> 00:29:50.810
So basically, essentially
we could think of

00:29:50.810 --> 00:29:53.120
inferior goods as goods--

00:29:53.120 --> 00:29:54.980
Once again, more is
always better.

00:29:54.980 --> 00:29:57.620
There's no goods
we don't like.

00:29:57.620 --> 00:29:58.740
More is always better.

00:29:58.740 --> 00:30:01.280
But there are goods we'd like
to substitute away from.

00:30:01.280 --> 00:30:03.060
We'd like to have
others instead.

00:30:03.060 --> 00:30:05.850
And goods you substitute away
from as you get richer are

00:30:05.850 --> 00:30:07.230
inferior goods.

00:30:07.230 --> 00:30:14.020
Goods you move towards as you
get richer are normal goods.

00:30:14.020 --> 00:30:16.270
Moreover, we can break
this down further.

00:30:16.270 --> 00:30:20.480
Within the class of normal goods
we can talk about gamma

00:30:20.480 --> 00:30:24.230
less than one and gamma
greater than one.

00:30:24.230 --> 00:30:26.370
Any guess as to what terms
I'll use for gamma?

00:30:26.370 --> 00:30:31.040
Any examples in, sort of, the
class of goods where it would

00:30:31.040 --> 00:30:32.530
be less than one versus
greater than one?

00:30:37.090 --> 00:30:38.970
What's an example of a good that
would be less than one?

00:30:38.970 --> 00:30:41.302
Think about what that means.

00:30:41.302 --> 00:30:42.294
Yeah.

00:30:42.294 --> 00:30:45.270
AUDIENCE: Perhaps food,
because if your income

00:30:45.270 --> 00:30:46.520
[INAUDIBLE PHRASE].

00:30:49.750 --> 00:30:50.120
PROFESSOR: Right.

00:30:50.120 --> 00:30:52.426
AUDIENCE: [INAUDIBLE]

00:30:52.426 --> 00:30:54.850
if your income increases.

00:30:54.850 --> 00:30:55.480
PROFESSOR: Excellent.

00:30:55.480 --> 00:30:57.566
So we call these necessities.

00:31:00.850 --> 00:31:04.120
And we call these luxuries.

00:31:04.120 --> 00:31:06.000
Goods where the income
elasticity is less than one

00:31:06.000 --> 00:31:06.610
are necessities.

00:31:06.610 --> 00:31:10.400
You want more of them
as you get richer.

00:31:10.400 --> 00:31:13.730
But you don't want as much
more as you get richer.

00:31:13.730 --> 00:31:16.630
So you've got the food as
the classic example.

00:31:16.630 --> 00:31:20.490
As your income doubles, you're
going to eat more food but not

00:31:20.490 --> 00:31:23.300
twice as much food.

00:31:23.300 --> 00:31:26.380
Likewise luxuries are things
where as your income doubles,

00:31:26.380 --> 00:31:28.170
you'll buy more than
twice as much.

00:31:28.170 --> 00:31:31.650
So you think about fancy cars.

00:31:31.650 --> 00:31:34.710
You might buy one with your
first million but three with

00:31:34.710 --> 00:31:36.950
your next million.

00:31:36.950 --> 00:31:39.830
So those are luxuries.

00:31:39.830 --> 00:31:41.487
They'll increase more than
proportionally as

00:31:41.487 --> 00:31:42.770
your income goes up.

00:31:42.770 --> 00:31:44.850
Necessities will increase less
than proportionally as your

00:31:44.850 --> 00:31:45.940
income goes up.

00:31:45.940 --> 00:31:48.070
Now, of course, it's a very
hard distinction to draw.

00:31:48.070 --> 00:31:49.800
And of course it varies
by person.

00:31:49.800 --> 00:31:51.470
So take clothing.

00:31:51.470 --> 00:31:53.800
Is clothing a necessity
or a luxury?

00:31:53.800 --> 00:31:55.990
Well, some clothing is probably
a necessity, and some

00:31:55.990 --> 00:31:57.375
clothing is probably a luxury.

00:31:57.375 --> 00:31:59.745
You know, Dolce and Gabbana
is a luxury.

00:31:59.745 --> 00:32:01.270
You know, Keds is a necessity.

00:32:01.270 --> 00:32:05.410
Or whatever, I don't know, what,
The Gap is a necessity.

00:32:05.410 --> 00:32:08.490
So basically we could think
about, it's actually a subtle

00:32:08.490 --> 00:32:10.890
distinction what makes
luxury and what's--

00:32:10.890 --> 00:32:13.700
Normal versus inferior
is kind of stark.

00:32:13.700 --> 00:32:17.270
Luxury versus necessity, that's
a little bit trickier.

00:32:17.270 --> 00:32:18.860
That's going to depend on
the person and depend

00:32:18.860 --> 00:32:20.090
on the type of goods.

00:32:20.090 --> 00:32:22.980
But it's important to understand
that concept.

00:32:22.980 --> 00:32:25.980
OK, questions about that.

00:32:25.980 --> 00:32:26.200
Yeah.

00:32:26.200 --> 00:32:27.646
AUDIENCE: I have a question
in general.

00:32:27.646 --> 00:32:30.538
Is there a way to relate income
elasticity to, like,

00:32:30.538 --> 00:32:31.788
own-price elasticity?

00:32:34.010 --> 00:32:35.400
PROFESSOR: Actually that's
a great question.

00:32:35.400 --> 00:32:37.530
That's a great segue to what
we're going to do next.

00:32:37.530 --> 00:32:40.060
It's actually a fundamental
determinant of own-price

00:32:40.060 --> 00:32:42.660
elasticity, and we'll
talk about why next.

00:32:42.660 --> 00:32:44.310
Other questions about
this concept?

00:32:44.310 --> 00:32:45.210
Yeah.

00:32:45.210 --> 00:32:47.730
AUDIENCE: [INAUDIBLE PHRASE]?

00:32:47.730 --> 00:32:50.240
PROFESSOR: Yes, income
elasticity is, once again,

00:32:50.240 --> 00:32:53.660
just like price elasticities are
not necessarily constant.

00:32:53.660 --> 00:32:56.180
You could have a constant income
elasticity curve or a

00:32:56.180 --> 00:32:57.490
non-constant income
elasticity curve.

00:32:57.490 --> 00:32:58.820
It can absolutely change.

00:32:58.820 --> 00:33:00.440
In general it will.

00:33:00.440 --> 00:33:03.730
But it might not.

00:33:03.730 --> 00:33:04.630
Good questions.

00:33:04.630 --> 00:33:07.040
But that other question in the
back, gee, how does this

00:33:07.040 --> 00:33:09.510
relate to own-price
elasticity?

00:33:09.510 --> 00:33:10.190
Great segue.

00:33:10.190 --> 00:33:13.310
In fact the income effect is
going to be one of two key

00:33:13.310 --> 00:33:16.210
determinants of own-price
elasticity.

00:33:16.210 --> 00:33:17.880
Now what we're going to do is
we're going to go even further

00:33:17.880 --> 00:33:21.110
behind the demand curve.

00:33:21.110 --> 00:33:23.190
We talked about the demand
curve comes from.

00:33:23.190 --> 00:33:25.650
We talked before about how the
slope of the demand curve is

00:33:25.650 --> 00:33:27.660
the price elasticity.

00:33:27.660 --> 00:33:29.370
Remember, the price elasticity
was the slope.

00:33:29.370 --> 00:33:30.790
Now we're going to talk
about where does

00:33:30.790 --> 00:33:32.350
the slope come from.

00:33:32.350 --> 00:33:33.330
Where do price elasticities
come from?

00:33:33.330 --> 00:33:34.940
So I've shown you where the
demand curve comes from.

00:33:34.940 --> 00:33:37.110
Now let's talk about what
determines the underlying

00:33:37.110 --> 00:33:38.230
slope of the demand curve.

00:33:38.230 --> 00:33:40.280
What determines the
price elasticity.

00:33:40.280 --> 00:33:43.240
And what's going to determine
it is two different effects

00:33:43.240 --> 00:33:47.500
which work generally together
but sometimes in opposition.

00:33:47.500 --> 00:33:52.320
There are two effects that
determine price elasticity.

00:33:54.890 --> 00:33:59.040
The first is the substitution
effect.

00:34:05.720 --> 00:34:16.900
The Substitution Effect is the
change in quantity demanded

00:34:16.900 --> 00:34:21.840
when price increases holding
utility constant.

00:34:21.840 --> 00:34:27.469
So the Substitution Effect
is delta p--

00:34:27.469 --> 00:34:30.110
it's in percentage terms-- delta
p over p, over delta q

00:34:30.110 --> 00:34:34.840
over q, holding utility
constant at a

00:34:34.840 --> 00:34:37.139
fixed level u bar.

00:34:37.139 --> 00:34:43.179
So given that your utility has
not changed, how does your

00:34:43.179 --> 00:34:44.429
demand for the good shift?

00:34:47.639 --> 00:34:49.260
We're now getting
kind of deep.

00:34:49.260 --> 00:34:52.940
Think of this as, as a good gets
relatively expensive, how

00:34:52.940 --> 00:34:56.580
do you shift away
from that good?

00:34:56.580 --> 00:34:57.710
Think of this as the
shift away from the

00:34:57.710 --> 00:34:59.420
good as it gets expensive.

00:34:59.420 --> 00:35:02.090
But at the same time, there's a
second effect which we just

00:35:02.090 --> 00:35:06.950
introduced, which is
the Income Effect.

00:35:06.950 --> 00:35:12.380
The Income Effect is the
complement of the Substitution

00:35:12.380 --> 00:35:16.840
Effect, which is the change in
quantity demanded because of a

00:35:16.840 --> 00:35:19.910
change in income holding
prices constant.

00:35:23.060 --> 00:35:26.700
So this is the Income Effect
we just introduced.

00:35:26.700 --> 00:35:33.700
So it's delta Q over Q
over delta y over y.

00:35:33.700 --> 00:35:35.985
But this is holding
prices constant.

00:35:39.530 --> 00:35:43.360
And these two put together
determine your

00:35:43.360 --> 00:35:45.650
own elasticity demand.

00:35:45.650 --> 00:35:46.950
Think of it intuitively.

00:35:46.950 --> 00:35:48.530
We'll do it intuitively,
and graphically, and

00:35:48.530 --> 00:35:49.580
mathematically.

00:35:49.580 --> 00:35:53.490
Intuitively, it's when a price
goes up, two things happen.

00:35:53.490 --> 00:35:56.840
On the one hand, you're like,
gee, at that different price

00:35:56.840 --> 00:35:58.980
ratio, I might want
to substitute

00:35:58.980 --> 00:36:00.800
my consumption bundle.

00:36:00.800 --> 00:36:02.520
The second is, gee, the
price just went up,

00:36:02.520 --> 00:36:04.190
I'm effectively poorer.

00:36:04.190 --> 00:36:07.040
And that's also going
to affect my demand.

00:36:07.040 --> 00:36:09.860
So to see that let's
go to the graphical

00:36:09.860 --> 00:36:11.410
analysis and figure 6-4.

00:36:16.710 --> 00:36:19.245
And we're going to actually
decompose income and

00:36:19.245 --> 00:36:20.180
substitution effects.

00:36:20.180 --> 00:36:22.140
This is one of the things that's
sort of hard to do it

00:36:22.140 --> 00:36:22.380
intuitively.

00:36:22.380 --> 00:36:25.480
The graphics kind of makes
it the most clear.

00:36:31.660 --> 00:36:35.420
We're going to start
at a point like a.

00:36:35.420 --> 00:36:37.940
Point a is our initial
equilibrium at our budget

00:36:37.940 --> 00:36:40.420
constraint one, where
we're choosing six

00:36:40.420 --> 00:36:41.650
movies and three pizzas.

00:36:41.650 --> 00:36:44.350
Once again, this is
pizzas not CDs.

00:36:44.350 --> 00:36:49.570
We're choosing six movies
and three pizzas.

00:36:49.570 --> 00:36:51.570
That's at point a.

00:36:51.570 --> 00:36:56.560
Now we're going to say imagine
the price of movies

00:36:56.560 --> 00:36:59.050
has risen to $12.

00:36:59.050 --> 00:37:02.560
Well we know that ultimately
you'll end up at

00:37:02.560 --> 00:37:03.920
a point like c.

00:37:03.920 --> 00:37:05.320
We demonstrated that before.

00:37:05.320 --> 00:37:08.100
That was when we derived the
demand curve for movies.

00:37:08.100 --> 00:37:11.880
We know that if the price of
movies rises from $8 to $12,

00:37:11.880 --> 00:37:14.100
your demand for movies
will shrink from six

00:37:14.100 --> 00:37:15.280
movies to four movies.

00:37:15.280 --> 00:37:17.910
We already established that.

00:37:17.910 --> 00:37:20.240
But now what we can see is
that that's actually a

00:37:20.240 --> 00:37:23.000
composition of two effects.

00:37:23.000 --> 00:37:25.500
The first effect is the
Substitution Effect.

00:37:25.500 --> 00:37:30.500
And the Substitution Effect is
the change in prices holding

00:37:30.500 --> 00:37:31.320
utility constant.

00:37:31.320 --> 00:37:33.140
How do we hold utility
constant?

00:37:33.140 --> 00:37:36.320
You have to be on the same
indifference curve.

00:37:36.320 --> 00:37:39.430
So the way to measure the
Substitution Effect is we

00:37:39.430 --> 00:37:43.090
effectively draw an imaginary
budget constraint.

00:37:43.090 --> 00:37:46.050
We say, look, imagine you're
on the same indifference

00:37:46.050 --> 00:37:50.100
curve, but prices changed.

00:37:50.100 --> 00:37:53.940
So we draw budget constraint--

00:37:53.940 --> 00:37:55.800
you're originally on budget
constraint one.

00:38:00.240 --> 00:38:04.180
Now we draw a new budget
constraint,

00:38:04.180 --> 00:38:07.380
budget constraint three.

00:38:07.380 --> 00:38:09.590
Budget constraint three, it's
sort of hard to see.

00:38:09.590 --> 00:38:14.050
Budget constraint three is drawn
so that it has the new

00:38:14.050 --> 00:38:15.460
price ratio.

00:38:15.460 --> 00:38:19.720
That is, it's parallel to the
final budget constraint bc2,

00:38:19.720 --> 00:38:24.730
but it's tangent to the original
indifference curve.

00:38:24.730 --> 00:38:27.150
This is hard, so let me just
walk this through again.

00:38:27.150 --> 00:38:30.070
You've got your original budget
constraint, bc 1.

00:38:30.070 --> 00:38:32.130
You chose your point a.

00:38:32.130 --> 00:38:34.980
Now the price of movies has just
increased, so you move to

00:38:34.980 --> 00:38:38.190
bc2 in reality.

00:38:38.190 --> 00:38:41.320
And at bc2, you choose
point c.

00:38:41.320 --> 00:38:45.870
But for the Substitution Effect
we're going to say,

00:38:45.870 --> 00:38:50.220
let's hold utility constant
and ask what package you'd

00:38:50.220 --> 00:38:52.370
choose at these new price
holding utility constant?

00:38:52.370 --> 00:38:54.780
Well the way to do that is to
draw an imaginary bc3--

00:38:54.780 --> 00:38:57.190
bc3 never existed in reality--

00:38:57.190 --> 00:39:00.530
that is parallel to the new
budget constraint, that is the

00:39:00.530 --> 00:39:04.320
new set of prices, but it's
tangent to your old

00:39:04.320 --> 00:39:08.490
indifference curve, that
is utility is constant.

00:39:08.490 --> 00:39:11.010
If that were the case,
you'd choose point b.

00:39:11.010 --> 00:39:14.240
You'd choose 4.89 movies.

00:39:14.240 --> 00:39:19.750
Therefore we say that the
Substitution Effect is 1.11.

00:39:19.750 --> 00:39:22.700
We reduce your movies by 1.11
through the Substitution

00:39:22.700 --> 00:39:24.270
Effect only.

00:39:24.270 --> 00:39:29.800
The price change only, holding
your income, holding utility

00:39:29.800 --> 00:39:33.360
constant, is 1.11.

00:39:33.360 --> 00:39:35.800
Then we say, well, what's
the Income Effect?

00:39:35.800 --> 00:39:39.830
The income effect is given
prices are fixed, what's the

00:39:39.830 --> 00:39:43.620
effect of just being poorer,
because movies

00:39:43.620 --> 00:39:45.010
are now more expensive.

00:39:45.010 --> 00:39:46.740
We know how to make
someone poorer, we

00:39:46.740 --> 00:39:47.990
just lower their income.

00:39:47.990 --> 00:39:52.940
So we shift from bc3 to bc2 and
from point b to point c,

00:39:52.940 --> 00:39:54.590
and we're done.

00:39:54.590 --> 00:39:56.390
We get the total effect
of the price.

00:39:56.390 --> 00:39:59.270
So we have to shift from a to
c is all that you see in the

00:39:59.270 --> 00:40:01.110
real world.

00:40:01.110 --> 00:40:04.450
But behind that is the
Substitution Effect which

00:40:04.450 --> 00:40:08.780
shifts you from a to b and the
Income Effect, which is this

00:40:08.780 --> 00:40:11.300
parallel shift inwards of the
budget constraint, which

00:40:11.300 --> 00:40:12.550
shifts you from b to c.

00:40:16.120 --> 00:40:21.490
I'm going to go through that
again in a minute.

00:40:21.490 --> 00:40:23.620
But let me first answer these
questions, then I want to show

00:40:23.620 --> 00:40:24.630
you some of the math of this.

00:40:24.630 --> 00:40:26.850
Are there questions about
what's going on here?

00:40:26.850 --> 00:40:27.370
Yeah.

00:40:27.370 --> 00:40:30.100
AUDIENCE: The income effect can
have either sign, right?

00:40:30.100 --> 00:40:30.380
PROFESSOR: The income effect
can have either sign.

00:40:30.380 --> 00:40:31.340
Excellent point.

00:40:31.340 --> 00:40:32.800
What have I illustrated here?

00:40:32.800 --> 00:40:36.150
What kind of good is this?

00:40:36.150 --> 00:40:37.220
This is a normal good.

00:40:37.220 --> 00:40:38.280
Excellent point.

00:40:38.280 --> 00:40:39.960
This is a normal good.

00:40:39.960 --> 00:40:42.700
I've assumed a normal good.

00:40:42.700 --> 00:40:44.750
And this should actually say
in the title income and

00:40:44.750 --> 00:40:46.550
substitution effects
for normal good.

00:40:46.550 --> 00:40:48.480
So I've assumed a normal good.

00:40:48.480 --> 00:40:51.010
And we know this student aptly
pointed out the normal good,

00:40:51.010 --> 00:40:55.100
because we know that as your
income fell, moving from point

00:40:55.100 --> 00:40:58.870
b to c, you consumed
less of it.

00:40:58.870 --> 00:41:02.350
This comes to the question that
was in the back, which is

00:41:02.350 --> 00:41:05.720
as the price changes, whether
we consume more or less is

00:41:05.720 --> 00:41:07.730
related to the Income Effect but
doesn't tell you precisely

00:41:07.730 --> 00:41:09.070
what the Income Effect is.

00:41:09.070 --> 00:41:11.440
But as your income changes, if
you consume more or less,

00:41:11.440 --> 00:41:12.360
that's directly Income Effect.

00:41:12.360 --> 00:41:13.330
So we see it's a normal good.

00:41:13.330 --> 00:41:16.895
So the Substitution Effect
moves us from a to b.

00:41:16.895 --> 00:41:18.480
The Income Effect,
from b to c.

00:41:18.480 --> 00:41:21.480
What I want to do is one more
thing before we stop.

00:41:21.480 --> 00:41:23.600
And then I'm going to come back
next time and go back

00:41:23.600 --> 00:41:26.160
over this and then talk about
some applications.

00:41:26.160 --> 00:41:28.620
What I want to do right
now is I want to--

00:41:28.620 --> 00:41:32.750
The Income Effect, say the
following, the sign is

00:41:32.750 --> 00:41:37.600
ambiguous, because it depends
on if it's a normal or an

00:41:37.600 --> 00:41:38.900
inferior good.

00:41:38.900 --> 00:41:42.930
The Substitution Effect
is unambiguous.

00:41:42.930 --> 00:41:47.430
Substitution Effects are
always negative.

00:41:47.430 --> 00:41:52.910
Holding utility constant, a
price increase of a good

00:41:52.910 --> 00:41:55.060
always shifts you away
from that good.

00:41:55.060 --> 00:41:57.550
Negative or less than or equal
to zero could have no effect.

00:41:57.550 --> 00:41:59.890
But once again, I always talk
in inequalities even though

00:41:59.890 --> 00:42:01.240
they're typically inexact.

00:42:04.480 --> 00:42:08.160
The key is the Substitution
Effect is always negative.

00:42:08.160 --> 00:42:10.890
We can think about this in two
different ways depending on

00:42:10.890 --> 00:42:12.620
how you want to think
about it.

00:42:12.620 --> 00:42:15.420
Graphically we could
show this by the

00:42:15.420 --> 00:42:17.880
fact that if the price--

00:42:17.880 --> 00:42:19.320
think about it graphically.

00:42:19.320 --> 00:42:20.980
If you're on the same
indifference

00:42:20.980 --> 00:42:22.820
curve as you were before--

00:42:22.820 --> 00:42:24.660
that's the definition of
Substitution Effect where I

00:42:24.660 --> 00:42:26.700
keep the same indifference
curve--

00:42:26.700 --> 00:42:31.840
but you're tangent to a steeper
budget constraint,

00:42:31.840 --> 00:42:34.720
which is what happened when
prices go up, you must be

00:42:34.720 --> 00:42:36.800
choosing less of the good.

00:42:36.800 --> 00:42:37.400
Think of it graphically.

00:42:37.400 --> 00:42:38.160
Just look at the graph.

00:42:38.160 --> 00:42:40.290
I'm in the same indifference
curve.

00:42:40.290 --> 00:42:43.160
But to be tangent to a steeper
curve, it's going to have to

00:42:43.160 --> 00:42:44.640
be to the left.

00:42:44.640 --> 00:42:47.320
So graphically it's going to
have to be that I'll choose

00:42:47.320 --> 00:42:51.330
fewer movies when the price
of movies goes up.

00:42:51.330 --> 00:42:52.480
That's graphically.

00:42:52.480 --> 00:42:56.240
Mathematically we can just say,
look, what do we know is

00:42:56.240 --> 00:42:59.120
our rule for utility
maximization?

00:42:59.120 --> 00:43:01.880
We know our rule for utility
maximization is that the

00:43:01.880 --> 00:43:05.130
marginal utility of movies over
the marginal utility of

00:43:05.130 --> 00:43:09.400
prices equals the price of
movies over the price--

00:43:09.400 --> 00:43:10.390
marginal utility of pizza.

00:43:10.390 --> 00:43:11.230
I'm sorry.

00:43:11.230 --> 00:43:11.450
Marginal utility of pizza.

00:43:11.450 --> 00:43:13.370
So the price of movies,
so the price of pizza.

00:43:13.370 --> 00:43:16.060
Or the MRS equals the MRT.

00:43:16.060 --> 00:43:17.230
We know that mathematically.

00:43:17.230 --> 00:43:20.350
That's our maximization
condition right?

00:43:20.350 --> 00:43:26.400
Well if the price of movies goes
up, holding the price of

00:43:26.400 --> 00:43:29.345
pizza constant, then the right
hand side has risen.

00:43:33.120 --> 00:43:37.770
If the right hand side rises,
then the left hand side has to

00:43:37.770 --> 00:43:40.680
rise to get this equality.

00:43:40.680 --> 00:43:44.090
How does the left
hand side rise?

00:43:44.090 --> 00:43:46.990
The left hand side rises by
either the marginal utility of

00:43:46.990 --> 00:43:49.300
movies going up or the marginal

00:43:49.300 --> 00:43:52.270
utility of pizzas falling.

00:43:52.270 --> 00:43:53.770
And how do you do that?

00:43:53.770 --> 00:44:00.340
By shifting away from movies
towards pizza.

00:44:00.340 --> 00:44:02.840
How do you make the marginal
utility of movies go up?

00:44:02.840 --> 00:44:05.000
Consume fewer movies.

00:44:05.000 --> 00:44:08.060
How do you make the marginal
utility of pizza go down?

00:44:08.060 --> 00:44:09.300
Consume more pizza.

00:44:09.300 --> 00:44:11.410
Now in this case, pizza
doesn't change in this

00:44:11.410 --> 00:44:13.740
particular case, but
in general it can.

00:44:13.740 --> 00:44:18.145
But the key point is you are
going to see this Substitution

00:44:18.145 --> 00:44:20.760
Effect is going to shift
you towards fewer

00:44:20.760 --> 00:44:24.170
movies to try to get--

00:44:24.170 --> 00:44:29.210
basically because given utility
constant, to try to

00:44:29.210 --> 00:44:31.870
equilibrate this, you're going
to have to move to a higher

00:44:31.870 --> 00:44:33.240
marginal utility of movies.

00:44:33.240 --> 00:44:35.430
Given that the price of movies
has gone up, you're going to

00:44:35.430 --> 00:44:38.310
have to move to worlds where you
care about movies more, or

00:44:38.310 --> 00:44:39.560
you're not in equilibrium.

00:44:41.370 --> 00:44:42.470
Think about it this way.

00:44:42.470 --> 00:44:49.170
If the price of movies goes to
$100 a movie, and you're going

00:44:49.170 --> 00:44:53.330
to be indifferent to where you
were before, on the same

00:44:53.330 --> 00:44:56.610
utilities you were before, it
can't possibly be true that

00:44:56.610 --> 00:44:59.030
you consume the same
number of movies.

00:44:59.030 --> 00:45:01.060
You'd have to be sadder if you
consumed the same movies.

00:45:01.060 --> 00:45:04.310
You're going to have to
move away from movies.

00:45:04.310 --> 00:45:05.590
And that's why the Substitution

00:45:05.590 --> 00:45:08.440
Effect is always negative.

00:45:08.440 --> 00:45:09.970
You're always going to move away
from the good where the

00:45:09.970 --> 00:45:13.370
price increases, holding
utility constant.

00:45:13.370 --> 00:45:17.360
But then we have the second
effect with the Income Effect,

00:45:17.360 --> 00:45:22.180
which then basically, if the
good is normal, reinforces

00:45:22.180 --> 00:45:23.420
that Substitution Effect.

00:45:23.420 --> 00:45:25.990
It says not only do you not want
movies because they've

00:45:25.990 --> 00:45:28.650
gotten more expensive, you also
don't want movies because

00:45:28.650 --> 00:45:30.410
you're poorer.

00:45:30.410 --> 00:45:31.790
Your opportunity set's
restricted.

00:45:31.790 --> 00:45:33.780
And when your opportunity set's
restricted, you buy less

00:45:33.780 --> 00:45:35.760
of everything including
movies.

00:45:35.760 --> 00:45:37.170
So there's two reasons.

00:45:37.170 --> 00:45:40.800
We only saw one own-price
elasticity demand, one shift.

00:45:40.800 --> 00:45:43.020
But there's two reasons
behind that.

00:45:43.020 --> 00:45:44.300
The first reason is because the

00:45:44.300 --> 00:45:46.370
relative prices have changed.

00:45:46.370 --> 00:45:49.080
And the other is because you're
effectively poorer.

00:45:49.080 --> 00:45:52.280
You add those up, and you
get the final effect.

00:45:52.280 --> 00:45:55.040
Now the first, as I said,
is unambiguous.

00:45:55.040 --> 00:45:57.410
That Substitution Effect will
always move you to the left.

00:45:57.410 --> 00:45:59.620
You're always going to want less
of a good if its price

00:45:59.620 --> 00:46:01.240
goes up from Substitution
Effect.

00:46:01.240 --> 00:46:03.180
The second is ambiguous.

00:46:03.180 --> 00:46:06.780
That depends on whether the good
is normal or inferior.

00:46:06.780 --> 00:46:09.690
Next time we'll talk
about what happens

00:46:09.690 --> 00:46:11.170
with inferior goods.

00:46:11.170 --> 00:46:15.390
With inferior goods, we can see
that we can actually get

00:46:15.390 --> 00:46:17.210
what we call a Giffen good.

00:46:21.570 --> 00:46:28.830
A Giffen Good is a good where
it's inferior, so the income

00:46:28.830 --> 00:46:30.820
effect goes the other way.

00:46:30.820 --> 00:46:32.900
And you can actually technically
get a good where

00:46:32.900 --> 00:46:37.105
when the price goes up, you
actually want more of it.

00:46:37.105 --> 00:46:38.750
When the price goes down
you want less of it.

00:46:38.750 --> 00:46:40.590
That is, you can get
a wrong sign, wrong

00:46:40.590 --> 00:46:42.370
slope demand curve.

00:46:42.370 --> 00:46:43.930
I say demand curves
always slope down.

00:46:43.930 --> 00:46:45.950
I might even refer to it
as the law of demand.

00:46:45.950 --> 00:46:48.070
That's not technically true.

00:46:48.070 --> 00:46:49.900
Technically there exist goods,
we'll talk about the next

00:46:49.900 --> 00:46:52.210
time, called Giffen Goods where
you can actually get the

00:46:52.210 --> 00:46:54.390
demand curve sloping
the wrong way.

00:46:54.390 --> 00:46:57.190
The price increase can actually
cause you to want

00:46:57.190 --> 00:46:58.820
more of it.

00:46:58.820 --> 00:47:00.350
In fact, these are based--

00:47:00.350 --> 00:47:02.680
I like the name Giffen because
it's close to griffin.

00:47:02.680 --> 00:47:05.520
And griffins are imaginary
and so are Giffens.

00:47:05.520 --> 00:47:08.670
In fact, there's no evidence
such goods exist, but it's a

00:47:08.670 --> 00:47:11.160
nice theoretical concept to
build your understanding of

00:47:11.160 --> 00:47:12.530
Income versus Substitution
Effects.

00:47:12.530 --> 00:47:14.900
So for next time think a bit
about how that could be.

00:47:14.900 --> 00:47:17.490
And we'll come and show you
graphically how you can get

00:47:17.490 --> 00:47:20.750
Giffen goods depending on the
sign of the Income Effect.