Everyone's checking
before or if we're all, you know,
getting the content, but I guess we'll
find out soon. So the plan today
is I've finally gone through the survey
data. I want to talk about that with
you a little bit. Then I'm going to
break down what we actually did on
Wednesday. As I said, there's purpose to
it. So I'll go through that and then we'll
get started on the next topic, which
is market failures. So thanks to
everyone who filled out the survey.
We have around, I think, 220 people
across the free sector. and we
got 119 responses. Given this survey
wasn't incentivized, like I didn't say I
would pay anyone to do it, I just wanted to
see what you thought, I do appreciate that.
So thanks everyone who put in the time
to give me feedback. So the main reason
why I asked it was I wanted to know what
people preferred. Option one, cancel
the classes on the day of the exam. Option
two, use the classes for revision, then
cancel three other classes at different
times, probably on Fridays, so we can
have longer weekends. Does anyone want to
take a guess? What was the more popular
one about how much? Yeah, what do you reckon? I'd say option two. Okay, that's
pretty hectic. Does anyone
disagree with that? Are we all similar? Okay. Well, you'd be pretty
much right. We had 95 .8% of people clearly
wanting option two. So I've decided I'm going
to do option one. No, I'm not going to
do that. That would be kind of fucked up,
wouldn't it? So we're doing option two. We're going
to do review classes on the day of the
exam. They will be valuable. I'll also try
and find a way to host a couple of hours review
after my 4.30 class before the exam I'll
need to find a big room But I like to give
people the opportunity to come and ask questions
on the day of the exam and I'll let you
know When I cancel classes we'll definitely
do one on the Friday before spring break
So some of you that might head off earlier
You won't miss anything and I'll let you know
what days they are update the syllabus and
make an announcement Okay, so we've done
that and I asked like what people's
backgrounds were I was kind of surprised
that the highest percentage is people
from engineering. I actually didn't
expect that. It's really interesting,
as we'll see on the next page.
It means we have large variants in
our class in terms of mathematical ability.
Our engineers are usually pretty great
with mathematics. That's why they're
here at Purdue. And probably a
lot of them are better at math
than I am, if I'm being
completely honest. Whereas there are other people, as we
can see here. This is a scale of 1
to 7. How confident are you in math? Where 1
is, oh God, no, not math. seven is I'm the
greatest mathematician alive and we have a
really skewed distribution we have five percent
of people who think they're the greatest
mathematician alive I don't know how
that's possible but it is and then we
have a few people down here like yeah
math not for me so this has kind of
been informative for the way I'll teach
this class I probably underestimated
people's mathematical ability but I don't
think I'm going to just gloss things over at
the end of the day like a lot of it is
understanding the math and at the same time
I want people not to fall behind so
if you're struggling with any of the math
come see me I'm happy to help and I do
apologize to those who are the greatest
mathematician alive if I'm explaining the
basics of calculus etc but this stuff is
important in terms of intuition and how to
solve these problems. I asked two
questions about GPAs the lowest grade
you are happy with and your
expected GPA. So this will be the
lowest grade people are happy with and
as you can see the distribution is quite
heavy to the left with a lot of people
expecting to get Sorry, a lot of people
send out only the happy and lowest with
an A, A minus, and I guess down to a
B. Does anyone know, based on the Mish
Daniels policy, what the average GPA should
be in this class? Yeah. 3.0. 3.0. So we have an average here of 3.38. And this gets even
more interesting on the next page when I say,
what do you expect? Average expected
GPA in this class is 3.78 with over
50% of people. it's actually over 60
% of people expecting an A or higher so
on one hand like it is possible
that everyone just smashes this course
I can't scale it down for whatever reason
and we all do well that's probably not
going to happen if I'm being honest but
I thought this was important to show so
we can recalibrate our expectations a little
bit but I do have a question there's some
great answers in the previous class why do
you think this is so high is there anything
going on here that we might not be taking
into account, or other psychological
factors, for example. No bad ideas. You
throw something out of your... Do you have
any thoughts here? Yeah, you. Yeah, what do you
reckon? We took the survey while we
were doing stuff. Okay, great. So the
content probably seemed easy at the time, and
that informed your view. We didn't start
off with doing, like, the heavy game theory
stuff, for example. Michael? We haven't had our
first exam. You haven't had your
first exam. Yeah, that could be a reality
check. like it was for my Behavioural
Econ students last semester, so
that's a good one. Anything else
that we've kind of covered this
semester already? Great, that's true. Like a lot of
people have taken either AP micro in
school or micro in first year. You need to do
well enough to get into this class so
the people that failed that aren't here.
That's a good one. On a similar note to
what Brooke just said, there's one key thing
I'm looking for here. Why do we see over
60% of people say A or above? Is there
something going on here? say a lot of people
feel like their lowest grade, their lowest
grade that they'd be happy with at like
maybe D plus or A minus. And I feel like
everyone kind of just naturally expects that
they're going to do better than what the
lowest grade they'd be happy with. That's why
we see so many more pain in the expected
person on the initial grade that they would
be happy with. This is a great point.
You should be in my experimental class.
This is something called order effects. The order
you present different questions can influence
how one answer is compared to another.
If we ask the other way around, it could
be vastly different. So it's a great
point. Not what I was personally looking for,
but fantastic point. I didn't even think of
that one. You should be thinking about it.
It's very real nice. Did you have something
different here? No? Anyone else think
about our first class when we spoke about
data at the end? Did every person
fill out the survey? No, we only had
119 out of 200. Could that be
important here? You're nodding? Is that just like an
affirmation nod? Or is that a, hey, there
is something going on here? It could
be from a standpoint of we're seeing skewed
data because people who are more willing
to put in effort towards classes in
forms of filling out surveys and putting
in extra time beyond what's required will
perform better and therefore expect themselves
to perform better because they can't
work. 100% correct. This is selective
data. But this isn't a random sample, I didn't
make everyone fill it out, only like just
over half the class did. Who's more
likely to fill out the survey that you don't
have to fill out for class? Probably people
in this room that have this motivation
to rock up for class, putting this extra
effort to help out. These are the types of
people that probably expect to get A's
and probably will get A's. If I got
everyone's data in this class, the people who
didn't fill out the survey may not think
that they'll get A's or may not be unhappy
with a C, for example. Their aims are different.
So we have selective data here. And this
is really important because you shouldn't
interpret data on the surface level. You've
got to think about who filled out your survey.
So even though we have a really high
expected GPA here, I'm actually not taking this
too seriously because I think on average people
that filled out the survey, also the people
in this room, expect to do better and
honestly probably will do better as we established
in the first class. Coming to class, paying
attention, keeping up, just means you'll
do better on average. So, great point,
really important to think about
selection when it comes to how you
get your data. So, what are people
looking forward to? A lot of people are
looking forward to the intuition behind
the ideas, which is kind of what we're
focused on in this class, as I said. The math
is the foundation, but the intuition
is really important. Game theory and
strategy, I was happy to say this. The last
third of our class will really dig into this
type of stuff. I think it's really cool and
I'm glad others think so as well as I said
I did actually get a chessboard for my
office but once I do anyone's welcome to
come in for a game when I'm around also just
the class experience I've heard the professor
is great in his other classes I've
taught one class here at Purdue so small sample
size future students will probably get a
better idea based on your experience you
can tell them whether you had a good time
or a terrible time so that will be a more
accurate description. So concerns, a lot of
people are concerned with the math, the
quantitative aspects. I get it. I think math
can be taught very poorly in a high school
all the time. So a lot of people believe
they're not good at math when they just haven't
been taught the right way. If anything
with the math is troubling you, please seek
help with me or our TA Ishita. It's what
we're here to help you with, so don't stress
too much about that. Pace, difficulty,
and workload. The two major
negative feedback, things of feedback
I got in my evaluation last
semester was one, he goes too fast. The
other is he goes too slow. So I don't really
know what to do with that. At the end of
the day, you can't make everyone happy.
I know I tend to have more of a faster pace,
but hopefully I give you enough to get the
foundation of Edge Concept where you can
review it yourself. Once again, if you don't
understand something i have three hours
dedicated to office hours a week and i'm happy
to meet with you at other times as well so
reach out use me it's what i'm getting paid
for i'm here to help also i did mention this
on the first day of class i think i know
i talk fast there's not much i can really
do about that um i just get excited by economics
what can i say um my accent can be
strange at times as well if english is not your
first language and you miss something that
i say don't hesitate to just tell me to
repeat it if you miss something i'm telling
you there's a very good chance that half the
class missed it as well you're not the only one
grades and performance totally understandable
like that stuff is just like what is
in the forefront of your mind as i said
um it's not really the greatest measure of
learning but hopefully by being here and you
know um coming to class doing the all the the
homeworks etc this will help and finally
exams and assessments exams are high variance
you could just have a bad day that's why
I'm letting you drop one of your exams where
we're not all you know infallible it happens
you just bomb an exam so hopefully that
helps you out a little bit also exam strategy
is important here for multiple choice
don't get bogged down on like a single question
you'll run out of time on the others so
simple things like that can really help I'll
give you more information about that later
finally some other comments which I thought
were interesting oh i should mention clearly
i used um gpt to summarize all this you
can probably tell by the the emojis very useful
it's synthesizing large amounts of data
so use it when you can when it helps does
winter get this cold in australia hell no it
doesn't get below 32 where i'm from so
this has been a really miserable week for me
um i'd recommend trying to do guided notes they
were helpful i don't know what guiding
notes are if any of you do and want to reach
out to me i'm very happy to implement it i'm
open to all ideas that you think can help
your learning process how much math should
i know for the actual solving of problems basic
algebra is sufficient however having a grasp
on calculus i think is really useful
especially the intuition behind it what is the
first derivative the second derivative tell
us things like why the marginal benefit
decreases over time why do we want the marginal
benefit equal to the marginal cost this idea
of optimization is so important in economics
which is what calculus is all about and
finally like why the parameter the sign in
front of a parameter or a term matters so much in
how we interpret what goes on in these
equations calculus makes it so much easier you
don't need to know it but if you can get
some sort of grasp on what's going on that will
really help and how do i prepare for exams
that's a great question when we get closer
we're still what three weeks away just under
three weeks So I'll give out some practice
exams, go over the lectures, the e-book. The
e-book's really solid. Sometimes, as you
know, I'll tangent a little bit of my
lectures. I'll let you know what tangents
are accessible and what aren't, so you
don't have to guess. For the ones that
are, I'll give you a little bit more
information, some extra questions
on it. And I'll probably generate some
more questions from the e-book as well
for you to do. And yeah, as I said,
multi-choice exams, so strategy is
important here as well. Any questions about
the survey? Yeah. So I'll be quite frank,
the reason why it's multiple choice is I have
250 students and five hours of a TA assigned
to me. If I do anything that isn't
multiple choice, I'll be spending three weeks
marking, and I just can't do that. So that's why
we're doing multiple choice. Like, if I
only had, like, you know, this one class,
I would do more things where you have to, like,
graph things yourself, et cetera. I do think
that's better in terms of learning.
But, yeah, pretty much. If I ask, I will ask
questions that involve some calculation
sometimes. like what is the consumer surplus based
on this what is the equilibrium price
and quantity based on these equations etc
and obviously i'll have five different answers
if i'm feeling nasty i'll have a none of
the above answer but i probably won't do that
to start with i want to see how people go on
the first exam if it's too easy then that's
when you start to expect it on the on the
second exam so um yeah that's hopefully a
sufficient answer yeah that's a good question
so i'm usually not a fan of a formula sheet
because i think the intuition of why
something is that way is important um for the
first exam really the only formula is to know
uh present value uh supply demand should be
intuitive based on the law of supply law of
demand and elasticity when we when we speak
about that next week That's probably going
to be the bigger one. Let me think about
it, let me think about it. I didn't form
a behavioural class last semester when
we were going through some hectic models,
so I'll probably lean towards no, but yeah,
I'll let you know. Okay, so thanks for
everyone who rocked up on Wednesday. I like having fun
doing these interactive things, but I won't
do them for no reason. You're paying
for an education here, they need
to have a higher purpose, and this one
specifically does. So before we dig into
it, I just wanted to get some of your
reflections. so have a think about how things
played out during class did you have a
strategy for the first trading period did
this strategy change over time if so why
and did you make any observation about others
behavior or just in general what you think
was going on during this five trade so
sorry i forgot your name over here dk so
you ended up winning a grand or three
dollars correct Do you remember your strategy
going into the game at all? Was it successful
or unsuccessful? What do you think was the barriers to you
doing better? I think my strategy
was just trying to... Were you a buyer or a
seller? I was a buyer. Okay, you were a
buyer. So if you have a two or three, you can
only buy below that. You found out you
were pretty screwed into a situation,
so fair enough. Did anyone have a
particular strategy they think worked
or didn't work? What about if
someone, our sellers are up here, yeah?
Everyone, you know, once you get
a seat in the class, you kind of sit
there forever. So I assume our sellers
were up here. Did anyone
here, like, what would you do if
you received a seven or an eight
as a seller? Were you here on
Wednesday? Yeah. Yeah, were you a seller? I was a
seller. What happened when you got a seven
or an eight or five? It was tough. It was
tough? Did he try or did he just walk down
the counter? I tried. Okay. I would call the big toe. I respect that. yeah exactly so
others with seven and eight up here or
like a two or three spies how many of
you just decided like in round four
or five to just like walk down the
front and back why yeah exactly that and what you saw on the board exactly so we learned
throughout like what prices since
we're being bought and sold out and
that informed a lot of people how
much effort you should put in Effort
isn't costless. You could have just
sat down and scrolled on your phones for the
five minutes if you really wanted to. So
for those of you, sorry, what was your name
in Snoopy's? Kevin. Kevin, so I applaud
you for putting in the effort even when the
probability was really low to kind of zero. But
others, I also respect that a lot. Like,
just being like, yep, this doesn't work.
I'm going to give up right now. So
giving up at the right time is a
really great decision. So for simplicity,
imagine we did this experiment with only 18
people nine buyers and nine sellers where
sellers got these these cards that showed
their minimum cost and buys their value so
sellers you know here can sell above two above
eight buyers can buy below 10 below four
etc given this alone can we make any predictions
about what price trades will happen
and how many trades should happen in each
period of this Three, four, five people. Six people. Hands up
if you think we can't. A couple of people.
Hands up if you've got no idea and
just want to see how it plays out. The
rest of you ask here. So the cool
thing is we can. So what is going
on here with these seller's costs? So
I'm going to pick on someone here. What's
your name? Sorry. Path. Path? Path. P-A-R-T-H. Path. Okay. Excuse my
accent here. Path. Okay. So imagine the
price of the good can only be bought
and sold at $3. actually let's go $2
.50 can only be sold at $2.50 how many
sellers are going to be willing to sell at $2
.50 only two of them so these two sellers
it cost them $2 and at $2.50 they make
a profit of 50 cents but the next seller
who costs $3 will make a loss if they
produce and sell so we know when the price
is $2.50 eat, we expect the supply of
two units of the good. Now, Parth,
what about when it's $3.50? How
many units are going to be
produced and sold? Exactly. So now it's
worth it for this seller to produce and sell,
so on and so forth. Now, Parth, I'm
going to come back to you here. If the price
of the good is $9 .50, how many buyers
are going to buy? Two. So two
buyers' value at their willingness
to pay is $10. And $9.50, they get
50 cents surplus from buying. So two,
we're going to demand it. Our quantity demanded
is two. As we move up, as the price goes
to $8.50, now it's worth it for this
person to buy as well. We have three units
demanded at $8.50. What we've done
here, all this is, this is our supply
curve and this is our demand curve. So
all we need to do is draw it up like
this on the next page as we see when
it is $9.50, we have, you know, two
people demanding it. Sorry, yeah, two people
demanding it, so on and so forth. At a
price of $8 here, every seller is going to
sell all nine. So it's at nine. And at a
price of $4, all buyers are going to want to
buy. We can just draw up our demand-of-supply
function in this way, and where they intersect
is our theoretical equilibria. So in
this situation, they intersect here at 6.5
units at a price of $6. So theoretically, we
should expect six to seven trades to occur
at a price of $6. And the effect of
a $1 tax shifts the supply curve up by
$1. If you originally had a minimal cost
of $2, it's now at $3. So everything's
just shifted up by one. And as you
can see, the new equilibria suggests
there's going to be less being supplied
at a higher price. These are the theoretical predictions
just from that. So, theory and
empirics can be divorced from each
other a lot of the time. That's what
we talk a lot about in my behavioural
economics class. Charlie Holt in 1996
ran this experiment in his class. This has been
done a lot actually. And the question is
how do people behave? so from charlie's data
he used those values that we discussed
you can see the equilibrium price where
supply and demand meet is at six dollars and
what you can see is in each round he did
seven rounds the prices that people traded
at and the dotted line is the equilibrium
price so what can people notice here
does anyone have any observations about how
things played out let's talk about the for
the first six rounds. Yeah. Sorry, what's your
name over here? Ryan. Ryan, what
do you notice? Is there something
different going on in the early rounds
and latter rounds? There's just a
lot more efficient in the rounds and
the last rounds. Perfect observation. We
have price dispersion. We have weird sales
going on at four and seven, and as
you'll see in our data, it's even more hectic
in the early rounds, but it flattens out.
Why do you think it's flattening out
of the latter rounds? And what do you
think they're basing those
strategies on? There's something else
here. That's a great point. There's
something else going on here, though. Is that
a hand here? Yeah, that's a great point.
How many of you looked at the board
to see what the prices of things being
bought and sold at was? And did you
condition your behavior and your offers
based on that at all? A lot of the
people here, yeah. This is why people
didn't give back in the first round or
the first two rounds. But in rounds four
and five, if you were a seller with like
an eight times ten or a buyer with like a
four and under, you looked at the board
and you're like, no one's buying and
selling like for a price like this, I'm going
to give it back. So yes, your own
experience mattered a lot. But the key thing here
in Holt's study and in ours is that people
were able to form expectations on what
others were doing. And what's really cool
about Holt's experiment is you see as people
learn and gain information, we get
pretty damn close to the equilibrium price on
average. round five and six especially you can
just see only one or two trades outside
the equilibrium price round seven he puts
in a two dollar tax so you can see it jumps
from two to four in this case from four
to six and as a result the equilibrium price
shifts up by one dollar here which actually
plays out in the in the data pretty nicely
as well so as as ryan said we see price
dispersion early on there's a few reasons
for this people don't know what's happening
really you don't know the data, and a seller
with an $8 cost may find a buyer with a
$10 value. They realise they can agree on a
trade, but the buyer doesn't think that
they could try and find a seller with a lower
cost and make more. So there's a lot
of things going on that creates
dispersion, but once people see what other
prices are available, it puts these
pressures on to trade ways you can
get more surplus. The data aligns
quite well with the competitive
prediction. However, there's
a big friction in the experiment we did. In the five
minutes of each round, how many
people on average do you reckon
you spoke to? Like, were you
able to speak to everyone in
class? Five, four? Can we have, like,
what, 20 sellers, 20 buyers, something
like that? You're not getting information
on what everyone has. You don't know what
the highest price someone's willing to
buy at and the lowest price someone's
willing to sell at. You only have limited
information in this regard. So this could
be a friction in getting towards our free
market equilibrium. Finally, what role
do incentives play? Holt and others have
paid everyone based on their earnings where
I, as I think was generous, but I only
paid one person 50% of their earnings. Do you
think your behaviour may have differed at
all if you knew you were going to earn what
your total surplus was no matter what? Do you
think you would have made different
decisions? Yeah, I see a few nods. When we
conduct experiments like this, it's important
to take into account incentives and how they
can change behaviour. Okay, across our three
classes, I teach one at 1.30, this class
one at all 30. Does our data look
similar to Holt's? An important caveat
is, like the way you usually do this is
you have a defined set of values, nine buyers,
nine sellers, but I wanted everyone to
have an opportunity to participate, learn
about the way things operate and what
happens when you get a seller's cost of eight
or nine, that you should just come to the front
and give it back. So based on the cards
we had, which was between ace and 10 there
was no aces for buys so we should expect
a price equilibrium price of around five
to six dollars it's not no scientific answer
it's more an estimate so this is our 130
class so the dotted line here is the average
price things were bought and sold at and as
you can see huge price dispersion in this one
early on and it kind of gets smaller and
we'd still have a couple of outliers but in
general the prices on average are around our
six dollar mark which is nice between five
and six and when we put in our one dollar tax
on our sellers you can see this increases
here so this does look pretty nice given all
the frictions all the chaos i'm pretty happy
with this data now our class here this is
i actually think better so in general we we
had like a couple of major outliers but if
you take that away once we get past the first
couple of rounds if we take away that like
one out like we're pretty close to what
the equilibrium price should be between five
and six there's a lot of trades happening around
here which is really nice and once again
once we increase the price based on the
tax you see everything shift up a little bit
most sales are occurring at six and above now
so played out pretty nicely in this class
well I don't know what I'm going tell my 4.30
class because it was pretty chaotic in this
class. Smaller class is you see a lot more
variance and at a time like I think people
started to get the hang of it. There's really
not that much dispersion here. It looks a
lot worse than it is but this one probably
doesn't confirm our predictions unlike all
the other data. But it's important to know
this is not the first time anyone's done this.
It replicates a lot and it gets pretty close
to the equilibrium. So the first time this
was run was in 1948 by chamberlain this was
published in the journal political economy and
was the first ever like economic experiment
and chamberlain's aim was to show that
markets didn't work when people actually played
them out in real life and the reason why he
showed that is because the way he conducted
the experiment was very different to how
i conducted here and i should say very there
was a few slight changes that had big effects
so it's similar to how he did it participants
were students in a classroom were split
into buyers sellers and given a value slash
cost card and they had a specific amount of
time to trade and there was three important
factors here first there were no announced
prices second there was only one round there
was no repetition and three he didn't pay
anyone anything there was no incentives so if we
go back to halt here if you just took the
first round you'd be like okay like yeah it
doesn't really confirm to the predictions it's
only once you get to four five six once people
have an opportunity to learn that we're able
to get the equilibrium prices based on
supply and demand here so repetition is
important secondly announcing prices is
really important if i didn't announce prices
people would have learned what to sell
at what to buy at sorry what was your
name at the back here. Ohm said something
really useful for us to think about. Ohm
saw that things were selling between 5 and
7 on the board or 5 and 6.50. So if he is a
seller who got 7 and above, he knew based
on this information, like, no chance of
him selling. But if he never got any other
information at all and it's only one round,
you know, you may walk into a buyer that
had a 9 and you might have been able to sell
at $8, which is just completely off the
equilibrium prediction so that's why chamberlain's
experiment had vastly different results
so this is um the big one so vernon
smith in 1962 vernon as i'll show in a second
is the father of the field of experimental
economics altered the design of
chamberlain's in a few very important ways and this
was the first experiment to show that behavior
um kind of to convert to the theory
of supply and demand. So rather than using
Chamberlain's pit market, and it's called
the pit market because all the trading takes
place in person, like a little pit
here, even though we're all in the aisles rather
than up the front, he used something
called a double auction. And in a double auction,
buyers make bids, and sellers make offers
asked, and everyone could see the highest
outstanding bid, which buyer was offering
the most, and the lowest outstanding
ask, which seller was offering the lowest
amount to sell it at. So everyone had full
information of the best deal that they
could get, which you didn't have because,
as you said before, you'll spend more than
four or five people rather than knowing
everyone's information. What this means is,
if a buyer says the current highest bid is
like $4, they could offer $4.50 and now be the
buyer with the highest bid. If a seller saw
the lowest sale price was $7.50, they could
offer $7 and they become the lowest sale
price. And once there was convergence between
a bid and an ask, you had a sale. So this bid, our
spread, would diminish. People would keep
undercutting each other or overbidding each other
until we had a deal. And it's called
a double auction because when you think
about it, there's two separate
auctions happening. Buyers are bidding
each other up and sellers are bidding
each other down. That's why it's
called a double auction. It's not
technical or anything. The way he did this
was as follows. So buyer two would be the
first to make a bid. They'd say, I'm willing
to buy it at $3. So that's the
highest bid. Sell of five would say,
I'm willing to sell it at $8. we still have
a gap of five dollars we don't have a deal
and seller six will undercut seller five
say i'm willing to sell at seven so this gap
is shrinking buyer one will offer four dollars
nothing will happen for a minute buyer one
will be like you know what i'm happy to offer
4.5 that's why they bid twice in a row
seller seven comes in and undercuts five and
six by offering six dollars then buyer two
comes back they could have a willingness to
pay of like ten dollars and then says yep I'll
accept six dollars a deal is done so this
is Vernon's version versus the one we did
and unlike Chamberlain Smith introduced this
repetition of trading so based on on the supply
and demand functions that Vernon was able
to create based on the the values he gave
people he said the equilibrium price should
occur at two dollars and if you look at the data
here here's a number of graphs in the
paper once we get to around three four and
five you see it converges really nicely with
the prediction of the equilibrium two dollars
so this is a really famous experiment like
the first experiment actually ever run like
we mean we discount change this is like the
initial experiment that started experimental
economics based on the methodology and it kind
of shows nicely that even if people don't
know what's going on you didn't know that
we were testing supply and demand maybe you
had an inkling but you just had some abstract
value you could make a bit of money and based
on and your behavior and the knowledge that
you have about what others are offering,
we're able to get this prediction to play out
in actual behavior. So Vernon Smith, that's
him. He's actually 99, still alive at
the moment. He won the Nobel Prize in
2002 along with Danny Kahneman for essentially
this experiment and found in the
field of experiment. Can someone tell me
anything interesting about the photo
with Vernon here? Does anyone
notice any of him? Yeah. kind of looks like it's new, it is the
Pudu colours. Vernon ran that 1962 experiment here at Pudu. Experimental economics,
this experiment was run here. It
all started here. When he won the
Nobel, he was actually at George
Mason University, but the way, you know,
academia works is if anyone was ever at
your, like, faculty or whatever, they'll
celebrate you. So that's why you see
those colours. And this is our lab on the seventh
floor of Krannert, where we run, like,
different but also similar experiments.
we pay participants a fair amount of money
has anyone participated in any experiments
here did you make money yeah nice so
would you recommend it to others i would
okay i'll give you a link so you can sign
up in the future i think it's great
experiments i think are really cool in general
i'll talk a little bit more about them
throughout the semester so hopefully this
drove home the point of the free market and
how it can be a great way to allocate
resources, and it's not just theory, we showed,
kind of showed, I guess, that it can
converge towards similar levels of what the
equilibrium predicts. So while the free
market can be quite good at allocating
resources efficiently, there are ways where it just completely fails. And this part
of the course is actually buried
in Module B, Part 6 of the
appendix, but I'm not having that. I love
the free market, clearly I think it's
great in a lot of ways but it can fail
spectacularly in a lot of ways and
it's important to teach you when and
where this occurs. So we have four
ways in which market failures occur. Market
power, we're going to discuss this in a few
weeks time when we talk about different
market structures. So we've all heard
of monopolies, we're going to look into
monopolies and how that can create
inefficiencies in markets. Today we're going
to get started on discussing
externalities, we're going to speak
about on Monday public goods and incomplete
information or asymmetric information when one
person has much more information than another
in a market it can cause what we call
unraveling okay so the myth of consensual free
market efficiency with externalities the buyer
consents the seller consents but there's
a third party who doesn't consent and as
the meme says isn't there somebody you forgot
to ask so what an externality is is when
the costs or benefits of a market activity affect
a third party so eric is eric yeah Eric,
you're selling cheese, for example, and I decide
to buy your cheese. We have a consensual
transaction, we shake hands. But this cheese
is like a moldy cheese, it's a blue cheese,
it's very smelly. And I keep this cheese in
class while I teach. So for all of you,
you have to smell this cheese throughout
class. And I'm sure a lot of you
don't like this smell. You didn't consent to
this activity of the transaction between
Eric and I, but you're being negatively affected
by the transaction. So your negative
well-being based on that transaction
is going to be a cost of this, which
is an externality. So they exist when
costs are borne by parties who are not
involved in production or consumption of
good or service. Other examples that are
quite famous are pollution, noise, or
public intoxication. The reason externalities
cause a market failure is the absence of
well-defined property rights. We don't know
who kind of owns this negative smell that
everyone's dealing with. there's no way to
internalize that on its own now this is where
government comes in government can be one of
the ways we solve these externality issues
there are other ways as well from private
bargaining and other mechanisms but one of
the big things government can do and the one
things that they arguably should do is deal with
externalities so as we can see here our
two black lines here we have our supply curve,
we have our demand curve, and B is the free
market equilibria. And this is for something,
I'll give another example, imagine a company
is producing tires. So they make tires,
people buy the tires for their cars, that's a
consensual transaction. But the production
of tires results in pollutants being put in
the air, which lowers the air quality of
everyone else living in the city, living in the
town. So that's an externality on all these
people so as production increases we also have
a marginal cost of pollution to society and
the more gets produced the higher this cost
is the worse the air quality is for everyone
so once we take into account this marginal
cost of the externality we get the marginal
cost of society producing steel which is
the internal cost which is just how much
production cost in general plus our external cost
which is this marginal cost of pollution
to society so we add them together and we
get this blue line here. This intersects with
the demand curve at point C at QS, so
this is our socially efficient equilibrium.
If we didn't want any extra costs, if we
wanted to be efficient based on the whole
production process, we would produce
at QS. But we're producing at QC because
we're letting the free market decide and
we're not internalizing these costs. We're
not taking into account these marginal
costs of pollution. As a result, this
beautiful triangle that I drew here is our
inefficiency, it's our dead weight loss.
This is the cost to society that
we're not taking into account. It's the
difference between what we're currently
producing at minus the socially optimal
amount, so that's the width of it, and it's
in between our two supply curves above
the demand curve. That's where the dead
weight loss occurs. So to solve the
externality problem caused by pollution, the US
passed a Clean Air Act in 1970 and made
amendments in 1990. In short, this was a
quota policy that you decided on a specific
number of polluting permits that you can
have, and you're only legally allowed to
pollute as much depending on how many permits
you have. And people would buy and sell
these permits as well. So this is called a trading
scheme, essentially, and they only
produced QS permits. So they knew that
people couldn't produce above QS without being
heavily reprimanded. Another thing that
governments can do is just put a tax on
pollution or production. So as you can see here
as we draw this up, here we have our
internal costs, and here we have the social
cost. so to move from internal to social
we can just put a tax to shift the supply
curve up and to the left which results in
us being at this new equilibrium which is
the social optimum so I've got a couple
of things left actually but if you want
to like play along we're going to do this
online but you can you can take a thing
a photo with your phone and do it along
while I'm doing it as well so I'll give
you 20 seconds to do that you don't have
to we're literally doing it up the screen
but i know people prefer to do things
on their own to learn so this is on a
website called marginal revolution by tyler
cowan and alex tabarok they're both amazing
they're both actually george mason where
vernon was when he won the nobel and they're
kind of polymaps in a way but they have a
lot of great resources for like first and
second year econ so definitely use them
so use the following table to graph their
supply curve for denim So when the price is $10, there's going to be
50 units supplied. When the price is $20, there's going to be
150 units supplied. And when the price
is $30, there's going to be 250
units supplied. So we can submit, and
we got it right. And now they say, with
denim dying, chemical used to help indigo
adhere to fabric can pollute nearby waterways
if the wastewater is untreated. This is a
negative externality. So now what they
want us to graph is the social cost. So
both the internal and external costs
together of the supply of denim. So
now at the social cost of $20, we
only produce $50. At $30, we only
produce $150. So everything is now
being produced less. And at $40, $250.
So like this. What is the
value to buyers of the 150th
unit of denim? So what are
buyers willing to pay for the
150th unit? So the 150th unit,
we look at the demand curve, and
it's at $20 here. That's what
buyers would be willing to pay
at most for it. What is the
cost of sellers to produce the
150th unit? Once again, go to 150,
see where it intersects with the supply
curve, so the equilibrium, so it's
also $20 here. It's the minimum amount they'll
be willing to accept. And what is
the social cost of producing
the 150th unit? So we intersect
with the social cost curve, and this is
at $30. dollars. Is it socially
optimal for the market to produce
150 out of 10? Yes or no? What do
people think? Hands up for yes? It is
socially efficient? Hands up for no,
it's not socially efficient. Great,
everyone knows what's going on
here. This is great. All right, fantastic.
We've got a couple left. What is the
quantity at the market equilibrium? It's
just 150. We can see that here. That's
fairly straightforward. And then at the
socially efficient equilibrium, is with our new social cost
curve, it's 100. Fairly straightforward. Okay. And as we said before, what an externality,
a negative externality means
is the market overproduces more
so than the socially optimal output. So
it's just A here. Okay, this is the
big one. Select the deadweight loss, the
inefficiency of the graph. So we have all these
areas here. And remember what I said,
that what's going to be the deadweight
loss is the difference between
what we're producing minus what the
socially optimal production point
is. So that means it can only
be this or this. It's between the two
supply curves and above the demand
curve. So it's this one here. This is
where the social cost is that we're not
taking into account. And this is may
or may not. It will turn up on
the Kahoot quiz. So important to pay
attention here. And using this graph,
what is the per unit tax amount that would
move the market to the socially optimal
level these two supply codes are parallel
that means there's an excise tax being
implemented so if we look at 150 units to
go from here to here we need to move from
20 to 30 so 10 but you can do that at
any point on these two lines and we'll give
you the same answer okay that's that
okay and finally i'm actually not going to get up to the other class So... Destinalities can
take many forms. One of them is
overcrowding. And subways are
famous for being crazy crowded as well.
Where a lot of the time, me going and
getting on a train means someone else can't
get on the train. In fact, there's a
famous Australian ad with one of our most famous rugby
players of all time. And in rugby, does anyone
know anything about rugby here by any chance?
there is something called a scrum where
they like all bend down like 10 meter each
way like 400 pounds each and they just
like drive forward it's kind of like the tush
pushing nfl at both sides are doing it if any
of you are nfl or not rugby fans and there's
an ad with him trying to get onto a train
but no one can get on so he starts to form
this scrum they push everyone on and half the
people get pushed out the other side so in
other words overcrowding creates negative
externalities on others there are ways to solve
this though you can use something like
congestion taxes this is internalizing the
externality of of traffic and also the pollution
caused by traffic so this is in sweden and
as you can see during non-peak time so between
6 30 pm and 6 30 am there's no tax but at
the peak driving times between 7 30 and 8 30
am and 4 and 5 30 pm yeah They finish work
at 4 in Sweden, not 5. It's more
expensive. It's 20 crores, which is,
I think, like $4 or something like
that in the US. So by increasing
the price, you're going to reduce the
amount of people driving on the road
because it's now more costly. A lot
of people won't. And that means
you're reducing traffic overall,
and you're taking into account
this externality. Finally, I want to
just introduce you to another way to solve
externalities without the government, known
as the Coase Theorem. Great. Woo! This is awesome! With pay a person from
US Bank, pay anyone, anywhere instantly
with just their mobile number. Deal. It's
just one of the ways the US Bank mobile
app lets you take care of business and get
on with the game. Woo! I promise. Okay, so it's pretty
cute, but essentially what's going on here
is this guy had a consensual transaction
with a seller to buy this comically large
helmet. But by wearing this, he's created a
negative externality for the woman behind
him. She can't see the gap. But what she did
there was she offered him some amount of
money to reduce this externality by getting
rid of the helmet. And this is known as
the Coase theorem. So if transaction
costs are low, with well-defined and
tradable property rights, parties can
bargain voluntarily to reach the
efficient outcome. The negative externality
can be internalised through monetary
payment. By paying this guy, she was able to get
rid of the externality, and as I'll show in
a moment, both are better off. There's a
number of examples of this. I'll actually get
back to the cattle one in a second. I think
this is a better one. So let's say your roommate
likes to play loud music, but disturbs
you. The table below reflects your happiness
under both scenarios. You, you get $50 of
happiness if there's no loud music, and $10 of
happiness if there's loud music. You're much
worse off with loud music. So what is the
socially optimal outcome? The roommate is $20
better off with loud music and worse off
with no loud music. So the socially optimal
outcome when we add up all the welfare, no
loud music has 100, loud music has 80. So your
roommate doesn't play loud music is the
socially optimal outcome. But assume your
roommate has the legal right to play music and
negotiation is costly so they can play and
they will play and this creates a
negative externality on you because when there's
loud music you're much worse off what
would be the maximum amount of money you'd
be willing to pay your roommate to stop
playing music yeah why fifty dollars it's a great point um there's
one thing you need to take into account here
what's your happiness with the loud music
is it zero great so if you pay fifty dollars
how much happiness to receive and then
you lose 50 from paying 50 so you end up with
negative so you're nearly there take another
stab at it 40 exactly that's the difference
between what you get with loud music and no
loud music so it pain anywhere between zero
to 40 will make you better off or no worse
off I should say so you'd be willing to pay
up to $40 exactly that what was your name
sorry Ivan okay great thanks Ivan well done
okay what's the minimum amount the roommate is
willing to accept to stop paying music
anyone but Ivan yeah $20 same reasoning yeah so
they lose $20 of happiness when they go from
loud music to no music so the only way that
accept that is if I get paid $20 or more
and someone asked a good question the other class
why do we have this point of indifference
why don't we say 39 .999 or 20.001 and the
way we do things in economics is if you're
indifferent between something we say you'd
still be like willing to do it like there's
no real difference in terms of what you you'd
say and because we say the maximum amount
you're willing to pay and the minimum amount the
roommate is willing to accept that's why
we have the point of indifference is the
answer okay so based on this is there a
possible solution can we internalize this externality
using the code here If you have any payment between $20 and $40, based on the
values here, it's going to solve
the externality. So imagine you pay $30. So you lose $30,
but you gain from $10 to $50,
$40 of utility. So $41 is $30 plus
$10, you're better off. Your roommate loses $20
from not listening to loud music, but they
receive $30. So now they're $10 better off
as well, both parties are better off, where
are the social option? Same thing that
happened in that ad, by paying
the guy in front of them to take
off the hat, he's worse off
because he was clearly enjoying wearing the
hat, but by receiving some compensation,
both parties are better off now. So when
property rights are well defined, in this
case it was legal to play loud music,
you can bargain. So if your doormate comes
home really drunk one night, you
technically can try and pay them to not
disturb you, but there are other things
besides these laws in, oh, I shouldn't say
laws, these theories.