Alright, we're
going to finish this topic today. I'm
going to release it tonight at 5.30 after
the next lecture. No rush, it's during
nine days. So I was going to originally
release it on Sunday after this one finished,
but I know some people have already finished
this homework and are eager to get a start
on the next one. So I wanted to give you that
opportunity. If you're someone like me, it
wouldn't matter when I release it. You'll
wait till the last hour to get it done anyway.
That's how I operated in college. so I get
that as well then next Monday so not this
Monday next Monday is going to be the exam as
I said on Brightspace it's going to cover the
rationality expected utility consumer
optimization and this lecture producer optimization
or cost minimization and the stuff we're
talking about on cost in the moment so those
three lectures I'm going to spend my
weekend writing the exam and then I'll write
up some practice exam questions which I'll
release next week as well and I'll do another
like two hour thing before the exam like I
did last time where I answer questions and
stuff like that so I'll book out a room and
let you know all about that any questions
regarding any of that? fantastic so we had I think
a really productive discussion on Wednesday
about the future of work and AI there's actually
some really big AI labour news that dropped
last night I don't know if anyone's
following did anyone see what announcement was
made by a company by any chance okay does
anyone know who Jack Dorsey is does that name
ring a bell to anyone Jack Dorsey Jack Dorsey
so Jack Dorsey was the founder of
Twitter and the CEO of Twitter for many years
before he sold it to Musk he's actually the
founder and CEO co -founder and CEO of another
company called Block now one of the subsidiaries
of Block you're probably all familiar
with is called Square Square is the the
company where you know you tap your credit
card on that little thing. That's what Square
is. And they made an announcement last
night that they were cutting 6,000 of their
10,000 employees. So 60 % of their workforce
they fired last night and in an announcement
ironically or ironically on Twitter, Jack
Dorsey said the reason for this is because
we're going to move forward with AI. So 60%
of the company got cut. And the question is,
is this like a canary in the coal mine moment
for the labor force? So I just wanted to
show you a couple of things we're going to
do the cahoot today that's not a surprise
the first thing is this is block share
price over the past five days so this is when
the announcement was made so you can see it's
jumped by 15 percent in the past like 24
hours so the market is responding positively
to this announcement it's one thing to take
into account however there's another really
important bit of information here to take
into account regarding block and this is a
question regarding if this is truly about
AI or something else. So if you go
to the entire history of Block
share price, you can notice something. Block hasn't been
doing well the past five years. That is
an incredible drop in share price, which
is a signal about how good a company
is in terms of the productivity and the
wealth it creates. So a lot of people are saying this
isn't about AI. This is just an
overblown company that had to cut costs
one way or another, and it was overdue
for them to fire their workforce.
So why did he come out and say it
specifically due to AI? So does anyone
kind of know why companies hire
management consultants to help them gut
their workforce? Can anyone give me a
psychological reason why you'd hire a
management consultant to do that if you're
a CEO of a company? Yeah? Yeah, you feel
less guilty essentially. If you have an
excuse or you can blame someone else,
I didn't do it, this was the management
consultants, you feel less guilty. So
using AI is a very convenient excuse to
go to workforce without feeling bad about it.
So whether this was actually a genuine
pivot to AI or just a way to get rid
of a whole bunch of people without looking
bad or feeling bad, I don't know.
This is hard to discern between
the two signals. And later in the
semester, if we still have time, we'll talk
about this signaling theory. When someone
says something, is it truthful or is
something else going on? So I don't know what
the answer is here. I just wanted to keep
you in touch with the news of what's
going on right now, especially after
Wednesday's conversation. So we finished up by
talking about opportunity cost. And we used
two examples of Mr. Beast videos to talk
about opportunity cost as both the explicit
cost of the video and the implicit cost of
the video. So in this OG Mr. Beast video from
nine years ago, the explicit costs are
literally like just the webcam and the
microphone. The implicit cost were what he
could do with 48 hours however if you remember
from the video he started off by saying like
please like and subscribe please give
me a shout out he wasn't a big channel there he
probably didn't have a lot of outside
options of the 48 hours it's probably worth a
couple of hundred dollars to him at most whereas
in the other video i showed his most
watched one um which cost 3.5 million dollars
in explicit costs so creating the the
extravagant set hiring all the the the employees
to help out with it probably a bunch
of legal flying people over someone in the
previous class actually told me he was in a
mr. beast video one of the twin challenge
runs I don't know if people have seen that
video he said like they got flown out and
everything like that so there's a lot of a lot
of costs involved in it and there's implicit
costs here as well But the implicit costs
here are probably much, much more expensive
than previously. He's a big shot here
now. So, you know, if this costs 24 hours
to plan and execute, he could have made
a whole bunch of other videos in his
time that would have made a lot of money.
So we're talking about probably hundreds
of thousands of dollars in implicit
costs. And I think if I remember correctly,
maybe I don't, would he make that
40 count to $100,000 number video today?
The answer is almost definitely not. His
implicit costs of making that video are so
much higher now than they were nine years
ago, he wouldn't do that. So opportunity
costs can change over time, given what your
outside options are. So I want to return back to this idea
of sunk costs. And the way we
rationally want to treat sunk costs is that
they don't matter. They shouldn't
matter to your decision making,
but I'll give you an example of how they
might, and do a little behavioral economics
twist on that. so imagine you have a
choice between between going to the football
or going to see a movie that you want to
you really want to watch and you have a
preference for seeing the movie if you haven't
bought a ticket for the football yet you're
going to go see the movie your marginal
benefit of seeing the movie i'm sorry yet seeing
the movie is greater than the marginal
benefit however imagine you bought the ticket
for the football but then only after you
realize a movie you really want to see is on
what should you do so the idea behind this is
you can't get a refund for the football
ticket no matter what decision you choose let's
say it's $100 you've spent $100 on the
football ticket if you go to the football or
you've spent $100 on the football ticket
if you go to the movie you shouldn't take that
into account the only thing that you should
compare is the pleasure you get from going to
the football compared to the pleasure of
going to the movie minus the cost of the
movie ticket if the latter is larger you
should do that but just think about this of what
most people would do they would think to
themselves oh I've already got the football
ticket i should go to the football and another
example in a similar way is for for
holidays you could have booked a holiday over
spring break cost you like a thousand dollars
or something and then a friend says hey
i've got a you know my family's beach house is
free or something come hang out with us for
a week um yeah you don't have to pay anything
you've got accommodation a lot of people
say oh i've already booked this holiday
but you shouldn't think like that you should
think hey is the marginal benefit of me going
with my friend rather than doing doing the
other thing higher if so you should do that
instead so that's how you should operate but
as you can probably imagine a lot of people
don't operate that way so how about for
for mr beast here imagine he's on 90 000
he spent what would it be like 42 hours counting
should he continue with counting or should
he stop counting and essentially it depends
on what his outside option is if at like
90,000, someone goes, hey, do you want to
like come work with me? We can make a couple
of thousand dollars. Then he should just
compare the marginal benefit of $2,000 to
the marginal benefit of counting the remainder
in 10,000 numbers minus the cost of
both. And in the end, this video got like 30
million views. So in hindsight, like
seeing it through was probably the right
thing economically, but there's a good chance
no one ever watched this video. So just counting
to finish something is not the right
way to operate here. So I'm going to give
you a textbook example and by textbook
example I mean an example from the textbook.
So as I say here, since sunk costs are
lost forever once they've been paid,
they are irrelevant to decision making like
what I explained before. So suppose you paid a
non-refundable amount of 10 grand to lease
a rail car for one month but immediately
after signing the lease you realise that
you do not need it. The demand for coal
is significantly lower than you expected,
so you don't need to use this rail car at
all, or if you use it, the benefit's pretty
low. A farmer approaches you and offers a
sublease of rail car from you for two
grand. If the terms of your lease permit you
to sublease a rail car, should you accept
the farmer's offer? So the answer
here is yes. You make more money
from subleasing it. If you don't sublease
it, the marginal benefit is going
to be really low. I don't know, a few
hundred dollars. But intuitively,
do you think a lot of people
might be hesitant to sublease it
in this scenario? Do you have any
intuition why you think people might
be hesitant to sublease? Yeah,
they wouldn't think about the same cost.
They'd be like, if someone could
buy it for $2,000, I bought it for $10,000. Yeah, yeah. Yeah,
so exactly. They're thinking about
the realized loss. They're like,
oh shit, I've thrown $8,000
down the train. And there's something
a bit more to that. You shouldn't fall
for the sunk cost fallacy, but there
are reasons for it. So let me give you an example that I cooked up. So Homer has investments in two different stocks. We have Flancrest
Enterprises. So when he invested
and bought one stock, it
originally was $200, but it lost $100
of value after he invested, and it's
now worth $100. The second stock
he invested in is Copy Global
HyperMegaNet, which was originally $50, and
this gained $50. So both of these
investments that he has are
now worth $100. Homer's friend
Mo offers to buy either of these
two stocks from Homer for $100,
but not both. With the caveat
that the outlook on the future of
both stocks is the same, there's no
difference in how you expect it to
grow or shrink, do you think Homer would be more likely to sell Flamcrest, CompuGlobal
HypermegaNet, or he's indifferent
between the two? Ivan, second one. Because he already
made a profit. I mean, at his point, he paid
$200 for the first one. So I think he holds
more value to the first one where he already
gained money on the second one. Yeah. I
think even though it wouldn't make a
difference about how much money he's getting if
he sold both of them, or either or, I think he would sell
the second one. That way he thinks he's
making money. Great. No, that's fantastic. He thinks he's making
a profit by selling the second one,
Whereas, what would he feel if he sold the
first one? He would be selling it out of
line. Yeah, great. Anyone want to latch
on to what Ivan's saying, or we all
agree with that? Great, got it.
Everyone in the other class said the other
thing, so I'm glad you got this one. It's
not a bad example. This is the idea.
There's a difference in how we value
what's called a realized loss versus
an unrealized loss. And I want to talk to you
about two different things from
behavioral economics. I usually spend a
couple of weeks talking about this, so
I'm going to rush through it a bit. You
can take behavioral economics with me
next semester if you want to learn more,
but there are two things here that
are going on. The first is something
called loss aversion. Loss aversion. What this means is
we feel more pain psychologically
from losses than the happiness we get
from the same gain. So we essentially feel
twice the amount of pain from losing $100
than the happiness we feel from gaining
$100. So an example, would you rather have
A, just nothing happens, or B, you win $100
only to lose $100. Also, how many people
do you think would accept the gamble where
there's a 50% chance they win $100 or a 50%
chance they lose $100? There's a great
Veritasium video actually where he does this
gamble and no one really wants to take
it up, even when he says, hey, there's
a 50% chance you win $20 and a 50% chance
you lose $10. This is a positive expected
value gamble, but many people don't want
to take it because they don't like
this idea of a loss. They put a lot
more weight on the loss than the gap.
So that's the first thing. People
don't like losses. The second thing
is something called mental accounting. This is a little bit
complex, but I'll try and break it down
as easy as possible. So in economics, in
your first ever econ class, one of the
things you're probably taught is the idea
that money is fungible. It doesn't matter if
you have $10 in your rainy day account,
your college fund, your savings account,
your check account, $10 under your
bed. $10 is $10. It doesn't matter
where it is. Mental accounting,
on the other hand, says, hey, it actually
does matter mentally what account you have
these $10 in. The way you value that
money and the way you choose to spend it
is going to differ dramatically. So, for
example, if you have money in like a college
fund growing up, you're probably less
likely to spend that on groceries or
something even though the money should be
fungible between them. So the idea is we value mental accounts
differently. And two mental
accounts that are different that matter are these ones that
I mentioned before, an account that
is realized losses and an account that
is unrealized. So I'll give you an example
of a realized loss. So if you keep $500
under your bed and you leave your window
open, the wind comes in and blows the $500
away, that $500 is just gone. that account is
closed you can't get that $500 back but
if you invest in the stock market for
example does anyone here invest by any chance
a lot of a lot of us here Christian what's
the stock you've invested in recently that is
not doing well okay have you sold that
yet okay why are you keeping it cool yeah
that's great but this is something we call
an unrealized loss yes you're in the
red right now but you haven't closed this
account yet once you sell it the account closes
and you realize the loss you get the cycle
that psychological disvalue from the
loss but before that is unrealized you don't
actually feel that pain just yet so
this is why there's this idea of like I'm
chasing your losses or throwing money
from good money after bad money because
when you're in the red you're thinking
I can leave this open until I'm back in
the black and then I'll sell it off and
I'll get a gain. So going back to our
original example here, Ivan is correct
that Homer is more likely to sell B than
A for two reasons. One, by selling B he
closes this account and he gets the positives
of getting a profit whereas by not selling
the first one he hasn't actually realized
that loss yet and there's still in his
mind the potential for it to go back into
the into the into the block so yeah there's
this psychological motivation for a lot
of gamblers who are in debt like if you're
playing roulette at the casino which i wouldn't
advise you to do and you're down 20 bucks
a lot of people say okay i'll keep going
until i can you know get my 20 bucks back
and then they end up you know 100 in the
hole or something like that so there is this
like value and this huge like you know um
uh shift we put on plus one dollar versus
minus one dollar. The human mind is pretty
wild in that respect. Are we all good on
opportunity costs and sunk costs? This
is kind of like the second time we've gone
over it so hopefully that sticks. So we
can now look at the algebraic forms of
cost functions. So here we have a cubic
cost function and the cubic cost function
we've actually kind of seen this is what
it looks like. As you can see it's not
constantly increasing at a decreasing rate
or increasing at increasing rate. It
goes back and forth. That's the cubic aspect of the cost function. So as you can see
here we have cost as a function of
quantity which is some intercept plus
A times Q plus B times Q squared plus
C times Q cubed. And A, B, C and F
are the constants. So as you can see here
we can run some sort of regression and we
get the values of F which is this and A,
B and C. So we get all those values and we
know how to get a marginal cost function
from our cost function which is take the
derivative of cost with respect to quantity
so that's what this is here as you can see Q
to the power of 1 take the 1 out the front
reduces by the power of 1 the Q disappears
we end up with A take the 2 out the front
reduces by 1 so it's Q to the power of 1 take
3 out the front and then reduce the power
by 1 we have our marginal cost function
and managers can use the regression analysis
to estimate these parameters and figure
out what the marginal cost is at any point.
So as you can see if we go back here, the
marginal cost is changing with certain output.
So you want to know where you are on the
marginal cost curve. So for example, at
Q equals 20, what is the cost to produce
the next unit? This is a marginal cost
question. What is the marginal cost of
producing the next unit when you're
producing 20 units. So we have our
marginal cost function, we have all our
parameters here, and we're being given
where we're at on the marginal cost curve,
which is 20. All we need to do is
plug in the numbers. So we plug in 20 for
Q, as you can see, our intercept here
is we don't have an intercept sorry
our a here is 21 .67 our b is 1.09
and our c is 0.087 just do the
calculations you get the marginal cost which is
173.27 this is just very formulaic nothing
fun or interesting about this something
a little bit more interesting is this idea
of long-run costs so far we've talked a
lot about short-run costs when capital is
fixed but what happens when capital isn't
fixed and you can allocate labour and
capital as you see fit. So we have something
which we call the long-run average cost
curve and this is a curve that defines the
minimum average cost of producing alternative
levels of output allowing for optimal
selection of both fixed and variable
factors of production. So let's focus on the
left for now, don't worry about the right.
So this is quantity and this is dollars per
unit and this is our average cost curve
our long run average cost curve and this
tells us that every point of output what is
the minimum amount of average cost we can
produce at when we can move around both labor
and capital so as you can see here q1 would
occur here on the long run average cost curve
q2 here and q3 up here so now we move to
the right and now we have this other function
as well and this is a short run average
cost curve when capital is fixed at one
whatever that means when capital is fixed at
one what that means is you can only move
labor around and not capital so there's a
couple of things going on here if you're producing
a q1 in the short run when capital is
at one then you are producing at the point
that is the minimum average cost in the short
run and the long run however at any other
point at any other point you can't move
capital around you can only move labor around
so you're making non -optimal substitutions
in the short run that you can only fix in
the long run so for example if you wanted
to produce a q2 q2 on the long run average
cost curve is down here but because you can't
move your capital around in the short
run it's going to be up here where q2 intersects
the short run average cost curve That's
because you're going to have to put less
productive labour into more output rather
than more productive capital because you can't
move capital around. So that's how the
short run differs from the long run in terms
of average cost. As you can see here,
in the short run, this is when capital is fixed
at K1. In the short run, this is when
capital is fixed at K2. So you can see for K2,
it is tangential to the long run average
cost curve at q2 so if you did want to move
to q2 and you were here in the short run
then you change your capital allocation at k2
and end up here in the long run and it's the
same issue here if you start with k2 and
you're at q2 originally but you want to move
to q1 in the short run your cost is going to
be up here and only in the long run you
can change your capital allocations to end up
down here and you can see it's the same thing
with q3 as well so you're bounded your
options in the short run. You can only move
your labor around so there's only ever going
to be one point where you're on this long
run average cost curve minimization and it's
only in the long run you can actually move
around essentially. Any questions on that? Great. So the long run average
cost curve tells us something about the
way costs are going and we call this either
economies of scale or diseconomies of
scale. So economies of scale are the declining
portion of the long run average cost curve,
and this tells us for each extra unit
of output that you create, the long run
average cost decreases. This also means that
the marginal cost is increasing at a
decreasing rate, if you remember the relationship
between our average cost and our marginal
cost curves. So the more you produce,
the cheaper it is per unit to produce, all
the way until Q star. This is what we call constant returns
to scale. The next unit you
produce will cost the same as the previous
unit and then as you get to the right hand side
we have dis economies of scale each extra
unit you produce is going to cost more
than the previous one so some examples here
let's go with bulk buying so an example will
be our lemonade stand you start a lemonade
stand you're going to buy like each lemon
individually that's quite costly but once
you get big enough you don't have to buy
individual lemons anymore you can get delivered
in bulk that's going to be much cheaper
per lemon than it was before that's a huge
cost saving exercise that's going to result
in the next unit being cheaper than the
previous also you might expand your lemon stand
and originally you have to get more laborers
but then eventually once you get a lemon
you know soda packing machine things are
going to be much cheaper you need less labor
and you can produce a lot more for less so
things like technology bulk buying learning
by doing so um you start off by doing
something you're not that efficient by with more
experience you can do it much more efficiently
so the fast food industry is a great
example of this like back in the day they
had a ton of workers but now they can optimize
with a few has anyone here shopped at Aldi
before by any chance kings of optimization
the absolute kings of optimization they
have like three employees in the store whereas
other grocery stores have like 15 20 so
learning by doing and just making things
more efficient is a way to decrease your cost
and more output. On the other hand, when you
start getting things like bottlenecks,
bloating in management, then things become
more expensive the more output you have. Things
like needing more transportation as well
for each extra batch, et cetera, would
result in more costs as well. So you can be in
either direction here. So a really good
example of economies of scale, I'm not going to
show you this video for two reasons the the the
first reason is it's like pretty dark um
but the second reason is the link's broken so
the the second reason is is the main reason
but um factory farming in modern agriculture
is a great example of economies of scale
so farming like you know a hundred years ago
you had like a few chickens to a large pan
a lot of management etc um to to look after
them so the cost per chicken or the cost per
egg was quite high but now if anyone's familiar
with modern factory farming and agriculture
things are just hyper optimized they have
like machines feeding the animals the animals
per square foot is quite low as well there's
a lot of processes I won't really get
into because I'm I find it a little bit
disheartening and you essentially have this
trade-off between the the welfare of the animals
and the cost so we could improve the welfare of
the animals but this would increase the
cost for your your products in the grocery
store as well so there's actually current debates
in congress right now about animal welfare
standards because i know if they put in
laws to increase the standards it's going to
increase the price and there are trade-offs
there but this is an example of economies of
scale over time another example of economies
of scale is is this um idea of ghost kitchens
so if you've ordered off uber eats or
doordash recently you probably ordered from
a ghost kitchen a ghost kitchen is a store a
restaurant that doesn't have like a front as in
they don't have seating or anything like that
so they can cut back on a whole bunch of
fixed costs they can cut back on waiters and
all that sort of thing and they're literally
just making the the food out of the back of
the kitchen sometimes they have multiple
stores in the same place as well so I don't
know if it's like this here but if you're from
a larger city you've probably seen a few like
different restaurants on Uber Eats and they
look similar photos etc they're all coming
from the same place so this is a way that
they they you know pack their costs together
and the more orders they get there the
cheaper it is on average. Also something like
food trucks as well the initial purchase of
the food truck is quite high but your expenses
are really low after that you don't
have to pay really rent or anything like that
you can just drive around and get you know
these cheap permits and the more people you
know demand from you the easier it is to
produce you've got your setup and everything
like that already. So before we put our
thinking hats on, you can also have a
long run average cost curve that doesn't
exhibit economies or diseconomies from scale.
So as you can see, this is constant
returns to scale no matter where you are.
So once again, you'll choose to optimise
in the long run based on whatever you want
your output to be when you can move your
capital around, but there's no place to
get economies of scale or remove from
diseconomies of scale. so let's put our thinking
hats on for a bit so for the following
scenarios do you think these businesses exhibit
economies of scale diseconomies of scale
or neither um there's no like real answer to
any of these it's just like whatever you
think so i'm going to pick on a few people
here and just just get their thoughts so what
about starting a cafe from scratch that first
like month of operating the cafe do you think
there's going to be economies of scale so
saving of cost more you produce dis-economies
of scale increases of cost as you produce
more neither or you're unsure was that a hand
yeah yeah yeah yeah exactly so whether
you're buying like small or by and large will
be a big thing to start with because if you're
you know selling and you buy small it's
going to probably be quite expensive what
about something like labor how do you think it's
going to be early on yeah I think you're
right and that's probably the the more robust
answer like my my counter argument would
be like usually a store is operated by
like a like family or something like that
they won't hire anyone else yet so the more
they produce they don't actually hire
labor to start off with so it can be
cheaper to start. Daniel? yeah that's good as we
talked about learning by doing is economies
of scale you're probably not even in
the learning by doing stage yet you're in
probably the making mistakes by doing so maybe
that first month is this economy is a scale
and after that when you learn the best
ways to go about things you start making um you
know these efficiencies appear so that's what
we want okay next one um dk yeah youtube
content creator starting off do you
think there's economies of scale this economy
is a scale or constant economies of scale i
think it's um economies of scale how come
you're learning by doing so like everything's
experimental and like you're purchasing
like technology and you like don't know yeah
like trying to make all the mistakes and
and if your outputs videos let's say how
much does it cost like from going for one video
to another after you do your initial
purchasing of everything it's pretty much free
yeah exactly yeah so like this is the
classic example of economies of scale what
about let's say five years into your YouTube
content creative career and you're
very successful now does anyone have any
let's talk about mr. beast so mr. beast
in his first couple of years definitely
economies of scale do you still think mr.
beast has economies of scale Oh, or dis
-economy is a scale. What do we think? Yeah, much more
elaborate. Yeah. And he's hiring a
lot of people now. Before, it was just
him. Now he's got video editors. He's giving
out a ton of cash as well. Probably
needs like legal, health insurance, a
whole bunch of things. So this probably means
for each extra video he produces, it's more
expensive than before. That's his business
model. He's bringing in a lot of revenue,
more revenue than before as well. but in terms
of his cost they're not getting cheaper
each video he produces also think about twitch
streamers and stuff when they start off
there's definitely economies of scale like
each you know extra hour you stream is probably
cheaper once you have your setup but at some
point when you get big enough if you stream
outside of your room you're going to need
like security guards you're going to need
people to video you as well so these costs
increase the bigger you get in these environments
i don't know this is a bit of a weird
one owning the nfl franchise i'm going to
skip it i don't like it let's talk about uber
or starting a ride share business so i
don't know if you're all too young to remember
when uber first started but starting from
scratch do you think that exhibited economies of
scale or diseconomies of scale ivan uh i
think it's probably diseconomies of scale
just because you're probably um well i guess
uber doesn't have that I would say you're
probably hiring, I would just go off the fact
that you're probably not having a lot of
customers, so you're not really optimizing, I
guess, and you're spending a lot of money
probably on labor for drivers to be out there
with their cars and for your business not to
be very popular yet. So there's just a slight
difference, I want to make sure we all
know, just because I don't have a lot of
customers, that says nothing about the cost. customers
is on the revenue side so this is the
thing with mr beast before he's probably
at this economies of scale but he's probably
making more of a profit now because the video
is bringing more and more dollars compared
to when he was at a stage before so this
economies of scale are not necessarily a bad
thing if you can get more revenue but yeah
i i don't know i think as you were saying
like the like they had probably you know a lot
more fixed costs per you know um output back
then as well so maybe maybe you're right
there i think there's there's definitely a
point to that take how you had your hand up
yeah yeah yeah so it's just like back stuff
really yeah they needed so yeah I don't really
know I think there's arguments in both
directions I think that one you can probably
like on google it and we'll give you an answer
but i i don't know uh so as we talked about
kind of with the the sorry we didn't talk
about with the cafe in five years for example
you might be really popular you might
need to expand so like like buy the property
next to you and expand you'll need to hire
more labor etc you might already be like optimizing
how much you can buy in bulk you can't
squeeze any more pennies in that regard so
you might start off you know this process
of learning by doing becoming more efficient
then at some point you don't get any more
economies of scale they start getting these
economies of scale same thing with streaming so
you can see here that early on when it when
a company's first producing they could be
at economies of scale so each unit they
produce the average cost in dollars per unit is
less and less and less until they reach q
prime here the the the minimum efficient
scale you can't get any lower than this it's
constant to q double prime, and then the
economies of scale, we get diseconomies of scale,
sorry, so each unit you produce after Q
double prime increases. So at different
stages of a company life cycle, and as
well they produce, you could get
something that looks like this, you can
get many shapes. Finally, I want to
talk about economies of scope and cost
complementarity. So economies of scope
exist when the total cost of producing two
goods, Q1 and Q2 together, is less than the total
cost of producing each of the type
separately so what this shows us is it's cheaper
to have a cost function where you produce both
using the same stuff rather than separating
it out and having the cost of producing
just q1 plus an isolated situation of producing
just q2 if this is more expensive then
you have economies of scope you should create
both together and an example over the next
page is you know we talked about a factory
that has both um the ability to make tractors
and cars you probably have machinery that's
able to make parts for both skilled labor that
you can use to like you know perform both
rather than have to hire double the people
with different skill sets etc and this idea
of cost complementarity exists when the marginal
cost of producing one type of output
decreases when the output of another good
increases so here we're looking at the marginal
cost of good one so the marginal cost of good
one the changing that given q1 and q2
decreases when the output of q2 increases so
you're not doing anything to q1 you're increasing
the output of q2 and this lowers the
marginal cost of good one so this is the economies
of scope example i gave before i don't
really like this cost complementarity
example but i'll say it anyway so let's
say you're producing two products donuts
and and donut holes the donut hole is
let's say you punch a hole in in the donut
i know that that's not how they're all
made but for convenience that's how they're
made. You punch a hole and you fry that
hole and you sell like a little board
or whatever as well. So every time you
make a doughnut you're also making a doughnut
hole essentially. So the firm can make
these products separately or jointly but
the cost of making additional doughnut holes
is lower when workers roll out the dough,
punch the holes and fry both the doughnuts
and the holes instead of making the holes
separately. Like this is pretty clear if
you just made doughnut holes you'd punch
the hole, you'd throw away the donut and
you'd fry the little dough. Obviously this
is going to be cost complementarity.
Another example would be making handbags out
of leather. If your company makes handbags
out of leather and you make steak as well,
the more steaks you make, the more leather
you'll have to make your handbags as well.
So that's probably a better example of
cost complementarity. Also this is archaic
but there's actually a Seinfeld episode
kind of like this with muffins, the muffin
tops and the muffin bottoms, and everyone
wants to eat the muffin tops, no one
wants to eat the muffin bottoms so there's a
little bit on that. Finally, we can look
at the algebraic form of a multi-product
cost function. So this is the cost
function of Q1 and Q2. You can see we have
an intercept, we have alpha here, which
is multiplied by producing Q1 and Q2
together, and then we have Q1 and Q2 kind
of in isolation here. the marginal cost of
producing q1 in this multi-product cost
function is just how does cost increase when we
increase the output of q1 so we take the
derivative q1 so here is just to the power
of one take one out the front we get rid
of q1 and end up with aq2 then here two out
the front of q1 and ends up with this so as you
can see the marginal cost of one depends
on both q1 and q2 so if this a here is
negative below zero what this tells us is as q2
increases the marginal cost of good one is
going to decrease is going to decrease
and that's our cost complementarity condition
essentially so if a is negative for both
the marginal cost of one of the marginal cost
of two if we produce the other good the
marginal cost of the other good is going to
decrease. That's all this is trying to say.
If a is greater than zero there are no cost
complementarities. Great, that's the
end of that lecture. We have all the
information for the exam now and
let's do the cahoot. Oh, let me get prizes. Questions while we wait? Oh, we lost someone. Yeah, anyone still
logging on? Okay, no rush. We're losing people. Are we back in Michael?
Okay, great, I'm starting now brilliant
six questions short run which input is
typically fixed labor capital both or neither
typically fixed yeah pretty much everyone
got that one right in my previous class
only like 30% of people got this right
so I don't know what's going on there I'm
glad you guys did well though right
indeed next question. If q equals the
minimum of 3k, 6l, and k
equals 4l equals 5, what is the
output? 12, 30, 42, 15. The majority of you got
that right. Remember, this is our Leontief
production function. So we'll have two
numbers inside of here, 3 times k, or 3
times 4 is 12, and 6 times l, 6 times 5
is 30. So we want to take the smallest
number out of those, that's what the minimum
function says, take it out, that's the
output. That's why I had the answer 42 there.
You don't add it together in Leontep.
That's the linear production function when
you add it together. Better to make
that mistake now than on the exam
when this will definitely make
an appearance. Okay. Third question. A profit maximizing firm hires labor until
the marginal product of labor equals
zero, the value of the marginal product
of labor equals wage, the average
product of labor equals wage, the marginal
product of labor equals the average
product of labour. Most people got that
one wrong, actually. So remember, the
value of the marginal product of labour is
just how much revenue the next unit of
labour brings in. So the output times
whatever we sell it at. So we want that to
be equal to the wage because if it's above
the wage, then we're leaving money on the
table by not hiring more workers or more work
hours. And if it's lower than the wage,
then we're hiring too many people. The last
unit of labour costs us more than we're
actually bringing in. So that's why we all
want them equal to each other. But we're
going a bit wild with the emojis right
now. So remember, the average doesn't
really matter in this case. Don't get
fooled by the average. It's the margin that
matters in a way. This is like social
contagion to a T. Again, three left. What is the marginal rate of technical
substitution? The slope of the
ISO cost line, the slope of the ISO quant
line, the ratio of output prices, and
none of the above. Marginal rate of
technical substitution. No? We don't like our answers over
here. Brooke? Okay. Yeah, cool.
Majority of people got that right. So
the marginal rate of technical substitution
is how much of labour and capital
you trade off that remain at the same
level of output. That's the slope of our
indifference curve. The slope of the
ISO cost line is actually the ratio
of the output prices of the rental rate of
capital and the wage. So those two are
actually the same thing. I should find a
way to disable these emojis
in the future. Okay, penultimate
question. It's pretty tight
up the top here. Fixed costs change
as output increases. True or false? Are you feeling
better about this one? Okay, good. Most people got
that right. Yes, the definition of fixed
costs is that it doesn't change when
an output changes. Okay. Okay, tied up the top
here, I don't want to pick on someone
because I did that last class and then
they lost the last question and this
is much closer, so you can remain
anonymous for now. Last question. Economies of scales
means long run average cost falls
as output increases. Long run average cost falls as output
increases. Yeah, most people
got that right. Great, let's see who
ended up winning. Yeah, well done everyone.
it's all about the extra credit and
learning but it's also about winning prizes
oh we had a switch up the tour maybe we did
is that is that you congratulations you
won one already no congratulations so I
had an extra prize today you can get a Purdue
stress ball bouncy ball as well or you
can take the most popular prizes obviously
yeah brilliant we can finish up a little
bit early today enjoy your weekend and I'll
see you on Monday congratulations next
time you're going to take 471 sorry you're
going to take 471 right so what do you cover
so remember all the stuff you're drawing on
rational decision yeah we're going to tear
that all down all the ways people don't have
rationally so we're going to look at a bunch
of different research papers showing how
people violate is it going to be a lot of case
studies more research papers and case studies
yeah so I'm going to teach different
research papers so there'll be a couple
of exams a couple of presentations as well
so for the exam is it no I mean like for the
chapters it's these three plus the one next week
no we're not doing the one next okay I've
got more than enough questions to ask for
these we'll see what the turnout is next
week also for the extra credit I messed up on
like these yes questions so I hate this okay
the fixed cost one I don't know I messed
up on like the fixed cost one oh really yeah
it's better to mess it up now so you know it
for the future then mess it up on the exam
right so this is like another good reason
to like do these extra credit quizzes like
people make dumb mistakes I make dumb mistakes
as well but at least you like get that
fixed like you will not forget that now because
of that so yeah don't worry about it exactly
good it should be so don't worry about
it all right have a great weekend yeah they
do we like that so it doesn't really matter
when you think about it all it means is you're
trading off one for the other yeah so like
it's just how it's defined if it's like
how much k do you trade off for l or how much
ld trade-off for k they have it weird in
the textbook where they mix it up like that
i'll make sure on the exam i i clarify um
what type of question it's asking for i think
it's also same for like the consumers at
the company you know you have to trade yeah
it's really annoying really annoying yeah and
then so on the um yeah because yeah you can
see here they've got k on top because it's
the y-axis where yeah here it's yeah so
but they are the same thing you know like
Yeah, so the trade-off is the same. So when you
think about it, you give up one unit of
labour for four units of capital, or you get
one unit of capital for 0.25 units of labour.
It's the same ratio. Yeah, that's
what matters. Okay, so the
expression doesn't matter, just
the way you're explaining it.
Yeah, yeah. Okay, and then, so on
the calculator, you said, like, what's
the... Marginal right of technical substitute.
It's a slope, but it's negative. It's
not only the slope. Shouldn't the opposite
of this... the the the slope is is like you
can take the absolute value of it so it's
still the slope yeah that's what it is and
the slope is changing across time it's the
same thing as um the slope the um the marginal
rate of substitution for consumer theory
how much do you trade off between the
two things okay that's all the same yeah okay
yeah no that was a clear answer that one
yeah yeah and i um well i think one more
like about the variable cost fixed cost The
variable labor cost? No, the value of the module product
of library.