pollev.com/chrismakler
Will you need a left-handed desk for exams?
Constrained Optimization When Calculus Works
Christopher Makler
Stanford University Department of Economics
Econ 50: Lecture 10
First Midterm Next Monday!
The homework due Saturday night has several old exam questions as optional exercises.
You don't have to do them all, but you should definitely review the solutions!
Section will be spent reviewing old exam questions.
I will run a review session on Sunday afternoon from 3-5pm, location TBD.
If you have special arrangements, do the paperwork today!
- OAE: upload your form to the link on the web site
- Traveling with a sports team - email me and copy your athletic coordinator
Choice space:
all possible options
Feasible set:
all options available to you
Optimal choice:
Your best choice(s) of the ones available to you
Constrained Optimization
Choice Space
(all colleges plus alternatives)
Feasible Set
(colleges you got into)
Your optimal choice!
Preferences
Preferences describe how the agent ranks all options in the choice space.
For example, we'll assume that you could rank all possible colleges
(and other options for what to do after high school) based upon your preferences.
Preference Ranking
Found a startup
Harvard
Stanford
Play Xbox in parents' basement
Cal
Choice space
Feasible set
Optimal
choice!
Found a startup
Stanford
Cal
Harvard
Play XBox in parents' basement
Optimal choice is the highest-ranking option in the feasible set.
Agenda for Today and Friday
Today: When Calculus Works
Wednesday: When Calculus Doesn't Work
Graphs: PPFs and indifference curves
Words: MRS vs MRT
Math: applying Lagrange
Graphs: PPFs and indifference curves
Words: MRS vs MRT
Math: applying Lagrange
Corner solutions
Kinks
Fish vs. Coconuts
- Can spend your time catching fish (good 1) or collecting coconuts (good 2)
- What is your optimal division of labor between the two?
- Intuitively: if you're optimizing, you couldn't reallocate your time
in a way that would make you better off.
Graphical Analysis:
PPFs and Indifference Curves
The story so far, in two graphs
Production Possibilities Frontier
Resources, Production Functions → Stuff
Indifference Curves
Stuff → Happiness (utility)
Both of these graphs are in the same "Good 1 - Good 2" space
Better to produce
more good 1
and less good 2.
MRS
MRS
>
>
MRT
MRT
Better to produce
less good 1
and more good 2.
MRS
MRS
<
<
MRT
MRT
Better to produce
more good 1
and less good 2.
MRS
MRS
>
>
MRT
MRT
MRS
MRS
<
<
MRT
MRT
“Gravitational Pull" Towards Optimality
Better to produce
more good 2
and less good 1.
These forces are always true.
In certain circumstances, optimality occurs where MRS = MRT.
Verbal Analysis: MRS, MRT, and the “Gravitational Pull" towards Optimality
Marginal Rate of Transformation (MRT)
- The number of coconuts you need to give up in order to get another fish
- Opportunity cost of fish in terms of coconuts
Marginal Rate of Substitution (MRS)
- The number of coconuts you are willing to give up in order to get another fish
- Willingness to "pay" for fish in terms of coconuts
Both of these are measured in
coconuts per fish
(units of good 2/units of good 1)
Marginal Rate of Transformation (MRT)
- The number of coconuts you need to give up in order to get another fish
- Opportunity cost of fish in terms of coconuts
Marginal Rate of Substitution (MRS)
- The number of coconuts you are willing to give up in order to get another fish
- Willingness to "pay" for fish in terms of coconuts
Opportunity cost of marginal fish produced is less than the number of coconuts
you'd be willing to "pay" for a fish.
Opportunity cost of marginal fish produced is more than the number of coconuts
you'd be willing to "pay" for a fish.
Better to spend less time fishing
and more time making coconuts.
Better to spend more time fishing
and less time collecting coconuts.
MRS
MRS
>
>
MRT
MRT
MRS
MRS
<
<
MRT
MRT
Mathematical Analysis:
Lagrange Multipliers
We've just seen that, at least under certain circumstances, the optimal bundle is
"the point along the PPF where MRS = MRT."
CONDITION 1:
CONSTRAINT CONDITION
CONDITION 2:
TANGENCY
CONDITION
This is just an application of the Lagrange method!
Example: Linear PPF, Cobb-Douglas Utility
Chuck has 150 hours of labor, and can produce 3 fish per hour or 2 coconuts per hour.
His preferences may be represented by the utility function u(x1,x2)=x12x2
To find the equation of his PPF, we invert the production functions and plug them in to the resource constraint:
L1(x1)=31x1
L_1(x_1) = {1 \over 3}x_1
L2(x2)=21x2
L_2(x_2) = {1 \over 2}x_2
31x1+21x2=150
{1 \over 3}x_1 + {1 \over 2}x_2 = 150
f1(L1)=3L1
f_1(L_1) = 3L_1
f2(L2)=2L2
f_2(L_2) = 2L_2
L1+L2=150
L_1 + L_2 = 150
Production function for fish
Production function for coconuts
Resource constraint
Example 1: Linear PPF, Cobb-Douglas Utility
Chuck has 150 hours of labor, and can produce 3 coconuts per hour or 2 fish per hour.
His preferences may be represented by the utility function u(x1,x2)=x12x2
s.t.
\text{s.t. }
L(x1,x2,λ)=
\mathcal{L}(x_1,x_2,\lambda)=
x1,x2max
\displaystyle{\max_{x_1,x_2}}
x12x2
x_1^2x_2
150−31x1−21x2
150 - {1 \over 3}x_1 - {1 \over 2}x_2
+λ (
+ \lambda\ (
)
)
OBJECTIVE
FUNCTION
CONSTRAINT
x12x2
x_1^2x_2
150−31x1−21x2=0
150 - {1 \over 3}x_1 - {1 \over 2}x_2 = 0
31x1+21x2=150
{1 \over 3}x_1 + {1 \over 2}x_2 = 150
UTILS
HOURS
What are the units?
UTILS
PER
HOUR
L(x1,x2,λ)=
\mathcal{L}(x_1,x_2,\lambda)=
x12x2
x_1^2x_2
150−31x1−21x2
150 - {1 \over 3}x_1 - {1 \over 2}x_2
+λ (
+ \lambda\ (
)
)
∂x1∂L=
\displaystyle{\partial \mathcal{L} \over \partial x_1} =
FIRST ORDER CONDITIONS
∂x2∂L=
\displaystyle{\partial \mathcal{L} \over \partial x_2} =
∂λ∂L=
\displaystyle{\partial \mathcal{L} \over \partial \lambda} =
2x1x2
2x_1x_2
x12
x_1^2
150−31x1−21x2
150 - {1 \over 3}x_1 - {1 \over 2}x_2
=0
=0
−λ ×
- \lambda\ \times
31
{1 \over 3}
21
{1 \over 2}
−λ ×
- \lambda\ \times
=0
=0
=0
=0
⇒λ = ×
\displaystyle{\Rightarrow \lambda\ = \ \ \ \ \ \ \ \ \ \ \ \times}
31x1+21x2=150
{1 \over 3}x_1 + {1 \over 2}x_2 = 150
⇒λ = ×
\displaystyle{\Rightarrow \lambda\ = \ \ \ \ \ \ \ \ \ \ \ \times}
2x1x2
2x_1x_2
x12
x_1^2
3
3
2
2
MU1
MU_1
MU2
MU_2
MPL1
MP_{L1}
MPL2
MP_{L2}
⇒
\Rightarrow
Utility from last hour spent fishing
Utility from last hour spent collecting coconuts
Equation of PPF
λ = ×
\displaystyle{\lambda\ = \ \ \ \ \ \ \ \ \ \ \ \times}
31x1+21x2=150
{1 \over 3}x_1 + {1 \over 2}x_2 = 150
= ×
\displaystyle{ = \ \ \ \ \ \ \ \ \ \ \ \times}
2x1x2
2x_1x_2
x12
x_1^2
3
3
2
2
MU1
MU_1
MU2
MU_2
MPL1
MP_{L1}
MPL2
MP_{L2}
Utility from last hour spent fishing
Utility from last hour spent collecting coconuts
Equation of PPF
31x1+21x2=150
{1 \over 3}x_1 + {1 \over 2}x_2 = 150
=
=
2x1x2
2x_1x_2
x12
x_1^2
3
3
2
2
MU1
MU_1
MU2
MU_2
MPL1
MP_{L1}
MPL2
MP_{L2}
Equation of PPF
TANGENCY
CONDITION
MRS
MRT
CONSTRAINT
31x1+21x2=150
{1 \over 3}x_1 + {1 \over 2}x_2 = 150
=
=
2x1x2
2x_1x_2
x12
x_1^2
3
3
2
2
x2=31x1
x_2 = {1 \over 3}x_1
31x1+21 =150
{1 \over 3}x_1 + {1 \over 2}\ \ \ \ \ \ \ \ \ = 150
(31x1)
({1 \over 3}x_1)
31x1+61x1=150
{1 \over 3}x_1 + {1 \over 6}x_1 = 150
21x1=150
{1 \over 2}x_1 = 150
x1∗=300
x_1^* = 300
x2=31
x_2 = {1 \over 3}
x2∗=100
x_2^* = 100
(300)
(300)
TANGENCY
CONDITION
CONSTRAINT
PLUG INTO
CONSTRAINT
PLUG BACK INTO TANGENCY CONDITION
31x1+21x2=150
{1 \over 3}x_1 + {1 \over 2}x_2 = 150
x2=31x1
x_2 = {1 \over 3}x_1
x1∗=300
x_1^* = 300
x2∗=100
x_2^* = 100
TANGENCY
CONDITION
CONSTRAINT
TANGENCY
CONDITION
CONSTRAINT
λ = ×
\displaystyle{\lambda\ = \ \ \ \ \ \ \ \ \ \ \ \times}
= ×
\displaystyle{ = \ \ \ \ \ \ \ \ \ \ \ \times}
2x1x2
2x_1x_2
x12
x_1^2
3
3
2
2
MU1
MU_1
MU2
MU_2
MPL1
MP_{L1}
MPL2
MP_{L2}
Utility from last hour spent fishing
Utility from last hour spent collecting coconuts
Interpretation of the Lagrange Multplier
At the optimum, we have x1∗=300 and x2∗=100:
2×300×100×3=180,000
2 \times 300 \times 100 \times 3 = 180,000
3002×2=180,000
300^2 \times 2 = 180,000
What does this mean?
31x1+21x2=150
{1 \over 3}x_1 + {1 \over 2}x_2 = 150
=
=
2x1x2
2x_1x_2
x12
x_1^2
3
3
2
2
x2=31x1
x_2 = {1 \over 3}x_1
31x1+21 =150
{1 \over 3}x_1 + {1 \over 2}\ \ \ \ \ \ \ \ \ = 150
(31x1)
({1 \over 3}x_1)
31x1+61x1=150
{1 \over 3}x_1 + {1 \over 6}x_1 = 150
21x1=150
{1 \over 2}x_1 = 150
x1∗=300
x_1^* = 300
x2=31
x_2 = {1 \over 3}
x2∗=100
x_2^* = 100
(300)
(300)
TANGENCY
CONDITION
CONSTRAINT
PLUG INTO
CONSTRAINT
PLUG BACK INTO TANGENCY CONDITION
\ \ \ \overline L\ \ \
\ \ \ \overline L\ \ \
\ \ \ \overline L\ \ \
\ \ \ \overline L\ \ \
\ \ 2\overline L\ \ \
\ (2\overline L)\
\ \ \tfrac{2}{3}\overline L\ \
x1∗(L)=2L
x_1^*(\overline L) = 2 \overline L
x2∗(L)=32L
x_2^*(\overline L) = \tfrac{2}{3}\overline L
u(x1,x2)=x12x2
u(x_1,x_2) = x_1^2x_2
=38L3
= {8 \over 3} \overline L^3
V(L)=u[x1∗(L),x2∗(L)]=(2L)2(32L)
V(\overline L) = u[x_1^*(\overline L),x_2^*(\overline L)] = (2 \overline L)^2(\tfrac{2}{3}\overline L)
How much utility do you get from that bundle?
What is the optimal bundle to produce if you have L hours of labor available to you?
What is the marginal utility from another hour?
V′(L)=8L2
V'(L) = 8L^2
x1∗(L)=2L
x_1^*(\overline L) = 2 \overline L
x2∗(L)=32L
x_2^*(\overline L) = \tfrac{2}{3}\overline L
u(x1,x2)=x12x2
u(x_1,x_2) = x_1^2x_2
=38L3
= {8 \over 3} \overline L^3
V(L)=u[x1∗(L),x2∗(L)]=(2L)2(32L)
V(\overline L) = u[x_1^*(\overline L),x_2^*(\overline L)] = (2 \overline L)^2(\tfrac{2}{3}\overline L)
How much utility do you get from that bundle?
What is the optimal bundle to produce if you have L hours of labor available to you?
What is the marginal utility from another hour?
V′(L)=8L2
V'(\overline L) = 8 \overline L^2
We found that when L=150, λ=180,000
8×1502=180,000
8 \times 150^2 = 180,000
L1=1001x12
L_1 = {1 \over 100}x_1^2
L2=361x22
L_2 = {1 \over 36}x_2^2
+
+
=100
= 100
1001x12
{1 \over 100}x_1^2
361x22
{1 \over 36}x_2^2
Remember this PPF...let's maximize the Cobb-Douglas utility function u(x1,x2)=alnx1+blnx2 subject to it...
Example 2: Curved PPF, Cobb-Douglas Utility
s.t.
\text{s.t. }
L(x1,x2,λ)=
\mathcal{L}(x_1,x_2,\lambda)=
x1,x2max
\displaystyle{\max_{x_1,x_2}}
alnx1+blnx2
a \ln x_1 + b \ln x_2
100−1001x12+361x22
100 - {1 \over 100}x_1^2 + {1 \over 36}x_2^2
+λ (
+ \lambda\ (
)
)
OBJECTIVE
FUNCTION
CONSTRAINT
alnx1+blnx2
a \ln x_1 + b \ln x_2
100−1001x12+361x22=0
100 - {1 \over 100}x_1^2 + {1 \over 36}x_2^2 = 0
1001x12+361x22=100
{1 \over 100}x_1^2 + {1 \over 36}x_2^2 = 100
First order conditions:
∂x1∂L=
\displaystyle{\partial \mathcal{L} \over \partial x_1} =
∂x2∂L=
\displaystyle{\partial \mathcal{L} \over \partial x_2} =
x1a
{a \over x_1}
x2b
{b \over x_2}
−λ ×
- \lambda\ \times
50x1
{x_1 \over 50}
18x2
{x_2 \over 18}
−λ ×
- \lambda\ \times
=0
=0
=0
=0
⇒λ =50ax12
\Rightarrow \lambda\ = {x_1^2 \over 50a}
⇒λ =18bx22
\Rightarrow \lambda\ = {x_2^2 \over 18b}
18bx22=50ax12
{x_2^2 \over 18b} = {x_1^2 \over 50a}
x22=25a9bx12
x_2^2 = {9b \over 25a}x_1^2
x2=ab×53x1
x_2 = \sqrt{b \over a} \times {3 \over 5}x_1
T.C.
CONSTRAINT
1001x12+361x22=100
{1 \over 100}x_1^2 + {1 \over 36}x_2^2 = 100
(let's square both sides to make substitution easier)
x22=ab×259x12
x_2^2 = {b \over a} \times {9 \over 25}x_1^2
1001x12+361[25a9bx12]=100
{1 \over 100}x_1^2 + {1 \over 36}\left[{9b \over 25a}x_1^2\right] = 100
1001x12+100abx12=100
{1 \over 100}x_1^2 + {b \over 100a}x_1^2= 100
ax12+bx12=1002×a
ax_1^2 + bx_1^2= 100^2 \times a
x1∗=100×a+ba
x_1^* = 100 \times \sqrt{a \over a + b}
[9/36 = 1/4]
[multiply both sides by 100a]
x12=1002×a+ba
x_1^2 = 100^2 \times {a \over a + b}
[factor out (a+b) and divide both sides]
[take the square root]
x2=ab×53x1
x_2 = \sqrt{b \over a} \times {3 \over 5}x_1
TANGENCY CONDITION
CONSTRAINT
1001x12+361x22=100
{1 \over 100}x_1^2 + {1 \over 36}x_2^2 = 100
x2=ab×53x1
x_2 = \sqrt{b \over a} \times {3 \over 5}x_1
TANGENCY CONDITION
1001x12+361[25a9bx12]=100
{1 \over 100}x_1^2 + {1 \over 36}\left[{9b \over 25a}x_1^2\right] = 100
1001x12+100abx12=100
{1 \over 100}x_1^2 + {b \over 100a}x_1^2= 100
ax12+bx12=1002×a
ax_1^2 + bx_1^2= 100^2 \times a
x1∗=100×a+ba
x_1^* = 100 \times \sqrt{a \over a + b}
x12=1002×a+ba
x_1^2 = 100^2 \times {a \over a + b}
x2∗=ab×53(100a+ba)
x_2^* = \sqrt{b \over a} \times {3 \over 5} \left( 100 \sqrt{a \over a + b}\right)
x2∗=60×a+bb
x_2^* = 60 \times \sqrt{b \over a + b}
x1∗=100×a+ba
x_1^* = 100 \times \sqrt{a \over a + b}
x2∗=60×a+bb
x_2^* = 60 \times \sqrt{b \over a + b}
Remember the setup...
x1=10L1
x_1 = 10\sqrt{L_1}
x2=6L2
x_2 = 6\sqrt{L_2}
We can rewrite our optimal bundle as
L=100
\overline L = 100
u(x1,x2)=alnx1+blnx2
u(x_1,x_2) = a \ln x_1 + b \ln x_2
x1=10a+ba100
x_1 = 10\sqrt{{a \over a + b}100}
x1=6a+bb100
x_1 = 6\sqrt{{b \over a + b}100}
x1=10a+baL
x_1 = 10\sqrt{{a \over a + b}\overline L}
x1=6a+bbL
x_1 = 6\sqrt{{b \over a + b}\overline L}
"Spend fraction a/(a+b) of your labor on good 1, and the rest on good 2."
x2=ab×53x1
x_2 = \sqrt{b \over a} \times {3 \over 5}x_1
Tangency Condition
pollev.com/chrismakler
Suppose your preferences may be represented by the utility function
u(x1,x2)=alnx1+blnx2.
What happens to the slope of the line representing the tangency condition if a increases?
x2=ab×53x1
x_2 = \sqrt{b \over a} \times {3 \over 5}x_1
Tangency Condition
Summary
- Some decisions are made by balancing your marginal benefits and marginal costs
- If you value something more than its opportunity cost, you should do more of it
- If you value something less than its opportunity cost, you should do less of it
- The Lagrange multiplier method is a mathematical implementation of this:
- The tangency condition finds all value at which MRS = MRT: that is, bundles where your willingness to give up good 2 to get another unit of good 1 (MRS)
is exactly equal to the opportunity cost of another unit of good 1 (MRT). - Adding the constraint condition finds the most affordable such bundle,
given your current resources. - The Lagrange multiplier captures how much more utility you would get from an additional unit of resources.
- The tangency condition finds all value at which MRS = MRT: that is, bundles where your willingness to give up good 2 to get another unit of good 1 (MRS)
- Next time: situations in which your optimal choice is not characterized by this kind of balancing, and Lagrange fails to find the optimum
pollev.com/chrismakler Will you need a left-handed desk for exams?
Econ 50 | Lecture 10
By Chris Makler
Econ 50 | Lecture 10
Constrained Optimization with Calculus
- 476
