EC 270 - 2026-10-01 - Lecture 07

EC 270 - Lecture 07 - 2026-10-01

Intermediate Microeconomics I · Logan McLeod · LH 1011

Week 4 · 11:30-12:47 · 77 min · Full transcript · Flashcards

Overview

This lecture finished Chapter 5 with non-convex preferences, the last family of preferences whose best bundle sits at a corner. It then started Chapter 6: Demand. Chapter 6 takes the ordinary demand functions from Chapter 5 and asks how they change when income, the good's own price, or the other good's price changes, one at a time. Today covered income changes in full (income offer curves, Engel curves, normal and inferior goods, homothetic preferences, luxury and necessary goods) and got as far as the ordinary demand curve for own-price changes. Cross-price changes and inverse demand are still to come.

Non-convex preferences

This finishes the corner solutions from last Tuesday. Perfect substitutes gave corners by comparing slopes, and non-convex preferences give corners for a different reason.

Non-convex preferences

Preferences whose indifference curves are bowed away from the origin. The consumer likes extremes: all of one good or all of the other, but not a mix. Higher indifference curves are still further from the origin.

Why: these preferences break the convexity rule from Chapter 3, where a weighted average of two bundles should be weakly preferred to the extremes. Here, if you draw a line between the two ends of an indifference curve, the line lies below the curve, so every mix is worse than the extremes.

Why the best bundle is a corner

The consumer still picks the affordable bundle on the highest indifference curve. With curves bowed away from the origin, that is always one of the two corners of the budget line.

Why: McLeod drew three bundles on one budget line. The corner on the x1 axis was on the highest indifference curve. The bundle in the middle, which looked like a tangency, was a mix of both goods and sat on the lowest indifference curve. With equal tastes for both goods and p1<p2, the cheaper good buys more units, so its corner wins.

When prices are equal

Even when p1=p2 (budget line slope of −1), the consumer still buys only one good. They pick the corner with the higher utility. If both corners give the same utility, they are indifferent between the two corners.

Why: equal prices do not change the preferences, and the preferences want an extreme. This is different from perfect substitutes last Tuesday, where equal prices made the budget line lie on top of an indifference curve. A student suggested something like distance to the cash register would break the tie. McLeod's answer was that such costs are really part of the price.

Solving for the best corner

Recipe: spot that the utility function is non-convex, so the answer is a corner. Find both corners: (m/p1,0) and (0,m/p2). Plug both into the utility function. The higher utility wins.

Why: the utility function ranks bundles, so it tells you directly which corner is on the higher indifference curve.

(mp1,0)versus(0,mp2)

where m is income, p1 and p2 are the prices, and each corner is what you get by spending all income on one good.

Example: Max utility with an income of 80

Utility is u(x1,x2)=max{x1,x2} and the budget line is x1+2x2=80, so p1=1, p2=2 and m=80. The budget line slope is −1/2, which is fairly flat.

  1. x1=m/p1=80/1=80 -- all income on good 1 gives the corner (80,0)
  2. x2=m/p2=80/2=40 -- all income on good 2 gives the corner (0,40)
  3. u(80,0)=max{80,0}=80 -- utility at the first corner
  4. u(0,40)=max{0,40}=40 -- utility at the second corner
  5. 80>40 -- the first corner is on the higher indifference curve

Answer: (x1∗,x2∗)=(80,0) -- the consumer spends all $80 on good 1.

Example: Circle utility with an income of 30

Utility is u(x1,x2)=x12+x22 and the budget line is 2x1+3x2=30, so p1=2, p2=3 and m=30. The budget line slope is −2/3.

  1. x1=m/p1=30/2=15 -- the corner with only good 1 is (15,0)
  2. x2=m/p2=30/3=10 -- the corner with only good 2 is (0,10)
  3. u(15,0)=152+02=225 -- utility at the first corner
  4. u(0,10)=02+102=100 -- utility at the second corner
  5. 225>100 -- the first corner is on the higher indifference curve

Answer: (x1∗,x2∗)=(15,0) -- the consumer buys 15 units of good 1 and none of good 2.

left=0; right=16; bottom=0; top=16
---
y=10-2x/3
x^2+y^2=225
x^2+y^2=100
x^2+y^2=900/13
(15,0)|label:Best corner
(0,10)|label:Other corner
(60/13,90/13)|label:Tangency

x is units of good 1 and y is units of good 2. The straight line is the budget line. The arcs are indifference curves with utility 225, 100 and about 69. The arc that just touches the line marks the tangency, which is the worst affordable bundle, not the best.

Example: Weighting good 2 more changes the answer

A student asked if the corner with the bigger quantity always wins. McLeod changed the utility to u(x1,x2)=x12+3x22 and kept the same budget line 2x1+3x2=30.

  1. (15,0) and (0,10) -- the corners do not change, because the budget line did not change
  2. u(15,0)=152+3(0)2=225 -- utility at the good 1 corner
  3. u(0,10)=02+3(10)2=3(100)=300 -- utility at the good 2 corner
  4. 300>225 -- now the good 2 corner is higher

Answer: (x1∗,x2∗)=(0,10) -- the bigger quantity does not always win, so always compare utilities.

Exam

To the student's question: "it's going to depend on the utility function... I would say always just check with the utility function to ensure that you know one of them provides the higher level of utility."

Not a Cobb-Douglas function

x12+x22 looks a lot like the Cobb-Douglas x12x22, but it adds instead of multiplies. Adding gives non-convex preferences and a corner answer. Multiplying gives Cobb-Douglas and a tangency.

Why: the plus sign changes the shape of the indifference curves completely. It turns them from bowed toward the origin into quarter circles bowed away from it, so the tangency becomes the worst bundle on the budget line.

Exam

McLeod said to put "a large asterisk" here: "let's hypothetically say on, I don't know, a midterm test there's a question like: Given this utility function, solve for the optimal consumption bundle. I guarantee you there's going to be more than one of you that's like, 'Oh, it's Cobb-Douglas. I'm going to go find a tangency.'... And that's wrong. Flat out wrong." Also: "make sure not to confuse a Cobb-Douglas utility function like x1 squared x2 squared with a non-convex preference x1 squared plus x2 squared... That's going to be critical." He circled the plus sign: "The plus is critical."

Four families and how to solve them

The goal is always the same: the affordable bundle on the highest indifference curve. How you find it depends on the family.

Why: every method ends with the same kind of result, the ordinary demand functions. Chapter 6 starts from those.

Comparative statics

Chapter 6 takes the ordinary demand functions from Chapter 5 and "plays around with them" to see what the model predicts.

Comparative statics

Changing one parameter of the model (income or one price), holding the others constant, and seeing how the optimal choice changes. There are three kinds:

Why: "It's not just building a model. It's actually seeing what the model can predict when we change some of the parameters in that model."

x1∗(p1,p2,m)x2∗(p1,p2,m)

where xi∗ is the optimal amount of good i, p1 and p2 are the prices, m is income, and p−i means the price of the other good.

Choice versus demand

A student asked how they differ. McLeod's short answer was "nothing", but the focus differs. Choice is the best bundle at one set of prices and income. Demand looks at how the quantity changes as prices change.

Why: this is building up one person's demand. Before the midterm, the course will add up individual demands into the market demand from EC 120 (for example, n identical people give n times one person's demand).

Income changes

Budget line shifts with income

When income rises with prices held constant, the budget line shifts out parallel to itself.

Why: the slope is −p1/p2, which has no m in it, so only the intercepts move. The old best bundle is no longer best, because a higher indifference curve is now affordable. The consumer re-optimizes at each income level.

Income offer curve

The line joining the best bundles as income rises (bundles A, B and C at incomes m1<m2<m3). It is also called the income expansion path. It shows the best amounts of both goods at each income.

Why: each point on it is the answer to the consumer's problem at one income level, so it traces out how the whole bundle changes as income expands.

Engel curve

A graph of the best amount of one good against income, with xi∗ on the horizontal axis and m on the vertical axis. It is named after the economist Engel. The slide graphs spell it "Engle", and even one slide title ("Income offer curves and Engle curve") does too, but the other slide titles spell it "Engel", which is correct.

Why: the income offer curve shows both goods at once, but it is often easier to see how income affects one good. Each point on the Engel curve takes the x1 amount from bundle A, B or C and pairs it with m1, m2 or m3. To write an Engel curve, take the ordinary demand function and solve it for m.

Engel curves for Cobb-Douglas

With u(x1,x2)=x1αx2β, both Engel curves are straight lines through the origin.

Why: prices are held constant and α and β are fixed preference parameters, so the bracket is a constant. A student asked about the derivative. McLeod said the derivative of m with respect to x1∗ is just that constant, so the slope never changes. It is also positive, because prices, α and β are all positive.

m=((α+β)p1α)x1∗m=((α+β)p2β)x2∗

where α and β are the exponents in the utility function, p1 and p2 are the prices, and x1∗ and x2∗ are the optimal amounts.

Example: Engel curve from Cobb-Douglas demand

Start from the Chapter 5 demand for good 1 and move m to the left-hand side.

  1. x1∗=αα+β⋅mp1 -- the ordinary demand for good 1
  2. p1x1∗=αα+βm -- multiply both sides by p1
  3. m=((α+β)p1α)x1∗ -- multiply both sides by α+βα
  4. m=((α+β)p2β)x2∗ -- the same steps on x2∗=βα+β⋅mp2

Answer: m=(constant)×xi∗ for each good -- both Engel curves are upward straight lines.

Made-up example: Engel curve with numbers

Take u(x1,x2)=x1x2, so α=β=1, and let p1=2.

  1. m=(1+1)(2)1x1∗ -- put the numbers into the Engel curve
  2. m=4x1∗ -- simplify
  3. At m=40: x1∗=40/4=10 -- the amount of good 1 at income 40
  4. At m=80: x1∗=80/4=20 -- the amount of good 1 at income 80

Answer: m=4x1∗ -- a straight line through the origin with slope 4, so doubling income doubles the amount of good 1.

Engel curves for perfect complements

With u(x1,x2)=min{x1,x2}, both Engel curves are straight lines through the origin with slope p1+p2.

Why: the best bundles all sit at the corners of the L shapes, along the ray from the origin where x1=x2. Raising income just moves the consumer further out along that ray.

m=(p1+p2)x1∗m=(p1+p2)x2∗

where p1+p2 is the cost of one unit of each good together, which stays constant while income changes.

Example: Engel curves from perfect complements demand

In Chapter 5 the consumer bought the goods one-to-one, so x1∗=x2∗=mp1+p2.

  1. x1∗=mp1+p2 -- the ordinary demand for good 1
  2. m=(p1+p2)x1∗ -- multiply both sides by p1+p2
  3. m=(p1+p2)x2∗ -- the same steps for good 2, because its demand is identical

Answer: Both Engel curves are straight lines with slope p1+p2 -- income buys the same pair of goods, just more of it.

Engel curves for perfect substitutes

With u(x1,x2)=x1+x2 and p1<p2, the Engel curve for good 1 is a straight line with slope p1. The Engel curve for good 2 is the vertical axis, x2∗=0.

Why: McLeod's example was Diet Coke (good 1) and Diet Pepsi (good 2). As long as Diet Coke is cheaper, all income goes to Diet Coke. More income never changes that, so Diet Pepsi demand stays at zero.

m=p1x1∗x2∗=0(p1<p2)

where p1 is the price of the cheaper good 1.

Example: Engel curves from perfect substitutes demand

Assume p1<p2, so the Chapter 5 demands are x1∗=m/p1 and x2∗=0.

  1. x1∗=mp1 -- all income goes to the cheaper good 1
  2. m=p1x1∗ -- multiply both sides by p1
  3. x2∗=0 for every m -- good 2 stays the more expensive good however much income rises

Answer: The good 1 Engel curve is m=p1x1∗, a straight line, and the good 2 Engel curve is a vertical line at x2∗=0.

Normal and inferior goods

Why: McLeod's inferior good was instant ramen. It was a big part of his diet as an undergrad, and he eats much less of it now that his income is higher. He also cited the Barenaked Ladies song "If I Had a Million Dollars", where the band argues over whether they would eat more or less Kraft Dinner. To him, they are arguing over whether Kraft Dinner is normal or inferior. He says he eats about the same amount now, so for him it is normal. "Inferior" is not meant as a put-down. It only describes what happens to demand as income rises.

∂xi(pi,p−i,m)∂m>0 (normal)∂xi(pi,p−i,m)∂m<0 (inferior)

where xi is the demand for good i, pi is its own price, p−i is the other price and m is income.

Made-up example: Is a Cobb-Douglas good normal?

Take u(x1,x2)=x1x2, so x1∗=12⋅mp1=m2p1, with p1=2.

  1. ∂x1∗∂m=12p1 -- the derivative of m2p1 with respect to m, holding p1 fixed
  2. 12(2)=14 -- put in p1=2
  3. 14>0 -- the sign is positive

Answer: Good 1 is normal -- each extra $1 of income buys a quarter of a unit more.

Homothetic preferences

Preferences are homothetic if scaling two bundles by the same positive amount never changes which bundle is preferred. Equivalently, the MRS (the slope of the indifference curve) is the same everywhere along a straight line from the origin.

Why: this is exactly why the Engel curves for Cobb-Douglas, perfect complements and perfect substitutes were all straight lines. Their income offer curves run along rays from the origin. For perfect complements the slope at the corner of the L is undefined, but the corners still line up on a ray.

(x1,x2)≻(y1,y2)⟺(kx1,kx2)≻(ky1,ky2) for every k>0

where ≻ means "strictly preferred to" and k is any positive scaling factor (for example, k=2 doubles both goods in both bundles).

Made-up example: Checking homothetic preferences

Take u(x1,x2)=x1x2 and compare the bundles (2,8) and (3,4), then double both with k=2.

  1. u(2,8)=2×8=16 and u(3,4)=3×4=12 -- so (2,8)≻(3,4)
  2. u(4,16)=4×16=64 and u(6,8)=6×8=48 -- after doubling, (4,16)≻(6,8)
  3. MRS=−x2/x1 -- the Cobb-Douglas MRS from Chapter 4 with equal exponents
  4. At (1,2): MRS=−2/1=−2; at (3,6): MRS=−6/3=−2 -- the same slope along the ray x2=2x1

Answer: The ranking survives scaling and the MRS is constant along the ray -- these preferences are homothetic.

Exam

On "is this true in general?": "whenever you hear questions like it's always the case or it's never the case, your spider-sense... should go off and be like, 'No, there has to be some exceptions.'" Engel curves are straight lines only for homothetic preferences.

Quasilinear preferences are not homothetic

With u(x1,x2)=f(x1)+x2, the Engel curves are not straight lines from the origin. They have a kink in them.

Why: quasilinear indifference curves are vertical shifts of each other, so the slope at a given x1 is the same on every curve. When income rises, the budget line shifts out without changing slope, so the tangency happens at the same x1∗, just higher up. At low income the consumer cannot reach that tangency yet, so extra income goes into good 1 only. Once x1∗ is reached, all extra income goes into good 2. So the Engel curve for x1 rises and then goes vertical at x1∗. The Engel curve for x2 only starts at a positive income and then slopes upward.

MRS=−f′(x1)

where f′(x1) is the slope of the f part of the utility function, so the MRS depends only on x1.

Made-up example: Quasilinear Engel curves

Take u(x1,x2)=4x1+x2 with p1=p2=1.

  1. f′(x1)=4⋅12x1−1/2=2x1 -- power rule, so MRS=−2x1
  2. 2x1=p1p2=1 -- tangency: the slope of the indifference curve equals the slope of the budget line
  3. x1=2, so x1∗=4 -- note that m never appears
  4. p1x1∗=1×4=4 -- the tangency costs 4, so for m≥4, x2∗=(m−4)/1=m−4
  5. m=10 gives (4,6) and m=20 gives (4,16) -- extra income goes only to good 2
  6. m=3 gives x1∗=3/1=3 and x2∗=0 -- below 4 the tangency is out of reach, so all income goes to good 1

Answer: The Engel curve for x1 is m=x1∗ up to m=4, then vertical at x1∗=4. The Engel curve for x2 is zero up to m=4, then x2∗=m−4.

Luxury and necessary goods

Why: McLeod raised this split for non-homothetic preferences. With homothetic preferences, every good's demand grows in exactly the same proportion as income (a straight-line Engel curve through the origin), so the split only shows up when preferences are not homothetic.

Made-up example: Luxury or necessary?

Income rises from $1,000 to $1,100. Demand for good A goes from 4 to 5 units, and demand for good B goes from 10 to 10.5 units.

  1. 1100−10001000=1001000=10% -- the proportional rise in income
  2. 5−44=14=25% -- the proportional rise in demand for A
  3. 10.5−1010=0.510=5% -- the proportional rise in demand for B
  4. 25%>10% and 5%<10% -- compare each good with income

Answer: A is a luxury good and B is a necessary good -- both are normal, but A grows faster than income and B grows slower.

Own-price changes

This topic does the same exercise as income changes, but changes p1 instead of m. McLeod stopped at the ordinary demand curve.

Budget line rotates with own price

When p1 rises (from p1′ to p1″ to p1‴, where more primes mean a higher price), the budget line rotates inward around its vertical intercept and gets steeper.

Why: the slope −p1/p2 changes, but the vertical intercept m/p2 is the most good 2 you can buy with all your income, and that does not depend on p1. Only the horizontal intercept m/p1 moves in. At each price the old best bundle becomes unaffordable, so the consumer re-optimizes, and x1∗(p1′)>x1∗(p1″)>x1∗(p1‴).

Price offer curve

The line joining all the utility-maximizing bundles traced out as p1 changes, holding p2 and m constant (bundles A, B and C).

Why: it does for price what the income offer curve does for income. It shows how the whole bundle moves as one price changes.

Ordinary demand curve

The plot of the x1 amount on the price offer curve against p1, with x1 on the horizontal axis and p1 on the vertical axis.

Why: each bundle on the price offer curve gives one point on the demand curve. For example, bundle A gives the amount of good 1 demanded at price p1′. As p1 rises, the quantity of good 1 falls, so the curve slopes down.

Made-up example: Demand points from Cobb-Douglas

Take u(x1,x2)=x1x2, so x1∗=m2p1, with m=60. Try p1′=1, p1″=2 and p1‴=3.

  1. x1∗(1)=602(1)=30 -- the point (30,1)
  2. x1∗(2)=602(2)=15 -- the point (15,2)
  3. x1∗(3)=602(3)=10 -- the point (10,3)

Answer: Quantity falls from 30 to 15 to 10 as the price rises -- these points trace out a downward-sloping ordinary demand curve.

Exam checklist

Admin & deadlines


Transcribed automatically from the lecture recording. Slides used: EC 270 - Ch. 06 - Demand (slides).pdf. Course info used: EC 270 - Math Appendix (slides).pdf, EC 270 - Rules of Derivatives.pdf, EC 270 - Syllabus - Fall 2026.pdf. Audio archived at /mnt/porsche/configs/lectures/archive/2026-10/EC 270 - 2026-10-01 - Lecture 07.opus.