EC 270 - 2026-10-08 - Lecture 09

EC 270 - Lecture 09 - 2026-10-08

Intermediate Microeconomics I · Logan McLeod · LH 1011

Week 5 · 11:30-12:23 · 52 min · Full transcript · Flashcards

Overview

This lecture finished Chapter 8: Slutsky Equation. It picked up from Tuesday's substitution effect and compensated demand. McLeod worked the milk example to calculate a substitution effect, then defined and calculated the income effect. He put the two together as the Slutsky identity, signed it for normal and inferior goods, and used it to explain the law of demand and Giffen goods. He finished with the Hicks substitution effect, the three kinds of demand curve and the Slutsky equation, and stopped before Chapter 15.

Calculating the substitution effect

On Tuesday the substitution effect was defined as the move from bundle X to bundle Y on the pivoted budget line. Today it gets calculated with numbers.

The substitution effect in numbers

The substitution effect is the change in demand for good 1 when its price changes and income is adjusted to m′, so that purchasing power stays the same. You get it by evaluating the ordinary demand function twice: at the new price with income m′, and at the old price with income m.

Why: Adjusting income to m′ removes the purchasing-power part of the price change. What is left is only the switch toward the good that is now relatively cheaper. For a price fall, compensating means income is taken away until the original bundle is just affordable at the new lower price.

Δx1s=x1(p1′,m′)−x1(p1,m)withm′=m+x1Δp1

where x1(p,m) is the ordinary demand function, p1 and p1′ are the old and new prices, m is the original income, m′ is the income that holds purchasing power constant, and x1 in the second formula is the original quantity demanded.

Example: Substitution effect for milk

Demand for milk is x1=10+m10p1. Income is $120, and the price of milk falls from $3 to $2.

  1. x1(3,120)=10+12010×3=10+12030=10+4=14 -- original demand (bundle X)
  2. x1(2,120)=10+12010×2=10+12020=10+6=16 -- demand after the price fall with the original income (bundle Z); this is needed later, but it is not part of the SE
  3. Δm=x1Δp1=14×(2−3)=14×(−1)=−14 -- the income change that keeps bundle X just affordable; it is negative because a price fall raises purchasing power
  4. m′=m+Δm=120−14=106 -- the income on the pivoted budget line
  5. x1(2,106)=10+10610×2=10+10620=10+5.3=15.3 -- demand on the pivoted budget line (bundle Y)
  6. Δx1s=x1(2,106)−x1(3,120)=15.3−14=1.3 -- bundle Y minus bundle X

Answer: Δx1s=1.3 units. Holding purchasing power constant, the cheaper milk alone makes the consumer buy 1.3 more units.

Exam

In class, a student first answered x1(2,120)−x1(3,120)=16−14=2. McLeod: "this is not our substitution effect. This is actually our total effect." The SE needs the demand at the new price with income m′. In his summary of the chapter, he called the milk example a good example of how to calculate these effects: "We need to figure out what quantity demanded is at the new price on that pivoted budget line holding purchasing power constant."

The income effect

The income effect is the second half of the decomposition: the move from bundle Y on the pivoted line to bundle Z on the final budget line.

The income effect

The income effect is the change in demand for good 1 when income moves from m′ back to m, with prices held at the new level. On the graph, it is the parallel shift out from the pivoted budget line to the final one (bundle Y to bundle Z).

Why: A price fall makes your income go further, which is like getting more income. The income effect measures how much of the change in demand comes from that extra purchasing power alone.

Δx1n=x1(p1′,m)−x1(p1′,m′)

where x1(p1′,m) is the demand at the new price with the original income (bundle Z), and x1(p1′,m′) is the demand at the new price with compensated income (bundle Y).

Example: Income effect for milk

Same setup as before: x1=10+m10p1, m=120, and the price falls from $3 to $2, so m′=106.

  1. Δx1n=x1(2,120)−x1(2,106) -- final demand minus demand on the pivoted line
  2. x1(2,120)=10+12020=16 and x1(2,106)=10+10620=15.3 -- evaluate the demand function at both points
  3. Δx1n=16−15.3=0.7 -- subtract

Answer: Δx1n=0.7 units. The rise in purchasing power from the cheaper milk adds 0.7 units of milk.

Exam

Do NOT get the income effect as "total effect minus substitution effect" (2 − 1.3), even though it gives the same number. McLeod raised this question himself, saying it usually comes up ("can we just calculate the income effect this way"), and answered: "the short answer is no, because I want on a midterm test for you to demonstrate your understanding of the model." He also said that if you already made an error in your SE or total effect, "your income effects can be wrong." Show x1(p1′,m)−x1(p1′,m′): "We're evaluating your ordinary demand functions." He then added: "assuming there'd be a question like that on the midterm, hypothetically."

Income effect by type of good

Whether the good is normal or inferior decides which way the income effect goes. This uses normal and inferior goods from Chapter 6.

Why: For a price fall, holding purchasing power means taking income away (m′<m). The income effect undoes that, raising income from m′ back to m.

Watch the sign words. McLeod and the slides call a normal good's income effect "negative". That means the quantity moves the opposite way to the price, not that the number is below zero. The slide's "Δx1n<0" for a normal good describes the compensation step x1(p1′,m′)−x1(p1′,m), which for milk is 15.3−16=−0.7. In the milk example itself, Δx1n=+0.7.

The Slutsky identity

The two effects add up to the whole change in demand, and that sum can be written in two ways.

The total effect

The total effect is the whole change in demand for good 1 after the price change, with income held at m. It goes from the old best bundle X to the new best bundle Z.

Why: These are the only two bundles you actually see in the real world. Everything in between is built by the model.

Δx1=x1(p1′,m)−x1(p1,m)

where both demands use the original income m, one at the new price and one at the old price.

In the milk example, the total effect was 16−14=2 units (step 2 of the substitution effect example).

The Slutsky identity

The total effect always equals the substitution effect plus the income effect. Written out in full, the bracketed terms are the SE and the IE.

Why: Y is just a stopping point between X and Z. Going from X to Y and then from Y to Z has to add up to going from X to Z, whatever the numbers are. That is why it holds "for all values of p1, p1′, m, and m′".

Δx1=[x1(p1′,m′)−x1(p1,m)]⏟Δx1s+[x1(p1′,m)−x1(p1′,m′)]⏟Δx1n

where the first bracket is the substitution effect (X to Y) and the second is the income effect (Y to Z).

Example: Checking the milk numbers

The milk example gave a total effect of 2, a substitution effect of 1.3 and an income effect of 0.7.

  1. Δx1s+Δx1n=1.3+0.7 -- add the two effects
  2. =2.0 -- the sum

Answer: 2=2. The sum equals the total effect, so the decomposition checks out.

Signing the identity

"Signing" means saying whether quantity demanded and price move in the same direction or opposite directions for each effect.

Why: The substitution effect always moves opposite to the price (cheaper good, buy more). Only the income effect can switch direction, depending on the type of good. So the good's type decides whether the total effect can be signed in advance.

Type of good Total effect Δx1 SE Δx1s IE Δx1n
Normal (−) (−) (−)
Inferior (?) (−) (+)

Here (−) means quantity and price move in opposite directions, and (+) means they move in the same direction. For an inferior good, the total effect is "?" because you do not know in advance whether the SE or the IE is bigger.

The identity as rates of change

The same identity can be written as changes per dollar of price change instead of absolute changes. To do this, first define Δx1m as minus the income effect, then divide by Δp1, and finally use Δm=x1Δp1 to turn the income term into a change per dollar of income.

Why: Dividing the income term by Δp1 looks odd, because the income effect is really about a change in income. The substitution Δp1=Δmx1 puts Δm under that term, which is easier to interpret.

Δx1Δp1=Δx1sΔp1−Δx1mΔmx1withΔx1m=x1(p1′,m′)−x1(p1′,m)=−Δx1n

where Δx1Δp1 is the total effect (price changes, income held at m), Δx1sΔp1 is the substitution effect (price changes, purchasing power held constant), and Δx1mΔmx1 is the income effect (income adjusted, prices held constant).

Made-up example: Milk as rates of change

Use the class milk numbers: price falls from $3 to $2, so Δp1=−1; m=120, m′=106, x1=14, and demands are 14, 15.3 and 16.

  1. Δx1Δp1=16−142−3=2−1=−2 -- total effect per dollar of price change
  2. Δx1sΔp1=15.3−14−1=1.3−1=−1.3 -- substitution effect per dollar
  3. Δx1m=x1(2,106)−x1(2,120)=15.3−16=−0.7 -- minus the income effect
  4. Δm=m′−m=106−120=−14 -- the income change
  5. Δx1mΔmx1=−0.7−14×14=0.05×14=0.7 -- the income term
  6. −1.3−0.7=−2 -- SE minus the income term

Answer: The right side is −2, which matches the total effect from step 1. Each $1 fall in price raises milk demand by 2 units: 1.3 from substitution and 0.7 from income.

Two parts of the income effect

The income term Δx1mΔmx1 is two pieces multiplied together. The first is how demand changes as income changes, Δx1mΔm. The second is the original level of demand, x1.

Why: A price change shifts your purchasing power more when you buy a lot of the good, because Δm=x1Δp1. So the same income response matters more for a big purchase. In the milk example above, the pieces are 0.05 and 14.

The law of demand and Giffen goods

The signing table explains both why demand curves usually slope down and when they do not.

The law of demand

If demand for a good rises when income rises, then demand for that good must fall when its price rises.

Why: This follows directly from the Slutsky identity. A good whose demand rises with income is a normal good. For a normal good, the SE and the IE reinforce each other, so a price rise unambiguously lowers quantity demanded. That gives a downward-sloping demand curve. Most goods are normal, and every graph so far in Chapter 8 was drawn for a normal good.

Inferior goods

For an inferior good, the substitution effect and the income effect move in opposite directions. In the slide graph for a price fall, the SE pushes demand right and the IE pushes it left. The total effect is still rightward there, but smaller than the SE.

Why: Taking income away (or giving it back) moves demand for an inferior good the "wrong" way. This is why the total effect for an inferior good cannot be signed in advance. It depends on which effect is bigger.

Giffen goods

A Giffen good is an extremely income-inferior good where the income effect is bigger than the substitution effect (|IE|>|SE|). Quantity demanded then moves in the same direction as the price: it falls when the price falls and rises when the price rises. (Slide 28's wording, "fall as own-price rises", reads the opposite way. McLeod's spoken version, quantity falling as the price falls, is the correct one.)

Why: McLeod's point is that Slutsky's decomposition explains both the law of downward-sloping demand and its violations. A Giffen good is possible only when the IE runs against the SE and outweighs it. His one real-world example, from Tuesday's class, was potatoes in the Irish potato famine: as the price of potatoes rose, people demanded more of them. He called it "a rare rare case".

The Hicks substitution effect and demand curves

So far the substitution effect has held purchasing power constant (the Slutsky version). Hicks defined it by holding utility constant instead.

The Hicks substitution effect

The Hicks substitution effect is the change in demand when prices change, holding utility constant. Instead of pivoting the budget line through bundle X, you "roll" the new, flatter budget line along the original indifference curve until it is tangent at a new bundle Yh. The move from X to Yh is the Hicks SE. The income effect is then the move from Yh to Z. In the slides' notation, the Slutsky SE is ∂xs(p1,p2,x¯1,x¯2)∂p1 and the Hicks SE is ∂xh(p1,p2,u¯)∂p1.

Why: Every bundle on one indifference curve gives the same utility. So staying on the curve through X holds the consumer exactly as well off as before, instead of holding their old bundle just affordable.

The total effect is the same

Slutsky and Hicks give different intermediate bundles, but the total effect from X to Z is identical.

Why: X and Z are what we actually observe: in the milk example, 14 units before and 16 after. The intermediate bundle (Y or Yh) is never observed. It is a theoretical construction, and the two decompositions are just two ways to place it. "The power of the model is really thinking about where this middle bundle is."

Exam

McLeod: "I think this is might be an important point to to flag." He asked the class whether demand before and after the price change would differ under a different decomposition. His answer was that it does not: "the total effect remains unchanged."

Three demand curves

The class has now looked at how quantity demanded responds to price in three settings, and each gives its own demand curve.

The Slutsky equation

The Slutsky equation is the Slutsky identity in rates of change, written with partial derivatives instead of triangles. It says the same thing: the effect of a price change on demand splits into a substitution effect minus an income term. Using the Hicks substitution effect gives an equivalent version.

Why: McLeod showed it "really more just for fun". As you go further in economics, you can say you have seen the Slutsky equation.

∂x1(p1,p2,m¯)∂p1=∂x1s(p1,p2,x¯1,x¯2)∂p1−∂x1m(p1,p2,m¯)∂mx¯1

where the left side is the total effect, the first right-hand term is the Slutsky substitution effect (this can be swapped for the Hicks term ∂x1h(p1,p2,u¯)∂p1), the last term is the income effect, and x¯1 is the original quantity.

Made-up example: Slutsky equation for milk at the starting point

Take x1=10+m10p1 at p1=3 and m=120, where x¯1=14. Find the substitution term.

  1. ∂x1∂p1=−m10p12=−12010×9=−12090=−43 -- total effect, by the power rule
  2. ∂x1∂m=110p1=130 -- how demand responds to income
  3. ∂x1∂mx¯1=130×14=1430=715 -- the income term
  4. SE=−43+715=−2015+715=−1315≈−0.87 -- rearrange the equation: SE = total effect + income term

Answer: The substitution term is about −0.87 units per dollar at p1=3. This is the slope at one price, so it is not equal to the class's −1.3, which came from a full $1 price drop.

Exam

Not on the midterm. McLeod: "for a midterm test in, you know, the end of October, you don't have to memorize that or use it or anything like that." He also said: "I'm not going to ask you to calculate substitution effects using partial derivatives. So the milk example is the way we're going to calculate those."

Exam checklist

Admin & deadlines


Transcribed automatically from the lecture recording. Slides used: EC 270 - Ch. 08 - Slutsky Equation (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 - 2026-10-08 - Lecture 09.opus.