Each objective states what you will be able to do after completing this chapter.
The map below shows how all topics in this chapter connect. Each node links to the section where it is developed in detail.
| Symbol | Meaning | Symbol | Meaning |
|---|---|---|---|
| X, Y | Quantities of goods X and Y | U(X,Y) | Utility function |
| Px, Py | Prices of goods X and Y | MUx | Marginal utility of X |
| I | Consumer income | MUy | Marginal utility of Y |
| X*, Y* | Optimal quantities | MRS | Marginal rate of substitution |
| IC | Indifference curve | BC | Budget constraint |
In the previous week, we introduced the concept of consumer preferences and indifference curves. To briefly recap: a consumer who faces two goods, X and Y, can be described by an indifference curve — a set of all bundles (X,Y) that provide the consumer with exactly the same level of utility (satisfaction). The consumer is, by definition, indifferent among all bundles on the same curve.
Several key properties hold for well-behaved preferences:
Indifference curves are the economist's way of representing preferences without asking people to assign dollar values to their satisfaction. Businesses use similar logic — through market research and conjoint analysis — to understand how customers trade off product features such as price, quality, and convenience. The same framework that describes a student choosing between food and clothing underlies how airlines design fare classes and how phone manufacturers bundle features.
The standard convex indifference curve is the right shape for a typical consumer who values both goods and prefers variety. But economics deals with many kinds of goods, and the shape of the indifference curve changes depending on the relationship between them. We examine four important cases.
Consider Maya, a college student with $60 per week to spend on coffee and textbooks. She values both, but they are neither perfect substitutes (she cannot read a cup of coffee) nor perfect complements (she can enjoy coffee without a textbook in hand). Her preferences are well-behaved: she prefers more of each, she prefers variety, and her willingness to give up coffee for an extra textbook declines as she already has many textbooks. Her indifference curves are the standard convex shape — downward sloping and bowing toward the origin. The concepts in this section explain how Maya's optimal choice changes if her budget grows or if textbook prices change.
| Good Type | IC Shape | MRS |
|---|---|---|
| Standard (both goods desired) | Downward-sloping, convex | Diminishing |
| Perfect substitutes | Straight line, slope −a/b | Constant = a/b |
| Perfect complements | L-shaped (kink on ratio ray) | 0, ∞, or undefined |
| Neutral good (Y useless) | Vertical line | ∞ |
| Economic bad (Y bad) | Upward sloping | Negative |
Perfect Substitutes: Two goods are perfect substitutes if the consumer regards them as identical — they serve the same function perfectly and the consumer is always willing to trade one for the other at a constant ratio.
The classic example is gasoline from two neighboring stations: Exxon gas and Mobil gas. For most consumers, a gallon of gasoline is a gallon of gasoline, regardless of brand.
Consider a consumer starting at bundle A = (2 Mobil, 2 Exxon) — two gallons from each station, four gallons total.
Because the consumer only cares about total quantity, all these bundles lie on the same IC. The willingness to trade is constant — always one-for-one. This constant trade-off rate means the IC is a straight line with slope −1. More generally, if the consumer trades a units of Y for b units of X, the IC has slope −a/b.
The utility function for perfect substitutes takes the linear form:
Because the consumer treats the two goods as identical, no amount of X or Y already consumed changes how willing they are to swap. The MRS is fixed. This is why the IC is a straight line rather than the bowed shape we see for ordinary goods.
Key feature: The ICs are straight, parallel, downward-sloping lines. The MRS is constant — the consumer is always willing to trade the two goods at the same fixed rate, regardless of how much of each they already have.
Retailers face perfect-substitute logic every time they launch a private-label product. If consumers genuinely view a store-brand paper towel as identical to the name-brand version, the IC between the two is a straight line with slope −1. The consumer will always choose whichever is cheaper — the budget line will be tangent to the IC at a corner, buying only the cheaper brand.
This is why Procter & Gamble and Unilever invest heavily in advertising: the goal is not just awareness but preference differentiation — shifting consumers away from a linear IC toward a bowed one, so that they prefer a mix rather than the cheapest-only corner solution.
Perfect Complements: Two goods are perfect complements if they must be consumed together in fixed proportions in order to generate utility. Extra units of one good, without the corresponding extra units of the other, provide no additional utility.
The canonical example is left shoes and right shoes. A consumer with 2 left shoes and 2 right shoes has 2 functional pairs.
The IC through A = (2,2) is L-shaped: a vertical arm at X = 2 and a horizontal arm at Y = 2. The kink is at (2,2). More generally, for goods consumed in fixed proportions a:b:
For equal proportions (a = b = 1): U = min(X,Y).
The kink of the L is the only point on each IC where both goods are being "used." Anywhere else on the L, one good is in excess. Extra units of the surplus good add no utility.
Key feature: L-shaped ICs with the kink on the ray X/a = Y/b. The MRS is not defined at the kink, and is either 0 (horizontal arm) or ∞ (vertical arm).
Perfect-complement preferences underlie one of the most successful pricing strategies in business history. Gillette sells razors cheaply (near cost) and charges a premium on replacement blades. Printer companies do the same with ink cartridges. Apple's ecosystem — iPhone and App Store — exhibits the same logic. When two goods are perfect complements, the consumer must buy them in a fixed ratio. A firm that controls supply of one complement can therefore extract value through the other. This is the foundation for understanding why firms in technology, gaming (consoles and games), and healthcare (devices and consumables) adopt two-part pricing structures.
Neutral Good: A good is neutral if it provides zero marginal utility to the consumer — consuming more or less of it leaves utility completely unchanged. Examples include class notes for a course you have never taken, or a book in a language you do not speak.
Suppose good Y is neutral and good X is a normal good. Starting at bundle A = (2,2):
The IC through A is therefore a vertical line at X = 2. Utility only changes when X changes, so utility increases as you move horizontally to the right.
Economic Bad: A good is an economic bad if consuming more of it reduces the consumer's utility. It confers negative marginal utility — also called disutility. Classic examples: pollution, noise, commute time.
Suppose good Y is an economic bad (e.g., air pollution) and good X is a normal good (e.g., food). Starting at bundle A = (2,2):
The IC through A is therefore upward sloping. Higher utility is reached by moving right (more X) or downward (less of the bad Y).
Economic bads appear constantly in policy debates. When a government evaluates a highway expansion, residents near the road experience noise and pollution as economic bads. To keep utility constant, they need compensation — more of some other good (income, parks, soundproofing). Understanding ICs for bads helps economists design sensible compensation policies and value environmental amenities.
Indifference curves describe what the consumer wants. But wanting something is not enough — the consumer also faces a resource constraint. This is where the budget constraint (BC) comes in.
Suppose a consumer has income I and faces prices Px for good X and Py for good Y. If the consumer spends all of their income:
Rearranging so that Y is on the left gives us the slope-intercept form:
From this form, we can read off the key properties:
The budget constraint separates the feasible from the infeasible. Any bundle on or below the line is affordable; any bundle above the line requires more money than the consumer has. The line itself represents all the ways the consumer can spend every last dollar.
Xmax = I/Px = 100/5 = 20 Ymax = I/Py = 100/2 = 50
100 = 5X + 2Y
Slope = −Px/Py = −5/2 = −2.5
In words: For every additional unit of food Sheela buys, she must give up 2.5 units of clothing.
Plot (0, 50) on the Y-axis and (20, 0) on the X-axis; connect them with a straight line.
Suppose Sheela's income rises to $150 but prices stay the same. Without calculating, predict what happens to the budget line. Then verify by computing the new intercepts.
Answer: both intercepts scale by 1.5; slope unchanged.
When income increases from $100 to $150 while Px = 5 and Py = 2 remain the same:
Both intercepts increase proportionally and the slope is unchanged. The budget line shifts outward in parallel.
Income change ⇒ parallel shift of budget line. The slope is set by prices alone, so as long as prices do not change, a higher income simply slides the entire line outward. Income up: BC shifts outward. Income down: BC shifts inward.
When Px decreases from $5 to $4, while income = $100 and Py = $2 remain the same:
The line pivots outward around the fixed Y-intercept. The X-intercept moves outward; the slope changes.
Own-price change ⇒ pivot of budget line. The intercept on the other good's axis is fixed; only the intercept on the good whose price changed moves. Px falls: pivot outward (flatter). Px rises: pivot inward (steeper).
Students often draw a parallel shift when a price changes, or a pivot when income changes. The rule is simple: income changes both intercepts proportionally and leaves the slope unchanged (parallel shift); a price change fixes one intercept and moves the other, changing the slope (pivot). Quick check: if the fixed intercept has moved, it is not a pivot.
| Change | Y-intercept | X-intercept | Slope | Type |
|---|---|---|---|---|
| Income ↑ | Moves out | Moves out | Fixed | Parallel shift out |
| Income ↓ | Moves in | Moves in | Fixed | Parallel shift in |
| Px ↓ | Fixed | Moves out | Flatter | Pivot out |
| Px ↑ | Fixed | Moves in | Steeper | Pivot in |
| Py ↓ | Moves out | Fixed | Steeper | Pivot out |
| Py ↑ | Moves in | Fixed | Flatter | Pivot in |
| All prices ×k | Moves in | Moves in | Fixed | ≡ income ÷k |
Suppose the government needs to raise $500 per household in revenue. It can do so either by levying a tax on petrol (pivoting the budget line inward around the non-petrol intercept) or by reducing all households' income by $500 (a parallel shift inward). Both policies raise the same revenue. But they affect consumer choices very differently: the fuel tax changes relative prices and discourages petrol consumption specifically; the income tax leaves relative prices intact and reduces consumption of all goods proportionally.
This is why environmental economists prefer fuel taxes to income taxes when they want to reduce emissions — the pivot, not the shift, does the work.
We now have two pieces: (1) indifference curves, which describe what the consumer wants, and (2) the budget constraint, which describes what they can afford. The consumer's goal is:
Choose the bundle that provides the highest utility while remaining within the budget.
Why interior points are not optimal. Consider a bundle A that lies strictly inside the budget constraint. The consumer is spending less than their income — some income is unspent. But since both goods are desirable, spending that remaining income on more X or Y would push the consumer to a higher IC. Therefore, the optimal bundle must lie on the budget line.
Why crossing is not optimal. Consider bundle B on the budget line, where an IC crosses (rather than is tangent to) the budget line. At a crossing, between the two intersection points there are affordable bundles that lie above the IC through B. The consumer can do better than B by moving along the budget line. Therefore, B is not optimal.
The tangency point is optimal. Point C is where the budget line is tangent to an IC. At this point: (1) the consumer is spending all income, and (2) there is no affordable bundle on a higher IC. Point C is the optimal consumption bundle.
From the geometric analysis, two conditions must hold at the optimal bundle:
In words: the market's trade-off between the two goods (the price ratio) must equal the consumer's personal trade-off (the MRS). If they differ, the consumer can always rearrange spending to reach a higher IC.
The tangency condition says: at the optimum, the market's exchange rate between X and Y (the price ratio) exactly equals the consumer's personal exchange rate (the MRS). If they differed, the consumer could rearrange spending to get more utility without spending more money.
The tangency condition holds only for smooth, strictly convex ICs (like Cobb-Douglas). For perfect complements, the optimum is at the kink of the L, where the MRS is undefined. For perfect substitutes, the optimum is typically at a corner. Always check the type of preferences before applying the tangency condition mechanically.
The optimality condition can be rearranged to yield a very intuitive interpretation. Rearranging Px/Py = MUx/MUy:
What does each side mean?
The optimality condition says: at the optimal bundle, the utility per dollar spent is the same for every good. If this were not so — say, MUx/Px > MUy/Py — the consumer could increase total utility by spending one less dollar on Y and one more dollar on X.
The bang-for-the-buck condition applies far beyond individual consumers. A firm deciding how to allocate a marketing budget across TV and digital advertising should equalize the return per dollar across channels. A hospital system allocating resources across departments, or a government allocating spending across programs, faces the same logic. Wherever resources are scarce and alternatives exist, equating marginal return per dollar is the path to efficiency.
A coffee shop owner has $10,000 to split between espresso machine upgrades and barista training. The last $1,000 spent on machines increases daily throughput worth $80. The last $1,000 spent on training improves customer satisfaction, increasing repeat visits worth $130 per $1,000 over a quarter. Since the return per dollar is higher for training ($130) than for machines ($80), the owner should shift spending toward training — and keep reallocating until marginal returns per dollar are equalized. This is the bang-for-the-buck condition applied directly to a capital budgeting decision.
Amanda has income I = $200. Px = $4 (food), Py = $5 (clothing). Utility: U(X,Y) = √(XY) = X1/2Y1/2.
200 = 4X + 5Y Slope = −Px/Py = −4/5
MUx = ½ X−1/2Y1/2 = ½√(Y/X) MUy = ½ X1/2Y−1/2 = ½√(X/Y)
MRS = MUx/MUy = Y/X (the slope of the IC at any bundle)
Condition 1 (budget): 200 = 4X + 5Y (i)
Condition 2 (tangency): 4/5 = Y/X ⇒ Y = 4X/5 (ii)
Substitute (ii) into (i): 200 = 4X + 5·(4X/5) = 4X + 4X = 8X ⇒ X* = 25
Y* = 4(25)/5 = 20
Verification: 4(25) + 5(20) = 100 + 100 = 200 = I ✓
New budget: 200 = 2X + 5Y. New tangency: 2/5 = Y/X ⇒ Y = 2X/5
200 = 2X + 5·(2X/5) = 2X + 2X = 4X ⇒ X* = 50, Y* = 20
Observation: When the price of food fell from $4 to $2, Amanda doubled her food consumption (25 → 50). Her clothing consumption stayed at 20. This reflects the equal-expenditure-share property of √(XY): exactly half of income is always spent on each good.
When two goods are consumed under Cobb-Douglas preferences, a fall in the price of one good raises its consumption but leaves consumption of the other good completely unchanged. The freed-up spending goes entirely into the cheaper good — a result that turns out to be useful for understanding demand elasticities in Chapter 3.
Generalising: U(X,Y) = XαYβ with budget PxX + PyY = I.
MUx = αXα−1Yβ MUy = βXαYβ−1
Tangency: MUx/MUy = Px/Py ⇒ αY/(βX) = Px/Py ⇒ Y = (βPx)/(αPy) · X
Substituting into budget constraint and solving:
Interpretation: The consumer spends fraction α/(α+β) of income on X and fraction β/(α+β) on Y. For Amanda (α = β = 1/2): X* = I/(2Px), Y* = I/(2Py) — exactly half of income on each.
Your one-page revision guide — everything you need before an exam or tutorial.
| Name | Formula |
|---|---|
| Budget constraint | PxX + PyY = I |
| Slope of BC | −Px/Py |
| Tangency condition | Px/Py = MUx/MUy |
| Bang for the buck | MUx/Px = MUy/Py |
| Cobb-Douglas demand | X* = [α/(α+β)] · I/Px |
| Preference | IC Shape | MRS |
|---|---|---|
| Standard | Convex, downward | Diminishing |
| Substitutes | Straight line | Constant |
| Complements | L-shaped | 0/∞/undef. |
| Neutral Y | Vertical | ∞ |
| Bad Y | Upward sloping | Negative |
Watch in order for the best results. Each video is directly aligned to the sections listed.
All 27 questions with full solutions — Level 1, 2 & 3