Here is the move that trips people up: you compute the slope of a straight-line demand curve, find it's constant (say −1 everywhere), and conclude elasticity must be constant too — so when a problem says the same line is elastic at the top and inelastic at the bottom, you decide one of those answers has to be a mistake. It's a tempting deduction. The slope genuinely is fixed, and slope sure looks like "how much quantity responds to price," which is exactly what elasticity sounds like. So the same number ought to come out everywhere, right?
No — elasticity is not the slope. A straight-line demand curve has one slope everywhere, but its elasticity changes at every point: elastic near the top, unit-elastic in the middle, inelastic near the bottom. That isn't a contradiction. Slope and elasticity measure two different things. Slope is the ratio of unit changes in price and quantity; elasticity is the ratio of percentage changes, and percentages depend on the price and quantity you happen to be standing on. One straight line keeps the slope fixed and lets the starting point — and therefore the elasticity — drift.
The recurring Quora question — "if the slope is constant, why does elasticity change?" — comes from people who did everything right and still got blindsided by a problem that called the same line elastic up top and inelastic down low.
Why it's so tempting to call them the same thing
Both slope and elasticity answer a question that sounds identical: how much does quantity respond when price moves? Both are computed as a ratio of a change in P to a change in Q. On a graph they even look like the same object — the tilt of the line. So it's reasonable to assume that a steeper line is "more inelastic" and a flatter line is "more elastic," everywhere, full stop.
The detail almost nobody emphasizes: slope uses raw unit changes, and elasticity uses percentage changes. That one word — percentage — is the whole story. A 1-dollar price change is a big percentage when the price is 2 dollars and a tiny percentage when the price is 90 dollars. Slope doesn't care; it treats every dollar the same. Elasticity does care, because it divides by where you started. (If raw-versus-percentage feels like a recurring source of pain in econ, it's the same fault line behind why you subtract for real GDP but divide by the deflator — two operations that get blurred because they answer similar-sounding questions.)
The actual definitions, side by side
Slope of the demand curve (drawn the usual way, with P on the vertical axis):
$$\text{slope} = \frac{\Delta P}{\Delta Q}$$
This is a fixed number for a straight line. It's measured in dollars per unit. It never changes as you slide along the line.
Price elasticity of demand:
$$E = \frac{\%\Delta Q}{\%\Delta P} = \frac{\Delta Q / Q}{\Delta P / P} = \frac{\Delta Q}{\Delta P} \cdot \frac{P}{Q}$$
Look at the last form. It's the reciprocal of the slope (ΔQ/ΔP) multiplied by the ratio P/Q. The ΔQ/ΔP part is constant on a straight line. The P/Q part is not. As you move down and to the right along the curve, P falls and Q rises, so P/Q shrinks continuously. That shrinking term is what drags elasticity down from "large" at the top to "small" at the bottom.
One straight line, three different elasticities
Take a clean demand curve:
$$Q = 100 - P$$
Rearranged as P = 100 − Q, this is a straight line with a constant slope. In terms of quantity responding to price, dQ/dP = −1 everywhere. There is exactly one slope on this entire line, and it never budges.
Now compute point elasticity, E = (dQ/dP)(P/Q), at three places.
Near the top — P = 80, Q = 20:
$$E = (-1)\cdot\frac{80}{20} = (-1)(4) = -4$$
The magnitude is 4, well above 1, so demand here is elastic. A small percentage cut in this high price brings a large percentage jump in this small quantity.
At the midpoint — P = 50, Q = 50:
$$E = (-1)\cdot\frac{50}{50} = (-1)(1) = -1$$
Magnitude 1: unit-elastic. The percentage changes in price and quantity match exactly. The midpoint of a linear demand curve is always the unit-elastic point.
Near the bottom — P = 20, Q = 80:
$$E = (-1)\cdot\frac{20}{80} = (-1)(0.25) = -0.25$$
Magnitude 0.25, below 1: inelastic. The price is already low and the quantity is already high, so a percentage move in price barely registers as a percentage move in quantity.
Same line. Same slope (−1) at all three points. Elasticity of −4, then −1, then −0.25. The slope term carried no information that distinguished these. The P/Q term did all the work: 4, then 1, then 0.25. Trace those numbers and you can see exactly where the change came from.
So when does "steeper" actually mean "more inelastic"?
The steepness shortcut isn't useless — it's just conditional. Steepness only tells you about relative elasticity when you compare two different curves that pass through the same point (same P and same Q). At a shared point, P/Q is identical for both, so the only thing left to differ is the slope. There, the steeper curve really is the more inelastic one.
The error is exporting that comparison to a single line. Within one curve, you're not holding P/Q fixed — you're changing it with every step. The slope stops being a reliable read on elasticity the moment you move along the curve instead of across curves.
The one-sentence version to keep
Slope is ΔP/ΔQ in raw dollars and units, and it's frozen on a straight line. Elasticity is (ΔQ/ΔP)(P/Q), and the P/Q rides along with you, so the same straight line is elastic above its midpoint, unit-elastic at the midpoint, and inelastic below. Catching yourself the moment you carry "slope is constant" into "so elasticity is constant" is the same discipline that keeps other intro-econ rules honest — the kind of careful flip-the-ratio reasoning that also separates which good a country should specialize in under comparative advantage.
Check yourself
For the demand curve Q = 100 − P, you're told two points: A at P = 90 (Q = 10) and B at P = 10 (Q = 90). Which point is more elastic, and why?
A) Point B — it's lower on the curve, and lower prices mean more responsive buyers.
B) Neither — the line is straight, so the slope is constant and elasticity is the same everywhere.
C) Point A — P/Q is 90/10 = 9 there versus 10/90 ≈ 0.11 at B, so |E| is 9 versus about 0.11.
D) Point A — it's steeper at the top of the curve.
Correct answer: C.
At A, E = (−1)(90/10) = −9, strongly elastic. At B, E = (−1)(10/90) ≈ −0.11, deeply inelastic. The slope is −1 at both, so D's "steeper at the top" reasoning is wrong — the steepness is identical. D actually names the right point (A) for the wrong reason: A isn't more elastic because the curve is steeper there, it's more elastic because P/Q is large (90/10 = 9) up top. B repeats the original misconception that a constant slope forces a constant elasticity. Option A picks the wrong point entirely and leans on the false intuition that low prices inherently make buyers responsive; it's the small P/Q ratio at low points that makes elasticity small, not the price level itself. The driver is always P/Q, and you can check it with the two numbers.
Close the gap
The slope-equals-elasticity confusion is sticky because the fix isn't a new formula — it's noticing that one familiar formula hides a moving part (P/Q) that the slope doesn't have. Seeing that once, in numbers you computed yourself, is what makes it stop feeling like a paradox. Gradual Learning is built to catch the exact point where a rule you trust gets carried into a case it doesn't fit, and to walk you back through it with values you check by hand.