Marginal cost and elasticity: what they mean

8 min · 1 drill

In the textbook: Ch. 3.3.7 Elastizitäten3.5 Anwendungen – Optimierung

In the last article the derivative was the slope of a curve. Here it gets a meaning in money. We start with a question that is easy to get wrong.

A joinery makes 20 stools a week and spends 700 € on them, that is 35 € per stool. A customer comes and offers 40 € for one more stool. It looks worth it: 40 is more than 35. But the joiner also works out the cost of 21 stools:

Cost per week
20 stools700 €
21 stools751 €
The 21st stool alone51 €
On average one of the 2035 €

The 21st stool costs 51 €, not 35: the more the workshop makes, the dearer each further piece becomes. For 40 € the joiner would lose 11 €. The average 35 € is misleading here; what you need is the increase in cost from one more piece. The derivative of the cost gives it almost exactly, 50 € here. This derivative is called marginal cost (“Grenzkosten”).

In the same way the derivative answers the joiner's other questions:

  • by how many percent demand falls if he raises the price by 1 % (this is elasticity, “Elastizität”);
  • at which output profit is largest (this is the maximum);
  • where marginal cost is lowest (this is the inflection point, “Wendepunkt”).

How to work out the derivative itself is explained in the article “What the derivative is: slope and the rules”.

Marginal cost and unit cost

The joiner worked out the costs only for 20 and 21 stools. To answer the same way for any quantity, the costs are written as a formula. The cost K(x) is everything a firm spends on x pieces. The fixed cost K(0) is paid even without production. This is explained in detail in the article “What a function is: cost, revenue, profit and demand”. Let K(x)=x2+10x+100 €. From this function you get three different numbers per piece:

  • Marginal cost K′(x)=2x+10. At 20 pieces K′(20)=50: the twenty-first piece costs about 50 €. Exactly: K(21)=441+210+100=751 and K(20)=400+200+100=700, the difference is 51.
  • Unit cost (“Stückkosten”) k(x)=K(x)x: k(20)=70020=35 € per piece, with the fixed cost.
  • Variable unit cost kv(x)=K(x)−K(0)x: 700−10020=30 €.

The three numbers 50, 35 and 30 describe one firm and one quantity. The task always says which of them it asks for.

Other functions are read the same way: the marginal revenue (“Grenzerlös”) E′(x) is the extra revenue from one more piece, the marginal profit is G′(x)=E′(x)−K′(x) (G for “Gewinn”, profit).

Elasticity: the change in percent

The derivative measures in pieces: the demand x(p)=120−4p at price p loses 4 pieces for every euro of price. But 4 pieces out of 40 and out of 400 mean different things. The elasticity (the letter ε, “epsilon”) translates this into percent:

εf(x)=f′(x)⋅xf(x)

At the price p=20 the demand is x(20)=40, the derivative x′(p)=−4, the elasticity −4⋅2040=−2. A price rise of 1 % (to 20.2) lowers demand by about 2 %: x(20.2)=39.2, which is 2 % less than 40. If the absolute value of the elasticity (the number without the minus sign: for −2 it is 2) is greater than 1, the function is called elastic (“elastisch”) at this point; if it is less than 1, inelastic (“unelastisch”).

Rising, maximum and minimum

Marginal cost and elasticity looked at the derivative at one point. Go through all the points and watch its sign, and you see where the function rises and where it falls. Take f(x)=x3−3x. Its derivative f′(x)=3x2−3=3 (x−1)(x+1) is zero at x=−1 and x=1. Work it out at one point in each piece, left of −1, between −1 and 1, right of 1: f′(−2)=9, f′(0)=−3, f′(2)=9. Left of −1 the derivative is positive, and the graph goes up: f(−2)=−2, f(−1)=2. Between −1 and 1 it is negative, and the graph goes down to f(1)=−2. Right of 1 it is positive again, and the graph rises again. At −1 and 1 the tangent is horizontal.

So the sign of the derivative tells you where the graph is going. This is called monotonicity (“Monotonie”):

  • f′(x)>0: the function rises at this point;
  • f′(x)<0: the function falls;
  • f′(x0)=0, and rising turns into falling: the top of a hill, a maximum; falling turns into rising: the bottom of a dip, a minimum. Both together are called extrema (“Extremstellen”).

For x3−3x the maximum is at x=−1, f(−1)=2, and the minimum at x=1, f(1)=−2. In the graph below the hill is at −1, the dip at 1 and the inflection point at 0 (more on it in the next section).

Graph of f(x) = x³ − 3x−2−1012−3−2−1123xmaximum (−1, 2)inflectionminimum (1, −2)

Instead of the signs you can look at the second derivative f′′(x)=6x. At x=−1 it is −6<0, a maximum; at x=1 it is 6>0, a minimum.

The stretches of the axis on which f rises or falls are written as intervals. On the same example:

  • (−1,1): all x between −1 and 1, the ends themselves not included (round brackets). Here f falls.
  • [−2,3]: all x from −2 to 3 including the ends (square brackets), a closed interval.
  • (1,∞): all x to the right of 1. The sign ∞ (“infinity”) means there is no right end. Here f rises.

Two places where people go wrong:

  • f′=0 is not yet an extremum. For x3 the derivative 3x2 is 0 at zero, but positive on both sides: the function rises and is horizontal for just a moment.
  • An interval has ends. On the interval [−2,3] the function x3−3x has its largest value f(3)=27−9=18 at the right end, more than the maximum 2. You look for the largest value among the extrema and the ends.

This gives two names:

  • local maximum (“lokales Maximum”): the peak f(−1)=2, near which all values are smaller;
  • global maximum (“globales Maximum”): the largest value on the whole interval [−2,3], here f(3)=18.

So a local maximum need not be global. Minima work the same way. The bottom of the pit f(1)=−2 is a local minimum and on [−2,3] also the global one: at the left end f(−2)=−2 as well, and no value is smaller.

Curvature and inflection point

The sign of the first derivative told you where the graph goes. How it bends on the way is told by the sign of the second, which the previous section used only at the peak and in the dip. Compare x2 and −x2.

  • For x2 the slope 2x keeps growing, the second derivative is 2>0, and the graph bends upwards like a bowl.
  • For −x2 the slope −2x keeps falling, the second derivative is −2<0, and the graph looks like a dome.

For x3−3x the second derivative is 6x: left of 0 it is negative, and there is a dome with the hill at −1; right of 0 it is positive, and there is a bowl with the dip at 1.

That is why the function is called convex (“konvex”) where f′′>0 and concave (“konkav”) where f′′<0. Where f′′ changes sign, one curvature turns into the other; that is the inflection point. For x3−3x it is x=0.

For cost the inflection point matters: the smallest marginal cost lies there.

Now a firm's cost, K(x)=0.25x3−3x2+19x+60. Its derivatives: K′(x)=0.75x2−6x+19, K′′(x)=1.5x−6. From 1.5x−6=0 follows x=4, and the smallest marginal cost is K′(4)=12−24+19=7 €.

Your turn1 of 1

Where does the function f(x)=x3−12x have a local maximum?

The topic in a minute

  • Marginal cost (“Grenzkosten”) K′(x): roughly what one more piece costs. Trap: this is not the unit cost K(x)x.
  • Variable unit cost (“variable Stückkosten”) kv(x)=K(x)−K(0)x: the unit cost without the fixed cost K(0). Trap: leaving K(0) in the numerator.
  • Elasticity (“Elastizität”) εf(x)=f′(x)⋅xf(x): by how many percent f changes when x grows by 1 %. Absolute value (the number without the minus) below 1: inelastic, above 1: elastic. Trap: f′f without x.
  • Rising: f′>0 on an interval means that f rises there, f′<0 means that f falls. Trap: the sign of f′ at one point says nothing yet about the whole interval.
  • Extremum (maximum or minimum): there f′(x0)=0, and f′ changes sign; or f′′(x0)<0 for a maximum, f′′(x0)>0 for a minimum. Trap: f′=0 without a change of sign, and forgetting the ends on an interval.
  • Curvature: f′′>0 convex, f′′<0 concave; where f′′ changes sign there is an inflection point (“Wendepunkt”). For the cost K, the inflection point xw carries the smallest marginal cost K′(xw). Trap: giving the point xw instead of the value K′(xw).

What comes next

Now practise it

Marginal cost and elasticity: what they mean

Trainer · Differential calculus

Practise
Next: The derivative at a point: roots, the chain rule, x to the power x
0/1

Was this article helpful?