Skip to content
Back to WU

Guide

The Lagrange method in CBK Mathematik at WU Vienna: partial derivatives and worked tasks

VWUPass Team8 min read

The Lagrange method and partial derivatives belong to calculus in two variables, Differentialrechnung mit Funktionen in zwei Variablen, one of the topics of the CBK Mathematik exam at WU Vienna. You look for an optimum when two quantities are tied by a condition: the highest utility for a given budget, or the lowest cost for a given output. Derivatives are taken on paper, because calculators that can differentiate are banned in the exam. The method itself is short. Wrong answers come from small slips: a fraction turned upside down, an exponent not brought down, a constraint never substituted.

Functions of two variables in the CBK Mathematik syllabus

In the syllabus (as of 28.06.2026), this is the eighth of nine topics. In the textbook «Mathematik für Wirtschaftswissenschaften» on mathe4wiwi.org it matches chapter 8 of the online training, Analysis von Funktionen mit zwei Variablen.

In our trainer the tasks of this topic ask for a partial derivative at a point, the derivative along a curve by the chain rule, the slope of an isoquant or indifference curve, the type of a stationary point, or an optimum under a constraint. All five rest on partial derivatives.

Formulas and their German names

The exam is in German, so these are the words you will meet in the tasks:

GermanEnglishFormula
partielle Ableitungpartial derivativefx: differentiate in x, treat y as a number
Grenzproduktivitätmarginal productQx, Qy
partielle Elastizitätpartial elasticityεf,x=fx⋅x/f
Isoquante, implizite Funktionisoquant, implicit functionslope dydx=−FxFy
Grenzrate der technischen Substitutionmarginal rate of technical substitutionQxQy
stationärer Punkt, Sattelpunktstationary point, saddle pointfx=fy=0, D=fxxfyy−(fxy)2
Lagrange-Funktion, NebenbedingungLagrangian, constraintL=f−λ (g−c)
Lagrange-MultiplikatorLagrange multiplierλ

For a constraint g(x,y)=c, the Lagrange method sets Lx=0 and Ly=0 and keeps the constraint itself. The first two give fx=λgx and fy=λgy, and dividing the first by the second removes λ:

fxfy=gxgy.

With a budget pxx+pyy=B the right-hand side is the price ratio px/py. That ratio only describes a curve; the point comes from putting it into the constraint.

A formula sheet comes with the exam tasks, according to the students' union ÖH WU for 2022/23. Your own notes and formula collections go in your bag, under WU's standard exam instructions. The CBK Mathematik theory, open to everyone, has that formula sheet typeset as it is handed out in the exam, so you practise looking formulas up where you will find them.

Partial derivatives: a worked example

Take f(x,y)=3x2y−2xy2+5y at the point (2,1). When you differentiate in x, y is a constant, so the term 5y drops out and the factor y in 3x2y stays:

fx=6xy−2y2,fy=3x2−4xy+5.

At (2,1) this gives fx=12−2=10 and fy=12−8+5=9, and f(2,1)=13. The level curve f(x,y)=13 through this point has slope −fx/fy=−10/9≈−1.11: if x grows a little, y has to fall about 1.11 times as much for f to stay the same. Turning the fraction over gives −0.9, losing the minus gives +1.11, and differentiating in the wrong variable gives 9 instead of 10.

Worked Lagrange task in German

The task is in German, as in the exam:

Ein Haushalt maximiert seinen Nutzen U(x,y)=x2y unter der Budgetbeschränkung 3x+6y=180 (Preise und Budget in GE). Welche Menge des Gutes y fragt der Haushalt im Optimum nach?

A household maximises utility x2y with prices 3 and 6 and a budget of 180 money units (GE, Geldeinheiten); the question is how much of good y it buys.

(a)(b)(c)(d)(e)
10152026.6740

The Lagrangian is

L(x,y,λ)=x2y−λ (3x+6y−180).

Setting the partial derivatives to zero gives Lx=2xy−3λ=0 and Ly=x2−6λ=0. Dividing one by the other, 2xyx2=2yx=36, so x=4y. Into the budget: 3⋅4y+6y=18y=180, so y=10 and x=40. The answer is (a).

As a check, x takes 3⋅40=120 GE and y takes 60, which is 180 in all. The budget also adds up for (b), (c) and (d), so the share decides: with utility xayb the share spent on x is a/(a+b), two thirds here, against one half in (b), one third in (c) and one ninth in (d).

The wrong options:

  • (b) 15: the exponent 2 not brought down, Lx=xy−3λ. Then y/x=1/2, x=2y and 12y=180.
  • (c) 20: the price ratio turned over, 2y/x=6/3. That gives x=y and 9y=180.
  • (d) 26.67: 2y/x=1/2 solved wrongly as y=4x. With 27x=180, x≈6.67 and y≈26.67.
  • (e) 40: the correct amount of the other good, the answer to the neighbouring question.

The tasks in the CBK Mathematik trainer are solved in the same blocks: what the task is about, the steps, the answer, a check, and a line for each wrong option saying which mistake leads to it or which step to recompute. The topic there is split into the exam's task types, so you can practise partial derivatives, isoquants and the Lagrange method separately.

Cost minimisation with the Lagrange method

The same method finds the cheapest way to produce a given output, a standard example in economics. A firm produces Q(x,y)=x0.5y0.5 from labour x and capital y. A unit of labour costs 3 GE, a unit of capital 12 GE, and it needs 30 units of output. Minimise cost C=3x+12y subject to x0.5y0.5=30.

The first-order conditions are 3=λ⋅0.5 x−0.5y0.5 and 12=λ⋅0.5 x0.5y−0.5. Their ratio gives y/x=3/12, so x=4y. From 4y2=2y=30 you get y=15 and x=60, at a cost of 180+180=360 GE.

Here the multiplier means something. The first condition gives 3=λ⋅0.5⋅15/60=0.25 λ, so λ=12, which is the marginal cost (Grenzkosten): each further unit of output costs 12 GE. At these prices cost grows linearly, C(Q)=12Q, and 31 units cost 372 GE.

Common mistakes with two-variable functions

  1. Giving the exponent instead of the derivative. For a Cobb-Douglas function Q=c xαyβ the partial elasticity in x is α, but the marginal product is α⋅Q/x. Both numbers can be among the options.
  2. Stopping at the ratio. x=4y is a line, not a point. Only the constraint gives the answer.
  3. The sign of the multiplier. Writing L=f+λ (g−c) gives the same point but λ with the opposite sign. If a task asks for λ, check which form it uses.
  4. Reading the Hessian test the wrong way. D>0 means an extremum, a maximum if fxx<0; D<0 means a saddle point.

How to practise the Lagrange method

In the CBK Mathematik theory every method has a card, saying when it is needed, which formula applies and where people slip, next to an example worked step by step. Start with the theory on partial derivatives and extrema: without confident partial derivatives the Lagrange method does not work. Then read the theory on implicit functions and the Lagrange method, which covers isoquant slopes and a budget problem with other numbers. Check every answer by putting it into the constraint: the budget must add up exactly and the output must be reached.

Once the topics are done, sit a CBK Mathematik mock exam in the real format: 20 tasks with five options in 90 minutes, scored by WU's rule of three free mistakes and minus a quarter point for each further one. The first eight mocks repeat no task, and the tasks of the first chapter, linear and quadratic functions, can be solved without signing up. Elasticity in one variable, where partial elasticity comes from, is covered in our article on elasticity in CBK Mathematik, and the exam as a whole in our overview of CBK Mathematik.

Sources

· 4 checked 15.11.2026
  1. Syllabus Mathematik, WU Wien (LEARN) · dated 28.06.2026
  2. Mathematik für Wirtschaftswissenschaften (Rudloff, Zeileis), mathe4wiwi.org
  3. ÖH WU Literaturliste, Wintersemester 2022/23 (Archivkopie)
  4. Leitfaden für Fachaufsichten: Ablauf und Anweisungen bei Prüfungen, Prüfungsorganisation der WU Wien · dated 27.05.2026

Ready to practice?

Go to trainer

Read next