Optimal Tax Theory
Winter term 2023/2024
Prof. Dr. Hans Fehr
Room 489
[Link]@[Link]
Topics
◮ Introduction: Basic microeconomic concepts
◮ Commodity Taxation
◮ Commodity tax rules
◮ Taxation of externalities and internalities
◮ Taxation with home production
◮ Income Taxation
◮ Linear income tax
◮ Non-linear income tax
◮ Family taxation
◮ Taxation and Uncertainty
◮ Taxation with risky labor income
◮ Optimal social insurance design
References:
Lecture Notes “Optimal Tax Theory , Winter term 2023/2024“.
◮ Boadway, R. (2012): From Optimal Tax Theory to Tax Policy,
Munich Lectures in Economics, Cambridge and London: MIT
Press.
◮ Jacobs, B. (2013): From optimal tax theory to applied tax
policy, FinanzArchiv 68(3), 338-389.
◮ Mankiw, N.G., M. Weinzierl and D. Yagan (2009): Optimal
taxation in theory and practice, Journal of Economic
Perspectives 23(4), 147-174.
◮ Piketty, T. and E. Saez (2013): Optimal Labor Income
Taxation, in: A. Auerbach et al. (eds.): Handbook of Public
Economics, Vol. 5, Amsterdam: Elsevier, 391-474.
Course organization and grading
◮ Lecture Notes, slides and exercise sheets will be provided
online every Tuesday!
◮ Lecture: Wednesday, 8:30 to 10:00
◮ Exercise class: Wednesday 16-18, SR 321!
◮ Lena Böswald, [Link]. (Room 490,
[Link]@[Link])
Course organization and grading
◮ Lecture Notes, slides and exercise sheets will be provided
online every Tuesday!
◮ Lecture: Wednesday, 8:30 to 10:00
◮ Exercise class: Wednesday 16-18, SR 321!
◮ Lena Böswald, [Link]. (Room 490,
[Link]@[Link])
◮ Grading: up to 15 bonus points (optional) + final exam 60
points !
◮ Bonus points can be obtained in exercise class by:
◮ preparing and presenting exercises (1-3 points each
presentation)
◮ multiple choice midterm (before christmas) (1-10 points)
◮ presenting paper at the end of the semester (1-5 points)
◮ Office hours by notice (send Email)
Basic Microeconomic Concepts
Direct utility, indirect utility and expenditure function
Homogeneous and homothetic preferences
Marshallian or uncompensated demand function vs. Hicksian or
compensated demand function
Decomposition of income and substitution effects
Social welfare functions
Basic Microeconomic Concepts
Direct utility, indirect utility and expenditure function
Homogeneous and homothetic preferences
Marshallian or uncompensated demand function vs. Hicksian or
compensated demand function
Decomposition of income and substitution effects
Social welfare functions
Preferences are represented by (direct) utility function U(x1 , x2 ),
where
∂U ∂2U
= Ui > 0 and = Uii ≤ 0 for i = 1, 2.
∂xi ∂xi2
Note: Uij has not been specified so far!
Utility maximization and Marshallian demand
max L(x1 , x2 , λ) = U(x1 , x2 ) + λ(m − p1 x1 − p2 x2 )
(Necessary) First-order conditions (focs) for optimum:
∂L
= U1 (x1 , x2 ) − λp1 = 0,
∂x1
∂L
= U2 (x1 , x2 ) − λp2 = 0,
∂x2
∂L
= m − p1 x1 − p2 x2 = 0.
∂λ
(Sufficient) Second-order conditions (soc’s):
U11 U22 + U22 U12 − 2U12 U1 U2 < 0,
U1 (x1 , x2 ) p1
= .
U2 (x1 , x2 ) p2
Figure: Utility maximization
X2
m
p2
B
x2 A
_
U
C
p
1 p
U
2 X
x1 m/ p 1 1
{
p2 x p
2 1
Homogeneous and homothetic preferences
U(κx1 , κx2 ) = κk U(x1 , x2 ).
Three very useful properties:
1. If U(κx1 , κx2 ) = κk U(x1 , x2 ) then
Ui (κx1 , κx2 ) = κk−1 Ui (x1 , x2 );
2. If U(x̃1 , x̃2 ) = U(x̂1 , x̂2 ) then U(κx̃1 , κx̃2 ) = U(κx̂1 , κx̂2 );
3.
U1 (κx1 , κx2 ) κk−1 U1 (x1 , x2 ) U1 (x1 , x2 )
− = − k−1 =− .
U2 (κx1 , κx2 ) κ U2 (x1 , x2 ) U2 (x1 , x2 )
Homothetic function: Parallel MRS!
Figure: Homogeneous utility functions
x2
. k x~
~
~
x .x
k^
2
.x
x^
2
.^x kk u
u
~
x x^
x1
1 1
Note: Homothetic preferences represent monotonic transformations
of a homogeneous utility function, i.e.
U(x1 , x2 ) = g (v (x1 , x2 )),
where g (·) is a strictly increasing function and v (·) is homogeneous
of degree k.
All homogeneous utility functions are homothetic, but not all
homothetic utility functions are homogeneous!
Combining foc’s and budget constraint give the Marshallian
demand functions: xi = xi (p1 , p2 , m).
Since
xi (κp1 , κp2 , κm) = κ0 xi = xi
so that
∂xi ∂xi ∂xi ∂xi p1 ∂xi p2 ∂xi m
p1 + p2 + m = 0 so that + + = 0.
∂p1 ∂p2 ∂m ∂p1 xi ∂p2 xi ∂m x
| {z } | {z } | {z i}
ǫi ,1 ǫi ,2 ηi
where ǫi ,j represent the uncompensated price elasticities of demand
for good i and ηi represents the income elasticity of demand for
good i .
Homothetic utility implies:
κxi = xi (p1 , p2 , κm) so that xi = xi (p1 , p2 )m.
and
∂xi ∂xi m
xi = m ⇒ ηi = =1
∂m ∂m xi
Indirect utility function
Substitute the Marshallian demand functions and the Lagrangian
multiplier back into the original Lagrangian:
L(p1 , p2 , m) = U(x1 (p1 , p2 , m), x2 (p1 , p2 , m))+
λ(p1 , p2 , m) m−p1 x1 (p1 , p2 , m)−p2 x2 (p1 , p2 , m) = V (p1 , p2 , m)
∂L ∂V (·) ∂x1 ∂x2
= = (U1 − λp1 ) + (U2 − λp2 ) +
∂m ∂m ∂m ∂m
∂λ
m − p1 x1 − p2 x2 +λ = λ,
∂m
and
∂L ∂V (·)
= = −λxi (p1 , p2 , m).
∂pi ∂pi
so that we get Roy’s identity:
∂V (·)/∂pi
xi (p1 , p2 , m) = −
∂V (·)/∂m