prac2
prac2
c2 w2 (1 − t2 )n2
c1 + = w1 (1 − t1 )n1 +
1+r 1+r
where ni , wi & ti are labor supply, wages, and tax in period i = 1, 2. The
consumer takes wages, taxes, and the rate of interest, r as given. Derive first
order conditions associated with the optimal choices and interpret them. How
do changes in the wages and taxes affect the inter-temporal pattern of labor
supply? Provide intuition.
Suggested Answer:
Let λ be the Langrangian multiplier associated with the b.c. Then, using
the f.o.cs you can show that
1
c1 : =λ (1)
c1
β λ
c2 : = (2)
c1 1+r
1
n1 : = λw1 (1 − t1 ) & (3)
1 − n1
β w2 (1 − t1 )
n2 : =λ . (4)
1 − n2 1+r
As usual you can interpret f.o.cs in terms of the MC and MB of the
choices. Using these conditions you can show that
1 − n2 w1 (1 − t1 )
= β(1 + r) . (5)
1 − n1 w2 (1 − t2 )
Now suppose w1 rises. Then other things remaining the same the LHS
term should rise as well. This will imply fall in n2 and rise in n1 i.e. in
1
response to increase in w1 the consumer will work more in the first period
and work less in the second period. Similarly you can derive effects of changes
in w2 , t1 , t2 , and r.
ct1−α (1 − nt )1−α
u(ct , nt ) = + ; α>0
1−α 1−α
where ct is consumption and nt is hours worked in period t. Suppose that
the representative consumer maximizes
2
X
U= u(ct , nt )
t=1
a2 = (1 + r)a1 + w1 n1 − c1 &
c2 = a2 (1 + r) + w2 n2 .
for all t > 0. Suppose that the real wage w and real interest rate, r, are
taken as given by the consumer:
Derive the Euler equations describing the trade-off between ct+1 and ct , nt+1
and nt , and ct and nt for t = 1, 2 respectively and interpret them.
Suggested Answer:
n1 : c−α
1 w1 = (1 − n1 )
−α
(1)
n2 : c−α
2 w2 = (1 − n2 )
−α
& (2)
a2 : c−α −α
1 = (1 + r)c2 . (3)
As before you can interpret f.o.cs in terms of the MC and MB of the
choices. Using these conditions you can show that
2
−α
1 − n1 w1
= (1 + r) (4)
1 − n2 w2
which equates the MRS between leisure in period 1 and 2 to the ratio of
real wages.
c1−α
1 c1−α
max + 2
c1 ,c2 ,s 1 − α 1−α
subject to
c1 + s = y1 & (1)
c2 = Rs (2)
where y1 & R are first period income and the rate of interest respectively.
Derive the effect of R on the optimal s.
Suggested Answer: Using the first order condition, you can show that the
optimal s
y1
s= α−1 . (3)
1+R α
From (3), it is clear that the effect of R on s will depend on α. For α = 1,
α−1
s = y1 /2 is independent of R. What happens when α 6= 1. Define H = R α .
Then
α−1 1
dH/dR = (4).
α R
Using (3) and (4) you can show that ds/dR > 0 if α < 1 and ds/dR < 0 if
α > 1. The reason is that an increase in R has a positive substitution effect
on s but a negative income effect. The strength of these two effects depends
on α.
3
From the lecture note we know that s will be higher compared to the
certainty case if
Uccc (c2 )
− Rs > 2. (1)
Ucc (c2 )
In this case c2 = Rs. Then (1) implies that if α + 1 > 2 or α > 1 saving
will be higher.
c2 y2 (1 − t2 )
c1 + E = y1 (1 − t1 ) + E
1+r 1+r
where yi denotes labor income and ti denotes tax rate for i = 1, 2. Sup-
pose that there is a linear technology in this economy, which determines the
constant interest rate r.
(a.) Solve for c1 and E(c2 ).
(b.) Suppose that y and r are constant. In period 1 the tax authority
announces the following rule for setting taxes in period 2:
t2 = λt1 + 2
where 2 is an unpredictable shock. Let E(2 ) = 0. Find the statistical
relationship between c1 and t1 .
(c.) In general, can we evaluate the effects of tax changes using decision
rules? Hint: Lucas Critique.
c1 = E(c2 ). (1)
By putting (1) in the budget constraint you can show that
1+r y2 (1 − t2 )
c1 = E(c2 ) = y1 (1 − t1 ) + E . (2)
2+r 1+r
4
By putting the policy rule in (2), you can show that
1+r y2 λ
c1 = y1 + − [y1 + ]t1 . (3)
2+r 1+r 1+r
In order to find out the relationship between consumption and tax nor-
mally, we regress consumption on tax rate. In this case, the estimated co-
1+r λ
efficient of t1 , B ≡ 2+r (y1 + 1+r ). In order to find out, the effect of tax
changes we simulate the model for various values of tax rate. However, this
procedure is valid only when the coefficient of tax rate B is constant or is
invariant to changes in tax rates/ policy/ regime . However, B depends on
λ and if there is tax changes, there might be changes in λ too and thus the
original estimate of B may not be valid under new tax policy/ regime. This
may lead to incorrect estimate of the effects of taxes. This is the essence of
Lucas critique. With the change in policies/regimes, the behavioral relation-
ships change (summarized in the coefficients like B). Historically observed
relationships may not be correct guide to the future if policies change. Thus,
using decision rules to gauge effects of policies can be problematic.