Tax Cuts, Redistribution, and Borrowing
Constraints
Tommaso Monacelli (Bocconi, IGIER and CEPR), Roberto
Perotti (Bocconi, IGIER, CEPR and NBER),
Fiscal and Monetary Policy Challenges in the Short and Long
Run, Hamburg May 19-20, 2011
Recent debate on the …scal stimulus
I
Higher spending vs. lower taxes
I
Tax changes: pro-poor or pro-rich?
Conventional wisdom
1. Lower taxes better because no implementation lags
2. But e¤ect on private spending can be minimal if households
decide to save
3. In a recession, should redistribute in favor of low-income
agents, because higher MPC
MPC higher for low income agents: evidence
I
MPC out of transitory income shocks (Parker 1999,
McCarthy 1995, Dynan, Skinner and Zeldes 2001)
I
Tax rebates (Parker 1999, Souleles 1999, Shapiro and
Slemrod 2003, Johnson, Parker and Souleles 2006).
More general questions
1. What are the aggregate e¤ects of redistributing income?
2. Are e¤ects of progressive tax cuts di¤erent from e¤ects of
regressive cuts?
I
Rarely addressed in a general equilibrium macroeconomic
model
Tax redistributions: a …rst look at the data
I
Each US tax bill since 1945
I
Assemble data on the level and composition of four
categories of taxes
1. personal income taxes
2. corporate income taxes
3. indirect taxes
4. social security taxes
Distributional impact of Personal Income Taxes
1. Employ original documentation by the Joint Committee of
Taxation
2. Provide narrative estimate of how each tax bill impacts on
the taxes paid by individuals in each income bracket
3. Data on the IRS Statistics on Income ! estimate the
number of individuals in each tax bracket, and the total
income in each tax bracket.
I
Measure how much of the total change in taxes from a given
tax bill will be borne by each decile or quartile of income.
Reagan 1981 Tax Cut
Clinton 1993 Tax Increase
Bush 2001 Tax Cut
Bush 2003 Tax Cut
"Poor-biased" tax change
I
The …rst two quartiles pay more than 50 percent of the
increase in taxes (or bene…t for more than 50 percent of the
decline in taxes).
Some theory
Our approach
1. Heterogenous agents: patient vs. impatient
2. Impatient agents face borrowing limit (as in classic
Bewley-Ayiagary-Hugget)
3. Impatience motivates borrowing (not idiosyncratic shocks)
Results
1. If prices ‡exible ! redistribution neutral or contractionary
2. If prices sticky ! redistribution (largely) expansionary
I
Address role of borrowing constraints, nominal rigidities,
persistence, govt. debt
Model: households
max E0
(
∞
∑
βtj
[u (cj ,t )
βs > βb
|{z}
|{z}
savers
cj ,t + rt
v (nj ,t )]
t =0
1 dj ,t 1
)
j = b, s
borrowers
= dj ,t + wt nj ,t
τ j ,t + σj Pt
|{z}
|{z}
lump-sum
db,t
d
| {z }
borrowing
constraint
pro…ts
share
E¢ ciency conditions
0
v (nj ,t )
= wt
λj ,t
cons/leisure
λs ,t = βs rt Et fλs ,t +1 g
λb,t = βb rt Et fλb,t +1 g + λb,t
ψt
|{z}
shadow
value
of
borrowing
Euler for savers
Pseudo-Euler for borrowers
Notice
1. If borrowing constraint binding
ψt > 0 !
λ > λs ,t
| b,t {z }
borrowers have
higher
shadow
value of
wealth
2. Credit premium
λb,t = βb
rt
1
ψt
Et fλb,t +1 g
Firms
I
Perfect competition
yt = F (nt ) = F
| {z }
∑ nj ,t
production
function
0
j
wt = F (nt ) = 1
| {z }
if CRS
!
Government
∑ τj ,t =
j
g
|{z}
…xed
govt.
spending
Neutrality
1. Perfect competition
2. Constant return to scale (CRS)
3. Steady state taxes are the same across agents
4. d = 0
cs ,t + τ s ,t
0
(rt 1 1)d = F (nt )ns ,t
|
{z
}
zero
0
cb,t + τ b,t + (rt 1 1)d = F (nt )nb,t
|
{z
}
zero
ϕ
0
ϕ
0
cs ,t ns ,t = F (nt )
cb,t nb,t = F (nt )
More generally
I
d >0
I
DRS or monopolistic competition ! Equilibrium pro…ts
deviate from zero
I
Natural assumption: savers hold shares of …rms
!Result: redistribution pro-borrowers is contractionary
Redistribution from Savers to Borrowers: Flex Prices and Decreasing Returns
Output
0
-0.05
-0.1
-0.15
-0.2
0
2
4
6
8
10
12
14
10
12
14
Aggregate Consumption
0
-0.05
-0.1
-0.15
-0.2
0
2
4
6
8
Intuition for contraction: asymmetry index
I
Endowment economy ! Each agent receive yt /2 in every
period
I
Resource constraint must imply!
ybt =
b
cb,t =
cs
y
ybt
(cb /y )
b
cs ,t +
b
cb,t
cs
b
cs ,t
cb
| {z }
asymmetry
index
cs > cb
| {z }
steady
state
cb
y
I
If savers’ss consumption larger
j∆b
cb,t j > j∆b
cs ,t j
j ∆b
n j
> j ∆b
n j savers’
| {zs ,t} l.supply
| {zb,t} borrowers’
l.supply
falls
I
rises
Asymmetric wealth e¤ect on labor supply
Nominal rigidities
I
New Keynesian setup + heterogenous agents + borrowing
constraint
I
Model inherently dynamic
I
Role of borrowing constraints in intertemporal substitution
Nominal rigidities
yt =
Z 1
yt (i ) = nt (i )
0
ε/(ε 1 )
yt (i )(ε
1 )/ε
i 2 [0, 1]
φ
(1 + it ) = r π t π
di
…nal good
pf. di¤erentiated varieties
monetary policy
Nominal rigidities
I
Suppose prices …xed for two periods (t and t+1) ! Riskless
real int. rate constant
I
Savers’Euler equation implies
cs ,t = cs ,t
1
=
cs
|{z}
savers’consumption
constant
I
Borrowers’consumption not constant
r βb Et
|{z}
constant
riskless
rate
cb,t
cb,t +1
= 1 ψt
| {z }
movements
in credit
premium
Nominal rigidities
yt = g + c s +
cb,t
|{z}
B.consumption
drives
aggr. output
Tax redistribution
∆τ s ,t =
I
∆τ b,t > 0
Transmission
# τ b,t !
# ψ ! " cb,t !
|{z}t
credit
premium
"y
|{z}t
output
expansion
Labor market
I
Aggregate labor supply
nt =
∑ nj ,t = ∑ l
j
I
cj ,t ,
j
wt
pt
Aggregate labor demand
nt = N
wt µ t
p
L cb,t , c s ,
wt
p
Labor market
Aggregate labor market e¤ects of a pro-borrower tax redistribution under
rigid prices.
Staggered prices
Output
0.4
0.3
0.2
0.1
0
0
1
2
3
4
5
6
7
8
9
10
7
8
9
10
Aggregate Consumption
0.4
0.3
0.2
0.1
0
0
1
2
3
4
5
6
Aggregate e¤ects of a pro-borrower tax redistribution: staggered
prices.
Responses to a Tax Redistribution from Savers to Borrowers
Consumption
Hours
2
2
savers
borrowers
1
1
0
-1
0
0
5
10
-1
0
Finance Premium
0
0.4
-0.2
0.3
-0.4
0.2
-0.6
0.1
-0.8
0
5
5
10
Real Riskless Rate
10
0
0
5
10
Responses to a tax redistribution from the savers to the borrowers: sticky
prices.
Temporary vs. Permanent Redistributions
Output Multiplier
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
-0.5
0
0.1
0.2
0.3
0.4
0.5
0.6
persistence in tax shock
0.7
0.8
0.9
1
Aggregate output impact multiplier of a tax redistribution that favors the
borrowers.
Extensions
1. Endogenous borrowing limit
2. Government debt
Endogenous borrowing limit
db,t
(1
|
χ)
Et fwt +1 nb,t +1 g
r
{z t
}
can collateralize
a fraction of
future L. income
Government debt
Savers
govt. bonds Bt
Fin. Intermediaries
st = d b,t + ∆ (db,t )
| {z }
intermed.
frictions
riskless deposits st
(1 +i dt )
= (1 + δt )
(1 +i t )
| {z }
spread
Borrowers
db,t
db
Debt-…nanced redistributions
gt +
(1 + it 1 )Bt
πt
τ j ,t = (1
1
= Bt +
∑
τ j ,t
govt. budget constraint
j =s ,b
ρτ )τ j + ρτ τ j ,t
1
+ φBj Bt 1 +εj ,t
| {z }
reaction to
govt. debt
Sharing the burden of debt stabilization
φBb = 0 φBs > 0 only savers’taxes adjust
φBb > 0
φBs > 0
both taxes adjust
Debt-…nanced redistribution
Flexible prices
Aggregate Output
0.02
φ
b
0.015
φ
b
0.01
φ
= 0.1
b
0.005
φ
b
=0
= 0.05
= 0.5
0
-0.005
-0.01
0
2
4
6
8
10
12
14
10
12
14
Aggregate Consumption
0.02
0.015
0.01
0.005
0
-0.005
-0.01
0
2
4
6
8
A tax cut to the borrowers under alternative values of φB
b : ‡exible prices.
Sticky prices
Aggregate Output
1.5
φ
b
φ
b
0.5
φ
b
0
φ
b
1
-0.5
0
2
4
6
8
=0
= 0.05
= 0.1
= 0.5
10
12
14
10
12
14
Aggregate Consumption
1.5
1
0.5
0
-0.5
0
2
4
6
8
A tax cut to the borrowers under alternative values of φB
b : sticky prices.