FullChap2

b00694343

December 17, 2020

1 Introduction

Introducing fiscal instruments to improve the competitive on the international market is not a new
idea. Back to the golden standard era, Keynes (1931) has proposed that a uniform ad valorem tariff
on all imports plus a uniform subsidy on all exports would have the same impact as an exchange
rate devaluation. The essense of a fiscal devaluation is to decrease the cost of domestic products
and increase the price of imports products. This gap will lead to an improvement of the trade
balance and the following production improvement, employment rate improvement and possible
production innovation. As part of their fiscal reforms, Denmark in 1988 and Sweden in 1993
(calmfors,1998) have already implemented fiscal devaluation and got good outcomes. More recent
standard examples are countries like Ireland, Spain, Germany and France, which implemented fiscal
devaluation around 2010, before or after the euro crisis, aiming at domestic fiscal reform and also
later getting out of the stillness.
In a monetary union, currency tools are no longer available. A general recognition has been
prompted in European Union that moving from the taxation of final goods, at a uniform rate,
on anorigin basis (according to where they are produced) to their taxation on a destination basis
(according to where they are consumed) is essentially equivalent to an exchange rate devaluation:
such a shift brings imports into tax, and takes exports out.(Calmfors (1998), Demooijkeen) After
the sovereign debt tensions in 2010, the loss of competitiveness in Southern European countries
and the following subsequent emergence of within-union external imbalances have been widely
regarded as important factors contributing to the euro area crisis. Countries in Figure 1 have been
suffering from a decreasing real effective exchange rate since 2008, especially Ireland and Greece.
Compared with Germany, countries like Greece, Portugal, Spain, Ireland and Italy are suffering
from huge current account deficits since the crisis(Figure 2). In these countries, fiscal devaluation
has come back to the forefront of policy discussion and been proposed as an option to enhance their
competitiveness and readjust external imbalance (Puglisi, 2014)
Farhi et al (2014) summarized the fiscal implementation in to two main types. One is a uniform
increase in import tariff and export subsidy, and the other one is a VAT increase and a uniform
payroll tax reduction, holding government budget constraints neutral. However, along with the
boom of international trade, the use of the first tool has been limited by the rules of not only the
WTO but also many newly emerging regional economic cooperation agreements in recent decades.
Based on this fact, my paper will focus on the second type of fiscal devaluation policy.
Some researches have discussed about the impact of fiscal devaluation in an open economy. Eagler
et al (2017) use a DSGE model to calibrate the fiscal devaluation implemented as a revenue-

1
 Figure 1: Real Effective Exchange Rate (2010=100)
fig1 Source: Eurostat, 2020

Figure 2: Current account balance (% of GDP)

Source: World Bank, 2020
fig2

2
neutral shift from employer’s social contributions to the Value Added Tax. They find that a fiscal
devaluation in ’Southern European countries’ has a strong positive effect on their output, but
mild effects on the trade balance and the real exchange rate. The ’Northern European countries’
face a negative effect on their output. Using OECD countries data, de Mooij and Keen (2012)
present empirical evidence which suggests that in Eurozone countries fiscal devaluation may indeed
improve the trade balance in the short-run, but the effects eventually disappear. They think that
VAT reform is not quite available in meeting fiscal consolidation needs. Lipinska and con Thadden
(2012) explores the fiscal devaluation under a monetary union characterised by national fiscal
policies and supranational monetary policy. They show that when financial market are perfectly
integrated, a unilateral fiscal devaluation gives positive effects on consumption and production on
both domestic and foreign market. They also show that short-run effects are not always amplified
by nominal wage rigidity and depend on the choice of the inflation index stabilized by the central
bank and on whether the tax shift is anticipated. Farhi et al. (2014) analyse fiscal devaluation in a
DSGE model with monopolistic competion and show that When the devaluations are anticipated,
fiscal devaluations need to be supplemented with a consumption tax reduction and an income tax
increase. Aurey et al (2017) quantify the effects of fiscal devaluations within a monetary union
model with endogenous entry and endogenous tradability. They find that permanent unilateral
fiscal devaluations have large and permanent positive effects on hours worked, consumption, GDP,
produced and exported varieties on both countries. Also the timing of reforms and their potential
preannouncements also matter for the resulting short term dynamics and welfare gains, but are
neutral in the long term.
This paper, to the best of our knowledge, is the first one try to discuss the effects of fiscal devaluation
on the innovation sector and the number of firms on both market. In the two-country economy,
consumers are characterized by CES-utility functions. Firm are devided into two types. Innovation
firms invent new varieties and producing firms bring them into production in the next term. By
investing in the innovation sector and lending money to both governments, households give their
bequest to the next generation. Both governments holding budget constraints neutral, the fiscal
devaluation starts from a decrease in labour tax and a compensated increase of consumption tax
in domestic country, following a government debt shift in foreign countries.
With a decrease in the social contribution or payroll tax, fiscal devaluation decreases producers’
labour cost, and stimulates firms to use more labour. This will eventually improve the employment
rate and increase production. At the same time, the reduction of cost will attract less productive
firms to enter the domestic market, innovation sector gets developed, providing consumers with
more choices. Prior to fiscal devaluation, these firms are not profitable enough to do so. Moreover,
firms already existing in the market but not able to export become able to enter foreign markets
as well. The decrease of labour cost is only for domestic producing firms, but the increase in
consumption tax targets both domestic products and foreign products. This will make imports more
expensive on the domestic market.Together with the decrease in labour cost of domestic products
on foreign markets, domestic firms will set lower final prices and win more competitiveness both at
home and abroad. And as a result the trade balance will improve.

2 Model

We consider a two-country model, where home country is labeled with H and foreign country with
F . Both countries are in the same monetary union.

3
2.1 Consumers

In each country, people live for one period and have one child. The number of households is then
constant in both countries, denoted by L in the domestic country H, and L∗ in the foreign country
F.
In period t, there exists a continum of differentiated commodities, indexed by ω ∈ Ωt , supplied by
monopolistic firms. Let pt (ω) and xt (ω) be the consumer price and the quantity of a given variety
ω ∈ Ωt .
Households care not only for their own consumptions, but also for the bequests they can give to
the next generation. Utility of a representative household that lives in the home country in period
t writes:
(1 − γ) ln Ct + γ ln at+1
where Ct aggregates consumptions in all commodities:
Z  σ
σ−1 σ−1
Ct = xt (ω) σ dω
Ωt

and at+1 is the bequest to the offsprings.

The overall household’s expenditure is
Z
Et = pt (ω)xt (ω)dω
Ωt

Let wt be the nominal wage rate in the home country (wt∗ abroad). Each household inelastically
supplies its unit time endowment in formal labor. Total resources of an household living in period
t are composed of labor income wt and inherited wealth at plus the associated interests it at , where
it denotes the nominal international interest rate. These resources are divided between commodity
expenditure Et and bequest at+1 to the offsprings. So the budget constraint writes:
Et + at+1 = (1 + it ) at + wt

To maximize the individual’s utility, we build the Lagrangian function using λt as the Lagrangian
multiplier:
Lt = (1 − γ) ln Ct + γ ln at+1 + λt [(1 + it ) at + wt − at+1 − Et ]

The first-order conditions are:

• with respect to xt (ω)

1−γ 1 1
(Ct ) σ (xt (ω))− σ = λt pt (ω) (1) FOC_x
Ct
• with respect to at+1
γ
= λt (2) FOC_a
at+1

Consider the usual price index Pt , characterized by

Z
Pt1−σ = pt (ω)1−σ dω (3) P_t
Ωt
consumer
We then get (see Appendix A)

4
 • the demand function for variety ω
 −σ
pt (ω)
xt (ω) = Ct (4) x_expression
Pt

• the following relationship between bequest and total consumption

γ
at+1 = γ ((1 + it )at + wt ) = P t Ct (5) a_tC_t
1−γ

2.2 Producing sector

There are two sectors in the economy: an innovation sector where new varieties are created and a
production sector where these varieties are produced. The production sector consists of monopo-
listic firms. Each firm produces and sells one variety that has been created in the innovation sector
one period before. All the varieties in the producing sector only exist for one period.
A firm that owns a blueprint can sell the corresponding variety at home and abroad. Consider one
market without mentioning if it is at home or abroad. The quantity of variety ω is produced using
linear technology with labor as the only input:

Lxt (ω) = z`t (ω)

where z represents labor productivity, and `t (ω) is the quantity of labor used into the production.
The firm maximizes profits
(1 + τtw ) wt
 
πt (ω) = (1 − τtc ) pt (ω) − Lxt (ω)
z
where τtc and τtw are respectively the tax rates on sales and labor paid by the firm (they differ
according as the locationsx_expression
where production and consumption take place). The firm takes account
of the demand function (4). The first-order condition implies
1 dpt (ω) 1 1 + τtw wt
xt (ω) + 1 =
pt (ω) dxt (ω) pt (ω) 1 − τtc z
With a constant price elasticity σ, the price of the variety is:
σ 1 + τtw wt
pt (ω) =
σ − 1 1 − τtc z
It is the same for all varieties ω demanded by a given representative consumer and produced with
the same labor cost and tax rates.
We then get four different prices according as the locations where production and consumption
take place:
w ∗ 1+τ w
σ 1+τt wt

pH
t = σ−1 1−τ c z , pH∗
t = σ∗σ−1 1−τ tc∗ wzt 
t t 
(6) prices
w∗ ∗ w∗ ∗
σ 1+τt wt σ ∗ 1+τt wt
pFt pFt ∗

= σ−1 1−τt z ,
c ∗ = ∗ c∗
σ −1 1−τt z ∗


z is labor productivity at home, and z ∗ , abroad. wt is the wage rate at home, and wt∗ , abroad. Tax
rate on sales is τtc at home and τtc∗ abroad. Finally, tax rate on labor (or subsidy rate if negative)

5
is τtw at home and τtw∗ abroad. Then, pH t is domestic product’s (H) price on domestic (H) market;
pFt is foreign product’s (F ) price on domestic (H) market; pH∗t is domestic product’s (H) price on
F ∗
foreign (F ) market; and, finally, pt is foreign product’s (F ) price on foreign (F ) market.
Profits write
πtH = LxH 1 c H πtH∗ = L∗ xH∗ 1 c∗ H∗

t σ (1 − τt ) pt , t σ ∗ (1 − τt ) pt 
(7) profits
πtF = LxFt σ1 (1 − τtc ) pFt , πtF ∗ = L∗ xFt ∗ σ1∗ (1 − τtc∗ ) pFt ∗


where −σ −σ∗ 

pH pH∗
 
xH = t
Ct , xH∗ = t
Ct∗

Pt∗
t t

Pt 

(8) demands
−σ ∗
−σ∗
pF pF
  
xFt = xFt ∗ = Ct∗

t
Pt Ct , t
Pt∗



2.3 Innovation sector

In the innovation sector, firms invent new varieties in period t − 1, and sell the blueprints to
producing firms that will be active only during one period (period t).
Innovating firms behave competitively and only operated in their own country. They produce
blueprints with a linear technology that uses only labor. The cost of a new blueprint is η labor
units at home and η ∗ labor units abroad. Both have to be multiplied by the associated unit labor
cost of each country in order to get the nominal cost of producing a new variety.
In period t − 1, producing firms located in the home country, which intent to be active in period t,
are buying blueprints from innovators at price vt (the price is vt∗ for firms that are located in the
foreign country). Market for blueprints is perfectly competitive, so that price is equal to marginal
cost:
wt−1 and vt∗ = η ∗ 1 + τt−1
e

e∗

∗
vt = η 1 + τt−1 wt−1
Labor in the innovation sector is taxed at rate τte in the home country and τte∗ abroad. To pay the
price of the blueprint (vt or vt∗ ) in period t − 1, the producing firm has to borrow money at the
nominal rate it . The differentiated blueprint gives the producing firm a monopolist market power.
A firm that has bought a blueprint in period t − 1 needs also to pay a fixed cost in period t to enter
the market:

• A firm located in country H pays fD units of labor to enter the domestic market, and fE
units of labor to export. Then, it is active in country H iff πtH ≥ fD (1 + τts ) wt . It is active
abroad iff πtH∗ ≥ fE (1 + τts ) wt .
∗ units of labor to enter its own market, and f ∗ units of
• If located in country F , a firm pays fD E
labor to export. Then it is active in the foreign market iff πtF ∗ ≥ fD∗ (1 + τ s∗ ) w ∗ . It is active
t t
in the domestic market iff πtF ≥ fE∗ (1 + τts∗ ) wt∗ .

Free-entry leads to the equality between the price of a blueprint in period t − 1 and the discounted

6
profits earned in period t:

max πtH − fD (1 + τts ) wt ; 0 + max πtH∗ − fE (1 + τts ) wt ; 0

 
vt =
1 + it
max πtF ∗ − fD
∗ (1 + τ s∗ ) w ∗ ; 0 + max π F − f ∗ (1 + τ s∗ ) w ∗ ; 0
 
t t t t t
vt∗ = E
1 + it

subsectionInternational capital market

There exists an international capital market, where firms, governments and households can borrow
and lend. At period t, the interest rate is it+1 . In period t, households in country H (resp. F )
invest their saving for bequest at+1 (resp. a∗t+1 ) in two kinds of assets: (i) loans to the firms that
need to finance blueprint purchases in order to produce in t + 1; (ii) government bonds of both
countries. Both kinds of assets offer the same nominal return it+1 . In aggregate, the financial
market equilibrium writes

Lat+1 + L∗ a∗t+1 = ∆t+1 + ∆∗t+1 + Nt+1 vt+1 + Nt+1

∗ ∗
vt+1

where ∆t+1 and ∆∗t+1 are the respective public debts of the domestic country and the foreign
country.

2.4 Government budget constraint

We write the government budget constraints in both countries, assuming that all producing firms
sell at home and abroad. Government in country H finances the return to public debt, (1 + it )∆t ,
through implementing labor taxes, consumption taxes and issuing a new debt:
∗ F F w wt ∗ H∗
∆t+1 + τtc L Nt pH H
Nt LxH



t xt + Nt pt xt + τt t + L xt
z
+ τt Nt (fD + fE )wt + τte Nt+1 ηwt
s

= (1 + it )∆t

In country F , the government budget constraint writes

∗
w∗ wt
∆∗t+1 + τtc∗ L∗ Nt pH∗ H∗ ∗ F∗ F∗
Nt∗ LxFt + L∗ xFt ∗



t x t + N p
t t x t + τ t ∗
z
+ τts∗ Nt∗ (fD
∗
+ fE∗ )wt∗ + τte∗ Nt+1
∗
η ∗ wt∗
= (1 + it )∆∗t

2.5 Equilibrium

At equilibrium, we need both financial market and labor market clearing equations, the free entry
condition, the government budget constraints and the equations that describe consumers’ behaviors.

7
 • households in Country H:1
γ
at+1 = γ ((1 + it ) at + wt ) = Pt Ct ,
1−γ
 H 1−σ  F 1−σ
pt ∗ pt
1 = Nt + Nt
Pt Pt
H
 H −σ F
 F −σ
xt pt xt pt
= , =
Ct Pt Ct Pt
H
pt w
σ 1 + τt wt pFt σ 1 + τtw∗ wt∗
= , =
Pt σ − 1 1 − τtc zPt Pt σ − 1 1 − τtc z ∗ Pt

• households in Country F :
γ∗
a∗t+1 = γ ∗ ((1 + it ) a∗t + wt∗ ) = P ∗C ∗
1 − γ∗ t t
 H∗ 1−σ∗  F ∗ 1−σ∗
pt ∗ pt
1 = Nt + Nt
Pt∗ Pt∗
 H∗ −σ ∗  F ∗ −σ∗
xH∗
t pt xFt ∗ pt
= , =
Ct∗ Pt∗ Ct∗ Pt∗
pH∗
t σ ∗ 1 + τtw wt pFt ∗ σ ∗ 1 + τtw∗ wt∗
∗ = ∗ , ∗ = ∗
Pt ∗ c∗
σ − 1 1 − τt zPt Pt σ − 1 1 − τtc∗ z ∗ Pt∗

• International capital market equilibrium in period t:

Lat+1 − Nt+1 η (1 + τte ) wt + L∗ a∗t+1 − Nt+1
∗
η ∗ (1 + τte∗ ) wt∗ = ∆t+1 + ∆∗t+1

• Labor market equilibrium in period t:

– Country H
LxH ∗ H∗
 
t + L xt
L = Nt + fD + fE + Nt+1 η
z
– Country F
LxFt + L∗ xFt ∗
 
∗
L = Nt∗ ∗
+ fD + fE∗ ∗
+ Nt+1 η∗
z∗
In both countries, the labor is devoted to the producing sector, the innovation sector and
those service parts supporting firms entering the markets.
• Free-entry conditions in the producing sector in period t:
We assume all producing firms are active on both markets. To satisfy the free entry condition,
the investment in the innovation sector (labor cost) should be equal to the net profits (profits
on both markets minus the entry cost):
– Country H
e H1
wt−1 = L (1 − τtc ) pH


(1 + it )η 1 + τt−1 t xt
σ
1
+ L∗ (1 − τtc∗ ) pH∗ H∗
t xt − (fD + fE ) (1 + τts ) wt
σ∗
1
The second line in both countries are deduced from the price index.

8
 – Country F
1
(1 + it )η ∗ 1 + τt−1

∗
e∗
wt−1 = L∗ (1 − τtc∗ ) pFt ∗ xFt ∗ ∗
σ
1 ∗
+ L (1 − τtc ) pFt xFt − (fD + fE∗ ) (1 + τts∗ ) wt∗
σ

• Government budget constraints

∆t+1 + τtc LPt Ct + τtw wt L

+ (τts − τtw ) Nt (fD + fE )wt + (τte − τtw ) Nt+1 ηwt
= (1 + it )∆t

∆∗t+1 + τtc∗ L∗ Pt∗ Ct∗ + τtw∗ wt∗ L∗

+ (τts∗ − τtw∗ ) Nt∗ (fD
∗
+ fE∗ )wt∗ + (τte∗ − τtw∗ ) Nt+1
∗
η ∗ wt∗
= (1 + it )∆∗t

3 Equilibrium with real variables

To make equilibrium analysis more convenient, we rewrite equilibrium equations in terms of the
following real variables, eliminating all the nominal ones:

• real prices
pjt pj∗
ρjt = and ρj∗
t = t
, for j = H, F
Pt Pt∗
• real wage
wt w∗
ωt = and ωt∗ = t∗
Pt Pt
• real exchange rate
Pt∗
Qt =
Pt
• real interest rates
∗
(1 + it ) Pt−1
(1 + it ) Pt−1
1 + rt = and 1 + rt∗ =
Pt Pt∗
Notice that
1 + rt∗ Qt−1
=
1 + rt Qt
• Real assets
at+1 a∗
bt+1 = and b∗t+1 = t+1
Pt Pt
∆ ∗
∆t+1
δt+1 = ∗
and δt+1 = t+1
Pt Pt

9
To further analyze the effects of fiscal devaluation, we assume that all the following instruments
are set to constant values:

• δt , τtw , τte , τts in the home country.

• τtc∗ , τtw∗ , τte∗ , τts∗ in the foreign country.

The tax rate on consumption τtc in the domestic country adjusts in order to balance the budget
constraint of the government. In the foreign country, since interests on public debt, wage rate and
consumptions will react to the change of tax rates in home country, we can obtain a change of the
∗ . The intertemporal equilibrium
tax collection and then, the adjustment of foreign public debt δt+1
is characterized by the following set of equations in real variables:

• Households
γ
bt+1 = γ [(1 + rt ) bt + ωt ] = Ct
1−γ
γ∗
b∗t+1 = γ ∗ [(1 + rt ) b∗t + Qt ωt∗ ] = Qt Ct∗
1 − γ∗
• labor market clearing condition:

−σ
∗
H∗ −σ L∗ C ∗
!
ρHt LCt + ρ t t
L = Nt + fD + fE + Nt+1 η
z

−σ
−σ∗ ∗ ∗ !
∗ ∗ ρFt LCt + ρFt ∗ L Ct ∗ ∗ ∗
L = Nt + fD + fE + Nt+1 η∗
z∗

• Government budget constraint

δt+1 + τtc LCt + τtw ωt L
+ (τts − τtw ) Nt (fD + fE )ωt + (τte − τtw ) Nt+1 ηωt
= (1 + rt )δt

∗
δt+1 + τtc∗ L∗ Qt Ct∗ + τtw∗ Qt ωt∗ L∗
+ (τts∗ − τtw∗ ) Nt∗ (fD
∗
+ fE∗ )Qt ωt∗ + (τte∗ − τtw∗ ) Nt+1
∗
η ∗ Qt ωt∗
= (1 + rt )δt∗

• International capital market equilibrium

Lbt+1 + L∗ b∗t+1 − Nt+1 η (1 + τ e ) ωt − Nt+1
∗
η ∗ (1 + τ e∗ ) Qt ωt∗ = δt+1 + δt+1
∗

• Free entry conditions

e

1−σ 1
ωt−1 = L (1 − τtc ) ρH


(1 + rt )η 1 + τt−1 t Ct
σ

1−σ∗ 1
+ L∗ (1 − τtc∗ ) Qt ρH∗
t Ct∗ − (fD + fE ) (1 + τts ) ωt
σ∗

1−σ∗ ∗ 1
(1 + rt∗ )η ∗ 1 + τt−1

∗
e∗
ωt−1 = L∗ (1 − τtc∗ ) ρFt ∗ Ct ∗
σ
1 1−σ 1 ∗
+ L (1 − τtc ) ρFt + fE∗ ) (1 + τts∗ ) ωt∗


Ct − (fD
Qt σ

10
 • Static relationships between prices
From the households behavior and price index, we can get:

1−σ
1−σ
1 = Nt ρH t + Nt∗ ρFt
σ 1 + τ w ωt σ 1 + τ w∗ Qt ωt∗
ρHt = , ρ F
t =
σ − 1 1 − τtc z σ − 1 1 − τtc z ∗
1−σ ∗ 1−σ ∗
1 = Nt ρH∗ + Nt∗ ρFt ∗



t
σ ∗ 1 + τ w ωt F∗ σ ∗ 1 + τ w∗ ωt∗
ρH∗
t = , ρ t =
σ ∗ − 1 1 − τ c∗ Qt z σ ∗ − 1 1 − τ c∗ z ∗

4 Steady state

Holding constant domestic debt (δt = δ), constant domestic taxes (τtw = τ w , τte = τ e , τts = τ s ),
and constant foreign taxes (τtc∗ = τ c∗ , τtw∗ = τ w∗ ,
τte∗ = τ e∗ , τts∗ = τ s∗ ), a steady state equilibrium
is composed of prices r, Q, ω, ω ∗ , ρH , ρF , ρH∗ , ρF ∗ , quantities (C, C ∗ , b, b∗ , N, N ∗ ) and instruments
(τ c , δ ∗ ), that is 16 variables characterized by 16 equations among the following 17 equations:

• Household budget constraints:

γ
b = γ [(1 + r) b + ω] = C (9)
1−γ
γ∗
b∗ = γ ∗ [(1 + r) b∗ + Qω ∗ ] = QC ∗ (10)
1 − γ∗

• Labor market equilibria

−σ
−σ∗ L∗ ∗
η + fD + fE ρH C + ρH∗ L C
N = 1−N (11)
L z
F

−σ L
∗
F ∗ −σ C ∗
∗ + f∗
η ∗ + fD ρ L∗ C + ρ
N∗ E
= 1 − N∗ (12)
L∗ z∗

• Government budget constraints

rδ = τ c LC + τ w Lω + (τ s − τ w ) (fD + fE )N ω + (τ e − τ w ) ηN ω (13) BCgov_home

rδ ∗ = τ c∗ L∗ QC ∗ + τ w∗ Qω ∗ L∗
+ (τ s∗ − τ w∗ ) N ∗ (fD
∗
+ fE∗ )Qω ∗ + (τ e∗ − τ w∗ ) N ∗ η ∗ Qω ∗ (14)

• International capital market equilibrium

Lb + L∗ b∗ − η (1 + τ e ) N ω − η ∗ (1 + τ e∗ ) N ∗ Qω ∗ = δ + δ ∗ (15) CapM

11
 • Free entry conditions

1−σ 1
(1 + r)η (1 + τ e ) ω = L (1 − τ c ) ρH C
σ

1−σ∗ 1
+ L∗ (1 − τ c∗ ) Q ρH∗ C∗ − (fD + fE ) (1 + τ s ) ω
σ∗

1−σ∗1
(1 + r)η ∗ (1 + τ e∗ ) ω ∗ = L∗ (1 − τ c∗ ) ρF ∗ C∗
σ∗
1 F
1−σ 1 ∗
+ L (1 − τ c ) ρ C − (fD + fE∗ ) (1 + τ s∗ ) ω ∗
Q σ

• Static relationships between prices:

1−σ
1−σ
1 = N ρH + N ∗ ρF (16)
σ 1 + τw ω σ 1 + τ w∗ Qω ∗
ρH = , ρF
= (17)
σ − 1 1 − τc z σ − 1 1 − τ c z∗
1−σ ∗ 1−σ ∗
1 = N ρH∗ + N ∗ ρF ∗



(18)
σ∗ 1 + τ w ω F∗ σ ∗ 1 + τ w∗ ω ∗
ρH∗ = , ρ = (19)
σ ∗ − 1 1 − τ c∗ Qz σ ∗ − 1 1 − τ c∗ z ∗

5 Fiscal devaluation

Fiscal devaluation implemented in country H runs in the following scenario: to keep the sequence
of public debt in country H unchanged, fiscal devaluation is consisted of a permanent decrease in
labor taxes (, which is a permanent decrease in at least one tax rate among τ w , τ s and τ e ), and it is
financed through an increase in the consumption tax τtc in each period. Government in country F
does not modify its tax rates. The foreign public debt adjusts to changes in endogenous variables
in all periods.
In the first step, we set a most simple case. We assume symmetric countries: same population size
(L = L∗ ), same utility functions (γ = γ ∗ , σ = σ ∗ ), same technologies in both the producing sector
(z = z ∗ ) and the innovation sector (η = η ∗ ), and same entry costs (fD = fD ∗ , f = f ∗ ).
E E

We also assume that

• In the foreign country, there is no tax: τ s∗ = τ w∗ = τ e∗ = τ c∗ = 0. Thus no public debt

before and after the reform: δ ∗ = 0.

• In the domestic country, we assume all labor taxes are equal to zero initially: τ w = τ e = τ s =
0, so as well as public debt: δ = 0. Then, the consumption tax is also initially equal to zero:
τ c = 0.

The fiscal devaluation starts from a simultaneous decrease in the three labor tax rates τ w , τ e and
τ s : dτ w = dτ e = dτ s < 0. Keeping the public debt δ at zero, the resulting consumption tax rate
τ c can then be deduced from the government budget constraint in the home country.

Proposition 1. With the assumption of symmetric countries, and assuming that fiscal instruments

satisfy the above restrictions, the steady state equilibrium values of prices r, Q, ω, ω ∗ , ρH , ρF , ρH∗ , ρF ∗ ,

12
quantities (C, C ∗ , b, b∗ , N, N ∗ ) and consumption tax rate τ c satisfy the following relationships:
w (1+τ w )ω

τ c = − τB , Q = 1 − τ c, (1−τ c )ω ∗ =1 





1

σ ω∗

ρH = ρH∗ = ρF = ρF ∗ = ρH = 1


σ−1 z ,
1−σ


2N 

  (20) SS
1−γ γ C∗ γ
1
1− C
b∗ = ∗ 
1+r = γ B , b= 1−γ Bω, ω∗ = ω = B, 1−γ BQω







   

B+τ w
N = N∗ = L 1 σ−1 
η+fD +fE 1− 2 σ 1+τ w +B 

where
σ−1 τ w
1 + τw 2− σ 1+τ w
B= (21) B
2 + τw
 
γ fD +fE
1−γ 1+ η + σ−1
σ

Proof.

• From the government budget constraint in country H

0 = τ c LC + τ w Lω

one gets
τw
τc = −
B
where
C
B≡
ω
• Consider the static relationships between prices.
We deduce
σ 1 + τw ω 1 + τw ω F
ρH = = ρ
σ − 1 1 − τc z Q ω∗
σ ω 1 + τw ω F∗
ρH∗ = (1 + τ w ) = ρ
σ−1 Qz Q ω∗
Replacing into

1−σ
1−σ
1 = N ρH + N ∗ ρF

1−σ
1−σ
1 = N ρH∗ + N ∗ ρF ∗

yields
" 1−σ # " 1−σ #
1 + τw ω 1 + τw ω
 
F 1−σ
∗ ∗

1−σ
ρF ∗


N +N ρ =1= N +N
Q ω∗ Q ω∗

13
 Consequently,
σ ω∗
ρF = ρF ∗ = (22) rhoF
σ−1 z
σ 1 + τw ω
ρH = ρH∗ = (23) rhoH
σ − 1 1 − τc z
Q = 1 − τc (24) Q
"  #
1 + τ w ω 1−σ


1−σ
1= N ∗
+ N ∗ ρF (25) N_Nstar
Q ω

• From the household saving behavior:

γ
b = γ [(1 + r) b + ω] = C
1−γ
γ
b∗ = γ [(1 + r) b∗ + (1 − τ c ) ω ∗ ] = (1 − τ c ) C ∗
1−γ
we get
−1
b∗ γ C∗

γ b 1
B= = − (1 + r) = =
1−γ ω γ Qω ∗ 1 − γ ω∗
and deduce   
1−γ 1 1−γ
B= 1−γ(1+r) ⇔1+r = γ 1− B








γ
b= 1−γ Bω



(26) SavSS
C∗ C

= =B


ω∗ ω 





γ
b∗ = ∗ 
1−γ BQω

Labor_homeLabor_foreign
• Labor market equilibria (11) and (12) rewrite
η + fD + fE
−σ C + C ∗
N = 1 − N ρH (27) LabM_H
L z
η + fD + fE −σ C + C∗
N∗ = 1 − N ∗ ρF


(28) LabM_F
L z

• The international capital market equilibrium becomes

η
b + b∗ = [N (1 + τ w ) ω + N ∗ (1 − τ c ) ω ∗ ] (29) CapMbis
L

• Free entry conditions become

C + C∗
 
η fD + fE
1−σ
(1 + r) + (1 + τ w ) ω = ρH (1 − τ c ) (30) FE_H
L L σ
∗
 
η fD + fE 1−σ C + C
ω ∗ = ρF


(1 + r) + (31) FE_F
L L σ
They imply
1−σ
ρH (1 + τ w ) ω

=
ρF (1 − τ c ) ω ∗

14
 rhoF rhoH
Using (22) and (23), one gets
(1 + τ w ) ω
=1
(1 − τ c ) ω ∗
Consequently
ρH
=1
ρF
FE_H FE_F
Moreover, from (30) and (31), we deduce

N = N∗

• Summary:

τw (1 + τ w ) ω
τc = − , Q = 1 − τ c, =1
B (1 − τ c ) ω ∗
σ ω∗
ρH = ρH∗ = ρF = ρF ∗ = , N = N∗
σ − 1 z
C∗
 
1 1−γ γ C γ
1+r = 1− , b= Bω, = = B, b∗ = BQω ∗
γ B 1−γ ω∗ ω 1−γ
N_Nstar
LabM_H CapMbis
where B, N , and ρH are solutions of the following system (using (25), (27) and (29))

1−σ
1 = (2N ) ρH (32) N_Nstar2
τw
!
η + fD + fE 1 σ − 1 H
1−σ 1+ B
= − ρ B +1 (33) LabM_H2
L N σ 1 + τw
γ 2 + τw η
B w
= 2N (34) CapMbis2
1−γ 1+τ L
The preceding 3 equations can be rewritten as

1−σ 1
ρH =
2N
w
!
η + fD + fE 1σ−1 1 + τB
N =1− B +1
L 2 σ 1 + τw
2 + τw 1 σ − 1 B + τw
  
γ η
B =2 1− +B
1 − γ 1 + τw η + fD + fE 2 σ 1 + τw

Then we get
σ−1 τ w
1 + τw 2− σ 1+τ w
B=
2 + τw
 
γ fD +fE
1−γ 1+ η + σ−1
σ
w
!
η + fD + fE 1σ−1 1 + τB
N =1− B +1
L 2 σ 1 + τw

1−σ 1
ρH =
2N

15
 FE_H
• Notice that the free-entry equation (30) can be deduced from the other equations. Indeed, it
is equivalent to

ω + ω∗
 
η fD + fE
1−σ
(1 + r) + (1 + τ w ) ω = ρH (1 − τ c ) B
L L σ
w
!
η η + fD + fE 1 1 + τB
⇔ rN + N = B +1
L L 2σ 1 + τw

1−σ 1
since ρH = 2N .
LabM_H2
CapMbis2
Using (33), (34) and  
1−γ 1
r= 1−
γ B
the free-entry equation becomes equivalent to
w w
! !
2 + τw 1 + τB 1 + τB
 
1−γ 1 1 γ 1σ−1 1
1− B w
+1− B +1 = B +1
γ B 21−γ 1+τ 2 σ 1 + τw 2σ 1 + τw

w
!
1 1 2 + τw 1 + τB
 
1
⇔ 1− B +1− B +1 =0
B 2 1 + τw 2 1 + τw

2 + τw 1 B + τw
 
1
⇔ (B − 1) +1− +B =0
2 1 + τw 2 1 + τw
− (2 + τ w ) τw
⇔ + 1 − =0
2 (1 + τ w ) 2 (1 + τ w )
which is always true.

From the above summary of equations
in session 4, we can not only get the equilibrium values
of prices r, Q, ω, ω , ρ , ρ , ρ , ρ , quantities (C, C , b, b∗ , N, N ∗ ), but also the amount of con-
∗ H F H∗ F ∗ ∗

sumption (xH , xF , xH ∗ , xF ∗ ), utility of consumers on both domestic and foreign market (U H , U F )

and trade balance of both countries.

• Individual consumption of domestic and foreign goods

−σ
pH

x H
= C = (ρH )−σ C
P
−σ
pF

x = F
C = (ρF )−σ C
P
 H ∗ −σ∗
H∗ p
x = C ∗ = (ρH ∗ )−σ∗ C ∗
P∗
 F ∗ −σ∗
F∗ p
x = C ∗ = (ρF ∗ )−σ∗ C ∗
P∗

16
 • Utilities of consumers on both markets
To compare the utilities of consumers on the two markets, it is equivalent to compare the
following two variables:
for consumers on home market:

U H = (1 − γ) ln C + γ ln b

and for consumers on foreign market:

U F = (1 − γ ∗ ) ln C ∗ + γ ln b∗ Q

We set parameter values as follows:

Parameter Values
σ 2.00 σ∗ 2.00
η 0.2558 η ∗ 0.2558
fD 0.0030 fD ∗ 0.0030
fE ∗
0.0037 fE 0.0037
L 1.00 L∗ 1.00
γ 0.30 γ∗ 0.30
z 1.00 z ∗ 1.00
τ c∗ 0 τ w∗ 0
τ e∗ 0 τ s∗ 0

We suppose the two markets are symmetric, so that the parameters with stars are the same as
those in home country. We set the total labour supply L to be 1, σ = 2, and bequest weights
30% in households’ utilities (γ = 0.30). In the innovation sector, η = 0.2558, this means each
blueprint requires 0.2558 unit of labour input in every period. In the producing sector, z = 1,
labour productivity is 1 in both markets. fD = 0.0030 and fE = 0.0037, which means the fixed
cost to enter domestic market and foreign market are 0.003 and 0.0037 unit of labour for each
firm. We first imply a one time labour subsidy, where labour tax rate decreases from τ w = 0 to
τ w = −0.1. The outcomes are shown in the following table.

17
 variables τ w = 0.0000 τ w = −0.1000 change
τc 0.0000 0.0965
Q 1.0000 0.9035 -0.096520
δ∗ 0.0000 0.0000
N 1.7827 1.8323 0.027778
N∗ 1.7827 1.8323 0.027778
ρF 3.5655 3.6645 0.027778
ρF ∗ 3.5655 3.6645 0.027778
ρH 3.5655 3.6645 0.027778
ρH ∗ 3.5655 3.6645 0.027778
r 0.1405 0.0812 -0.421902
ω 1.7827 1.8393 0.031752
ω∗ 1.7827 1.8323 0.027778
C 1.8969 1.9057 0.004601
C∗ 1.8969 1.8983 0.000731
B 1.0641 1.0361
b 0.8130 0.8167 0.004601
b∗ 0.8130 0.7350 -0.095859
xH 0.1492 0.1419 -0.048968
xF 0.1492 0.1419 -0.048968
xH ∗ 0.1492 0.1414 -0.052632
xF ∗ 0.1492 0.1414 -0.052632
ΠH 0.2660 0.2349 -0.116894
ΠF 0.2660 0.2600 -0.022551
ΠH ∗ 0.2660 0.2340 -0.120295
ΠF ∗ 0.2660 0.2590 -0.026316
N etΠH 0.2607 0.2300 -0.117827
N etΠF 0.2607 0.2545 -0.023583
N etΠH ∗ 0.2594 0.2279 -0.121538
N etΠH ∗ 0.2594 0.2522 -0.027691
UH 0.3861 0.3906 0.011890
UF 0.3861 0.3259 -0.155858
NetSH 0.0000 0.0408
NetSF 0.0000 -0.0408
TBH 0.0000 -0.0033

We then imply a continuous labour subsidy increase of 0.01 for 25 steps. This can allow the interest
rate to stay positive, which is more interesting to analyze. The outcomes are shown in the following
graphs.

18
The outcomes confirm our analysis of the innovation in the former paragraphs. The development
of innovation sector will induce spillover effect, thus decrease the cost and introduce more varieties
at the same time. This is reflected by a growing number of commodities in the market. With
the implement of fiscal devaluation, the numbers of firms increase in both markets, showing that
the innovation sector performs well. With a more multiple choices, consumers with a preference
of more balanced consumption basket will decrease their individual consumption of each variety.
This is observed in both home and foreign markets and for both domestic and import goods (shown
in the 5th line). Facing the pressure of consumption tax increasing and individual consumption

19
decreasing, firms turn to increase their prices to protect the profits. This can also be interpreted as
the fiscal base is not large enough to cover the labour subsidy, requiring an increase of consumption
tax and higher prices at the same time. However, the increase of prices are much smaller than
the increase of consumption tax, which means that part of the ’cost’ of fiscal devaluation has been
paid by the the firms and only part of the tax burden has been transferred to the consumers. And
this can also help explain why the individual profit of firms on both markets decrease when the
level of fiscal devaluation improves, which is shown in the second graph. Another reason why the
price increases is that the labour cost, real wage increases. Since there are more firms entering the
market, the total demand for labour in innovation sector and supporting parts is increasing, while
the total labour supply is fixed, this not only lead to a lower labour stock in the producing sector,
but also an increase in the real wage. The lower labour stock in the producing sector, increasing
prices and increasing number of firms and decreasing individual consumption are consist with each
other. Besides, since the foreign goods on home country are more expensive due to the increase
of consumption tax and domestic goods are more cheaper on foreign market thanks to the labour
subsidy, the general price index of foreign market decreases and the one of home country increases.
This leads to a decrease of real exchange rate Q, which is also consist with the expression Q = 1−τ c .
Another interesting finding is that the implement of fiscal devaluation differentiates the allocation
of consumption and investment of the two countries. The real wage increase in both markets
during the fiscal devaluation process due to the increase demand of labour. The change of total
consumption is a balanced outcome of number of firms, individual consumption and individual
prices. For home market, the bequest is changing at a fixed ratio to the total consumption, so that
it is easy to understand that the bequest and utility change in the same way as total consumption
does in home country. We find that when τ w reaches around -0.15, the utility for households
in home country reaches its peak, this is an ideal level. As for foreign households, the bequest
changes at a fixed ratio to the total consumption multiplied by the real exchange rate. Since the
real exchange rate decreases a lot, the foreign bequest also falls seriously, leading the consumer
utility decreasing continuously. We take the home market capital market for example since it is
no need to consider the real exchange rate. Due to an increase supply of capital (real asset), the
price of capital (real interest rate) is decreasing. And since domestic consumption curve is concave
while the foreign one is convex, the domestic consumption increases greater and faster than the
foreign ones, leading to a negative trade balance. It finally recover to positive due to the change of
relationship between consumptions on both market.
Now, let us begin to release the assumption that labour taxes are all equal to zero. We assume
that in foreign country, subsidy for working labour in the producing sector is τ w∗ = −0.1, for
labour in the innovation sector is τ e∗ = −0.11, for labour in the supporting parts is τ s∗ = −0.09,
and consumption tax is τ c∗ = 0.1, thus we can get the first graph. Then, we modify the labour
subsidy in home country. τ w , τ e and τ s still start from zero, but we subsidize more on the labour in
innovation sector and less in the supporting parts. In each term, the labour subsidy of innovation
sector (τ e ) is 1.1 times of producing sector, and the subsidy for supporting sectors (τ s ) is 0.9
times.The outcomes are shown in the second graph.

20
Similar to the basic case, the consumption tax rate, real exchange rate, number of firms, prices and
individual consumptions change in the same way. What is different is that the total consumption

21
has not reach its peak in this two settings and the consumer utility in home country is able to
reach a higher level, where the case with a differenticated labour subsidy is more benefitial for
home households. As to the foreign country, the fiscal revenue keeps increasing and it reaches a
higher level when the subsidy is different among labour types. This outcomes indicated that a
differentiated fiscal devaluation gives positive outcome for both markets.
Now we keep these changes in taxes and suppose the households on two markets have different
weights on the bequest in their utilities. Withγ ∗ = 0.3 and γ ∗ = 0.4, it describes a case that foreign
households care more on the bequest they can give to the next generation tan home households.
Here we got the outcomes:

An interesting finding is that consumer utility for foreign households is the only variable that
changed. With a higher weighting on bequests, the utility for households decreases in foreign
country. This is because that real asset in F drops sharp than the one of home country.
Based on the basic case, we then modify the foreign country as a

22
 A Consumer
consumer

Eliminating λt in the two above equations, one gets

1−γ 1 1 γ
(Ct ) σ (xt (ω))− σ = pt (ω) (35) eq1
Ct at+1

Multiplying both sides by xt (ω) and summing over ω, we deduce

Et
1−γ =γ
at+1

Combining with the budget constraint of the consumer, we get

γ
at+1 = γ ((1 + it )at + wt ) = Et (36) a_t
1−γ
eq1 P_t a_t
Let us divide equation (35) by Pt (defined as in equation (3)). Using equality (36), we get
  −1
1 xt (ω) 1 pt (ω)
σ
= (37) eq2
P t Ct Ct Et Pt
 −σ
σ xt (ω) σ pt (ω)
⇔ (Pt Ct ) = (Et )
Ct Pt
 1−σ
σ pt (ω)xt (ω) σ pt (ω)
⇔ (Pt Ct ) = (Et )
P t Ct Pt

Summing over ω implies

(pt (ω))1−σ dω
R
σ Et
(Pt Ct ) = (Et )σ ω
P t Ct Pt1−σ
and leads to
Et = Pt Ct
a_t eq2 x_expression
a_tC_t
Then (36) and (37) lead to (4) and (5).

B Net External Liability and Net Export

Summing the budget constraints of the consumers in country H leads to

LPt Ct + Lat+1 = (1 + it ) Lat + Lwt (38) AgBC

where
∗ F F
Pt Ct = Nt pH H
t xt + Nt pt xt

From the labor market equilibrium in country H

Lxt + L∗ xH∗
 H 
t
L = Nt + fD + fE + Nt+1 η
z

23
 AgBC
one can rewrite equation (38) as follows

LxH ∗ H∗
 
t + L xt
Nt pH H
t Lxt + Nt∗ pFt LxFt + Lat+1 = (1 + it ) Lat + wt Nt + fD + fE + wt Nt+1 η
z
⇔
  H    ∗ H∗ 
Lxt H∗ ∗ H∗ L xt
Nt pH
t LxH
t − w t + fD + Nt p t L x t − w t + fE
z z
+ Nt∗ pFt LxFt − Nt pH∗ ∗ H∗
t L xt + Lat+1 − wt Nt+1 η = (1 + it ) Lat (39) AgBC2

Profits earned by H-firms are

(1 + τtw ) wt H
πtH = (1 − τtc ) pH H
t Lxt − Lxt
z
∗ H∗ (1 + τtw ) wt ∗ H∗
πtH∗ = (1 − τtc∗ ) pH∗
t L xt − L xt
z
These equations are equivalent to
τtw wt H
 
 H s c H H wt H
Lxt + τt wt fD = pH
s H

πt − (1 + τt ) wt fD + τt pt Lxt + t Lxt − Lxt − wt fD
z z
τtw wt ∗ H∗
 
 H∗ ∗ H∗ H∗ ∗ H∗ wt ∗ H∗
πt − (1 + τts ) wt fE + τtc∗ pH∗ s

t L x t + L x t + τt t E = pt L xt −
w f L xt − wt fE
z z
AgBC2
Then (39) rewrites

Nt πtH − (1 + τts ) wt fD + πtH∗ − (1 + τts ) wt fE

 

τtw wt H τtw wt ∗ H∗
 
c H H
+ Nt τt pt Lxt + s
Lxt + τt wt fD + L xt + τt wt fE + Nt∗ τtc pFt LxFt
s
z z
+ Nt∗ (1 − τtc ) pFt LxFt − Nt (1 − τtc∗ ) pH∗ ∗ H∗
t L xt + Lat+1 − wt Nt+1 η = (1 + it ) Lat (40) AgBC3

The free-entry condition and the government budget constraint imply

πtH − (1 + τts ) wt fD + πtH∗ − (1 + τts ) wt fE = (1 + it ) η 1 + τt−1

e


wt−1

τtw wt H τtw wt ∗ H∗
 
Nt τtc pH H
t Lxt
+ s
Lxt + τt wt fD + L xt + τt wt fE + Nt∗ τtc pFt LxFt
s
z z
= (1 + it )∆t − ∆t+1 − τte Nt+1 ηwt
AgBC3
Then (40) becomes
e


(1 + it ) Nt η 1 + τt−1 wt−1 + (1 + it )∆t − ∆t+1
+ Nt∗ (1 − τtc ) pFt LxFt − Nt (1 − τtc∗ ) pH∗ ∗ H∗ e
t L xt + Lat+1 − (1 + τt ) wt Nt+1 η = (1 + it ) Lat

Lat+1 − (1 + τte ) wt Nt+1 η − ∆t+1

e


= (1 + it ) Lat − 1 + τt−1 wt−1 Nt η − ∆t
∗ H∗ ∗
+ Nt (1 − τtc∗ ) pH∗ c F F
t L xt − Nt (1 − τt ) pt Lxt

24
Notice that Nt (1 − τtc∗ ) pH∗ ∗ H∗ corresponds to after-tax exports (exports after consumption
t L xt
tax payments in the foreign country), while Nt∗ (1 − τtc ) pFt LxFt is after-tax imports (imports after
consumption tax in the home country). The difference between these two terms is the (after-tax)
trade balance.
e


Let us call Lat − 1 + τt−1 wt−1 Nt η −∆t , the net external liability of country H (difference between
households’ assets and the debt level of the firms and the government).
The net external liability at the end of period t is equal to the one at the end of period t − 1, plus
interest payments, plus the trade balance.
To get the preceding equation in real terms, let us divide by Pt on each side:

Lbt+1 − (1 + τte ) ωt Nt+1 η − δt+1

e


= (1 + rt ) Lbt − 1 + τt−1 ωt−1 Nt η − δt
∗ H∗ ∗
+ Nt (1 − τtc∗ ) ρH∗ c F F
t L xt Q − Nt (1 − τt ) ρt Lxt

since 1 + rt = (1 + it ) PPt−1
t

In steady state, the preceding equality becomes

r [Lb − (1 + τ e ) ωN η − δ] + N (1 − τ c∗ ) ρH∗ L∗ xH∗ Q − N ∗ (1 − τ c ) ρF LxF = 0

The real trade balance is the opposite to real interest payments on the net external liability.
If the net external liability is negative, then (i) the trade balance is positive, (ii) real interests paid
to the foreign agents are equal to the trade real balance.

25