Monetary Policies for Developing Countries:

The Role of Institutional Quality

Haizhou Huang (Barclays Capital) and Shang-Jin Wei (IMF, NBER and CEPR)
Revised August 29, 2005

Abstract
Weak public institutions, including high levels of corruption, characterize many developing countries. We demonstrate that this feature has important implications for
the design of monetary policymaking institutions. We nd that a pegged exchange rate
or dollarization, while sometimes prescribed as a solution to the credibility problemy,
is typically not appropriate for countries with poor institutions. Such an arrangement
is inferior to a Rogo-style conservative central banker, whose optimal degree of conservatism is proportional to the quality of institutions. Finally, we cast doubt on the
notion that a low inationary framework can induce governments to improve public
institutions.
JEL Classication Numbers: E52, E58, E61, E62, H50
Keywords: Corruption, central bank, ination, exchange rate

Corresponding author: Shang-Jin Wei, phone: (202) 623-5980, fax: (202) 589-5980, and swei@imf.org.
Mailing address for both authors: Research Department, International Monetary Fund, 700 19th Street
NW, Washington DC 20431, U.S.A. A longer version of the paper can be found as CEPR discussion paper
4911, February 2005. The authors would like to thank Peter Clark, Ke-young Chu, Anne-Marie Gulde, Paul
Masson, Sunil Sharma, Danyang Xie, seminar participants at HKUST and IMF, and participants at the
2002 LACEA meetings for comments; two anonymous referees and the editor Charles Engel for very helpful
suggestions; and Yuanyuan Chen and Hayden Smith for capable research assistance. The views expressed
are those of the authors, and do not represent those of their a liated institutions.

Introduction

Textbook discussions of monetary policies do not usually separate developing from developed
countries. Are there important features about developing countries that might suggest that
the optimal design of monetary policies should be dierent? In this paper, we study one
particular feature that is prevalent in developing (and transition) economies, namely weak
public governance. Obviously, developed countries are not immune to this problem, but it
is far less prevalent than in many developing countries. Surprisingly, the consequence of this
feature on the design of monetary policy has not been systematically examined. This paper
aims to ll this void, and to demonstrate that the eect is not trivial.
As many developing countries lack credibility in their monetary policy, a subject heavily
studied in the literature, a conventional wisdom is that these countries should peg their
currency to a major currency from a low-inationary country, adopt a currency board, or
dollarize. Our analysis in this paper, however, will show that when weak institutions are
considered these policies are not necessarily appropriate.
Our theory combines useful ingredients from two dierent strands of the literature.
The rst strand is on the design of monetary policy, which is too voluminous to be referenced completely here, but includes, as seminal and other important contributions, Kydland
and Prescott (1977), Calvo (1978), Barro and Gordon (1983), Rogo (1985), Barro (1986),
Alesina and Tabellini (1987), Cukierman (1992), Svensson (1997), Walsh (1995), and Benigno and Woodford (2003).1 In this paper, we make use of a framework developed by
Alesina and Tabellini (1987), where the governments objective function includes provision
of public goods in addition to minimizing ination and output uctuations. The second
strand studies the causes and consequences of weak institutions, in particular, corruption.
This literature includes work on the eects of institutions on development (Rose-Ackerman,
1975, Shleifer and Vishny, 1993, and Mauro, 1995). Wei (2000 and 2001), Bai and Wei
(2000), Fisman and Wei (2004) and Du and Wei (2004) investigated the consequences of
corruption for international capital ows, tax evasion, and stock market volatility. As far as
we know, these two strands of the literature have not been married before. In other words,
none of the papers in the literature that we know of has examined the implications of weak
institutions, including widespread corruption, for the design of monetary policies.
For the purpose of our paper, we model weak institutions as an erosion of a governments
ability to collect revenue through formal tax channels. This may arise through outright theft
1

See Persson and Tabellini (1990) and Berger et al (2001) for surveys of the literature.

by tax o cials or practices whereby tax inspectors collude with taxpayers to reduce the
latters tax obligation in exchange for a bribe. Under an ination targeting framework, we
study how the socially optimal level of the ination target is aected by weak institutions.
We further examine the implications for the design of several other monetary frameworks,
including a currency board, dollarization, and a Rogo-type conservative central banker,
and rank them in terms of social welfare. We also examine the authorities incentive in
strengthening institutions from a political economy perspective.
Several interesting results emerge from the analysis. First, the optimal ination target
is higher for a country with poorer institutional quality. Hence, an ination target of 1-4
percent, that is common among advanced industrialized countries and might be called international best practice,is generally not something to be emulated by developing countries.
Second, pegged exchange rate, currency boards, or dollarization are often prescribed as
ways to solve the lack of credibility problem. However, these monetary regimes are typically
not very credible themselves and are likely to fail (often associated with a currency crisis)
in countries with weak institutions.
Third, a Rogo-type conservative central banker is generally preferable to a mechanical
ination target of 1-4 percent and to most exchange-rate-based monetary arrangements. In
equilibrium, the optimal degree of central bank conservatism is proportional to institutional
quality. Thus, developing countries with lower institutional quality should have less conservative central bankers, and in the limit, when weak institutions make collection of tax
revenue infeasible, the optimal degree of conservatism is zero.
Fourth, we consider the political economy of strengthening institutions. In particular, we
ask whether forcing a government not to rely too much on the ination tax through external
pressure (e.g., conditionality in an IMF program) could induce it to improve institutional
quality, e.g., to ght corruption. The answer is probably not. One interesting result is a
poor-institution trap. That is, when the initial quality of institutions is su ciently low, it
would be di cult to induce the authorities to devote any eort to strengthen them.
The paper proceeds as follows. Section II sets up the model. Section III compares various
popular frameworks that implement a commitment regime, namely ination targeting, xed
exchange rates, currency boards, and dollarization. It is shown that the relative desirability of
these frameworks depends on the institutional quality. Section IV analyzes the discretionary
regime and examines a conservative central banker framework. Section V extends the basic
model to allow for a Laer curve eect in seigniorage revenue. Section VI endogenizes the

institutional quality from a political economy perspective. Section VII concludes.

Basic Setup

Our model utilizes a framework developed in Alesina and Tabellini (1987), which we think
has been insu ciently appreciated in the literature. The governments objective function
includes public goods provision in addition to stabilizing ination and output:
V( ; )=

1
2

+ ky 2 + l(g

g)2 ;

(1)

where denotes the ination rate, y the log of real output, and g the ratio of expenditure on
public goods to output. k > 0 and l > 0 are the weights on output and public expenditure
stabilities, respectively. In this objective function, the target levels for ination and output
are normalized to zero. In addition, the government aims to minimize the deviation of public
goods provision from a nonnegative target g.
To generate an ination bias under a discretionary regime, the original Barro-Gordon
(1983) model has to assume that a governments targeted output level is systematically above
the long-run equilibrium. An interesting property of the Alesina-Tabellini reformulation is
that the need to provide public goods (g > 0) is enough to generate an ination bias. This
is demonstrated below.
For simplicity, we consider a deterministic
mand. A modied Lucas supply curve governs
and government policies: unexpected monetary
distortionary tax rate reduces aggregate supply.

economy with no shocks to aggregate dethe relationship between aggregate output

growth increases aggregate demand, but a
To be more precise, output is given by:
e

y= (
where e is the expected ination rate, and
> 0 are coe cients.2

is the tax rate on total output.

(2)
> 0 and

To nance the public goods provision, the government has two sources of revenue: an
output tax ; and an ination tax . There is ample evidence suggesting that seigniorage is an
important source of government revenue for developing countries. For example, Cukierman,
2

Equation (2) implicitly assumes that money demand is not aected by scal policy and, therefore, that
scal policy is not subject to time inconsistencies. Otherwise, an independent central bank could not directly
control ination, since it would be jointly determined by the money supply and the tax rate.

Edwards and Tabellini (1992, Table 1) show that over 1971-1982, seigniorage (dened as an
increase in base money) as a share of total government revenue could account for more than
ten percentage points (e.g., 21.6% for Bolivia, 28.0% for Ghana, 13.1% for India, 23.9% for
Mexico, and 24.8% for Uganda).
A crucial assumption that we make is a connection between the governments scal
capacity and the quality of institutions. More precisely, weak institutions (e.g., corruption)
are assumed to cause a leakage of the tax revenue: the lower the institutional quality, the
greater the leakage. If the private sector pays a tax in the amount of ; only
accrues to
the government, where 0
1.
can be thought of as an institution-quality index. If
= 1, then the quality is the best and there is no leakage of tax revenue.3 If = 0, then
there is complete leakage, and the government collects no tax revenue.4 The governments
budget constraint can be written as:5
g=

(3)

Note that when = 1, our specication becomes that in Alesina and Tabellini (1987).
Our model abstracts from public debt and leakage in the collection of ination tax.

Commitment and Its Implementations

3.1

The Commitment Regime

We consider an institutional setup in which monetary and scal authorities each control a
single policy instrument (an ination rate, , by the central bank, and a tax rate, , by the
scal authority), but share a common objective function dened by Equation (1). The two
branches of the government solve a noncooperative game. The equilibrium ination and tax
rates are given by the Nash equilibrium of the game.6
3

Here, the normal administrative cost for tax collection is not considered a leakage.
Although we focus on corruption as the main reason for tax leakage, weak institution and its associated
low scal capacity can be attributed to other factors, such as a large informal sector in many developing
countries, a tradition of outing government regulations, little experience with voluntary compliance, and
constantly changing procedures in transition economies.
5
Equation (3) can be obtained from a two-step derivation as in Alesina and Tabellini (1987). First, the
government budget constraint in nominal terms is: Gt =
Mt 1 , where G denotes public
t Pt Xt + Mt
spending, P price level, X real output, and M equilibrium money supply, respectively. Second, dividing both
sides by nominal income Pt Xt and approximating (Mt Mt 1 ) =Pt Xt as t , we have gt = t + t .
6
In this setting, a cooperative game would yield the same result.
4

This subsection focuses on the case in which the central bank sets an ination rate and
can credibly commit to it. In this case, y =
. The Nash equilibrium monetary and
scal policies can be directly obtained from solving two rst-order conditions, i.e., the two
reaction functions.7 This yields the equilibrium ination and tax rates:
C

kl 2 g
=
(1 + l) k 2 + l

2;

(4)

l g
(1 + l) k 2 + l

2:

(5)

A number of observations can be made. First, if there is no need to provide public goods
(g = 0), then the equilibrium ination and tax rates under the commitment regime would
be zero. Second, it is straightforward to see that the ination rate goes up as the quality of
institutions becomes worse ( goes down). As the shadow cost of raising revenue through
regular tax channels rises vis--vis the ination tax, a higher ination becomes optimal.
Third, as institutional quality worsens, the tax rate C can go either up or down, depending
on the combination of other parameters.8
The equilibrium levels of public expenditure and social welfare are g C =

(k 2 + 2 )lg
(1+l)k 2 +l 2

2 2

kl g
; respectively. Therefore, lower institutional quality unambiguously
and V C = 21 (1+l)k
2
+l 2
reduces social welfare.

Proposition 1 Under a commitment regime, as the institutional quality becomes poorer,

(1) the equilibrium ination rate goes up; (2) the tax rate can go either direction, but (3) the
social welfare declines unambiguously.

3.2

Comparing ination targeting, exchange rate xing, currency

board, and dollarization

Four popular frameworks have been developed to implement the commitment regime: ination targeting, exchange rate xing, currency board, and dollarization. It is easy to compare
their relative desirability based on the insights from this model.
7

The second-order conditions associated with this problem (as well as with the discretionary case later)
are trivially satised since V ( ; ) is globally concave.
8
See Huang and Wei (2005) for more discussion.

Ination targeting is a monetary arrangement in which the central bank announces (or
is asked to follow) a target level (or range) for the ination rate.9 In principle, this system
can be used to achieve the desirable outcome ( C , C and g C ).
A number of countries, including Australia, Brazil, Canada, Finland, New Zealand,
Norway, Poland, Sweden, and the United Kingdom, have adopted ination targeting. In
practice, they either target their ination rates to a point (e.g., U.K.) or to a relatively
narrow range, typically within 1-4 percent. It is often thought that a similar level of ination
target would benet developing and transition economies as well. The empirical evidence,
however, shows that ination targeting has been less successful in developing and transition
economies. In fact, many of them are reluctant to adopt it, even though a lack of credibility
is a clear concern for them. We believe that the poorer quality of institutions provides one
important reason.10
It is useful to make a distinction between a mechanical ination target and an optimally
chosen target. A mechanical ination targeting is a framework that advocates developing
countries to do what developed countries are doing, namely to target a low ination rate
such as 4 percent (or a narrow range around that). An optimal ination targeting is an
arrangement that is consistent with the optimal ination level discussed earlier. An immediate implication is that the poorer the institutional quality, the higher the optimal level of
the ination target. A mechanical ination target could reduce the welfare of countries with
lower quality institutions.
Under either a xed exchange rate or a currency board arrangement, there is an implied
ination target which is the anchor countrys ination rate. Generally speaking, the anchor
country tends to have higher quality institutions than most developing countries. Therefore,
a developing economy with lower institutional quality tends to have a lower ination rate
than is optimal, and a higher tax rate than is optimal.11 Its welfare is thus lower than it
could be.
9

See Bernanke and others (1999) for recent international experience of ination targeting.
Masson, Savastano, and Sharma (1997) and Eichengreen, Masson, Savastano, and Sharma (1999) stated
that a monetary authority free of scal dominanceis a precondition for the success of an ination targeting
regime. Our model can be viewed as a formalization of this argument. Cukierman (1992, p. 445-452)
suggested that limited access to capital market by developing country governments is another possible
explanation for the apparent reluctance in adopting the system.
11
In the context of this model, a crawling peg of the type used in Chile and Israel in the past is better
than a conventional currency board for a high-corruption country, as it allows for more seignorage revenue.
See Ghosh, Gulde, and Wolf (2003) for more discussions on currency board and its problems.
10

It is sometimes thought that developing countries can import credibility through a

xed exchange rate or currency board. Our discussion suggests that weak institutions may
limit or even destroy the imported credibility as the implied ination rate is too low and the
incentive to deviate from it is too strong.
We have so far assumed away stochastic shocks to the aggregate Phillips curve. Without
these shocks, a xed exchange rate, a currency board, and a mechanical ination target are
qualitatively equivalent. However, if shocks are introduced, an ination targeting framework
can dominate the other two arrangements as it allows for the exibility to respond to shocks
that are specic to the domestic economy. As noted in Cukierman et al (2004), an exchange
rate band also permits an optimal tradeo between exibility and commitment.
Dollarization, or more generally, the adoption of a foreign currency, is a monetary
arrangement that involves an even stronger commitment to low ination assuming that
the anchor country has low ination than a currency board arrangement. Unlike a currency board, the national currency disappears. The commitment is stronger because the
cost to the government of reversing such an arrangement is higher. While the underlying
ination rates of the two regimes are the same, the government in a dollarization regime
has to forgo seigniorage revenue associated with the issuance of domestic money. Hence, the
tax rate under dollarization is higher, but the social welfare is lower, than under a pegged
exchange rate or currency board.
Proposition 2 The optimal commitment regime dominates a mechanical ination target,
which (weakly) dominates a xed rate or currency board, which in turn dominates dollarization.

4
4.1

Discretion and Conservative Central Banker

Conventional Discretionary Regime

If a central bank cannot precommit, the ination rate (and correspondingly the tax rate)
derived for a commitment regime would not be time consistent. As is well known in the
literature, if the expected ination were at the commitment level ( e = C ), the central
bank would always nd it optimal to raise ination unexpectedly. Hence, such ination
expectation would not be rational. The time-consistent policy mix, ( D ; D ), is the Nash

equilibrium solution to the noncooperative game played by the central bank and the scal
authority, who take the expected ination rate as given.
The solution is characterized by the two rst-order conditions associated with (1), with
the requirement that the expected ination rate equals its equilibrium value. Solving the
two rst-order conditions, we have a Nash equilibrium policy mix:
D

=
D

where

=(

+ )

g
)

g
k

(1
=

+ g

+g

(6)
(7)

1:

C
C
It is easy to verify that D
, and D
, with the equality holding if and
only if
= 0. Thus, the ination under discretion is higher than under commitment,
and the tax rate is lower. It is straightforward to work out the level of the social welfare,
[kl( + )2 +k2 2 +l2 2 ]kl 2 g2
V D = 12
, as well as the public goods provision and the output.
2
[kl ( + )+k 2 +l 2 ]

From (7), it is clear that the derivative of the discretionary tax with respect to has
the same sign as the derivative of the commitment tax with respect to . For moderate
p
institutional quality ( =
(1 + l) k=l), an optimal response to a decrease in quality is to
p
raise the tax rate. But at a poor quality level ( = < (1 + l) k=l), the optimal response
would be to lower the tax rate.
From (6), one sees that a decrease in has two osetting eects on discretionary ination:
it increases the commitment ination level C but decreases the inationary bias captured
by . Thus in contrast to the commitment case, the optimal response of monetary policy
in discretion to a decrease in institutional quality is nonlinear. If the quality of institutions
p
is relatively modest ( =
1 + k 2 =l 1), then the optimal response to a decrease in
quality is to raise the ination rate. On the other hand, if the quality of institutions is
p
already poor ( = < 1 + k 2 =l 1), then the opposite response would be optimal.
This makes an interesting comparison with the commitment case. For example, starting
p
from a poor quality of institutions ( = < 1 + k 2 =l 1), the optimal monetary policy
response to a decrease in the institutional quality is to lower the ination rate under a
discretionary regime, but to raise it under a commitment regime. For the intuition behind
this dierence, one rst recalls that the a decrease in the institutional quality implies a rise
in the shadow price of collecting regular taxes relative to collecting the ination tax. This
explains why the ination rate under commitment goes up. In the discretionary case, while
8

this force still operates, there is an additional, opposing force at work. The equilibrium
ination rate under discretion is already higher than that under commitment, especially
when the initial level of the institutional quality is low, resulting in a lower social welfare.
When the marginal cost of raising regular taxes increases yet again, in order not to further
reduce the social welfare by raising the ination rate from an already high level, cutting
public goods provision and cutting both the regular and ination taxes now become more
attractive. Despite this non-linear response pattern, it is easy to see that V D V C , where
the equality sign holds when = 0. Therefore, consistent with the literature, an ability to
commit (to the optimal level of ination) raises welfare.
Proposition 3 The optimal commitment regime generates a lower ination rate, a higher
tax rate, and a higher social welfare than the discretionary regime.

4.2

Rogo-type Conservative Central Banker

If, for whatever reason, a commitment regime is not available, Rogo (1985) suggested that
delegating the monetary policy to a more conservative central banker (still with discretion)
can improve upon the social welfare relative to the conventional discretionary regime. Here,
more conservative means that the weight in the loss function on ination placed by the
central banker is greater than that by the social planner.
In this section, we examine whether and how the optimal degree of central banker conservatism is aected by the presence of weak institutions. As a by-product, we also examine
how the inclusion of public goods provision in the social welfare function may modify our
understanding of the role of a conservative central banker.
Consider a modied central bankers problem. Let S denote the weight on the ination
rate placed by the central banker. The central bankers objective function is given by
V CC ( ; ) =

1
S
2

+ ky 2 + l(g

g)2

(8)

If the central banker cares about ination as much as the social planner, then S = 1. We
can measure the degree of conservatism of the central banker by the excess weight she places
on the ination term relative to the social planner, i.e., conservatism = S 1.
The central banker and the scal authority still play a noncooperative Nash game. The
time-consistent policy mix in this case, labeled as ( CC ; CC ), is characterized by the rstorder conditions associated with (8), with the requirement that the expected ination rate
9

equals its equilibrium value. This yields:

CC

CC

(1

1
S(1

g
S )

+S g

(9)

;
k

(10)

+g

where = = ( + )
1 as dened earlier. It is clear that with a more conservative
central banker, the ination rate comes down, but the tax rate goes up. One can work out
[kl( + )2 +S 2 (k 2 +l 2 )]kl 2 g2
the level of social welfare as V ( CC ; CC ) = 21
.
2
[kl ( + )+S (k 2 +l 2 )]
Suppose the social planner can choose any value of S, then what degree of conservatism
would maximize the social welfare? To answer this, we maximize V ( CC ; CC ) with respect
to S. The rst order condition leads to12
Proposition 4 The optimal degree of central bank conservatism is given by S = 1 +

A number of observations can be made. First, generally speaking for 0 <

1,
S > 1, thus a central banker more conservative than the social planner should be appointed.
Second, the optimal degree of conservatism is proportional to the quality of institutions in
the economy. The poorer the quality of institutions (i.e., a lower value of ), the stronger
is the eect of distortionary tax on output (i.e., a higher ), or the weaker is the eect of
surprising ination on output (i.e., a lower ), the less conservative the central banker should
be.13 Third, in the extreme case in which weak institutions prevent the working of the tax
system completely (i.e., when = 0), the optimal degree of conservatism is zero.
When the central banker is optimally chosen, it can be veried that CC = C , CC = C ,
and V CC = V C . That is, this (modied) discretionary regime restores the rst-best solution
under commitment. This result is somewhat surprising and worth some further elaboration.
There are a number of dierences between our model and Rogos framework. First, we
have added public goods provision as part of the objective function. Although a more
conservative central banker can reduce ination further, it would not be optimal to do that
given the increasing costs of collecting taxes. Second, we do not have stochastic shocks to
the aggregate supply/demand. Third, we do not have the equivalent of the labor market
12

It can be veried that V ( CC ; CC ) is indeed convex in S.

As a possible extension, one might endogenize and make it a decreasing function in . That is, when
the leakage in tax collection increases, the tax distortions could become smaller. In this case, the eect of
surprising ination on output ( ) becomes relatively more important in determining S .
13

10

distortion that causes the social planner to attempt to stabilize output at a level above its
natural rate.
In our setting, the welfare under a conservative central banker dominates that of a
currency board or dollarization. If installing a conservative central banker requires fewer
technical preconditions than implementing an ination target (e.g., due to the principle of
contract implementation a la Moore, 1992, and Maskin and Moore, 1999), then a conservative
central banker framework would also be better than an ination target. Without a scal
policy or with only a passive scal policy, the Walsh (1995) contract can implement the
commitment solution under a discretionary regime. However, once strategic manipulation
by the scal authority is introduced, the Walsh contract could be suboptimal (Huang and
Padilla, 2002). As a result, the discretionary tax may be too high while the ination rate
may be too low. By this logic, the conservative central banker arrangement may outperform
the Walsh-type incentive contract.

A Laer Curve Eect in Seigniorage

The discussion so far has ignored a possible nonlinear, Laer curve eect of a rise in the
ination rate on seigniorage revenue. As Cagan (1956) has shown, the semi-elasticity of the
demand for money aects the ability of government to extract seigniorage: when ination
rate exceeds a threshold, the seigniorage starts to decline. In other words, seigniorage revenue
and ination rate are not proportional to each other. In this section, we extend our analysis
to feature such an eect.
A simple way to capture the Laer curve eect is to replace in (3) by exp (
),
where the parameter
0 can be thought of as the semi-elasticity of the demand for real
money balances and captures the strength of the Laer curve eect. When
1= , an
increase in ination leads to an increase in the total seigniorage. When ination exceeds a
threshold, or > 1= , however, any further increase in ination rate leads to a decline in the
total seigniorage. Indeed, in the limit, when ination approaches innity, everyone avoids
using the domestic currency, and the government collects no seigniorage revenue. This is
represented by lim !1 exp (
) = 0.
With the introduction of this possibility, the government budget constraint becomes:
g=

+ exp (

11

):

(11)

The basic setup in the earlier sections is a special case in which

= 0.14

Under the commitment regime, the two rst-order conditions are:15

= l [g
k

exp (
2

)] (1

= l [g

) exp (

exp (

(12)

);

(13)

)] :
C

We denote the Nash equilibrium ination and tax rates by (

( ),

( )).

Although a closed form solution for ( C ( ), C ( )) is not possible, we can still obtain
several interesting results concerning the equilibrium. Note rst that with > 0, condition
(12) implies that to have
> 0, it has to be the case 1
> 0. That is, the Nash
equilibrium ination C ( ) has an upper bound, 1= , which is a decreasing function of .
This suggests that the optimal ination under commitment never goes beyond the threshold,
to the wrongside of the Laer curve.
C

Second, it is easy to show that @ @ ( ) < 0 for all

0. That is the ination rate goes
down as the quality of institutions improves ( goes up) regardless of the strength of the
Laer curve eect.
Third, it can be veried that

C(

< 0 if k

+l

(2

C(

[1

)) exp(2
C(

)]

C(
2

))

> kl

. Two

su cient condition for this to hold are l 2 and 2 > k . The rst su cient condition
corresponds to the case of a government that is not overly concerned with public goods
provision. The second su cient condition corresponds to the case when the quality of public
institutions exceeds a minimum threshold. Under either of the these conditions, the stronger
is the Laer curve eect, the lower is C ( ). In these circumstances, a mechanical ination
targeting, xed exchange rate, and currency board would be more acceptable than without
considering the Laer curve eect.
And fourth, for

0,
C

exp
( )
=
C( )
1

( )
C( )

The last result can be understood intuitively. As the Laer eect raises the marginal cost
14

It is possible that tax revenue collection also has a Laer curve eect. It can be analyzed by replacing
in (3) with
exp (
), where > 0. Other things equal, this complication tends to tilt the choice
between ination and tax towards the former, especially when the need for public goods provision is high.
We leave this for future research.
15
The second-order conditions associated with this problem (as well as those in the discretionary problem
below) are satised.

12

of seigniorage, the mix of the revenue collection shifts toward greater reliance on tax. This
leads to a higher ratio of tax to ination rates.
We have also analyzed the discretionary regime. We nd that in the discretionary case,
we cannot say unconditionally that the Laer curve eect would lead to a lower ination.
What we do know for sure is that the Laer curve eect tilts revenue collection away from
ination toward tax. Interested readers can nd the analysis in the working paper version
(Huang and Wei, 2005) for detail.

Strengthening Institutional Quality

So far we have treated institutional quality, , as exogenously given. Eorts in strengthening

institutional quality, e.g., through ghting corruption and improving scal capacity, should
increase the value of . In this section we endogenize the quality of institutions and ask
when a government would be willing to undertake reforms to strengthen institutions. Our
modeling strategy is similar to the case of tax reform in Cukierman, Edwards and Tabellini
(1992).
To start with, we observe that a governments eort to improve institutional quality is
likely to come with a cost. The cost could be in the form of a loss of economic rents that
o cials enjoy, or a stiened resistance from powerful special interest groups that have been
beneting from corruption and lost tax revenue. To capture this observation, we assume that
the institutional quality and the authoritieseort go hand in hand and their relationship
can be captured in a multiplicative form:
p
= 0 f;
(14)
where f 2 [1; 1= 20 ] denotes the level of eort by the authorities,

is the initial value of .

We assume further that the authorities share the preference of the social planner except
that they also bear the cost (disutility) of the eort, which is proportional to their eort,
C = (f
where

1) ;

(15)

> 0 is the cost per unit of the eort.

With this simple setup, the equilibrium eort level, and thus the equilibrium value of
can be solved in two steps in a principal-agent framework. Our analysis focuses on the
commitment case, but it is straightforward to extend it to the discretionary case. The policy
13

game is the same as before, except that the authorities need to choose their level of eort
rst.
Since the authorities share the preference of the social planner, their utility net of the
eort cost is
1
kl 2 g 2
C
C
VA (f ) = V
C=
(f 1) :
(16)
2 (1 + l) k 2 + l 20 f
Taking rst derivative of (16) with respect to f , one gets16
kl2
2 (1 + l) k

2 2 2
g 0
2

+ l 20 f

(17)

= 0:

Examining this rst-order condition (17), we have the following proposition.

Proposition 5 For

<

< , where
2 2 2
g 0
2

kl2

2 (1 + l) k
kl2

2 2 2
g 0
2

2 (1 + l) k
an interior optimal solution 1 < f < 1=
r
f =

2
0

+l
+l

2;

;
2 2
0

exists, and

k
2

1+lk
l

2
2
0

The equilibrium level of eort (f ) goes up, if the marginal cost of eort is lower (smaller
), or more importance is placed on public goods provision (higher l or g).
The equilibrium ination and tax rates are
p
2k
g
;
=
2
2
+ l 0f
0
sr
p
2 g
1+l2
l 0 f g
C
(f ) =
=
2:
2
2
k
l
(1 + l) k + l 0 f
0
0
C

kl
(f ) =
(1 + l) k

It is easy to see that

C

16

(f ) =

kl
(1 + l) k

kl 2 g
g
<
2
(1 + l) k 2 + l
+ l 20 f
2

It is easy to verify that the second-order condition holds.

14

2
0

(18)

(19)

l 0 f g
=
(f ) =
(1 + l) k 2 + l 20 f

p
, if 2 (1 + l) 2 k l 0 g;
p
> C , if 2 (1 + l) 2 k > l 0 g:
C

Further examining the rst-order condition (17), we have the following corollary, which
suggests that the cost coe cient, , is a key parameter that aects the authoritiesincentive
to strengthen institutions.
Corollary 6 If
, then the authorities would have no incentive to devote any eorts
to strengthen institutions; If < , however, the authorities would have incentive to devote
su cient eorts to strengthen institutions so that = 1.
We note that lim 0 !0 = 0. In other words, when the initial quality of institutions is
very poor, such that 0 has a very low value, most values of would be greater than .
In this case, the authorities would have no incentive to devote any eort to strengthen the
institutions17 . This is because a very low initial level of institutional quality (e.g., a very high
initial level of corruption) means a massive leakage of tax revenue. Under such circumstances
even with a lot of costly eort, the authorities would not be able to raise enough revenue to
make the eort worthwhile. Thus they would choose not to invest in any eort any all.
In this case, setting a low ination level through ination targeting or appointing a
Rogo-type conservative central banker would not by themselves induce the government
to devote more eort to strengthen institutions. Perhaps reforms to improve institutional
quality should be taken before adopting a monetary regime aiming for a low level of ination.
If the initial quality of institutions is moderate, such that < holds, then the authorities
would have incentive on their own to devote eorts to strengthen institutions. Setting a low
ination level through ination targeting (to induce corruption ghting) would not hurt,
though it is by no means a decisive tool to strengthen the institutions.

Concluding Remarks

In this paper, we examine the eects of institutional quality on the desirability of several popular monetary regimes, including ination targeting, exchange rate xing, currency
17

It is possible that changing the assumed relationship between institutional quality and eort could
modify this result.

15

board, and a conservative central banker. The simple model of a monetary policy game,
whereby institutional quality adversely aects the taxable revenue, has generated a number
of interesting results.
First, we cast doubt on the conventional wisdom that prescribes pegged exchange rate
regimes, currency boards and dollarization as means to increase the credibility of a governments resolve to maintain low ination. Our analysis suggests that these monetary regimes
may not be very credible themselves and can fail in countries where institutions are seriously weak. Second, an optimally chosen conservative central banker is generally preferable
to a mechanical ination target of 1-4 percent and to most exchange-rate-based monetary
arrangements. The optimal degree of conservatism is proportional to the quality of institutions in the economy. Third, the presence of a Laer curve eect on seigniorage revenue
likely lowers ination and raise tax rate, although in some cases it may raise both ination
and tax rates. Fourth, the notion that a low ination target or a currency board can be used
as an instrument to induce governments to strengthen institutions is questionable. These
ndings are important in the design of monetary policies for developing countries.
A number of further extensions can be made. First, the government can be allowed to
borrow in domestic bond market or international capital market. The interactions among
institutional quality, debt, and monetary policies can be explored. Second, the eect of a
Laer curve in tax revenue as well as in seigniorage on ination and tax rates is also interesting. Third, a systematic empirical examination can be illuminating. For example, is there
support for Proposition 4 in the paper that the optimal degree of central bank conservatism
depends positively on the institutional quality, and the elasticity of the aggregate demand
with respect to the ination surprise, but negatively on the elasticity of the aggregate supply to the distortionary tax rate? Does the dispersion in the experiences of developing and
transition economies with ination targeting reect the insights of this model? These can
be interesting and important topics for future research.

16

References
[1] Alesina, Alberto, and Guido Tabellini, 1987, Rules and Discretion with Noncoordinated Monetary and Fiscal Policies,Economic Inquiry, Vol. 25, No. 4, pp. 61930.
[2] Bai, Chong-En, and Shang-Jin Wei, 2000, Quality of Bureaucracy and Open-Economy
Macro Policies,NBER Working Paper No. 7766 (June).
[3] Barro, Robert, 1986, Reputation in a Model of Monetary Policy with Incomplete Information,Journal of Monetary Economics, Vol. 17, No. 1, pp. 320.
[4] Barro, Robert, and David Gordon, 1983, Rules, Discretion, and Reputation in a Model
of Monetary Policy,Journal of Monetary Economics, Vol. 12, No. 1, pp. 10122.
[5] Benigno, Pierpaolo, and Michael Woodford, 2003, Optimal Monetary and Fiscal Policy:
A Linear-Quadratic Approach, in M. Gertler and K. Rogo, eds., NBER Macroeconomics Annual 2003, vol. 18.
[6] Berger, Helge, Jakobde de Haan, and Sylvester C.W. Eij nger. 2001, Central Bank
Independence: An Update Theory and Evidence, Journal of Economic Survey, pp.
3-40.
[7] Bernanke, Ben, Thomas Laubach, Frederic S. Mishkin, and Adam S. Posen, 1999, Ination Targeting: Lessons from the International Experience, Princeton University Press.
[8] Cagan, Phillip, 1956, The Monetary Dynamics of Hyperination, in Studies in the
Quantity Theory of Money, ed. Milton Friedman. Chicago: University of Chicago Press.
[9] Calvo, Guillermo, 1978, On the Time Consistency of Optimal Policy in a Monetary
Economy,Econometrica, Vol. 46, pp. 141128.
[10] Cukierman, Alex, 1992, Central Bank Strategy, Credibility and Independence. Cambridge: MIT Press.
[11] Cukierman, Alex, Sebastian Edwards, and Guido Tabellini, 1992, Seigniorage and
Political Instability,American Economic Review. Vol. 82: 537-555.
[12] Cukierman, Alex, Yossi Spiegel and Leonardo Leidermna, 2004, The Choice of Exchange Rate Bands: Balancing Credibility and Flexibility, Journal of International
Economics. Vol. 62(2): 379-408.

17

[13] Du, Julan, and Shang-Jin Wei, 2004, Does Insider Trading Raise Stock Market Volatility?Economic Journal, 114(October): 927-956.
[14] Eichengreen, Barry, Paul Masson, Miguel Savastano, and Sunil Sharma, 1999, Transition Strategies and Nominal Anchors on the Road to Greater Exchange-rate Flexibility,
Essays in International Finance No. 213 (Princeton: Princeton University).
[15] Fisman, Raymond, and Shang-Jin Wei, 2004, Tax Rates and Tax Evasion: Evidence
from Missing Importsin China,Journal of Political Economy, 112(2): 471-496.
[16] Ghosh, Atish R., Anne-Marie Gulde, and Holger C. Wolf, 2003, Exchange Regimes:
Choices and Consequences, MIT Press.
[17] Huang, Haizhou, and Jorge Padilla, 2002, Fiscal Policy and the Implementation of the
Walsh Contract for Central Bankers Annals of Economics and Finance, Vol. 3, pp.
2742.
[18] Huang, Haizhou, and Shang-Jin Wei, 2005, Monetary Policies for Developing Countries: The Role of Institutional Quality," CEPR Discussion Paper No. 4911, February.
[19] Kydland, Finn E., and Edward C. Prescott, 1977, Rules Rather than Discretion: The
Inconsistency of Optimal Plans, Journal of Political Economy, Vol. 85, No. 3, pp.
47391.
[20] Masson, Paul R., Miguel A. Savastano, and Sunil Sharma, 1997, The Scope for Ination
Targeting in Developing Countries, IMF Working Paper 97/130 (Washington, DC.:
International Monetary Fund).
[21] Maskin, Eric, and John Moore, 1999, Implementation and Renegotiation, Review of
Economic Studies, Vol. 66, No. 1, pp. 3954.
[22] Mauro, Paolo, 1995, Corruption and Growth, Quarterly Journal of Economics, Vol.
110, pp. 681712.
[23] Moore, John, 1992, Implementation, Contracts, and Renegotiation in Environments
with Complete Information,in Advances in Economic Theory: Sixth World Congress,
Vol. 1, pp. 182282, ed. by Jean-Jacques Laont (Cambridge: Cambridge University
Press).
[24] Persson, Tosten, and Guido Tabellini, 1990, Macroeconomic Policy, Credibility and
Politics. London: Harwood Publishers.
18

[25] Rogo, Kenneth, 1985, The Optimal Degree of Commitment to an Intermediate Monetary Target,Quarterly Journal of Economics, Vol. 100, No. 4, pp. 116990.
[26] Rose-Ackerman, Susan, 1975, The Economics of Corruption, Journal of Public Economics, Vol. 4, No. 2, pp. 187203.
[27] Shleifer, Andrei, and Robert Vishny, 1993, Corruption, Quarterly Journal of Economics, Vol. 108, pp. 599617.
[28] Svensson, Lars, 1997, Optimal Ination Targets, Conservative Central Banks, and
Linear Ination Contracts,American Economic Review, Vol. 87, pp. 98114.
[29] Walsh, Carl, 1995, Optimal Contracts for Central Bankers, American Economic Review, Vol. 85, No. 1, pp. 15067.
[30] Wei, Shang-Jin, 2000, How Taxing is Corruption on International Investors?Review
of Economics and Statistics, Vol. 82, No. 1, pp. 111.
[31] Wei, Shang-Jin, 2001, Domestic Crony Capitalism and International Fickle Capital: Is
There a Connection?International Finance, Vol. 4, No. 1, pp. 1545.

19