domestic wage hike is exactly offset by the (positive) response of the ultra-populist CB at
Home or is perceived nil by atomistic wage setters34.
As to inflation, equation (39) and (40) reveal the following proposition.
Proposition 6 (i) in the absence of different strategic effects, i.e. ηu = η n and τff = g*N,
the inflation rate under a NMP regime is always lower than in a MU for a given level of
employment. (ii) A higher degree of the CBC, β, reduces the inflation bias (< 0).
Proof. The first part of the proposition is immediately proved by observing equation (39)
and (40). For the second part, see equation (53). ■
The motive the shift to a MU raises inflation is the different trade-off between employ-
ment and inflation faced by the common CB. Since the Phillips curve in a MU is flatter
than under a NMP regime, the common CB has stronger incentive to resort to surprise in-
flation (Rogoff, 1985b). Unions anticipate this inflationary inducement and strive to keep
CB from modifying their real wages which culminates in a higher inflationary bias. In this
respect, Proposition 6 explains why the ECB has a statute that is more conservative than
the one of the pre-MU Bundesbank considered the most conservative CB in Europe (Piga,
2000).
Contrary to Coricelli et al. (2004) where a higher degree of CBC is always associated
with lower inflation and unemployment, a more conservative CB in this model does curb
inflation while reduces unemployment only if the adverse output effect is stronger than the
adverse competition effect. The different upshot in Coricelli et al. (2004) is mainly due to
the absence of labor substitutability in the production function. Thus, the adverse output
effect always dominates the adverse competition effect and a more inflation averse CB
makes unions perceive higher labor demand elasticity, which results in lower real wages.
Now according to Proposition 5, the impact of CBC on labor elasticity depends on the
predominance of the adverse output or the competition effect. Hence, employment will be
an increasing function of CBC if the labor market distortion are high. Since the adverse
output effect is always larger in a MU relative to the adverse output effect under a NMP
regime, three cases are feasible35.
First, the adverse competition effect may prevail in both regimes (Figure 1). Employ-
ment is therefore a decreasing function of CBC and an ultra-populist CB is the first best
for the economy in terms of employment. Note that, under a NMP regime, an increase in
Foreign CBC boosts employment at Home by raising labor demand elasticity. However,
employment level is unambiguously higher in a MU when the CB is inflation averse.
Second, the adverse competition effect may prevail under a NMP regime but not in a
MU (Figure 2). In such a case labor market power is low at country level but relatively
34When the domestic CB does not care about inflation and wage setters are atomistic, the labor demand
elasticity is equal to σ.
35Analytically proved in the Appendix. As for the following simulation, we let n∣∣ = 3, 7 = 1/2, к = 1
and a = 3/4.
19