choice of the other region. Analyzing this non-cooperative policy is the sub-
ject of this section. The cooperative case and ist consequences will be treated
in the subsequent sections.
In the non-cooperative case, maximizing welfare is synonymical to maxi-
mizing regional real income:
where nominal income Y is divided by the regional price level. For any given
max
TI ,TA
YP-γ
(1-TA)γ-1
(11)
foreign tariff TI* and tax rate TA optimal policy can be now derived by setting
the partial derivatives of the real income with respect to TI and TA equal to
|
zero (cf. appendix). |
Then, domestic optimal tax and tariff rates are given |
|
TI,Opt |
_ σP*1-σT* - γ(1 -L) - γLΘ* = σ2P*1-σT* - γσ(1-L) - γLΘ* > , (12) |
|
TA,Opt |
= 1 > 0. |
As long as the standard assumptions about the parameters of the previ-
ous section are fulfilled γ < 1 < σ both optimal rates are always positive,
independent of the policy choice of the other region (cf. appendix). Re-
garding the optimal tariff, TI,Opt , this result here for a model of economic
geography confirms the well known home market effect of the new trade the-
ory (e.g. Brander (1995)). By a tariff, demand is shifted from foreign to
domestic varieties such that the domestic backward linkage is strengthened.
The additional demand generates additional profits of the monopolistic firms
resulting in higher wages and, therefore, in higher income. But, although
this nominal wage effect is always positive, the welfare gain is restricted by
the simultaneously increasing price level.
Regarding the optimal tax rate, TA,Opt , the result corresponds to the so-
cial optimum analysis of Dixit and Stiglitz (1977) in their original paper on
monopolistic competition. Taxes on agricultural consumption shift demand
towards industrial varieties. By this way the monopolistic market distortion
of supplying too expensive and too less varieties can be at least partly cor-
rected. Partly, because the resources for subsidizing the industrial sector are
10