adjust. Note first that
>
(tc,τc)
.
1 + ' (tN ,τN )
(15)
To verify the inequality note that the SMCPF is independent of τ since flk ≡ 0 and thus kτi ≡ 0.
Also, the SMCPF is higher the higher the level of symmetrically chosen capital taxes, because
kii ∣(tc τc) = kii ∣(tN τN) and hence tckii ∣(tc τc) < tNkii ∣(tN τN). Followingthefirst-Ordercondition
b0(дг) = i+i (Eq∙ (9)) the inequality (15) implies
b0(gi)∣(tc,τc) > b0(gi)∣(tN,τN)
By the strict concavity of b(gi), we can conclude that gi decreases relative to the Nash-level, i.e.
Public consumption becomes more downward distorted in the centralized leadership game.
A lower amount of labor-tax financed transfers exposes public consumption levels more to states’
incentives to compete for mobile capital - with the familiar consequence of a downward distortion
in public consumption.
In equilibrium, the total amount of tax revenues decreases, but taxes are more efficiently
collected. In spite of the efficiency trade-off, the welfare effect is ambiguous in sign. The federal
government is able to replicate the policy outcome under Nash-behavior by setting τc = τN .
Consequently, revealed by the policy preference of the federal government (τc < τN), welfare
improves relative to welfare under Nash behavior.
6 Sequential Policy Choice: Pre-Commitment by State Govern-
ments
In this section, states governments have the ability to pre-commit toward the federal government
(decentralized leadership). The sequence of decisions becomes:
16