The government's objective function is
L = 2 [θG∏2 +(y∣- У* ) + λs2
(3)
which is an augmented Barro and Gordon (1983) utility function. The government wishes
to minimize inflation π , reduce the gap between given and desired output (yi - y*i ), and is
averse to structural reforms si. Because governments are politically constrained in structural
reform they must fear to lose political support through liberalizing labor markets,
privatizing enterprises, reducing subsidies to certain industries, or constraining possibilities
of appropriating rents. Too many reforms may risk losing elections or being otherwise
forced out of office.
The other national decision maker is the central bank. Its preferences are
LB = 2 Pb+(у,- У* )
(4)
which is the standard assumption about monetary policy makers' preferences (Barro and
Gordon 1983). It can be reasonably assumed that the central bank does not care about
structural reforms (and in particular one would not assume that they, like governments,
oppose reform). In addition, I assume θB ≥ θG which captures the fact that central banks
usually put more relative weight on avoiding inflation than unemployment, if it can pursue
its own preferences. As indicated above, the status of central banks in West Africa is not
clear, meaning that their independence can be doubted (Fouda and Stasavage 2000; Masson
and Patillo 2001b). Where important, I will discuss the influence of differences between
government and central bank preferences. Notice, however, that the relative weights the
authorities put on the several aims (θB, θG, λ) are not indexed and are assumed to be equal
across countries. There are hence internal conflicts but not across countries. This
assumption allows to abstract from effects that stem mainly from the fact that policy