- Fourth, delegation should not give rise to major policy coordination problems. If a policy in a
particular area or some aspect of it is delegated, it should not create conflicts with policymakers in
another area that is not delegated. Otherwise, the coordination difficulties could outweigh any
benefit from delegation. Likewise, delegation would be undesirable if there are significant
economies associated with complementarities between prerogatives to be delegated and
nondelegated ones.^3
The delegation of monetary policy fits in fairly well with the above analytical criteria: (1) the
politicization of monetary policy is generally perceived as a key source of macroeconomic instability; (2)
there is relatively little variance in opinions about what constitutes sound monetary policy; (3) monetary
policy is not primarily distributive; and (4) a clear-cut assignment of responsibilities between monetary
and fiscal policies is in fact one way to deliver a well-coordinated policy mix (Box 6.1; see also Dixit and
Lambertini, 2003). (Rogoff (1985) first made a case for an independent “conservative” central bank that
would preserve policy discretion and still reduce the inflation bias.)
At the same time, the above framework shows why structural policies are unlikely to be delegated to an
independent agency. This is despite the fact that structural policies are often characterized by biases in
favor of the status quo due to political distortions (such as the action of well-organized special interests),
and reflects the following: (1) there is no consensus on an ideal economic “model”—mainly because
structural policies often involve difficult trade-offs between efficiency and equity; (2) structural reforms
often have deep distributive implications; and (3) they have far-reaching implications for other policy
areas.
1.2. Delegation of fiscal policy
With regard to fiscal policy, delegation might appear possible in some areas. The above analytical
framework suggests that specific areas of fiscal policy particularly susceptible to government failure
could be delegated to an independent agency. This concerns especially the overall fiscal balance, as
supported by the four criteria elaborated above: deficit bias and procyclical fiscal policies constitute
socially harmful distortions in policymaking; there is broad agreement that sound fiscal policy should not
create unsustainable deficits; the fiscal balance does not have direct distributional consequences, except
across generations;^4 and delegating the setting of the fiscal balance can reduce the problem of
macroeconomic policy coordination, especially with monetary policy.
However, some other areas of fiscal policy should clearly remain under elected officials’ control.
Specifically, those serving primarily distributive objectives, such as the progressivity of the tax system or
the size of social transfers, are not good candidates for delegation. Even though political decisions on
them might create economic inefficiencies, there is no broad consensus on what constitutes sound policy
in these areas. Aspects of fiscal policy that are so highly dependent on social preferences should clearly
remain under the control of the political process. (Nevertheless, there might be technical aspects of tax
and expenditure policies where delegation could be considered, as discussed below.
(^3) Some observers also point out that delegation is particularly useful in cases where policy choices involve a lot of
technical expertise with respect to other dimensions of the decision process (e.g., Blinder, 1997; and Alesina and
Tabellini, 2003). However, this does not appear to be a key discriminating criterion, as most well-designed economic
policy decisions arguably require a significant input from highly skilled professionals
(^4) Intergenerational redistribution actually adds a political economy argument for delegation: future generations are
generally not sufficiently represented by elected policymakers. Additional distributional aspects of the fiscal balance
concern its financing and the implications for expenditure composition of rapid adjustment. However, these
distributional repercussions are not necessarily greater than those associated with monetary policy changes.