We argue in this paper that validity of discretionary macro-economic policy should be discounted from the point of view of politico-economics.
In section 2, We present a fundamental illustration, in which economic managers (bureaucrats) execute a feedback control of economic system and the incumbent party (polititians) pressures them. Particularly, we call the pressure“political noise”.
In sections 3, 4, 5 and 6, we formulate a model, and derive some propositions.
NOTATION:
Y (real national income)
p (rate of inflation)
pe (expected rate of inflation)
m (increase rate of money supply)
U (
Y, p) (evaluation function)
i (time preference rate)
ASSUMPTION:
(1)
Y,
p,
pe and
m are twice differentiable with respect to the time.
(2)
U1>0
U is a strictly quasi-concave function.
∂ (-
U2/
U1) /∂
p>0
(3) We assume an economic system described by the following equations.
Y=YN+a (p-pe) ;
YN,
a>0
(aggregate supply function of the Lucas type)
pe=
b (
p-pe) ; b>0
(adaptive expectation hypothesis)
p=m(quantity theory of money)
(4) The increase rate of money supply is a control variable of the economic system.
PROBLEM: The economic manager would consider the following problem.
Max
m∫
+∞0U(
Y, p)
e-itdtsubject to
Y=
YN+
a(
p-pe)
pe=
b(
p-pe)
p=
m.
THEOREM: The steady state solution of PROBLEM has the following properties.
(1)
Y*=
YN(so-called“normal output level”)
(2)
p*→0 as
i→0
(3)
dp*/
di>0
(4)-
U1/
U2(
Y*,
p*)→1/
a as
i→+∞
Politicians are generally more myopic than bureaucrats because of the existence of elections, so that the incumbent party would pressure economic managers into modifying their time preference rates. This mechanism inevitably accelerates inflation by the THEOREM. Hence we can conclude that discre-tionary macro-economic policy has a bias to-ward inflation in democratic systems. In particular, as it is plausible that political noise is increasing over the legislation period, this mechnism also suggests the existence of political business cycle.
By the way, the magnitude of inflation generated by political process would depend on the“autonomy”of bureaucratic organiza-tion. Therefore the difference in economic performance between Japan and the United States today may reflect the one in autonomy of civil service between the two. Namely, we can conjecture that economic manage-ments of the United States are too sensitive to the political noise to stabilize the economy.
In section 8, we discuss some policy im-plications in the context of theories of con-stitutional contract. Particularly we reevaluate the fixed control (as opposed to the feedback control), for instance, the Friedman's
k% rule, as a second-best policy.
抄録全体を表示