An Assessment
The contributions made by leading new classicists such as Lucas, Barro,
Sargent and Wallace dominated macroeconomics discussion throughout the
1970s, particularly in the USA. In particular the business cycle research of
Lucas during the 1970s had an enormous methodological impact on how
macroeconomists conducted research and looked at the world (Lucas, 1980a,
1981a; Hoover, 1992, 1999; Chapter 6). For example, although the idea that
all unemployment should be viewed as voluntary remains controversial, economists
after the ‘Lucasian revolution’ have been much less willing to accept
uncritically Keynes’s idea of ‘involuntary unemployment’ (see Solow, 1980;
Blinder, 1988a; Snowdon and Vane, 1999b).
However, by the close of the 1970s, several weaknesses of the new classical
equilibrium approach were becoming apparent. These deficiencies were
mainly the consequence of utilizing the twin assumptions of continuous
market clearing and imperfect information. By 1982 the monetary version of
new classical equilibrium models had reached both a theoretical and empirical
impasse. For example, on the theoretical front the implausibility of the
assumption relating to information confusion was widely recognized (Okun,
1980; Tobin, 1980b). With sticky prices ruled out on methodological grounds,
new classical models were left without an acceptable explanation of business
cycles involving money-to-output causality. Furthermore, the doubts cast by
Sims (1980) on the causal role of money in money–output correlations raised
questions with respect to monetary explanations of the business cycle. On the
empirical front, despite some early success, the evidence in support of the
proposition that anticipated money was neutral did not prove to be robust (see
Barro, 1977a, 1978, 1989a). According to Gordon (1989) the influence of the
first phase of new classical theorizing peaked in the period 1976–8. Gordon
also dates the downfall of this phase ‘precisely at 8.59 a.m. EDT on Friday
13th October, 1978, at Bald Peak, New Hampshire’ for it was here that
Robert Barro and Mark Rush (1980) began their presentation ‘of an empirical
test of the policy-ineffectiveness proposition on quarterly US post-war data
that was not only severely criticised by three discussants, but also contained
dubious results that seemed questionable even to the authors’ (see Hoover,
1992, Vol. 1). Thus the early 1980s witnessed the demise of the mark I
(monetary surprise) version of the new classical approach in large part due to
the implausibility of supposed information gaps relating to aggregate price
level and money supply data, and the failure of empirical tests to provide
strong support for the policy ineffectiveness proposition (Barro, 1989a). The
depth of the recessions in both the USA and the UK in the 1980–82 period
following the Reagan and Thatcher deflations provided the critics with further
ammunition. As a consequence of these difficulties the monetary surprise
model has come to be widely regarded as inappropriate for modern information-
rich industrial economies.
Meanwhile Stanley Fischer (1977) and Edmund Phelps and John Taylor
(1977) had already shown that nominal disturbances were capable of producing
real effects in models incorporating rational expectations providing the
assumption of continuously clearing markets was abandoned. While accepting
the rational expectations hypothesis was a necessary condition of being a
new classicist, it was certainly not sufficient. Following the embryonic new
Keynesian contributions it was quickly realized that the rational expectations
hypothesis was also not a sufficient condition for policy ineffectiveness. As a
result the policy-ineffectiveness proposition was left ‘to die neglected and
unmourned’ and ‘Into this vacuum stepped Edward Prescott from Minnesota,
who has picked up the frayed new classical banner with his real business
cycle theory’ (Gordon, 1989). Thus Lucas’s MEBCT has been replaced since
the early 1980s by new classical real business cycle models emphasizing
technological shocks (Stadler, 1994), new Keynesian models emphasizing
monetary disturbances (Gordon, 1990), and new neoclassical synthesis models
combining insights from both approaches (see Lucas, 1987; Goodfriend
and King, 1997; Blanchard, 2000).
Economists sympathetic to the new classical approach (such as Finn Kydland
and Edward Prescott) have developed a mark II version of the new classical
model, known as real equilibrium business cycle theory (REBCT, see Figure
5.7). While proponents of the REBCT approach have abandoned the monetary
surprise approach to explaining business cycles, they have retained
components of the equilibrium approach and the propagation mechanisms
(such as adjustment costs) used in mark I versions. Responding to the Lucas
critique was also a major driving force behind the development of REBCT
(see Ryan and Mullineux, 1997).
Despite the controversy that surrounds the approach, new classical economics
has had a significant impact on the development of macroeconomics
over the last decade and a half. This impact can be seen in a number of areas.
First, it has led to much greater attention being paid to the way that expectations
are modelled, resulting in a so-called ‘rational expectations revolution’
in macroeconomics (Taylor, 1989). For example, the rational expectations
hypothesis has been widely adopted by new Keynesians and researchers in
the area of the ‘new political macroeconomics (see Chapters 7 and 10). It also
formed a crucial input to Dornbusch’s (1976) exchange rate overshooting
model (see Chapter 7). Second, the insight of rational expectations that a
change in policy will almost certainly influence expectations (which in turn is
likely to influence the behaviour of economic agents) is now fairly widely
accepted. This in turn has led economists to reconsider the role and conduct
of macroeconomic stabilization policy. In particular, the modern emphasis on
‘policy rules’ when discussing the stabilizing role of monetary policy has
been heavily influenced by the idea of rational expectations.
Much of the controversy that surrounds new classical macroeconomics is
directed, not at the rational expectations hypothesis per se, but at the policy
implications that derive from the structure of new classical models. In this
context it is interesting to note that Keynesian-like disequilibrium models
(where markets do not clear continuously) but which allow agents to have
rational expectations, as well as incorporating the natural rate hypothesis,
still predict a role for demand management policies to stabilize the economy.
If, in the face of random shocks to aggregate demand, the government is able
to adjust its policies more quickly than the private sector can renegotiate
money wages, then there is still a role for aggregate demand management to
stabilize the economy and offset fluctuations in output and employment around
their natural levels. As Buiter (1980) summed it up, ‘in virtually all economically
interesting models there will be real consequences of monetary and
fiscal policy–anticipated or unanticipated. This makes the cost–benefit analysis
of feasible policy intervention the focus of the practical economist’s
concern.’ There should therefore be no presumption that ‘a government that
sits on its hands and determines the behaviour of its instruments by the
simplest possible fixed rules is guaranteed to bring about the best of all
possible worlds’. Furthermore, given the gradual adjustment of prices and
wages in new Keynesian models, any policy of monetary disinflation, even if
credible and anticipated by rational agents, will lead to a substantial recession
in terms of output and employment, with hysteresis effects raising the
natural rate of unemployment (see Chapter 7).
Finally, in trying to come to an overall assessment of the impact of new
classical macroeconomics on the debate concerning the role and conduct of
macroeconomic stabilization policy, three conclusions seem to suggest themselves.
First, it is fairly widely agreed that the conditions necessary to render
macroeconomic stabilization policy completely powerless to influence output
and employment in the short run are unlikely to hold. Having said this, the
possibility that economic agents will anticipate the effects of changes in
economic policy does imply that the authorities’ scope to stabilize the economy
is reduced. Second, new classical macroeconomics has strengthened the case
for using aggregate supply policies to stimulate output and employment.
Lastly, new Keynesians have been forced to respond to the challenge of new
classical macroeconomics and in doing so, in particular explaining why wages
and prices tend to adjust only gradually, have provided a more sound microtheoretical
base to justify interventionist policies (both demand and supply
management policies) to stabilize the economy.
Before discussing new Keynesian economics we first examine in the next
chapter the evolution of the Mark II version of new classical economics, that
is, real business cycle theory.
The contributions made by leading new classicists such as Lucas, Barro,
Sargent and Wallace dominated macroeconomics discussion throughout the
1970s, particularly in the USA. In particular the business cycle research of
Lucas during the 1970s had an enormous methodological impact on how
macroeconomists conducted research and looked at the world (Lucas, 1980a,
1981a; Hoover, 1992, 1999; Chapter 6). For example, although the idea that
all unemployment should be viewed as voluntary remains controversial, economists
after the ‘Lucasian revolution’ have been much less willing to accept
uncritically Keynes’s idea of ‘involuntary unemployment’ (see Solow, 1980;
Blinder, 1988a; Snowdon and Vane, 1999b).
However, by the close of the 1970s, several weaknesses of the new classical
equilibrium approach were becoming apparent. These deficiencies were
mainly the consequence of utilizing the twin assumptions of continuous
market clearing and imperfect information. By 1982 the monetary version of
new classical equilibrium models had reached both a theoretical and empirical
impasse. For example, on the theoretical front the implausibility of the
assumption relating to information confusion was widely recognized (Okun,
1980; Tobin, 1980b). With sticky prices ruled out on methodological grounds,
new classical models were left without an acceptable explanation of business
cycles involving money-to-output causality. Furthermore, the doubts cast by
Sims (1980) on the causal role of money in money–output correlations raised
questions with respect to monetary explanations of the business cycle. On the
empirical front, despite some early success, the evidence in support of the
proposition that anticipated money was neutral did not prove to be robust (see
Barro, 1977a, 1978, 1989a). According to Gordon (1989) the influence of the
first phase of new classical theorizing peaked in the period 1976–8. Gordon
also dates the downfall of this phase ‘precisely at 8.59 a.m. EDT on Friday
13th October, 1978, at Bald Peak, New Hampshire’ for it was here that
Robert Barro and Mark Rush (1980) began their presentation ‘of an empirical
test of the policy-ineffectiveness proposition on quarterly US post-war data
that was not only severely criticised by three discussants, but also contained
dubious results that seemed questionable even to the authors’ (see Hoover,
1992, Vol. 1). Thus the early 1980s witnessed the demise of the mark I
(monetary surprise) version of the new classical approach in large part due to
the implausibility of supposed information gaps relating to aggregate price
level and money supply data, and the failure of empirical tests to provide
strong support for the policy ineffectiveness proposition (Barro, 1989a). The
depth of the recessions in both the USA and the UK in the 1980–82 period
following the Reagan and Thatcher deflations provided the critics with further
ammunition. As a consequence of these difficulties the monetary surprise
model has come to be widely regarded as inappropriate for modern information-
rich industrial economies.
Meanwhile Stanley Fischer (1977) and Edmund Phelps and John Taylor
(1977) had already shown that nominal disturbances were capable of producing
real effects in models incorporating rational expectations providing the
assumption of continuously clearing markets was abandoned. While accepting
the rational expectations hypothesis was a necessary condition of being a
new classicist, it was certainly not sufficient. Following the embryonic new
Keynesian contributions it was quickly realized that the rational expectations
hypothesis was also not a sufficient condition for policy ineffectiveness. As a
result the policy-ineffectiveness proposition was left ‘to die neglected and
unmourned’ and ‘Into this vacuum stepped Edward Prescott from Minnesota,
who has picked up the frayed new classical banner with his real business
cycle theory’ (Gordon, 1989). Thus Lucas’s MEBCT has been replaced since
the early 1980s by new classical real business cycle models emphasizing
technological shocks (Stadler, 1994), new Keynesian models emphasizing
monetary disturbances (Gordon, 1990), and new neoclassical synthesis models
combining insights from both approaches (see Lucas, 1987; Goodfriend
and King, 1997; Blanchard, 2000).
Economists sympathetic to the new classical approach (such as Finn Kydland
and Edward Prescott) have developed a mark II version of the new classical
model, known as real equilibrium business cycle theory (REBCT, see Figure
5.7). While proponents of the REBCT approach have abandoned the monetary
surprise approach to explaining business cycles, they have retained
components of the equilibrium approach and the propagation mechanisms
(such as adjustment costs) used in mark I versions. Responding to the Lucas
critique was also a major driving force behind the development of REBCT
(see Ryan and Mullineux, 1997).
Despite the controversy that surrounds the approach, new classical economics
has had a significant impact on the development of macroeconomics
over the last decade and a half. This impact can be seen in a number of areas.
First, it has led to much greater attention being paid to the way that expectations
are modelled, resulting in a so-called ‘rational expectations revolution’
in macroeconomics (Taylor, 1989). For example, the rational expectations
hypothesis has been widely adopted by new Keynesians and researchers in
the area of the ‘new political macroeconomics (see Chapters 7 and 10). It also
formed a crucial input to Dornbusch’s (1976) exchange rate overshooting
model (see Chapter 7). Second, the insight of rational expectations that a
change in policy will almost certainly influence expectations (which in turn is
likely to influence the behaviour of economic agents) is now fairly widely
accepted. This in turn has led economists to reconsider the role and conduct
of macroeconomic stabilization policy. In particular, the modern emphasis on
‘policy rules’ when discussing the stabilizing role of monetary policy has
been heavily influenced by the idea of rational expectations.
Much of the controversy that surrounds new classical macroeconomics is
directed, not at the rational expectations hypothesis per se, but at the policy
implications that derive from the structure of new classical models. In this
context it is interesting to note that Keynesian-like disequilibrium models
(where markets do not clear continuously) but which allow agents to have
rational expectations, as well as incorporating the natural rate hypothesis,
still predict a role for demand management policies to stabilize the economy.
If, in the face of random shocks to aggregate demand, the government is able
to adjust its policies more quickly than the private sector can renegotiate
money wages, then there is still a role for aggregate demand management to
stabilize the economy and offset fluctuations in output and employment around
their natural levels. As Buiter (1980) summed it up, ‘in virtually all economically
interesting models there will be real consequences of monetary and
fiscal policy–anticipated or unanticipated. This makes the cost–benefit analysis
of feasible policy intervention the focus of the practical economist’s
concern.’ There should therefore be no presumption that ‘a government that
sits on its hands and determines the behaviour of its instruments by the
simplest possible fixed rules is guaranteed to bring about the best of all
possible worlds’. Furthermore, given the gradual adjustment of prices and
wages in new Keynesian models, any policy of monetary disinflation, even if
credible and anticipated by rational agents, will lead to a substantial recession
in terms of output and employment, with hysteresis effects raising the
natural rate of unemployment (see Chapter 7).
Finally, in trying to come to an overall assessment of the impact of new
classical macroeconomics on the debate concerning the role and conduct of
macroeconomic stabilization policy, three conclusions seem to suggest themselves.
First, it is fairly widely agreed that the conditions necessary to render
macroeconomic stabilization policy completely powerless to influence output
and employment in the short run are unlikely to hold. Having said this, the
possibility that economic agents will anticipate the effects of changes in
economic policy does imply that the authorities’ scope to stabilize the economy
is reduced. Second, new classical macroeconomics has strengthened the case
for using aggregate supply policies to stimulate output and employment.
Lastly, new Keynesians have been forced to respond to the challenge of new
classical macroeconomics and in doing so, in particular explaining why wages
and prices tend to adjust only gradually, have provided a more sound microtheoretical
base to justify interventionist policies (both demand and supply
management policies) to stabilize the economy.
Before discussing new Keynesian economics we first examine in the next
chapter the evolution of the Mark II version of new classical economics, that
is, real business cycle theory.
No comments:
Post a Comment