Keynesian Economics Without the LM Curve
The modern approach to stabilization policy outlined in section 7.11 above is
now reflected in the ideas taught to students of economics, even at the
principles level (see D. Romer, 2000; Taylor, 2000b, 2001). The following
simple model is consistent with the macroeconomic models that are currently
used in practice by the US Federal Reserve and the Bank of England (see
Bank of England, 1999; Taylor, 1999; Clarida et al., 2000). Following Taylor
(2000b), the model consists of three basic relationships. First, a negative
relationship between the real rate of interest and GDP of the following form
y = −ar + μ (7.20)
where y measures real GDP relative to potential GDP, r is the real rate of
interest, μ is a shift term which, for example, captures the influence of
exogenous changes to exports and government expenditures and so on. A
higher real rate of interest depresses total demand in an economy by reducing
consumption and investment expenditures, and also net exports via exchange
rate appreciation in open economies with floating exchange rates. This relationship
is ‘analogous’ to the IS curve of conventional textbook IS–LM
analysis. The second key element in the model is a positive relationship
between inflation and the real rate of interest of the form:
r = bP˙ + v (7.21)
where ˙P is the rate of inflation and v is a shift term. This relationship, which
closely mirrors current practice at leading central banks, indicates that when
inflation rises the monetary authorities will act to raise the short-term nominal
interest rate sufficient to raise the real rate of interest. As Taylor (2000b)
and D. Romer (2000) both point out, central banks no longer target monetary
aggregates but follow a simple real interest rate rule. The third key relationship
underlying the modern monetary policy model is a ‘Phillips curve’ type
relationship between inflation and GDP of the form:
P˙ P˙ cy w = t−1 + t−1 + (7.22)
where w is a shift term. As equation (7.22) indicates, inflation will increase
with a lag when actual GDP is greater than potential GDP (y > y*) and vice
versa. The lag in the response of inflation to the deviation of actual GDP from
potential GDP reflects the staggered price-setting behaviour of firms with
market power inducing nominal stickiness. While this aspect indicates the
new Keynesian flavour of this model, the relationship also allows for expectations
of inflation to influence the actual rate.
From these three simple relationships we can construct a graphical illustration
of the modern approach to stabilization policy. Combining equations
(7.20) and (7.21) yields the following equation:
y = −abP˙ + μ − av (7.23)
Equation (7.23) indicates a negatively sloped relationship between inflation
and real GDP, which both Taylor and Romer call an aggregate demand
(AD) curve. Figure 7.15 illustrates the derivation of the aggregate demand
curve.
For simplicity, if we assume that the central bank’s choice of real interest
rate depends entirely on its inflation objective, the monetary policy (MP) real
rate rule can be shown as a horizontal line in panel (a) of Figure 7.15, with
shifts of the MP curve determined by the central bank’s reaction to changes
in the rate of inflation. Equation (7.20) is represented by the IS curve in
Figure 7.15. In panel (b) of Figure 7.15 we see equation (7.23) illustrated by
a downward-sloping aggregate demand curve in inflation–output space. The
intuition here is that as inflation rises the central bank raises the real rate of
interest, thereby dampening total expenditure in the economy and causing
GDP to decline. Similarly, as inflation falls, the central bank will lower the
real rate of interest, thereby stimulating total expenditure in the economy and
raising GDP. We can think of this response as the central bank’s monetary
policy rule (Taylor, 2000b).
Shifts of the AD curve would result from exogenous shocks to the various
components to aggregate expenditure, for example the AD curve will shift to
the right in response to an increase in government expenditure, a decrease in
taxes, an increase in net exports, or an increase in consumer and/or business
confidence that leads to increased expenditures. The AD curve will also shift
in response to a change in monetary policy. For example, if the monetary
authorities decide that inflation is too high under the current monetary policy
rule, they will shift the rule, raise real interest rates and shift the AD curve to
the left (see Taylor, 2001).
The Phillips curve or inflation adjustment relationship, given by equation
(7.22), is represented by the horizontal line labelled IA0 in Figure 7.16. Following
Taylor (2000b) and D. Romer (2000), this can be thought of as the aggregate
supply component of the model, assuming first that the immediate impact of an
increase in aggregate demand will fall entirely on aggregate output, and second
that when actual GDP equals potential or ‘natural’ GDP (y = y*), inflation will
be steady, but when y > y*, inflation will increase and when y < y*, inflation will
decline. Both of these assumptions are consistent with the empirical evidence
and supported by new Keynesian theories of wage and price stickiness in the
short run (Gordon, 1990). When the economy is at its potential output the IA
line will also shift upwards in response to supply-side shocks such as a rise in
commodity prices and in response to shifts in inflationary expectations. Figure
7.16 illustrates the complete AD–IA model.
Long-run equilibrium in this model requires that AD intersect IA at the
natural rate of output (y*). Assume that the economy is initially in long-run
equilibrium at point ELR 0 and that an exogenous demand shock shifts the AD
curve from AD0 to AD1. The initial impact of this shift is an increase in GDP
from y* to y1, with inflation remaining at P˙0 . Since y1 > y*, over time the rate
of inflation will increase, shifting the IA curve upwards. The central bank will
respond to this increase in inflation by raising the real rate of interest, shown
by an upward shift of the MP curve in the IS–MP diagram (Figure 7.15). The
IA curve continues to shift upwards until the AD and IA curves intersect at the
potential level of output y*, that is, where AD1 and IA1 intersect. The economy
is now at a new long-run equilibrium shown by ELR 1, but with a higher steady
rate of inflation of P˙1. The central bank has responded to the demand shock
by increasing the real rate of interest from r0 to r1. If the central bank decides
that the new steady rate of inflation is too high (that is, above its inflation
target), then it would have to take steps to shift the AD curve to the left by
changing its monetary policy rule. This would lead to a recession (y < y*) and
declining inflation. As the IA curve shifts down, the central bank will reduce
real interest rates, stimulating demand, and the economy will return to y* at a
lower steady rate of inflation.
The simple model described above gives a reasonably accurate portrayal of
how monetary policy is now conducted. In Taylor’s (2000b) view this theory
‘fits the data well and explains policy decisions and impacts in a realistic
way’. Whether this approach eventually becomes popularly known as ‘new
Keynesian’ (Clarida et al., 2000; Gali, 2002) or as ‘new neoclassical synthesis’
(Goodfriend and King, 1997) remains to be seen. David Romer (2000)
simply calls it ‘Keynesian macroeconomics without the LM curve’.
The modern approach to stabilization policy outlined in section 7.11 above is
now reflected in the ideas taught to students of economics, even at the
principles level (see D. Romer, 2000; Taylor, 2000b, 2001). The following
simple model is consistent with the macroeconomic models that are currently
used in practice by the US Federal Reserve and the Bank of England (see
Bank of England, 1999; Taylor, 1999; Clarida et al., 2000). Following Taylor
(2000b), the model consists of three basic relationships. First, a negative
relationship between the real rate of interest and GDP of the following form
y = −ar + μ (7.20)
where y measures real GDP relative to potential GDP, r is the real rate of
interest, μ is a shift term which, for example, captures the influence of
exogenous changes to exports and government expenditures and so on. A
higher real rate of interest depresses total demand in an economy by reducing
consumption and investment expenditures, and also net exports via exchange
rate appreciation in open economies with floating exchange rates. This relationship
is ‘analogous’ to the IS curve of conventional textbook IS–LM
analysis. The second key element in the model is a positive relationship
between inflation and the real rate of interest of the form:
r = bP˙ + v (7.21)
where ˙P is the rate of inflation and v is a shift term. This relationship, which
closely mirrors current practice at leading central banks, indicates that when
inflation rises the monetary authorities will act to raise the short-term nominal
interest rate sufficient to raise the real rate of interest. As Taylor (2000b)
and D. Romer (2000) both point out, central banks no longer target monetary
aggregates but follow a simple real interest rate rule. The third key relationship
underlying the modern monetary policy model is a ‘Phillips curve’ type
relationship between inflation and GDP of the form:
P˙ P˙ cy w = t−1 + t−1 + (7.22)
where w is a shift term. As equation (7.22) indicates, inflation will increase
with a lag when actual GDP is greater than potential GDP (y > y*) and vice
versa. The lag in the response of inflation to the deviation of actual GDP from
potential GDP reflects the staggered price-setting behaviour of firms with
market power inducing nominal stickiness. While this aspect indicates the
new Keynesian flavour of this model, the relationship also allows for expectations
of inflation to influence the actual rate.
From these three simple relationships we can construct a graphical illustration
of the modern approach to stabilization policy. Combining equations
(7.20) and (7.21) yields the following equation:
y = −abP˙ + μ − av (7.23)
Equation (7.23) indicates a negatively sloped relationship between inflation
and real GDP, which both Taylor and Romer call an aggregate demand
(AD) curve. Figure 7.15 illustrates the derivation of the aggregate demand
curve.
For simplicity, if we assume that the central bank’s choice of real interest
rate depends entirely on its inflation objective, the monetary policy (MP) real
rate rule can be shown as a horizontal line in panel (a) of Figure 7.15, with
shifts of the MP curve determined by the central bank’s reaction to changes
in the rate of inflation. Equation (7.20) is represented by the IS curve in
Figure 7.15. In panel (b) of Figure 7.15 we see equation (7.23) illustrated by
a downward-sloping aggregate demand curve in inflation–output space. The
intuition here is that as inflation rises the central bank raises the real rate of
interest, thereby dampening total expenditure in the economy and causing
GDP to decline. Similarly, as inflation falls, the central bank will lower the
real rate of interest, thereby stimulating total expenditure in the economy and
raising GDP. We can think of this response as the central bank’s monetary
policy rule (Taylor, 2000b).
Shifts of the AD curve would result from exogenous shocks to the various
components to aggregate expenditure, for example the AD curve will shift to
the right in response to an increase in government expenditure, a decrease in
taxes, an increase in net exports, or an increase in consumer and/or business
confidence that leads to increased expenditures. The AD curve will also shift
in response to a change in monetary policy. For example, if the monetary
authorities decide that inflation is too high under the current monetary policy
rule, they will shift the rule, raise real interest rates and shift the AD curve to
the left (see Taylor, 2001).
The Phillips curve or inflation adjustment relationship, given by equation
(7.22), is represented by the horizontal line labelled IA0 in Figure 7.16. Following
Taylor (2000b) and D. Romer (2000), this can be thought of as the aggregate
supply component of the model, assuming first that the immediate impact of an
increase in aggregate demand will fall entirely on aggregate output, and second
that when actual GDP equals potential or ‘natural’ GDP (y = y*), inflation will
be steady, but when y > y*, inflation will increase and when y < y*, inflation will
decline. Both of these assumptions are consistent with the empirical evidence
and supported by new Keynesian theories of wage and price stickiness in the
short run (Gordon, 1990). When the economy is at its potential output the IA
line will also shift upwards in response to supply-side shocks such as a rise in
commodity prices and in response to shifts in inflationary expectations. Figure
7.16 illustrates the complete AD–IA model.
Long-run equilibrium in this model requires that AD intersect IA at the
natural rate of output (y*). Assume that the economy is initially in long-run
equilibrium at point ELR 0 and that an exogenous demand shock shifts the AD
curve from AD0 to AD1. The initial impact of this shift is an increase in GDP
from y* to y1, with inflation remaining at P˙0 . Since y1 > y*, over time the rate
of inflation will increase, shifting the IA curve upwards. The central bank will
respond to this increase in inflation by raising the real rate of interest, shown
by an upward shift of the MP curve in the IS–MP diagram (Figure 7.15). The
IA curve continues to shift upwards until the AD and IA curves intersect at the
potential level of output y*, that is, where AD1 and IA1 intersect. The economy
is now at a new long-run equilibrium shown by ELR 1, but with a higher steady
rate of inflation of P˙1. The central bank has responded to the demand shock
by increasing the real rate of interest from r0 to r1. If the central bank decides
that the new steady rate of inflation is too high (that is, above its inflation
target), then it would have to take steps to shift the AD curve to the left by
changing its monetary policy rule. This would lead to a recession (y < y*) and
declining inflation. As the IA curve shifts down, the central bank will reduce
real interest rates, stimulating demand, and the economy will return to y* at a
lower steady rate of inflation.
The simple model described above gives a reasonably accurate portrayal of
how monetary policy is now conducted. In Taylor’s (2000b) view this theory
‘fits the data well and explains policy decisions and impacts in a realistic
way’. Whether this approach eventually becomes popularly known as ‘new
Keynesian’ (Clarida et al., 2000; Gali, 2002) or as ‘new neoclassical synthesis’
(Goodfriend and King, 1997) remains to be seen. David Romer (2000)
simply calls it ‘Keynesian macroeconomics without the LM curve’.
No comments:
Post a Comment