Ngân hàng, tín dụng - Chapter 28: Empirical evidence of the effectiveness of monetary policy
These models suggest that fiscal policy has a multiplier
of about one without any help from monetary policy,
but an accommodating monetary authority can make
fiscal policy even more effective
• Empirical evidence on fiscal policy confirms the
crowding-out effect, but only if the contractionary
effects on private spending are given substantial time to
work themselves out
12 trang |
Chia sẻ: huyhoang44 | Lượt xem: 549 | Lượt tải: 0
Bạn đang xem nội dung tài liệu Ngân hàng, tín dụng - Chapter 28: Empirical evidence of the effectiveness of monetary policy, để tải tài liệu về máy bạn click vào nút DOWNLOAD ở trên
1Copyright © 2009 Pearson Addison-Wesley. All rights reserved.
Chapter 28
Empirical
Evidence of
the
Effectiveness
of Monetary
Policy
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-2
Learning Objectives
• Understand the historical volatility in velocity
and money demand and the implications of that
volatility for Keynesianism and monetarism
• Explain the fiscal and monetary policy lags and
resulting difficulties in enacting these policies
• Analyze the empirical evidence relating
monetary and fiscal policies to GDP,
investment, consumption, and interest rates
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-3
Introduction
• Previous chapters dealt with theoretical concepts
of monetary theory on economic activity
• This chapter explores empirical evidence
– How are interest rates affected
– What categories of spending are most influenced
– Does monetary policy alter real economic
outcomes—output and unemployment—as
compared to changes in prices
2Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-4
Living with Velocity
• Much of Monetarist-Keynesian debate hinges on
the behavior of velocity
– Monetarists—relatively stable and changes are
highly predictable
– Keynesians—neither contention is true, velocity can
vary unpredictably
• While velocity is not constant, changes do not
appear to be obviously random or perverse
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-5
Living with Velocity (Cont.)
• If the Federal Reserve could discover the
underlying determinants of these changes, it
might still be able to coexist with such a moving
target
• Figure 28.1 plots the historical movements of
measures of velocity for two definitions of the
money supply—M1 and M2—over time
– Velocity of M2 has been relatively stable
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-6
FIGURE 28.1 Historical movements
in M1 and M2 velocities.
3Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-7
Living with Velocity (Cont.)
• Figure 28.1 (Cont.)
– Velocity of M1 appears highly unpredictable
• Reached a peak of about 4 at the beginning of the 1920s
• Fell almost continuously during the Great Depression and
World War
• Experienced a dramatic increase since the low in 1949
reaching almost 9 during the early 2000s
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-8
Living with Velocity (Cont.)
• Figure 28.1 (Cont.)
– Main reasons for the post-World War II rise in
velocity of M
• The relatively narrow definition of M1
• Increasing attractiveness of other categories of financial
assets as prudent/desirable places to invest excess cash
• Higher interest rates experienced during the 1970s and
early 1980s increased opportunity cost of holding cash
• This lead to an increase in velocity by inducing business
firms and households to economize on money
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-9
Living with Velocity (Cont.)
• Figure 28.1 (Cont.)
– However, since the mid-1980s, M1 velocity has
reversed the post-World War II uptrend forcing the
Federal Reserve to abandon setting M1 targets and
focus on the more stable behavior of M2
4Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-10
The Demand for Money
• The historical movements in the velocity of M1
are a result of many factors
– Movements in interest rates
– Technology
– Innovations in financial markets
• In order to be able to focus more precisely on
the relative importance of each, must be able to
disentangle their separate influences
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-11
The Demand for Money (Cont.)
• Formal application of statistical techniques to historical
data of economic variables will permit identification of
individual relationships
• Statistical studies show that interest rates and the level
of GDP influence money demand
– Higher rates reduce demand for cash balances
– Contradicts extreme forms of monetarism, which assume zero
interest-sensitivity of money demand
– However, none of the empirical studies isolate the Keynesian
liquidity trap
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-12
The Demand for Money (Cont.)
• More sophisticated approach focuses on the
stability of the money demand
– Recognizes that specific numbers are less important
than ability to forecasting future money demand
– If this relationship can be established, Federal
Reserve can gauge proper amount of money to
add/subtract to hit a target of economic activity
– If this relationship is unpredictable, it provides no
useful guidance to the Federal Reserve
5Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-13
The Demand for Money (Cont.)
• Empirical evidence
– Demand for money was quite stable until the mid-
1970s
– Deteriorated after the middle of 1974 with people
holding smaller money balances than the historical
relationship suggested
– The observations in the 1980s were the opposite—
holding of too much money reflecting the decline in
M1 velocity
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-14
Econometrics and Time Lags
• The simple velocity and money demand approaches
leave much to be desired as a guideline for Federal
Reserve policy-making.
• Ignores time lags between changes in monetary policy
and impact on economy
• Ignores more sophisticated computer driven
statistical methodology that can simulate the impact of
monetary policy via formal econometric modeling
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-15
Econometrics and Time Lags (Cont.)
• Lags in monetary policy
– Prompt recognition of what the economy is doing
can be difficult
– Available data are often inadequate and frequently
mixed
– The economy rarely proceeds on a perfectly smooth
course
6Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-16
Econometrics and Time Lags (Cont.)
• Lags in monetary policy (Cont.)
– Recognition lag
• Time associated with getting an accurate understanding of what is
happening or is likely to happen in the future
• Data suggests the Federal Reserve generally starts to tighten only a
few months after a business cycle has reached its trough, while the
move toward easing is somewhat more delayed
• This leads to the conclusion that the Federal Reserve is typically
more concerned with preventing inflation than with avoiding a
recession
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-17
Econometrics and Time Lags (Cont.)
• Lags in monetary policy (Cont.)
– Impact lag—time span from when the central bank starts to
use one of the monetary tools until an effect is evident on the
ultimate objective—aggregate spending in the economy
• It may take weeks before interest rates change
• Changes in credit availability also take time
• Delay before actual spending decisions are affected
• Once monetary policy starts to influence spending, it will most likely
continue to have an impact well into the future
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-18
Econometrics and Time Lags (Cont.)
• Econometric Models
– A mathematical-statistical representation that
describes how the economy behaves
– How do different economic actors (consumers and
businesses) respond to economic stimuli
– Once the relationships are formalized in
mathematical representations, data on past
experience in the real world are used to estimate the
precise behavioral patterns of each sector
7Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-19
Econometrics and Time Lags (Cont.)
• Econometric Models (Cont.)
– The computer model simulates the economy in action and
makes predictions based on the formal relationships
embedded in the model
– However, depending on the theoretical propositions used
to construct the model, the data can produce different
results
– Since Monetarists and Keynesians use different theory,
predictions of models based on their respective theories
would be significantly different
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-20
The Impact of Monetary Policy on GDP
• The Federal Reserve, working with several
universities, has developed econometric models
of behavior of economic aggregates in the U.S.
• These econometric models are an evolutionary
phenomena, constantly being revised and
updated to reflect new and different perspectives
about economic reality
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-21
The Impact of Monetary Policy on GDP
(Cont.)
• Outcomes of these models will vary with
specific conditions of current economic activity
• The Federal Reserve model articulates rather
carefully the impact of monetary policy on
various categories of spending
8Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-22
The Impact of Monetary Policy on GDP
(Cont.)
• Fed’s model shows that a 1% increase in the money
supply raises real GDP by about ½% after one year,
rising over the next two years to reflect the full 1%
increase
• Model demonstrates that monetary policy has a
significant impact even within a year
• However, because of various lags, the continued
impact may cause difficulties for policymakers in
subsequent years
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-23
Fiscal Policy and Crowding Out
• The models test the Monetarists-Keynesian
distinction of the effect of fiscal policy and
crowding out
– Monetarists—increased government spending will
merely displace private spending, leaving little net
impact of fiscal policy on GDP
– Keynesians—“crowding-out” effects are
incomplete, implying that fiscal policy generates
much of the traditional multiplier effect on GDP
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-24
Fiscal Policy and Crowding Out
(Cont.)
• The Federal Reserve model
– Holding the money supply constant
• This results in an increase in the interest rates caused by
additional government borrowing
• An increase in government spending by 1 percent of GDP
increases the level of real GDP by about 1 percent for at
least two years
• It takes more than three years for crowding out to reduce
the impact on real GDP
9Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-25
Fiscal Policy and Crowding Out
(Cont.)
• The Federal Reserve model (Cont.)
– Allowing the money supply to increase
• In this case, the Federal Reserve finances the additional
borrowing/spending by printing money and interest rates
do not increase
• In this case the multiplier effects on GDP are substantial
• The effects on GDP continue even after 4 years
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-26
Fiscal Policy and Crowding Out
(Cont.)
• These models suggest that fiscal policy has a multiplier
of about one without any help from monetary policy,
but an accommodating monetary authority can make
fiscal policy even more effective
• Empirical evidence on fiscal policy confirms the
crowding-out effect, but only if the contractionary
effects on private spending are given substantial time to
work themselves out
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-27
Interest Rates
• Monetarists and Keynesians believe the initial liquidity
impact of an expansionary monetary policy reduces the
level of interest rates
• Eventually both recognize inflationary expectations
generated by excessive expansion of money will raise
interest rates
• Difference between the two rests on how long it takes
for inflationary expectations to counter the initial
liquidity effect
10
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-28
Interest Rates (Cont.)
• Figure 28.2
– Illustrates that levels of short-term interest rates have been
closely related to actual movements in the rate of inflation
– However, the chart also shows substantial intervals of
independent movements in the level of interest rates
– Most econometric models report that interest rates decline
and remain below their original levels for six months
following an expansion of money
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-29
FIGURE 28.2 Interest rates move
with the rate of inflation.
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-30
Interest Rates (Cont.)
• Figure 28.2 (Cont.)
– Similarly, interest rates are above their original
levels for a similar period after an contractionary
monetary policy
– After a year, however, the initial liquidity effect is
reversed, and interest rates move in the opposite
direction
11
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-31
Interest Rates (Cont.)
• It should be emphasized that the response of
inflationary expectations depends crucially on
the initial state of the economy
– Responding very quickly when the economy is very
close to full employment
– Also, a quick response when there is high degree of
concern about possible inflation
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-32
Business Investment
• Theory would suggest interest rates and all types
of investment spending to move in opposite
directions
• However, historically the two almost always
move in the same direction
– This conflicting result stems from the fact that
business investment is dependent upon a number of
factors besides interest rates
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-33
Business Investment (Cont.)
– The model, therefore, violates the assumption of
holding all other influences constant and just
changing the rate of interest
– An increase in interest rates may inhibit investment,
but an increase in investment caused by other factors
may offset the interest rate effect
12
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-34
Business Investment (Cont.)
• Fed’s econometric model can sort out effects of
individual variables and does demonstrate the negative
relationship between interest rates and business
spending
• Most of the effects occur during the last year of a 3
year time period
• The time delay relates to the fact that current business
decisions are not executed usually until several years
later
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-35
Residential Construction
• The impact of monetary policy is felt more promptly
and powerfully on residential construction
expenditures—higher interest, lower residential
construction
• Part of this effect occurs through credit rationing
activities by financial institutions engaged in mortgage
lending
• Construction expenditures by state and local
governments also appear sensitive to the actions of the
monetary authorities
Copyright © 2009 Pearson Addison-Wesley. All rights reserved. 28-36
Consumer Spending
• The Federal Reserve model includes the wealth
effect of modern Keynesians
• Lower interest rates raise the value of financial
securities which increase the level of consumer
spending
• According to the Fed’s model, the importance of
the wealth effect in the overall impact of
monetary policy is quite substantial
Các file đính kèm theo tài liệu này:
- ch28_9544.pdf