Transcript Supply Side - Vincent Hogan

Review of
Aggregate Supply
& Aggregate Demand
Overview of AS-AD
• Simple model of Aggregate Supply and
Aggregate Demand
• Same idea as supply and demand
• AD slopes down:
– As price rises AD falls
– summarizes the whole of the IS-LM-BP model
• AS slopes up:
– As price rises AS increases
– Expectations are important
AS
P
AD
Y
Example: Fiscal Policy
P
AS
B
A
AD1
ADo
Y
• If G increases then AD
shifts to right
• At any P, AD will be
higher because G is
higher
• P and Y rise
• Why P rise?
– Need to pay higher w
to get higher output
Aggregate Supply
• Q: What underlies the Aggregate Supply
Curve?
• A: Labour market
• Why? – to increase production
– hire more people
– Pay higher wages
• More of other inputs also
– Focus on labour market
Labour Market
W
D1
Do
L
• Firms have a demand
for labour
• Decreases as wage
rises
• Increases as Price of
output rises
– Curve shifts to the
right
– At any wage firm will
want more workers
• Workers supply labour
• Form expectation
about cost of living
W
– Price expectation (Pe)
• Higher wage will
induce more work(ers)
• Increased (Pe)
L
– Curve shifts up
– Higher wage for any
level of work
Labour market Equilibrium
W
S(Pe)
• Put two curves
together
• Lab Mkt eqm
– Firms and workers
plans agree
– Wage and employment
• Also agree on price
D(P)
L
– P=Pe
– Rational expectations
• Implies agree on real
wages
Derive the Aggregate Supply Curve
• We can use the labour market to derive the
AS curve in short run and in the long run
• First the short run:
– Assume that expectations do not change
– Pe = P
– SL will not move
• Suppose start at A which represents lab mkt
eqm
– P level rises
– DL increases
– New eqm at B
• Employment increases
• Wage increases
– Goods market
• Output increases
– Upward sloping supply curve
• Note that AS depends on expectations
SL(Pe)
W
P
AS(Pe)
B
B
A
DL(P1)
DL(P0)
A
L
Y
• Note that all this implies something weird
about workers
– The work more, for less!
• W has increased, but P has increased by
more
– Real wages have declined
– See diagram
• Supply of labour was conditional on prices
being at a certain level
– Pe = P
– But this no longer true
• With P>Pe we have
• Demand w increase to restore living
standards
– they adjust expectations upwards
– Supply curve shifts up
– Shifts up so that increase in W is same as
increase in P
• New equilibrium is a point C
– Real wage same at C as at A
– Employment same at C as at A
SL(Pe)
W
C
SL(Pe)
P
AS(Pe)
C
B
A
AS(Pe)
B
DL(P1)
DL(P0)
A
L
Y
• As expectations change there is a new AS
curve
• Eqm at C has same output as A
– Reflects the fact that employment is the same
• Net effect is that output and employment
remain the same
• P and W go up by the same amount
Summary of AS
• We have a distinction between short run and long
run
• The Short run is for fixed expectations
– AS(Pe)
– SRAS
– Quite flat: Explains why ISLM works as approx
• LR is how long it takes for real wages to adjust
– Expectations adjust
– Workers to act on exp
– ISLM wont work in LR
• In LR Y is unaffected by P
LRAS
P
AS(Pe)
Y*
Y
• What determines Y*?
–
–
–
–
–
Natural rate
Incentives
Technology
“growth”
Not anything that just affects price
Policy in AS-AD Model
• Suppose there is an increase in G
• AD shifts right
– For all P, there is higher AD, because govt component
has risen
– Could derive this from IS-LM
– Same for MP
• For fixed expectations i.e. SR
–
–
–
–
Move along AS
New (temp) eqm at B
Y increases
P increases (but not by much)
• P rising implies real wage falling
– P>Pe
• Pe will adjust upwards
– W increase
– SRAS shifts up
• Keep going until output returns to “natural
level”
• How long does transition take?
– Theory: depends. Instantaneous?
– Empirics: about 2 years – see diagram
LRAS
P
C
AS(Pe)
B
A
AD0
Y*
Y
AD1
• Be clear on the reasons why there is no long
run effect
– In order to get more output need to pay more
people higher wages
– Higher wages imply firms need to charge
higher prices
– Higher prices negate the higher wages as far as
workers are concerned
– We go back to original values of real variables
– Only affect nominal variables
• Policy is ineffective!
• We can only get an increase in Y in long
run i.e. increase in Y*
–
–
–
–
–
If induce people to work more
Need increase in real wage
Technology
Efficiency
Lower taxes?
• Reganomics
• Supply side economics
• Voodoo economics
Reagan Style Tax Cut
• Cut personal taxes
–
–
–
–
–
Idea is that this will improve incentives
People will work more
Shift the LRAS to the right
Increase Y* and reduce P
Note that SRAS shifts also as expectations adjust to the
new lower level
• But cutting taxes will shift the AD curve to right
– SR boom
– LR return to Y* with higher P
• Which happened?
– Both
– Demand effect larger
P
LRAS
AS(Pe)
AS(Pe)
A
B
AD0
Y*
Y
Dealing With Shocks
• The AS-AD diagram shows how an economy will
automatically adjust to a shock
• Re-adjust to be in terms of inflation
• Start from LR eqm
– Y=Y*
– pe=p
• Suppose there is a fall in AD
– Eqm moves from A to B
– Y<Y*
• This can only be a temporary eqm
• At B, p<pe
– Real wages are higher than expected
• Prices fall, but by more than nominal wages
• See labour market diagram
– Workers are expensive
– Explains the decline in output
– Over time workers will
• Reduce price expectations
• Reduce wage demands
• SRAS shifts down
• Process continues until LR eqm is restored at C
– Real wage returns to original level
– Y=Y*
 pe=p but at new lower level
LRAS
p
SRAS(pe)
A
B
C
AD0
AD1
Y*
Y
• So the economy will automatically work itself out
of recession
• Mechanism depends on wage adjustment
– Mirror image of previous discussions
– Workers respond to lower prices by demanding lower
wages
– Reasonable?
• Yes real wages return to normal
• No long term decline in real wages
– Realistic?
• No! see data
• Nominal wages are rigid
• Have to wait for productivity
– Have lower wage increases than otherwise
• All this takes time
– 3+ years
• Alternative is for Government to expand AD
– Shift AD back
– Return to long run equilibrium A
• Rationale for stabilization policy
– After WTC, cut interest rates
– Enough? Or too much?
• Debate over which is best
– Policy: “long and variable lags”
– Automatic: “long run we are all dead”
– Calls for “flexibility” after EMU
Disinflation
• We can use the model to analyse the issue of
deflation
– i.e. how do we lower inflation without causing (much)
unemployment (lower output)
• This is the flip-side of what we just looked at
– What happens when we try to reduce unemployment
• We already know part of the answer
– LR there is no trade-off
• No LR increase in unemployment (or loss of output)
– SR there is a trade off
• Reducing inflation will mean higher U (lower Y)
– Expectations play a role
LRAS
p
SRAS(pe)
A
B
C
AD0
AD1
Y*
Y
• Suppose we are at A
– Unemployment equals its natural rate (6%)
– Inflation is 12%
– Too high want to reduce it
• Reduce AD (increase interest rates)
– Unemployment rises: 8% > 6%
– Move down the SRAS to B
– Actual inflation falls
• B cannot be LR eqm (why?)
–
–
–
–
Actual Inflation (9%) is lower than expected (12%)
Real wages higher than expected
Expectations adjust down –
New SRAS
• New SR eqm at C
– Still not a LR equilibrium
– Inflation will keep falling for as long as U is
kept above the natural rate
• When the gov. is satisfied with the level of
inflation,
– allow unemployment to return to the natural
rate
– Inflation stable
• Disinflation achieved at a cost
– Y<Y* for several years
– Sacrifice ratio
• The increase in unemployment (above the natural
rate) needed to reduce inflation by one percentage
point in one year.
• My example: 0.66
• Example Volker disinflation
– Inflation fell by 6.7% over 4 years (1982-85)
– Unemployment was 9.5%, 9,5%, 7.4% and
7.1%
– Natural rate was 6%
– Sacrifice ratio 1.4
Painless Disinflation
• Previous example assumed expectations
adjust gradually
• In reality expectations could adjust a lot
quicker
• Think of extreme case
– Expectations adjust fully and immediately
– Full immediate fall in inflation without any
increase in unemployment
– Sacrifice ration of zero
• What is required for this?
– Policy must be announced
– Policy must be credible
• Willing to cause pain
• Intermediate case is more realistic
– More credible policy maker faces a lower
sacrifice ratio