Output, Inflation and Monetary Policy

Output, Inflation and Monetary Policy

Output, Inflation and Monetary Policy

- Mishkin and Serletis Ch. 21-24

- A variation on the ECON 1100 Aggregate Demand – Aggregate Supply model.

- Differences?Focus on inflation rather than the price level.

Central bank:sets interest rates inpursuit of an inflation target.

This aspect of monetary policy is built into the AD curve.

Aggregate expenditure (AE) and the IS Curve (see Ch. 21)

- Aggregate expenditure (AE): planned spending on an economy’s final goods and

services.

- Components of AE:

AE = C + I + G + NX

C = Consumption spending

I = Investment spending

G = government spending

NX = net exports (Export spending – Import spending)

Consumption spending (C)

- Spending on final goods and services by the household sector.

e.g. food, clothing, leisure, consumer durables (furniture, cars etc.)

- Roughly half of AE in Canada.

- What determines C?

Household disposable income (YD): Income (Y) minus taxes net of transfers (T)

i.e. income in household sector’s hands.

e.g, in the textbook algebraic model:

C = C*+ m ∙YD = C*+ m ∙ (Y-T) ‘m’ is a constant (0<m<1)

Marginal propensity to spend (m): share of an additional $1 of YD

spent on C

C* is ‘Autonomous consumption spending’ (C*): consumer spending that

doesn’t depend on disposable income.

- Other determinants of C (these work through the value of C*):

- Household sector wealth (Assets-Debts): another measure of household

sector financial resources.

- Real interest rates (r)

- Saving: an alternative to spending: high r, save more, spend less

- Borrowing costs: high r, less borrowing, less spending.

- Expectations of future income, jobs prospects, etc.

- Demographics (age structure of population)

- Preferences (a microeconomic story)

Investment spending (I):

- Spending on investment goods (adds to physical capital stock):

- plant and equipment

- new residential housing

- inventories

- Roughly 20% of AE in Canada (but volatile!)

- Some key determinants:

- borrowing conditions:

- level of real interest rates (r): cost of borrowing.

- the real (inflation-adjusted) interest rate is the relevant rate

(r = i-i=nominal interest rate; =inflation rate)

- credit market constraints, financial frictions (f*).

- business expectations: is it profitable to expand? (animal spirits)

what return is expected on capital investments?

Optimistic? Pessimistic?

- taxes on business income, tax incentives for investment.

- Text model: I a simple function of I = I* -d∙ (r+f*)

- depends on interest rates (r) and financial frictions (f*):

d is a constant that measures the effect of these variables on I.

- effects of other factors (business expectations, taxes etc) are captured by

I*.

(I* is ‘autonomous’: independent of Y)

Government expenditures (G):

- Public sector spending on goods and services (all levels of government)

- infrastructure spending, office supplies, bureaucrat salaries.

(much of government budgetary spending is on transfers:

- this spending is not part of G as it becomes someone else’s income

and is spent by them.)

- Roughly 25% of AE.

- Determinants:

- Much of this spending is at the discretion of the government.

- So treat it as autonomous (unaffected by Y): G*

- Fiscal policy acts through G as well as through T (taxes net of transfers)

Net export spending (NX): X-M= Exports-Imports

- Spending by foreigners on Canadian goods and services (exports) minus

Canadian spending on foreign goods and services (imports).

(note: some of the spending in C, I and G will be on foreign goods so

import spending is deducted to obtain spending on Canadian output)

- X and M each around 30%-35% of AE.

- Some key determinants of NX:

- Exchange rates: affects the cost of Cdn. vs. foreign goods.

- Cdn. prices vs. foreign prices.

- Level of Cdn. Incomes (affects import spending)

- Level of foreign income (affects foreign spending on Cdn.

goods and services)

- Text equation is very simple:NX = NX* - x∙ r x=a constant>0

- NX* captures the effect of variables other than ‘r’.

- Interest rate (r) has a negative effect via the exchange rate

- rise in Cdn. r (assume foreign r constant)

- Canadian financial assets more attractive.

- foreigners demand more Cdn. $ to buy Cdn. assets.

- Canadians buy less foreign currency to buy foreign assets.

- Extra demand for, less supply of Cdn. $ in foreign exchange

markets: Cdn. dollar appreciates.

- Cdn goods are more expensive (X falls)

- Foreign goods now cheaper for Canadians. (M rises).

- The rise in r has lowered NX.

Equilibrium Aggregate Spending

- Goods market equilibrium:

Aggregate Spending = Output (Y)

AE = Y

- if not then output is changing:

AE > Y more spending than output; firms raise output to satisfy extra demand

AE < Y more output than spending, unsold output, firms respond with lower output.

- In the text model AE depends on Y (via effects of disposable income on C)

- Equilibrium builds in the interdependence of spending, output and income.

i.e. morespending, more output, more jobs and incomes gives still more

spending.

- A change in autonomous spending or a change in the real interest rate cause equilibrium Y to change in the same direction.

- size of the change is larger the initial change in spending due to multiplier

effects.

- Textbookalgebraic version of the model:

AE = C + I + G + NX

C = C*+ m ∙(Y-T)

I = I*- d∙(r+f*)

G = G*

NX = NX* -x∙ r

(m, x, d are positive constants measuring the effect of a RHS variable on a

LHS variable, e.g. if ‘r’ rises by 1 then NX falls by ‘x’)

So:

AE = (C*+I*+G*+NX*) – (d+x)∙r -d∙f* + m ∙(Y-T)

Equilibrium requires:

Y = AE

Y= (C*+I*+G*+NX*) – (d+x)∙r -d∙f* + m ∙(Y-T)

Solving for Y gives the equilibrium level of spending:

Y = (C*+I*+G*+NX*-d∙f* -m∙T )∙ – r

Autonomous spending multiplier = 1/(1-m)

Interest rates have a negative effect on equilibrium Y

( ↑r by 1 then ↓Y by (d+x)/(1-m)

IS curve (next page) plots this negative relationship.

IS Curve:

- Text model: Y = (C*+I*+G*+NX*-d∙f* -m∙T )∙ – r

- IS curve shows the Y vs. r relationship for which Y=AE.

- Negative slope reflects effect of the interest rate (r) on I, NX in the

textbook model (could also affect C).

At a given level of r:

- if at a point to right of IS: Output (Y) > Spending so output falls.

- if at a point to left of IS: Spending > Output, output rises.

- must be on the IS curve for Y to be unchanging.

Shifts in the IS curve:

- Increases in autonomous spending, fall in financial frictions, fall in taxes could all

shift IS right.

- Decreases in autonomous spending, rise in financial frictions, rise in taxes could

all shift IS left.

Appendix: An Extended Version of the Textbook Model

- Allows: C to depend on ‘r+f*’; allow import spending to depend on Y-T.

AE = C + I + G + NX

C = C*+ m ∙(Y-T) - a∙(r+f*) (added term)

I = I*- d∙(r+f*)

G = G*

NX = NX* -x∙ r - b∙(Y-T) (added term)

Solution:

AE = (C*+I*+G*+NX*) – (a+d+x)∙r –(a+d)∙f* + (m-b) ∙(Y-T)

Equilibrium requires:

Y = AE = (C*+I*+G*+NX*) – (a+d+x)∙r –(a+d)∙f* + (m-b) ∙(Y-T)

Solving for Y gives the equilibrium level of spending:

Y = [C*+I*+G*+NX*-d∙f* -(m-b)∙T ] ∙ – r

(this is an IS curve with slope: - )

Interest Rate Determination in the model:

- How are interest rates determined in this type of model?

LM Curve approach: Money Demand = Money Supply

- interest rate settles to the level where this is true: driven by shifts

between financial assets and money.

- Central bank conducts monetary policy by changing money supply

which affects interest rates.

- this is probably what you did in first year (origins: Hicks (1937))

MP Curve approach: (textbook, a typical modern approach)

- Treat the central bank as an interest rate setter.

- Central bank follows some policy rule when setting interest rates.

- This is often how central banks talk about the monetary policy.

- Bank of Canada: targets the overnight interest rate.

( )

- a change in the overnight rate can shift the entire structure of

interest rates (substitutes story).

Monetary Policy Curve (MP Curve):

- Behaviour of the central bank?

- Seeks to control the inflation rate(π)

- may have an explicit target rate (Bank of Canada: 2%)

- may be trying to just keep π stable (text p.551 calls this the Taylor

Principle)

- Key link:Interest rate affects spending, which affects inflation.

↑r → ↓Y (move along IS curve, less demand) → ↓π

↓r → ↑Y (move along IS curve, more demand) → ↑π

- Central bank then sets r according to a relationship like:

r = r* + ∙π where >0 measures how much the

central bank changes r in response to inflation.

(Plotting this gives the MP Curve (next page))

- Why might this be a sensible way to model how ‘r’ is determined?

- empirical: it is what many central banks do (since 1990s)

e.g. Bank of Canada 2% inflation target 1-3% range.

- underlying reason for inflation targeting?

- inflation is viewed as costly.

- theoretical viewpoint: in the long-run central bank can affect

inflation and nominal variables but not real variables (like real GDP, unemployment)

MP Curve:

- Moving along MP: if inflation rises the central bank counters by

setting a higher interest rate.

if inflation falls the central bank counters by

setting a lower interest rate.

- Focus is on the real interest rate (r).

- The real rate is most relevant for spending.

- In practice Central bank policy focuses on setting nominal interest

rate levels (iNominal).

- Approximation:iNominal=r + πexpected

- the two will move together if expected inflation ( πexpected) is stable.

- given knowledge of expected inflation the central bank can change

nominal interest rates to achieve a desired real interest rate.

(Technical point: both r and  can sometimes be negative. Soimagine that the

axes scalesallows for this, i.e. 0 is part way along each axis)

Shifts in the MP curve:

- Why might MP shift?

- Central bank tightens or loosens monetary policy.

- changes the interest setting rule: (tighter then r* rises; looser

then r* falls.)

- this may be due to a change in its inflation target

e.g. 2% (midpoint of 1%-3% range) since early-1990s;

implicitly higher  target before this (so MP

curve was lower before 1990s).

- lower inflation target requires less AD, so higher r at

any given π.

- A change in the structure of the economy may require the bank to set

higher ‘r’ for any given inflation rate if it is to achieve its inflation goal.

The MP Curve in Practice:

- A common alternative way of writing the MP curve:

r = rn +  (-T) where T is the inflation target of

the central bank.

(this is the same as earlier if we let: r*=rn - T – this version makes

it explicit that r* depends on the inflation target)

- In practice the behavior the ‘MP curve’seeks to capture can be a quite complex:

- Bank of Canada: forecasting models of the entire economy tell it what

r is needed for any given target inflation rate.

i.e. it takes into account the structure of the rest of the model (exact

relationship of AE to its determinants, input price and inflation expectations determination, value of Yat full employment).

- The simple linear equation in the just text represents the idea that interest rate

setting is done with an inflation objective in mind.

The Aggregate Demand Curve

- Relationship between value of Aggregate Demand (Y) and the actual inflation

rate (π).

- downward sloping relationship: higher  means lower Y.

- in the text version of the model this is due entirely to monetary policy.

Monetary policy:

- MP curve says that ↑π causes central bank to ↑r, this ↓AE and

so lowersY (a move up the IS curve).

Some other reasons for a negatively sloped AD curve?

(2) Wealth effects:

- ↑π reduces real value of money stock and financial assets

denominated in nominal (money) terms.

- lower asset values, people not as wealthy, spend less, ↓AE

(3) International substitution:

- ↑π domestic goods more expensive vs. foreign goods

- ↓Exports, ↑Imports so ↓AE

(4) Uncertainty:

- ↑π may raise uncertainty, people may save more leading to

less AE.

Shifts in the AD curve?

- Shifts in AE (rise in AE, causes rise in AD)

- A decrease in spending creates shifts in the opposite direction.

- AE falls, IS shifts left, AD shifts left.

Monetary policy changes (shifts in MP curve) and the AD curve:

- Looser monetary policy (lower r for any value of )

- lower r

- raises interest sensitive spending (move down the IS curve)

- this gives more spending at the current inflation rate (AD shifts right)

- Tighter monetary policy (higher r for any value of ): reverse the shifts above

- higher r for give  (MP shift up), less spending, AD will shift left.

Aggregate Supply

- Shows relationship between the value of output (real GDP) supplied and the

inflation rate.

Long-run Aggregate Supply (LRAS): vertical at ‘potential output’ (YP)

- Potential output: determined by ‘real’ factors

- quantities and qualities of productive inputs available at full

employment;

- technologies dictating how inputs can be turned into output.

- ‘frictions and imperfections’ in the economy that determine the level

of full employment of inputs and the efficiency of their allocation.

- Inflation does not affect YP.

- LRAS shifts: more inputs, technological improvements, fewer frictions and

imperfections all shift LRAS right over time.

Fewer inputs, greater frictions and imperfections (more

regulation?) can shift LRAS left.

(climate change and LRAS?)

Short-run Aggregate Supply (AS in text, SRAS here):

- Short-run aggregate supply: upward slopingin  vs. Y space

 = e+ ∙(Y-YP) + with >0

- inflation ():

- increases/decreases one-for-one with expected inflation (e)

e.g. if everyone expects 5% inflation then everyone (workers,

businesses) plan to raise their wages and prices by 5% to keep pace.

- inflation also depends on the ‘output gap’ (Y-YP)

- inflation is higher if output is greater than potential

(low unemployment, high input demand drives up input prices, higher costs lead businesses to raise output prices)

- lower inflation if output is below potential (higher

unemployment, weak input demand, leads to low wages, low input prices, low costs and lower prices)

- strength of this effect determines the slope of SRAS.

- increasing in ‘shocks’ (), e.g. shocks to costs of production (oil

price shock; bad harvest)

SRAS curve:

- upward slope reflects effect out output gap in inflation.

- changes in expected inflation, inflation shocks or changes in YPshift

SRAS.

(ASIDE: SRAS in the graphs above are linear – this need not be. A common argument is that SRAS is flatter at low Y, i.e. inflation response to low output is weak especially around zero inflation:

)

SRAS and LRAS:

- Say there is a persistent output gap (so Y≠YP):

- if Y-YP>0 for an extended period people will come to expect

higher inflation (↑e) – this raises inflation and shifts SRAS

upward.

i.e. at Y>YP actual inflation > expected inflation; if this

persists people update their expectations.

- if Y-YP<0 for an extended period people will come to expect

lower inflation (↓e) – this lowers actual inflation and shifts SRAS downward.

- SRAS is only stable if Y=YP, i.e. if on LRAS.

(inflation is at its expected level, neither e or  changes)

Equilibrium in the Aggregate Demand – Aggregate Supply Model: Short-run

- Short-run equilibrium: AD = SRAS (0, Y0 in diagram)

- on SRAS (inflation-output combination consistent with how inflation is

determined).

- on AD curve: inflation-output combination consistent with

- central bank is setting r according to its MP curve; and

- goods market equilibrium (on IS curve): else Y is changing.

(say instead output is Ylow then SRAS says inflation must be low but this implies that AD is YHigh in which case businesses will be raising output to meet this high demand, as they do so inflation rises, etc. – move toward the

intersection)

Equilibrium in the Aggregate Demand – Aggregate Supply Model: Long-run

- A short-run equilibrium is only a long-run equilibrium if Y0=YP

- If Y≠YPactual inflation differs from expected inflation. In the long-run

expected inflation changes and SRAS shifts:

i.e. SRAS shift raises inflation which decreases AD along the AD curve.

i.e. SRAS shift lowers inflation which increases AD along the AD curve.

Effects of a Fall in AD:

- Could be due to fall in autonomous spending; increased financial frictions;

a more anti-inflationary (tighter) monetary policy.

- Short-run: recession – output and inflation both fall (pt. 0 to pt. 1)

i.e. after shift AD<YP at 0, firms ↓output then as ↓Y, ↓ along SRAS.

- Long-run: at pt.1 output is below YP so inflation is below its expected level.

- expected inflation falls, SRAS shifts down, inflation falls still

more and output grows back towards YP.

- new equilibrium is at pt. 2:

Effects of a Rise in AD:

- Short-run: inflation and output both rise (boom!), leaving output above YP (pt.1)

(process: after AD shifts AD>YP at 0, businesses ↑Y, this causes↑ along SRAS)

- Long-run: high output, causes ↑expected inflation, SRAS shifts up. Inflation

rises still further, output returns to YP at pt. 2.

SRAS shifts up (negative supply shock):

- Could be due to a shock to input prices or a rise in expected inflation.

- Short-run (pt. 0 to 1): inflation rises, output falls below YP (stagflation of

mid-late 1970s!)

- Long-run (pt.1 to 0): since Y<YPexpected inflation falls, SRAS shifts back

down, return to pt. 0.

LRAS shifts left:

- Possible cause? fewer inputs; more inefficient production (regulation?):

result potential output is lower.

- Short-run: with lower YP inflation is higher at any Y (SRAS shifts up) and

the economy goes from pt.0 to pt.1.

- Long-run: at pt.1,Y>YP1 so actual inflation exceeds expected inflation,

And in the long-run expected inflation rises shifting SRAS still

more. This continues until Y is at the new lower potential level

of output (pt. 2)

Applications: Text uses the model to think about some business cycle episodes

- Early 1980s disinflation (Figure 23-10)

- Bank of Canada shifts AD left in an effort to lower inflation.

- Recession results, expected inflation falls shifting SRAS down.

- US 2001-2004 (Figure 23-11)

- Same diagram as Canada in the early 1980s.

- Difference? AD shift left is caused by a combination of shocks

(Stock market decline at the end of the ‘tech bubble’, confidence shocks to consumer and business spending from 9/11 and Enron scandal)

- Text shows SRAS shift as moving the economy back to potential at

a lower inflation rate.

- A possible complication? Did the Federal Reserve also shift AD

right? i.e. were the interest rate cuts made larger than those suggested by MP curve?

- Oil shocks of the 1970s (Figure 23-13)

- Rise in oil prices shifts SRAS up.

- Stagflation results (recession, higher inflation)

- Expected inflation fall, shifting SRAS back down.

- US financial crisis 2007-09 (Figure 23-16):

- Increase in financial frictions (AD shifts left) coinciding with rising

oil prices (SRAS shifts up).

- Recession (both shifts reduce Y). Muted effect on inflation (AD

shock lowers it, SRAS shock raises it).

- Further shift left in AD as crisis worsens in August 2008.

- Recession deeper, inflation lower by 2009.

- See later discussion of what happens next:

- inflation expectations should be falling (SRAS will shift

down)

- Federal Reserve is attempting to shift AD right; fiscal policy

is mixed (raising AD at Federal level; reducing AD at state level).

- Complications outside the model arise due to zero lower

bound on interest rates.

- China and the financial crisis:

- AD shifts left (fall in spending on exports as trading partners

experience recession)

- Chinese government counters this with fiscal policy: more

government spending which shifts AD back to the right.

Monetary Policy in the AD-AS Model

- Chapter 24 looks at using monetary policy to stabilize the economy.

- MP curve and monetary policy.

- MP curve: the interest rate set by the central bank depends positively on

the inflation rate.

- this already builds in the idea that the central bank is conducting

monetary policy to stabilize inflation.

- so without a change in policy the central bank moves the economy

along the MP curve in response to changes in the inflation rate.

- this chapter is mainly concerned with shifts in the MP curve, i.e. when the

central bank changes its interest rate setting rule.

- so if MP curve is:r = r* + ∙

- concern is with changes to r* (autonomous changes in monetary

policy)

- rise in r*: ‘tightening’ – a more anti-inflationary policy: shifts AD