Keynesian Macroeconomics: Aggregate Supply Mankiw Chapter 13 Williamson Chapter 12

Keynesian Macroeconomics: Aggregate Supply Mankiw Chapter 13 Williamson Chapter 12 1 Aggregate Supply • • • • 2 Frictions in Three Models So far ...
4 downloads 3 Views 214KB Size
Keynesian Macroeconomics: Aggregate Supply Mankiw Chapter 13 Williamson Chapter 12


Aggregate Supply • • • •


Frictions in Three Models

So far focus on Aggregate demand Time to look at Aggregate Supply Curve A bit more careful treatment of SRAS curve (so far an extreme assumption is made: prices are fixed in the short run!)

• Sticky Wage Model • Sticky Price Model • Imperfect Information Model (is in fact a market clearing model) – All three models predict upward sloping SRAS


What determines SRAS?


What determines SRAS? • When the actual price level deviates from its expected level, output deviates from its natural (trend level)

Y = Y + α (P − Pe ) 1/ α : slope of SRAS − Curve 5



Sticky Wage Model

Sticky Wage Model

• In the micro-founded models we assumed that labour market clears in each instant • Here, it is not anymore the case • Longer term wage contracts • Sluggish adjustment of nominal wages • When nominal wage is inflexible in the SR, a rise in the price level lowers the real wage • More hiring • More output

• Nominal wage bargaining: labour unions and efficiency wages keep the target real wage higher than the market clearing wage • W: nominal wage ω: target real wage • After the nominal wage is set and BEFORE hiring materialized, firms observe the actual price level P 7


Labour demand & supply • Hiring is based on the actual real wage • L=Ld(W/P)

• Nominal Wage

– In the micro-founded models Ld was solely determined by marginal productivity of labour NOT Price level)

• Production takes place: Y=F(L)

W = ω Pe

• Model neglects the labour supply decision (once nominal contract is signed, labour supply is guaranteed even if actual real wages are lower than the target real wages! • It implies countercyclical real wages! (opposite of the stylized facts!) • Role of productivity shocks

• Real Wage

W/ P=ω*Pe / P 9





Imperfect Information Model (Lucas Island Model)

Table 10-1

• Goods and labour markets clear in each instant • However SRAS and LRAS differ because of temporary misperception about prices • Suppose an overall price level increases unexpectedly • As a producer (worker) you have to decide how much to produce (offer labour as in the micro-founded model) • You have to make a decision based on the prices you observe around you (e.g. market price of the good that you produce (or help to produce via labour supply))


Imperfect Information Model


Table 11-1

• Mechanism: Nominal wage rise is perceived as being real wage riseÎ Labor supply increasesÎ Labor market eq´m: higher employment and lower real wages ÎLabor realizes that prices have also risenÎ labor supply curve shifts backÎ initial eq´m reestablishes • Confusion between overall change in prices with relative prices

Y = Y + α (P − P e)



Pricing Decision

Sticky Price Model • Firms do not adjust their prices instantaneously w.r.t. AD shocks

• Depends here on two macro-variables • 1. Aggregate price level: the higher the aggregate price level the higher the costs of production, the higher the price of the individual good • 2. Aggregate Income: means also higher demand for the individual good, higher final prices

– Menu costs – Imperfect competition (firms are price setters)




Pricing Decision

Pricing Decision • Desired prices

• Aggregate Prices will thus be

P = sP e + (1 − s )[ P + a (Y − Y )]

p = P + a (Y − Y ) • Two types of firms (flexible and rigid firms) • Flexible firms (1-s):

• Subtract (1-s)P from both sides and rearrange for P


p = P e + a(Y e − Y )

P = P e + [(1 − s ) a / s ] (Y − Y ) 


• Rigid firms (s):

p = Pe 19




• When firms expect higher inflation (higher costs) set their prices high, others follow

• When output (Y) is high, demand is high. Flexible firms set prices high (s matters)


Evidence II

• If there is learning about AD fluctuations or if AD is quite volatile (any change in Y-Ybar reflects rather aggregate price level change instead of relative price changes) α is expected to be smallÎ flat SRAS Curve • If AD is rather stable, a change in Y-Ybar is expected to reflect relative price change , α is expected to be large Î steep SRAS Curve • Lucas (1973): finds evidence in support of imperfect information model, i.e. AD changes have largest impact on Y only in those countries where Prices and AD are in general stable! 23

• Sticky price model: high rate of inflation calls for more frequent price adjustments (high inflation countries face a large α (steep SRAS) (low inflation countries face a low α (flat SRAS) – Confirmed by data



Main Conclusion • Imperfections cause SR monetary nonneutrality • Compatible with LR monetary neutrality


Inflation, Unemployment and Phillips Curve


Deriving the Phillips Curve

• Policymakers objectives may conflict • Three factors drive PC

– Low inflation and high employment – Phillips Curve

– Expected future inflation – Cyclical unemployment – Supply shocks – i,.e.

π = π e − β (u − un ) +υ


• Subtract Pt-1 from both sides

Deriving PC

P − P−1 = P e − P−1 +

• Rewrite AS equation as a function of P

P = Pe +



(Y − Y )





(Y − Y ) + υ

• i.e.

• Add υ to capture an exogenous AS shock (e.g. oil prices, deregulation etc)

P = Pe +


π =πe+



(Y − Y ) + υ

(Y − Y ) + υ 29



• Use Okun’s Law that is



Essential Issue (Y − Y ) = − β ( u − u n )

• PC and /or SRAS brake the classical dichotomy in the Short run

• rewrite AS equation to obtain Phillips Curve such that

π = π e − β (u − u n ) + υ

• policy (more precisely monetary policy) effectiveness

• Classical dichotomy breaks down in the short run!



How to make PC Workable: Adaptive Expectations • backward looking behaviour

• Inflation inertia

• Agents form expectations by looking at the previous period inflation data

• Past inflation affects current expectations that affects wages and prices!

π = π t −1 • • Rewrite Phillips Curve accordingly (NAIRU) e

π = π−1 − β (u − un ) +υ



Causes of Rising and Falling Inflation • Demand pull inflation • Cost push inflation


− β (u − u n )




SR Trade-off between Inflation and Unemployment β determines the trade off between inflation and unemployment

• Early evidence indicated a SR trade off • Now very little support if any! 37


Disinflation • Under adaptive expectations disinflation is costly! • In the absence of favourable supply shock (decline in u via say a decline in energy prices) lowering inflation requires an increase in unemployment and decline in real output . • If agents are rational, (as in the micro-founded models) expectation formation should use all available information. Less pain in terms of output loss! 39


Disinflation and Sacrifice Ratio • Sacrifice ratio: percentage of a year’s GDP (or unemployment: by Okun’s Law) that must be foregone to reduce inflation by 1% point. (estimates: for 1% inflation decline sacrifice about 5% of a year’s GDP)

π = π e − β (u − un ) +υ 1


(Y − Y ) = − β ( u − u n ) 41





• Characterizes long lasting influence of AD on natural rate.. • E.g. A long lasting recession

• • • •

– May cause loss of skills – Reduction in desire to find a job – Insiders/outsiders

Three theories of SRAS Phillips Curve Inertia: adaptive expectations Natural rate hypothesis

– Permanent implications on the labour market – Hysteresis raises the sacrifice ratio (output is permanently lost even if disinflation is over) 43



Suggest Documents