Macroeconomicshomepage.univie.ac.at/robert.kunst/macrob2.pdf · 2017. 10. 22. · there is little unemployment, workers receive an incentive to work harder, as they lose much if they

. . . . . . 1/25

The demand side. . . . . . . . . . . . . . . . . . . . . . . .The supply side

MacroeconomicsBased on the textbook by Karlin and Soskice:

Macroeconomics: Institutions, Instability, and the FinancialSystem

Robert M. [email protected]

University of Viennaand

Institute for Advanced Studies Vienna

October 22, 2017

Macroeconomics University of Vienna and Institute for Advanced Studies Vienna

. . . . . . 2/25


Unemployment

Unemployment: empirical evidence I

1950 1960 1970 1980 1990 2000 2010

24

68

1012

14

%

Austrian unemployment rate according to national definition,monthly since 1950.


. . . . . . 3/25


Unemployment

Unemployment: empirical evidence II

Current (2016) unemployment rates across Europe. (Source Eurostat)


. . . . . . 4/25


Unemployment

Unemployment rates

The Austrian definition of the unemployment rate is the simpleratio

unemployment rate =unemployed

labor force=

unemployed

unemployed+ employed,

with employment excluding self-employment (lawyers, farmers,entrepreneurs), and all numbers taken from registers.

The international definition (EU, OECD) is based onquestionnaires. It includes self-employment. It can be larger(Spain) or smaller (Austria) than the ‘registered rate’.


. . . . . . 5/25


Unemployment

Facts on unemployment

Contrary to the goods and financial market, the labor market doesnot clear. There is unsatisfied supply of labor (unemployment) andunsatisfied demand for labor (vacancies).

Involuntary unemployment can be a matter of strong concern.Unemployed persons are often unhappy or tend to be unhealthy.Employers can use the fear of losing one’s job to discipline workers.


. . . . . . 6/25


Wage and price setting

The wage-setting curve

Currently, most economists describe the labor market by the wage-setting(WS) curve. The WS curve is an increasing function in the (N,W /P)(employment – real wage) space. We use symbols N (employment), W(nominal wages), P (prices).

WS looks like a labor supply curve, with higher wages convincingunemployed persons to intensify their job search. Most economists,however, see WS as distinct from labor supply.

In the efficiency wage interpretation, the WS curve connects those realwages that are paid to discourage shirking. If the economy is ‘good’ andthere is little unemployment, workers receive an incentive to work harder,as they lose much if they get fired.

In the market power interpretation, workers or trade unions can demand a

higher real wage when unemployment is low, as employers fear losing

workers.


. . . . . . 7/25



The wage-setting curve and labor supply

W/P

N

labor supplywage settingWS

res.w.

A

B

If workers receive more than their reservation wage, they supply afixed amount of labor N. With low unemployment as in B, theyget higher wage offers than in A. Unemployment is the differencebetween the green line and the red vertical.


. . . . . . 8/25



Reservation wage and efficiency wage

The reservation wage is the wage that workers require tocompensate the disutility of work and the unemployment benefit.

A minimum wage is a legal minimum for wages.

An efficiency wage is a wage paid in excess of the reservation orminimum wage, in order to discourage shirking and layoffs byworkers and generally to boost workers’ morale.

The vertical distance between the WS curve and the reservationwage represents the cost of losing one’s job.

Labor supply need not be vertical, but it is typically quite‘inelastic’. There is little empirical evidence on a dependencebetween labor supply and real wages.


. . . . . . 9/25



The price-setting curve and labor demand

Under competition, each production factor earns its marginalproduct, i.e. the derivative of the production function w.r.t. thefactor. Labor receives the marginal product of labor as its realwage.

Firms are assumed to be imperfectly competitive. Their price P isa mark-up (µ) on wages. With a simple linear production function,the derivative is constant in the amount of labor demanded N.This horizontal line in the (N,W /P) diagram is called theprice-setting curve (PS).

The interpretation may be that labor demand does not depend onthe amount of labor. Any labor is demanded at a given wage.


. . . . . . 10/25



The wage-setting and the price-setting curve

W/P

N

labor supplyWS

PS

unemployment

labor force

The vertical labor supply line (gray) separates involuntary andvoluntary unemployment.


. . . . . . 11/25



WS and PS: higher unemployment benefits

W/P

N

WSWS1

PS

unemployment

labor force

More generous unemployment benefits increase the reservationwage and shift up the WS curve. This results in moreunemployment. The additional bargaining power will not yieldhigher real wages.


. . . . . . 12/25



WS and PS: less competition, higher markup

W/P

N

WS

PS

PS1 unemployment

labor force

An increase in the markup µ shifts the PS curve down. This resultsin more unemployment and a lower real wage. A higher µ is theconsequence of less competition among producers in the goodsmarket.


. . . . . . 13/25



WS, PS, sometimes not in equilibrium

W/P

N

WS

PS

unemployment

labor force

A

B

A sudden positive demand shock requires more employment thanat the intersection of PS and WS. This situation is in line with theshort-run IS world, but not with the labor market. The short-runpoint will be somewhere between PS and WS, between A and B.


. . . . . . 14/25



How long does the economy live out of equilibrium?

If there is more employment than at the WS-PS intersection, realwages negotiated in the wage rounds are higher than the realrevenues for the producer. Producers tend to increase prices. Therewill be more inflation than anticipated. The economy will moveback to its ‘medium-run equilibrium’ at the WS-PS intersection.

If there is a negative demand shock, there will be deflationarytendencies. Falling or unexpectedly slowly rising prices push theeconomy back to its equilibrium.


. . . . . . 15/25



A formal model for the WS equationThe wage-setting equation can be formalized as

W = PE · B(N, zw ),

where PE is the expected price that serves as the basis in wagenegotiations and zw summarizes wage-push variables that mayreflect the bargaining power of the workers, such as unionization orunemployment benefits. B(., .) is some function that increases inits arguments.

Re-writing this equation for the real wage w as

wWS =W

PE= B(N, zw )

involves approximation, as the expectation of 1/P is not usually1/PE .


. . . . . . 16/25



A formal model for the PS equationThe price-setting equation can be formalized as

P = (1 + µ)W

ΠN,

where µ is the markup and ΠN is the marginal productivity oflabor. This property follows from economic theory: withcompetition, the production factors are paid their marginalproduct. With imperfect competition, there is a markup. Theequation is re-written as

wPS =W

P=

1

1 + µΠN ,

which for a linear production function is a horizontal line in the(N,w) diagram.


. . . . . . 17/25



Equilibrium in the labor marketSuppose price setting can be described more generally by afunction of price push variables, such as a tax ‘wedge’:

wPS = ΠNF (µ, zp),

with F (., ) some function that decreases in its first argument µ (asin F (µ, .) = 1/(1 + µ)). Equilibrium in the labor market holds iff

wPS = wWS ⇒ ΠNF (µ, zp) = B(N, zw ).

With these assumptions, labor-saving technological progress yieldsprofits that are directly distributed to workers. Some economistsmaintain that productivity affects the wage-setting function.

The equilibrium defined by this condition can be expressed as theequilibrium employment Ne , the ‘natural’ unemployment rate ue ,or as the equilibrium or natural output ye .


. . . . . . 18/25



Labor force, labor supply, and output

The labor-market models use N for actual employment (employedpersons) and L for the labor force (potentially available labor),with labor supply typically less than the labor force:

N < labor supply < L

If the production function y = Π(K ,N) is approximately linear inN, then output y will grow and fall with N. In 1962, Okun foundempirically that a 1% growth in y above ‘normal growth’ causesthe unemployment rate to drop by 0.5 percentage points (Okun’sLaw). The Okun’s coefficient of 0.5 may have increased since then.


. . . . . . 19/25


The Phillips curve

Phillips curves

The original Phillips curve from 1958 (after A.W. Phillips) wasan inverse relation between wage inflation and the unemploymentrate: episodes with high unemployment typically have less inflation.

The curve was also presented for price inflation and for changes inthe unemployment rate. Its validity was often debated ontheoretical and empirical grounds.

Current Phillips curves often substitute output or the output gapfor unemployment. Output and unemployment are inverselyrelated, such Phillips curves are usually upward sloping.


. . . . . . 20/25


The Phillips curve

Phillips curve shifting

π

y

PC(πE=2%)

PC(πE=4%)

2%

4%

6%

y0 yH

A

B

C

C0

In A, the economy is in its medium-run equilibrium, WS=PS. Ademand shock moves the economy to B, with more output and 4%inflation. In the next wage round, the 4% are expected and thePhillips curve shifts up. If the higher output yH is targeted again,it comes with an inflation of 6% in C.


. . . . . . 21/25


The Phillips curve

An equation for the Phillips curve

Modern Phillips curves are derived from WS and PS.

Inflation πt evolves from bargaining at wage rounds, where workersnegotiate for a money wage that takes price expectations intoaccount. Their bargaining power increases with the output gapyt − ye , demand in excess of equilibrium output at the intersectionof WS and PS:

πt = πEt + α(yt − ye)

Phillips curves will shift when inflation expectations or equilibriumoutput change by forces exogenous to the curve itself. Withadaptive expectations πE

t = πt−1, the curve becomes

πt = πt−1 + α(yt − ye)


. . . . . . 22/25


The Phillips curve

Can a Phillips curve exist?

Many economists do not like the idea of adaptive expectations.They argue that bargainers learn from observed behavior. Ifinflation increases every year, then this must be accounted for. Ifexpectations are rational, then π − πE can only be a smallunpredictable random error:

πt − πEt = εt = α(yt − ye),

such that α = 0 and there will be no Phillips curve.

A Phillips curve can only exist in the short run, when expectationsof bargainers are ‘fooled’. There can be no systematic relationshipbetween something real (an output gap) and something purelymonetary (inflation).


. . . . . . 23/25


The Phillips curve

An aggregate demand shock

r

yy0 yH

A Bre

IS

IS1

w

N

WS

PS

Ne NH

A Bwe

wH

π

y

PC(πE=2%)

PC(πE=4%)

PC(πE=6%)

2%

4%

6%

y0 yH

A

B

A demand shock shifts the IS curve out to a new short-runequilibrium output yH . The disequilibrium in the labor market canonly be sustained by repeated upward shifts of Phillips curves andinflation expectations. Workers will never receive wH .


. . . . . . 24/25


The Phillips curve

A shock to bargaining power

w

N

WS

WS1

PS

Ne NH

A Bwe

π

y

PC(πE=2%,y0)

PC(πE=2%,yH)

2%

0%y0 yH

A B

A sudden reduction of workers’ bargaining power shifts the WScurve down, a new equilibrium yH ,NH forms. If output does notincrease immediately, a discrepancy y < yH shifts the Phillips curvedown. At the end of the day, the medium-run equilibrium willprevail.


. . . . . . 25/25


The Phillips curve

Equilibria in conflict

In the joint model of IS and WS-PS, the equilibrium in the goodsmarket (IS) will often differ from that in the labor market(’supply’,WS-PS). In the longer run, the labor-market pointWS=PS is ‘stronger’. It is attained by price inflation or deflationvia shifting Phillips curves. Only supply-side policy, such asreducing unemployment benefits, affects this WS=PS point.

Usual short-run economic policy, fiscal and monetary, can play animportant role in this process.

On the other hand, trying to keep output above the medium-runequilibrium (or below, but this target is unlikely) is only possible bynever-ending shifts of Phillips curves. In the long run, it isimpossible.


Documents

Macroeconomicshomepage.univie.ac.at/robert.kunst/macrob2.pdf · 2017. 10. 22. · there is little unemployment, workers receive an incentive to work harder, as they lose much if they