CHAPTER 6 – WAGE SETTING AND UNEMPLOYMENT (Why is there always unemployment in a market

economy?)

The labour market: Introduction

 Why study unemployment?

 Unemployment means that total production and income are lower than they could have been
 Unemployment causes poverty, inequality and social problems
 Why is there always unemployment in a market economy?
 Theories of wage setting and unemployment
 Efficiency wages (setting higher wages than equilibrium so would stay in the market) (low
turnover)
 Search/matching frictions
 Labour unions and bargaining
 Minimum wages
 Skill-biased technical change
 Two types of unemployment:
Focus here is on equilibrium unemployment: the natural rate of unemployment u n (“average rate of
unemployment over the business cycle”).
Cyclic unemployment: u-un will be analysed later
 Labour force surveys:
 N – number employed: whose who have a job
 U – number unemployed those who don’t have a job and are actively seeking
 L=N+U – labour force = employed + unemployed
 OLF – outside the labour force: those who do not have and are not actively looking for work
(retired, students, ‘discouraged workers’)
 u = U/L – unemployment as a share of the labour force
 Stocks and flows – individuals move between jobs and different labour market statuses (flows).
Important to understand stocks and flows

Most “not employed” are OLF

The labour market: stocks and flows

 individuals move between jobs and between different labour market statuses (N, OLF, U)
 we disregard flows in and out of OLF
 notation:
  s = share leaving job for exogenous reasons and seek another
 Z = share of employed workers whose switch if they find another job
 f = probability to find job for employed job seeker
 λf = probability to find job for unemployed
 flows:
 employment -> unemployment: Ns(1-f) (quit but won’t find another job)
 unemployment -> employment: λUf
 employment -> employment: N(s+Z)f

 Let’s first analyse the probability to find a job, f

The labour market: the chance to find a job

 What determines the chance to find a job for someone looking for work? – this depends on how many
people are looking for work but also on how many jobs become available
 s: share of employed who apply for other jobs
 Z: share of employed who apply only if they find a job
 f: probability to find a job
 how many job openings are there each month?
 how many seek a job?
 chance to find job: number of openings divided by number of job seekers:

 the approximation holds since L/N is close to one.

 Example: what’s the probability to find a job if L=100, N=96, U=4 or 10 and s=0.01? Result: the
higher the unemployment rate, the lower the chance to find a job.

The labor market: theories of unemployment

 Our model of “turnover and the job-finding rate” does not say anything about the level of
unemployment. And why do we always have unemployment? We need to analyse wage-setting and
labour demand.
 Five different aspects/theories will be discussed:
 1. Efficiency wages (model of “wage-setting and unemployment”, chapter 6.3)
 2. Search/matching frictions
 3. Labour unions and bargaining
 4. Minimum wages
 5. Technical unemployment

The labor market: a model of wage setting and unemployment

 What wage does the firm set? As low as possible? No, turnover is costly…
  Some employees look for other work and the number depends on the firm’s wage relative to other
companies
 Higher wage: higher wage cost but lower staff turnover
 The company weighs the cost against the benefit of setting a higher wage
 Labor cost of a firm
 hW is a hiring cost proportional to the average wage level
 Z(Wi/W) (relative wage) is the share of on-the-job seekers. Decreases if Wi goes up

 The turnover cost for firm I is then given by:

 The per worker cost is turnover cost + direct wage:

 Costs:
 The wage, Wi, is just a 45-degree line
 Turnover cost, TC, falls as the wage increases
 Per worker cost, Wi + TC is the sum of these two
 What happens if unemployment drops?
 Easier to find other jobs
 Turnover costs increase
 Firm will raise its wage to counteract the increase in turnover costs
 Low unemployment: all firms want to raise wages relative to other firms
 High unemployment: all firms want to reduce wages relative to other firms

Wage line shows how W=MPL,

when wage is bigger, turnover is lower
and cost for it as well.
Z function is decreasing so TC
downward sloping.

 Equilibrium (symmetry):
 All companies set the same wage (Wi/W=1)
 The unemployment rate must be so high that no company wants to change its wage relative to
other companies
 Adjustment: if u<un it is optimal for companies to raise wages and reduce the number of
employees, which means that N will fall (and vice versa)
 
 We can summarize this theory in a wage setting equation:
 Wd=(1+α-bu)W
 Ws is the wage the company wants to set and W is the average wage  in equilibrium, all

companies set the same wage

 Inter-temporal version:

 (u-un) – cyclical unemployment

 Desired wage increase depends on three factors:

 Average wage increase
 Cyclical unemployment (u-un) = deviation from equilibrium unemployment
 The sensitivity of the desired wage level with respect to cyclical unemployment
 Now we know the wage-setting behaviour of the firm. But what are the equilibrium employment and
wage levels?

 Remember:

The labor market: search frictions

 Matching frictions:
 It takes time for an unemployed person to find a job and for a company to find someone to employ
 You also need to find the right worker who want to take the job
 If the matching process works poorly, unemployment will be higher
 Factors that influence the search for employment and the matching process:
 How intensively unemployed workers look for work
 How picky the unemployed are when they choose what jobs they will look for and accept
 Institutions: u-benefits, labor market policies, educational system
 The job-finding rate with search effort:

 where λ is the proportion of the unemployed who search effectively

 If λ falls (workers compete less), f increases and wages rise (wage-setting schedule in 6.10 shifts up
and u increases)

The labor market: labor unions

 Workers from unions and threaten to go on strike if they don’t get the wage be higher wages and
higher unemployment

The labor market: minimum wages

 In many countries there is a statutory minimum wage and most collective agreements include a “floor”
for the wage
 High minimums wage can cause high unemployment, especially among low-skilled groups
 The problem becomes more difficult as recent technical developments seem to favour the highly
educated (skill-based technical change)
 In the early 1900s, assembly lines raised the productivity of low and middle skilled workers
 Today, automatization means lower demand for manual workers and higher demand for very specific
skills
 Dilemma for government and unions: larger income differences or higher unemployment?

The labor market: technical unemployment

 So far, the creation of jobs have outweighed the labour-saving impact of technology, but there is
increasing polarization in the labor market
• A broader range of tasks than just routine tasks may be performed by computers or robots in the future
• Frey and Osborne (2013) estimate that 47% of US workers are in occupations that could be performed
by computers within the next 20 years. Lower according to Arntz, Gregory and Zierahn (2016) (figure
below):



The labor market: why does unemployment differ between countries? – empirical methods

Our theories show that unemployment is affected by the unemployment benefit system, the education
system, the wage-negotiation process, the union structure etc.
• Many of these factors are hard to measure and compare across countries
• Macro studies of differences in unemployment between different countries (‘cross-section’)
• Macro studies of changes over time in unemployment in different countries (‘panel studies’)
• Micro studies using individual data, often studying the effects of specific changes in policy (‘panel
studies’ or ‘repeated cross sections’, ‘difference in difference’)
 Countries with generous unemployment compensation tend to have higher unemployment
• Rise in unemployment when compensation became more generous (early 1970s) and decline when
conditions became stricter (80-90s)
• Micro studies show an elasticity of about 0.5: increase in benefit by 10 per cent increases
unemployment period (duration) about 5 per cent
• Cross-country evidence: strong unions raise unemployment but coordination of wage bargaining
reduces unemployment (The Calmfors-Driffill hypothesis)
• Employment protection and taxes seem less important for unemployment
• Labour market programs show mixed evidence: some programs work, some don’t
• Much focus now on skill-biased technical change and globalisation

The labor market: why has unemployment gone up over time in many countries?

 Persistence: high unemployment in itself seems to lead to continued high unemployment

 Loss of human capital during unemployment period
 Insiders and outsiders
