5. Wage setting

KAT.TAL.322 Advanced Course in Labour Economics

Nurfatima Jandarova

September 8, 2025

Market imperfections?

Source: Occupational Employment and Wage Statistics (US)

Another possible explanation for variation in observed wages could be that wages in some sectors are affected by market imperfections.

First, we could think about market power. When employer is operating as a monopsonist (sole buyer of labour), then wages in that sector of the economy may be dampened. For example, compare the distribution of wages between workers in fast-food restaurants, waiters and food servers. Fast-food sector is more likely to contain large fast-food chains, which may be operating as monopsonists. (However, note that this can be very difficult to show). We see that hourly wages of fast-food workers are typically lower and more compact.

Second, market imperfections may also arise due to asymmetric information. When employers don’t observe exact productivity types of their workers, they will need to pay wages based on some kind of average information. It means that these wages may be above marginal productivities for some workers, and below - for others. It is again difficult to find a direct evidence in the data that we could attribute to this phenomenon. As a proxy, we could look at the wages of workers in food sector to those of food delivery workers (yellow line in the above figure). It may be interesting because productivity of delivery workers may be easier to measure, especially nowadays with delivery apps and rating scales so ubiquitous. In particular, interesting comparison is between wages of delivery workers and fast-food counter workers. We see that lowest wages of delivery workers are below the lowest wages of fast-food counter workers, and highest wages are above. This could point to the possibility that wage scheme of fast-food counter workers suffer from imperfect information.

Of course, there are many things happening at the same time, and none of the graphs above isolate one specific channel. But they give us some ideas of what might be happening.

We will now study the impact of compensating differentials and market imperfections on wages. In the end, we will review some of empirical papers that attempt to estimate the importance of different channels.

Jobs of varying difficulty

Continuum of jobs with varying difficulty
Productivity such that
also corresponds to effort worker puts in if employed
Compensating wage differentials:

Now, we can consider a case where jobs do vary in the type and difficulty of tasks they involve. We assume that all relevant characteristics of the job can be summarised in a uni-dimensional variable .

Important addition is that now the characteristic is relevant for production function. In particular, marginal productivity of labour is no longer constant, but increasing function of : . Therefore, every job with different will have their own wage in the equilibrium. The labour demand of a firm with job of difficulty will still look very similar to the demand curve we’ve seen on the previous slide. If , then firm wants to employ as many people as possible. If , then firm is indifferent between any amount of labour as FOC is satisfied at any . If , then firm has no interest in employing anyone.

The worker chooses which job to supply her labour to. Once she has chosen a job with characteristic , she has to do all the required tasks. So, in a way, the variable also captures the effort she has to exert to do this job. Her utility function has the same form as before. Therefore, if she works at job she gets utility and if she doesn’t work she still gets utility of zero. Therefore, she would only work if . This time, apart from deciding to work or not, she can also choose the job at which she wants to work. The optimal job should satisfy the FOC

To make sense of this condition imagine what happens if she works at jobs with slightly lower or slightly higher . With slightly lower , she is in a situation where the job is not too difficult for her. Her disutility parameter allows her to tolerate even more difficult jobs, and by doing so, she could also earn higher wages. Therefore, it is not optimal for her to work at less difficult jobs. If she, however, works at a job with a higher , then it is too difficult for her. Even though she’s getting higher wages, her disutility from that job outweighs higher consumption utility.

We can also describe the relationship between and . Using implicit function differentiation of the FOC above, we can show that

Meaning that higher is the disutility parameter of the worker , the less difficult her optimal job is. It makes sense: if the worker really hates putting in a lot of effort, she doesn’t want to choose very difficult jobs.

Notice also that wages are increasing in difficulty . This is because in the equilibrium, and we have assumed that marginal productivity is increasing in . This is called compensating differentials. Jobs with “worse” characteristics have to compensate their workers with higher wages.

Barriers to entry: monopsonistic employer

Start from baseline model

Continuum of workers with utility
Monopsonistic employer

Equilibrium wage where

Equilibrium employment

Labour markets may not operate under perfect competition. In this case, wages do not only reflect marginal productivities and individual preferences.

One example of imperfect competition is when the assumption of free entry fails. For example, workers may not be able to choose jobs and firms because there is only one employer in her labour market. Barriers to entry may arise due to geographic isolation or licencing costs of heavily-regulated jobs.

We will continue with a simple linear utility function of the worker. For simplicity, we again assume that is constant and normalize it to 1. So, we know that the agent works if her and doesn’t work otherwise.

On the firm side, we again assume constant marginal product of labour . So, the firm’s profit is . Because firm is a monopsonist, its labour demand will determine the equilibrium employment in the economy. We already take that into account by using labour supply directly in its profit equation. Then, the firm needs to set the wage that would maximise its profit.

Monopsonistic wage derivation (FOC)

Since (thanks to being a CDF), the labour supply elasticity to wage is also positive .

This means that . Wages paid to workers by a monopsonistic employer are lower than wages paid in perfectly competitive labour market.

We can use this expression to define monopsony power as the ratio between marginal productivity (i.e., wage under perfect competition) and monpsony wage, . It is also clear that the monopsony power clearly depends on the elasticity of labour supply with respect to wages. If it is highly elastic , then monopsonist cannot offer wages that are too low and has to take into account the falling employment. If labour supply is highly inelastic , then monopsonist can set wages “arbitrarily” low.

Imperfect information and adverse selection

Workers are now described by their ability with CDF
- produce units of good
- enjoy leisure utility such that and
Workers enjoy utility
Firms now offer identical jobs and
Firms do not observe true of workers (only see the distribution )

Another form of market failure that can affect wage setting arises due to imperfect information.

We have seen in the previous lecture the case where workers have imperfect information about jobs. There workers could end up with different wages even though their marginal productivities and preferences are identical. We also briefly discussed the implications of an extension where firms are added to the search model. In particular, one of the implications was that firms were willing to offer higher wages when competition for workers is high. This could also be reflected in the wages of a worker with longer working history being much higher than wages of a new entrant, even if their marginal productivities are exactly the same.

This time we will consider what happens to the firm behaviour when they don’t have enough information about workers.

In this model, workers have different abilities . You can think that worker of type has marginal productivity . Firms do not know the true type of worker, but they do know the distribution of worker types in the economy. We also assume that a productive worker (high ) is also more productive at home. Therefore, her leisure utility is also an increasing function of . To ensure that workers have incentives to be employed, we also assume .

Can you describe the equilibrium in this model under perfect perfect competition and full information?

Firms now offer identical jobs (). Because they don’t observe true marginal productivities of their labour force, firm’s have to maximise expected profit. There exist equilibria in this model where the wage schedule depends on type . But in the following slides we will consider pooling equilibrium, where firms do not attempt to discern worker types.

	Thaler and Rosen (1976)	Hwang et al. (1992)
Age	3.890	4.500
	(0.800)
Age	-0.048	-0.096
	(0.009)
Education	3.400	4.870
	(0.550)
Risk	0.035	0.302
	(0.021)
R2	0.41	0.31
Price of life saved (in years of wage)	26.54	227.67
Mean weekly wage	132.65	132.65

	Finland
Type of work	0.016	0.107
	(0.180)	(0.040)
Working conditions	0.070	0.004
	(0.080)	(0.030)
Working times	-0.016	0.048
	(0.070)	(0.040)
Distance to work	0.162	-0.031
	(0.060)	(0.040)
Job security	0.537	0.068
	(0.220)	(0.040)

A	Variance
Total	0.104
No learning by doing	0.096
No monopsony	0.093
No premarket skill variation across jobs	0.05
No premarket skill variation at all	0.008
No search frictions	0.007

B	Variance
Total	0.104
No learning by doing	0.096
No monopsony	0.093
No search frictions	0.086
No premarket skill variation across jobs	0.049
No premarket skill variation at all	0.007

C	Variance
Total	0.104
No learning by doing	0.096
No monopsony	0.093
No nonpecuniary aspects of jobs	0.087
No premarket skill variation across jobs	0.048
No premarket skill variation at all	0.006

D	Variance
Total	0.104
No learning by doing	0.096
No monopsony	0.093
No nonpecuniary aspects of jobs	0.087
No search frictions	0.061
No premarket skill variation across jobs	0.047

1 / 35

5. Wage setting KAT.TAL.322 Advanced Course in Labour Economics Nurfatima Jandarova September 8, 2025

5. Wage setting
Today
Why do wages differ...
Stylised facts
Wage dispersion
Variation by occupation
Market imperfections?
Perfect competition
Jobs of equal difficulty
Jobs of equal difficulty
Jobs of varying difficulty
Jobs of varying difficulty
Workplace safety regulation
Workplace safety regulation
Perfect competition: summary
Imperfect competition
Barriers to entry: monopsonistic employer
Barriers to entry: monopsonistic employer
Monopsonistic employer and minimum wage
Monopsonistic employer and minimum wage
Imperfect information and adverse selection
Imperfect information and adverse selection
Imperfect information and adverse selection
Imperfect competition: summary
Empirical evidence
Estimates of compensating differentials
Estimates of compensating differentials
Estimates of compensating differentials
Estimates of compensating differentials
Estimates of compensating differentials
Determinants of wage inequality
Determinants of wage inequality
Determinants of wage inequality
Summary
References

5. Wage setting

Stylised facts

Wage dispersion

Variation by occupation

Market imperfections?

Perfect competition

Jobs of equal difficulty

Jobs of equal difficulty

Jobs of varying difficulty

Jobs of varying difficulty

Workplace safety regulation

Workplace safety regulation

Perfect competition: summary

Imperfect competition

Barriers to entry: monopsonistic employer

Barriers to entry: monopsonistic employer

Monopsonistic employer and minimum wage

Monopsonistic employer and minimum wage

Imperfect information and adverse selection

Imperfect information and adverse selection

Imperfect information and adverse selection

Imperfect competition: summary

Empirical evidence

Estimates of compensating differentials

Estimates of compensating differentials

Unobserved heterogeneity in productivity

Estimates of compensating differentials

Hwang, Reed, and Hubbard (1992)

Estimates of compensating differentials

Bonhomme and Jolivet (2009)

Estimates of compensating differentials

Lamadon, Mogstad, and Setzler (2022)

Determinants of wage inequality

Taber and Vejlin (2020)

Determinants of wage inequality

Taber and Vejlin (2020)

Determinants of wage inequality

Firm-specific wage premiums

Summary

References

	Finland
	MWP	Wage differentials
Type of work	0.016	0.107
	(0.180)	(0.040)
Working conditions	0.070	0.004
	(0.080)	(0.030)
Working times	-0.016	0.048
	(0.070)	(0.040)
Distance to work	0.162	-0.031
	(0.060)	(0.040)
Job security	0.537	0.068
	(0.220)	(0.040)