What Do We Know About the

Effects of Minimum Wages?

Attila Lindner
University College London
Impact of Minimum Wages

• Great interest for more than a century now (Leonard, 2000)

• Fierce debate that was always about more than simply minimum wages

• Debate between Stigler (1946) and Lester (1947) on the impact of the 1938
federal minimum wage

• George Stigler and the Chicago-style approach (or so-called “price theory”)
• Low skilled workers are homogenous
• Many firms compete for the same type of workers
• Low skilled jobs are standardized
“Neoclassical” Firm’s Problem

• Firms maximize their profits, taking market-level wage as given:

max 𝑝𝑓 𝐿" − 𝑤𝐿"

!!
• The FOC:
𝑝𝑓 # 𝐿" − 𝑤 = 0
or
𝑝𝑓 # 𝐿" = 𝑤
Marginal Revenue Marginal
Product Cost
# $% &
And so 𝐿" = 𝑓
'
“Neoclassical” Firm’s Problem

Firm-level Wage

Firm-level Labor
Supply Curve (is flat)
𝑤∗

Firm-level Marginal
Revenue Product

O 𝐿!∗ Firm-level Employment

Market-level Equilibrium

Market-level Wage Market-level Labor Supply

Curve

w*

Market-level Labor Demand

(∑ 𝑳∗𝒋 )

O L* Market-level Employment
The Effect of the Minimum
Wage

Firm-level Wage

Firm-level Labor Supply

MW Curve at MW
Wages increase
𝑤∗ Firm-level Labor
Supply Curve no MW

Firm-level Marginal
Revenue Product

O 𝐿!∗ 𝐿!∗ Firm-level Employment

Firm-level employment declines
The Effect of Minimum Wages
on Market-level Outcomes

Market-level Wage
Market-level Labor Supply

MW
Wages increase
w* Efficiency Loss

Market-level Labor Demand (∑ 𝑳∗𝒋 )

O L* L* Market-level Employment
Employment declines
The Debate
George Stigler (1946): price theory
• No free lunch
• Minimum Wage decreases employment and leads to efficiency loss
• This is a clear testable prediction of the theory

Richard Lester (1947): common sense approach

• Instead of analysing models, look at the data
• Ask managers about what determines employment
• Most important factor: the output demand
• Less than 20% reported the labor cost as most important
• Conclusion: (small) increase of labor cost will have a limited employment effect
The Consensus
Milton Friedman (1953): “Essays in Positive Economics”
• Assumptions can be unrealistic and simplifying

The billiard player analogy:

Consider the problem of predicting the shots made by an expert billiard player. It
seems not at all unreasonable that excellent predictions would be yielded by the
hypothesis that the billiard player made his shots as if he knew the complicated
mathematical formulas that would give the optimum directions of travel, could
estimate accurately by eye the angles, etc., describing the location of the balls,
could make lightning calculations from the formulas, and could then make the balls
travel in the direction indicated by the formulas.
Our confidence in this hypothesis is not based on the belief that billiard players,
even expert ones, can or do go through the process described; it derives rather
from the belief that, unless in some way or other they were capable of reaching
essentially the same result, they would not in fact be expert billiard players.
The Consensus

Milton Friedman (1953): “Essays in Positive Economics”

• Assumptions can be unrealistic and simplifying

• A good model focuses on the most important mechanisms, instead of

discussing all irrelevant ingredients
• A good model can explain and predict what happens if circumstances
change
• Evidence produced over the 60-80s supports declining employment
• Chicago-style approach/price theory can capture what happens in response
to the minimum wage
Consensus Breaks

Card and Krueger (1994, 1995) and the “credibility revolution”

• Study the impact of increasing the minimum wage in New Jersey
• Difference-in-differences approach (check out last year’s Nobel prize!)
• Control group: Pennsylvania, where there was no minimum wage hike
Effect on Wages
Consensus Breaks

Card and Krueger (1994, 1995) and the “credibility revolution”

• Study the impact of increasing the minimum wage in New Jersey
• Difference-in-differences approach (check out last year’s Nobel prize!)
• Control group: Pennsylvania, where there was no minimum wage hike
• Diff-in-diff
Δ𝐸& = 𝑎 + 𝑋&' 𝑏 + 𝑐𝑁𝐽& + 𝜀&
where 𝑋& is baseline store characteristics
• Exposure design
Δ𝐸& = 𝑎, + 𝑋&' 𝑏- + 𝑐𝐺𝐴𝑃
̃ & + 𝜀2&
(.*(+,!"
where 𝐺𝐴𝑃& = 𝑁𝐽& 3 ,0
,!"
Consensus Breaks
Consensus Breaks

Card and Krueger (1994, 1995) and the “credibility revolution”

• Study the impact of increasing the minimum wage in New Jersey
• Difference-in-differences approach (check out last year’s Nobel prize!)
• Control group: Pennsylvania, where there was no minimum wage hike
• Result: the minimum wage increases wages and employment
• Initial reactions were not welcoming

“Just as no physicist would claim that ‘water runs uphill’, no self-respecting economist
would claim that increases in the minimum wage increase employment. Such a claim, if
seriously advanced, becomes equivalent to a denial that there is even minimal scientific
content in economics, and that, in consequence, economists can do nothing but write as
advocates for ideological interests.”
James Buchanan, Nobel prize in Economics
Consensus Breaks

• Scepticism comes from the fact that the finding contradicts “basic” theories
(e.g. “law of demand”)
• Dangerous implication of this argument:
• Arguing on one hand that our simplifying assumptions might be unrealistic
• Arguing on the other hand that model predictions should be taken seriously, and that
we should not trust data if our ideas are rejected
What is the Lesson?

• Economic models are important as they can capture the most important
mechanisms at play
• Simplicity is value
• But we should not be dogmatic about our models
• We need “credible” research designs, that could be used to reject or accept the
key predictions of our models

• Alan Krueger (2018)

“The idea of turning economics into a true empirical science, where core theories
can be rejected, is a BIG, revolutionary idea.”

• Nobel Prize in Economics in 2021 (David Card, Joshua Angrist, Guido Imbens)
New Minimum Wage
Research
• New minimum wage research shifted the focus from theory to “reduced”
form empirical evidence on employment and wages

• U.S. has been a fertile ground since state-level variation can be exploited
• Two-way Fixed Effects estimation (e.g. Neumark and Wascher, 1993)
• Usage of administrative data (e.g. Card and Krueger, 2000)
• Border-discontinuity design (e.g. Dube, Lester and Reich, 2010)

• But most of the evidence focuses on specific demographic groups (e.g.

teens) or sectors (e.g. restaurants)
• It is possible that MW increases employment in some sectors and decreases it in
others
Frequency Distribution-Based
Approach
Cengiz et al. (2019) develop a novel method to assess the overall employment change
Frequency Distribution-Based
Approach
Frequency Distribution-Based
Approach
Frequency Distribution-Based
Approach
Frequency Distribution-Based
Approach

Assess the change in missing and excess jobs

Frequency Distribution-Based
Approach
• Key empirical challenge: how to get the counterfactual wage distribution
• Apply a Diff-in-Diff idea
• Case study: Washington State
• During 1999-2000, WA raised its MW from $7.51 to $9.18 (in 2016)
• A large, 22% increase in MW
• WA has administrative data on hourly wages
Frequency Distribution-Based
Approach
• Key empirical challenge: how to get the the counterfactual wage distribution
• Apply a Diff-in-Diff idea
• Case study: Washington State
• During 1999-2000, WA raised its MW from $7.51 to $9.18 (in 2016)
• WA has administrative data on hourly wages
Frequency Distribution-Based
Approach
Frequency Distribution-Based
Approach
Frequency Distribution-Based
Approach
Frequency Distribution-Based
Approach
Frequency Distribution-Based
Approach

Cumulative
employment
changes up to that
wage bin
Frequency Distribution-Based
Approach
• Apply the same idea in a difference-in-differences style estimation
• Exploit 136 state-level minimum wage changes
• Use wage bin-by-state-by-period state-panel data from 1979-2016
Frequency Distribution-Based
Approach
Frequency Distribution-Based
Approach
What is the Problem with the
Standard “Price Theory”?
• For an average minimum wage change, we see a clear wage increase and no
disemployment effect
• If anything, employment increases (but the change is statistically non-
significant) à cannot rule out small negative disemployment effects

• But our estimates can rule out modest negative disemployment effects

• What about heterogeneity?

Heterogeneous Responses by the
Level of the Minimum Wage
What is the Problem with the
Standard “Price Theory”?
• For an average minimum wage change, we see a clear wage increase and no
disemployment effect
• If anything, employment increases (but the change is statistically non-significant)
à cannot rule out small negative disemployment effects

• But our estimates can rule out modest negative disemployment effects
• Major challenge for the standard neoclassical model of the labor market

• What about heterogeneity?

• No indication that disemployment effects are more prominent for larger changes
Heterogeneous Responses By
Unemployment Rate
What is the Problem with the
Standard “Price Theory”?
• For an average minimum wage change, we see a clear wage increase and no
disemployment effect
• If anything, employment increases (but the change is statistically non-significant)
à cannot rule out small negative disemployment effects

• But our estimates can rule out modest negative disemployment effects
• Major challenge for the standard neoclassical model of the labor market

• What about heterogeneity?

• No indication that disemployment effects are more prominent for larger changes
• No indication that disemployment effects are more prominent when unemployment rate is
high
What is the Problem with the
Standard Model?
• Workplaces are differentiated (and not homogenous), like products on the market
• The differentiation leads to market power, like “monopolistic competition” in the
product market

The idea:
• Since some workers want to work at my firm, I can hire them even if I pay lower
wages
• This creates a wedge between productivity and wages
• This means I can “exploit” workers by paying them a wage lower than the value of
their work
• In equilibrium, I hire fewer worker and pay them less
Firm’s Problem in the Presence
of Labor Market Power
• Firms maximize their profits:

max 𝑝𝑓 𝐿" − 𝑤 𝐿" 𝐿"

!!

• The FOC:
𝑝𝑓 # 𝐿" − 𝑤 𝐿" − 𝑤 # 𝐿" 𝐿" = 0
or
𝑝𝑓 # 𝐿" = 𝑤 𝐿" + 𝑤 # 𝐿" 𝐿"
or
𝑤 # 𝐿" 𝐿"
𝑝𝑓 # 𝐿" = 𝑤 𝐿" 1+
𝑤 𝐿"
Firm’s Problem in the Presence
of Labor Market Power

• The Elasticity of Firm-level Labor Supply is defined as:

𝜕𝐿" 𝑤 1 1
𝜂= = = #
𝐿
𝜕𝑤 𝐿" 𝜕𝑤 " 𝑤 𝐿" 𝐿"
𝜕𝐿" 𝑤 𝑤(𝐿" )

Note that: 𝜂 > 0

Firm’s Problem in the Presence
of Labor Market Power
• In the presence of imperfect competition, we then have:

1
𝑝𝑓 # 𝐿" = 𝑤 𝐿" 1+
𝜂
Marginal Revenue Marginal
Product Cost

• Compare this to the neoclassical framework FOC!

𝑝𝑓 # 𝐿" = 𝑤
Firm’s Problem in the Presence
of Labor Market Power
• Mark-down:
!
1
𝑝𝑓 𝐿" $ = 𝑤 𝐿"
1
1+
𝜂
Marginal Revenue Mark-down Wages
Product

In the presence of labor market power, there is a wedge between marginal

revenue product and wages
Firms earn “extra profit” on each worker
The extra profit depends on 𝜂 – the firm-level elasticity of labor supply
If 𝜂 is large → we get back the neoclassical framework
Each Employer Operates as a
Small Monopsony
Marginal Cost
Firm-level Wage Upward Sloping Labor
Supply Curve

𝑀𝑅𝑃!∗

Mark-down

𝑤!∗

Firm-level Marginal
Revenue Product

O 𝐿!$∗ Firm-level Employment

The Model of Workplace
Differentiation
Main predictions of the model:
• People feel they are underpaid at their workplace relative to their true
productivity – wages are “marked-down”
• People feel they could find a better paying workplace, but those are not
ideal (they are too far from home, not nice co-workers)
• Workers with similar skills and ability are paid differently at different firms
(there is a firm-level skill premium)
• Efficient firms are larger and pay more to workers with similar skills
The Model of Workplace
Differentiation
What is the impact of minimum wages in this framework?
• The least efficient firms will close
• The most efficient firms increase wages, the mark-down (or profit per worker)
falls
• Since employing is still profitable (though the wedge between productivity and
wage falls), it is still profitable to employ workers
• In fact, firms want to hire more at lower mark-down
Each Employer Operates as a
Small Monopsony
Marginal Cost
Firm-level Wage Upward Sloping Labor
Supply Curve

𝑀𝑅𝑃!∗

𝑀𝑅𝑃!∗
MW
Wages increase
𝑤!∗

Firm-level Marginal
Revenue Product

O 𝐿!$∗ 𝐿!$∗ Firm-level Employment

Employment increases
The Model of Workplace
Differentiation
What is the impact of minimum wages in this framework?
• The least efficient firms will close
• The most efficient firms increase wages, the mark-down (or profit per worker)
falls
• Since employing is still profitable (though the wedge between productivity and
wage falls), it is still profitable to employ workers
• In fact, firms want to hire more at lower mark-down

Employment from the least efficient firms moves to more efficient ones
Reallocation improves efficiency
The new firms pay more (though workers might be made worse off)
Reallocation

Historically, the role of reallocation featured prominently in the

minimum wage debate:
• First (modern time) minimum wages in New Zealand and Australia
• Advocates sought to stop the proliferation of “sweatshops” in the 1890s
(Nordlund, 1997)
• Sweatshops mainly employed women and young workers and paid them substandard
wages
• Many efficient and worthwhile companies employing working class breadwinners lost
market share
• The minimum wage sought to reverse these trends.
Winston Churchill – Trade
Boards Act in 1909
• Trade Boards Act 1909: piece of social legislation passed in the United Kingdom in 1909
• It provided for the creation of boards which could set minimum wage criteria that were
legally enforceable
• Winston Churchill, MP, put the argument for the legislation as follows:
It is a serious national evil that any class of His Majesty's subjects should receive less than a living wage
in return for their utmost exertions. It was formerly supposed that the working of the laws of supply and
demand would naturally of regulate or eliminate that evil. The first clear division which we make on the
question to-day is between healthy and unhealthy conditions of bargaining. […]
[…] But where you have what we call sweated trades, you have no organisation, no parity of
bargaining, the good employer is undercut by the bad, and the bad employer is undercut by the worst;
the worker, whose whole livelihood depends upon the industry, is undersold by the worker who only
takes the trade up as a second string […] where those conditions prevail you have not a condition of
progress, but a condition of progressive degeneration.
Reallocation and the
Minimum Wage
• Dustmann, Lindner, Schönberg, Umkehrer, and vom Berge (2022) provide direct evidence on the
impact of the minimum wage on reallocation

• Reform: Introduction of the minimum wage in 2015

• Data: administrative employer-employee database covering the life history of all German private
sector workers

• Empirical analysis: how wages, employment, and quality of the firm evolved following minimum
wage hikes
• “Control” group
1. What trends do we observe before the introduction of the minimum wage?
2. How did wages evolve at the higher end of the wage distribution?
Wage Growth as a Function of
Wages 2 Years Before
Employment Growth as a
Function of Wages 2 Years Before
Firm Wage Premium as a
Function of Wages 2 Years Before
Firm’s Productivity as a Function
of Wages 2 Years Before
The Effect of the Minimum
Wage by Firm Size
Reallocation: Summary of
Findings
Key Findings:
• Positive and significant effect on wages, no disemployment effect
• MW leads to reallocation of workers to
• firms paying higher wage premium
• firms with lower turnover
• larger and more productive firms

Further Evidence
• Increases commuting time
• MW at highly exposed locations led to
• a decrease in the number of firms
• an increase in average firm size
• an increase in average firm wage premium and productivity
Reallocation: Efficiency and
Welfare

Efficiency
• MW led to worker reallocation from less to more productive firms
• Productive efficiency of worker-firm sorting was increased

Welfare
• Workers’ commuting distance was also increased
• The effect on worker’s welfare is unclear, but our model suggests that
it was increased on average
Conclusions

• MW has limited employment effects at the current levels

• Affects the worker-firm allocation, and could potentially increase
productivity in certain cases
• Prominent interest in MW, since it allows “testing” competing
explanations on how the low-wage labor market operates
• Standard competitive model fails to capture key aspects
• Note: such model is widely used to study the impact of taxes and transfers on low-
wage workers