Microeconomics Final Study Guide

Anti-trust / competition policy

Definition: government regulation limiting negative effects of market power; decides what firms can/cannot do, merger approval, and sector concentration

Terms: U.S. = Antitrust; Europe = Competition Policy

Main Areas:

• Price fixing → collusion, cartels

• Merger policy → prevent harmful concentration

• Abuse of dominant position → exclusionary/monopolistic behavior

• Other areas → market regulation, firm regulation, consumer protection

U.S approach vs. Europe approach

U.S: Consumer-focused

• Antitrust decisions revolve around impact on consumers

• Department of Justice (DOJ) asks:
  “Will this lead to higher prices or worsen outcomes for consumers?”

• Merger approval depends heavily on consumer impacts

• Historically:

  • U.S. has become lenient toward mergers and large dominant firms.

  • Exception: “Trust-busting” periods (e.g., Biden administration taking a more aggressive stance)

Europe: Competitor-Focused

• Focuses on impact on rival firms and overall market structure, not just consumers

• Especially through DG Comp (Directorate-General for Competition, equivalent to the DOJ in the U.S.)

• Known for aggressive treatment of dominant firms, including:

  • Microsoft

  • Apple

  • Google

  • Facebook

  • Intel

  • etc.

The great reversal

Historically, the U.S. and Europe switched roles:

20th Century

• U.S.: Leader in competition policy; actively broke up monopolies

• Europe: More focused on industrial policy (“national champions”)

Today

• U.S.:

  • Loose merger policy, tolerant of dominant firms

  • Occasional trust-busters (like Biden) are exceptions

• Europe:

  • Much stricter and more aggressive toward dominant firms

  • DG Comp leads strong enforcement on tech giants and merger control

Market power leads to/involves

General Trend:
A wide range of indicators show that market power has increased significantly in the 21st century, especially in the United States

Evidence / Indicators:

• Higher estimated markups (firms charging above marginal cost more than before)

• More concentrated ownership (e.g., same investors holding many firms in a sector)

• Higher industry concentration → fewer competitors per industry

• Large number of mergers, many of them unchallenged (e.g., hospitals)

• Rise of superstar firms (esp. big tech): Amazon, Meta, Google, etc

What drives markups up?

Relationship:
Higher markups → higher market power
Firms can charge more above marginal cost

What drives markups up:

1. Low marginal cost (direct effect)

   1. Low marginal cost doesn’t make firms raise price; it directly increases markup because markup = P−MC. When MC falls, the gap P−MC widens automatically, so markup rises even if price stays the same

2. High fixed cost (indirect effect)

   1. High fixed cost doesn’t enter the pricing formula, but it pushes firms to want higher markup because they need more revenue to cover large long-run costs—this only works if the firm has market power

3. Lower elasticity of demand (consumers less sensitive to price changes)

   1. When consumers are less sensitive to price changes, a firm can charge their markups at a higher price without impact.

Historical Forces Behind Rising Markups:

1980s — Reagan Era

• Deregulation across major sectors like transportation

• Much looser antitrust enforcement

• Allowed many mergers

• Belief: bigger firms = more efficient (economies of scale) and more competitive globally

• Result → higher concentration and larger firms

1990s — IT Revolution

• Internet, computers, mobile phones

• Tech firms with high fixed cost + very low marginal cost

• Network effects increased their market power

Right Before COVID

• Even greater concentration in big tech

• Superstar firms dominating key markets

Market power, prices, and industry differences

Concentration vs Prices:
Hard to prove strict causality between higher concentration (more market power → fewer firms in a market, more dominant firms) and higher prices
But strong correlation exists

Why markups increased:
Digital firms like Microsoft and Google have

• High fixed costs

• Extremely low marginal costs

• Network effects
  This structure naturally leads to high markups

But not just tech:
A long-term increase in concentration and prices is also seen in traditional industries such as

• Housing construction

• Health care

Concentrated ownership

Trend

• More concentrated ownership in the 21st century

• Even when the number of competitors stays the same, ownership overlap between competing firms is increasing

What this means
Many major investment funds own shares in multiple competitors within the same industry
Examples

• Berkshire Hathaway

• JP Morgan

• PRIMECAP

• Vanguard

Why overlapping ownership increases market power
When the same investors own multiple competing firms, those firms have weaker incentives to compete aggressively (“rival” firms don’t compete because they share the same shareholder: Pepsi and Coca Cola)
Higher overlapping ownership is associated with

• Higher prices

• Lower output

• Higher market value for firms

Evidence
When institutional investors for unrelated reasons increase their ownership of multiple competitors
→ market prices tend to rise
→ firm values tend to rise
Anecdotal evidence suggests shareholders may pressure managers toward raising prices and reducing output

Causality vs. Reverse causality for healthcare in the U.S.

Healthcare in the US

• Health care is taking up a growing share of the US economy

• Health expenditures are about 17 percent of GDP, which is extremely high compared to other developed countries

Why Americans spend so much

• One explanation is that US healthcare might be higher quality

• But health outcomes do not justify the huge spending gap

• Economists generally agree that market power plays a major role

Mergers and market power

• Healthcare prices have increased alongside rising concentration in hospital ownership

• Market power from mergers is a likely contributor

• Hard to separate correlation from causation

Causality vs reverse causality
Causality:

• When hospitals merge, they become larger and gain pricing power

• This allows them to raise prices

Reverse causality:

• Some hospitals may have already raised prices because they improved performance

• Higher prices made them more attractive merger targets (another hospital wants to merge with them because they look successful)

• This creates an association between mergers and higher prices even if mergers did not directly cause the price increase

Effects of airline deregulation diagram

Deregulation in the airline industry

• Deregulation means the government removes rules about how firms set prices and which routes they can operate

• Firms can now choose prices freely and compete across routes

• More competitors enter the market because of less rules. More competition pushes prices down and expands output

  • When more firms enter and compete, they lower prices to attract customers, and lower prices increase the quantity demanded

Price and quantity effects

• Price falls from p1 to p2

• Quantity increases from q1 to q2

Area A

• Transfer from producers to consumers because consumers pay lower prices

Area B (allocative inefficiency) → decrease

• Deregulation reduces deadweight loss by allowing trades that were previously restricted

  • DWL: value of trades and transactions that are forgone/given up due to monopolies/ tax imposed in the market

Area C (productive inefficiency) → decrease

• More competition pressures firms to lower marginal costs because if they were to lower price to attract consumers, they need to reduce marginal cost for profit

• As marginal cost decreases, the supply curve shifts down from S to S’

Merger policy

Merger policy applies when two firms want to merge, especially if they are large firms or operate in the same industry because a reduction in the number of competitors may reduce competition and harm consumers through higher prices, less choice, and less innovation

• large firms

• operate in same industry

Efficiency gains (Pro-Merger) include merger synergies such as cost savings, economies of scale, combined capabilities, and operational improvements

• merger synergies (cost savings)

• economies of scale

• combined capabilities

• operational improvements

Market power threats (Anti-Merger) include reduced competition, higher prices, less innovation, and reduced consumer choice

• reduced competition

• higher prices

• less innovation

• reduced consumer choice

Property rights

Property rights = who legally controls a resource. Competitive markets only maximize surplus if property rights are well-defined

• competitive equilibrium depends on property rights

• total surplus maximization assumes property rights

Property rights fail when a resource is common and non-excludable → overuse (tragedy of the commons)

• tragedy of the commons

  • England’s enclosures (17th century) →

  • China’s responsibility system (1970s) → overusing farmland

  • Newfoundland cod collapse (1990s) → overfishing

Property rights also fail with externalities → no one owns the harmed resource, so the true social cost isn’t priced

• climate change

Negative externality

A negative externality occurs when an action imposes a cost on another party and leads the market equilibrium activity (overproduction) level to be greater than the socially optimal level

• pollution

• loud music

• smoking a cigarette where smoke harms others

A firm choosing output only considers its private marginal cost MC and ignores its marginal external cost MEC so it equates demand with MC and produces at p0 and q0

Marginal social cost MSC = MC + MEC and when MEC is added to MC the MSC curve identifies the socially optimal point at p* and q*

True cost to society (MSC) = cost to firm (MC) + harm to others (MEC)

Negative externalities → DWL because of too much production (value of trades and transactions SHOULD NOT happen because they harm society)

L represents deadweight loss, the value of the negative externality and the aggregate social cost created by producing between q0 and q* because too much is produced

• MSC > MC

• markets overproduce

• MSC > willingness to pay for units between q* and q̄

• area L = deadweight loss

Positive externality

A positive externality occurs when an action creates a benefit for others and the market equilibrium activity level is smaller than the socially optimal level

• yard work

• public art

• vaccinations

In the positive externality diagram marginal social benefit MSB = D (marginal private benefit) + MEB

• Total society benefit = marginal private benefit + marginal external benefit

The socially optimal output occurs where supply intersects MSB at p* and q*

Underproduction: you only see demand (marginal private benefit) you don’t see marginal society benefit

• Because buyers don’t get the extra benefit they’re creating for society, so they’re not willing to consume as much as society wants

  • Private benefit (you benefit) and social benefit (society benefit)

In contrast to negative externality, DWL because positive externality → too little output → DWL from missing beneficial trades and transactions that SHOULD happen. L represents the aggregate social benefit lost when output is at q0 instead of q*

• MSB > demand

• market underproduces

• efficiency at q*

• deadweight loss between q̄ and q*

Positive externality example - flu vaccination

• Flu vaccination
  • private benefit: you don’t get sick
  • social benefit: you don’t infect others

• A 1% increase in vaccination saves 795 lives and 14.5 million work hours valued at approximately $150 per vaccination
  • 795 lives saved
  • 14.5M work hours saved

• Research question: do parts of the US with higher flu vaccination rates display better health outcomes

• Omitted variable bias: latent variables (e.g., health consciousness) influence both vaccination and health outcomes

• Identification: relevant flu strains are unpredictable at vaccination time creating exogenous variation in effective vaccination

• Market price of a flu shot: $20 to $40

Pigou tax

For a negative externality, a Pigouvian tax equal to the marginal external cost shifts the supply curve upward, causing firms to internalize the external cost and reduce excessive activity.

• Pigouvian tax outcomes include decreased emissions, decreased traffic delays, increased speeds, and reduced congestion in Manhattan’s congestion zone

• emissions decrease

• traffic delays decrease

• travel speed increases

• congestion improves

For a positive externality, a Pigouvian subsidy equal to the marginal external benefit shifts the supply curve downward (or the demand curve upward), causing firms or consumers to internalize the external benefit and increase insufficient activity to the socially optimal level.

Social pressure can also correct externalities

• yard work

• community organizations

• water rights

Pigou taxes make private cost equal social cost without banning activity or micromanaging behavior

Requires government to know correct tax t which is difficult especially for carbon

Starting from p0 and q0 a Pigou tax T representing external harm shifts the supply curve upward until it aligns with the marginal social cost curve MSC where supply meets demand at the socially optimal output

Negative externality:

• Market overproduces

• Fix = Pigouvian tax

• Tax = MEC

• Shifts MC → MSC

• Supply curve shifts up

• Reduces quantity to q* and increases price

Positive externality:

• Market underproduces

• Fix = Pigouvian subsidy

• Subsidy = MEB

• Shifts Demand → MSB

• Supply curve shifts down

• Increases quantity to q* and decreases price

Climate change

• Carbon emissions and global temperatures have increased and although correlation is not causation strong arguments suggest causal links

• Climate change is a global tragedy of the commons because the atmosphere is non-excludable and emissions anywhere affect everyone

• CO₂ emissions strongly correlate with GDP

• To reduce emissions

  • reduce economic activity

  • increase efficiency

  • switch to cleaner energy

• To reduce CO₂ stock

  • reforestation

  • carbon capture

  • direct air capture

Asymmetric information

Asymmetric information is a situation where one party (buyer or seller) knows more than the other about a key characteristic of the good, the transaction, or themselves which affects how the market functions

• buyer knows their health risk better than insurer

• seller knows car quality better than buyer

• worker knows effort level better than employer

Asymmetric information causes mispricing, mis-sorting, market unraveling, and welfare losses because one side has superior private information and the other side cannot correctly price or screen

Two major types of asymmetric information failures

• Adverse selection (hidden information): the problem happens before the transaction because one party has hidden type

  • One side has private information before the deal

  • Hidden type/quality

  • Problem: the people who benefit most from a deal are the ones more likely to participate → leads to "bad selection" into the market

  Example:
  • Unhealthy people are more likely to buy health insurance.
  • Sellers hide car problems when selling a used car → buyers lower price → good cars leave the market.

• Moral hazard (hidden action): the problem happens after the transaction because one party takes hidden actions

  • Happens after the deal is made

  • Hidden behavior/actions

  • People take more risks once they are protected by insurance/contract

Example:
• After getting phone insurance, people take less care of their phone.
• Insured drivers drive less carefully because repairs are covered.

Examples

• adverse selection example: unhealthy buyers are more likely to buy insurance

• moral hazard example: insured people take less care of their phones

Key conceptual difference

• adverse selection = hidden type

• moral hazard = hidden behavior

In strategic situations one player may have more information than the other which changes how both parties play and creates incentives to exploit informational advantages

Economic principle: when people have superior information expect them to try to use it to their advantage

Hidden actions

Hidden actions occur when the choice of one player is not observed by the other and this leads to moral hazard and principal–agent (agency) problems

Example: employer vs worker (sales assistant)

• it is hard for employer to monitor worker’s effort

• worker may shirk

• employer anticipates shirking (neglect responsibility) and offers a low wage

• hidden action: employer cannot observe worker’s actions

• result: moral hazard

LEADS to moral hazard (people behaving more riskier because they’re protected, or in this case, not monitored by employer)

Hidden information

Hidden information occurs when one player knows more about the game or payoffs than the other and this leads to adverse selection and the lemons problem

Example: seller of used cars vs buyer

• buyer is uncertain about quality

• low-quality seller more likely to accept buyer’s offer

• buyer anticipates this and only offers low prices

• market becomes full of lemons

• hidden information → adverse selection (only bad left in the market, full of lemons which are the cheap cars as expensive cars can’t be sold)

• too few transactions of used cars

Adverse selection

Adverse selection occurs when one side of the market possesses private information about their type (quality or risk) and the uninformed side must make an offer based on averages, causing the offer to be disproportionately accepted by worse types and driving better types out

• unhealthy buyers more likely to buy insurance

• sellers more willing to sell low-quality used cars

• workers with low effort more likely to accept low monitoring jobs

When the uninformed party sets a price for the average participant, good types find the offer unattractive and exit first which worsens the composition of the remaining pool

As the pool worsens, the uninformed side must adjust the price in a way that makes even more good types leave, creating a downward spiral that can unravel the market entirely

The general effect of adverse selection is that low-quality goods or high-risk customers disproportionately populate the market, causing trade to shrink or disappear even though mutually beneficial trade existed under full information

Markets combat adverse selection using mechanisms that reveal or certify true quality or risk

• warranties (consumer confidence)

• reputation and branding

• verification such as medical examinations

• mandatory insurance

Adverse selection in health insurance

In health insurance, adverse selection occurs when buyers know whether they are low-risk or high-risk but insurers cannot distinguish types, so a premium based on the average will be selectively accepted by high-risk individuals and rejected by low-risk individuals

When neither buyers nor sellers know the buyer’s type, the insurer uses expected cost and expected willingness to pay to price the contract

• expected cost = 1,000 × 90% + 30,000 × 10% = 3,900

• expected WTP = 3,000 × 90% + 20,000 × 10% = 4,700

• since expected WTP > expected cost, gains from trade exist

• a premium around 4,650 is acceptable to both insurer and buyer

When buyers know their own type but the insurer cannot distinguish them, the same premium of 4,650 is rejected by low-risk buyers (WTP = 3,000) and accepted only by high-risk buyers (WTP = 20,000)

Because only high-risk buyers accept the contract, the insurer infers that acceptance reveals high-risk status
• conditional expected cost = 30,000
• premium of 4,650 produces large losses for insurer

To avoid losses, the insurer must raise the premium to at least 30,000, but at that price even high-risk buyers drop out (WTP = 20,000), causing the entire market to collapse

Hidden information → Insurer charges one price → Low-risk exit → High-risk stay → Insurer raises price → Everyone leaves → Market fails

Adverse selection - death spiral

A death spiral occurs when successive rounds of adverse selection cause healthier or lower-risk individuals to exit an insurance plan, leaving behind sicker and costlier individuals and forcing premiums continually higher until the market collapses

Dynamic adverse selection happens because each premium increase drives out the remaining healthy people, which raises average cost again and triggers yet another premium increase

Empirical example: Harvard University Case

• in 1994, Harvard had about 10,000 full-time employees

• employees could choose between low-cost HMO plans and a higher-cost PPO plan

• 18% of employees enrolled in the PPO in 1994

• the university reduced the employer subsidy for the PPO for 1995

• employee PPO cost difference rose from $361 (1994) to $731 (1995)

• healthy PPO enrollees dropped the PPO because it was no longer worth it

• sicker employees stayed in the PPO, raising its average cost

• premiums increased again the next year, causing even more healthy people to leave

• the PPO market moved toward collapse due to repeated adverse-selection cycles

Adverse selection in credit card

• Credit card markets also exhibit adverse selection because high-interest credit offers are disproportionately accepted by borrowers with bad credit rather than borrowers with good credit

• When only credit-unworthy borrowers accept high-APR offers, expected repayment quality declines and issuers raise interest rates further, pushing good borrowers out and worsening the pool even more

• Adverse selection demonstrates again that with asymmetric information, low-quality participants can drive high-quality participants out of the market

Hidden actions

Hidden actions occur when a principal hires an agent to perform a task but cannot fully observe the agent’s actions, creating misaligned incentives and the risk of moral hazard

In a principal–agent relationship the outcome depends partly on the agent’s effort, but because effort cannot be perfectly monitored the principal must design compensation to align incentives

Example: American Express

• AmEx (principal) wants stores to accept AmEx cards

• consultants (agents) must convince retailers to install equipment

• AmEx cannot observe how hard consultants work because stores are spread out

• the payment scheme must choose between hourly pay (low-powered) or commission (high-powered)

Other relationships also involve a principal, an agent, and a hidden action, and each requires identifying who cannot observe whom

Hidden actions in consulting

• In consulting, moral hazard arises when firms pay consultants hourly because clients cannot observe how much real work the consultant performs and expect the consultant to bill more hours or shift effort toward tasks that benefit the consultant rather than the client

• A solution is to require consultants to submit daily or weekly activity reports so that payment is tied to documented work rather than unobserved effort

Moral hazard

Moral hazard occurs when one party changes its behavior after a contract or agreement because the other party bears the cost and the first party’s actions cannot be observed

Key idea: hidden actions distort incentives, causing individuals to behave more carelessly or take more risks once protected by insurance or coverage

Difference from adverse selection

• adverse selection = hidden type (before the contract)

• moral hazard = hidden action (after the contract)

Example: cracked iPhone screens

• full insurance would cause people to take less care or even intentionally damage screens

• such insurance exists but is extremely expensive because moral hazard would increase claims

Moral hazard in health insurance (ex ante & ex post)

• There are two types of moral hazard on the demand side of health insurance: ex ante and ex post

• Ex ante moral hazard occurs when individuals behave more riskily because they know insurance protects them from financial loss

  • insurance → riskier behavior → greater chance of sickness or accident

  • BEFORE ILLNESS: You take more risks (get more injuries) because you know insurance will cover you

• Ex post moral hazard occurs when individuals change their consumption of healthcare after becoming sick because insurance lowers the marginal cost of care

  • insurance → lower out-of-pocket cost → choose more expensive drugs or surgery

  • AFTER ILLNESS: You know that insurance will pay so you take advantage by using more like taking more tests, etc

• Solutions for insurers include

  • deductibles

  • copays

  • coinsurance

Moral hazard in financial crisis

• Moral hazard in the financial crisis arose because large financial institutions expected government bailouts if they failed, encouraging excessive risk-taking

• The belief that failure would be rescued (“too big to fail”) created the incentive for firms to take on more risk since losses would be socialized while gains would be privatized

• This perverse incentive contributed to the buildup of risky positions before the crisis

• Banks engage in risky behavior because they know the government will incentivize and resolve issues

High-powered vs low-powered incentives

Low-powered incentives such as fixed wages or hourly pay provide stable income but weak incentives for effort because the agent is paid the same regardless of performance

High-powered incentives such as commissions, bonuses, or performance pay tie compensation strongly to measurable output, giving the agent stronger effort incentives

• Advantages of high-powered incentives
  • strong effort alignment
  • can motivate high productivity

• Disadvantages of high-powered incentives
  • high income variability
  • unfair outcomes when results are influenced by luck
  • performance may be difficult to measure or easy to game
  • tasks may be multidimensional and not captured by a single metric

Moral hazard in trucking industry

• In the trucking industry, some firms contract with independent owner-drivers rather than hiring employee drivers because owner-drivers have stronger incentives to maintain their own trucks

• When the trucking company owns the trucks, employee drivers may take less care because they do not bear maintenance costs, creating a hidden-action problem

• Trucking companies evaluate performance based on factors such as

  • driving speed

  • care of the truck

Income distribution

China income distribution:

• China experienced a major shift in income distribution as rapid economic growth pulled hundreds of millions of people out of poverty and expanded the size of the middle class

• The growth of China’s middle class reshaped global income distribution by transforming the world from a bimodal distribution (a two-peak world with rich vs poor countries) into a more unimodal distribution with a large global middle

• China’s economic rise pulled the global income distribution upward, reducing global inequality even though inequality within some countries may still remain high

Within-country inequality:

• Within-country inequality refers to the gap between rich and poor individuals inside the same country regardless of the country’s overall income level

  • the United States has one of the highest within-country inequality levels despite being a high-income nation

• A central insight is that global inequality has decreased because China’s middle class expanded dramatically even though regional inequality across some areas (such as Sub-Saharan Africa vs rich nations) has not fallen

• Across-country inequality measures gaps between nations, while within-country inequality measures gaps within a nation’s population

Income distribution - Sub-Saharan Africa:

• Sub-Saharan Africa shows only a slight shift in income distribution compared to China, and the change is not as large or transformative

Lorenz curves and the Gini index

The Lorenz curve measures the distribution of income within a country by plotting cumulative population on the horizontal axis and cumulative income on the vertical axis to show how equally or unequally income is shared

• Brazil’s Lorenz curve lies far below the diagonal and shows high inequality

• the United States has a curve below the diagonal but not as extreme as Brazil

• the further the curve bows away from the diagonal the greater the inequality

• The 45° diagonal line represents perfect equality because each x% of the population would receive x% of the income

The Gini coefficient is a summary measure of inequality ranging from 0 to 1 and is derived from the area between the Lorenz curve and the diagonal

• the area between the Lorenz curve and the diagonal is multiplied by 2 because the whole unit square has area 1

• a Gini of 0 represents perfect equality

• a Gini of 1 represents maximum inequality

Cross-country inequality refers to income gaps between nations while within-country inequality refers to gaps among individuals in the same country and most modern inequality debates focus on within-country inequality

Income and income inequality

• On average the richer a country is the lower its income inequality because higher-income nations tend to have stronger institutions and redistribution that compress the income distribution

  • Ukraine has low per-capita income but also a low Gini coefficient meaning low inequality

  • Norway and Sweden have high per-capita income and low inequality

  • South Africa has relatively low per-capita income and extremely high inequality

• The United States is an outlier because despite being a rich high-income country it has far more income inequality than most countries at its income level

• On average countries with more income equality have higher hourly productivity and the relationship is positively sloped because equality is associated with better human capital investment and more efficient use of labor

  • countries like Denmark and Norway are highly equal and highly productive

  • countries with low equality indexes tend to have lower productivity levels

Top 1% income

From 1920 to 1980 the share of total income earned by the top 1% generally fell across many countries and then from 1980 to the present day the top 1% income share increased

• the rise after 1980 is associated with policy changes such as Ronald Reagan’s administration in the United States which induced more wealth inequality

• the increased income concentration in the top 1% and in the top .01% and .001% began in the late 20th century

The rise of top income shares took place in many different countries with different varieties of capitalism which shows that the trend is broad and not limited to a single economic model

In the United States the top 1% holds extremely high wealth relative to other countries and wealth inequality is far more extreme than income inequality

Expenditure and income matter less than wealth because expenditure is much less concentrated while wealth is heavily concentrated in the top

Expenditure (consumption, money you spend) is less concentrated than wealth because of life-cycle smoothing which reduces measured expenditure inequality

Life-cycle smoothing explains why expenditure inequality is lower than income inequality because income varies a lot over a lifetime while expenditure changes much less

• young workers earn less

• middle-aged workers earn the most

• retirees earn little income

• but borrowing saving and smoothing behavior keep expenditure relatively stable

• therefore measured expenditure inequality ≈ lower than income inequality because some observed income inequality reflects temporary life-cycle differences rather than permanent class differences

Sources of income inequality

• Capital vs. labor

  • Returns on capital accrued faster than returns on labor

  • Rich people compared to poor people, they save more money in proportion (so the non-consumed wealth becomes more wealth as it accumulates) in contrast to poor people who spend most of their income

• Skilled-biased technical change

  • Ai improvements will favor the rich compared to the poor

  • AI complements high-skilled labor and therefore, replaces low-skilled labor

  • Robots and automation substitute for them. This reduces demand → lowers wages and employment.

  • AI, software, data systems complement them. This increases demand → raises wages

• Market power (monopoly and monopsony)

  • Decline in the power of unions

  • Monopoly of companies like Apple

  • Superstar firms:

    • Very high fixed costs + very low marginal costs allow “winner-take-most” outcomes

    • Firms like Apple, Microsoft, Amazon dominate markets and employ few people relative to revenue

• Globalization

  • Firms have to compete, so shifting from using low-skill workers to high-skill workers

  • Firms hire from poor countries that pay much less (U.S. $15 minimum wage whereas Vietnam is $1-3)

• Superstar effects

  • Ex: the potential to earn power in the 60s (The Beatles) in contrast to now (Taylor Swift) is huge 

    • In the 60s, it’s more limited but now, it’s more available because of the internet (views, social media)

    • Celebrities nowadays have more earning power 

    • Small differences in talent translate into huge differences of income. In 2025, that happens because of technology

    • Actors, athletes, CEOs, authors — tiny quality differences → massive income gaps

Love and inequality

The rich marry the rich (people usually marry into the people who are similar)

• The rich who marry the rich will combine their wealth which perpetuates even more wealth inequality

Identifying discrimination

• Economists have shown that discrimination is present in hiring by examining how women and minorities are treated in competitive settings such as orchestra auditions

• By as late as 1970 only about 10% of musicians in each of the US “big five” orchestras were women despite the large pool of trained female musicians

• In July 1969 during the civil rights movement two Black musicians in the New York Philharmonic a double bassist and a cellist accused the orchestra of racial discrimination which brought attention to bias in the audition process

• In response various symphony orchestras began gradually introducing blind auditions where musicians perform behind a screen so the jury cannot see them and this gradual adoption allows researchers to test whether blind auditions affected female hiring

• The historical data shows a strong correlation between the introduction of blind auditions and the increase in female musicians because as more auditions became blind more women were hired

• Researchers note possible endogeneity since more female musicians being hired could also cause orchestras to adopt blind auditions but the overall timing still supports the presence of bias

• By 1970 about 10% of orchestra members were female while by the mid-1990s that figure increased to about 35%

• Goldin and Rouse estimate that approximately 30% of the increase in female hires was attributable specifically to blind auditions

• These findings confirm that discrimination against women existed when juries could see the candidate because eliminating visual cues increased female hiring and revealed prior biases against women in orchestras

U-Shaped female labor force participation

• Goldin’s research combined economics and history to create a coherent framework for understanding female labor market outcomes and challenged the conventional belief that female labor force participation always rises with development

• Goldin (1990) showed that the long-run pattern of US married women’s labor force participation is U-shaped meaning women worked a lot in early agricultural society participation declined during industrialization and then rose again in the modern service economy

• 1790–1850 Agriculture era had high married women’s participation because rural farm work and household production overlapped and women’s work was informal but economically significant
  • women milked cows tended crops and contributed directly to household production

• 1820–1880 Industrialization reduced married women’s participation because men working in factories earned enough money for the household and factory jobs required high working hours which is not possible for women taking care of children at home, which caused the downward slope of the U-shape

• 1910–1940 Service society expanded and women’s participation began rising again due to major structural and social changes
  • World Wars increased female employment as men left the workforce
  • the 19th Amendment expanded political rights
  • women entered teaching nursing clerical work

• Post-1940 the rise accelerated due to changes that increased women’s autonomy and human capital
  • contraceptive pill allowed women to control timing of childbirth (less childbirth)
  • changing expectations shifted norms around women working
  • expanding education increased skills and career opportunities

• By the mid-20th century women’s participation rose sharply because higher education fertility control and changing social norms enabled long-term careers forming the upward part of the U-shape

Pill and delayed marriage

• The introduction of the contraceptive pill in the 1960s allowed women especially U.S. college-educated women to delay marriage and childbirth because fertility control gave them more autonomy over timing and long-term planning

• Between the 1930 and 1970 birth cohorts the fraction of college-graduate women married by young ages declined showing a clear trend toward later marriage

  • women born in 1930 had high marriage rates by ages 20–24

  • women born closer to 1970 married significantly later

• Access to the pill made it easier for women to invest in education graduate school and early-career development because they could avoid early unplanned births and coordinate schooling with future family goals

• The pill additionally contributed to rising female labor force participation by enabling women to delay family formation which supported long-run career continuity and allowed entry into higher-earning and professional occupation

Education

• The introduction of the contraceptive pill in the 1960s allowed women to delay marriage and childbirth which increased women’s ability to invest in education and career planning

• Access to the pill and expanding education helped women enter professional programs in large numbers after the early 1970s

  • medicine

  • law

  • dentistry

  • MBA programs

• The modern gender wage gap is not mainly about human capital differences because women today have similar or higher education levels compared to men

• Today the gender wage gap is driven by how the labor market rewards long hours rigid schedules and low flexibility

  • women still face more household and caregiving responsibilities

  • parenthood is the proximate cause of the wage gap

• Workplace inflexibility amplifies the parenthood penalty because high-paid jobs disproportionately reward continuous long-hour work

• Working from home may reduce the gender wage gap by increasing flexibility and reducing penalties for caregiving interruptions

Some philosophical questions and quotes

• The mainstream economics paradigm of rational behavior assumes that individuals are selfish, in the sense that they maximize their wellbeing (their utility)

• Is this a normative statement (“greed is good”) or a positive statement?

• What does an individual’s utility include? Specifically, do individuals care about the wellbeing of others? If so, as much as their own wellbeing? Is there an individual demand for fairness?

• Should a democratic society care about fairness, equity, solidarity? If so, by how much?

• How much value do we need to lose so as to achieve lower inequality?

• How much value are we willing to lose so as to achieve lower inequality

• We are now entering the realm of political economy

Adam Smith

“How selfish soever man may be supposed, there are evidently some principles in his nature which interest him in the fortunes of others, and render their happiness necessary to him, though he derives nothing from it except the pleasure of seeing it.” (Adam Smith)

• People care about others, the fortune of others

Francis Edgeworth

“The first principle of economics is that every agent is actuated only by self-interest.” (Francis Edgeworth)

• People are pure self-interested

Ultimatum bargaining game

• Ultimatum game is probably simplest experiment to test trade-off between personal gain and fairness

• Often repeated laboratory experiment

• Subjects include students, farmers, warehouse workers, and hunter-gatherers

How we play:

• Proposer is temporarily given 100

  • Must choose a share s to offer the responder

  • Proposer keeps 100 − s if offer is accepted

• Responder chooses to accept or reject

  • Accept → payoffs become (100 − s , s)

  • Reject → both receive (0 , 0)

• s can be any value from 0 to 100

  • Represents how much proposer is willing to give player 2

• Game theory prediction from pure rationality

  • A rational responder with no fairness concerns accepts any s > 0

  • Proposer anticipates this and works backwards

  • Proposer’s optimal move under pure rationality is to offer the smallest positive amount

• What the game tree shows (image 1)

  • Player 1 chooses s

  • Player 2 chooses accept or reject

  • Accept branch leads to (100 − s , s)

  • Reject branch leads to (0 , 0)

  • Highlights the take-it-or-leave-it nature of the game

Ultimatum bargaining game: results in Emory University

Player 2 acceptance rates

• Offers near 0 are accepted 0% of the time

• Acceptance rises with s

• Around 30 → ~82% acceptance

• Around 40 → ~94% acceptance

• At 50 → 100% acceptance

• Shows responders have strong fairness preferences

Offer frequency patterns

• 21% of proposers offer 0

• 11% offer 10

• 22% offer 20

• 36% offer 30

• Only 5% offer 40 and 5% offer 50

• Most proposers cluster around 30 because it balances selfishness with a high chance of acceptance

Strategic interpretation

• Proposers are not being altruistic (selfless)

• They are avoiding rejection by offering enough to satisfy fairness expectations

• They expect responders to reject unfairly low offers

• Offers below ~20–30 risk rejection and lead to (0 , 0), which proposers want to avoid

Intemporal favor-exchange game

Stage game

• The game is repeated many times

• Each round, nature generates a potential payoff pair for player 1 and player 2

• The sum of payoffs is usually positive, but one player may get a negative payoff while the other gets a positive one (example 8 , −3)

• Both players simultaneously choose accept or reject

• If either player rejects, both get zero

• This makes it a veto game

Graph explanation

• Left graph shows accepted proposals

• Right graph shows rejected proposals

• The diagonal L-shaped boundary shows all payoff combinations where one player gains and the other loses, or both gain

Accepted proposals (left graph)

• Points above and to the right of the L shape represent cases where both players gain and are accepted

• Points to the left of the vertical boundary represent cases where player 2 takes a negative payoff while player 1 gets a positive payoff

• Points below the horizontal boundary represent cases where player 1 takes a negative payoff while player 2 gets a positive payoff

• Shows that players sometimes willingly accept negative payoffs

Rejected proposals (right graph)

• Inside the L shape, some proposals are rejected even though the sum of payoffs is positive for both players

• Shows that players sometimes reject mutually beneficial outcomes

• Most rejected proposals follow a pattern where one player wins heavily and the other loses heavily

Intemporal favor-exchange game - Why

Ways people care about others’ wellbeing

• altruism

  • I get satisfaction when you benefit

• intrinsic reciprocity

  • I treat you well because you treated me well earlier

• instrumental reciprocity

  • I treat you well now so you will treat me well in the future

Evidence from the game

• Instrumental reciprocity:

  • Players sometimes accept negative payoffs because they expect future favors

  • This behavior collapses in the final round because there is no future opportunity to reciprocate

• The last-period effect shows instrumental reciprocity is strongest

• Intrinsic reciprocity comes next

• Altruism is weakest

Overall conclusion

• Instrumental reciprocity explains most of what we see (majority of favors are driven by people paying you back, not kindness)

• The favor-exchange game supports the idea that people are strategic rather than purely altruistic

• Edgeworth’s idea that people act largely out of self-interest gets support again

Society’s feasible set

It shows all possible welfare (surplus) combinations for two people, A and B

• The shaded area = everything society can achieve

• The curved boundary (E₁, E₂, E₄) = Pareto-efficient points — society is using resources fully; you can’t make one person better off without hurting the other

• Attainable combinations of individuals A and B

  Pareto optimal points (efficient):

  • Ex: at E₁, any movement will make one person worse off

    • Moving in one direction makes A better off but B worse off

    • Moving in the opposite direction makes B better off but A worse off

  • Ex: at E₂, same idea: moving in either direction along the frontier helps one person but hurts the other — no way to make both better off

  • Ex: at E₄, same applies

  Not Pareto optimal (inefficient):

  • Ex: at E₃, society can move outward toward the frontier and make both A and B better off

    • From E₃ → E₂ improves both

    • From E₃ → E₁ improves both

    • From E₃ → E₄ improves both
      → E₃ wastes surplus The interior point E₃ = inefficient — society could make both A and B better off by moving to the frontier

So:

• E₁, E₂, E₄ = efficient

• E₃ = inefficient (wasting surplus)

Pareto allocations

Pareto optimal allocation

• an allocation is Pareto optimal if you cannot make one person better off without making someone else worse off

• these allocations lie on the frontier of the welfare feasible set

First Welfare Theorem

• in a competitive market economy, the equilibrium outcome will lie on the Pareto frontier

Pareto frontier

• competitive markets naturally produce Pareto-efficient outcomes

• when markets are competitive, the equilibrium distribution lies on the frontier of the welfare feasible set

  • When we’re on points like E₁, E₂, and E₄, these are all Pareto optimal.

  • Being on the frontier satisfies the first welfare theorem (any point on the frontier is efficient).

  • This means A and B are maximizing their surplus and the economy is in equilibrium.

Second Welfare Theorem

• Starts at inner points of inefficient points

• any point on the Pareto frontier can be achieved by redistributing initial endowments

• After redistribution, markets will automatically push the economy to that chosen point

• Society can choose ANY point on the Pareto frontier (E₁, E₂, or E₄)

Potential vs. strict Pareto moves

strict Pareto move

• makes at least one person better off and makes no one worse off

  • Ex: from E₃ to E₂, we move onto the frontier and both A and B end up better off (higher surplus)

  • Ex: from E₃ to E₄, both A and B increase their surplus

  • Ex: from E₃ to E₁, both A and B become better off
    → all moves from E₃ outward to the frontier are strict Pareto improvements

• potential Pareto move

  • Policy increases total surplus (A + B both increase), but some individuals lose

  • The effect is big enough that the winners could compensate the losers

  • important in policy because many real policies create winners and losers even when the overall pie grows

Moves like E₃ → E₁ are strict Pareto improvements because both individuals gain, while moves like E₁ → E₂ are only potential Pareto improvements since one person loses but total surplus increases.

Economic and political thought

Libertarians

• respect private property

• want minimal government

• believe helping others is a personal choice, not government’s role

Market-based, socially concerned

• government gives income support or redistributes money

• markets handle most other decisions

• focus on fairness through redistribution + efficiency through markets

Regulated markets, socially concerned

• believe many markets fail or are dominated by big firms

• support government regulation plus redistribution

• mix of markets with strong government rules

Socialists

• think markets often fail

• support heavy government involvement

• prefer public provision of many goods and strong redistribution

Income density distribution

• The #1 determinant of one’s social and economic opportunity is probably the country where one’s born

• By being born in the US, on average, one is 20 or 30 times “better” (in terms of per-capita GDP) than by being born elsewhere

Graph:

The U.S. has much of a higher annual income in compared to the world

• The tail, the lowest of the distribution also has a huge gap

Intergenerational mobility

• Rank elasticity is a measure of intergenerational mobility

• It tells us how much children’s income rank tends to move with their parents’ rank

• If rank elasticity is high (close to 1) → low mobility: kids usually end up in a similar position to their parents

• If rank elasticity is low (close to 0) → high mobility: parents’ income rank doesn’t say much about where kids end up

• In the U.S., rank elasticity is fairly high compared to countries like Denmark or Canada, meaning less mobility

• A transition matrix (or heat map) shows, for each parent income group, where their children end up in the income distribution as adult

• For the U.S., the matrix is darkest along the diagonal, meaning:

  • Kids from poor families usually stay poor

  • Kids from rich families usually stay rich

• This diagonal pattern shows that the U.S. has strong intergenerational persistence and limited upward mobility, especially from the bottom decile

The Great Gatsby curve

• The Great Gatsby Curve shows a positive cross-country relationship between:

  • Income inequality (e.g., Gini coefficient)

  • Intergenerational income persistence (e.g., rank-rank elasticity)

• In countries with higher inequality, children’s income ranks are more strongly determined by their parents’ income ranks → lower mobility

• Interpretation:

  • When inequality is high, advantages and disadvantages solidify across generations, making it harder for children from low-income families to move up

  • Conversely, more equal countries tend to exhibit greater intergenerational mobility

• The U.S. stands out among developed economies for having both very high inequality and very low mobility, placing it toward the upper-right corner of the curve

• Rank elasticity meaning:

  • Elasticity = 1 → parents’ rank fully predicts children’s rank (perfect persistence)

  • Elasticity = 0 → parents’ rank has no predictive power (high mobility)

The fading American dream

• The “Fading American Dream” refers to the sharp decline in the probability that children will earn more than their parents in adulthood

• For those born in 1940, about 92–94% exceeded their parents’ income (even children of high-income parents reached ~88%)

• For those born in 1984, this probability fell to ~50%, meaning upward mobility is now essentially a coin flip

• Part of the decline is due to slower economic growth in recent decades.

• But the drop also reflects a substantial fall in intergenerational mobility itself — where you end up now depends more heavily on where you start

• This decline occurs at every parental income level: children from low-, middle-, and high-income families all saw reduced chances of surpassing their parents