Labor Economics

X is in logarithms, Y is not
a 1% change in X is associated with a change in Y of 0.01B1

Y is in logarithms, X is not
a one-unit change in X is associated with a 100xB1% change in Y; B1(100%)

Log-Log Model
a 1% change in X is associated with a B1% change in Y

The NIT Experiment
Negative Income Tax System (1968-1976; Johnson/Office of Economic Opportunity)
Treatment–Replaced existing tax schedule w/ fixed income guarantee and benefit reduction/tax rate
Looking at how income and variation in the after-tax wage affect labor supply
Evidence primarily focused on males
Overall responsiveness of male labor supply is small, with the elasticity of labor supply wrt after-tax wages estimated at around 0.1.

Elasticity of Male Labor Supply
Income Effect Dominates

Elasticity of Female Labor Supply
Female LFP rates are very responsive to changes in the wage
Among working women, hours of work are not very responsive to changes in the wage
Substitution Effect Dominates

Aid to Dependent Children
Established in 1935 as a minor part of Social Security Act (signed into law by FDR)
$30/month for the elderly; $18/month for the first child and $12/month for the second child

ADC renamed to Aid to Families with Dependent Children in 1962
Categorical (must have dependent child), Means-Tested (income and asset levels can’t be too high), Cash
Yn varies by state and family size–Family of 3 in 2009: $204 in Arkansas, $1464 in Alaska

Established in 1962
Extended welfare benefits to two-parent families in which both parents were unemployed and the principal earner had a significant work history (only 5.3% of recipient families have two parents present)

Marginal Tax Rate
% of the next dollar that you earn that you pay in taxes
% of the next dollar that you earn that you don’t get to keep
(1-change in income/change in earnings)100%

Work Incentive Problems w/ AFDC
1) Single mothers have no incentive to work on welfare (face 100% marginal/implicit tax rate)
2) Some single mothers are induced to exit the labor force (would be working but for welfare)

Welfare Reform (TANF)
1996–Clinton signs PRWORA
New welfare program known as TANF (Temporary Assistance for Needy Families)
1) Cash welfare changed from Entitlement to a Block Grant
2) 5-yr lifetime time limit on benefits
3) 2-yr continuous use time limit (introduces work requirement)
4) Expand EDP
5) States permitted to exempt 20% of recipients from 5-yr lifetime time limit for “hardship” reasons

Work Requirement Tradeoff
1) Will induce some to work while on welfare
2) Will not induce all who were on welfare to work (those with steep indifference curves)

EDP/EDP Tradeoff
Allows welfare recipients to retain some of their earnings while on welfare
brr=implicit tax rate (percentage by which first dollar in welfare benefits is reduced when first dollar is earned)
1) Will induce some to work while on welfare
2) Will not induce some to work while on welfare (those with steep indifference curves)
3) Will give single mothers who were working previously (not on welfare) and who become eligible for benefits after the implementation of an ED program an unambiguous incentive to work less (but they will attain higher utility)

Effects of 1996 Welfare Reform
1) Dramatic decline in welfare caseloads (more than 50% nationally)–about 1/3 of the decline can be attributed to reform
2) Rise in single mother labor supply partly due to reform
3) Single mothers as a group have not seen a drop in their consumption or income as a result of reform (increase in labor supply–rising earnings fully offset falling welfare benefits)

Grogger (2004)–“Time Limits and Welfare Use”
All other things equal, the presence of time limits reduces the welfare participation of families where the youngest child is 3 years old by about 8 percentage points relative to the welfare participation of families where the youngest child is 10 years old.

Gruber (Summary of 1996 Reform)
These reforms, and state reforms that preceded them, appear to have lowered welfare rolls without lowering the incomes of single mothers in the U.S.

Subsidizes wages of low-income wage earners
It is a tax-credit
It is refundable (tax credit>tax liability government sends you a tax rebate)
Introduced in 1976; Government now spends $45 B/yr, compared to $15.2 B in 2011 for TANF

EITC Eligibility
1) Have to have earned income
2) Can’t have too much income (couple w/ 2 children–receive no credit when earned income reaches $43,000)
3) Amount of tax credit depends on family structure

1) Only receive EITC if working; can receive TANF if not working
2) Only receive TANF if you have dependent children; can receive small EITC tax credit if childless)

EITC Region 1 (Increase in Wn)
1) Initially not working–Can induce entry; may not if steep IC
2) Initially working–May increase H if SE>IE; May decrease H if IE>SE
3) Cannot induce exit–If at corner solution when face BC2, must also be at corner solution when face BC1

EITC Region 2 (Increase in Y, no change in Wn)
Will decrease H

EITC Region 3 (Increase in Y, decrease in Wn)
Will decrease H

EITC Tradeoff
1) Will induce some single mothers to enter the labor force
2) Will simultaneously create unambiguous negative work incentives for some single mothers already working a positive number of hours
3) But, after implementation, every single mother who qualifies for the program has higher utility and higher income, for a given number of hours worked, now vs. before

Eissa and Liebman (1996)–Labor Supply Responses to the Earned Income Tax Credit
Effect of EITC expansion (1986) on LFP of single mothers
Statistically significant increase in LFP for single mothers

Eissa and Liebman (1996)–Labor Supply Responses to the Earned Income Tax Credit
Effect of EITC expansion (1986) on the hours of work of single mothers already in the labor force
1) Competing Incentives–SE>IE, work more
2) Not understanding disincentives in the phase-out region
No effect on average hours worked (failed to reject null that population change in hours was 0)

Eissa and Hoynes (1998)–The EITC and the Labor Supply of Married Couples
Effect of the EITC expansions (1984-1996) on the LFP of low-education married women
Effects of the EITC more negative because the EITC is computed on the basis of total family earnings, so the wife is likely to face only the downward-sloping phase-out range in her decision on how much to work
Decreased low-education married women’s LFP in a statistically significant manner
Borjas–a 10% increase in the husband’s wage lowers the participation rate of women by about 5% points

Short-run labor demand elasticity
-0.4 to -0.5

Long-run labor demand elasticity

Elasticity of substitution
Percent change in (K/E)///Percent change in (w/r)

Borjas & the capital-skill complementarity hypothesis
1) unskilled labor and capital gross substitutes, skilled labor and capital gross complements
2) debated
3) at the very least, unskilled labor and capital are much more substitutable than skilled labor and capital

First minimum wage
1938 as part of the Fair Labor Standards Act
Set at 25 cents/hour
Only 43% of nonsupervisory workers were covered
Now the minimum wage is $7.25 and most workers are covered (has been since 2007)

Mike Munger on minimum wage
Opponent of MW because it reduces opportunities for low-skill workers
When wage is artificially high, excess supply of workers, so those receiving $7.25 in a position to be exploited
Finding a job will be hard
MW pushes down non-monetary compensation
There are other channels of adjustment the firm can take (like reducing turnover)

Non-compliance with MW law
2006–MW was $5.15
2.2% of workers earned %5.15 or less
75.8% of THESE workers paid less than $5.15

Covered sector vs. uncovered sector
Less than universal coverage may mitigate adverse employment effects of MW
Free migration b/t sectors can equilibrate real wages in an economy despite the intentions of policymakers

Elasticity of teenage employment (wrt MW)
-0.1 to -0.3
Increase in MW led to statistically significant decrease in teenage employment
2003–10% of workers ages 16-19 earned MW or less
2003–1.7% of workers >age 25 earned MW or less

Card and Krueger (1994)–Minimum Wages and Employment
1992–NJ increased MW to $5.05
1992–PA kept MW at $4.25 (federally mandated MW)
Difference in Difference estimate of the impact of the MW on employment was an increase of about 2.7 workers in the typical fast-food restaurant
Total employment change in NJ was 0.6
Employment change in NJ from economy was -2.1
Employment change in NJ from higher MW was 2.7

MC under Monopsony
MC=new wage+(change in wage)(# previous workers)

Generates and increase in employment
Card and Krueger’s results are therefore theoretically plausible

minimum wage>VMP monopsony>Wpc
Decrease in employment

fast-food supply curve
The supply of labor curve to firms becomes upward sloping when employee mobility is assumed to be costly
Supply curves to firms that typically employ teenagers are upward-sloping and monopsonistic conditions prevail
Card and Krueger–Increase in MW, Increase in employment (if MW exceeds mobility costs)

Employment in monopsonistic sectors in the long run (Russ Roberts)
Could fall if a MW is imposed
With a mandated wage that is not too high, the firm’s MCe is reduced, causing a Sub effect of labor for capital in the SR.
Labor’s average cost, however, has increased.
Now more expensive to produce the same level of output than before, so profits decline, some firms leave the market, which puts downward pressure on employment .
If the latter scale effect is large enough, employment could fall in the LR.

John Schmitt (2013) summarizing Card and Krueger
They found “no evidence that the rise in NJ’s MW reduced employment at fast-food restaurants in the state”

Doucoullagos and Stanley (2009)
1972-2007 meta-study on teenage employment
The most precise estimates were heavily clustered at or near zero employment effects
“corroborate Card and Krueger’s overall finding of an insignificant employment effect (practically and statistically) from MW raises”
1) MW may simply have no effect on employment
2) MW effects might exist, but may be too difficult to detect and/or are very small

Sabia, Burkhauser, and Hansen (2012)–Are the Effects of MW Increase Always Small? New Evidence from a Case Study of NY State
MW increases in NY raised the wages of less-skilled younger workers relative both to similar workers in the control states and to better-educated workers of the same age in NY state.
But they also found robust evidence that raising the NY minimum significantly reduced employment rates of less-skilled, less-educated New Yorkers.
Their estimates implied a median elasticity of around -0.7.

Burkhauser et al. (1996)–Who Gets What from MW Hikes
1989-1992–$3.35 to $4.25
1990–only 7.1% of workers earned between these wages
Many of these workers are teenagers from wealthy households
Only about 19% of the increase in income generated by the higher MW accrued to poor households
More than 50% of the income increase went to households with incomes that were at least twice the poverty threshold
Divided distribution of family incomes into 10 deciles
Among adults in the lowest decile, 80% were below the poverty line and only 1/4 of them worked
Of those who did work, <1/3 earned wages that were less than the MW Thus, even without any loss of employment opportunities, less than 10% of those in the lowest income decile stood to benefit from the 1990-1991 MW increases.

EITC–much better targeted program at low-income working families; funded with general tax revenue so firms aren’t burdened
MW–some benefits going to middle-class teens that we don’t care about; burdens firms

MW recent developments
Channels of adjustment to a higher MW:
1) decrease employment
2) decrease earnings of high-wage workers
3) increase prices
4) lower profits

PBS clip about Seattle MW
both models (PC and monopsony) depict a large decrease in employment if there is a HUGE increase in the MW
Obama proposed raising MW to $10.10 but was shut down
29 states will exceed the federal MW in 2015
Federal contract workers’ wage is $10.10
Seattle MW now $15

Social Security Payroll Tax
12.4% of wages (workers pay 6.2%, firms pay 6.2%)
2013–Social Security Wage Base–Tax levied on first $113,700 of wage income
Ceiling is automatically increased each year in line with wage income in the economy
Tax is not levied on capital income
Fraction of wage income high-income taxpayers pay in SS payroll taxes <<< Fraction of wage income low-income taxpayers pay in SS payroll taxes

Medicare Payroll Tax
Increase and Expansion went into effect in 2013 as part of the Affordable Care Act
2.9% of wages (workers pay 1.45%, firms pay 1.45%) initially
3.8% of wages (workers pay 2.35%, firms pay 1.45%) on income above $200,000 (single) or $250,000 (married)
Tax is levied on ALL wage income
A new 3.8% Medicare tax on “Net Investment Income” on income above $200,000 (single) or $250,000 (married) (previously Medicare taxes only applied to wages)

Rules of Tax Incidence
1) The statutory burden of a tax does not describe who really bears the burden of the tax
2) The side of the market on which the tax is imposed is irrelevant to the distribution of the tax burdens (if there are no impediments to wage adjustment)
3) The tax burden falls more heavily on the side of the market that is more inelastic
4) If there are impediments to wage adjustment (MW), then the side of the market on which the tax is imposed might influence the distribution of the tax burdens (MW in place, tax firms–they bear the whole burden and there is a bigger employment loss than had we taxed workers)

The Targeted Jobs Tax Credit Program
Targeted disadvantages youth, the handicapped, and welfare recipients
Provided employers with a tax credit that lasted one year
Average duration of jobs under this program was 6 months
The subsidy reduced employer wage costs by about 15% for jobs of this duration
Eligibility requirements were stigmatising
When 23-24 yr olds were removed from eligibility for the TJTC in 1989, employment of disadvantaged youths of that age fell by over 7%.

Mandated Benefit vs. Public Provision
1) Issuance of mandated benefit might be more politically feasible than public provision
2) Expect bigger employment loss under public provision vs. mandated benefit (Ls does not shift down w/ public provision because you get the benefit whether or not you work)

Public Provision vs. Mandated Benefit
1) Mandated Benefit only helps those who are employed
2) Because of wage rigidity, those who most need the benefit are the least likely to get hired and receive the benefit when a mandated benefit is implemented
3) Mandated Benefit might fuel the growth of government because their costs are relatively invisible (could lead to excessive spending on social programs because mandated benefits do not appear in the government’s budget)

Borjas–Rate of Return to Schooling in the US in the 1990s

Gruber–Each year of education raises wages by X

Average wage is X% higher for a group that attains 1 yr of college vs. a group of HS graduates

Average wage is X% higher for a group that attains 4 yrs of college vs. a group of HS graduates
80% (sheepskin effect–extra boost to wages from getting degree)

GED graduates vs. HS dropouts
HC–the learning in HS that matters, not the degree
Signaling–GED signals something very different

Which is Bryan Caplan a proponent of?
The Signaling Model
Having a degree in hand matters

Borjas on Signaling vs. HC
Education does more than just signal a worker’s productivity; it also must alter the human capital stock
We would have come up with a less expensive signal by now

For every X females, only X males attain a college degree
Tyler Cowen on WHY
140, 100
People who are conscientious will do better with more free educational resources; women are more conscientious

Charles Murray–Academic Ability and Likelihood of Graduating from College
17/20 white HS seniors @ 90th %ile enter college; 20% expected to fail
2/3 white HS seniors @ 75th %ile enter college; 40% expected to fail
1/2 white HS seniors @ 60th %ile enter college; 52% expected to fail
2/5 white HS seniors @ 50th %ile enter college; 60% expected to fail

Arnold Kling
Large # of college grads doing jobs that don’t require degrees; If there really were excess demand, we wouldn’t be wasting resources like this
We are trying to send too many people to college
Top 100 schools–Completion rate >50%; Why aren’t more people finishing? Not everyone should go!

Bryan Caplan
Think about the marginal student
Bottom 10% get into college–Marginal student might make <90% of the average college kids (who make more) It's not clear that they should be in college

Advantages of Educational Vouchers
Consumer Sovereignty
1) More discretion/flexibility to maximize utility
2) Do not induce family Y to spend less on education
1) Low quality schools forced out of the market
2) High quality schools thrive (more students attend)
3) Level playing field b/t private and public schools (removing financial advantage that public schools get under free public education)

Problems with Educational Vouchers
May Lead to Excessive School Specialisation
1) “Football Schools” w/ little educational provision
2) If regulations become too onerous, defeating the purpose of vouchers (allowing schools to play with the delivery of educational content)
May Increase School Segregation
1) (benefit) Allow minority students w/ more ability/motivation to mix with students at high quality schools (less segregation based on income)
2) (drawback) Separates the education system into higher and lower ability/motivation schools (more segregation based on ability/motivation)
May be an Inefficient and Inequitable use of Public Resources
1) Family Z now receives a subsidy, so they spend less of their own resources on private education
2) (solution) Make vouchers means tested
Government may be reluctant to allow certain schools to go out of business
1) If a large inner-city school closes due to lack of demand, those students might not have other options, so the government is reluctant to let the school close
2) If the government says this school is too important to fail, then the competitive pressure on under-performing schools will be mitigated
The Cost of Special Education
1) Children w/ diagnosed disabilities have much higher costs associated w/ their need for special education
2) The government could pay the difference in costs b/t the types of students or increase the size of the voucher for families w/ disabled student

Rouse (1998)–Private School Vouchers and Student Achievement
Treatment group saw an increase in academic performance
There was a rise in math test scores of 1-2%/yr relative to the control group
No difference in reading scores across the two groups

1) Good information system
2) Provide direct incentives to reward higher-value-added schools
3) Use school funding system in a way that rewards performance; Difficult bc we want to ensure that different students have the resources they need

Characteristics of Age-Earnings Profiles
1) Average earnings of full-time workers rise w/ level of education
2) The most rapid increase in earnings occurs early, thus giving a concave shape to the age/earnings profiles of both men and women; Returns to human capital investments are generally larger when the post-investment period is longer, so we would expect workers’ investments in training to be greatest at younger ages and to fall gradually as they get older
3) Age-earnings profiles tend to fan out, so that education-related differences later in workers’ lives are greater than those early on; HC Theory leads us to expect that workers who invested more in schooling will also invest more in post-schooling job training
4) The age-earnings profiles of men tend to be more concave and to fan out more than those for women; Returns to human capital investments are generally larger when the expected work life is longer, and on average, women are less likely than men to be in the labor force and, if employed, are less likely to work full-time

Why is the age-earnings profile flatter for women than men?
If women themselves expect shorter work lives, they will be less inclined to seek out jobs requiring high levels of training
If employers expect women workers to have shorter work lives, they are less likely to provide training to them
If women expect employers to bar them from occupations requiring a lot of training, they have less of an incentive to seek out those types of jobs
The age-earnings profiles of women have become steeper in recent decades

The National Supported Work Demonstration
Guaranteed people in the treatment group a job for 9-18 months
It takes longer than a decade for the program to reach its breakeven point ($12,500/worker)
There is still a self-selection problem

Inequality in the 1980s
Unambiguous increase
Earnings in the upper and lower halves became more dispersed (increase in 80:50, 50:20, 90:10, 50:10 for men and women)

Inequality in the 1980s (bottom)
Especially pronounced fall in relative earnings at the very bottom of the distribution, indicated downward pressures on the earnings of the lowest-skilled workers
Bigger increase in 50:10 vs. 50:20 for men and women

Inequality from 1990-2011
Earnings generally became less dispersed in the lower half of the earnings distribution
Men–slight decrease in 50:20, slight increase in 50:10
Women–Decrease in 50:20, decrease in 50:10

Inequality from 1990-2011 (upper)
Earnings generally became more dispersed in the upper half of the earnings distribution
Earnings at the 90th percentile for both men and women have pulled farther away from the median than have earnings at the 80th percentile
Men–increase in 80:50, bigger increase in 90:50
Women–no change in 80:50, increase in 90:50

Men (1980-2011)
The real earnings of HS dropouts and HS graduates decreased, and the real earnings of those with a college of graduate education increased
The earnings advantages of acquiring a bachelor’s degree or a graduate degree are higher in recent years than in 1980
The earnings advantages of obtaining a HS degree (as opposed to dropping out) increased slightly over this period

Women (1980-2011)
The real earnings of HS dropouts decreased, but the real earnings of HS graduates increased slightly
The real earnings of those with a college or graduate education increased
All three earnings ratios (HS/Drop, Bachelor’s/HS, Grad/Bachelor’s) increased over this period.

Growth of Earnings Dispersion within Human Capital Groups
Among male college graduates, earnings disparities grew throughout the three decades in each age group
Earnings disparities also grew among male HS graduates in the 1980’s, but afterward, they tended to stabilize or even shrink
So it is likely that the growth in earnings disparities within human capital groups played at least some role in generating overall earnings inequality during the 1980s and afterward

2010–membership rate was 11.4%
Unionized workers tend to have earnings in the middle of the distribution, so the decline in unions could have served to increase the 80:50 or 90:50 ratios
The effects of declining unionization on wage inequality explain about 20% of the growth in inequality for men (but not women) in the 1980s; Played no important role after 1990

Minimum Wage
1981–$3.35; 45% of the average wage for nonsupervisory workers in manufacturing
The legal minimum had fallen to about 1/3 of the average wage by the time it was again increased in the 1990s
This decline in the real minimum wage appears to explain the sharp fall in relative earnings at the very bottom of the earnings distribution during the 1980s
However, the increasing equality in the lower half of the earnings distribution and the relative wage growth in the very upper tail suggest that declines in the real minimum wage, which were again marked after 1997, did not play much of a role after 1990

International Trade
1970–exports/imports was 8%
1996–ratio had risen to 19%; nearly 40% of imports came from less-developed countries
Borjas–Increased foreign trade contributed modestly to the rise in wage inequality, probably accounting for less than 20% of the increase

Reasons for earnings gap between males and females
1) Education
2) Experience–training explains why wage profile is flatter for women
(a) employers provide women w/ less training (unstable)
(b) women expect shorter work lives, so they are less likely to seek out jobs requiring a lot of training
(c) women expect employers to bar them from jobs requiring a lot of training
3) Occupational Choice–women choose lower wage occupations more on average than men (select into jobs w/ less training)
(a) control for occupation and the earnings gap b/t men and women is much smaller

June O’Neill (2003)
Differences in labor market experience explained the largest part of the observed gender gap in earnings
Unadjusted gender gap may be w/ us for a while bc the roles of women are not changing–0.2463
Unexplained gap (diff. due to discrimination)–0.0253

Assessing the Oaxaca Decomposition
Our estimate of wage discrimination (log wage differential of 0.05) might understate the true impact of discrimination by not taking into account the feedback effect
Our estimate of wage discrimination (log wage differential of 0.05) might overstate the true impact of discrimination by not taking into account unobserved differences associated with gender that might affect earnings.

Betrand, Goldin, Katz (2010)–Dynamics of the Gender Gap
Among MBA graduates–women and men earn about the same at first, but after 15 years they earn 40% less
Some of this difference associated w/ fewer hours of work, but most was associated w/ less accumulated experience

In the labor force
Employment-to-population ratio
LFP rate
Unemployment rate
employed + unemployed
in the labor force/population
unemployed/in the labor force

Black-White Differences in Labor Market Outcomes
1) LFP rate for black women higher than LFP rate for white women 1970-2012
2) LFP rate for black men much lower than for white men (7.4% points in 2012) and employment ratio much lower for black men than for white men (11.7% points in 2012)
3) Employment ratio and LFP rate declined 1970-2012 for both black men and white men (reductions greater for black men)

What caused the decline?
1) Increase in college enrolment (lower avg. education for black men)
2) Earlier labor force withdrawal among older men
3) Increase in number of discouraged workers
4) For both men and women, the unemployment rate among blacks is approximately twice that among whites; This constant ratio means that blacks suffer disproportionately in a recession

Lang and Manove (2011)–Education and Labor Market Discrimination
2/3 of the pay gap b/t black and white men could be explained by these observable variables (levels of education, AFQT scores, etc.)
A gap of 11% could not be explained
1) might understate the true impact of discrimination by not taking into account the feedback effect
2) might overstate the true impact of discrimination if the measures of school quality are inadequate (on average whites attend better schools)

Race related wage gaps under employer discrimination
1) Holding human capital constant, race-related pay gaps will be greater when the black population in a region is greater
2) Pay gaps will be larger, other things equal, when the prejudice of the white employers who hire blacks is greater (d value of the marginal discriminator determines relative wage)
3) Pay gaps will be unaffected by the level of prejudice of the most prejudiced employers (those who do not hire blacks)

Employee discrimination model
1) implies a completely segregated workforce
2) unlike employer discrimination, it does not generate a wage differential between equally skilled black and white workers
3) It does not affect the profitability of firms. All firms pay the same price for an hour of labor. There are no market forces, therefore, that will tend to diminish the importance of employee discrimination over time

Customer discrimination model
1) It does generate a wage differential between equally skilled black and white workers

Why is the female Ls higher?
1) Firms must increase wages by a lot to attract more female workers (high search costs)
2) If firms decrease wages a lot, few females will leave (difficult to find another job, might encounter discriminatory firms); VMPm=VMPf but workers with higher search costs are paid less (higher costs because they might encounter discriminatory firms)

An applicant’s expected productivity
w=alphaT + (1-alpha)(avgT)
higher alpha–higher predictive power of the test

Bertrand and Mullainathan (2004)
Applicants w/ white-sounding names got one callback for every 10 resumes sent
Applicants w/ black-sounding names got one callback for every 15 resumes sent
This 50% gap was statistically significant

Drawbacks of audit studies
1) don’t know if its taste discrimination or statistical discrimination
2) even if researchers find evidence that there are discriminating firms in the labor market, the wage gap could still be 0; wage gap depends on interaction of supply and demand
3) if we observe a wage gap in the data, audit studies cannot tell us if discrimination is driving the result. The wage gap could be driven by the different productive characteristics of black and white workers.

Equal Pay Act 1962
Overturned old laws about women’s total weekly hours of work, lifting heavy objects, etc.
Outlawed separate pay scales for men and women using similar skills and performing work under the same conditions
Said nothing about discrimination based on race, color, religion, etc.
Said nothing about banning discrimination in hiring or promotion
If labor market discrimination is to be eliminated, legislation must require BOTH equal pay and equal opportunities in hiring and promotions for people of comparable productivity

Civil Rights Act 1964
Made it unlawful for any employer to refuse to hire or to discharge any individual…because of race, color, sex, etc.
Disparate treatment standard of discrimination–occurs if individuals are treated differently bc of race, sex, color, etc. and if it can be shown that there was an intent to discriminate
Disparate impact standard of discrimination–it is the result, not the motivation that matters; policies that appear to be neutral but lead to different effect by race, gender, etc. are prohibited unless they can be related to job performance; violations include–word of mouth recruiting and statistical discrimination

Affirmative Action
Federal civil rights program strengthened in the 1960s by Executive Orders that prohibited discrimination by race and sex among government contractors
Compel federal contractors to 1) not discriminate and 2) take affirmative action to ensure that they do not
There is little operational difference b/t establishing employment “goals” and “quotas” which require that x percent of new workers belong to a particular group

Black-white male median earnings ratio
The relative wage of both black men and women is substantially higher today than it was in the late 1960s

Why has the male black-white wage gap narrowed in recent decades?
1) Increase in educational attainment of blacks–1940 gap was 3.9 yrs, 1980 gap was 1.4 yrs (20-25% of post-1960 gain attributable to this)
2) Increase in school quality for blacks–rate of return to schooling higher for black males vs. white males in the late 1970s (15-20%)
3) Large decrease in LFP of black (distribution no longer includes artificial increase in black wage) (10-20%)

1/3 of the improvement in the black/white earnings ratio for men remains to be explained
Discontinuous increase in B/W earnings ratio during same time that government policy was implemented

Heckman (1991)
Southern employers were eager to employ blacks if they were given the proper excuse
Not clear that the programs were successful after 1980, when the market for less-educated workers turned poor.
Once the most blatant forms of discrimination were attacked, the effects of federal efforts have weakened.

What are the most important factors that affect the size of the SE towards foreign labor?
1) SE greater if the supply of americans to relevant occupations is elastic
2) Size of SE depends on ease w/ which foreigners can be substituted for Americans; manufacturing jobs at risk and local service jobs much safer

Trefler (2004)
1989 Canada-U.S. Free Trade Agreement
Reduced Canadian tariffs on imports from the U.S.
Canadian employment fell by 12% in industries whose tariffs fell the most
Overall employment rate same in 2002 vs. 1988–displaced workers found new jobs rapidly

Bhagwati et al. (2004)
as of 2002, 300,000 americans could expect to lose jobs each year (25000/month) owing to offshoring
from 2001-2002, 1.6 million americans/month lost jobs and another 2.9 million voluntarily quit
jobs lost to offshoring only 1.5% of all jobs lost in a month

Blinder (2006)
Almost 30-40 million jobs at risk because of offshoring (1/4 of those currently employed)
Jobs in trouble–manufacturing, computer programming, financial services
Safe jobs–repair, maintenance, education, health care, leisure, retailing services

Policy Issues (1)
1) Removing the barriers to transactions across an international border can enhance aggregate consumption in both countries through greater specialisation and the exploitation of comparative advantage
2) The movement of resources w/in each country that is needed to adjust to greater specialisation imposes costs on workers whose jobs are displaced; generally the less-skilled w/in a country that are most at risk of losing from expanded trade

Policy Issues (2)
We can only conclude that a society as a whole is made definitively better off by a policy change if (a) everyone gains from it, (b) some gain and no one else loses, (c) some gain and some lose but the winner fully compensate the losers

Revenue Side
High-skill workers gain from international trade, so increase taxes to compensate losers from trade

Spending Side
1)Subsidising Human Capital Investments
government training programs not very effective
those who benefit are A) young (long time to recoup ROI) B) low psychic costs of learning C) relatively low discount rates (not too present-oriented)
for displaced, older workers the PV of future benefits might fall short of the costs of training
2) Income Support Programs–EITC (even though it doesn’t specifically target displaced workers from trade)
3) Subsidised Employment–offer target payroll subsidy for firms who hire workers displaced by trade or offshoring

Most importantly, the government should extend and improve schooling.

Borjas (1)
Moreover, catering to customer tastes does not reduce the firm’s profits.
58% of newly hired workers are black in contact firms where most customers are black.
9% of newly hired workers are black in contact firms where most customers are white.
Customer discrimination reduces the fraction of black among newly hired workers by 49% points.
The fraction of newly hired workers who are black falls 46.6% to 12.2% as the customer base shifts from being mainly black to mainly white, a reduction of 34.4% points–this estimates what might happen to black employment, even in the absence of customer discrimination, when a firm caters mainly to black customers, when a firm caters mainly to black customers, perhaps bc this shift requires that the firms open up shop in black neighbourhoods and hence attract many black applicants.
The difference in differences estimate of the impact of customer discrimination would then be given by 14.6%. In other words, face to face contact b/t blacks workers and white customers substantially lowers the probably that the firm hires black workers.

Borjas (2)
3% of all games played during a season have 0 white referees, 21% have one white referee, 47% have two, and 29% have an exclusively white referee team.
If all three are black, the avg. foul rate for white players is 5.25 and 4.42 for black players (0.83 difference in the foul rate).
If all three are white, the avg. foul rate for white players is 4.90 and 4.32 for black players (0.58 difference in the foul rate).
Difference i difference estimate of the impact of having an all white referee team for white players is -0.25 (all white referee team leads to fewer fouls for white players).
The number of fouls received by black players is essentially constant regardless of the racial composition of the refereeing team.
Nepotism–white referees go “easy” on white players
Nepotistic behaviour leads to about 4-5% fewer fouls for players that have the same race as the referees, and this translates into roughly 2-3% more points/game.

Borjas (3)
After applying for the job, the white applicant was 33% more likely to be interviewed and 52% more likely to receive a job offer.
8/10 job offers made by low-priced restaurants were made to women, whereas 11/13 job offers made by high-priced restaurants were made to men.

Borjas (4)
1920s–pupil/teacher ratios in southern states were 50% higher in black schools than in white schools. This had disappeared by the late 1950s.
As a result, the racial gap in the rate of return to school also vanished.
The rate of return for white workers who entered the labor market around 1940 was 9.8%, only 4.7% for black workers.
In the 1970s, blacks actually had a higher rate of return to schooling (9.6% vs. 8.5%)
The increasing quality and quantity of black schooling contributed to the narrowing of the black-white wage gap.
It has been estimated that at least half of the increase in the black-white wage ratio in recent decades can be attributed to the increase in black human capital.
Affirmative action programs have influenced employment decisions.
In 1966, black men were 10% less likely than white men to work in firms that were federal contractors. By 1980, there were 25% more likely to work in covered firms.
1910-1964 South Carolina–black employment in the textile industry stood at 4-5%, even though it sold 5% of its output to the U.S. government.
By 1970, nearly 20% of the workers in the industry were black.
No consensus on whether these programs have increased the black WAGE–black relative wage was rising even prior to the 1960s.
The number of blacks employed by large firms increased substantially in the 1970s, raising the average black wage.
It is estimated that the increasing representation of blacks in the workforce of large firms accounts for about 15% of the increase in the black-white wage ratio over the period.

Borjas (5)
Hiring quotas first imposed in 1973.
Late 1960s–Detroit entrance exam had a passing rate of 44% for blacks and 81% for whites
1970 exam in NY–STILL the passing rate was 55% for blacks and 82% for whites
1970s–use of these exams in police hiring was ruled discriminatory on the theory of disparate impact
1970 two sets of cities roughly similar–blacks made up 5-7% of the police force. By 2000, the black share in employment had risen to 23% in litigated cities but only to 13% in the unlitigated cities.
It has been estimated that the 25-yr gain in the black share due to the involvement of the federal judiciary was around 10-percentage point gain.
This change in the racial composition of the city’s police force was accomplished without any corresponding change in the city’s crime rate.