Prof. Bryan Caplan

bcaplan@gmu.edu

http://www.gmu.edu/departments/economics/bcaplan

Econ 321

Fall, 2000

Answer problems #1, #2, #12, AND your choice of 7 other questions.

I.                     Provide an explanation other than pure discrimination that could account for the following group differences.  (You do not need to agree with your explanation; just try to make it plausible).

A.                Blacks with college degrees earn less than whites with college degrees.

Probably the most empirically important is differences in measured intelligence, which persist controlling for education.

B.                Married men without children earn more than single men.

Married men are more conscientious - they are both more able to commit to their job and to another person.

C.                People with criminal records earn less than other people with identical education and intelligence.

They have less (normal) job market experience.

D.                There are so few Asians in professional basketball.

One possibility: group differences in average height.

E.                Japanese-Americans earn more than white Americans.

Their average education level is higher.

F.                 Teachers earn less than equally-educated non-teachers.

Teachers on average come from the bottom half of their graduating class.

II.                   In chapter 16 of Landsburg's Fair Play, he works through two arithmetic examples on discrimination, and shows that a non-discriminating employers can increase stockholders' returns by 150%.  Using the numbers from his first example, recalculate the percentage increase in stockholders' returns if:

A.                (Sample problem)  All of Landsburg's numbers stay the same, except that blacks earn 90% as much as equally productive whites.

 9 white workers @ \$100,000  \$900,000 1 black worker @ \$90,000     \$90,000 Bondholders               \$240,000 Stockholders              \$210,000   Total                    \$1,440,000 10 black workers @ \$90,000 \$900,000 Bondholders                \$240,000 Stockholders               \$300,000   Total                      \$1,440,000

The profits from hiring an all-black workforce are thus (\$300,000/\$210,000)=1.43 as large as those of a regular workforce; switching to an all-black workforce raises stockholder returns by 43%.

B.                All of Landsburg's numbers stay the same, except that bondholders and stockholders each receive half as much as the workers.

 9 white workers @ \$100,000  \$900,000 1 black worker @ \$60,000     \$60,000 Bondholders               \$480,000 Stockholders              \$480,000   Total                    \$1,920,000 10 black workers @ \$60,000 \$600,000 Bondholders                \$480,000 Stockholders               \$840,000   Total                      \$1,920,000

The profits from hiring an all-black workforce are thus (\$840,000/\$480,000)=1.75 as large as those of a regular workforce; switching to an all-black workforce raises stockholder returns by 75%.

(I would also accept an answer that scaled down all of the numbers to keep the total fixed at \$1,440,000).

C.                All of Landsburg's numbers stay the same, except that bondholders receive 40% as much as workers, and stockholders receive 10% as much.

 9 white workers @ \$100,000  \$900,000 1 black worker @ \$60,000     \$60,000 Bondholders               \$384,000 Stockholders              \$96,000   Total                    \$1,440,000 10 black workers @ \$60,000 \$600,000 Bondholders                \$384,000 Stockholders               \$456,000   Total                      \$1,440,000

The profits from hiring an all-black workforce are thus (\$456,000/\$96,000)=4.75 as large as those of a regular workforce; switching to an all-black workforce raises stockholder returns by 375%.

D.                All of Landsburg's numbers stay the same, except that the firm initially employs 9.9 white workers and only .1 black workers.

 9.9 white workers @ \$100,000  \$990,000 .1 black worker @ \$60,000     \$6,000 Bondholders               \$240,000 Stockholders              \$204,000   Total                    \$1,440,000 10 black workers @ \$60,000 \$600,000 Bondholders                \$240,000 Stockholders               \$600,000   Total                      \$1,440,000

The profits from hiring an all-black workforce are thus (\$600,000/\$204,000)=2.94 as large as those of a regular workforce; switching to an all-black workforce raises stockholder returns by 194%.

III.                  Briefly (1-2 sentences each) explain what the preceding four examples show about the costliness of discrimination.

A.  The lower the level of discrimination, the less costly it is to discriminate.

B.  The higher the share of capital to labor income, the less costly it is to discriminate.

C.  The higher a firm's debt to equity ratio, the more costly it is to discriminate.

D.  The lower the discriminated group's current share of the workforce, the more costly it is to discriminate.

IV.               Discrimination is less costly when wages exceed market-clearing levels.  Is there any incentive for employers to avoid discrimination when wages are too high? (2-3 sentences)

Yes.  As long as employees differ in QUALITY, it still pays employers to try to hire the most competent ones.  They can't get the same workers for less money, but they can get higher quality for the same amount of money.

V.                 Sowell (Race and Culture) discusses segregation of Jewish and Gentile workplaces.

A.                What efficiency rationale for this form of segregation does Sowell offer?

Sowell notes that Jews and Gentiles wanted different days off because they worshipped on different days of the week.

B.                What happened to this labor market in fields with small Jewish populations?

In these cases, Jews had to adapt to the workweek of the majority, or else limit their occupational choice.

VI.

A.                Briefly, carefully, and neutrally state one stereotype that you are familiar with (but not necessarily in agreement with).

One stereotype that I am familiar with is that males are more logical and less emotional than females.

B.                Would you expect this stereotype to be "self-fulfilling"?  Why or why not?

I would not expect this to be self-fulfilling.  Instead I would expect people unhappy with their gender stereotypes to try to overcome them by being, for example, the man who "isn't afraid to cry," or the woman who "only cares about the bottom line."

C.                Search the world-wide web or consult another reference source to check the accuracy of your stereotype.  How close to the truth was your stereotype?

In the book Gifts Differing, by Myers and Briggs Myers, they provide extensive data showing large personality differences between men and women on the Thinking-Feeling dimension.  If anything I underestimated how large the difference is: empirically, men are 60% T, 40% F, whereas women are 30% T, 70% F.

VII.              Suppose male and female employers are equally productive when employed, being worth \$15/hr, and work 2000 hours/year.  But 10% of female employees (and 0% of males) under the age of 30 plan on leaving work to have children, costing their employers \$15,000 in re-training expenses.

A.                Diagram the market for male and female labor if employers statistically discriminate.  How does the equilibrium female wage compare to the equilibrium male wage?

Males earn \$15/hr, or \$30,000/year.  Women earn (\$15*2000-.1*\$15000)=\$28,500/year, or \$14.25/hr.

B.                How much are women who don't leave work underpaid relative to their productivity?  How much are women who do leave work overpaid relative to their productivity?

Women who don't leave at underpaid by \$.75/hr, since they are worth \$15/hr but only earn \$14.25.  Women who do leave are only worth \$15,000 per year, or \$7.50/hr; they are thus overpaid by \$6.75/hr.

C.                What would happen to the male-female average wage gap if female employees could give their employer a "money-back guarantee"?

The AVERAGE gap would not change.  Underpaid women would get a \$.75/hr raise; overpaid women would get a \$6.75 pay cut.  The raises and cuts exactly balance.

VIII.            Suppose that workers who speak fluent English are more productive than workers who don't speak fluent English.  Using S&D curves for both types of labor, show the effect of banning wage differences as "discriminatory."

This creates a surplus of non-fluent labor (or a shortage of fluent labor).

IX.               Thinking in terms of the "market for mates," try to explain why men generally marry later than women. (2-3 sentences)

One idea: men's mate value peaks later in life, so if they were to marry early, they would often have to settle for a "low-quality" spouse.  In contrast, women who marry earlier usually get higher-quality spouses; the longer they wait, the lower their own mate value, and the lower quality spouse they have to settle for.

X.                 Discuss the impact of technological change since 1900 on BOTH supply AND demand for female labor. (2-3 sentences)

Technological progress in household appliances and the like have enabled women to care for a home and children using far less time than in 1900.  At the same time, fewer and fewer jobs call for physical strength, so one of the main historical advantages of men over women in terms of marginal productivity matters less and less.

XI.               Give an example of statistical discrimination in the dating market and discuss its effects. (2-3 sentences)

Many teenage girls won't date a guy who doesn't own a car.  This could easily be seen as statistical discrimination - guys with cars are, on average, richer, more successful, and more independent from their parents.  (Of course, let's not overlook the direct advantage - they are more able to take a date to various fun places).  Guys perceived as "losers" might then work extra hard to get their own car in order to overcome this perception.

XII.              Find an opinion piece in a newspaper or magazine that relies on either "the standard history of discrimination" or "the standard history of gender."  What aspects of the opinion piece would Caplan agree with?  What precisely would he disagree with? (1 paragraph)

At http://www.eeoc.gov/35th/pre1965/index.html, the Equal Employment Opportunity Commission discusses its historic role:

EEOC has worked tirelessly to eliminate discrimination from America's workplaces since its creation. The hard work, idealism, and commitment of EEOC employees has been instrumental in widening the doors of employment opportunity for all Americans and helping to create a standard of living for this nation's diverse citizenry that is the envy of the world. But challenges still abound. In far too many workplaces, old ways die hard. Discrimination, while often boldly evident, persists now in subtle forms as well. The need for an agency like EEOC is as evident today as before 1964. This is the unfortunate but true state of affairs as the nation enters the new Millennium.

There is very little for Caplan to agree with here.  He would certainly doubt that discrimination (as distinct from difference) is "boldly evident," and would probably expect the "subtle forms of discrimination" to actually reflect "subtle differences" between groups.  And even if discrimination were an appreciable force, it should at least be mentioned that what there are two possible errors: one is missing genuine discrimination, the other is mis-interpreting merit-based hiring as somehow discriminatory.  "Old ways die hard" because group differences are - for diverse reasons - frequently quite stable over time.  To Caplan, the need for the EEOC is about as "evident" as the need for a law barring employers from burning their own money.