PUBPOL 5072  Practice Midterm Exam

None of these questions are essay questions. Answers should be very succinct; all can be answered in 4 sentences. Feel free to use bullet points and pictures.

1. (10 points) True or false? Students from high-income families have higher attendance rates at Ivy-Plus universities primarily because they apply to these schools at higher rates, compared to students from lower-income families with the same test scores. Explain.

2. (10 points) True or false? Nearly all differences in the economic connectedness (EC) of individuals across schools can be accounted for by differences in exposure to high-SES children at the same school. Explain.

3. In this class, we have used at least four different statistics to describe intergenerational mobility.

(a) (10 points) Define two of the measures we have used in this class.

(b) (10 points) For both of the statistics you provided in (a), what would each equal in a world where parents’ income is totally irrelevant for their children’s income as adults (i.e., a world of perfect mobility and no intergenerational transmission of income, ability, etc.)?

4. Below are questions on inequality and intergenerational mobility.

(a) (8 points) What is the difference between economic inequality and intergenerational mobility?

(b) (8 points) Do countries with more economic inequality tend to have higher or lower levels of relative intergenerational mobility, on average?

(c) (8 points) Recall the following definitions:

• Absolute mobility: the share of children who earn more than their parents in real terms.

• Relative mobility: how children’s rank in the income distribution compares to their parents’ rank.

(d) (8 points) Give one reason we might expect a causal relationship between economic inequality and relative mobility.

(e) (8 points) Give one explanation that would generate the negative relationship in part (b) without a causal effect of inequality on mobility.

5. Consider the following graph, which displays estimates of the causal effect of receiving various CMTO moving supports on a recipient’s likelihood of using the housing voucher in a high-opportunity neighborhood.

1. Treatment 1 (T1): Financial assistance only.

2. Treatment 2 (T2): Limited rental search support, limited landlord services, and limited financial assistance.

3. Treatment 3 (T3): Full program supports (rental search support, landlord services, and financial assistance).

(a) (15 points) Which of the three treatments had an effect on moving to opportunitythat is statistically significantly different from zero? Explain.

(b) (15 points) Describe the key takeaways from this graph to a mayor interested in implementing CMTO in their city who is smart but not trained in statistics. Avoid technical or statistical jargon.

(c) A critic raises the following concern about Phase 2 of the CMTO experiment (whose results are shown above):

• Seattle had previously run Phase 1 of the CMTO experiment before starting Phase 2.

• As a result, potential voucher recipients in Phase 2 may have already been more aware of high-opportunity neighborhoods and therefore more likely to move there, even without program supports.

(i) (10 points) If this concern were valid, how would it affect your interpretation of the estimates? Explain.

(ii) (10 points) Propose a specific way to check whether participants actually differed in this way, using the numbers in the graph above and comparable numbers from the original experiment. (You do not need to actually conduct this test.)

6. Suppose the federal government is considering funding an intensive tutoring program at Lansing High School for students from households with parent income below $60,000. Use the information below to compute the marginal value of public funds (MVPF) for this program.

• Eligible students receive free access to after-school academic support throughout all four years of high school.

• Total cost of providing the program to one student for four years: $8,000.

• Each graduating class has 40 eligible students.

• Before the government program: 12 of the 40 eligible students were already paying to attend out of pocket.

• After the government begins covering the cost: 20 of the 40 attend.

• A new paper (RDD) estimates that completing the program increases adult earnings by $30,000 in PDV of lifetime pre-tax income.

• Students face a 25% tax rate, so the PDV increase in post-tax lifetime income is $22,500.

Using this information, answer the following questions.

(a) (10 points) What goes in the numerator and denominator of the MVPF? Why is ituseful?

(b) (5 points) Does the MVPF tell you what the government “should” do?

(c) (5 points) What about cases where the net cost to the government is negative?

(d) (5 points) Among eligible students, what is the effect on tutoring program partici-pation in percentage points?

(e) (5 points) What is the percent increase in tutoring program participation among eligible students?

(f) (5 points) What is the total upfront cost of the policy per graduating class? (“Up front” = total tutoring dollars paid out by the government.)

(g) (10 points) What is the net cost to the government in the long run? Compute: upfront cost − increase in future tax revenue from induced participation.

(h) (10 points) What is the beneficiaries’ willingness to pay (WTP) for the policy? Split into two groups:

• Already attending: The 12 students who were already paying value the pro gram at its dollar cost ($8,000 each).

• Induced to attend: The 8 students induced by the free program value it at the increase in post-tax lifetime income ($22,500 each).

(i) (5 points) Compute the MVPF of this policy. MVPF = WTP / Net government cost.

(j) (5 points) Interpret the MVPF.

(k) (5 points) Do policies targeted at children or adults tend to have larger MVPFs? Why?