1. Associated with the name of Simon Kuznets is the idea that the relationship between GDP and inequality is nonlinear. Kuznets hypothesized three phases in economic development. In the first stage of development, incomes and inequality are both low. During the process of modernization and industrialization, income and inequality both rise. Once a country has become a modern, industrialized economy, inequality declines as income continues to rise. (As you might suspect, the last 30 years of U.S. economic experience have not provided much evidence to support the Kuznets conjecture!)

Which of the following equations would support the Kuznets hypothesis? The dependent variable is the Gini coefficient, a measure of inequality that ranges from 0 to 1, with 0=perfect equality (everyone has the same income) and 1=perfect inequality (one person has all the income). PCINC is "per capital income" (measured in thousands of dollars).

a. GINI = -0.04 + 0.120 PCINC - 0.004 PCINC 2 Adj-R2 = 0.86
b. GINI = -0.04 + 0.004 PCINC - 0.120 PCINC 2 Adj-R2 = 0.84
c. GINI = -0.04 + 0.120 PCINC + 0.004 PCINC 2 Adj-R2 = 0.83
d. GINI = -0.04 - 0.120 PCINC + 0.004 PCINC 2 Adj-R2 = 0.92

2. You run two regressions on the same data set:

Y = 0.56 + 2.3 X1 + 40.3 X2 + 0.46 X3 + 1.1 X4, Adj-R2=0.89
t=2.0 t=6.2 t=4.1 t=2.6

Y = 1.12 + 42.3 X2 + 0.48 X3 + 0.2 X4, Adj-R2=0.88
t=5.9 t=3.5 t=3.5

a. Which explanatory appears to be correlated with X1? ______________
b. If the estimated coefficients in the first regression happen to be the true betas (call it beginner's luck), what is your best guess about the size of omitted variable bias in the second regression?
i. -0.9
ii. 1.1
iii. 2.0
iv. 0.2
v. None of the above.
c. What is your best guess about the value of the coefficient you would get if you were to regress X1 on the explanator you named in 5.a?
i. 0.39 ii. -0.39 iii. -0.9 iv. 0.9 v. None of the above.