Use the data in OKUN to answer this question; see also Computer Exercise C11 in Chapter 11.

(i) Estimate the equation pcrgdpt 5 b0 1 b1cunemt 1 ut and test the errors for AR(1) serial correlation, without assuming 5cunemt

: t 5 1, 2, p6 is strictly exogenous. What do you conclude?

(ii) Regress the squared residuals, u^

2 t , on cunemt

(this is the Breusch-Pagan test for heteroskedasticity in the simple regression case). What do you conclude?

(iii) Obtain the heteroskedasticity-robust standard error for the OLS estimate b^

1. Is it substantially different from the usual OLS standard error?