In studying the movement in the production workers' share in value added (i.e., labor's share) in manufacturing industries, the following regression results were obtained based on the U.S. data for the years 1949 to 1964 (t ratios in parentheses): Model A: Ŷt = 0.4529 - 0.0041t; r2 = 0.5284; d = 0.8252
t = (- 3.9608)
Model B: Ŷt = 0.4786 - 0.00127t + 0.0005t2; R2 = 0.6629; d = 1.82
t = (- 3.2724) (2.7777)
where Y = labor's share and t = the time.
a. Is there serial correlation in Model A? In Model B?
b. If there is serial correlation in Model A but not in Model B, what accounts for the serial correlation in the former?
c. What does this example tell us about the usefulness of the d statistic in detecting autocorrelation?