In a 1988 article, Josef Brada and Ronald Graves built an interesting model of defense spending in the Sovietic Union just before the breakup of that nation. The authors felt sure that Soviet defense was a function of U.S. spending and Soviet GNP but were less sure about whether defense spending also a function of the ratio of Soviet nuclear warheads to U.S. nuclear warheads (NR/NU) was. Using a double-log functional form, the authors estimated a few alternative specifications one of there is (file Defend):

1. lnSDHi=0+1USDi+2lnSYi+3lnSPi+i

Where: SDH_i= the CIA's "high" estimate of Soviet defense expenditures in yeari

USD_i= U.S. defense expenditures in yeari

SY_i= Soviet GDP in yeari

SP_i= NR_i/NU_i.

1. Estimates the regression equation
2. Test the equation for positive first-order serial correlation (You must include all the steps of the hypothesis test, as well as the operations you used in the test statistic)

obs NR NU SDH SDL SFP SY USD 1960 415 1734 31 23 7.03 232.3 200.54 1961 445 1846 34 26 6.07 245.3 204.12 1962 485 1942 38 29 3.9 254.5 207.72 1963 531 2070 39 31 2.97 251.7 206.98 1964 580 2910 42 34 1.4 279.4 207.41 1965 598 4110 43 35 1.87 296.8 185.42 1966 674 4198 44 36 4.1 311.9 203.19 1967 1058 4338 47 39 4.9 326.3 241.27 1968 1270 4134 50 42 4.07 346 260.91 1969 1662 4026 52 43 2.87 355.9 254.62 1970 2047 5074 53 44 4.43 383.3 228. 19 1971 3199 6282 54 45 3.77 398.2 203.8 1972 2298 7100 56 46 2.87 405.7 189.41 1973 2430 8164 58 48 3.87 435.2 169.27 1974 2534 8522 62 51 4.3 452.2 156.81 1975 2614 9170 65 53 6.33 459.8 155.59 1976 3219 9518 69 56 0.63 481.8 169.91 1977 4345 9806 70 56 2.23 497.4 170.94 1978 5097 9950 72 57 1.03 514.2 154.12 1979 6336 9945 75 59 0.17 516.1 156.8 1980 7451 9668 79 62 0.27 524.7 160.67 1981 7793 9628 83 63 0.47 536.1 169.55 1982 8031 10124 84 64 0.07 547 185.31 1983 8730 10201 88 66 1.5 567.5 201.83 1984 9146 10630 90 67 1.63 578.9 211.35