“…The results of the Wald test to evaluate the null hypothesis of k 1 = k 2 = _ = k 12 , yields an F value of 27.722. The null .023540 h 6 .648036 h 7 .886481 h 8 .225282 h 9 .902247 h 10 .521423 h 11 .220639 h 12 .636066 k .000001 f .000000 .000290 # f 0 .000007 .044617 à b 1,8 .044893 à b 1,14 .046623 # b 1,15 .045322 # b 7, 16 -.038984 # b 9, 16 -.038324 # b 7, 18 -.027254 # b 9, 18 -.025562 # b 7, 19 -.024061 à b 9, 19 -.024353 à c 7,1 -.013037 à c 1,2 .025986 à c 11,2 -.013052 à c 7,3 -.002352 à c 1,5 .024671 # g 1 0.078578 g 2 0.089664 g 3 0.072775 g 4 0.081545 g 5 -0.055284 g 6 0.110166 g 7 0.218938 g 9 0.176231 g 10 0.146975 g 11 -0.033631 g 12 0.101146 hypothesis is rejected, implying that the time trends do vary across the panels. Note however that while the coefficients are statistically significant and do vary across panels, the coefficient values themselves are very small.…”