Table 1. — New England trawler production functions: alternate specifications. 



Dependent variable 



INDhPENDENT VARIABLE 



LOG DAYS LOG DAYS LOG LOG LOG LOG CONSTRUC- DUM DL M 

 ABSENT FISHED GRT' HP^ CREW AGE TION^ 65" kT^ 



INT ^ 



Problem I 



Log total value (All years) 

 Reg. Coef. 

 f ratio 

 Part. Cor. Coef. ' 



1.080 



47.600 



.815 



.409 .038 -.410 -.240 



6.340 .525 5.160 4.540 



.184 .016 -.151 -133 



.429 .002 



7.580 .037 



.219 .001 



.365 .113 



10.800 2.980 



.305 .088 



-2.66 

 4.040 

 -.119 



-.207 

 4.470 

 -.131 



-.002 -.107 

 .062 3.860 

 -.001 -.113 



.373 .074 .347 -.129 



11.000 1.940 8.830 4.660 



.309 .058 .253 -.136 



-.024 

 .752 

 -.022 



-.043 

 2.280 

 -.067 



.095 



5.000 



.146 



4.69 .405 98.70 



.011 -.059 

 .533 2.750 

 .015 -.081 



3.39 .542 170.28 



-.024 

 1.920 

 -.057 



.0006 2.43 



.0500 



.0010 



.023 .010 



1.790 .855 



.053 .025 



,834 724.34 



1.44 .833 718.97 



'Gross registered tonnage. 



^Horsepower. 



^Construction; equals one if wood, zero otherwise. 



''Dummy variables for year of observation. 



'Partial correlation coefficient. 



may be less productive: (1) Older vessels might 

 tend to have more breakdowns and equipment 

 that was not in the best working order; (2) older 

 vessels might have poorer working conditions 

 and accommodations and, therefore, attract less 

 able crews; (3) older vessels may embody older 

 technologies. If the last hypothesis is dominant, 

 vessels do not become less productive as they 

 get older, rather old vessels are less productive. 

 This would have different implications than 

 the first hypothesis when fishing power factors 

 are computed. 



The dummy variable created for hull con- 

 struction took on the value 1 if the hull was 

 wood and if steel. The results using this vari- 

 able were mixed. In Problem 4, using total 

 value and days absent, it was positive and 

 significant. This may mean that ceteris paribus 

 wooden hulls are 25% more productive.^ There 

 is no theoretical reason why these results 



2 The antilog of 1 is 10. We have 10."»'* which equals 

 1.2.5. Therefore, a wooden hull is 2.59^ more productive. 



should be obtained. The data in Appendix Table 

 3 show that the large vessels in the fleet are 

 steel and the small ones wood, with a very small 

 overlap. We may be observing an upward 

 adjustment for the wood vessels because they 

 fish many fewer days during the most productive 

 portion of the year. 



The tests for locational differences in produc- 

 tivity were made by creating an array of six 

 dummy variables, one for each of the major 

 ports in New England. A "one" was placed in 

 proper location in the array corresponding to a 

 vessel's home port and a "zero" in all the 

 others. Equations showing the results of these 

 tests are given in Appendix Table 1. In the 

 logarithmic forms of the equations, there are 

 no consistent differences between ports when 

 total value is the dependent variable, the ports 

 designated "Maine" appear to catch significantly 

 more and "Boston" significantly less (Problem 

 10). These differences appear because. Maine 

 specializes in low value species and Boston in 



48 



