Regression modelling of the employment durations of UK seafaring officers in 1999

This paper presents the results of an econometric analysis of the duration of voyages undertaken by NUMAST deck and engineer officers in 1998-1999. Earlier papers examined the employment profile of the officers, and compared voyage patterns across organization type, company nationality, flag of registry, and ship type. Analysis of the various cross-tabulations of employment duration appeared to show that the key driver is ship type. Using regression analysis, it is shown that ship type does indeed have a significant influence on duration, as does ship size. Flag, organizational type and company type were all found to be statistically unimportant.