Level III was based on a much broader sample of ships (1,487), but was
rather drastic in the simplifying assumptions it made in order to work with this sample.
In fact, in Level III maiming costs and capital costs were the only categories of cost
that were taken into account, and these costs were estimated rather than actual. Then
comparisons were made among (a) each of the parent ships selected, (b) an ATOMOS-type
ship, and (c) a cheap-crew ship, all appropriately defined. Details follow.
The total number of ships in the world ship database at our disposal
(supplied by Lloyds Register) was 25,058. From these ships we selected those that flew (a)
the 12 flags of all EU countries, as defined earlier, that is: Belgium, Denmark (including
DIS), France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain,
and United Kingdom; or (b) the flags of Finland, Norway (including NIS), Sweden, Japan,
and the US.
We next discarded all ships for which the Lloyds ship database had no
information on crew size or horsepower, and the ones which had crew size less than 11. The
remaining ships after all these screens were 1,487, Each of these 1,487 ships was labeled
a "parent ship". For each parent ship, two equivalent (identical in terms of
capacity and speed) hypothetical ships were considered:
a) An "ATOMOS-10" ship, manned by 6 officers and 4 ratings.
It flies the parent ship flag, and is manned by flag nationals. Such a configuration
resembles closely the one used in the Lauritzen reefers.
b) A "cheap-crew" conventional ship. This cheap-crew ship
flies also the flag of the parent ship. It differs from the parent in that only the
captain and the first officer are flag nationals, the rest of the crew being (non-EU)
nationals paid a very low salary. The reason for examining this alternative (which may be
unrealistic or even illegal in some countries under current national legislation- the most
notable example being the US) was to assess it as an alternative to an ATOMOS-type ship.
Such an alternative is under discussion in several countries as a means to control manning
costs, and so we thought it would make sense to see how it would compare with an ATOMOS
alternative.
In Level III maiming costs for each of the crew ranks and nationalities
assumed were based on information on collective labor agreements in various countries,
multiplied by a user-defined "surplus factor" (whose default value is assumed to
be 1.3 to account for the fact that actual wages are typically higher than those