 ppendix A: Deriving state market shares
We derive market shares for ballasts using the approximations
detailed in Table A-1. We distinguish between new additions to the commercial
stock and existing commercial stock. The top part of Table A-1 calculates
the average ballast market share for new and existing buildings over the
1985 to 1993 period. It uses existing stock floor area from Koomey et al.
(1995 and US DOE (1995) with an eight year lifetime (the average from Table 4) to estimate the amount of the floor area
retrofitted each year. It uses new floor area additions from Census (1994).
These two numbers are calculated each year from 1985 to 1993 as a fraction
of total, and then the simple average is calculated for use in the bottom
of the table.
The available state data was the value of construction
contracts in 1993 (which represents new construction) and population (which
is a reasonable proxy for existing stock). Both are taken from Census (1994).
The distributions by state are comparable. We weight the two distributions
by the fractions derived in the top of the table to get a weighted average
share of ballast sales by state. We then correct those shares to reflect
the fact that California had a 20 percent saturation of efficient ballasts
in 1980 (CEC 1982) and that the saturation for the US as a whole was 12.5%.
This calculation implies that all states but California had (on average)
an 11.5% share of efficient magnetic ballasts. | 
