firms, as Audretsch and Fritsch (1994) points out, a measure of mean establishment size
(MEANSIZE) was used. The other variables included are both the proportion of firms in
which 10 to 24 persons engaged (SMFIRMS) and the proportion of firms in which more
than 200 persons engaged (LMFIRMS) in total manufacturing sector. These rates are
calculated for the number of persons engaged also (correspondingly, SMPEREN and
LMPEREN). All these variables corresponding to the 1985 figures are disregarded for
CHANMIC, except MEANSIZE.
8- Production structure. Although this paper deals with why change in firm population
differs within the country, the impact of sectoral concentration can be important with the
assumption that a high specialization index of industrial employment would result in high
rates of new firm formation like in The Third Italy (Garofoli 1994). To measure this
variable, the number of persons engaged in each manufacturing industry is divided into total
manufacturing employment. Then, they were squared, summed and the square root taken of
the sum (Reynolds 1994). However, this operation is repeated for number of firms also
(SPECFIR). While the former has a national average of 0.479, and changes from 0.770 to
0.392; the corresponding values for the latter are 0.469, 0.691 and 0.415.
9- Government investments. To determine the relationship between government investments
and firm growth, five variables are included. These are the total government investments
between 1985 and 1992 (GOVERN), the investment incentives between 1985 and 1992
(INCENTIVE), the number of workshops in small industrial estates (SIESHOP) and the
number of parcels in the organized industrial districts (PARCEL).
The matrix of correlation coefficients between independent and dependent variables is
shown in table 4. Although the correlation coefficients have similar figures for CHANTOT
and CHANMIC, CHANSM is rather different. While there is the highest positive
relationship between population growth rate and the former two dependent variables (around
0.42), it is achieved by the share of employment in small firms for CHANSM (0.46).
12