The first issue considered here is the extent to which a detailed statistical analysis supports the proposition in chapter 6 that the pattern of regulation and ownership had little effect on the level of infrastructure development. The second issue is what factors determined the pattern of regulation and ownership. I analyse the cases of three sectors, railways, telephones and gas, as they had developed by the early 1900s. The data are good for these services and a cross-section comparison is made of eight European countries: Britain, Denmark, France, Germany, Sweden, Italy, Norway and Spain. The main sources are given in Tables 2.1, 2.2, 3.1, 4.1 and 7.1.

The level of infrastructure development, such as rail track per head of population, may be affected by the level of GDP per head (GDPPH in Table A.1 below) and by energy sources (COALPH, coal output per head), as well as the proportion of mileage in public ownership (OWN). Countries with high population density (POPDEN) may need well-developed infrastructures. Finally, when military matters loom large in a nation's thinking, a well-developed infrastructure may be seen as an important strategic weapon. Here I measure the defensive element by military expenditures as a percentage of all central government expenditures (MILB). The measures of the dependent variable, infrastructure development, are gas supplies, rail track and telephone numbers all expressed per head of population. These are shown as PENLEV in column 2 of Table A.1.