ized by an attempt to equalize the distribution of wealth, the leftist end of the continuum consists of the most severe attempts at transferring a nation's wealth vertically, from the rich to the poor. In many states characterized by an unequal distribution of wealth, a party's position toward government activity in redistributing wealth is a good measure of the party's attitude toward equality of distribution. A position that favors no government activity at all, however, does not anchor the rightist end of the continuum, because it is at least theoretically possible for a party to favor redistribution from the poor to the rich.

Operational Definition. The pro and con positions are as follows:

PRO-strong +5	Advocates severe vertical redistribution from rich to poor; suggestions for major land reforms and equalization of all incomes should be included in this category; demands that redistribution be immediate. Combined seizure and redistribution by government included here.
PRO-moderate +3	Advocates a major redistribution of wealth which is not as severe as abolition of personal property or major land reform, but is more severe a vertical redistribution than an income or property tax. A redistribution plan over time would be included in this category, as would government seizure but failure to redistribute.
PRO-weak +1	Advocates graduated income tax and/or graduated personal property tax.
NEUTRAL 0	Includes ambiguous or contradictory positions.
CON-weak -1	Opposes the enactment of new legislation, such as higher progressive taxation rates, that would erode the differences between rich and poor but accepts legislation now in effect.
CON-moderate -3	Accepts government inactivity in wealth redistribution in a country that features extreme inequalities in wealth.
CON-strong -5	Urges new policies that would enhance the income acquiring capacity of the wealthy at the expense of the poor.

Coding Results. According to the way we operationalized "redistribution of wealth," the parties of the world moved toward more redistributive policies from the early to the late 1950s, as shown by the higher mean values in Tables 6.4a and 6.4b for BV503. This shift runs counter to the trend noticed for the first two issue orientation variables. There is no significant correlation between BV503 and AC503, but note that about 5 percent fewer parties were scored on this variable than the previous two.

In looking at the United States, Canada, and Australia, Birth (1955) argues that welfare measures may be divided into two classes: (1) employment and health serv-

go to page 59