Carsten Anckar, "Size and Party System Fragmentation," Party Politics, 6 (July 2000), 305-328.

First Paragraph:
In their seminal work Size and Democracy, Robert Dahl and Edward Tufte (1973) discussed, among other things, the link between size and party systems, suggesting that the size of political entities affected the degree of fragmentation of the party system. The explanation for this assumed relationship was twofold. One argument was derived from social-psychological theory-building, more specifically from the well-known experiments on conformity conducted by S. E. Asch. The other reason for the expected link between size and party system fragmentation is about diversity. Large size, the authors argue, leads by necessity to organizational diversity and complexity. A large unit presupposes a wide range of organizations and institutions, which produce a high degree of specialization and complexity. Therefore it is reasonable to expect that an increase in size leads to an increase in attitudinal diversity (DahI and Tufte, 1973: 30-40). From this discussion of the arguments that DahI and Tufte lay down, two hypotheses emerge, which will be tested in this study:

Figures and Tables:
Table 1: Strength of association between size, socio-economic differentiation, number of parties and party system fragmentation in Finnish local elections, 1992 and 1996 (multiple regression)
Table 2: Strength of association between size, number of parties and party system fragmentation in British local elections, 1988-1992 (sample of 400 shires) (multiple regression)
Table 3: Strength of association between size, effective threshold, presidentialism, urbanization, ethnic-religious fragmentation, number of parties and party system fragmentation in 77 countries with free
party systems (multiple regression)
Table 4: Strength of association between size, effective threshold, presidentialism, urbanization, ethnic-religious fragmentation, number of parties and party system fragmentation in countries with free party systems and plurality electoral system (multiple regression)
Table 5: Strength of association between size, effective threshold, presidentialism, urbanization, ethnic-religious fragmentation, number of parties and party system fragmentation in countries with free party systems and proportional electoral systems (multiple regression)
Table A1: Size, degree of urbanization and level of ethnic-religious fragmentation in countries with free party systems
Table A2: Electoral system, effective threshold, form of government, number of parties, electoral support for the leading party and effective number of parties in countries with free party systems

Last Paragraph:
Thus, the contribution of this present study to empirical theory-building can be expressed in the following statements.
1 -- The larger the size of a country, or a unit within a country, the higher the number of parties. The rule applies irrespective of electoral system.
2 -- The larger the size of a country, the lower the electoral support for the leading party. The rule applies irrespective of electoral system.
3 -- The larger the size of a country, the higher the effective number of parties. The rule applies irrespective of electoral system.