In his 1993 paper, Richard Matland argues that party magnitude — the district magnitude divided by the number of parties — is causally closer to levels of political representation than the underlying district magnitude. The concept of party magnitude is used from time to time in the literature, but usually little attention is paid as to how it works.
In addition to the effects of district magnitude, assuming that men are more likely to take the top spot of party lists than women are, where there are fewer parties competing, the likelihood that a woman is elected is increased by reaching further down the party list. Combining the two effects, it is also apparent why, as Matland acknowledges, the effects of party magnitude are temporarily limited. The association is weak where the proportion of women in parliament is low. In this case the likelihood of that a woman is elected is low in any case. As the proportion of women in parliament increases, so does the association. Once women are as common as candidates as men are, and they are equally likely to appear at the top of party lists, the association once again decreases. The likelihood that a woman is elected in this case approaches 50%.
Matland, R. (1993). Institutional variables affecting female representation in national legislatures: The case of Norway. Journal of Politics, 55, 737-55.