and gender. Afghan rules—as outlined initially—lean toward privileging
connections between voters and individual candidates, but try simultaneously
to guarantee minority representation and representation according to at least one
prominent form of collective identity: gender. Reconciling individualistic and
collective representation in a set of workable electoral rules has proven diYcult in
the Afghan context.
With respect to legislative representation more generally, these cases suggest that
the individualistic vs. collective representation distinction may be as important
as the principle characteristic by which electoral systems are more frequently
distinguished—whether elections are winner-take-all in single-member districts
(SMD), or proportional (PR). The characteristics and relative merits of SMD vs.
PR are central to a long-standing literature on legislative elections, the predominant
conclusion from which has been that PR is normatively superior to SMD elections
(Sartori 1976 ; Lijphart 1994 ; Huber and Powell 1994 ; Colomer 2001 ). This conclusion
rests on some key assumptions however: that political parties are fundamental units
of legislative representation, and that a left–right spectrum meaningfully describes
the ideological arena of party competition. In the industrialized, long-
standing democracies, where most studies of legislative representation have been
conducted, there is solid empirical evidence for these assumptions (Powell and
Vanberg 2000 ). They are open to greater skepticism in other environments, however,
particularly where party systems are more volatile or party reputations less stable.
The point here is that the foundation on which the conventional SMD vs. PR
debate has been conducted is weak in many political environments where the most
critical choices about how to organize legislative representation remain open. The
complete absence of established party systems in the Iraqi and Afghan cases are
extreme examples, but it is worth noting that SMD versus PR was not central to
debate in either context; winner-take-all rules gained traction in neither case.
Rather, the critical distinction in these cases is over whether electoral rules ought
to prioritize collective vs. individualistic representation. This theme has been
central to debates over reforming legislative representation much more widely
during recent decades, particularly with respect to mixed-member electoral
systems that combine SMD with list PR elections within the same legislative
chamber, variants of which were adopted in the 1990 s by over twenty countries
(Shugart and Wattenberg 2001 ; International Institute for Democracy and Electoral
Assistance 1997 ; Culver and FerruWno 2000 ; Carey 2003 ).
To sum up, legislatures oVer the promise of representing the diversity of
the polity, but electoral rules aVect the dimensions along which diversity can
be translated into representation. Although the diVerences between SMD and
PR elections have traditionally been essential to the study of comparative legis-
latures, this distinction is growing less central relative to that between individu-
alistic and collective representation, which is quite a diVerent matter, both
theoretically and empirically (Carey and Shugart 1995 ). Whereas the literature
legislative organization 437