17
17
designers openly acknowledge (Jaggers and Gurr 1995: 471)—is that Polity pays no
heed whatsoever to the incidence of political freedoms.
Polity’s strength is that displays its components in disaggregated fashion, country by
country, so that the material can then be re-processed using other methods of
aggregation. It also shows how the coding (with different units of scale for each
component) is carried out. As a consequence, this index is often given high marks in
the methodological area (see, e.g., Munck and Verkuilen 2001: 19–21). It bears
pointing out, however, that the available codebook (Marshall and Jaggers 2000) is
hardly an example of clarity or systematic presentation.
Polity’s prime defects lie on the conceptual level. The connection between operative
measurements and basal democratic criteria is weak. In addition, the weighting given
to various components seems arbitrary. We get no information as to what justifies the
varying point totals, which is handed out. It is also worth noting that tests of the
index’s internal composition have shown that one factor above all others influences
the overall rating of countries, namely: constraints on the executive (Gleditsch and
Ward 1997). This is also that component among the ones employed which seems most
dubious from the standpoint of construct validity.
Vanhanen
This index is based on two components: electoral participation and party competition.
The former is measured by calculating (with the help of electoral statistics) the
proportion of the population who vote in elections; the latter by calculating the share
of parliamentary seats going to the largest party (the smaller the share, the better). The
figures derived in this manner are then multiplied by each other—for both
components are, in the view of the designer, necessary democratic features (Vanhanen
2000).
The strength of this index lies in its simplicity. It has only a few components, and it is
relatively easy to get information about them. It has accordingly proved possible to
extend this index over an unusually long period. It is furthermore based on simple
statistical data, which makes the coding easy and reliable. Thus, Vanhanen’s trump
card is the limited resort to subjective judgements (2000: 257). What is gained in this
Convention