such formal concepts for the intellectual clarification of party com-
petition if opinion surveys revealed even a modest tendency on the
part of respondents to employ left–right placements of self and others
(parties) with specific policy positions in mind.”^8 As we show below,
this is exactly what happens with citizens and elites of the region.
Insofar as they appropriate the left–right distinction, and around
70 percent do, respondents from the former Soviet sphere of influence
do so in conformity with our expectations.
Among Islamic countries, the high level of non-response probably
has more to do with the lack of democracy than with cultural or
religious orientations. Bangladesh, Indonesia, and Turkey have non-
response rates below or near the world proportion. The populations
of these nations are overwhelmingly Muslim (from 83 percent for
Bangladesh to 99.8 percent for Turkey), but benefit from political
regimes that are open and democratic compared to other Muslim
countries.^9 Algeria, Iran, Jordan, Morocco, and Pakistan, by contrast,
all countries whose populations are more than 90 percent Muslim,
have more authoritarian regimes and very high rates of non-response.^10
Likewise, the rate of non-response for Mexico could be explained
by the non-consolidated nature of democracy until 2000. India, an old
and well-established democracy, appears more puzzling and would
require further inquiry. All in all, however, these proportions of non-
responses confirm the importance of democratic debates in con-
structing meaningful left–right oppositions, across various cultural
and geographical areas.
If we turn to the left–right means, the country with the most leftist
average is Zimbabwe (3.66), and the one furthest on the right, if we
leave Vietnam aside, is Bangladesh (7.56). Zimbabwe, Vietnam,
Bangladesh, and Tanzania are the only cases with means more than
1 point away from the world mean, which implies that all countries
share broadly similar distributions. There are nevertheless important
remaining differences. At one end, Zimbabwe, Montenegro, Spain,
(^8) Herbert Kitschelt, Zdenka Mansfeldova, Radoslaw Markowski, and Ga ́bor
To ́ka,Post-Communist Party Systems: Competition, Representation, and Inter-
9 Party Cooperation, Cambridge University Press, 1999, p. 283.
Freedom House rates Bangladesh, Indonesia, and Turkey as partially free, with
combined scores, respectively, of 8, 7, and 7 (www.freedomhouse.org).
(^10) Freedom House classifies these countries as non-free or partially free, with
combined scores at or above 10 (www.freedomhouse.org).
A worldwide value divide 39