To Peter Cave:
PC:
> I have always found it baffling why the only default
question in
> political opinion polls in Japan is whether the voter
'supports' the
> current cabinet or not. Not only is the phrase vague
(what on earth
> does it mean to support the cabinet?), but it
generally always yields
> the same result: at the start, higher ratings ('give
them a chance'),
> gradually or rapidly becoming lower ('yappari dame'),
though there are
> of course rare exceptions.

ESK: Absolutely. I think you, me and Paul all agree that it is a very flawed question but it is the one that Japanese press uses, so we are stuck with it as the closest we can get to supporting the PM.

PC:
> British political opinion polls might well produce
the same results if
> the voters were ever asked this question, but
instead, the default
> question is on voting intentions: specifically, if a
general election
> were held tomorrow, who would you vote for? This
produces results
> which seem more varied and also more relevant than
the question asked
> in Japan - more relevant because the question is
clear and focused and
> also related to something entirely practical that
voters will
> eventually get the chance to do, i.e. vote.

ESK: Again. Agreed. Closest they seem to come in Japan is to ask "Which party do you support?" But that always depresses vote intention since so many voters are "floating voters" and even those who consistently or almost always vote for one party don't necessarily say they are party supporters. That is why in US survey research they usually ask both about party support but also about voting record in past and also likely voter in next election.
Best regards,
Ellis