>>> [EMAIL PROTECTED] 5/31/2004 2:06:42 PM >>>
Gitlin is a repulsive character, but everything he says in this
passage is, sadly, true:
>But there is no evidence that nonvoters differ from voters in any
>ideological way.
poli sci guy stephen earl bennett's 1990 'deconstruction' ('the uses
and abuses of registration and turnout data', _ps_) of above exposed it
for canard it actually is...
'truism' gained prominence with couple of 1988 post-election surveys
indicating that non-voters would have cast ballots for bush (54%) in
roughly same percentage as voters did, bennett
disclosed extent to which after an election even folks who didn't vote
say they prefer winner...
2000 national election study (nes) included question about what to do
with ostensible 'budget surplus' available at that time, there was
significant difference between voter and non-voter responses with former
favoring tax cut and latter favoring spending for education, health
care, etc...
actual turnout among lower-income folks increases in rare instances
where candidates (dems in most places) are perceived as concerned with
their needs, even when chances for victory are only modest (not,
however, when hopeless as in too many minor
party/independent/alternative campaigns)...
upper income voters three times more likely to vote for reps than lower
income voters, helps explain why effort is put into attempts to
discourage latter from voting, no need to 'purge' voter rolls as florida
does under rep. governor bush if larger low-income electorate wouldn't
make difference...
in any event, asking certain questions of separate and distinct
individuals and then aggregating responses creates opinion that wouldn't
otherwise exist, polling organizes 'publics' in ways that they wouldn't
on their own...
in sum, example of what c. wright mills called 'crackpot realism'...
michael
