Among the string of extracts The Australian has published from its own book, The Howard Factor, one of the more controversial was by Caroline Overington
on the supposed phenomenon of young people voting for Howard. She calls
him the “Prime Minister of choice for people aged 18 to 24”, and says
“Howard’s new constituency, the ‘young fogies’, adore him the way their
parents loved to smoke dope.”

But the evidence on which this thesis rests is incredibly thin, and Peter Brent at Mumble has given it both barrels. Last week he accused
Overington of “trawling the hundreds of opinion polls published over
the last decade to find two or three showing youth favouring the
government (and ignoring the other several hundred surveys showing the

This week he’s returned to the attack with the aid of ACNielsen survey figures provided by Tim Colebatch at The Age.
Nielsen’s polling in 2004 was very accurate; the average of its last
four pre-election polls picked the two-party-preferred vote spot on. So
there’s plenty of credibility behind its contention that the youth vote
(18 to 24-year-olds) split almost 55%-45% in favour of Labor.

Here are the numbers:

Age group Coalition vote (2PP) Swing from 2001
18-24 45.4 1.1% to Labor
25-39 49.0 0.5% to Coalition
40-54 52.6 2.8% to Coalition
55+ 59.3 0.3% to Coalition
Total 52.7 0.9% to Coalition

(As Brent notes, the total swing is too low, because Nielsen overstated
the Coalition vote in 2001, so all the swing figures should be about 1%
better for the Coalition.)

So if these figures are right – and it’s both a much larger sample and
a better predictive record than the AES survey Overington uses – then
the Coalition vote is quite heavily concentrated among the over-40s,
and becoming more so rather than less. Which, given the government’s
policies, is not at all surprising; as Colebatch says, it’s “what
anyone with their head right way up would expect.”

Mythology, however, is often stubbornly resistant to hard data. Let’s
hope that’s not the case with the youth-enthusiasm-for-Howard thesis.