Some Labor campaign leaders now blame their 2004 election loss on a fear of increased interest rates under a Labor Government, as if somehow, like security issues after the arrival of the Tampa in election 2001, countering fear campaigns is beyond their control.
At first glance the fear of higher interest rates appears plausible, given the fact Australian households currently owe $431 billion, of which $279.4 billion was outstanding on mortgages, according to data provided by the AMP.NATSEM Report on Debt in Australia.
Before drawing any conclusions, it is useful to take an objective look at this question by examining the NATSEM report in the context of the post election demographic modeling for 2004, done by John Lockwood and me.
This election modeling uses computer based comparisons, or correlations, of the election results with our Census-based demographic model, including variables covering age, income, occupation, industry, family type, education and mortgages.
It does indeed show a swing towards the Coalition from suburbs dominated by homebuyers, those with most to fear from high mortgage interest rates.
But, when broken down further, to take into account the size of each homebuyer’s mortgage, this Coalition swing was seen to be concentrated amongst those voters with lower mortgages, typically of less than $1000 a month. This represented just over half of those paying mortgages.
In the upper mortgage ranges, the top one-third of homebuyers paying $1200 or more a month, actually swung to Labor and against the Coalition. This trend was confirmed by correlations for the average size of the mortgages in each electorate, which showed seats with the highest average mortgages swung strongly to Labor.
If mortgage payers were afraid of higher interest rates under Labor, we would expect to see the reverse of this pattern: the bigger the mortgage, the bigger the swing against Labor. We saw the reverse.
But perhaps it is a function of the ratio of household debt to income, and lower income earners have a higher proportion of debt to income? Well, the NATSEM report’s very first table shows households with the lowest positive incomes, from $1 to $20 000, have the lowest total debt of $9700, roughly equal to the mid point of this income range. This percentage actually rises, as income increases, so that those households earning more than $100 000 are carrying debt of $143,200, including mortgages of $86 200.
So the rich have bigger mortgages and more to lose, if interest rates were the spur, so again, we see evidence contradicting the argument that fear of interest rates cost Labor the election.
But wait, perhaps the fear of interest rate rises meant more to the lower income earners, as their disposable income, after servicing debt, was much more tightly stretched than for the rich? This is pretty much the last straw for those supporting the argument that interest rates cost Labor the election.
However, when we rank the correlations, and run further, more sophisticated statistical modeling, to see which variables were simply descriptive, or ecological, and which highlighted stronger, underlying behaviour, we see the swing took place amongst formerly rusted on Labor voters - those in lower to middle income, unskilled or blue collar trades jobs, who just happened to be paying lower to middle sized mortgages.