This was the article I had in the Crikey mail earlier today:

In terms of electoral demography — the who lives where, earning what, and how they vote, framework — the two big issues coming out of the budget are the $150,000 household income threshold being the primary means testing weapon on the one hand, and the increase in the aged pension on the other.

To look at the former, we can break down each electorate by household using the 2006 Census data, and measure the proportion of households in each electorate that are couple households with dependent children earning around the $150K per annum mark or greater — we’ll call them FDC150’s for brevity. This gives us a rough idea of the proportion of each electorate that might develop some animosity toward the government as a result of the stricter means testing for not only Family Tax Benefits part A and B, but the means testing of the private health insurance rebate as well.

If we then rank those seats from the highest to the lowest — we find that the ALP has little to fear. Of the top 15 seats where these FDC150 voters are highly clustered, 14 of them are Coalition held with only Canberra being Labor territory.

In order from highest to lowest the seats are: Bradfield, Berowra, Mitchell, Kooyong, Warringah, Goldstein, Menzies, Mackellar, Hughes, Moore, Ryan, Tangney, Canberra, Curtin and North Sydney — quite the blue ribbon collection. The average proportion of these top 15 FDC150’s is 18.5%, compared to the national seat average of 9.6%, while the average two party preferred vote in those 14 Coalition held seats is 58.6%. The means testing contained in the budget has only created large clusters of potential animosity for the government in seats where, apart from North Sydney and Ryan on an extremely fortuitous day, the ALP couldn’t win in a pink fit.

If we just look instead at very marginal seats which sit on a two party margin of 2% or less, Bennelong and Flynn (held by Labor), and the Coalition held seats of McEwen, Bowman, Dickson, La Trobe, Macarthur and Cowan are the only ones where the FDC150’s are in the double digits — and then only barely with all being 12% or less.

In fact, only 9 of the 30 most marginal seats in the country have FDC150’s in double digits.

Yet most of these marginal seat FDC150’s are clustered in areas within those seats where the closest booths to those wealthier neighbourhoods are generally strong ones for the Coalition already, suggesting that most of any potential animosity won’t play out at the ballot box since they already vote Blue rather than Red.

If you were a highly political administration that had to choose a means testing threshold that delivered the most political and economic benefit for the least political pain, 150K per year is pretty much bang on the mark you would choose. If you chose any threshold lower than that, say 120K per annum, you would start to hit the hip pockets of significant numbers of people in large numbers of swing seats. Instead of having only 9 of the 30 most marginal seats having double digits of FDC150’s, there would be around 22 to 23 of the 30 most marginal seats being in double digits if the group became FDC120’s.

Two charts here will increase the cynicism even further. If we run a scatter plot and a regression line of FDC150’s against Family Tax Benefit Part A recipients across all 150 electorates, we see that there’s an obvious relationship between income and Family Tax Benefit qualification.

If we keep that chart in our thought orbit while we look at another chart — this time the ALP two party preferred swing at the 2007 election against FTB part A recipients, it shows us that the ALP will not, under any circumstances, alienate too many of the FTB A set — which a cynic might think is why the 150K threshold was chosen.