There's a highway that stretches from Burnie to Launceston in Tasmania that's so smooth your wheels almost sing when they hit it. Keep driving further south, though, and the music stops.
Despite years of pleas, the road between Launceston and Hobart has only just received meaningful upgrades, even though it connects the two largest population centres on the island state. The difference? The first connects two marginal seats - Launceston in Bass, and Burnie in Braddon. Braddon has switched between Labor and the Liberals at six of the last eight elections. Bass has changed hands in seven of eight.
Hobart, meanwhile, sits in the electoral division of Clark. It was comfortably held at the 2019 election by independent Andrew Wilkie, who's represented the area since 2010. Before that, it had been Labor territory since 1987.
Is it fair that seats like Braddon get prime rib while others beg for scraps? Perhaps not. But Australians appear to have long ago accepted that this is how the system works.
While there's been much anger about the alleged misuse of the $100 million Community Sport Infrastructure Grant program in the lead-up to the 2019 election, as revealed last week in an explosive report from the Australian National Audit Office, we've long understood that it's not need that determines which places get gold-plated highways. It's the quirks of electoral distribution. And instead of crying out for reform, we cry "make it marginal".
Pork-barrelling is not an Australian phenomenon. It's even been argued the construction of the Egyptian pyramids was a form of pork-barrelling, designed to keep the peasants from rebelling.
The practice is rooted in the United States, where politicians who bring home the bacon have long been rewarded at the ballot box. But as David Denemark wrote in 2000, pork-barrelling is a "decidedly non-partisan" practice in the US. There, the carve-up generally reflects an individual politician's committee or leadership power.
One of the most notorious examples of this is the "Big Dig" project in Boston. United States House of Representatives Speaker Tip O'Neill successfully lobbied for federal funding to send 5.6 kilometres of his local highway system underground in 1982. The project was supposed to be finished in 1998 and cost $2.8 billion in 1982 dollars. It was actually completed in 2007, and cost more than $8 billion (not adjusted for inflation). The Boston Globe in 2008 estimated that, with interest, the project would actually cost $22 billion and would not be paid off until 2038. Members of Congress are said to have cheered when the US government vowed to stop cutting cheques for it in 2000.
But in the Australian parliamentary system - with its ultra-disciplined parties, and where seats won on the slimmest of margins can mean the difference between forming government and defeat - pork-barrelling becomes an entirely new beast.
"Unlike the sharp dichotomy between the interests of the individual and the party that fuels constituency activities in America, parliamentary parties in government confront a collective electoral imperative to assure the victory of their most vulnerable party colleagues in marginal seats," Denemark writes. "In short, the parliamentary gap between individual and collective interests is virtually non-existent."
That's why you hear squeals during every Australian election, even though only 30 per cent of seats are considered marginal. It's also why Canberra remained largely dry while Bridget McKenzie was splashing cash around in what has become known as the "sport rorts" scandal.
There were 13 sporting clubs that received grants in the ACT. Woden Valley Gymnastics actually managed to snag a $200,000 grant in both round one and round three of the program (a total of $400,000), even though Sport Australia quietly planned on limiting organisations to one grant each. Collectively, Canberra clubs got around $1.6 million. Tasmania received about $3 million across 23 clubs. Sure, Tasmania has about 130,000 more people than the ACT. But three of its five seats are also marginal. Canberra, meanwhile, is largely safe ground for Labor. The last Liberal politician elected to a lower house seat here was Brendan Smyth in 1995.
Ironically, we only got Smyth because the Member for Canberra, Labor's Ros Kelly, was forced to resign due to the first sport rorts scandal. In 1993-94, Kelly was found to be using a whiteboard in her office to divvy up grant funding to marginal Labor electorates. While the auditor-general of the day was unable to prove the funding decisions were politically motivated - because the evidence had been wiped away with a dry-eraser - two separate studies have concluded the money was directed towards swing voters.
Kelly protested that her allocation of the money reflected social need, not sandbagging. While Clive Gaunt in his 1999 study found a slight correlation between grant funding and areas of high unemployment, it did not explain "variations in the magnitude" between each electorate. What he did find was that the pattern of spending was consistent with pork-barrelling. A very marginal Labor-held electorate would receive $172,952 more than the equivalent safe Coalition-held seat, according to his analysis. A very marginal Coalition seat received around $71,000 more in grants than the equivalent safe Coalition electorate.
McKenzie, meanwhile, has claimed she was performing the "reverse pork-barrel". She said her intervention meant more money went into Labor electorates than would have if she'd followed Sport Australia's recommendations. As our current Auditor-General all too inconveniently points out, this was because the senator's office had identified them as targets through a colour-coded spreadsheet.
But how effective is pork-barrelling as an electoral tactic? Another Canberra MP has helped shed some light on this. Andrew Leigh, in a former life as an Australian National University professor, uncovered "robust" evidence that you can in fact buy votes. In a 2008 paper, Leigh examined four discretionary grants programs (read: slush funds) used during the 2001-04 election cycle. He found that for every extra $1 million spent on Roads to Recovery funding, for example, the Coalition increased its vote share by between 0.06 and 0.37 percentage points. Since the average number of votes cast per electorate in 2004 was 82,367, this indicates each extra vote gained through the Roads to Recovery program cost between $3281 and $20,234.
Ultimately, McKenzie could be forced to pay the price for all the largesse in the lead-up to the 2019 election. But given Australians have demonstrated that their votes can be hamburgled over a long period of time, she will not be the last to try it.