Parallels between Howard's November 2001 win and Keating's in March 1993. And some differences.
Both were "remarkable victories" by governments seeming to not have a prayer just months earlier.
Both governments had scraped through at the previous election (1990 and 1998) with a minority of the two party preferred vote.
In both cases the government achieved a swing to win. This is rare; two party preferred support usually ebbs governments out of office.
Furthermore, given the terms of the electoral contest, each needed a decent swing to hang on. Keating's 1.5% turned out to be 0.5% more than he needed, ie had he gotten under one percent Hewson would have won.
Howard's 2.1% is looking to contain about a 1.5 percent safety margin; had the Coalition gotten only a half percent or less national swing, it would have lost office on November 10.
Both won on a single issue that exemplified voter misgivings about the Opposition leader. John Hewson the heartless "feral abacus" with his manic ambition to turn Australia into an economic test-tube; Kim Beazley: a weak leader who flip flopped on the tough issues and didn't stand up for his country against foreigners in boats and UN bureaucrats.
Both wins were "battler" heavy. (I use the term "battler" advisedly; it can be patronising. Here it means poorer, less educated urban-dwellers - yes, still patronising.) That is, they predominately comprised big swings among people at lower end of socio-demographic range, many of these in Labor's "heartland". These people's votes tend to have low bang for the buck (as they reside disproportionately in safe seats).
Hence their "needing" to get decent swings to retain office. The 2.1% swing to Howard gave him just 3 extra seats (increasing his majority by four). Keating's 1.5% swing gave him 3 seats (increasing his majority by six.)
To understand how low three seats from a 2.1 percent swing is, plotting it on a national pendulum prior to November 10 would have increased Coalition's majority increase by 16 seats. (Don't have a copy of pendulum going into 1993 election, but imagine a similar situation.)
Both were of course politically clever election campaigns run by good communicators.
Both tried afterwards to claim the result was actually not about that single issue, but because they ran such wonderful governments.
Throughout 1993 campaign, most expected government to lose, while in the campaign period 2001 (though not earlier in the year) most thought, correctly, that the government would win.
Labor in 1993 had been in for 10 years, while Coalition this time only 5 and a half.
Coalition's 49% win in 1998 had an aberrational quality to it, being in large part a GST referendum.
"Battlers" are Labor's natural constituency, and despite much recent chatter in newspaper columns regarding the western suburbs of Sydney, continue to reside overwhelmingly in Labor held seats. So Keating's battler-swing saw him wasting votes with high majorities in safe seats, Howard similarly got big swings in safe Labor seats, not enough to give him these seats.
In both cases the "battlers" in more marginal seats swung sufficiently to save the government.
Keating waited a respectable interval (a year or so) before trying on his brand new version of history: that the result was actually an indication of voter support for his government. This was universally pooh-poohed by the commentocracy.
Howard and team by contrast started rewriting at close of polls on November 10, with much assistance from friendly editorials and commentary. The Australian in particular is pulling out all stops; it's to give the government its mandate so it can just get on with the job and continue making the tough decisions for the betterment of the nation, you see.
Broadsheet editorial opinion unanimously opposed Keating in 1993, close to unanimously endorsed Howard in 2001