The hard numbers of politics

What happens if no party wins 76 seats on election night?

Anne Twomey, writing in The Conversation, explains what happens if the 2022 election results in a hung parliament.

The short answer is “nothing much” immediately, because according to convention a prime minister continues in office, albeit in a “caretaker” capacity, until he or she resigns, or loses support in the House of Representatives in a no-confidence vote.

If the prime minister resigns or is booted out in a no-confidence vote after a party secures an agreement with other parties or crossbenchers, the governor-general simply appoints a new prime minister from that party. Only in a very unusual situation would the governor-general have to exercise any discretion in appointing a new prime minister. (One such situation may be the crossbenchers accepting that the current government can stay in office, but not with the current prime minister, as was the case in 1922.)

By convention, a minority government can stay in office so long as parliament grants it supply, and it does not lose a vote of no-confidence in the House of Representatives.

Twomey dispels the notion that a minority government leads to chaos. A parliament with a minority government can help bring about a centrist consensus on policies, but it can also be unnecessarily conservative and cautious about confronting hard issues. She also reminds us that it is not unusual, particularly at a state level, for a government to hold office without holding a lower-house majority. (As a case in point, the Coalition government in New South Wales holds only 45 seats in the 93-seat Legislative Assembly.)

She could have gone on to explain that on the conservative side of politics a minority government is the norm. The Liberal Party rarely, if ever, holds a majority in the House of Representatives. The “Coalition” is actually two parties, and in only 3 of the 18 parliaments where the Coalition has held office has the Liberal Party had a majority of seats. The “Coalition” is a secret deal between two parties with different ideologies and different constituencies, and as any observer of the climate wars over its latest term of office can observe, it is hardly free of chaos.

Polls – no sign of a Coalition comeback

As election day approaches there will be a frenzy of polling. William Bowe’s Poll Bludger site can keep us informed with his reporting and well-considered analysis.

Newspoll, which has been reporting weekly since the beginning of April, suggests that Labor’s support has bounced around in the last couple of months, but is back on trend, while the Coalition’s support is stuck in the doldrums.

The situation as reported by major polls at the start of the week is shown below. Compared with the last election support for the Coalition is down by 5 or 6 percent, support for Labor is up 3 or 4 percent, support for the Greens and right populist (UAP and One Nation) is up, and support for independents and “others” is up.

(Anyone wishing to dig into these three polls can find figures on Poll BludgerMorgan in one link, Newspoll and Ipsos in another.)

On the basis of these figures it would take an extraordinary turn of events for the Coalition to have any chance to win a majority of seats in the House of Representatives.

There are estimates of two-party-preferred outcomes, showing a Labor lead of about 8 percent, but at each election, as support for the two main parties falls and as support for independents and minor parties rises, these TPP estimates are becoming less useful in their predictive power, particularly when there are strong independent candidates in particular seats. The hard job for pollsters is to predict what will happen in individual seats.

The YouGov MRP poll, released on Wednesday and Thursday (links are to Poll Bludger’s summaries) attempts to fill that gap by providing electorate-by-electorate forecasts. The sample size is huge, but it has to be when looking at 151 electorates. Notably it combines established sample survey methods with another established method, the use of regression analysis based on electorate demographics (age, education, country of origin etc), a common tool in marketing. Those who find delight in advanced mathematics can read about this combined technique, multi-level regression with post-stratification, in the American Journal of Epidemiology.

Most of us are more interested in the results, rather than the method. This poll yields results predicting that Labor will win 80 seats, the Coalition 63, and others 8. On his Poll Bludger site Bowe provides state-by-state estimates of party support based on the YouGov MRP work, which confirms the results of other polling: Coalition support has fallen with some going to Labor, some to far-right populist parties, some to independents. Support for far-right populists is around 8 percent in most states, and 13 percent in Queensland.

Because this poll has commanded widespread media attention, particularly with its 80-seat prediction for Labor, it is likely to influence some voters’ behaviour, but there is no way a clear prediction can be made. For example, some may be (irrationally) fearful of a “hung” parliament, and switch their vote to Labor, while some may welcome a minority government because it keeps the bastards honest and switch from Labor to independents.

With so much attention focussed on the competition for the House of Representatives, it is easy to overlook the contest for the Senate, where in each state six senators are elected. Four of those places are generally taken by the major parties, but the other two are up for grabs, and are contested by major parties, minor parties and independents. Also the ACT is voting for two senators in a contest that has no assurances because of the presence of strong independents. William Bowe has an article in Crikey taking the reader through possible outcomes and it is summarised on his Poll Bludger site. The short conclusion is that even if Labor wins a majority in the House of Representatives, there is no guarantee that it will be able to wave legislation through the Senate.