Like what you've read?

On Line Opinion is the only Australian site where you get all sides of the story. We don't
charge, but we need your support. Here’s how you can help.

  • Advertise

    We have a monthly audience of 70,000 and advertising packages from $200 a month.

  • Volunteer

    We always need commissioning editors and sub-editors.

  • Contribute

    Got something to say? Submit an essay.


 The National Forum   Donate   Your Account   On Line Opinion   Forum   Blogs   Polling   About   
On Line Opinion logo ON LINE OPINION - Australia's e-journal of social and political debate

Subscribe!
Subscribe





On Line Opinion is a not-for-profit publication and relies on the generosity of its sponsors, editors and contributors. If you would like to help, contact us.
___________

Syndicate
RSS/XML


RSS 2.0

How to predict the 2001 federal election outcome (maybe)

By Justin Wolfers and Andrew Leigh - posted Sunday, 30 September 2001


The latest turnaround in the polls has focused attention squarely on the forthcoming election. For the first time in months, it appears that the there is real uncertainty over the mood of the electorate. Should Kim Beazley start measuring curtains for the Lodge? Or will John Howard add to his 1996 and 1998 successes to make it a hat-trick?

No-one knows for sure, but there are several schools of thought on predicting elections. Pollsters argue that they can measure the "pulse" of the electorate more accurately than ever before. Economists contend that the business cycle is actually a better way of predicting when governments will be ousted. And the bookies, who make a living out of predicting races better than the rest of us, argue that they have a pretty good idea. Interestingly, when we apply each of these the three models to the 2001 poll, they yield surprisingly different results.

First, consider the most common way of picking an election winner – the polls. Every two weeks, the major polling firms ask about 700 Australians: "If a Federal election were held tomorrow, which party would you vote for?" This graph shows that since the 1998 election, Howard has trailed Beazley on the two-party preferred poll – with the gap widening to as much as 26 per cent. Yet in recent weeks, the Tampa incident and the American terrorist crisis has put Howard into the lead. Although he currently enjoys a healthy margin, it is hard to believe that these two incidents will continue to support the Coalition through to the election.

Advertisement

But even if polls taken during the election campaign suggest that one party is cruising to victory, should we trust them? Through most of the 1993 campaign, polls suggested that John Hewson was likely to defeat then Prime Minister Paul Keating. Yet Labor not only won that election, it also increased its majority. In 1999, all the major polling organizations predicted that the cocksure Jeff Kennett would defeat Steve Bracks in the Victorian State election, leading the ABC’s Kerry O’Brien to assert at the start of the campaign that a Labor win was "almost impossible". But to the surprise of O’Brien, Kennett - and probably even Bracks himself - the polls were proved wrong.

There are two reasons why polls might err. One is straightforward - in our system of government, winning an overall majority of the electorate does not guarantee victory. In the 1998 federal election, Labor won more votes than the Coalition. But unfortunately for Beazley, winning some seats by large margins doesn’t help if your party ends up losing a whole host of seats by small margins. So a poll that simply measures the electorate as a whole may end up incorrectly predicting the result.

The other reason opinion polls aren’t always reliable is a little more tricky. Some researchers believe that merely asking people for their opinion is flawed because the respondent need not put their money where their mouth is. Psychologists argue that when asked to choose between candidates, we often are expressing a preference – although not necessarily a preference for one candidate over the other. For example, at a time of national crisis, it is easy to understand why we might express a preference for stability.

So if pollsters can get it wrong, why not turn to see what the economists have to say? The second theory of predicting elections starts from a simple maxim: historically, the people most likely to lose their jobs in a period of rising unemployment are the Prime Minister and Treasurer. Those who study the effect of economic cycles on elections explain this in two ways. On one hand, when voters learn that the government is not a particularly adept economic manager, they bring in new talent. On the other hand, perhaps the game is about retribution: when voters believe that economic opportunities have been squandered, they exact their revenge at the polls. Whichever theory holds, politicians have a strong incentive to manage the economy well.

The current state of the Australian economy doesn’t seem to augur well for Howard’s job prospects. Although unemployment fell through 2000, it has risen in the past six months, and election studies have typically found that voters have short memories. And it doesn’t matter that much of the present slowdown is part of a global recession – our own research has shown that voters don’t distinguish between a downturn caused by poor economic management, and one caused by these external factors.

Moreover, it seems unlikely that the economic indicators will turn around between now and the end of the year. Given its single-minded pursuit of low inflation, there’s little chance of the Reserve Bank throwing the government a lifeline. To make matters worse for the government, consumer confidence has dipped dramatically in light of the terrorist crisis.

Advertisement

But how much does the fate of governments really turn on the state of the economy? As a way of checking the theory, this graph shows the swing against incumbent governments against the change in the unemployment rate over each electoral cycle since 1949. The trend line up until 1990 shows a fairly direct relationship. Prior to the last few elections it was clear that rising unemployment spelled doom for the incumbent government.

Unfortunately for the economists, more recent data is less clear. Despite the recession of 1990-91 and its after-effects, Keating was re-elected in 1993. Yet three years later, after presiding over a robust economic recovery, he was summarily booted out of office. Although joblessness did not increase in the following two years, the Coalition barely scraped home in 1998. On the face of it, one might think that unemployment was no longer as important to voters. Yet in the 1987 and 1990 elections, less than half of all Australians rated unemployment as an extremely important factor in their voting decisions – in the subsequent three elections, this rose to well above 60 percent

What is going on here? One explanation is that the elections of the 1990s were just aberrations. Another is that the traditional relationships between election outcomes and unemployment started breaking down. The 2001 election will help us understand which is true. No doubt Howard will be hoping that economic indicators matter less today than they once did, while Beazley will be on the side of the traditionalists. If Howard squeaks home this year, researchers will need to start explaining why electoral fortunes are not as tightly linked with the hip pocket nerve as they once were.

  1. Pages:
  2. Page 1
  3. 2
  4. All

An edited version of this article appeared in the Sydney Morning Herald on September 29, 2001.



Discuss in our Forums

See what other readers are saying about this article!

Click here to read & post comments.

Share this:
reddit this reddit thisbookmark with del.icio.us Del.icio.usdigg thisseed newsvineSeed NewsvineStumbleUpon StumbleUponsubmit to propellerkwoff it

About the Authors

Dr Justin Wolfers is an Assistant Professor of Economics at Business and Public Policy Department of the Wharton School, University of Pennsylvania.

Andrew Leigh is the member for Fraser (ACT). Prior to his election in 2010, he was a professor in the Research School of Economics at the Australian National University, and has previously worked as associate to Justice Michael Kirby of the High Court of Australia, a lawyer for Clifford Chance (London), and a researcher for the Progressive Policy Institute (Washington DC). He holds a PhD from Harvard University and has published three books and over 50 journal articles. His books include Disconnected (2010), Battlers and Billionaires (2013) and The Economics of Just About Everything (2014).

Other articles by these Authors

All articles by Justin Wolfers
All articles by Andrew Leigh
Related Links
Andrew Leigh's Web page
Justin Wolfers's Web page
Photo of Justin WolfersJustin WolfersPhoto of Andrew LeighAndrew Leigh
Article Tools
Comment Comments
Print Printable version
Subscribe Subscribe
Email Email a friend
Advertisement

About Us Search Discuss Feedback Legals Privacy