Oz Politics Blog

Some movement in the betting market for the WA election. Centrebet is paying $1.57 for a Labor win and $2.25 for a Coalition win. The implied win probabilities are 59 and 41 per cent respectively. We are back to where we were one week ago. Earlier odds here.

Update 16 Feb 05 @ 9 pm: We are back to $1.40 for Labor and $2.70 for the Coalition. This gives implied probabilities of 66 and 34 per cent respectively.

Update 17 Feb 05 @ 7 pm: Now paying $1.38 for Labor and $2.80 for the Coalition. This gives implied probabilities of 67 and 33 per cent respectively.

The Centrebet Manager has this analysis dated 11 February (slightly out of date now but still interesting):

With only a week to go in the Western Australian election, things have gone pear shaped for Opposition leader Colin Barnett following the release of his controversial plan to spend billions on a canal running from the Kimberley to Perth. It amazing how this revelation has turned the market upside down, and following on from that it is obvious that the residents of WA might not think it is such a good idea.

When betting first opened on the WA election, we posted both Labor (Geoff Gallop) and the Coalition at $1.85. Opinion polls had Barnett ahead, but we know how hard it is to remove incumbents, and were happy to lay them as they got shorter.

A couple of the bigger bets included $10,000 from a Sydney punter at $1.80, and $5000 from a Melbourne man at $1.60. The money from the West was scattered between both candidates, so we still had an open mind on which way it would go, and adopted a ‘wait and see’ attitude. Normally, unless something major happens, betting doesn’t change much until the final few days before polling as things become clearer. Well the press release regarding the canal became a major event as in the following four days we took close to $20,000 worth of bets for Labor, and less than $500 for the Coalition. If that wasn’t damning enough, all the bets came from Western Australia, so we have a similar position with the betting on this election as we did when Mark Latham released his forest policy in last years federal election, and we know what happened after that!

The fluctuations this week have seen Labor come from $2.30 into $1.57, with a recent bet of $10,000 coming at $1.65. Now that the Coalition have drifted out to $2.25 we have seen a sprinkling of money for them (including one bet of $1200 ), but at the moment it is all one way traffic, and while this flies in the face of what the polls say, it is our suggestion that Labor will be re-elected for another term.

Today’s SMH has the headline, Beazley for PM? Not yet, but it’s narrowing. News began its hyperbolic report with

FEDERAL Opposition Leader Kim Beazley had narrowed the gap between the Labor Party and the Government by 10 percentage points since mid-January

Absolute rubbish! The January ACNielsen poll was anomalous - completely out of step with both Newspoll and Morgan for January. The dramatic improvement in Labor’s February ratings compared with the January poll are an artefact, nothing more.

The crux of today’s story is that the previous ACNielsen poll (also here) gave the Coalition a primary vote of 52 per cent, and current poll has the Coalition primary vote at 47 per cent. The previous poll had the two-party preferred vote for the Coalition at 57 per cent. Last weekend’s poll had it at 52 per cent.

Beazley’s first ACNielsen poll results are only marginally better than Latham’s performance at the 2004 Federal Election (this difference is not statistically significant and it does not warrant the subheading from the SMH). They are also essentially unchanged (some slightly worse) than his own performance immediately prior to and at the 2001 election, which he lost and after which he resigned as Opposition Leader.

According to ACNielsen, the Prime Minister’s approval rating has fallen from 67 per cent in January to 62 per cent last weekend. At the time of ACNielsen’s January poll Latham had resigned but the Labor Party had not selected its new leader. Beazley’s approval rating in the February poll is 50 per cent (just one point lower than his final approval rating according to ACNielsen before the 2001 election).

As an aside, since 1996 Newspoll has only once had the Prime Minister’s satisfaction rating at a record 67 per cent, associated with his response to the 1996 Port Arthur massacre. The next highest rating was 65 per cent in November 2002. Although Newspoll’s question is probably different to ACNielsen’s, it further suggests that ACNielsen’s January poll result was an anomaly.

On the question of who was preferred prime minister, 56 per cent backed Howard and 35 per cent preferred Beazley (unchanged for Beazley since November 2001).

The SMH and the Age reported that the poll found 72 per cent of those surveyed thought the Government should apologise to Cornelia Rau (the German-Australian women with a psychiatric condition detained by immigration officials when she claimed to be here illegally) and her family. However, on the question of the broader policy, 53 per cent believed the system of holding asylum seekers in detention until their applications were processed was “about right” or “too lenient” (with nine per cent for this latter category). Forty-two per cent believed it was “too harsh”. It would appear that five per cent were undecided.

The Age at least had a sensible subheading: Beazley’s approval has a whiff of the past.

Another point of interest in the Age’s report was the breakdown by states.

On a two-party basis, Labor leads the Coalition 51 per cent to 49 per cent in WA.

While (I gather) the WA results relate to the Federal sphere, they augur well for Dr Gallop and Labor in the upcoming state election. We might see some movement in the betting market today on back of these results.

It is a shame that ACNielsen does not publish a regular poll report on its website. It makes it difficult to interpret the reported poll findings. Sample sizes are often not reported. Furthermore, not all of the results are reported from month to month, so time series graphs cannot be maintained on the more esoteric matters like approval ratings and the performance of the minor parties.

We are off and racing. The writ for the Werriwa by-election was issued today.

On a matter of pedantry, I would like to know which is correct. Is it by election, by-election or byelection? Your advice on this matter is appreciated. (Although my spell checker accepted the first two, it rejected byelection and suggested ‘bye lection’ - whatever that is).

There are three great predictive indicators for electoral outcomes:

• the opinion polls;
• the betting market; and
• the key economic indices (especially unemployment and inflation rates, but also interest rates, real wages, and growth in Gross Domestic Product (GDP)).

While psephologists argue about which of these three is the more reliable predictor, one thing is certain: on the economic indices Howard has had a dream run. The last recession (two quarters of negative GDP growth) occurred in 1990-91. The recession we had to have was almost 15 years ago. The Australian unemployment rate has been falling steadily since its peak in the early 1990s and is now comparable with the rates in Japan and the United States. Inflation has been relatively stable in the 2 to 3 per cent band over the past 5 years. Real wages have been growing at a faster rate than under the previous government. And interest rates have been stable and low since the mid 1990s. (You can check out the relevant charts from the Reserve Bank of Australia).

According to the SMH, Labor MP Rod Sawford predicted the 2004 outcome on the basis of his simple model.

[Sawford] developed a model for predicting election outcomes, which he says is infallible. It demonstrates the veracity of the political maxim, “it’s the economy, stupid”, beloved of the Clinton campaign in 1992.

Looking at just three indications of economic health - interest rates, inflation and unemployment - Sawford says election results can be predicted. If all three go down over a term, the Government will be returned; if all three go up, the Opposition will win.

Last year all of these economic factors were lower than in 2001, so Sawford was confident the Government would be returned.

This week the Reserve Bank of Australia noted that “the likelihood of further monetary tightening being required in the months ahead had increased.” Decoded, this means an interest rate increase is likely in order to fight the risk of inflation associated with full employment and the country running out of available workers. The market has already factored into its thinking two increases of a quarter of one percent each between now and 30 June 2005.

Because of the link between economic indices and electoral outcomes, a few commentators have wondered whether the star of electoral fortune may finally shine upon Labor and Kim Beazley. Many in the Labor party feel cheated. They believe Howard has only had a dream run in the economy because of the economic foundations established by the Hawke-Keating governments. Doubtless there is some truth in this assessment (see here for example). But it would be equally wrong to attribute all of Howard’s success to the Hawke-Keating years.

In the sometimes unfair game of politics, bad economic performance typically precedes a change government (regardless of whether the performance was locally induced or driven by international factors). Remember Whitlam to Fraser, Fraser to Hawke, and (albeit delayed by three years) Keating to Howard. Some in Labor are now hoping the vicissitudes of the market will herald a transition from Howard to Beazley. In terms of the key indicators noted above, this could happen in one of two ways. An overly tight setting for monetary policy may induce a recession or at least an increased unemployment rate. However, insufficient tightening of monetary policy could see inflation break out of the 2-3 per cent per annum target range as the economy approaches full employment.

Personally, I am not convinced that the threat to national economic wellbeing is as big as some predict (or at least hope for). The Australian unemployment rate is currently running at 5.1 per cent. In New Zealand it is 3.6 per cent and inflation is still within the 2 to 3 per cent band. (Admittedly interest rates in New Zealand are around 1.25 per cent higher than Australia). This suggests that while there may be some sectoral hot-spots, there is some way to go before a low unemployment rates begin to drive significant inflationary pressures in Australia.

Second, there is a heightened sensitivity to interest rate movements in the economy. Because of an unusually high level of household debt, any interest rate increase is likely to result in a higher reduction in aggregate demand than has been the case in the past. This means the Reserve Bank need only apply a modest interest rate increase, with a correspondingly modest impact on business cost structures, but a larger impact on aggregate demand. In plain-English, when it comes to managing inflationary pressures, interest rate adjustments are not as blunt an instrument as they have been in the past.

All in all, I think it is far too precipitous to suggest that the Reserve Bank’s foreshadowing of the possibility of a small tightening in monetary policy heralds a change of government in 2007. No one is suggesting a return to the interest rates of the late 1980s and early 1990s. While there are risks, they currently appear easily manageable with limited impact on the economy.

Apologies if this service is interrupted, just plain ugly, or refuses to let you make a comment.

I have decided to trial the new Wordpress 1.5 software (in gamma development prior to release). It means I will spend the next few days trying to get the blog looking decent as I develop a presentational theme with a sufficiently wide column to hold some of my bigger tables and images.

Update: I have now have the production version of WordPress 1.5 up and running, with a wide column.