Dr Kevin Bonham
MOST OF THE DUST having finally settled (except perhaps for the residents of McEwen and Bowman) it is time to consider what we can learn from the results of the federal election just completed. This piece is a wrap-up of the success or otherwise of my own projections and those of others, plus extensive analysis of the Tasmanian results. In summary, both Labor’s recovery of Braddon and the large swing to Labor in Lyons were primarily due to the reversal of the Latham forests policy disaster from 2004, while the causes of Labor’s recovery of Bass appear to have little or nothing to do with the infamous pulp mill.
My predictions published just before the election was called can be found in my article, It Ain’t The Economy Stupid. Nationally, Labor has won 83 House of Reps seats, the Coalition 65 (provided McEwen is not overturned) and Independents 2. My own prediction, 87 seats, was over by four, but I regard this as a rather good result given that it is fairly hard to predict exactly where a big swing will land, and given that a narrow majority of serious psephologists were further out than that, with several tipping in the very high 80s or even well into the 90s. While there are so many people posting predictions on psephological websites that a small proportion of them are bound to get the seat total right, I’m not aware of any psephologist with their own column anywhere who was absolutely spot-on. Peter Tucker, who, one week out from the election predicted 85 on his website, Tasmanian Politics, was the nearest I remember, while Adam Carr’s notoriously cautious prediction of 80 seats for Labor scrubbed up a lot better than most in the end.
Alas, I could have been a lot closer … on 7 September I posted a prediction of 84 seats on various forums, only to raise that to 87 a few days out in the face of evidence that the much-discussed “narrowing” of the polls was yet to appear to any great degree. I am in good company here as, quite independently, William Bowe ( The Poll Bludger ) did exactly the same thing. In the end, the margin narrowed in the final days despite the Coalition running a generally poor campaign. I suspect that the last of the late-deciding voters saw insufficient reason to ditch the incumbent Government, unlike in 1996 when they too had ultimately had enough.
Generally, those pundits who were too slavishly devoted to opinion polls and projections based on them tended to overestimate the number of seats Labor would win, although not hugely, while those who were within five seats were often taking the polls with some degree of caution and using other factors (including, in my case, the betting markets, which are actually very useful once you get very close to the election) to work out which individual seats might or might not fall.
In my article, I gave a list of 25 seats where I considered Labor favourites to win. Labor won twenty of those (an 80% strike rate being no great predictive feat since several of those were no-brainers). The Coalition retained Bowman by next to nothing, LaTrobe, Sturt and Paterson with varying degrees of closeness, and the one where I wasn’t even in the ballpark was McMillan, where Russell Broadbent performed brilliantly to suffer hardly any swing at all. There were five Labor gains I did not have on my list: Forde, Longman and Dawson in Queensland, and Bennelong and Robertson in NSW. In three of those cases I was cagey because of the high profile of the incumbents for the size of swing required. However, this line of thinking was fallacious. As Charles Richardson (who provocatively called Bennelong as a “definite” Labor gain before the election) pointed out post-poll, whatever factors could have assisted John Howard to resist a swing in his seat were already factored into his vote in 2004, so there was actually no reason to expect that he would do so again.
I mentioned it was possible Labor might lose up to three seats in Western Australia; they lost two. In the Senate, my prediction that the Greens would again not meet some expectations was correct, but they actually had even less success than I predicted, winning only three seats instead of my expected four or five. However, at least they managed to not have Kerry Nettle returned, which in the long term will probably work in their favour.
There has been much discussion of whether this result is a “landslide” for Labor. In my view, it was merely a comfortable victory. There is one school of thought that says a landslide is a big swing in your favour and another that it entails victory by an overwhelming margin. Some may argue that a landslide represents a great movement in the electorate, but I prefer the latter version: the burial of the opposition. In that light, I suggest victory by more than a quarter of the parliament, or a 2PP result of 53.5% or better, as useful criteria for a “landslide” win. This wasn’t such, and indeed, Labor has won this election less emphatically than the Coalition won the last one.
Tasmania
A general feature of the Tasmanian results was the very strong performance by the Liberals in keeping the House of Representatives swing down, although it was not enough to save their two marginal seats. Both Bass and Braddon are now rather marginal Labor seats, Michael Ferguson having put up a much stronger fight than expected to hold the swing below four percent.
I gave predicted totals including 2PP for the five Tasmanian House of Representatives seats and the Tasmanian Senate. In all I made 31 predictions, of which twelve were accurate to within one percentage point, ten to within two, four to within three and two to within four. The three I got more wrong than that were the size of the Liberal votes in Bass and Franklin and the two-party preferred vote in Franklin.
Of the 21 tweaks I applied to the combined EMRS poll figures to get my predicted total, 14 made my projections more accurate and seven didn’t. Of those that worked, in all cases bar one I went in the right direction from the polling, but not far enough; I will be bolder next time! The items where my tweaks were most successful were the size of the Green vote, the size of the vote for “others”, and the two-party preferred vote. It appears that any polling that gives a list of parties including the Greens and “others”, will overestimate the Green vote and underestimate the vote for “others” – by about 2.5 and 2 points respectively in this case.
I was most pleased with my Senate prediction for the state, where, having no real idea exactly where the clearly substantial swing would land, I took a blind stab “with much trepidation and an expectation of being way out” at Labor 39.5 (actual 40.1), Liberal 37.5 (37.4), Green 17 (18.1), Family First 3 (2.05), others 3 (2.35).
I was not so pleased when, immediately following the election, bogus claims that the Greens stood a chance of a second Senate seat on the back of their greater share of below-the-line votes began to be made both in the Sunday Tasmanian/Mercury and on 7ZR.
It turned out that in both cases the main source of the false claims was not the Greens’ habitual optimism-beyond-the-call-of-reason, but Associate Professor Richard Herr mistakenly believing that below-the-line votes were not included in the AEC’s election-night Senate totals. In fact (and I called the AEC in 2004 to check this), on the night of the poll, the AEC’s totals include both above-the-line and below-the-line votes, sorted by party, but otherwise described as “unapportioned”. As counting progresses, postal, pre-poll, absent and provisional votes are added and sorted, but the ATL and BTL votes counted on the night are already included in the unapportioned tally, and are sorted into ATL votes by party and BTL votes by candidate. The sorting of unapportioned into ATL and BTL does not increase the party’s totals, and thus a party with a high proportion of BTLs (eg the Greens) will only improve its position compared to polling night if declaration votes (postal, pre-poll, absent and provisional) are in its favour. A party that has a high proportion of BTL votes is probably in a worse position than its election-night result reveals, because it is more likely to lose preferences through leakage. Despite the Greens’ excellent Senate vote Andrew Wilkie never had a ghost of a chance of election in the cut-up.
It was odd that substantial derision was rightly directed at Michael Ferguson for failing to concede Bass on the night when he was over 1000 votes behind and hence had no realistic chance, but less was directed at Bob Brown for not writing off Wilkie, who was clearly trailing on the night not by a thousand or so, but by over ten times as many. Indeed in the end, Andrew Wilkie was defeated because he trailed the third Liberal Don Morris by 18,726 votes at the point where he was excluded. Morris himself was marginally closer to winning than that (losing to Bilyk by 18,370 votes), but it was one of the widest final-seat margins in Tasmania’s Senate history, and to have multiple sources fanning the belief that it could somehow be close after polling night was one of the stranger displays of cluelessness I have seen in the public coverage of election results … and that’s saying something!
I sent a brief letter to the Mercury pointing out the blunder, which so far as I’m aware was never published, and an SMS to 7ZR – I am not sure if the latter was ever read out, but if so it was certainly not done promptly.
Bass
Bass was one of the two most-discussed seats on this site in the leadup to the election, the other being Lyons, in both cases because of the pulp mill. Not surprisingly, Bass had the largest swing to the Greens (7.17%) in the country, but the Green vote was still only just over 15% and not into the twenties as some had expected. The result was well in line with my prediction of “a significant but not massive increase in the Green vote (perhaps into the mid-teens but no higher)” in my article, The shrinking pulp mill backlash. The other prediction that article made was that the pulp mill would have at most a modest impact on the two-party preferred result. If anything, it had even less than I expected.
If we consider the “anti-mill index” (AMI) to consist of the vote for anti-mill independent Sven Wiener plus the swing to the Greens (although the swing to the Greens in Tasmania was not entirely pulp-mill driven, and also Wiener’s vote was exaggerated by a donkey component) then the following eighteen booths have the highest AMI:
Weymouth* 19.8, Hillwood* 19.5, Trevallyn 16.9, Blessington Upper 16.3, Dilston* 15.3, Lilydale* 14.8, Launceston Central 14.2, LGH 14.15, E Launceston 13.9, S Launceston 13.4, Lebrina* 13.1, Pipers River* 12.5, Cosgrove Park 12, Riverside West 11.7, Ringarooma 11.5, W Launceston 11.4, Windmill Hill 10.9, Fiveways 10.3, Legana(Bass)* 10.2.
The booths with asterisks are East Tamar booths (for completion, the others listed in Peter Henning’s article at Rudd and the pulp mill were Karoola 9.1, George Town 9.5 and George Town South 4.5 – the latter was below the seat average AMI of 8.9). But what is also noticeable is the large AMI in urban Launceston generally, which accounts for most of the remaining booths in this list. (Blessington Upper has only 71 votes and hence can be ignored on the basis of small sample size. Not sure what’s going on in Ringarooma.)
The important question is, did this voting pattern translate into an impact on the two-party preferred result between Jodie Campbell and Michael Ferguson, and hence deprive the latter of his seat? Looking at the data I can find no real evidence that it did. The 2PP swing to Labor within Bass was 3.63%, but of the nineteen booths listed above, eight of them, including three of the seven East Tamar booths, had 2PP swings to Labor that were either less than this amount, or else swung to the Liberals. The seven East Tamar booths with the highest AMIs delivered 2PP swings ranging from 3.1% to the Liberals at Pipers River to 7.1% to Labor at Lebrina. This is much as would be expected if the pulp mill had no net influence on two-party preferred vote between the major parties.
Indeed I have painstakingly crunched the AMI compared with the 2PP swing for the entire electorate of Bass and detected not even a whiff of a significant correlation. This is remarkable, given that polling booths are not static in their voter base; demographic trends (in particular the greening of the inner cities) should lead to increases in the Green vote in some but not others, and these trends should then be encountered in the 2PP vote as well (potentially producing a correlation for reasons unrelated to the pulp mill), but even that does not show up across the seat as a whole. Possibly factors in the opposite direction, or good old-fashioned sample size noise, mask a swing against Ferguson caused by the pulp mill, but if so, it would have to be a tiny swing in the first place.
Looking at where Ferguson did lose the election, the following were the booths delivering the biggest swings against him:
Winnaleah 11.7, St Leonards 8.6, S Launceston 8.1, Newnham 8.1, Ringarooma 7.6, Summerhill 7.6, Prospect 7.2, Karoola* 6.9, Fiveways 6.3, Riverside West 5.9, Trevallyn 5.3, Hillwood* 5.3
Ignoring outliers like Winnaleah (all 207 formal votes of it), we can see that Michael Ferguson was mainly thumped not in the impact zone of the pulp mill, nor in the timber towns, but in the suburbs of Launceston (and not the very poor ones where he never got many votes anyway, but those where he had polled competitively to well in 2004.) The list includes some Launceston booths that had high anti-mill votes (those that also appear on the list above) but also includes booths like St Leonards, Prospect and Newnham where the AMI was below the seat average (the 3.7 point AMI at Newnham being the third lowest in the whole electorate, ahead of only Scottsdale at 2.6 and Myrtle Park, the only booth with a swing against the Greens in the whole of Bass, at minus 2.2).
The conclusion I draw is that Michael Ferguson’s demise had very little to do with either the pulp mill or the forestry debate, but rather that he was primarily a victim of the same low-to-middle-income economic concerns that made up a large part of his party’s nationwide defeat. If there was a pulp mill factor in his defeat at all, I believe that it was smaller than the margin by which he lost.
Braddon
The following were the biggest-swinging booths for Labor on a 2PP basis:
Sulphur Creek 14.7, Calder 12.6, Marrawah 11.8, Miandetta 11.2, Irishtown 10.8, Stowport 10.6, Smithton 10.1, Forest 9.5, Highclere 9.1, Brooklyn 8.2, Havenview 7.9. Boat Harbour 7.5.
Eight of those booths had swung by between nine and 15% to the Liberals in 2004, well exceeding the seat average. Furthermore, if one considers the twenty booths that swung most to the Liberals in 2004 (headed by Togari 19.4, Heybridge 16.7, Highclere and Irishtown 15, Smithton 14.8 and Forest 13.8), seventeen of them swung back to Labor by more than the seat average of a modest 2.57%, eleven swinging back by more than double that.
Explaining the result in Braddon is a relative no-brainer. Just as Labor lost Braddon entirely because of the Latham forestry policy disaster of 2004, Labor regained Braddon by ditching that policy in 2007.
The swing to the Greens in Braddon was 2.53 points in the House of Reps and 2.29 points in the Senate. The latter was the lowest Senate swing in Tasmania by a considerable margin, and the former would have been but for Ben Quin competing for the anti-pulp-mill vote in Lyons. Most likely, this swing had more to do with the similarity of the two major parties on forestry policy than the pulp mill, and non-forestry issues like climate change (a major driver of Green voting at this election among first-time voters) played a role. Indeed, I feel quite safe in stating categorically that the Braddon 2PP result had nothing at all to do with the mill and that all spinning of the Braddon result as pulp-mill-driven was nonsense.
Lyons
As I predicted Dick Adams was very easily returned in Lyons. Also as I predicted he appears to have obtained the lion’s (pun unintended) share of Green preferences. Although the full distribution of preferences is not yet available, we do know that 24.31% of the vote did not go to either major party, and of this, Adams obtained 15.63 points, or 64.3% of the preferences. This is despite around 85% of these preferences coming either from the Greens or Ben Quin. Indeed, even if every CEC or Family First voter has preferenced Adams, we still know that Adams has over 58% of the combined preferences of Ben Quin and the Greens, and it would be very surprising if he did much better on the preferences of a renegade Liberal than a Green. This all flies in the face of many claims made by posters on this site that anti-pulp-mill voters in Lyons would not preference Dick Adams. They did, and it is no surprise because the Liberals were also pro-mill, so much so that they would not permit their original candidate to run on an anti-mill platform. Furthermore, a single issue will never completely annul Green voters’ very strong tendencies to preference Labor ahead of the Coalition.
Teasing out the impact of the pulp mill on the Lyons vote is made more difficult by Ben Quin’s candidacy. The Senate vote probably gives us a better idea of the real impact of the issue across Lyons. The swing to the Greens in the Senate for Lyons was 4.6 points. This compares with 5 points in Franklin (swing in Reps 3.3%), 5.9 points in Denison (Reps 4%), 6.7 points in Bass (Reps 7.2%) and 2.3% in Braddon (Reps 2.5%). On average the swing to the Greens in the Senate was just over one percent higher than in the Reps, and this is consistent with the national average (and no surprise, since some voters only vote for minor parties in the house where they have a chance of election). On that basis, had Ben Quin not stood, the swing to the Greens in Lyons would probably have been around the same as the swings to the Greens in Denison and Franklin – say, 3.5%. I haven’t crunched the Senate swings by booth but I suspect this shows a strong pulp mill factor in and around the West Tamar primarily, and no real pulp mill factor across much of the electorate. If this is correct, Ben Quin took around three-quarters of his votes from the major parties and only around one quarter from the Greens. Because the swing to “anti-mill candidates” in the Reps was much stronger than that in the Senate, the swing to Quin was not primarily an out-and-out anti-mill vote. Rather, it is more likely to mainly reflect either a personal vote, or the votes of voters who were opposed to the pulp mill but even more strongly opposed to the Greens.
Some strange things happened with the Quin vote. Peter Henning mentioned that Ben Quin won two booths in the West Tamar, these being Gravelly Beach and Legana, although, as I pointed out in the comments section, Henning’s conclusion that Ben Quin is the hypothetical member for West Tamar is most likely incorrect. What is not so well known is that Quin also won two other booths – you would not guess them – Chudleigh and Rosebery. In Rosebery there was a massive 2PP swing against Dick Adams, despite Queenstown (just down the road) returning a massive swing to him. I am unaware of the reason for this.
If we add the Quin vote to the swing to the Greens and call this the AMI for this seat (despite the reservations stated above) then the West Tamar seats (Beaconsfield 16.4, Exeter 22.4, Frankford 18.4, Glengarry 17.5, Gravelly Beach 32.1, Kelso 18.8, Legana 18.3 and Sidmouth 33) are all predictably at or near the pointy end. Other booths with a high AMI are Beauty Point 20.7, Carrick 11.9, Chudleigh 31.1, Claude Road 12, Deloraine 14.2, Elizabeth Town 13.6, Orford 13.9, Rosebery 31.5, Tullah 15, Westbury 13.2. Orford is near Quin’s home base and the Rosebery anomaly (which spills over into Tullah) has already been mentioned – the other thing we can see from this list is a mild inland central north factor. The latter is also present in the Greens’ Senate results, for instance swings of 6.6% to the Greens at Chudleigh and 8% at Claude Road where the Green Senate vote was 27.7%. There is possibly some forestry dimension to this central north factor then, but I would not assume it was necessarily the pulp mill.
Once again, what we really need to know is whether the Quin/Green vote contributed anything to the 2PP result. Considering the eight West Tamar seats mentioned above, not one exceeded the seat-wide swing to Labor, and one (Sidmouth) returned a 2PP swing to Geoff Page, one of just twelve booths to do so. Furthermore, of the other booths with a high Ben Quin vote, some (eg Chudleigh, Claude Road, Deloraine) swung strongly to Dick Adams, while others (Beauty Point, Carrick, Tullah) swung to Page.
Some might suggest from the West Tamar vote that the small swing to Dick Adams suggests an anti-pulp-mill backlash harmed his 2PP results in that area. This would, however, be a fallacy, because what is really going on is that Adams picked up a larger swing elsewhere in Lyons for reasons that do not apply in the West Tamar.
The following is the list of booths that swung to Adams by 9% or more: Buckland 22.3, Tunnack 21.4, Tunbridge 19.5, Bothwell 19.1, Triabunna 16.2, Port Arthur 14.1, Brighton 12, Queenstown 11.5, Campania 11.3, New Norfolk 11.1, Magra 10.8, Forcett 10.1, Mole Creek 10.1, Taranna 10, Maydena 9.8, Pyengana 9.6, New Norfolk North 9.5, Glenora 9.4
Timber towns overwhelmingly dominate the list, and a look at the 2004 figures shows a similar pattern to Braddon. Two-thirds of these booths had swung to the Liberals by more than 7% in 2004. Furthermore, almost 80% of the booths that swung to the Liberals that heavily in 2004, returned to Labor by above-average margins in 2007. I have mentioned previously that Dick Adams’ rejection of his party’s forestry policy at the 2004 poll probably had the effect of dampening the swing against him in some booths. At this election, with that policy gone, the many timber booths where he did not manage to dampen the swing last time have generally returned to him. The same swing is not seen in the West Tamar booths generally, not because Dick Adams is on the nose over the pulp mill in those booths, but because those booths are not timber communities and hence are not affected so much by the Latham forestry policy reversal. Again, if there is evidence that the pulp mill had any effect on the 2PP result in Lyons, it is hiding.
Denison
Probably the most interesting thing about Denison is the size of the swing to the Greens, the second-highest in the state in both houses. This swing was very strongly concentrated in the inner suburbs and the known green-leaning satellite suburbs of Fern Tree and Taroona – in short, the areas that were already very Green became more so, and may well continue to do so. It was less apparent in traditional Labor areas in the northern suburbs and the Liberal enclaves around Sandy Bay. The pulp mill, climate change or demographic factors? It’s difficult to say. All of the above?
Franklin
Franklin also displayed a big swing to the Greens, and this was concentrated in areas where they were already doing well. Booths with swings in the range of five to nine percent to the Greens included Sandford, Opossum Bay, Kingston Beach, Cygnet, Franklin, Glen Huon, Grove, Howden, Middleton and Snug. Most of these swung to the Greens in 2004 as well, and as in 2004 I suspect a demographic basis, in particular the ongoing intake of sea-changers.
Franklin delivered the largest swing to the Liberals of any seat in the nation. While Kevin Harkins was Labor’s candidate this was hardly an implausible outcome, but once they switched to Julie Collins, polls forecast a swing to the ALP. The combined 400-vote EMRS sample was out by an amount that exceeded its margin of error, which is unusual given its proximity to the election. So what happened?
I do believe Vanessa Goodwin was a good candidate for the Liberals, but I also expected that Labor’s selection of Julie Collins should have more or less neutralised her impact, and under normal circumstances Labor would get a swing of a few percent in line with the rest of the state. For one suggestion as to what happened, I am very grateful to poster Scotty from The Poll Bludger, who suggested the real cause of the swing to the Liberals was “a cynical demand for pork” and that southern Tasmanians “resent all the goodies” the northern marginal seats receive. Another poster, Tassieannie, backed this up by observing that a Liberal HTV card distributor in Warrane had said “make Franklin a marginal seat” while handing out his cards.
Swings over two points to Labor were recorded at Glen Huon (12 points), Franklin (6), Geeveston (4.3), Judbury (11), Opossum Bay (4) and swings of one to two points at Howrah, Huonville, Kettering, Port Huon and Seven Mile Beach. There is probably a demographic component in some of these (Opossum Bay and Kettering especially) but the major factor in the few booth swings to Labor that did occur was our familiar friend, the reversal of the Latham forests policy. The southern forests timber booths swung heavily against Labor almost without exception in 2004, and at this election, despite the swing to the Liberals overall, they generally swung back. Without that, Franklin would now be even more marginal.
These were the booths with the biggest swings to the Liberals in Franklin:
Adventure Bay 12.5%, Southport 8.7, Barnes Bay 8.4, Sandford 7.9, Bridgewater 7, Snug 5.9, Margate 5.5, Clarendon Vale 5, Geilston Bay 4.9, Mornington 4.9, Kingston Beach 4.6, Rokeby 4.6, Montagu Bay 4.4, Howden 4.3, Lauderdale 4.3
There’s quite a lot going on here. Around Kingston and Margate demographic change runs in a pro-Liberal direction through the evangelical Christian population (even if Vanessa Goodwin’s lack of children offends the family values of some of the dimmer minions of Eric Abetz), and I’m not sure exactly what the story is in Sandford or on Bruny Island. What’s notable is how many of these booths are the kinds of low-income high-unemployment suburbs where Labor should not have gone backwards. Perhaps some of this reflects Vanessa Goodwin’s local policing background (with crime a major issue in some of these suburbs), but it is astonishing all the same that Bridgewater can swing by seven points in one direction while just up the road (but in a different electorate) Brighton goes twelve points in the other.
State implications
Something is making the Green vote in many parts of Tasmania rise. We have seen this not only in this federal election, but also the recent Council polls. Whether it’s the pulp mill or not, there’s no reason to believe it’s only interested in federal elections. It may have gone away by 2010 when the next State election will fall due, but what will happen if it hasn’t?
To model this, I have used the swings to the Greens from the House of Reps seats at this election, and assumed they are 2PP neutral (which means that as primaries they come mainly at the expense of Labor, but as preferences they distribute as per normal for a state election). I have used a swing to the Greens of 3.5% for Lyons as the likely swing had Ben Quin not stood, and I have also looked at two scenarios, one in which there is no 2PP swing between the major parties, and one in which there is a 3% 2PP swing from Labor to Liberal.
In Bass, a 7% swing to the Greens leaves state figures of roughly 44-32-21 (ALP-Liberal-Green respectively) in the first scenario and 41-35-21 in the second. In either case the seat splits 2-2-1, which is no surprise as Labor almost got an extra seat at the Greens’ expense last time.
In Lyons, a 3.5% swing to the Greens results in roughly 49-29-19 in the first case and 46-32-19 in the second. The first produces no change while the second would very likely result in Labor losing a seat to the Liberals.
In Braddon, a 2.5% swing creates around 49-36.5-13 in the first case or 46-39.5-13 in the second. The first again produces no change while the second places Labor in a close fight for the final seat with the Greens.
In Denison, a 4% swing throws up 44-25-28 or 41-28-28. The first places Labor in a close fight for the final seat with the Greens, but the latter loses Labor’s third seat and the Liberals and Greens fight for it.
Lastly, in Franklin, a 3% swing produces 45-30-23 or 42-33-23. In the former case, Labor retains its current position, while in the second their third seat is taken by the Liberals.
Adding all this up, if the swings to the Greens seen in the Federal election are repeated in the next State election with no swing between the major parties, then the Lennon government will retain majority status with at worst the loss of one seat. But if the same swings to the Greens are accompanied by moderate swings to the Liberals, then Labor’s majority will be gone – the party loses two seats and may lose another two. Interesting times ahead … but then again, they said that last time, didn’t they?
At the most recent federal poll, Kevin Bonham voted for Duncan Kerr with extreme reluctance and apathy. He would have voted informal but wished to take some token and meaningless part in sending Howard packing, and also to place the Family First party at the bottom of his ballot. In the Senate he voted, quite enthusiastically, below-the-line for the Secular Party, which finished last.
