Category Archives: Voters


Over the weekend I was asked by the Election Watch team at the University of Melbourne to comment on the polling conducted for the Brexit referendum.

The article that came out of this can be found somewhere on their blog, I’m sure. Here are the comments I provided in full (questions are bolded):

1. What do the Brexit polls tell us about the accuracy of polling generally?

Not a lot (about polling for elections in Australia, at least).

There were two complications pollsters trying to understand public attitudes towards Brexit that are generally not faced by those trying to understand vote intention in Australia.

First, voting in Britain is not compulsory, as it is in Australia. Therefore, not only do organisations conducting public opinion surveys have to try and get a representative sample of voters, they need to work out what a representative sample is. This can be quite difficult to do, as turnout can be quite volatile.

In Australia, the Australian Electoral Commission collects data on the age and gender breakdown of everyone on the electoral rolls of each division. Since approximately 95 per cent of voters actually vote, pollsters in Australia just need to make sure their sample of respondents matches the electoral roll and they can be fairly confident it will be close to the makeup of the electorate that actually turns out to vote (this is harder than it sounds, but much easier than the context faced by those operating in Britain).

The difficulties encountered by pollsters in a country with voluntary voting are compounded in a referendum. With elections we have a number of reference points to work from (past elections). Pollsters know which groups of voters are more likely to turn out and vote in a generic election, and even roughly how those groups tend to vote.

Britain has never conducted a referendum on leaving the EU before, so knowing which groups — rural or urban, university graduates or those that had not finished high school, or the young and the old — would turn out and vote, is difficult. Nor have they had a chance to check their result with actual outcomes, as is the case with elections.

2. What do they tell us about voter volatility?

Public opinion on Brexit, at least as it was measured by surveys, appeared to move around a lot, with a fair amount of volatility. This may be the result of a lack of fixed opinion on the question. Topics like health and education spending, and taxes, for instance, tend to be real and immediate concerns for people. As a result they tend to have pretty fixed opinions on these kinds of policy areas, and their opinions are more difficult to change. Brexit was probably a little more abstract, and as a result a strong argument either way could have probably changed some voters’ position

3. Do the Brexit polls, or voting patterns, tell us anything useful about how voters in Australia can be expected to behave?

Close to the vote the polls indicated the result would be roughly 50 per cent for both leave and stay, with a slightly larger number voting stay than leave. However, several per cent were undecided, and it appears most of these voters broke towards leave at the last minute. This to me suggests the polls did reasonably well under difficult conditions, but in a close race where we have little relevant comparisons with which to calibrate our data, it can be difficult to know what the eventual outcome will be. We could expect that polling on similar issues here would be comparably difficult (with the benefit of compulsory voting though).

4. Any other observations about what it could mean for Australia?

We saw a similar process at work during the republican referendum in 1999. For years surveys had said a majority of Australians were in favour of a republic, but then they voted to stick with the monarchy. A possible reason may have been the issue was relatively abstract for most people, so they could be swayed either way. The pro-monarchy campaign in that case was able to argue the republican model being voted on (with a president elected by the parliament) was not a good outcome. They were successful. We could imagine similar tactics could be used in future referenda here, for instance the same-sex marriage plebiscite, or for Indigenous recognition, with those without strong opinions either way swayed by a convincing argument for one side or the other.

As a follow up I was also asked The vote was 48-52. If I it had been that result in favour of Stay, it would be accurate to say it was close. However, referenda (in Australia anyway) almost always fail – that is, voters decide to stick with the status quo. Do you therefore think it’s fair to say that actually the Stay vote suffered a huge defeat, and that in that sense, since the polls predicted a win for Stay, they’ve really got it wrong.

I would characterise the outcome as reasonably close. If two per cent of voters who chose leave had chosen stay instead, the latter would have won. The polls in aggregate generally had leave and stay neck and neck all month (with leave ahead for most of June). Right before the vote the polling aggregators generally had leave at 44 per cent and stay at 46 per cent, and around 10 per cent undecided. While many commentators may have concluded this meant stay was ahead, I would never have characterised the polls as saying this. The polls suggested it was too close to call (with perhaps a slightly higher chance of stay winning).

An interesting anecdote to help make my case. Apparently there were quite severe storms and flooding across much of the south, the Midlands and the east, with London and some other pro-stay areas hit at peak voting times. Some polling stations were flooded and closed for hours as a result. This may have hampered turnout in these areas and, with two percentage points in the result, changed the outcome. I’m not saying it did change the outcome (I don’t know who chose not to vote as a result, but it may have been young people living in urban areas without a car who were more likely to vote stay and may have been more likely not to brave the storm as a result) but it was close enough that it could have.

Additionally, after I commented to Election Watch I found this blog post on the same topic by Andrew Gelman. He probably did a better job on this than I did (no surprise).

Are there political differences between Sydney and Melbourne?


Even a casual observer of Australian politics will likely know that the Coalition tends to do better in some areas, and Labor in others. However, some important patterns are missed, either because geographic variation is often examined at the electorate level (sometimes obscuring important variation within divisions, which can be relatively large sptial areas) or because the mapping is not done as well as it could be.

For instance, this Australian Electoral Commission map leaves something to be desired. In particular, I don’t like how they divide up the Coalition parties by different colours (making the patterns between Coalition and Labor areas harder to identify), and that these are not shaded by the level of support, so a safe Labor seat looks the same as a marginal one. Combined with the focus on the division rather then smaller units, these decisions reduce the amount of information we can gain from the map.

Interested in how vote choice maps across different parts of Australia at a finer-grained level than the electoral division, I grabbed the two-candidate (Coalition v Labor) results from every polling place in the country from the 2013 federal election. These polling places were then spatially placed at the local neighbourhood level,* and mapped used ggmap() in R.

Here, for instance, are the results for the Sydney (left) and Melbourne (right) metropolitan areas. They are shaded so that an area shaded blue had a Coalition majority (the darker the blue the higher the Coalition vote) and red a Labor majority.

Two-party vote by neighbourhood in the Sydney (left) and Melbourne (right) metropolitan areas. A blue area is one in which the Coalition won a majority. The darker blue a neighbourhood is shaded, the higher the Coalition vote. A red area is a Labor-majority area. Neighbourhoods are defined here as ABS Statistical Area 2 regions.

These maps make it clear there are important differences within and between the two cities. In particular, the Coalition appears to do better in Sydney and Labor in Melbourne.

In Sydney, the Coalition enjoys large majorities across much of the north, the inner-east, and the fringes of the south and south-west. The Labor vote is largely confined to concentrated regions in the inner west around Newtown and outer west around Liverpool, Cabramatta and (to a lesser extent) Mt Druitt. It’s not as bad as it looks for Labor, though, as their strongest areas of support tend to have higher population densities than many of the Coalition–supporting neighbourhoods.

Labor appears to be more successful in Melbourne, winning large majorities throughout much of the north, west and outer-east. There are Coalition majorities throughout large parts of the east, but these majorities are much slimmer than those enjoyed by Labor.

What drives these differences in vote choice? I suspect it’s likely a mix of economic and social differences. For instance, as the plots below show, neighbourhoods with older average residents and higher income households tend to have higher Coalition votes. This is particularly the case in suburban and urban areas, where neighbourhoods with median ages closer to 50 tend to have ~20 per cent higher Coalition votes than those closer to 30; and those with median incomes around $150,000 per year tend to have Coalition votes 30-40 per cent higher than those with median incomes around $50,000.

Of course, you cannot be certain that these are the causes of different electoral outcomes (we might be making an ecological fallacy), but a quick check of the Australia Election Study data makes it pretty clear these socioeconomic differences are fairly important to vote choice (maybe more on this is another post).

Coalition two-party vote by neighbourhood location and median age of residents. Each open circle representative a neighbourhoods, defined here as state suburbs/localities. These are scaled by the resident population. The curve is a LOESS regression showing trends in the relationship between median age and party support.



Coalition two-party vote by neighbourhood location and median household income. Each open circle representative a neighbourhoods, defined here as state suburbs/localities. These are scaled by the resident population. The curve is a LOESS regression showing trends in the relationship between median income and party support.


*More specifically, the ABS Statistically Area 2 level, which are areas of approximately 10,000 people. There are several thousand of these neighbourhood areas across the country.


Albrechtsen misses part of the story on the difficulties of reform

In a recent article in the Australian ($),*  columnist Janet Albrechtsen claims the primary reason the current federal government is having trouble getting it’s economic reforms through Parliament is that people are not willing to make the needed sacrifices required to balance the budget and make government spending sustainable.

Although I don’t disagree with everything in her article, I think this is being a little harsh on the Australian people.

Bernard Keane had a much more balanced piece on the difficulty government’s face implementing a reform agenda in Crikey.  He made the very good point that one factor that can make reforms difficult to pass is the perception of fairness. Will this reform help the majority of the population, or will most be forced to pay for the benefit of a few?

And this is one of the main problems the federal government faces in having its reform agenda passed. It is viewed as unfair by many, and an objective analysis of the outcomes of the (largely unpassed) 2014-15 federal budget indicate its impact would fall hardest or low and middle income households (hard to sell that as fair).**

At a time when inequality of wealth, income and opportunities is increasingly a public concern, a government budget that gives all appearances of increasing these inequalities was going to be a difficult sell, even if it also provided greater budget sustainability.

I didn’t see any of this mentioned in Albrechtsen’s piece. It would have been far more well-rounded if it had discussed the difficulties of getting hard reforms through parliament (and elections), balanced with the need for those reforms to be balanced and of benefit to the majority of the population.

* Although paywalled, if you google ‘We, the people, are the threat to fiscal reform’ you may be able to access it.

** Phillips, B. 2014. National and Regional Analysis of the 2014-15 Federal Budget. NATSEM, University of Canberra.

^ An earlier version of this post listed the last federal budget as the 2015-16 Budget. This was incorrect and has been amended.

The only way to save the democracy is to destroy it

Or so says this guy in the Herald Sun…

Now I don’t know who he is or what his deal in, but  (big surprise) I disagree.

The problem? For starters, who decides what a good outcome is? It could be economic growth, it could be environmental protection, or greater income equality, or more opportunities for the disadvantaged, or it could be more substantial rewards for personal effort.

These are all reasonable debates. Disagreements our democracy allows, and that are be worked out right now through our state and federal parliaments. The committee whatshisname is suggesting would stifle that debate and attempt (unsuccessfully I believe) to find some kind of pseudo-objective consensus where none exists.

Would the committee (lets call it the Committee of Public Safety,* for arguments sake) be rewarded for building a new freeway, or punished for not reducing car dependency and increasing public transport patronage?

Would our Comité de salut public (I’ve made it French for extra flair**), which increases the use of renewable energy, be lauded, or punished, if this meant slightly higher power prices? And if they are to be given incentives for good decisions (how we objectively decide these good decisions, I do not know) who decides how they get the incentives? And how do we ensure these incentives encourage better decision making of currently Parliamentarians?

Now, I know I’m banging on about the same issue about (two of the last seven posts have been on some derivation of this issue). However, one upside is it gives me a chance to make some sweet graphs. For instance, the figure below shows voters attitudes towards immigration (as measured in the Australian Election Study) conditional on their occupation.

Why might we care about this? Because how people view policy outcomes (for instance, whether more immigration is good or bad) will depend on their life experience. As we can see below, voters with professional and managerial occupations are more likely to have positive views on immigration. From 1990 to 2010, 20% of professionals  and 15% of managers said the number of migrant arrivals had not gone far enough, compared with 11% of voters in general, 10% of those without employment and 8% of those in manual occupations. Conversely, 34% of professionals and 41% of managers said immigration to Australia had gone too far over this period, compared with 49% of the electorate in general, 54% of those without employment and 62% of those in manual occupations.


The relationship between attitudes towards immigration and occupation, 1990-2010. Each plot shows the proportion of survey respondents in six occupational categories who responded to the question “The statements below indicate some of the changes that have been happening in Australia over the years. For each one, please say whether you think the change has gone too far, not gone far enough, or is it about right? The number of migrants allowed into Australia at the present time”.

The relationship between attitudes towards immigration and occupation, 1990-2010. Each plot shows the proportion of survey respondents in six occupational categories who responded to the question “The statements below indicate some of the changes that have been happening in Australia over the years. For each one, please say whether you think the change has gone too far, not gone far enough, or is it about right? The number of migrants allowed into Australia at the present time”.


Why might this be? There is some evidence the benefits of immigration may flow disproportionately to the owners of capital and higher income households. This potential distribution of the benefits of immigration reveals itself in public opinion, with voters in professional and managerial occupations (generally higher income jobs) tending to hold more positive attitudes towards immigration.

A technocrat could look at the objective evidence, but on many policy issues we deal with in politics – from immigration to redistribution to something else – the data is open to interpretation and objectivity is thin on the ground. So while immigration may boost GDP growth, for instance, that boost in growth may mostly accrue to the immigrants themselves, and certain segments of the population, and have more limited benefits for other groups. The research is not conclusive on that, but that’s the whole point: you’re not necessarily going to find some sweet spot in the middle everyone agrees on. That’s why we have politics and parliament and elections; so the contentious issues that divide us can be thrashed out. We might not all like the outcomes of this contest all the time, but it’s better than someone deciding outcomes without the accountability that comes with elections.

*Prize for the person who gets the reference.

** A second hint. Come on, it’s easy.

Measuring Australian voters’ ideology over 44 years

What do we want to do if we wish to measure voters’ ideology (or issue preferences, for those who don’t like that word)?

Unfortunately, here in Australia the we don’t really have survey data available which asks a large number of consistent questions over time, which would allow us to compare voters of, say, the 1960s, with the voters of today.

Additionally, there’s also the question of how useful is survey data for measuring voter attitudes? Based on his analysis of survey repsonses, Philip Converse didn’t think most voters’ had a comprehensive ideological framework within which they organised their political preferences; although later research has indicated he might have been a little harsh, with measurement error obscuring the real political preferences of the voters he was examining.

One way of dealing with the lack of consistent questions over time and issues of measurement error, is to model the latent ideological traits using graded Item Response Theory (IRT) models.

Although much of the original work on IRT estimation was conducted by education scholars to measure latent traits such as intelligence or personality, more recently these methods have been developed to measure the latent ideological attitudes of political actors, such as justices of the United States Supreme Court (Bafumi et al. 2005) and voters (Treier and Jackman 2002, Treier and Hillygus 2009). The model I look at here was fit to 34 issue items from a combined file of all 12 Australian National Political Attitudes Survey (ANPAS( and Australia Election Study AES) surveys using the mirt package (Chalmers 2012) in R (R Core Team 2013). These surveys provide information on the policy preferences, voting behaviour and demographic backgrounds of 24,393 respondents sampled from 1967 to 2010.

I measure the latent issue preferences of voters on a left-right spectrum (standardised to have a mean of 0 and variance of 1) and two dimensions (effectively an economic and social issue dimension), with the results for the economic dimension shown below.


The first preference vote choice of ANPAS and AES survey respondents and their economic issue preferences, 1967-2010. Each party is measured by the difference between the issue preference of their median voter, and the electorate’s median voter.


Each plot shows the economic issue preferences of the median voter of  five of the more significant parties to contest Australian federal elections over the 44 years from 1967-2010, compared to the actual median voter for each survey. The higher scored preference, the further to the right each party’s median voter is. I’ve plotted a linear trend line for each party to make patterns clear.

The results are not particularly surprising. Greens and Labor voters are left of centre. Coalition voters are right of centre, with One Nation slightly more centrist on economic issues. What is interesting is that the Democrats start at the centre (not surprising) and then shift well to the left over a 20 year period. Also interesting, the median Coalition and Labor voter has moved to the right compared to the electorate median over this time (likely as both lost their more leftwing voters to the Democrats and then Greens).