Marlboro Man or Winnie Blue?

The new political debate over tobacco taxes has got me wondering: are smokers more likely to vote for parties of the right (because they believe in individual liberty) or parties of the left (because they tend to be poorer than non-smokers)?

The problem is, I can’t find a dataset that contains information on voting and smoking. Any suggestions? (You can rule out HILDA and ASSA, btw.)

Update: Regular reader Thinking in old ways finds some evidence in the NESSTAR data archives.

The June 1994 Saulwick poll reports that the smoking rates were:

27% labor
17% liberal
24% national
12% democrat
25% independents
29% other parties

So at least on these data, it looks like smokers are more likely to support Labor than the Coalition.

This is also consistent with Don Arthur’s findings from a survey in Budbury, WA.

This entry was posted in Uncategorized. Bookmark the permalink.

7 Responses to Marlboro Man or Winnie Blue?

  1. David says:

    It is an interesting question Andrew. In the absence of any data why don’t you start a ‘book’. Based on the revealed odds you would have your answer.

  2. Dominic says:

    You could look for some second stage correlations, like the prevelance of smoking in various electorates. Control for other common vote determinants like education, age, income, gender, minority etc, then see if per cent of voting age population who are smokers has any predictive power for the outcome of elections.

    The advantage obviously is that you don’t need smoking and voting in the same dataset.

    You would need prevalance of smoking by post code, and I’ve no idea if such a thing exists having never done anything with public health data.

    If that isn’t available, you might try lung cancer deaths as a proxy for smoking. The data should be more meticulous and possibly contain associated post codes.

    If that still doesn’t work, try asking online opinion to include a question on smoking in their next survey.

    I don’t know if I would accept your explanation of smokers voting conservative. Do people really smoke as an expression of individual liberty?

    Good luck

  3. Don Arthur says:

    I’m not aware of any national data set. But there is a small-scale study from Budbury in WA that reported:
    “Interestingly Liberal Party and Green Party supporters were much more likely to be in favour of a total ban [of smoking in hotels, vars and nightclubs] than Labor Party supporters. These differences reflect smoking status rather than political affiliation (i.e. more Labor supporters than Liberal and Green supporters are smokers).”

    Click to access 021120.pdf

  4. Dave Bath says:

    You are talking about branding differences. So if you were to get sales figures somehow (probably not available from the supermarket, more likely from owners of smaller shops) broken down by brand, then you might be able to figure something out.

    e.g. “Holiday”, “Peter Jackson” v “Dunhill”, “Benson and Hedges”

    My guess is that Dunhill and B-and-H would sell more in conservative-voting electorates.

    The other thing is where the cigarettes are carried. If most carry them in their shirt sleeves up near the shoulder, you are probably looking at a different type of voter.

    You could probably get similar results from beer brands.

  5. Thinking in old ways says:

    There is some data up on the NESSTAR site however most of it is fairly old.

    The June 1994 Saulwick poll reports that the smoking rates were:

    27% labor
    17% liberal
    24% national
    12% democrat
    25% independents
    29% other parties

    I wouldn’t be surprised if some of the commercial polling organisations may have more up to date data which they have not archived for public access.

    There may also be some other more recent survey in the ASSDA – I only had a quick look

  6. Because of the links between smoking and class, there is no reason to think that the results reported so far would not represent the current pattern. But smoking is one of those issues which has ideologically aware Liberals conflicted, given the extraordinary levels of state harassment of smokers.

  7. Don Arthur says:

    Credit for the Bunbury findings should go to Rob Donovan, Owen Carter and Geoffrey Jalleh.

Comments are closed.