What did Galaxy's poll tell us about freedom of speech and 18C? Not what the IPA said it did

In evidence to the parliamentary inquiry into freedom of speech on Tuesday, the Institute for Public Affairs (IPA) think-tank tendered a statement based on a survey it had commissioned from Galaxy Research. The Australian newspaper covered this polling as a front-page “exclusive”.

The second paragraph in the IPA’s media release claimed – without evidence – that there was set to be much surprise among the media and the political class that 95% of Australians think “free speech matters”.

The release then reported that 48% of people supported removing the words “insult” and “offend” from Section 18C of the Racial Discrimination Act.

The IPA actually wants the whole of 18C removed. But the way forward since then-prime minister Tony Abbott baulked at the gate on the changes proposed by his attorney-general, George Brandis, in 2014 has been this apparently minor surgery to the less serious end of the unlawful quartet (the others being “humiliate” and “intimidate”).

But is the IPA’s statement a fair reading of the Galaxy polling? And was the research fair to start with?

Questions of methodology and polling

According to a Galaxy spokesperson whom I spoke with on Tuesday, no attempt was made to ensure the sample included a representative component of Indigenous and non-Anglo or overseas-born Australians.

Such data was not collected as part of the study as the client (the IPA) had not asked for it, so the results could not be profiled on these criteria. Yet these are the people 18C is mostly designed to protect.

Chances are that an average online panel (the Galaxy polling was done mainly online) won’t include many Indigenous people, people with poor English, or people from minority refugee communities – that is, the primary targets of race hate speech. We had to work hard to ensure our online survey on a similar issue included enough minority-group Australians to ensure statistical accuracy.

The IPA research was two questions in the regular Galaxy omnibus survey, which seeks to control only for age, gender and region. It also looks at shopping patterns.

The first question was:

How important is freedom of speech to you?

This was designed to position the respondent positively to the question and its point of view.

The second question was:

Do you approve or disapprove of the proposal to change the Racial Discrimination Act so that it is no longer unlawful to “offend” or “insult” someone because of their race or ethnicity? It will still be unlawful to “humilitate” or “intimidate” someone because of their race or ethnicity.

This aims to deliver the coup de grâce that reinforces the desired outcome.

So, more than 95% of those polled thought freedom of speech was important. This is a no-brainer. Had the question been – as other surveys have put it – “Is freedom of speech more important than freedom from hate?”, the percentage of those in favour may well have come down significantly. Or if that question were to be reversed, even more so. But we will never know.

Then the question of removing “insult” and “offend” was put. Less than half of any group supported this. Given the preparatory question and the lack of information about the implications or impact, this is less than one might have predicted.

However, neither Galaxy nor the IPA discussed the most interesting data.

Youth responses show IPA conclusions invalid

In the Galaxy poll, the 18-24 age group had the highest commitment to freedom of speech but the lowest support for removing “insult” and “offend” from 18C – by a long way.

So, a suggestion that a commitment to freedom of speech necessarily carries with it support for amending 18C is simply false. There is no simple correlation. They appear to be independent variables, though mediated by some other factor – probably social media use.

There is a much better explanation which neither Galaxy nor the IPA evoked.

The 18-24 age bracket comprises the true digital natives; a very high proportion are regular users of social media. Our research shows they have the highest rate of encounters with racist hate speech. They are usually witnesses, though sometimes are targets. Most encounters with online hate happen on Facebook (40%), YouTube (20%) and in comment threads on news media site (15%).

Digital natives value freedom. But they also want vulnerable people protected and civility enhanced. And they don’t trust sites like Facebook, YouTube or Google to do that – nor, it must be said, government.

In our research, young people were among the least likely to want offending someone on the basis of race to be lawful, just like those surveyed by Galaxy for the IPA. However, they were more likely to hold a neutral position than older people; they were more reluctant to force regulation, but more aware of what racism did to its targets.

The people most in support of retaining 18C in our study but not in Galaxy’s were the older group, who are far less likely to use social media and thus encounter cyber-racism. In our study, the people most likely to want the right to offend people were those who identified themselves as authors of racist material.

So, it follows that the less racism you encounter that you don’t want to see, the less likely it is that you’ll worry about it. The more you want to freely offend people, the more likely it is you author racist material.

Lucky I read the report – or you’d never have known quite what the IPA was selectively trying to slip through to the inquiry and the press. Be sure, though, that the claim most Australians want 18C gutted in the name of freedom of speech simply is not supported by evidence.