There has been growing controversy about Australia’s level of overseas immigration. In the year to March 2017 Australia’s population is estimated to have grown by a massive 389,100, some 231,000, or 60 per cent of which was due to net overseas migration. For the last few years around two thirds of the net growth in migrants have been locating in Sydney and Melbourne.
The consequences are becoming obvious and are being reflected in increased public concern about urban congestion and other quality of life issues.
But are these consequences resulting in increased opposition to high migration? In order to explore this issue The Australian Population Research Institute (TAPRI) commissioned a national online survey of Australian voters in August 2017 ( ‘Australian voters’ views on immigration policy’.)
The survey found that 74 per cent of voters thought that Australia does not need more people. Furthermore, 54 per cent wanted a reduction in the migrant intake. TAPRI also found that big majorities think that population growth is putting ‘a lot of pressure’ on hospitals, roads, affordable housing and jobs. Thus it seemed reasonable to conclude that these concerns were manifesting in concern about migration levels.
This conclusion has been challenged by the 2017 Mapping Social Cohesion report from the Scanlon Foundation, which surveyed Australians at about the same time as the TAPRI survey. Scanlon reports that only 37 per cent thought immigration levels should be reduced (up just three percentage points from 2016).
When interviewed by David Marr on these findings, the author of the report, Professor Andrew Markus, said: ‘On one level, we’re doing really well as a society… There are all these stories about overcrowding, public transport, housing and everything. That could have gone negative on immigration and so on, but it hasn’t’ .
Who is right? The answer is of major consequence for Australia’s political class. In Western Europe concern about immigration levels has manifested in anti-migration parties gaining 15-20 per cent of the total vote. Is Australia an outlier, immune to sentiment of this sort?
Australian political elites appear to believe that they have little to fear on this front. This is because their main source of information about public opinion on the issue has been the Scanlon Foundation.
For instance, Labor’s shadow Deputy Treasurer, Andrew Leigh, has recently asserted that Australian attitudes to migrants are warm and ‘becoming warmer over time’ (Choosing Openness). According to David Marr, ‘more than almost any people on earth, we are happy for migrants to come in big numbers’ (The White Queen <>). Both sources draw these conclusions from successive Scanlon reports.
This remarkable outcome, at least by comparison with the anti-immigration protests across Europe, has prompted a special report in The Economist magazine. The report notes recent efforts, as by Dick Smith, to sound the alarm about Australian migration levels. Yet, so the magazine judges, relatively few Australians seem to be concerned. The authors’ main source, once again, is the Scanlon Foundation. The Economist states that:
Regular surveys conducted by the Scanlon Foundation, which works to integrate immigrants, show that the sense that immigration is too high has fallen substantially since the 1990s.
Why the difference in results?
The TAPRI survey was completed online by a random sample of 2057 voters, (with quotas set with a 10% leeway, in line with ABS distributions for age, gender and location). The sample was drawn from a panel of 300,000. Thus TAPRI used the same methodology as is now employed by Newspoll and by Essential Media.
It is true that, despite the demographic weighting, the panels in question may not be representative of the overall population of voters. For example, the TAPRI sample had a higher representation of graduates than that of the voting population as a whole. This means it probably underestimated the opposition to migration since we found that only 41 per cent of the graduates amongst our respondents favoured a reduction in migration levels, compared with 61 per cent of non-graduates.
However, there are at least as many problems with probability samples done by telephone. The Scanlon poll was based on a telephone sample of 1,500 Australian residents drawn from the entire population of residents. It therefore included many respondents who are not citizens and therefore not eligible to vote.
Citizenship requires a four-year stay in Australia, at least one year of which must be as a permanent resident and, of course, the desire to make the application. As the TAPRI survey conformed, Australian-born persons are much more likely to take a tough line on immigration numbers than are overseas-born persons (unless they are UK-born).
There are also significant issues concerning the reliability of telephone interviews when probing sensitive issues. As the highly credible Pew Research polling organisation has indicated, respondents may be more likely to provide socially undesirable responses in the relative anonymity of the internet.
Research by Scanlon supports this point. The 2017 report got quite different answers to the question ‘Is your personal attitude positive, negative or neutral towards Muslims?’ when the question was asked in its telephone survey and when asked in a separate online survey that Scanlon funded. In the telephone survey 25 per cent said ‘negative or very negative’, while 41 per cent responded this way in the online survey .
Similarly, Scanlon found a much larger share of respondents favoured a reduction in immigration numbers in a different online survey that it funded which used methodology similar to that used by TAPRI. In the telephone survey 37 per cent said that immigration was too high. In contrast, 50 of this online sample agreed that the immigration intake was too high, rising to 53 per cent when the findings were limited to those who were Australian citizens.
This result is almost identical to the TAPRI finding. It may well be that because attitudes to immigration numbers, like attitudes to Muslims, are sensitive, voters responding online feel freer to express negative opinions.
If TAPRI’s and Scanlon’s findings when using panel methodology are reliable they have great political significance. The TAPRI report found that 57 per cent of Liberal voters and 46 per cent of Labor voters thought that the immigration intake should be reduced.
If the immigration issue were to be contested at the next federal election both parties would be vulnerable. One Nation or any other party with a fierce low-migration agenda could draw voters from both the Liberal and Labor parties. Alternatively, should the Liberal party stake out a low migration agenda, it could draw votes from the Labor Party.
Katharine Betts and Bob Birrell are with the Australian Population Research Institute, a non-profit think tank.