Using the ABS to conduct a same-sex marriage poll is legally shaky and lacks legitimacy

The Turnbull government’s decision to put the Australian Bureau of Statistics (ABS) in charge of a voluntary postal plebiscite on same-sex marriage has left many scratching their heads. It was expected the Australian Electoral Commission (AEC) would run the vote, not the body responsible for the five-yearly census.

By giving the job to the ABS, the government has sidestepped questions about its constitutional authority to pay for an AEC-run plebiscite. But it has opened up new avenues of legal challenge and established a process that lacks legitimacy.

Further reading: Explainer: with no free vote for now, where next for marriage equality?

Legal questions

On Wednesday, Treasurer Scott Morrison directed the Australian Statistician to ask the ABS to collect statistical information about the proportion of electors who are for or against the law being changed to allow same-sex couples to marry. This information is to be requested on a voluntary basis.

The same day, independent MP Andrew Wilkie and two others announced they would launch a High Court challenge against the ABS poll. It is likely that at least two grounds of challenge will be put.

The first concerns the power of the ABS to run the plebiscite. Under the Census and Statistics Act, the Australian Statistician can, if directed by the minister, collect “statistical information” on a range of prescribed matters, including “births, deaths, marriages and divorces”, “law”, and “population and the social, economic and demographic characteristics of the population”.

Separately, the Australian Bureau of Statistics Act provides that one of the functions of the ABS is “to collect, compile, analyse and disseminate statistics and related information”.

The key question is whether information about Australians’ opinions on same-sex marriage is “statistical information”.

Surveying people on their views about marriage is very different from collecting factual data about, say, the numbers of marriages taking place in Australia.

And given the postal ballot will be voluntary, the views of some social groups (like those more likely to use postal services) will feature more heavily than others. Arguably, such an “unweighted” data set falls short of the rigorous standards of “statistical information”.

More broadly, if opinions about marriage law are considered “statistical information”, it is hard to think of what sorts of information do not fall into that category. Is it the case that any collection of data is a statistical exercise? If so, the ABS’s powers are very broad.

The second legal question concerns the government’s authority to spend money on an ABS-run postal vote. Finance Minister Mathias Cormann acted quickly to source the A$122 million required from a little-used “advance” appropriated by parliament in this year’s budget.

However, the minister’s advance fund is reserved for “urgent” and “unforeseen” expenditure. It is doubtful that spending on a postal plebiscite falls into either of these categories.

By way of comparison, in 2016 the government gave the AEC $101 million in “advance” funds to help it implement major changes to Senate voting in time for the July federal election.

There is no similar urgency in a vote on same-sex marriage. And, if anything, the need to spend money on same-sex marriage poll was entirely predictable. The government’s budget papers anticipated spending $170 million for this purpose.

The ABS is poorly equipped

Legal questions aside, the ABS is poorly equipped to run a plebiscite on same-sex marriage.

It has significant expertise in collecting factual information on the economy, housing, crime and many other matters that is of immense value to governments and researchers. But it has never run a poll of this kind.

The closest precedent occurred more than 40 years ago. In 1974, the ABS conducted a telephone survey of 60,000 Australians, asking for their preferences on the national anthem. But that is altogether different from the massive logistical exercise of administering a postal ballot for 15 million voters on a contentious social issue.

The stakes are higher and the risk of mistakes is greater. And there will be little tolerance for error.

For the ABS, even the basic task of sending out ballot papers will not be straightforward. Unlike the AEC, it does not have direct access to the electoral roll.

The Commonwealth Electoral Act sets down rules about who can access the electoral roll and for what purpose. Under regulations, the AEC “may” provide the ABS with information on the roll for the purpose of “collecting, compiling, analysing and disseminating statistics and related information”. It would therefore be open to the AEC to refuse the ABS access to the roll, including on the basis that a poll on marriage is not about “collecting statistics”.

Assuming that the ABS gains access to the roll, it is unclear whether it will be able to send ballot papers to all registered voters. The position of silent electors is particularly uncertain.

The addresses of silent electors are not displayed on the roll: to do so would put their safety, or the safety of their family, at risk. Also, the AEC is not permitted to provide information about silent electors to agencies such as the ABS. As a result, silent electors may wonder if they will be able to participate in the poll.

The addresses of eligible overseas voters also do not appear on the roll. However, Cormann has said they will receive ballot papers provided they have “registered as an overseas voter and provided their overseas address”.

More generally, the ABS lacks the AEC’s institutional capacity when it comes to conducting nationwide votes. These are highly complex exercises, which involve distributing, collecting and transporting ballot papers, and then counting them quickly and securely.

Cormann has said AEC officers will be seconded to the ABS to “assist” the process.

Other problems

ABS involvement in the plebiscite raises particular problems, but the shortcomings of a voluntary postal vote go well beyond who is administering it.

Even if turnout is high, we cannot be confident that the result is representative of community opinion. For example:

• Young people move address more frequently than the rest of the population and are less likely to receive the ballot papers.

• Homeless electors and grey nomads may also find it hard to participate.

• Communities with limited access to reliable postal services (including Indigenous people living in more remote parts of Australia) may be disadvantaged.

Also worrying is that the postal vote will take place without the usual protections of election law. Campaigners will be able to circulate unauthorised material – including posters and pamphlets with harmful messages about same-sex couples and their families – without fear of legal consequences.

And, if the result is close, there will be no clear process for resolving claims about the formality of votes and other contentious administration issues.

It should be abandoned

The voluntary postal vote on same-sex marriage should be abandoned. Not only does it rest on shaky legal foundations, it risks damaging the standing of two of our most trusted national institutions.

The absence of standard legal protections is worrying, and the polling method is so flawed that neither side can have confidence in the outcome.

And, at the end of it all, the result is non-binding. The federal parliament is the only institution that can resolve the issue of same-sex marriage definitively. It should do so now, without resorting to such a flawed and expensive venture.