The National leader says census data is so ropey we should halt the redrawing of electorate boundaries. Down that path lies a dangerous world of ‘alternative facts’, writes Andrew Geddis

Let’s be clear – the 2018 census turned out to be a good, old-fashioned omnishambles. The just released report of an independent review into the processes Statistics NZ adopted sets out that conclusion in clear and unambiguous terms. And the fact this report has occasioned the rare but commendably responsible resignation of a government department chief executive underlines the seriousness of the issue.

However, there are degrees of shambolicness – yes, it’s a word – even among omnishambles. So where does the current example sit on that scale?

In the eyes of Simon Bridges the 2018 census reached such complete clusterfuck status that it really isn’t any good for anything at all. In particular, the data it obtained is so ropey that it shouldn’t be used to redraw the electorate boundaries for the 2020 election. Instead, Bridges thinks the existing boundaries used in 2014 and 2017 ought to be retained in place until a new, proper census can be held.

I think he’s wrong about this. What is more, the only way that he could try to force the issue risks damaging a set of very important unwritten societal, political and constitutional presumptions. In other words, this is one of those issues where the long-term consequences matter more than the immediate politics of the matter at hand.

Before getting into the guts of the issue, let’s recognise just how relatively unimportant this electorate boundary drawing exercise actually is. Sure, it sounds crucial: it decides whom people can cast their electorate vote for as their electorate MP! And electing MPs to parliament is really important! Democracy itself is at stake!

That may be true for individual voters who want the chance to support a particular individual into parliament. If I want to help ensure Jacqui Dean gets her seat back in 2020, it really matters to me whether I have to vote in the Waitaki or Dunedin North electorate. But in terms of determining election outcomes at a national level, my (and your) electorate vote hardly features at all.

Because, as everyone ought to know, it’s the party vote that really counts under MMP. This vote is what ultimately determines each party’s overall share of MPs in parliament. And party votes are cast and counted the same irrespective of which electorate a voter resides in.

So, any dark mutterings that a flawed electorate boundary drawing exercise inevitably calls an election result into question need to be taken with something of a grain of salt. Once upon a time, under our old first past the post voting system, this was true. Today, not so much.

Nevertheless, reasons remain to care about having solid data when determining electorate boundaries. Determining the total number of electorates requires an accurate population count. It is especially important that the Māori population is got right, as this ultimately determines the number of Māori seats in parliament. And there are some electorates (think Epsom, for instance) where a major boundary change could have national implications.

Those reasons do lend the issue some importance, albeit without overly catastrophising it. So, does that make Bridges right when he argues that caution requires calling a halt to the electorate boundary drawing exercise?

The first problem is that it actually can’t be stopped without some specific legislative change. Electorate boundary drawing isn’t a government exercise at all, but rather the responsibility of an independent Representation Commission which has various specific duties laid out in the Electoral Act.

Were the government to even try and tell this body to stop its work, that itself would be unlawful. As such, unless and until parliament changes the law, the Representation Commission must carry on and determine how many electorates there will be in 2020, then draw up new boundaries for them, which automatically will apply on election day.

(Note also as an bonus fun constitutional fact that 75% of MPs would have to agree to any law change to postpone this electorate boundary drawing exercise, as the process is one of the matters protected by the Electoral Act’s “entrenchment provision”!)

Of course, in order to carry out its various duties the Representation Commission first must obtain “the results of the census” from the government statistician. And if Bridges is right that the 2018 census was a complete clusterfuck of an omnishambles, then there simply may be no such “results” to give. In which case, the Representation Commission could not do its job.

Here is where Bridges’ view of the quality of census data departs from those involved in the census process itself. According to Statistics NZ, as well as the report of the independent reviewers of the census process, “the counts required for Statistics NZ to satisfy its statutory requirements will be met.” Provided that is so, the Representation Commission can be given the required numbers and the process will automatically continue as per the statute.

Bridges and his National colleagues may try and cast shade over this, but they cannot stop it. Nor, it should be noted, can any government minister. The process deliberately has been set up to prevent such direct political interference in the Representation Commission’s work.

There is, however, one way that Bridges and his National Party colleagues still could try and disrupt proceedings. In a throwback to our old first past the post days, two political representatives must sit on the Representation Commission: “one of those members being nominated to represent the government and one to represent the opposition.”

In the past, these members have been ex-Labour or ex-National MPs. As such, the opposition (meaning National – ACT’s David Seymour doesn’t get a look in here) might be tempted to appoint as their representative someone determined to wreck the process from the inside.

Given that this possibility hasn’t yet been formally floated, I won’t spend too much time on it now. For one thing, the tactic simply may not work given that “the [Representation] Commission may make such rules for the conduct of its business, not inconsistent with the provisions of this Act, as it thinks fit.”

For another, it would very much be a nuclear option. As Elizabeth McLeay lays out in her very good book on the subject, our current Representation Commission and its procedures represent a long-term compromise designed to manage partisan wrangling. A party trying to use its representative to stymie the Commission’s work risks destroying the basis of that compromise. And once it is gone, then it is gone.

As such, this is one of those times where it might be nice for political actors to take a deep breath and step back to look at the big picture. Yes, the census was screwed up. Yes, that was a bad thing. There’s certainly still blame to go around beyond the department’s chief executive; and it is properly the opposition’s job to see just how much of that blame sticks to its minister, James Shaw.

But there’s also been a process put in place to try and recover from that screw up. This doesn’t simply involve employees from Statistics NZ pulling numbers out of thin air. It draws on various data sources held by the government, using procedures that are being overseen by “an external data quality panel of experts” and that will be “subject to peer review by international experts”.

So, when Statistics NZ assures us that this process will produce some (but not all) census data that is as good as (if not better than) that from previous censuses, why wouldn’t we believe this? Because if we’re not prepared to believe it, then we’re solely left with a world of “alternative facts” and individual reckons in which there is no common basis for either agreement or dispute.

The international news pages of our newspapers (ask your parents, kids) are giving us a good, real time look at how societies that make decisions based on such assumptions work out. They don’t look all that great to me. And it would be nice if we could avoid importing the worst of their mistakes to our shores.

Because that really would represent an omnishambles of epic clusterfuck proportions.