Burton on the OIA

Tony Burton, ex-Treasury Deputy Chief Economist, lays out some of the problems in the OIA over at The Spinoff.

He describes an incident in which a minor letter from him to DoL in preparation of the annual minimum wage review became, after an OIA request, well, this:

As a Treasury adviser on labour market and welfare issues I was asked to see if DoL could be encouraged to improve the report. I thought, naively as it turned out, that if I discussed evidence in bite-sized chunks some of it might sneak into the report. This was the content of my email. Labour market economists would regard what I wrote as anodyne. They would not all agree with it – academics never all agree on any matter – but it would be unexceptionable. The DoL officials ignored my email and the subsequent meeting when writing their report.

Described this way I imagine the eyes of most readers glazing over. “Conversation on academic evidence between unimportant people makes no difference” is not an attention grabber. But when released through the OIA, in the midst of an election campaign, the issue turned into “Mr Key ‘sat on’ the advice for 18 months and ‘tried to fool people’ by using only a later Labour Department review to back his argument”.

This version became one of the issues in a television debate between the party leaders.

A cursory glance at the emails would have made clear they were not sent to ministers. They did not include any reference to a Treasury report to ministers because there was no report.

Tony links through to some reporting by Andrea Vance on it, but that's perhaps because linkrot has set in on TV3's coverage. It was a TV3 OIA.

Here's how Gower reported it at the time, which I'd then blogged. This is the blockquote from Gower:

Everyone knows it’s got bloody hard to live on the minimum wage - even John Key admits that. His defence is that a rise from $13 to $15 an hour will cost jobs.

Key has used this defence in a televised debate, and he's used it to workers on the shop floor at McDonald's as seen in my story last night. But what Key doesn't want to admit, is that this claim is not the full picture and may just be fear-mongering - a rise may not cost jobs at all.

That's what Treasury says in this advice from March 2010 obtained by 3 News under the Official Information Act.

SEE THE TREASURY EMAIL THAT SAYS A MINIMUM WAGE RISE WON'T COST JOBS

Yes, that's right - the Treasury.

The Treasury are the big guns - Government's quasi-independent economic advisers.

And the Treasury says the "claim" (yes Treasury calls it a "claim") a minimum wage rise may cost jobs - "has not been true in the past".

I'd then written:

Gower tries to frame this as a scrap between Treasury and the Ministry of Labour over estimates of the disemployment effects of minimum wages.

So the Department of Labour report actually mentions about 6000 forecast new jobs that might not happen under a theory that the Treasury doesn't believe.

But absolutely nothing in the OIAed email from a single Treasury analyst speaks to the Ministry of Labour's estimates. Nothing.

Gower wanted a gotcha, and he framed what he had as a gotcha.

When I asked about that a couple years later when the issue popped up again, I got this answer (which I still don't really understand).

@EricCrampton @helenkellyCTU @TrevorMallard you are selective, labour/national selective, Treasury selective #MinimumWage

— Patrick Gower (@patrickgowernz) August 1, 2014

The term 'fake news' hadn't yet been invented, but Gower's reporting here surely would have made the cut. An email between analysts hardly constitutes an agency's position.

What are the consequences of this kind of use of the OIA? Burton describes a culture akin to prey animals approaching a watering hole in the savannah, constantly scanning the horizon for predators.

Public servants experience the OIA the way savannah animals experience crocodiles lurking under the surface of a river. The animals have to go to the river but do so aware that random attacks are a moment of inattention away. If this metaphor seems over the top, I invite the reader to look over the last few weeks of everything they wrote, typed or texted. Imagine someone had a legal right to publish any three consecutive words, without context or explanation, and with the potential that you might lose your job. Would that make you a little more guarded about what you wrote? Within the public service, versions of this thought experiment are called the “Dom Post test”.

Unsurprisingly public servants take steps to avoid the crocodile. The OIA is meant to include verbal exchanges, but in practice that is hard to enforce. The result is the habit of minimising the written record if there is a risk of failing the Dom Post test. This habit is so endemic it is applied semi-consciously and only noticeable when someone, usually a junior official who has not been fully acculturated, needs to be reminded to “take the discussion off-line”.

Of course there are many occasions when it is common sense to have a quick chat rather than to draft documents and set up meetings. (And anyone who has wasted time in pointless bureaucratic meetings will wish for more of this common sense.) However, extending this to replacing written comments is not in the spirit of the OIA.

In truth, any criticism of OIA avoidance of this sort needs to recognise it is a pragmatic response by people who want our government institutions to function. Public servants are meant to both serve the minister and have some level of independence. In practice this means some of what they do is not what the minister would want them to do were they simply serving the minister. Strict adherence to the spirit of the OIA would see these differences continuously used to attack the minister. It is hard to see how ministers could use their ministries effectively when the process of generating advice would be so damaging to what they were elected to achieve.

In as much as the OIA puts pressure on public servants to only write what they are personally prepared to defend in the full glare of the public they serve, it’s a good thing. The problem is that’s not the reality of how it is experienced, which means people do not just respond by improving what they write but by trying to avoid the lurking crocodile.

The OIA is broken, but so are the underlying incentives. Trying to fix the OIA without addressing the underlying incentives won't work well.