Chief Statistician resigns: But is that the end of the story?

Chief Statistician of New Zealand, Liz McPherson handed in her resignation to Minister of Statistics James Shaw on . Whilst Ms McPherson’s departure is definitely a step in the right direction, can we trust Statistics New Zealand and their treasure troves of statistical data? And if the answer is no, what will it take to restore that trust?

To briefly recap Ms McPherson’s chief task as the boss of Statistics New Zealand was to make sure that the 2018 Census ran smoothly. As is now very clear, that was anything but the case. Ms McPherson is now gone, but there are huge holes in the 2018 edition and considerable uncertainty no doubt exists among the many ministries and departments whose planning for the next few years has been thrown into turmoil.

Would it be better to simply organize a full brand new census for the earliest possible date and assume that the 2018 data is not able to be properly used? Possibly, but a new Census is not a cheap, logistically easy or rapid task to be carried out in terms of organizing it, never mind Census Night.

If not, do we know how much of the data is usable? How much data is missing? What data is missing? Are there minimum data amounts that must be maintained in order for particular data sets to be of use, and if so what are they? No doubt these questions have been taxing a lot of minds in Statistics New Zealand, as well as its Minister in Charge.

Let us assume the worst. The data sets for multiple agencies simply do not permit them to carry out appropriate planning and the next few years will be based on educated guesses rather than hard data. We will assume that it affects the Ministry of Health, Ministry of Education, Ministry of Social Development among other significant users and providers of data. I cannot help but wonder what health, social, and education programmes are going to be underfunded, under resourced or simply stop working because of this ineptitude.

To be painfully honest, I have no confidence in Statistics New Zealand. Unless Mr Shaw reruns the Census I believe he should hand back his ministerial warrant for that particular portfolio. In April of this year Ms McPherson acknowledged no Iwi related data would be released. She also admitted that that data collection was insufficient for what was expected of the set. She should have handed her resignation in at the same time.

Where is National on this? Where is A.C.T.? I would have expected both parties to be busy making political hay out of this. I would have thought that A.C.T. leader David Seymour would have been left, right and centre telling us what A.C.T. would do and why this Government is not fit for the tasks ahead.

Stuff reporter Thomas Manch suggested that it might take years to restore the level of trust we need to have in Statistics New Zealand and its leadership. Given that even with the handing in of her resignation, Ms McPherson was still trying to play down the gravity of the situation, I am inclined to agree.



Statistics New Zealand’s potential Census emergency

Some time ago I made mention of a major failing of Statistics New Zealand in their internal operation and failure to successfully run the 2018 New Zealand Census. In April 2019 the Chief Statistician Liz McPherson admitted 1 in every 7 New Zealanders failed to complete the compulsory survey that happens every 5 years and is essential for planning government services, spending priorities and performance targets.

At that time it was discovered that the agency responsible for collecting statistical data on New Zealanders had filled a hole in its finances by funnelling $10 million from insurance payouts and capital. Upon realising that it was short, S.N.Z. asked for another $20 million for the 2019-20 financial year, which was to fill in a funding short fall of 15%.

Now it has been found out that S.N.Z. actually needs considerably more money – between $33-$43 million more each year for subsequent years. As a result the Minister for Statistics, James Shaw is having to ask the Treasurer Grant Robertson for millions more in funding that no doubt both of them would have hoped they would not have to fork out.

If the money is not stumped up, S.N.Z. has a list of ten products it was going to cull or severely restrict. They included surveys for research and development, land occupancy/transfers, energy use among others. This would affect planning and spending priorities for a multitude of agencies and items in the budget.

Allowing it to continue suggests lax responsibility by the Minister of Statistics in overseeing the agency. It suggests that the incompetence of Ms McPherson is going to be tolerated. Sure it might not be the biggest mismanagement crisis we have had in a New Zealand government department, but after telling the Government a second time in less than a year that its financial problem is worse than it thought, can we really be expected to trust Ms McPherson and her senior S.N.Z. staff to know what is going on?

Ms McPherson is contracted to S.N.Z. until the end of 2021, but one has to ask whether that should still be the case. If priority targets were set and closely monitored with the threat of sacking hanging over her head, could we rely on Ms McPherson to display the necessary honesty when she originally tried to hide the issue? I am not wholly sure we can.

Mr Shaw needs to make a couple of tough decisions and he needs to make them quickly. The first one is whether Ms McPherson is worth the risk that goes with forking tens of millions of dollars more in terms of making sure that they do not end up being wasted. The second is – assuming the full extent of the problem has now been revealed – whether the data provided from the Census by those who did manage to complete is enough to fend of an emergency Census.

Drastic? Yes. Unnecessary? I hope so, but if agencies on whose well being people depend such as M.S.D. and the Ministry of Health suddenly find themselves unavoidably short on critical data, do we have a choice?

Possibly not.

South Pacific development report: Not Yet Complete (and nor is N.Z.’s)

This week five South Pacific nations including New Zealand are turning in report cards on their development progress. But even before the United Nations receives them, we know what will be written on them by the U.N. Development Programme.

Not Yet Complete.

Not surprisingly this will attract a response asking how it has come to this. Some of the answers are so obvious we have somehow been blinded by them. One great example of this is the sheer number of indicators that need to be reported on. 230 have been set. In the South Pacific, which comprises two wealthy first world nations – Australia and New Zealand – and a bunch of small island nations so tiny most printed maps struggle to show them, a combination of administrative difficulties, lack of resources, tiny populations and corruption mean reporting on them all is highly improbable on the best of days.

Perhaps it is time to refine the Voluntary National Review – as these exercises are known – to something more realistic. From an office somewhere in New York it is highly unlikely unless the people designing these programmes have visited small island nations like Tonga, the Solomon Islands or Vanuatu, that they have any idea about the logistical, cultural and political challenges of carrying out such in depth data gathering.

Let us take Papua New Guinea for example. Remote, once you get outside Port Moresby, towns like Lae, Wewak or Rabaul on New Britain, one might have to travel for days on dirt roads across rivers with no bridges, few airfields to land aircraft. A translator fluent in the many dialects would be needed as well as a cultural advisor to navigate local customs.

Anyone who has done statistical research will know and understand the challenges of creating, maintaining and manipulating data sets. They will understand that that whilst the set needs to be comprehensive enough and deep enough to work in, there needs to be a degree of refinement about what types of data one is after. In recognizing these challenges perhaps the biggest problem for me though, is, whether all of this is actually necessary? Of course the United Nations needs to know how its individual members are getting on, but how much of this can they not gain from simply requesting that their member states focus on a simplified range of indicators?

New Zealand is lucky. With a comparatively well working system for gathering statistical data, our biggest problem might be more whether it is still current. Being one of the larger nations by resources, population and wealth, we could establish those 230 indicators. But, just as with Fiji, Papua New Guinea and elsewhere, a bigger question might be do we really need all of them? Our 2018 Census left some major gaps in what it was meant to supply those doing the analysis for various government departments. The Chief Statistician was found wanting in terms of her management of Statistics New Zealand’s biggest single (five yearly)task.

Before any of this happens though, New Zealand needs to be honest with itself. Our statistical system, whilst in good order compared to say, the Solomon Islands or P.N.G., is not good enough if it cannot conduct a Census properly. Maybe the Voluntary National Review of New Zealand should be postponed until we get our own statistics in order.