We need to learn what happens when public spending is repressed. It does not lead to efficiency gains. Sometimes the consequences are disastrous.

The private enterprise failures in the Global Financial Crisis led to public expenditure changes. The Americans poured them into subsidies supporting the bankers, the Europeans went for Austerianism cutting back in the public spending on the ‘undeserving’, The Chinese – the last Keynesians? – splashed out on infrastructure, which has enhanced their growth prospects.

New Zealand chose to give big permanent income tax cuts, as well as a $2b financial sector bailout. The result was that, under the Key-English Government, public spending was severely constrained – reinforced by its choice to deal with the Canterbury earthquakes from within the public spending budget.

Inevitably much public service delivery deteriorated. The worst example was the Pike River disaster. The failure of the Department of Labour to maintain an effective mining inspectorate led to the death of 29 miners. Now we are spending $37m on recovering their bodies. Talk about the lack of a ha’p’orth of tar sinking the ship.

As terrible as this single instance was, it was not unique. Others died in unnecessary industrial accidents. And it was not just the Department of Labour whose services suffered from inadequate funding (and a penchant for light handed regulation which justified the cutbacks).

This is nicely illustrated by a recent paper by Don Gilling, a retired professor of accounting (and a friend) with a passion for libraries. He used the Official Information Act to work out that the annual operating spending of the National Library fell 9 percent in the four years following its merger with the Department of Internal Affairs, that of the Alexander Turnbull Library fell 10 percent and that of Archives New Zealand fell 19 percent following a similar forced merger. Meanwhile consumer prices rose 12 percent.

The easy explanation is that the central administration of the DIA raided its operating divisions by inflating the charges for overheads. (We know that happened to Archives New Zealand in the 1990s.) The heads of the two units were too low in the hierarchy to resist and the DIA did not give high priority to their responsibilities among its numerous and diverse activities.

I am told the DIA claims to have faced severely rising costs for its overheads. We shall see when they go before a select committee. They were under funding pressures and perhaps were more benign towards central administration compared to the operating divisions. Even so there are two curious consequences.

The first was that when the mergers were imposed in 2010 there was a promise that there would be savings on overheads. It will be interesting to see how the promises are reconciled with the outcomes.

Second, Gilling used the Official Information Act to obtain the information. The responses were grudging and one is left puzzling whether the DIA accounting system was not up to it, or whether they were being bloody-minded or even hiding something. What is surprising is that the information was not already in the public domain. Before the mergers it was in reports to parliamentary select committees. The effect of the mergers was to reduce public monitoring.

This is particularly ironic since the focus of the two agencies is about public access to information. Archives New Zealand is the keeper of the public record. Without it, we could never have fairly settled Treaty of Waitangi claims of the nineteenth century, while the Official Information Act would be a nullity since a requested document could legally be destroyed. A library is the private sector equivalent.

You may not care much about these things but the Czech novelist, Milan Kundera, wrote ‘[t]he struggle of man against power is the struggle of memory against forgetting.’ A reason for New Zealand’s regime being less repressive than the Czech one (Kundera was exiled by it) is that we do not have to struggle as much. It is not surprising though, that a department of state should be so dismissive of fostering memory even if it was its responsibility.

We are indebted to Gilling for doing that detailed and tedious forensic work. My guess is that were someone to similar work on a host of other government activities they would come to similar conclusions. (Even Gilling, though, would blanch at doing the same job for that vast mysterious conglomerate, the Ministry of Business, Innovation and Employment, which has the rump of the Department of Labour merged into it.)

The conclusion has to be that the Key-English experiment of restraining government spending in the expectation it would find efficiencies has been a failure – any experiment which caused 29 deaths has to be a failure.

We have been seeking such efficiency gains for at least 50 years. The reasonable assumption has to be that most of the easy gains have been found. Such fat that is left is like that stippled through a quality steak. Remove it and you destroy the texture of the meat.

There is the story of a farmer, who thinking his horse was eating too much, steadily reduced its feed. He had finally succeeded in getting the feed down to zero when the bloody thing went and died on him.

The same is true for the public service. Of course we should insist on its improving its performance but squeezing it of funds is not a means. My guess is that we need to review what we want it to provide: there are activities which reflect hysteria, fashion and irrationality or are not evidence-based.

(For example, I am not a criminologist, but I puzzle over our incarceration rate which is high by international standards. At $100,000 plus a prisoner a year, I am glad to see this government systematically reducing numbers; I hope it is done rationally.)

We may not have a list of what the public service should do just because ‘power’ prefers to forget, to not think systematically. The danger is any review would find wide public support for a bigger public sector and we cannot have that, can we?

The previous government tried to forget Pike River; the families would not let them. We learned about the need for mine inspectors with the Brunner mine disaster in 1896 – 65 miners died (of little consolation to the 29's families). But mine inspectors were not its only legacy. It led to the Workers Compensation Act of 1908, which developed into our accident compensation scheme which despite its defects, is the best example in the world that I have seen of tacking the problem of industrial accidents.

A consequence of the Pike River disaster was the implementation of extensive changes to workplace health and safety practices (WorkSafe ). Sometimes you cannot hear the clash of the stable door being closed for the sound of the hoof beats of the galloping horses.

We may find out what happened to the 29 men, their families have had some financial compensation, we may exact retribution against some – but only some – of those at fault, WorkSafe is in place, and we have given the 29 a physical memorial. Learning and applying the larger lesson of the dangers of underfunding government services would ly be a greater memorial.