Disputing Figures

Listener: 11 June, 1990

Keywords: Statistics;

A Treasury working paper, released under the Official Information Act, appeared to indicate that government expenditure was out of control. It said that between 1984-85 and 1988-89 ‘net expenditure at 1989/90 prices’ had grown from $23,594 million to $27 ,877 million. That is an increase of 18.1 percent, or 4.3 percent a year, while the economy grew annually in volume terms only 0.4 percent.

Obviously social security spending had gone up dramatically because of unemployment, and debt servicing had gone up because of higher interest rates and burgeoning debt. But the Treasury paper seemed to say that real government expenditure on both hospitals and schools had gone up 20.3 percent.

A good number of people would say immediately that they always knew Labour could not control government expenditure, and this was further evidence of its incompetence. One contribution came from Raglan MP Simon Upton, who focused on the data in a chapter in Rogernomics: Reshaping New Zealand’s Economy (published by the state-owned Government Printer for the private enterprise-promoting Centre for Independent Studies). Upton argued that one of the main reasons for New Zealand’s dismal economic performance was the failure of the Labour Government to keep to its planned (almost zero increase) spending targets.

Some Labour politicians leapt at the figures with equal alacrity, although reaching the opposite conclusion. The big increases showed that the Government really cared, and that it was committed to the welfare state. When the Government began heavily cutting the welfare budget, they argued that these cuts – of five percent in the case of the health budget – had to be set against the huge expenditure increases already made. (Indeed, one may wonder whether the Treasury figures had the effect of convincing cabinet members that they bad been over-generous, easing any pain they must have felt from accepting severe expenditure cuts to their portfolios. Subsequent statements by the Ministers of Education and Health suggest this may be true.)

Now the reader may be a bit puzzled about where the extra money went. If you worked in the education and health sector, or were a client of either, you probably do not remember any blowout of munificence between 1984-85 and 1988-89. You are more likely to remember severe restraint under rising demand pressures.

My suspicions of the figures came from examining the volume of output figures published by the Government Statistician. He estimated a decrease for general government services of three percent. Now the official figure covers slightly different spending in comparison with the Treasury figure, but there is no way that the two figures can be reconciled.

So I looked at the Treasury figures and I was amazed by how crude the adjustments were. They merely compared (deflated) the amounts spent in nominal (ie, current price) terms with the consumer price index (CPI).

It makes no economic sense to use the CPI to deflate government spending. Every bundle of spending is different, and the prices that apply to make up its nominal value all change differently; every spending group requires its own price index. The CPI is designed for measuring consumer spending. It would be a poor indicator of price changes for exports, imports, investment or government spending. Each needs its own specific price deflator, and the Department of Statistics goes to a lot of trouble to construct them. Using the CPI to deflate government spending is about as sensible as using the price of carrots.

This was especially true over the period 1984-85 to 1988-89, because the Government changed the spending rules dramatically. Departments suddenly found themselves paying fringe benefit tax, GST, income taxes on social security, rents for their buildings, superannuation for their workers, rates, higher interest rates, charges for other departments and so on. These were transfers within government spending, or between spending and revenue. They are not reflected in the CPI, so it is a totally invalid deflator. (The Treasury official did adjust for GST, but for none of the other changes.)

As it happens, the area health boards could not believe the Treasury figure either. They commissioned a consultant, Guy Scott, to do the job properly for their component of government spending. This is easier than for many other items because there is a specifically designed hospital cost index, although changes in coverage confuse the story. Scott’s paper has now been released. It reports that funding to hospitals increased by 2.9 percent, not the 20.3 percent the Treasury paper would have us believe. Population rose 2.1 percent over the period, so per capita funding rose 0.2 percent per year (probably less than the additional demand pressures). That is all that the hospitals had left to purchase staff and supplies, maintain buildings and so on, after allowing for changes in the prices they faced.

Readers in the hospital sector, either as workers or patients, will find that number much more credible than the Treasury figure. I suspect if Scott were asked to do the same job on the schools, he would find a similar minimal increase.

Little can be done to rectify the mischief that the faulty Treasury figures have done to the Labour Government’s social policy. But those who devise the next one would do well not to be overly dependent upon Treasury statistical advice.