The cost of the state

The government’s share of the economy may have less effect on growth than previously thought — but the expenditure and taxing mix may be holding us back.

These are among findings in a study by Dr Arthur Grimes, a former head of economics at the Reserve bank and director of the Institute of Policy Studies at Victoria University, now with Motu Research, a think tank, and adjunct economics professor at Waikato University.

Grimes found that lifting the government’s share of the economy does slow economic growth, as many economists reckon, but not by as much as some say.

This will be music to the ears of Finance Minister Michael Cullen, who has argued that what governments do is shaped by how wealthy their countries are, not the other way round.

But Grimes’ other finding will not be comforting: that what the government spends its money on and the choice of tax may play an important part in how much a change in government share affects growth — and New Zealand’s particular mix is likely a negative.

Cullen stated his belief in a speech in late August in which he also said: “We should never apologise for having an active role for the government in providing personal supports that it can afford. That is a right of progress,” he said.

That assertion flew in the face of what National, ACT and many in business have taken as a golden rule: that cutting government spending will increase economic growth and without growth social services will suffer.

Their contention draws on a much-quoted 1998 historical economic analysis* showing that countries which at the start a decade spend less as a percentage of GDP grow on average faster during that decade than do those with a higher share of the economy at the start of a decade.

The 1998 study, which examined 23 OECD countries in decades from 1960 to 1996, concluded that a 10 per cent increase in government spending as a percentage of GDP reduces a country’s annual GDP growth rate by 1 per cent. Over a decade that cumulatively reduces wealth heavily.

And this country seems to have demonstrated the thesis in reverse: spending cuts in the 1980s and early 1990s were followed by a decade of higher growth than in the 1980s.

Not so fast, says Grimes. There is more to slower growth rates than higher government spending.

In a paper for the Ministry of Economic Development (MED), Grimes quotes research showing on the one hand that “demand for government services tends to rise as countries become richer” and on the other that the richer a country is, the lower its subsequent growth rate tends to be. Cullen would readily endorse that.

Also, Grimes says, the 1998 study omitted “decade-specific effects”. Though the two oil shocks of the 1970s contributed to lower world growth in the 1970s than in the 1960s, the 1998 study attributed the slower 1970s growth wholly to the fact that governments occupied a larger government share of the economy at the beginning of the 1970s decade than at the beginning of the 1960s.

Moreover, Grimes says, the 1998 study dealt only with overall GDP growth, whereas a more relevant indicator of growth is GDP per capita. Overall GDP can rise just because there are more people even if individuals’ income and wealth don’t rise. (In fact, part of New Zealand’s high growth rate in the mid-1990s and the early 2000s was due to high immigration.)

Turn that around: if population falls or the growth in population slows, the growth rate will be slower. So slowing population growth rates from the 1960s would have contributed to a lower total GDP growth rate in later decades, Grimes says. The 1998 study omitted this population factor.

Grimes fed all of those factors into the 1998 study’s equations. He found that a 10 per cent lift in the government’s share of the economy would cut annual GDP per capita growth by between zero and 0.5 per cent — not the 1 per cent the 1988 study found.

But Grimes also examined the impact of the mix of government spending and tax.

He quotes a study showing that a higher proportion of direct (income) tax to indirect tax (GST and excise) reduces per capita income. “Increasing that ratio [of direct taxes as a proportion of total tax] from 50% to 60% is estimated to decrease per capita income by 3.3%,” Grimes says. But other studies show this can be partially offset by government investment spending — which is what MED programmes are supposed to be.

“The overall implications of this recent international evidence is that a higher tax burden (particularly through direct taxes), higher government transfers [welfare payments] and to a lesser extent higher government consumption expenditures (excluding education and health) are all associated with poorer growth outcomes,” Grimes concludes.

“Government investment expenditures (for example, infrastructure and education) contribute positively to a country’s growth rate if, and only if, financed by non-distortionary taxes or through cuts to unproductive expenditures.”

How does that apply here? Grimes says that a true comparison rates New Zealand as having a relatively small government sector in the OECD, 7 per cent less than the weighted OECD average (if state and local taxes are included in the calculation). That should be a plus for growth relative to the OECD average.

But he says New Zealand has a large share of GDP paid in distortionary — labour and capital income — taxes: 18.3 per cent, compared with the OECD average of 14 per cent. The gap is wider when measured as a proportion of government revenue.

And, Grimes says, a relatively low share of government spending in New Zealand is productive expenditure.

“Overall,” Grimes says, “while New Zealand has a relatively small government sector, government revenue flows are skewed (relative to the OECD average) towards income taxes and away from productive expenditures. Both features may be detrimental for economic performance.”

So there is work for Cullen to do yet. But he has a consolation prize in his determination to run budget surpluses: Grimes’ inquiries also showed that “an increase in the fiscal surplus of 1 per cent of GDP has a partial effect of increasing average annual growth by 0.11 percentage points”. Prudent budgeting counts.

*James Gwartney, Randall Holcombe and Robert Lawson, “The Scope of Government and the Wealth of Nations”, Cato Journal 18(2).