As usual, an election campaign brings out a number of false or misleading statistics, plus a general discomfort with issues that involve numeracy skills. While sometimes they are genuine mistakes, usually the inaccuracies are biased towards the political interests of those who perpetrate them. Ironically, given their emphasis on reading, writing and 'rithmetic as the core of education policy, National and Act are the most guilty parties when it comes to arithmetic deception.

I will note three examples of National's deception here:

1. We keep being told that economic growth in New Zealand is running at four percent. I downloaded the latest GDP production data from Statistics New Zealand's website yesterday (24 November). The most recent annual figure for real GDP growth per capita is zero. Further, the growth in the most recent (June) quarter for which we have published data is negative.

It may be true that Mrs Shipley has advance knowledge of the yet to be released September 1999 GDP datum. But it is really foolish to draw any annual growth conclusions from a single quarterly figure. Mrs Shipley is, it would seem, citing forecasts of economic growth for the year ending March or June 2000.

We might remind her that the All Blacks were forecast to defeat the French on October 31, but did not. We might also note that Donald Brash has already acted on the basis of four percent forecast growth to tighten monetary policy; a self-unfulfilling prophecy. Brash's actions to pre-empt growth in 2000 will ensure that New Zealand's growth will be even further behind Australia's by the 2002 election.

New Zealand missed the growth bus in the 1990s; National have not given New Zealand growth. If we take the important real GNP figure (gross national product), we find that New Zealand has had in effect a zero-sum economy since 1990, having got off to a really bad start with the severe Ruth Richardson inspired recession of 1991-1992. With statistical hindsight, the growth that really only happened in 1994 was choked off by the combined weight of the 1989 Reserve Bank Act, the 1994 Fiscal Responsibility Act, and the inadequate purchasing power of the bottom 80% of New Zealanders.

2. Mrs Shipley has on a number of occasions told us that New Zealand's unemployment rate is the same as Ireland's and lower than Australia's and Finland's.

While the definitions of unemployment of all these countries conform with the ILO standard, there are important differences in the way each country compiles its unemployment statistics. Anyone who has tried to compare unemployment in New Zealand and Australia through the 1980s will know that New Zealand's 1986 official (ie surveyed) unemployment data significantly understate the true unemployment rate.

In Australia, the Bureau of Statistics adjust their household survey data to match the more accurate census data. In New Zealand, we took the survey data as correct and then conducted hamfisted attempts to generate census unemployment numbers that would match those of the survey. We changed the way we calculated census unemployment in 1991 and again in 1996, making it impossible to calibrate our survey data in the way that the Australians do.

You don't have to be a rocket scientist to know that the Australian labour market has been more buoyant than New Zealand's has been in the 1990s. To compare New Zealand's unemployment with Australia's, add at least one percentage point to the New Zealand figures.

Unemployment statistics are extremely sensitive to national institutions; in particular the dynamics of the contrasting income support systems. A large proportion of New Zealand's unemployed are counted as employed because unemployment beneficiaries are allowed to work a few hours part-time while receiving a benefit. Other countries' income support systems are less puritanical; more accepting of the legitimacy of non-workforce status. Hence, other countries have relatively more fulltime students and relatively fewer people employed on a casual or part-time basis. As a result, such countries (of which I assume Finland is one), have fewer "employed" people relative to their unemployed.

The difference in unemployment rates in Europe compared to New Zealand lies mainly in the denominator (the workforce) and not the numerator (the unemployed).

3. On Crossfire (Sunday) Jenny Shipley said that she thought a person grossing $10,000 per annum would get a weekly tax cut of $2.50 if National is returned to power. On the Holmes' leaders debate (Tuesday), she said that a family on $12,000 would get a tax cut of $2-$3 per week. Surprisingly, Helen Clark did not contest Mrs Shipley's claim.

In fact, a person on $10,000 will get a tax cut of ten cents per week. A family on $12,000 may or may not get a tax cut, depending on how many people are involved. If there is just one earner, such a family will get a tax cut of fifty cents per week.

A simple rule is to add 19c per week for every thousand dollars of individual income in the annual range of $10,000 to $38,000. Hence, a person grossing $20,000 would expect to get a weekly tax cut of 10c+(19c*10) = 200c = $2.

I don't know if any of this deception on taxes will affect people's votes. What I am sure of is that a person on $10,000 per annum (a typical annual income for New Zealand's many part-time workers) would be rather upset if they voted for National on the expectation of an extra $2.50 per week in the hand, to find that, on April 1 next year, they only get 10 cents. A mean joke.

In addition to problems with getting statistics right, we get challenges to do arithmetic things such as "name one country in which tax increases led to more growth". Helen Clark finds this difficult to answer because very few countries raised taxes in the 1990s, thanks in the main to pressures arising from globalisation. Nevertheless, the answer is "every single country in the OECD in the quarter-century after World War 2". Growth rates were higher than ever before as taxes were raised to fund collective goods.

Perhaps more pertinently for New Zealand, real per capita GNP doubled in the 10 years from 1932 to 1942. In 1939 - before the war - we were each paying much more tax than we were in, say, 1931.

A combination of better numeracy skills and a better appreciation of history would do wonders for the standard of political debate.

ends

Deaf Ears: Speaker On Support For Mojo Mathers

Scoop Audio: In a press conference this afternoon Speaker Lockwood Smith defended his handling of requests for support for profoundly deaf MP Mojo Mathers' participation in Parliament, and said he was "deeply concerned" by the way the issue had been portrayed.

Earlier today the Greens said they had been told they would have to fund support Mathers requires out of their own budget. More>>