Inland Revenue is conducting a survey on how the rich get richer, but already it has exposed itself to the sense it is searching for what it is looking for.
The department is conducting a survey of the rich, in an attempt to plug what is a clear gap in New Zealand's official knowledge on how wealth accumulates among the wealthy.
But, it would appear, it has little interest in looking at what happens when the wealthy fall on harder times.
Having written to more than 400 people who it estimated have a net wealth of more than $20 million warning them that they are going to be part of the study, IRD began scrubbing names from the list soon after. So far, the list facing information demands has dropped by "about 20".
As well as determining that some on the list are so ill they cannot provide the information required, or only recently in New Zealand, IRD has confirmed some of the original 405 were removed after convincing the department they are not, in fact, highly wealthy.
The fact an individual is not a multimillionaire may seem like a good reason to exclude them from the study into the very wealthy, but it is likely to skew the results.
"It shows that the methodology of this whole thing is totally unscientific," Act leader David Seymour said.
IRD says the study is meant to help assess the effective tax rates of wealthy people on their economic measures of income, not just their taxable income.
For wealthy people whose assets have surged in value over the past decade, it may well show that the wealthy pay an effective tax rate (when, for instance, the unrealised increases in the value of their land and property assets are included) below that paid by even those on low incomes.
But if the survey excludes even some of those who have lost their fortune, it will inevitably give a rosy picture of the finances of the wealthy, some who rise or fall on the back of taking huge amounts of risk.
"If they're discounting people who have lost money, of course they're only going to find people who have only got richer," Seymour said.
The Epsom MP claims written Parliamentary questions filed by the Act Party are further evidence of a lack of scientific rigour to the study.
In the answers, Revenue Minister David Parker acknowledged formation of the list included scanning public records, including the NBR List.
The questions also revealed the 405 was not a subset of a larger group it had been considering, but the list it already had.
Seymour said he would have expected the survey might have been drawn from a much larger sample and used variables to select which of the group was most relevant for the policy purpose.
"It's an amateur hour, avaricious fishing expedition and it can't seriously be seen as a policy initiative, it's an envy initiative."
Parker's office referred questions about the survey to IRD. It issued a statement which suggested the list had been built over the past 20 years.
"Inland Revenue monitors a range of information sources including public information (e.g. media), major transactions, and information through the tax system," it said.
"It is not possible for IR to know the entire population of high-wealth individuals, and sample from it, due to information shortfalls in New Zealand's statistical base. This is one of the reasons for the project."
While it has not said as much, the group is said to have long been quietly monitored by a special team in Hamilton.
Many of those included in the list are said to see it as an invasion of their privacy.
Some see their inclusion in the survey as an act of bad faith because, in effect, they had earlier volunteered details of their wealth to IRD, unaware it would be used be used as the basis for inclusion in the wealth survey.
One tax adviser said some of his clients had volunteered to provide much more information to IRD than was legally required, in an attempt to head off any possible concerns about the way they had treated business transactions.
For those clients, the tax adviser said, it was likely the only way IRD would have known about the extent of their wealth. Other clients who were quite clearly more wealthy appeared to be unknown to the department.
The IRD's decision to investigate people who had volunteered information meant future clients were unlikely to make the same decision again.
"The chances of us convincing clients to do this again is now basically zero."