Of the very strong June-quarter employment growth relative to the March figures, Statistics New Zealand was unable to distinguish whether it was genuine employment strength or merely the impact of survey changes.
''As such, we cannot read into the second-quarter results any implications about the strength of the economy or implications for the interest-rate outlook.
''It is best to treat the new HLFS survey as a structural break from the previous one,'' Mr Tuffley said.
Statistics NZ figures showed the unemployment rate fell to 5.1% in June from a revised 5.2% in March. Compared with the March quarter, 1000 fewer people were unemployed in New Zealand at the end of June.
Statistics NZ labour and income statistics manager Mark Gordon said compared with June 2015, there were 8000 fewer unemployed women and their unemployment rate fell from 6.2% to 5.4%.
In June, results indicated 66.2% of the working age population were employed. Compared with March, that meant 58,000 more employed people, up 2.4%.
However, Mr Gordon acknowledged part of the increase reflected changes to the redeveloped HLFS, alongside any real-world increase in employment during the quarter.
Key changes included: improved questions about undertaking paid work, which now identified more self-employed people; and the inclusion of members of the armed forces who lived in private dwellings.
Mr Tuffley said the new HLFS was a better labour market measure, as it now captured self-employed people better.
Including armed service people in private residences could add more than 10,000 people alone.
The inclusions, as well as underlying changes in employment, added 58,000 to the number of employed people compared with the last quarter.
''Beyond the known armed forces impact, it is impossible to break out the drivers of the remaining 48,000 lift in employment between survey changes and underlying employment changes.''
The level of employment, the labour force and the participation rate all showed signs of major change and it would take time before the picture became clearer, Mr Tuffley said.
''It may take another three quarters for the data to settle down. In the meantime, the monetary policy significance of the survey will be low.''
The Reserve Bank would rely heavily on its Labour Utilisation Composite Index, he said.
Labour finance spokesman Grant Robertson said on Tuesday National appeared to be actively massaging official unemployment statistics by changing the measure for joblessness to exclude those looking online.
He was rebuked by government statistician Liz MacPherson, who said her department was politically neutral. That did not stop other Labour MPs from using social media to question the department's neutrality.
Mr Robertson said yesterday he accepted Ms MacPherson's assurances on the reason for the change in criteria.
