Privacy concerns in the lead-up to the 2016 Census have taken a toll, with some people giving fake names and withholding their date of birth.
There's also been a sharp drop in the number of respondents allowing authorities to keep their data archived for 99 years.
RELATED STORIES:
Census 2016: Five ways Australia is getting more diverse
The first batch of data from last year's bungled census was released by the Australian Bureau of Statistics on Tuesday with authorities insisting the information collected is useful.
That's despite a fall - from 96.5 per cent in 2011 to 95 per cent last year - in the participation rate.
Privacy concerns plagued the half-billion-dollar exercise in the lead up to Census night on August 9 with several politicians, including independent senator Nick Xenophon, vowing to risk a $180-a-day fine by refusing to provide their names and addresses.
The exercise was then marred by cyber attacks which prompted the bureau to shut it down for almost two days.
IT company IBM later paid out millions of dollars in compensation for its role in the mess.
Senator Xenophon at the time said the ABS had failed to explain why it needed names, warning the Census would morph into a "mobile CCTV that follows every Australian".
Chief statistician David Kalisch insists the ABS is precluded from sharing the data with other agencies.
But he admitted it was used to evaluate the performance of health measures like mental health programs.
"We did some work for the national mental health commission over the past couple of years which looked at participants in various mental health programs and matched that with some of the census information around employment," he told reporters.
An independent panel appointed to examine the quality of the data collected concedes the privacy concerns had an impact.
Panel chair Sandra Harding told AAP one per cent of respondents gave no name or a fake name, while three per cent chose to provide their age instead of date of birth.
"It doesn't really impact on data quality so much as it does potentially impact data integration or matching activities," she said.
The proportion of respondents allowing the ABS to archive their data for 99 years fell from 63 per cent in 2011 to 50 per cent in 2016, despite having previously been on an upward trend.
Professor Harding said it might be a sign of growing concerns within the community about online privacy.
