The Australian Bureau of Statistics estimated Tasmanian employment decreased by 650 people in August after an increase of 730 in July.

Mr Eslake said full-time employment fell by 2000 - reversing most of a gain of 2490 in July - while part-time employment increased by 1360.

"Almost certainly, these movements are largely the result of people working fewer hours in August than they had done in July - and hence moving from full-time work to part-time - as a result of the impact of the closure of Tasmania's borders to travellers from lockdown-affected states on employment in tourist-related businesses," Mr Eslake said.

"This observation is supported by looking at the gender composition of the change of employment.

"Women's employment fell by 2910 in August, of which 2790 were full-time jobs, whereas men's employment rose by 2260."

Mr Eslake said the decline in August left employment about 1020 people ahead of its pre-recession peak in February 2020.

"Total hours worked fell by 0.2 per cent in Tasmania in August, compared with increases of 0.4 per cent in each of June and July," he said.

"Despite the decline last month, total hours worked in Tasmania were still 3.1 per cent above the pre-pandemic peak, which is a better outcome than for any other state or territory except South Australia.

"In New South Wales, hours worked last month were 12.1 per cent below their pre-pandemic peak, while in Victoria hours worked were 2.0 per cent below the highest pre-pandemic level.

"Tasmania's labour force increased by 1960 people in August, pushing the labour force participation rate up by 0.4 pc pts to 61.7 per cent, its highest since March.

"With employment falling and the workforce increasing, inevitably the unemployment rate rose, by 0.9 percentage points to 5.5 per cent, from 4.6 per cent in June and July, which had been the lowest since before the Global Financial Crisis.

"Tasmania's unemployment rate is thus once again above the national average (of 4.5 per cent), although not too much should be read into that, given the distorting impact on official unemployment rates of big swings in labour force participation rates."