Dr Cooper's report also questions a key plank of the case for the Gungahlin light rail line, which equates higher density living with higher use of public transport.

The presumption that there was a direct correlation between transport and demand for higher density land use was contestable, she said.

"Dr [Geoffrey] Clifton, a transport expert engaged by the Audit Office, has advised that it is possible to create higher density urban development that does not lead to a significantly greater demand for sustainable transport."

Dr Clifton warned that unless frequent transport was put in place early "it is difficult to create dense low-driving neighbourhoods. New residents will be more likely to develop the habit of using the car as their primary means of transport," he said.

Dr Cooper has investigated the frequent bus network, which was the backbone of the government's 20-year transport strategy released in 2012.

She found significant weaknesses.

Oversight was not effective, she reported. A working group set up in 2012 to oversee the strategy had not met as often as it was supposed to, and was disbanded in September. Responsibility had been shifted to the parking co-ordination group and the roads co-ordination group, but the minutes of both groups showed no evidence that either had considered the strategy.

Although the government was to report each year from 2012 on progress towards the frequent network, it had only reported once, in late 2014. The reporting was "ambiguous and in some instances inaccurate".

"Some actions that have not been completed have been reported as being "achieved", and some actions reported as being "on track to be achieved" were actually not achieved; many actions were reported in the report card as "on track to be achieved", yet no evidence is presented to support this claim," Dr Cooper said.

A commitment to "'include seven day network in ACTION enterprise bargaining agreement in 2013" was reported as "needs improvement" whereas it had not been achieved and was unlikely to be.

The 40km/h operating speed was reported as "needs improvement" but actual speeds were just 32km/h. "Grow the bus fleet" was reported as "on track to be achieved" but progress had been slow and replacement of buses would need to be accelerated.

Dr Cooper said a significant number of new buses was needed, but she pointed to confusion over just how many. A government masterplan in November 2014 said the fleet needed to increase from 411 to 562 by 2031, but the MRCagney review put the number much higher, at 1007 buses.

Targets for shifting people from cars to public transport for trips to work had not been met.

In 2011, 7.8 per cent of people used public transport for their journey to work, well short of the 9 per cent target. Just 2.8 per cent of people were cycling against a target of 5 per cent, and 4.9 per cent walking against a target of 6 per cent.

The 2016 and 2026 targets (a 10.5 per cent share of work trips in 2016 and 16 per cent in 2026) remained "a major challenge", Dr Cooper said. The MRCagney report said the 16 per cent target was "plausible but ambitious" and would take many more buses, higher density living and changes to "road pricing" and "parking pricing".

Dr Cooper said implementation of the frequent network was inadequate in parts and "not being progressed in a timely manner".

The frequent network aims for buses every 15 minutes, or less. In 2015, the only routes meeting the definition were the rapid buses between Belconnen, the city, Woden and Tuggeranong; and between Gungahlin, the city, Russell, Barton, Kingston and Fyshwick.

The plan was to have frequent services also on at least 10 smaller routes, including in the parliamentary triangle, to the airport, and in the inner north by 2016, but so far none of the local services had met the target.

The audit found major holes in data needed to check whether targets were being met. The government was not able to report on how many people were switching to public transport, walking or cycling because there was no updated census data and it hadn't yet developed an alternative measure. It was also unable to report on transport network performance, including travel time and wait time for connections because methods of counting such data hadn't been designed.