The main value of the report is the wealth of statistics it contains, drawn from the APS employee census and other data sources. They reveal the public service continues to grow, up 2.3 per cent from the previous year. Women are now a majority of the workforce at almost all levels, including SES Band 1 (and creeping up towards equal numbers at SES Band 2 and 3).

The tables and graphs are great; the commentary is mixed. Parts are new and welcome, including a focus on mental health and suicide prevention and encouragement for flexible working and working from home. Others could have been written any time in the past three decades - digital government, effective use of data, recruiting staff with a diverse range of skills - but are still important.

There is however an incongruous element - the claim that Australia "now faces the most challenging and complex strategic environment since the 1930s and early 1940s". Really? Compared with the tensions of the cold war in the 1950s and 1960s? The Vietnam War? The Korean war? The Cuban missile crisis that brought the world to the brink of nuclear war? Indonesia's Konfrontasi, serious conflict on our doorstep? The Global Financial Crisis?

Despite the evidence of history, the idea that we are under increased threat has become widespread. It is not altogether surprising the report has picked it up. The narrative is being fostered actively by a government that sees an electoral advantage in raising a sense of threat, because the Coalition is traditionally seen as stronger than Labor on defence and security.

Much of the media coverage of the report focused on an increase in complaints about harassment and bullying, including a more than doubling of complaints about sexual harassment (from 32 complaints in 2019-20 to 78 in 2020-21). This is an underestimate - the report notes more than half of census respondents who had seen bullying or harassment did not report it, the most common reason being they did not think action would be taken.

There has also been an increase in potential corruption witnessed by census respondents, from 3.5 per cent to 3.8 per cent. Cronyism and nepotism are the main types of corruption witnessed. While this percentage number might seem low, it is higher than it should be. Any corruption is too much. Assuming the census is representative, 3.8 per cent of the APS workforce of 153,945 has witnessed potential corruption - some 5800 cases. Way more than a democracy should find acceptable.

The report is complacent about corruption. It cites a report from a former public service commissioner that the integrity framework was robust and needed little change. That conclusion was reached after consultations with senior public servants and so-called "externals" who comprised, in all but one instance, former senior public servants. The self-serving conclusion was that the public should move along - nothing to see here.

This in a public service that implemented the unconscionable and unlawful robodebt scheme (for which it has still not apologised or disciplined the senior public servants responsible); paid ten times too much for land for the western Sydney airport; told public servants that reporting ministerial corruption to the media could be unlawful; helped implement a $660 million car park grants scheme that a former judge labelled corruption; and is still not subject to oversight by an independent anti-corruption agency.

This last omission is vital. We do not even know how much corruption is out there. What little we know is mainly due to the excellence of work by community groups and the underfunded Australian National Audit Office. Until there is a Commonwealth equivalent to the various state anti-corruption bodies such as the NSW ICAC it is possible that large amounts of corruption continue unreported and unscrutinised. Or there may be relatively little. We just don't know. However, experience in other jurisdictions is that where disclosure laws and investigation are weak, corruption flourishes. It is a significant weakness in Australian democracy that the Commonwealth remains complacent about corruption.

Former APS commissioner, now academic, Andrew Podger suggested earlier this year that this year's State of the Service report might identify lessons from the robodebt fiasco with a focus on behaviours required by the APS values. His suggestion was not taken up. That is characteristic of today's public sphere. Learning from failure requires owning up to failures. The public service is taking its signals from the government and ministers; in our current political environment, admitting to mistakes is career ending. As a result, learning is harder and future mistakes more likely.

So what does the year ahead hold for the public service? The pandemic will continue to dominate. Vaccination rates are now high, but that may itself lead to dangerous complacency. Vaccines are not 100 per cent effective. They reduce the chance of acquiring COVID-19 and significantly reduce the risk of death, but don't eliminate risks. Some fully vaccinated people will still acquire the disease, and some of those will die. Managing this will remain a priority. The government indicated in its last budget that it would maintain the COVID-19 economic recovery plan "until the economic recovery is secure" - when that might be is undefined. If there are further lockdowns arising from the spread of new variants, measures to deal with their adverse economic impacts will still be needed.

That is one area where the public service has done well - managing economic impacts. In other respects, it might have done better. The chopping and changing advice from the Australian Technical Advisory Group on Immunisation, ATAGI, on the AstraZeneca vaccine contributed to community distrust. The issues were complex: the balance of risks between the dangers of COVID-19 for older Australians versus the tiny risk of adverse events from the AstraZeneca vaccine. Standard stuff for public health immunology folks, but not for the public.

There's a known cognitive bias among experts - the tendency to think that because they know what they are talking about, other audiences will too. Unaware of this blind spot, the health messages put out were complicated and confusing. This slowed the vaccine rollout and according to the ABC continues to contribute to low vaccination rates in remote indigenous communities. This might have been avoided if behavioural economists and broader public policy analysts were involved rather than reliance purely on unmediated technical advice.

At state level there are lessons from the contrast between different approaches to lockdowns and movement restrictions. It seems that harsh messaging and prominent use of troops and police is counter-productive, producing mistrust and fear of government that gets in the way of effective public health measures. If we want 2022 to be better our public services need to learn from experience and continue to evolve and adapt practice in dealing with the pandemic.

More immediately there is the little matter of a federal election, likely in the early part of 2022. Timing depends on a finely calibrated calculus of multiple factors - polling, economic data, social trends, the preparedness of the party machine, sentiment, and ultimately a prime minister's gut feel for the public mood. He will announce it when he feels the time is right.

Post the election, what happens next will hinge on COVID-19. If we suffer new waves or vaccine regimes fail, COVID response will remain the key challenge for the public service. If, however, the pandemic diminishes, the present government's fiscal strategy to stabilise and then reduce debt signals an end to the recent trend of growth in agency budgets, and cuts for many. If there is a change of government, there will be a whole new set of programs and priorities to deliver. Under any of these scenarios next year will be even more challenging than 2021.