Australia's budget is presented to the Parliament on the second Tuesday of May, regular as clockwork.
It is a convention that serves Australia well - a predictable budget time reassures markets and helps businesses and community groups with forward planning.
That is, except - under the present government - in an election year. The most recent budget was delivered on March 29. It was similarly early in 2019, on April 2; just one week early, on May 3, in 2013 (when Scott Morrison was treasurer in the Turnbull government).
What's going on? It seems an unanticipated consequence of legislation designed to promote budget transparency.
The Charter of Budget Honesty Act 1998 requires the secretaries of the Commonwealth Treasury and Finance departments to release an independent report on the state of the budget and the economy within 10 days of the calling of a federal election.
It is called the pre-election fiscal and economic update, or PEFO.
Because of the public nature of PEFO the government has an incentive to bring down its budget as close as possible to the calling of an election. In each of 2022, 2019 and 2016 the election was called shortly after the budget was handed down.
Minimising the time between budget and PEFO helps ensure the numbers in the two publications are as far as possible the same.
Governments do not want to be accused of "cooking the books": massaging budget spending or revenue estimates, or economic forecasts, for political ends. That is frowned on by the electorate.
Australians still prefer their budget documents to be as honest as possible.
There is always politics in a budget's new tax or spending measures. That's expected - they reflect government political priorities.
However, the budget process is designed to ensure the numbers associated with measures - how much revenue they raise or lose, how much of the expenditure budget they cost or save - are as accurate as possible.
The same applies with economic forecasts. They are developed by a joint economic forecasting group; Treasury is responsible for publishing them in the budget.
They are not infallible, forecasts have a margin of error, but as good as the officials can make them based on modelling and data.
The government accepts the official forecasts for the budget; if it were to try to substitute its own forecasts based on political considerations it would suffer a voter backlash.
Nevertheless, the budget is presented by the government of the day.
The names on the front of the budget documents are the treasurer and minister for finance. There is always scope for a government's political opponents to accuse it of leaning on officials to produce favourable forecasts or changing the numbers to understate the true costs of controversial new measures, overstate the impact of tax cuts, and other forms of spin.
The accusation would certainly be made if PEFO had different numbers to those in the budget. This would be presented as "proof" that the government had manipulated the budget figures.
A sad reflection on the combative nature of current politics is that the accusation would likely be made even if there were good reasons for PEFO to be different - for example, if the Australian Bureau of Statistics were to put out new statistics on economic growth, employment, or inflation, that changed the official forecasts and in turn the forward estimates, which are highly sensitive to these parameters.
The government would attempt to explain why the PEFO was different but doing so would chew up precious campaign time. The explanation would also of necessity be complicated, so voters might be left with doubts about the government's budget honesty.
For that reason, it is in the governments interests to have a budget as close as possible to the PEFO to minimise the chance of updates changing the numbers.
It came close this year. Just before PEFO new unemployment numbers were published by the ABS. They were - fortunately for the government - close to those that applied at budget time, so the budget and PEFO forecasts remained the same.
It illustrates why the documents are produced so close together; a couple of months apart and likely the unemployment figures would have diverged significantly.
While the government controls the timing of both the election and the budget, the trend of the budget coinciding with the start of an election campaign is likely to continue (or, if a government chooses to call an election later in a calendar year, with the mid-year economic and fiscal outlook [MYEFO] playing a similar role to the budget).
MYEFO is malleable in terms of timing. The Charter of Budget Honesty requires release of MYEFO annually by the end of January, or six months after the last budget: giving a government many months to play with.
In a normal year MYEFO is in December, but the middle of the year has frequently fallen in November instead, and as early as October under both Labor and Coalition governments.
Timing it to come before an election announcement is relatively easy. The implications for budget honesty are worrying. If budgets or mid-year updates are institutionalised as a precursor to an election campaign, officials will come under more pressure to adjust the numbers to favour the government of the day, at the expense of budget honesty.
The problem would of course disappear if the Commonwealth, like a majority of states and territories, had fixed-term elections; it's been proposed in the past, but does not seem a prospect any time soon.
Under current arrangements, our best safeguard is a professional public service free of political appointments and undue ministerial influence.
Sign up for our newsletter to stay up to date.