The disease known as "confirmation bias" is endemic among economists. They have a marked tendency to remember events that seem to confirm the correctness of long-held beliefs, but forget developments that challenge their prejudices.
So their pre-existing notions about how the world works become ever more firmly held.
In which case, let me remind them - and you - of the view-changing lessons about tax reform they probably haven't learnt from the results of recent elections.
Economists, politicians and business people have long-held views about which tax reforms are relatively easy to bring about politically, and which aren't.
But some of these assumptions have been turned on their head by the Turnbull government's poor showing at the election in July, and by the comfortable re-election of the ACT Labor government last month.
In this year's federal election, tax reform was perhaps the biggest single issue. It was, you recall, the one to which the Coalition government was expected to bring a package of comprehensive tax changes, following a green paper and white paper decision process.
The process didn't happen, nor were either Tony Abbott or Malcolm Turnbull willing to propose the much-mooted increase in the goods and services tax.
There was no package as such, just a collection of tax measures announced in the budget brought down just before the election campaign started.
The big one was a plan to reduce the rate of company tax from 30 per cent to 25 per cent, phased in over 10 years, with smaller companies going first and the cuts not reaching big business until 2024.
This was the centrepiece of the government's claim to have a "plan for jobs and growth". To this it added a tiny income tax cut of up to $6 a week for the top 20 per cent of taxpayers, earning more than $80,000 a year.
But the budget included various tax increases to help pay for these tax cuts. It pinched Labor's plan for a further big, phased increase in tobacco excise, and adopted its own versions of Labor's plans to cut back tax concessions on superannuation and extract more tax from multinational corporations.
Labor had been first to put its tax reform cards on the table. It proposed also to phase out negative gearing of property investments and cut the discount on capital gains tax.
The government considered its own measure to reduce negative gearing, but finally decided to do nothing, thus leaving itself free to claim Labor's plan would wreck the housing market. Product differentiation.
But here's our first lesson on the politics of reform: it's a lot easier for governments to propose possibly unpopular reforms when the opposition has already stuck its neck out, or when cabinet has reason to believe the opposition won't attack it for acting.
We can deduce from opinion polling, from the debate during the campaign and from the election outcome how these various reform measures went down with voters.
The tobacco excise increase, the crackdown on multinational tax avoiders and the tiny tax cut hardly rated a mention in the campaign.
Had the small but expensive tax cut not be made, it's doubtful if the Coalition would have lost many votes. Bracket creep is rarely a biting election issue.
The crackdown on multinationals was probably intended to answer the criticism that companies hardly need a tax cut when they were already paying very little, but it could just as easily have reminded voters of this argument.
The super changes attracted little discussion publicly, but did anger some well-lined Liberal supporters. After the election the measures were toned down accordingly.
But it's hard to believe the Libs lost many votes over it when Labor had similar proposals. Nor that many people loaded enough to have a problem with the super changes would have switched their vote to One Nation, as some claim.
It seems pretty clear the cut in company tax wouldn't have gained the Coalition many votes it didn't already have, but probably lost it quite a few.
The public has little sympathy for big business - can't think why - and the claim that the benefit of company tax cuts would trickle down to the rest of us wasn't believed.
Even the government's own modelling showed the belated effect on "jobs and growth" would be minor. For the punters, the link was impossible to see.
By contrast, the government's attack on Labor's negative gearing policy didn't stick and the policy may have gained more young voters than it lost from older property investors.
Economic theory tells us taxes on land are about the most economically efficient - doing least to distort the choices people make about working, saving and investing - of all taxes.
They're particularly attractive when the increasing ease with which financial capital can be moved between tax jurisdictions is used as a key argument for reform, including increasing the GST. Land is immovable.
Land tax is also much fairer - "progressive" - than "regressive" GST, which takes a higher proportion of lower incomes than higher ones.
The value of the land people own tends to be highly correlated with their overall wealth.
But many reform advocates say raising land tax would be even harder politically than raising the GST.
Well, they should note the case of the Labor government in the ACT, which got comfortably re-elected even though it has been implementing a reform long advocated by tax economists: slowly phasing out stamp duty on property conveyances while phasing in a universal land tax.
Of course, there are no controlled experiments in economics, and many factors - notably, perceptions of a government's general competence - play a part in election outcomes.
Even so, this year's elections cast doubt on some supposedly self-evident truths in the politics of tax reform: that company tax cuts won't be a problem, whereas negative gearing, superannuation and land tax are untouchable.
