Politics is about trust.
Prime Minister Malcolm Turnbull has been claiming for weeks that Labor's plans for negative gearing would smash house prices.
"The 70 per cent of Australians who own houses will see the value of their single most important asset smashed to fulfil an ideological crusade," he told parliament.
His Attorney-General George Brandis has made it sound even worse. "There is one thing we know about the negative-gearing debate," he told us. "If the Labor Party were to implement its policy, the value of most Australians' homes would collapse".
His assistant treasurer Kelly O'Dwyer briefly said the opposite. Labor's policy would "increase the cost of housing for all Australians; for those people who currently own a home and for those people who would like to get into the housing market".
And then his treasurer Scott Morrison latched on to a "credible report" that said Labor's policy would have "a significant impact on property values".
He latched on too quickly. The report, by BIS Shrapnel, said no such thing. Prices would continue to rise in all but two of the next 10 years under the scenario it modelled, just as they would if negative gearing was maintained. After a decade, they would have climbed 15 per cent. That's less than with full negative gearing, but its still an increase.
The report explained that house prices are typically "sticky in a downwards direction," unable to fall lower than the cost of construction plus a markup. When new attempts at negative gearing were temporarily suspended between 1985 and 1987 real estate prices continued to climb.
While new investors would be less keen to buy if Labor's policy stopped them negatively gearing, existing investors would be also less keen to sell, because they could only continue to negative gear if they hung on to the properties they had. Prices wouldn't be smashed.
It's all there in the report Morrison lauded as credible (because it said rents would rise), but appeared not to properly read.
Certainly his eyes appeared to glaze over the howling error on page one. The report said Australia's national income would average $190 billion over the next ten years when it meant $1.9 trillion.
And they appeared not to be troubled by its suggestion that a measure that raised around $2 billion per year would shrink the economy by $19 billion per year. That's $9 of economic damage for every $1 collected, a sum so big as to be way out of the ballpark of anything his department has ever modelled.
When Treasury modelled a range of taxes for its tax discussion paper, it found the worst of them, stamp duty, did 70 cents of economic damage for each dollar collected. Yet first thing Thursday morning on AM Morrison described as "credible" a report that found removing negative gearing would create multiples of the biggest damage his department could find.
The Grattan Institute's John Daley says the finding doesn't even pass the giggle test. Try it for yourself. Attempt to say: "a tax that raises $2 billion will shrink the economy by $19 billion" without laughing.
What's really odd about the report is its false precision. Limiting negative gearing would create 175,000 fewer jobs over ten years. The unemployment rate would settle at 5.9 instead of 5.8 per cent.
And its woolliness. It assumes away the role of the Reserve Bank in stimulating the demand as economic growth slips, and also the role of state governments in controlling the release of land to regulate the housing market.
The oddest thing is its origin. Who commissioned it? BIS Shrapnel won't say. Why did it release it instead of the client? And was the whole idea to get a gullible politician to swallow and regurgitate it so that the public became even more confused and decided any change was too risky?
It's happened before, in the mining tax debate, in the carbon tax debate and whenever anyone suggests anything that might hurt the superannuation industry.
Economic modelling is the cheapest and dirtiest way to muddy a debate. It lends an appearance of authority to what amounts to guesswork, with key mechanisms often deliberately or accidentally left out. The Australia Institute wants a code of conduct for economic modellers. There's one for auditors and accountants. They'd have to spell out their assumptions and who was paying them.
Right now, with the enthusiastic assistance of people who should know better, we're being had. And we don't even know by who.
Peter Martin is economics editor of The Age.