That new E-tax for Mac application from the Australian Taxation Office (ATO) is hilarious. Its failure to use a developer certificate — meaning that OS X running under its default security settings would reject it as suspicious software from an untrusted source — seems to indicate poor testing. Surely, trying to install and run the app on a stock OS X installation just once would have detected this. But that's just the start.
The ATO's reported AU$4.7 million spend has delivered an old-fashioned clunker that's so unlike actual OS X software that it looks as out of place as a stripper at a funeral.
The alien appearance comes from it being a straight port of Windows software. It arrives as a Windows-style installer rather than a .pkg file, for instance, and that installer asks whether it should install into the Applications directory or somewhere else. Mac users don't expect to be bothered with such things. And don't get me started on the arrangement of the menus.
But E-tax is also infested with usability glitches. The installer asks which version of the software to install, for example, when there's only one.
My favourite, though, is when E-tax asks for my Tax File Number (TFN). I copied my TFN from my secure records, where it was written in the usual format "123 456 789", and pasted it in. That was rejected: "Error. Result contains invalid characters." Normal humans find that sort of language completely opaque. Still, I re-typed my TFN without spaces. It was accepted — and then reformatted to display with spaces again!
I could go on. There's much to laugh at in E-tax for Mac. But let's make some serious observations.
The big question is why E-tax is being done as downloadable applications specific to each operating system.
That approach would have made sense when the ATO started the E-tax project, back in the days of dial-up internet. A downloadable app could do all the grunt work locally, only transmitting the completed tax return at the end as a small chunk of data.
But these days, we have things called "broadband" and "web applications", making everything independent of the user's operating system, and the security problems are well understood. Dial-up internet is all but consigned to history, and even banks use web apps — and they tend to think about security.
Maybe it's a budget problem. While AU$5.2 million seems a lot for porting an existing app, that figure also includes support. But I wonder whether cultural factors are at play — and, specifically, a tendency for government agencies to fall for the sunk cost fallacy.
Sunk costs are money that has already been spent. The sunk cost fallacy is when you think that because you've already invested a lot of money in a project that then runs into problems, you need to spend more money to fix it, to prevent "wasting" the money you've already spent. The fallacy is that the money you've already spent is already gone.
The clear-headed way forward is to look at the situation as it stands today, and assess how to most effectively reach your goal. Maybe that means dumping everything you've done so far.
A classic example of the sunk cost fallacy at work is Tcard, the failed electronic ticketing system for Sydney's public transport system. Originally intended to be used during the 2000 Olympic Games, it still wasn't finished a decade later — and eventually, it was scrapped. It should have been scrapped much earlier.
Government agencies like the ATO are particularly prone to the sunk cost fallacy, for a number of reasons.
They tend to have a more consensus-based management style, rather than a commanding leader with the power to pull the plug. Their communication style is indirect, with reports speaking of "continuing difficulties" rather than "unmitigated disasters".
No public servant wants to end up fronting a Senate Estimates committee to explain why they "wasted taxpayers' money". No government minister wants to give their political opposition the opportunity to ask the same question. Neither wants to be seen to be making "courageous decisions", as the euphemism goes.
I also wonder whether the best and most forward-looking programmers have heads filled with Silicon Valley dreams of becoming billionaires, and would rather work in the stereotypical tech startup, rather than in government. Put less politely, I wonder whether the government doesn't end up being stuck with the B Team most of the time. After all, it doesn't offer the most lucrative salaries.
Whatever the cause, in the case of the ATO's E-tax program, millions of taxpayer dollars have been spent on something that looks out of date on day one. If that embarrassment is all that existing processes can deliver, then clearly, the processes need changing.