Monthly Archives: August 2016

Destroying dreams: #Censusfail reveals the danger of penny-pinching

ImperialstarDestroyer480ppx (1).png

With #censusfail shaping up to be another case study in the false economy of cut-price outsourcing, it reminds us of the need to balance price considerations with quality and risk considerations.

Not everyone thinks like this.

Back as a Finance Department grad in 2009, I knew what my job was: a funding destroyer.

Ian Watt, the then Finance Secretary, publicly described the Finance role as “the destroyer of dreams”. Because, he said, “most those dreams are not worthy of public funding”.

Perhaps drawing on my childhood upbringing in austere Protestantism, I took the calling up with a righteous zeal, meticulously trawling over every costing line item to expose and challenge spending that had a sniff of being surplus to immediate need. It was an approach that won kudos and status as a Finance grad.

You can imagine the departmental culture. In fairness to Finance, their internal ethos develops as a kind of battlefield solidarity necessary to resist the continuous onslaught of rubbish that comes across their desk: bullshit thought-bubbles, cynical pork-barrelling, naked rent-seeking and shameless portfolio empire building. A classic of the genre came from John Howard, as he desperately tried to buy himself the 2007 election, throwing $250,000 at some bloke who wrote in asking for money to make a golden carriage for the Queen (no joke).

Agencies adapt to Finance’s punctilious penny-pinching by becoming even more crafty in their funding proposals, which Finance adapts to with even more vigorous cynicism and reflexive nay-saying. The resultant feedback loop leads to a subgame Nash equilibrium of distrust, in which excessive and even sneaky funding requests are countered by an agency mentality that consciously stylises itself with apocalyptic dream crusher associations sounding like they’re straight off a Megadeth album.

(When I say distrust, I mean at the margin. Australia is still much better than most of the world on public servant social norms.)

Anyway, I moved off to Treasury the next year and accordingly my strategic focus shifted from the thousands to the billions. Since this time,  I’ve grown increasingly conscious about the dangers of underfunding projects of high priority and high risk. Sometimes it’s wise to build slack into a project if the stakes are high enough. Alas Finance-style thinking sometimes fails to see the forests (social output, catastrophic risk mitigation) for the trees (saving a buck).

One policy risk with extreme stakes is the potential for the community to lose trust and confidence in public institutions, such as the Census. The Census is crucially important given its role informing policy-making. The accuracy of ABS statistics depends on the public trusting the ABS enough to provide honest answers.

But the importance of institutional trust may run far deeper than this. The thesis of renowned MIT macroeconomist Daron Acemoglu is that institutions – including informal institutions like trust – are the central driver of a society’s long-run economic growth.  In particular, a society’s political and economic establishments must be trusted and perceived as legitimate by the community for that society to function efficiently.

A loss of trust in government is remarkably persistent. A working paper from late last year by Yuyu Chen of Stanford found that Chinese families who were affected by the Great Chinese Famine still – half a century later –trust government substantially less than do other Chinese families. The distrust carried across generations. The paper’s methodology is still being discussed and debated among economists, but given the stakes involved, it would be wise to err on the side of caution when it comes to preserving trust in the mean time.

In a similar vein, the citizens of my current home – Indonesia – remain deeply sceptical about economic reform after mismanagement by the IMF during the 1997-98 crisis. “Neo-liberalism” is to this day regarded as a dirty word across walks of life here. We’ve also seen disenchantment in Greece after the troika continuously promulgated blatantly wrong forecasts overstating the benefits of austerity and economic reform.

As I discussed in my post on Indonesia and Greece, it’s all too easy for wonks to sneer at the irrational masses, exuding over-confidence in their own expertise while forgetting the critical imperative of maintaining perceptions of democratic legitimacy. Trust in experts is damaged when the experts turn out to be wrong, but especially when they’re wrong on policy changes that are rammed through without the community having a voice in the decision-making process. That’s a theme that applies to the Census changes, but I digress… Back to underfunding.

Given that a reputation takes a lifetime to build but a moment to destroy, experts should err heavily on the side of averting the risk of threats to public confidence. Building in excess capacity works as an insurance device against the unexpected. Nassim Taleb – who is 75% ranter now but formerly 50% ranter/50% genius – illustrates this intuition well:

Let me summarise my ideas of how Mother Nature deals with the Black Swan. First, she likes redundancies. Look at the human body. We have two eyes, two lungs, two kidneys, even two brains (with the possible exception of company executives) – and each has more capacity than is needed ordinarily. So redundancy equals insurance, and the apparent inefficiencies are associated with the costs of maintaining these spare parts and the energy needed to keep them around in spite of their idleness.

The exact opposite of redundancy is naive optimisation. The reason I tell people to avoid attending an (orthodox) economics class and argue that economics will fail us is the following: economics is largely based on notions of naive optimisation, mathematised (poorly) by Paul Samuelson – and these mathematics have contributed massively to the construction of an error-prone society. An economist would find it inefficient to carry two lungs and two kidneys – consider the costs involved in transporting these heavy items across the savannah. Such optimisation would, eventually, kill you, after the first accident, the first “outlier”. Also, consider that if we gave Mother Nature to economists, it would dispense with individual kidneys – since we do not need them all the time, it would be more “efficient” if we sold ours and used a central kidney on a time-share basis. You could also lend your eyes at night, since you do not need them to dream

Taleb overdoes the economist bashing (everyone bashing) but, as a description of some government budget guardians, the comment rings true.

The thing about the #censusfail is that it wasn’t even a Black Swan, it was quite predictable. You don’t need to have any deep Talebian scepticism about prediction; even a ‘naive’ analysis should have encouraged caution.

It’s easy to see how penny-pinching contributed to the #censusfail disaster.

  1. The ABS evidently lacked the funding for a proper public consultation and information campaign about sensitive changes. The changes were announced late in the five-year window, and the timing of the announcement suspiciously resembles a strategy of hoping the community wouldn’t notice. Given the inherently controversial nature of changing the parameters of government compulsion, the apparent lack of transparency and failure to engage the public in the decision-making process gave the appearance of pulling a swifty on the community. As such the public smelled a rat (even if there was no rat). In truth, the ABS would not have had much option if they couldn’t afford any serious public consultation and advertising.
  2. I suspect the ABS consciously avoided raising awareness about the shift from paper to online because they wanted to minimise printing costs.
  3. The website crashing on the one day it really needed to work, despite a predictable amount of traffic, suggests insufficient capacity to handle a small variation from expected load due to bare bones funding. [Update: apparently it failed to withstand some form of attack.]

Unfortunately, the ABS’s credibility is now in tatters. The fervent claims from the ABS about its robust data security become questionable after their newly established record of bungling critical IT projects. In a context of information asymmetry between government and citizen, it is rational to assume that if a government underfunds a publicly visible output, it is even more likely to underfund an unseen output, such as information security.

The ultimate responsibility for #censusfail does not lie with any public servant, but rather with the elected representatives within government. It’s the politicians who ultimately make funding decisions, and it’s the politicians who pressure economic agencies to identify savings. Nevertheless, it’s not hard to imagine how the Finance ethos interacts destructively with the current government’s ethos (cut public funding, except to the rich). They reinforce, rather than check, each other’s biases.

Spending even as much as $500 million more on the Census would correspond to only 0.0061% of Australia’s national income over the past five years (That’s a rough estimate – I tried to get precise GNI statistics but unfortunately the entire ABS site is still down…). Presumably that would have averted the disaster. It is value for money to allocate a small fraction of our income to data that is crucial for improving the quality of our national public policy. Unfortunately, the government must cut corners because it has failed to wisely allocate its spending and/or has ideological problems with increasing public revenue.

Let’s hope this is a lesson. Stop the false economy of penny pinching to the detriment of vital public institutions.