Why Uber-style public services are not the answer to the burden of bureaucracy

There is an argument for embracing technologies to improve public services – but only if they are carefully managed

A Centrelink office
Centrelink introduced an algorithm that cross-referenced government benefits with people’s tax records, which produced a large number of false positives. Photograph: Julian Smith/AAP

In the film I, Daniel Blake, a middle-aged man is rendered unfit for work by a heart condition. As he undertakes the tortuous process of navigating the welfare system, the film captures the way in which the privatised bureaucracies of the modern neoliberal state are every bit as awful, soul-destroying and Kafkaesque as the government bureaucracies they replaced.

As it stands, too much of what bureaucracy concerns itself with is the monitoring and punishing of ordinary citizens, and modern technologies can make matters worse.

Just ask all the Australian “Daniel Blakes” harassed by Centrelink for debts they didn’t owe because of the introduction of an algorithm that cross-referenced government benefits with people’s tax records. The idea was to detect any undeclared income but, because the data wasn’t adequately “cleansed” – checked for errors in format, duplications and the like – it produced a large number of false positives. This meant people were sent letters demanding they pay back money they hadn’t received.

Senator Jenny McAllister described one example in parliament this month, in which a 67-year-old pensioner was falsely billed $36,000 and had her pension cancelled. The mistake was corrected but it is an error that will keep on happening as long as technology is imposed thoughtlessly in bureaucracies designed to discipline rather than help.

It is a good reminder of something the scholar Mark Fisher argued in his book Capitalist Realism – namely, that any sort of leftwing populism likely to challenge the rising tide of rightwing populism needs to be committed to getting rid of this sort of dehumanising bureaucracy.

But how do you do that?

Technology could help, if implemented properly. A report released by the UK thinktank Reform suggests that robots and other forms of artificial intelligence might be able to replace up to 250,000 bureaucrats within the next 15 years. It makes the point that “the demands on public services are changing rapidly” and that an “ageing population, with increased prevalence of chronic conditions, requires a new way of delivering health and social care”.

All true but, as the Centrelink example exposes, some scepticism is warranted.

For all its quite reasonable analysis, the report exhibits the worst sort of techno boosterism, with a good dash of neoliberal groupthink thrown in for good measure. Not only does it presume job losses are sexy – efficient! streamlined! empowering! – the report is soaked in the hubris that assumes matters of governance can be reduced to something neat and clean like a new online platform.

As noted in an article in Politico, this is a pathology straight out of Silicon Valley: “Whenever the tech world turns its attention to politics, there’s always the hint of this nerdish fascination for system: an inattention to what politics actually is or does but a fetishisation of efficiency, the latent notion that all these 18th-century structures really should just be replaced with something you can download on your phone.”

This nerdish fascination for system is nowhere more apparent than in the report’s suggestion that we introduce Uber-type platforms into government processes.

“‘Contingent labour’ platforms,” it suggests, “may suit hospitals and schools as an alternative to traditional agency models. It may also suit organisations who face seasonal peaks of demand … Using such platforms in the public sector would show its commitment to delivering working practices fit for the 21st century.”

Well, yes, but what is neatly skipped over is that with the rise of contingent labour comes a concomitant loss of wages and conditions. Do we really want hospitals and police forces staffed with Uber nurses and cops (or support staff) struggling to earn a living as they string together various “gigs”?

The report also leans towards more privatisation, suggesting for example that the “efficiencies” they recommend will likely require “strong leaders … drawn from the private sector, to change organisational culture”. But Australia has seen the logic of this approach, where privatisation of employment services has already gone further than it has in Britain.

In the book Getting Welfare to Work, the authors note successive Australian governments introduced into the system “private agencies who were thought to have better links to employers. Those agencies would also have greater scope to decide how to assist each individual in a model designed to give the jobseeker more choice.”

But there was a hitch: because neither the government nor the agencies themselves can guarantee someone a job, they use the only metric of success they have, which is to ensure that the unemployed are “actively seeking work”. As the book notes: “[Job]seeker motivation was to be viewed as the primary driver of outcomes.”

In other words, far from being more efficient, such changes merely enabled the sort of insanity dramatised in I, Daniel Blake and in Centrelink’s robo-call debacle, a system of monitoring and control that becomes draconian.

There is no doubt employment conditions are changing. As the structure of the economy shifts, traditional models of work are being rendered irrelevant, something reflected in the growing levels of under- and over-employment distorting the traditional labour market. Under such circumstances, there is an argument – a leftwing argument – to embrace technologies as a way of improving services. But it has to be done in a way that doesn’t leave workers worse off and airy-fairy notions of turning governments into Uber risk just that.

Still, a system that combined better technologies with a reduction in compliance rules would help and the most obvious way of achieving that is by introducing a universal basic income.

A UBI takes away the endless layers of compliance now demanded (à la Daniel Blake) for anyone applying for welfare. It reduces, almost at a stroke, not just the intrusiveness of the state but the need for an army of bureaucrats to administer payments.

Sure, it can’t simply be used as a way of replacing other social benefits including those around housing, education and health. But, as economist John Quiggin argued: “Social democratic parties need to break with their current role as the responsible managers of the status quo and offer a radical vision for the future. An expanded, and ultimately universal, basic income is such a vision.”

There is no panacea here. Technology alone will not solve our problems. But we need to call the bluff on the neoliberal promise that privatisation and other anti-state measures reduce bureaucratic sclerosis. They don’t, they just change its form, emphasising monitoring and control in the name of efficiency. A system that actually used technology (and a UBI) to reduce this bureaucratic burden would go a long way to empowering ordinary people in a way that leftwing populism takes as axiomatic.