The one-hour week and other damned lies in our job statistics


By: Dr Nick Dyrenfurth

For too long front-page statistics – the ones that affect people’s lives – have been hiding our real economic story.

Consider the nation’s headline statistic regarding who is looking for a job, the monthly and quarterly unemployment rate published by the Australian Bureau of Statistics.The latest ABS unemployment data is ostensibly good news.

The December 2018 rate fell to 5 per cent, the lowest since June 2011, on the back of what some term a “buoyant jobs market” .

ABS figures should be treated with caution; they are a “survey” based on a sample of 50000. Unemployment is likely to curve upwards next month, or, if it falls, be driven by lower labour force participation, and population growth.Full-time work fell by 3000 in December, while part-time work rose by 24,600. The long-term trend growth is geared towards the “gig economy”.

Less than half of workers hold a full-time job. Casuals, part-timers and contractors are on the rise – denied job security, sick leave, holiday pay, and superannuation. All on top of stagnant wages growth.

The Uberisation of work can be discerned from other data. Total hours worked rose by 1.5 per cent in 2018. Yet hours worked per person declined. Underemployment – employees working but who would like to work more hours – is rising and reached 8.4 per cent in December. The under-utilisation rate — unemployment and underemployment combined — was steady at 13.3 per cent.

The correlation between unemployment and underemployment – historically moving in tandem whereby the latter is two points higher than the former – is weakening. The differential is now about  3 per cent. As economics writer Greg Jericho has warned, this indicates a permanent structural shift towards higher underemployment: bad for young or old, male or female. Youth unemployment is unacceptably high.

There are three kinds of lies: lies, damned lies, and statistics. It’s a familiar refrain. Yet people pay attention to the ABS’ published unemployment data. Unemployment down is “good news”; unemployment up, not so much.


How we measure and talk about unemployment matters. Yet many politicians, pundits and the public remain fixated on pure data, measured by a near 60-year-old International Labour Organisation standard – if you work for at least “one hour” a week you are “employed”. Granted, this data provides us with internationally comparative and historically trackable data. The ABS acknowledges one hour’s work a week is “insufficient to survive on”. Yet the idea that unemployment is 5 per cent and employment is equal to one hour’s work is laughable. Real fake news.

Our economy and society have changed dramatically since the 1960s. Yet the labour force statistics and the ways they are used by politicians and media haven’t. ABS figures don’t really take into account the increasing divergence between the new “haves” and “have nots” in the Australian economy. The “haves” enjoy secure full-time employment and the financial and emotional benefits it brings; “have nots” are subject to vagaries of insecure work, underemployment, and lower pay packets.

Statistics matter in view of declining trust in our institutions. When the perception of reality on the street is so out of step with the Canberra beltway, cynicism and anger grows. It breeds the sort of nihilistic fury that saw Donald Trump elected.

Federal Labor leader Bill Shorten put it well in his speech to the party’s 2018 national conference. His party’s biggest challenge was not simply fighting the Coalition, but restoring faith in democracy. “Our deeper opponents are distrust, disengagement, scepticism and cynicism,” he said. “Our Labor mission is not just to win back government; it is to rebuild trust in our very democracy, to restore the meaning to the fair go.”

It’s time to talk honestly about unemployment data and insecure work, how we measure it, and whether it is fit-for-purpose in 2019.

The one-hour-a-week definition of employment should be swept into the statistical bin of history. Any new definition of employment must be based on the ability of a citizen and her or his dependants to earn a “living wage”, not some outdated “minimum wage”. Rather, it should be a version of our nation’s famous 1907 Harvester judgment, that a “fair and reasonable” wage must take into account the “normal needs of the average employee regarded as a human being living in a civilised community”, crafted for this precariat age.

Our job statistics are broken. It’s time to #changethestats.

(Visited 6 times, 1 visits today)

Publications Views: 152

Tocsin Special Edition

By: The John Curtin Research Centre

Read More

Newsroom Views: 42

Liberals future must rest on recognising the past

By: Nick Dyrenfurth

Read More

Publications Views: 925

Tocsin #20 Special Edition

By: The John Curtin Research Centre

Read More

Donate to the John Curtin Research Centre

At the John Curtin Research Centre, we rely on the generosity and support of individuals like you to advance our mission of policy development and advocacy.