Inequality in Australia: four ways technology is creating a new divide
Originally posted by Nicholas Davis @ abc.net.au
How we analyse and talk about inequality reveals as much about ourselves as it does the state of the world.
In the Committee for Economic Development of Australia's (CEDA) latest research report, How Unequal? Insights on Inequality, Nicholas Rohde and Lars Osberg argue that different approaches to measuring inequality can lead to different views on its level and extent.
However, I'm certain of one thing: In order to appreciate the future evolution of inequality in Australia, we need to be equally conscious of the political nature of emerging technologies, given that we are currently using them to reshape entire industries, transform labour markets and change the way we relate to one another — and even see ourselves as human beings.
Five decades of rising capabilities and falling costs in digital communications, information storage and data processing have given rise to a number powerful, multi-use technologies, such as machine learning algorithms, secure and distributed forms of data sharing and management, advanced materials, biotechnologies and neurotechnologies, among many others.
At the World Economic Forum, we call this the Fourth Industrial Revolution: a period of economic and social transformation on par with the industrial revolutions that ushered in the factory system, mass transport, electricity and sanitation and the digital economy.
These technologies could significantly impact inequality in at least four different ways.
Automation threatens low-income workers
First, Australia could experience widespread unemployment driven by the rapid adoption of artificial intelligence and other new and powerful technologies which promise to replace labour with ever-smarter capital.
The 2015 CEDA report Australia's Future Workforce? featured work by Hugh Durrant-Whyte et al, who found that "40 per cent of current jobs have a high probability (greater than 0.7) of being computerised or automated in the next 10 to 15 years".
The timeframe — which is almost always the least certain aspect — is actually a crucial component. It's not job destruction per se that we worry about. Rather it's the speed at which it occurs, and the relative rate of job creation that matters.
Inequality would, of course, rise if extended periods of technological unemployment overwhelmingly impacted lower-income workers — something that most studies find.
Second, the widespread adoption of new technologies across industries will affect inequality if their productive use requires skills that are available only to certain segments of the population.
Laura Perry points out that Australia's primary and secondary education systems are far less equal than we realise, particularly for Indigenous students, students from lower socioeconomic status backgrounds and students who reside in rural/remote areas.
So-called skill-biased technological change would exacerbate educational inequalities and lower the value of workers and jobs that lack technological complementarity.
Third, more productive capital in the form of robots, algorithms and automated factories could further shift returns in the economy away from workers, recalling Thomas Piketty's theoretical framework that reignited the global inequality debate in 2014.
As Laura Tyson, chair of the Council of Economic Advisers under former US president Bill Clinton remarked, the critical question quickly becomes, "who owns the machines?" If ownership of smart capital is concentrated, wealth and income will follow.
Technologies encode our biases
Fourth, while we may not think of highways, smartphones or algorithms as inherently political, they are.
Technologies encode our biases, they incorporate and reflect our values and reveal how we think about the world. They also shape our behaviour and determine the opportunities we have and the outcomes that are available, often for generations after.
Given how powerful today's emerging technologies, denying this aspect of their nature could be particularly dangerous for future inequality.
Langdon Winner, who famously made this point in his 1980 article Do Artifacts have Politics? gives the example of how building low highway bridges in New York meant that poorer communities were denied access to Long Island pools and beaches, a reflection of the racial bias of the master planner.
More recent examples are offered by Virgina Eubanks in her book Automating Inequality, where the use of algorithms to decide who gets welfare are applied without thinking about how marginalised populations are affected.
The Centrelink automated debt recovery scandal rhymes with Eubanks' harrowing case studies, indicating how procedural fairness can be undermined for those least able to assert their rights. The Cambridge Analytica controversy, meantime, highlights how globally significant political and social outcomes can be influenced via digital platforms.
Technology governance being tackled
I'm confident that Australia can manage all four sources of technologically driven inequality.
Focusing on augmenting rather than replacing workers, co-investing in buffers for the most affected workers, innovating the way we educate, devising new ways to define and collect tax in a new economy, and addressing the sources of discrimination in machine learning — these are all areas where we can draw on historical Australian experiences of reform and innovation.
As the world wakes up to the implications of the Fourth Industrial Revolution, governments, civil society organisations and multinationals are launching exciting experiments in technology governance to tackle similar challenges. Indeed, the World Economic Forum's Centre for the Fourth Industrial Revolution is a global resource in this regard.
However the hardest aspect is likely to be the politics.
None of these solutions can be realised by a single party, sector or organisation alone. Yet, to work together productively, to grasp the benefits and manage the risks associated with technology, we have to be open and honest about both shared and conflicting values.
It's time to step up constructive political debate on current and future inequality — not to prove anyone wrong, but rather to make sure we set in place systems and technologies that are both perceived and experienced as giving both current and future generations a fair go.