Are employers, not robots, to blame for job losses?

Newspapers and conferences are abuzz with how machines are taking over the world of work. The predictions about how many jobs will be lost or skills rendered obsolete are often dire; some suggest that more than half of today’s jobs could be lost to automation, for example.

The problem with much of this analysis, however, is that it glosses over how machines have no agency. People make decisions about who stays employed and who does not. Rather fixating on the future of work as technology versus humans, it is more useful to think about how and when employers may adopt and adapt to technology, and what that means both for them and for their workers.

Consider the statement that ‘robots will steal our jobs.’ The problem is that this phrasing misses out who buys, installs, operates and maintains those robots. A more explicit formulation could be, ‘employers will use robots to replace human workers, taking away their jobs.’ Already, this statement adds multiple dimensions of complexity. The result? Clarity that the decision to cut or create jobs is not made by the robot, but by an employer.

The story of technological unemployment is then about more than the unilateral entry of technology into our lives. It becomes about politics, profits, negotiation, emotion, and culture. It becomes about human choices, not some abstract or universal technological tsunami.

Employers matter

Making this explicit also has an important function for the ongoing public policy debate. This construction positions employers—who have been a missing middle in the conversation about the future of work—as the channel through which jobs will actually be created or lost. This allows us to go beyond the horizontal public policy responses to thinking more specifically about if and how governments might support employers.

Such support might be needed because even though technology is now available almost anywhere, its diffusion is unpredictable and varied. New devices, tools and techniques now rapidly spread around the globe thanks to the communications and trade links. But even if a technology reaches a country, its diffusion within that economy might take much longer.

A mix of competitive market pressures (at the local, regional, or global levels), information and awareness, financial and regulatory considerations (costs of labor, profitability), and social factors (the emotional attachment to workers, interest in retaining traditional techniques) will drive an employers’ decisions about how they adopt technology. The literature on the diffusion of innovations among businesses has studied the range of reasons well.[i]

What this means is that technological advancement and technological adoption diverge. Hence, employers—and their workers—will (and do) adopt technologies at varying speeds for many reasons. Some firms will be early adopters while others will be laggards, to use Everett Rogers’ terms from the theory of the diffusion of innovation.

We might also recognize that while some employers will chose to wait, others might be forced to wait. Moreover, it might be that employers who are better connected—in every sense of that term—would access, adopt and benefit from technology. Others might be left behind, in spite of their interest. The result could be a bifurcation of markets, where monopolies are created or sustained, and others are disadvantaged.

The trends and divergences in how businesses adopt Internet technologies are telling. As of 2013, more than 85% of small businesses in Slovenia, the Czech Republic and Slovakia had their own websites, compared with 20% in Nepal and about 12% in the DR Congo or Madagascar. Certainly, this is because of differences in access to technology, with developing countries lagging richer economies.

But even in developing countries, between 80% and 90% of larger businesses have a websites (although only 38% in DR Congo). Even with email, accessible to anyone with a feature phone, the gaps are evident across countries and across firm size especially in the developing world. This means that differences are due to more than access (see below).

The effect on employment will thus vary significantly. One furniture manufacturer might sack most of their designers first, replacing them with computers and 3D printers; another might let go of most shop floor workers and replace them with robots. Some might outsource, while others might re-shore as costs and preferences evolve. And some might be left behind.

How might public policy respond?

Governments will need to accelerate efforts to get the business climate right through appropriate regulation and investments, in particular with measures to promote and protect competition. But going beyond those horizontal efforts, there might be a case to be made for firm-level interventions to help them be more aware of technological advancement and adapting. This means providing employers—for example, those in the informal sector, in rural areas, or poorer settings—access to infrastructure, to finance, and to business development support. These efforts could build on well-tested SME support programs, but with a sharper focus on enabling employers to overcome existing and future technological divides.

Until now, we have rightly focused on how workers might better prepare for the future. But we must also think about preparing employers, and deciding what role public policy might play in assisting employers and markets evolve to remain competitive and grow, creating new jobs. Preparing better for the future of work will mean bringing more of the world into it.