When algorithms go to war in the workplace

Businesses crunch data to gain power; workers should bend it to their own ambitions


One constant in public debate is the assertion that the world of work is on the cusp of unprecedented change.

Amid the hype, one genuine source of flux is the manner in which data and technology are combining to alter workplace management and control.

To date, this has been a tale of the good, the bad and the ugly. At one end of the spectrum are those bringing fresh rigour to “people management”: multiple data sources interrogated to challenge the false assumptions and biases that often distort personnel decisions. Then there are those dressing up standard human resource insights as if they are part of the “big data” revolution. Worst are those touting sinister schemes for controlling worker behaviour ever more tightly: mood swings monitored, toilet breaks timed and social media scrutinised.

The rise of the tech-controlled workplace matters. From the voice in our ear telling us which task to perform next to the software determining who gets on a shortlist or works an extra shift, data-driven decisions are shaping our working lives.

Yet the algorithms responsible can only ever be as good — and fair — as the numbers funnelled into them. And this has been subject to very little scrutiny. The questions are not only whether workers will push for better safeguards, but whether they may also go on the offensive, using data imaginatively to tilt power their way.

It should, after all, be relatively straightforward to equip workers with personalised real-time information on wages and prospects taking account of age, stage, skills and sector. The large informational advantage that employers hold over workers in pay discussions could be eroded. Worker-friendly algorithms might nudge us when the time is ripe to push for a pay rise, look for a new role, or move to a different platform with lower fees. Nor would it be that surprising if those at the sharp end of the jobs market generate ever more of their own data to highlight the people and practices that cause workplace misery.

The future of collective action could also be altered. Data could be used to calibrate the optimal moment for workplace disruption and “go-slows”. Post-industrial disputes could take a digital turn: algorithm v algorithm.

Whether this comes to pass depends on institutions emerging that workers trust to defend and deploy their information. Might this offer a 21st-century lifeline to trade unions whose roots lie in the 19th century?

The UK Trades Union Congress is exploring aspects of this agenda. To get anywhere it will need to push hard since to date there has been little recognition from most union leaders that even the data on their 6m-plus members could be a precious resource. Scepticism runs deeps about the value of workers seeking to bend data to their own ends.

To some degree that is understandable. Information is not the same as power. A sea of data will not do much for the zero-hour contractor toiling for an unscrupulous employer who withholds shifts as a disciplining device. What is known as “digital Taylorism”, an updating of the management techniques propounded in the early 20th century by Frederick Taylor, needs to be combated by traditional means as well as new ones. Data-activism will not remove the need for collective organisation, but it might give it a new lease of life.

For now we are witnessing a one-sided contest. Business is perpetually crunching numbers to help gain power, both in the market and over workers. Organised labour is only starting to stir. The opportunity for unions is closing fast. They also face a demographic crunch: to survive they need younger recruits.

That means winning over a generation who are economically precarious, tech-savvy and socially networked. It is a group that urgently needs some smart workplace muscle.

Originally published in the Financial Times