The article will be open access in English, Spanish, and French, and I’ll share it here once it’s out - stay tuned!

I explore how Spain’s efforts to bring platform couriers into standard employment relations have affected their working and living conditions (#LeyRider but not only), and how workers’ representatives and public authorities have (not) been able to leverage the new right to #algorithmic transparency.

health outcomes (provided there are non-tech risks also going on), but it is impossible to say such pattern is causal or just correlational only. In any case, the best combination seems to always be no AM + no other risks.

Thanks Tyler. In my qualitative studies, no. The “best” integration is the one where workers do not perceive AM as enhanced managerial control but act in ways that are detrimental to them anyway. In the quanti approach, things change. In some cases AM presence seems to offer better occupational

Curious about the whole dissertation? Don't hesitate to drop me a line – I will gladly share it!

#PhD #algorithmicmanagement #AI #employment

This means that we cannot simply assume what will come out of the adoption of these tools. Importantly, however, in the specific case of occupational health, the potentially positive effects of algorithmic management cannot compensate for the bad (non-technological) practices of managers.

5) Finally, looking at the specific case of occupational health and safety (again through a large-N analysis), I show how the outcomes from deploying algorithmic management are not pre-determined. Rather, they depend on pre-existing or parallel non-technological characteristics of management.

In such cases, managers and employers’ authority ends up being reinforced by the internalisation of the meritocratic ideals and notions of machinic superiority so familiar to the development of AI.

They experience it not as a form of insidious control (which it is!) but as a tacit incentive to improve their work performance.

this event is largely contingent on how algorithmic brokerage is conducted by managers. In some instances, algorithmic management may lead to the intensification of work rhythms, increased surveillance, and decreased autonomy of workers, but – nonetheless – be internalised by workers.

This suggests there's a perception of violation of norms by employers and, thus, resistance emerges. But is this the whole story?

4) No, things can be more complicated. Here's how: despite some workers resisting (or at least showing dissatisfaction) with algorithmic management,

But is this accepted by workers?

3) To a large extent, probably not. As I found through a large-N analysis (and confirmed in qualitative case studies), on average, the presence of algorithmic management is associated with various forms of workplace turbulence.

2) A crucial part of the change this brings to the employment relationship is that it becomes a further uneven playing field, marked by deeper forms of information asymmetry between workers and managers/employers. This paves the way for (more) opaque decision-making by managers.

Now, this doesn't mean there's no change. Rather, instead of algorithmic management equating to the automation of managers, it transforms them into algorithmic brokers (or translators of algorithmic output), paving the way for new forms of human-machine complementarity.

Concisely, my findings can be summarised in five points :

1) Contrary to platform work (where much of management seems to be run by algorithms), in conventional employment settings, things are very different. This means that human managers are still present and actually quite salient.

*four, obviously 🥲

Hopefully, the coming months will bring more pieces on #platformwork and #algorithmicmanagement (also in conventional hashtag#employment settings) that people can read and enjoy.

If after all this you are curious to read the article, you can find it here: onlinelibrary.wiley.com/doi/full/10....

I suppose like any article that one writes and some years later goes back to, I would now do a lot of things differently (starting from finding a much cooler title - jeez, this one is really boring!), but it is still nice to know that people find it interesting and useful.