Why isn't Uber worker-managed? : a model of digital platform cooperatives

Thanks to algorithmic management, the digital platform sector does not require sophisticated governance structures and labour intensity tends to be higher than in traditional sectors. So, why aren't usually digital labour platforms worker cooperatives? We develop a simple model to study the comparative viability of a worker-managed (WM) via-app labour platform firm vis-à-vis a capital-managed (CM) counterpart. Firms compete over workers by choosing the optimal size and (CM firms only) the pay policy. Given the size of the market, we show that WM platforms maximize per-capita incomes over a middle range interval of firm size. At the equilibrium size, viability of WM firms may be impeded by the costs of the external capital, no matter how low, which enable CM firms to pay a wage premium. The worker payoff in CM firms is higher in the presence of higher unit revenues and network effects (which improve the ability to pay of WM firms, thereby stimulating pay competition between platforms) and lower when WM platforms need to charge new members a fee to overcome free-riding problems faced by those who fund the initial investment. The model also shows that the conditions for worker buyouts are weaker than those required for WM platform creation from scratch, and that group incentive mechanisms allow WM platforms to better pursue quality improvements than CM firms, when digital techniques make the cost of effort relatively low.