Beschreibung:
The practice of outsourcing labour and vehicle capacity to independent freelancers in ride-hailing is known as ridesourcing. We present a conceptual representation of the interaction between supply and demand in the ridesourcing market to understand why these markets may be prone to evolve towards particular - potentially socially undesirable - equilibrium states. In addition, we propose an agent-based model that allows simulating the effect of market conditions and platform strategies on system performance. This day-to-day model accounts for the sequential decentralised processes that characterise both sides of the two-sided ridesourcing market: (i) information diffusion, (ii) platform registration, (iii) platform participation, and (iv) learning based on experience. We apply the model on a case representing Amsterdam, the Netherlands. We show that there are network effects in the ridesourcing market associated with both matching time (time between requesting / dropping off previous passenger and being assigned) and match quality (time between assignment and pick-up). An equilibrium state with matching time asymmetry - i.e. a market that is either considerably over- or undersupplied - yields high-quality matches, mitigating matching time disutility on the competitive side of the market. Our simulation results suggest that a profit-maximising ridesourcing platform may trade-off market transaction volume for higher earnings on successful transactions, a strategy that is harmful to the interests of travellers and drivers, and possibly of (very) limited benefit to the platform. Moreover, we find that ridesourcing operations may be viable even when potential supply and demand in an area are limited