In recent years, new regulations have been proposed in different parts of the world to re-define the relationship between online service platforms and their workers. These regulations aim to give gig economy workers flexibility in how they conduct their work, and put a limit on how much control platforms can impose over them. Focusing specifically on control over workers’ ability to set their prices and their ability to work across multiple platform apps, this dissertation aims to understand the key operational trade-offs platforms face when choosing how much flexibility to give to its workers. In the first chapter Pricing Control and Regulation on Online Service Platforms, I build a model of a monopolistic service platform to identify when it is in the best interest of a platform to give pricing control to its workers. Taking a mechanism design approach, I identify the optimal fee structure that a platform can adopt to maximize its profits. In the second chapter Multi-homing Across Platforms: Friend or Foe?, I build a model of two platforms where each can control the monetary terms it offers to its users and also decide whether to allow its workers to accept work from the competing platform. I find that some platforms benefit from letting their workers use the competing platform. In markets that exhibit economies of scale, workers’ flexibility to use multiple platforms reduces each platforms’ ability to differentiate themselves through service quality, dampening price competition between platforms. In the third chapter Centralized or Decentralized Pricing With Platform Competition, I build a model of two competing service platforms that can individually choose whether to give pricing control to its workers or not. I identify a key trade-off: when platforms take charge of the prices in the market, they can control for the adverse effects of competition among servers, but they expose themselves to price competition with other platforms. Giving pricing control to the servers makes platforms susceptible to server competition, but dampens the competition between platforms.