Conventional wisdom views authoritarian legislatures as 'rubber stamps'. According to this model, non-democratic parliaments are entirely subservient to dominant executives, having no influence on the development of policy; as a result, all bills introduced into the legislature become laws without amendment. Although these bodies might perform other functions, they serve - according to this account - a purely ceremonial function in the policy-making process. There is evidence, however, inconsistent with this portrayal from a range of non-democracies, including evidence of executive bill failure and bill amendment. Existing attempts to explain these apparently deviant observations refer to some degree of legislative autonomy - bills fail and change as a result of legislator influence. According to these accounts, authoritarian elites use legislatures to co-opt members of the opposition and to gather information about citizen grievances. This dissertation, in contrast, argues that legislative activity in non-democracies can be driven by executive concerns. Whereas the 'rubber stamp' model infers from executive dominance an absence of legislative activity, the approach proposed by this dissertation suggests there are a variety of reasons why executive actors might want to amend or kill off their own bills in the legislature. In particular, these legislative policy developments can result from clashes between executive factions, which use legislative institutions to monitor, challenge, and amend each others' proposals. This dissertation proposes and assesses this new approach using fine-grained data on legislative processes and outputs from the contemporary Russian State Duma. The analysis draws on a variety of data sources, using both qualitative and quantitative methods. The findings suggest that legislative institutions can still 'matter' in non-democracies, even with an entirely subservient body of legislators.