Participatory budgeting has been widely discussed, and often celebrated in scholarly literatures on public policy and democratic practice. It is now a part of governance in at least 1,500 cities worldwide. Yet the enthusiasm for it belies important democratic challenges, particularly, those tensions arising between resource distribution through participatory budgeting and through existing budget allocation mechanisms. This tension is more clearly resolved when the latter process is compromised, for example, in the presence of corruption. But a puzzle emerges when existing budgeting processes are functioning well. Why do politicians supplement representative institutions with those more directly dependent on citizen priorities? Focusing on the case of New York City, where capital projects related to school investment are chosen and funded through a Participatory Budgeting process, my argument considers how citizens might reward politicians based on perceived competence, leading to politicians with strong reputational incentives to signal competence rather than implement genuinely beneficial policy. In the years following a major educational reform designed to strengthen accountability within the New York City school system, I use a difference-in-differences approach to analyze how citizen participation mitigates the extent to which certain decisions are governed by traditional accountability mechanisms. Drawing on data from 1,821 NYC schools, 2,181 participatory budgeting proposals, and 97,001 earmarked projects spanning 2009 to 2020, I demonstrate that school performance does not significantly influence funding allocation through participatory budgeting. This finding suggests that participatory processes may retain legitimacy in environments characterized by heightened accountability pressures.