This dissertation examines how nonprofit organizations and local governments allocate resources under different constraint configurations. The three essays test how externally imposed, externally absent, or self-imposed constraints shape recipient allocation behavior, drawing on conditional grant theory, principal-agent and stewardship theory, and new institutionalism. The essays contribute to theoretical insights in nonprofit management and local government finance, with policy implications for ongoing debates concerning trust-based philanthropy, conditional grants, and municipal fiscal rules. The first essay examines how nonprofit recipients of a conditional grant targeting an operational input allocate resources across targeted and non-targeted spending categories. Using a doubly robust difference-in-differences estimator on a panel of 7,114 U.S. nonprofits from 2015 to 2023, with the Paycheck Protection Program as a natural experiment, the essay finds that PPP receipt produced positive effects on payroll (the targeted category) and on nonpayroll program spending (a non-targeted, mission-related category), with no comparable spillover to management or fundraising. These findings are consistent with the fungibility prediction and the nonprofit-specific institutional features that channel freed resources toward mission-related uses. The second essay examines how nonprofit organizations exercise discretion through unrestricted grants. Using system Generalized Method of Moments estimation on a panel of arts and culture nonprofits from 2009 to 2023, the essay tests competing predictions from stewardship theory and principal-agent theory. Unrestricted grants are found to be positively associated with administrative compensation, concentrated in compensation of existing personnel, rather than with increases in program spending, organizational capacity, workforce, or program reach. The third essay shifts gears to self-imposed rules in the context of local governments. This essay examines whether voluntarily adopted municipal debt limits, established through municipal charter or ordinance, are associated with lower debt outcomes and borrowing costs. Using a Correlated Random Effects model on a panel of U.S. municipalities from 2017 to 2023, supplemented by a 20-year long-panel covering 1997 to 2017, the essay empirically tests competing predictions from new institutionalism and from strategic financial management literature. The findings reveal that self-imposed debt limits show no significant association with debt outcomes; in fact, ordinance-based limits show a positive association with debt level. Taken together, the three essays indicate that recipient-side institutional features do not automatically produce mission-aligned or fiscally disciplined outcomes. They function as complements to external accountability rather than substitutes for it, suggesting that the design and type of constraints shape the outcomes that funders and policymakers care about.