This dissertation comprises three essays on how institutional constraints shape contracting, monitoring, and organizational performance across financial and public sector settings. The first essay examines how collateral protection reshapes bank monitoring as firms approach bankruptcy. Using hand-collected quarterly data on loan agreements and liquidation valuations from Chapter 11 filings between 2007 and 2022, I document that secured lenders systematically shift from performance-based covenant monitoring toward liquidity and collateral-focused oversight as borrowers near insolvency. This shift is concentrated among overcollateralized lenders, who abandon comprehensive oversight once their recovery is fully protected by collateral values. Using instrumental variables that exploit exogenous variation in local real estate prices and aggregate secured lending market conditions, I provide causal evidence for this monitoring reallocation. While it is individually rational for senior lenders, it generates significant negative externalities: going-concern value declines by 71%, unsecured creditor recovery rates fall by 41 percentage points, and fire sales increase. Causal mediation analysis confirms that the monitoring channel—not collateral itself—drives these adverse outcomes. The second essay, co-authored with Brian Dillon, John Mulenga, and Twivwe Siwale, studies tax evasion in relational contracts. Using a randomized controlled trial with 1,083 small retail firms in Lusaka, Zambia, we incentivize firms to request formal VAT invoices from their suppliers. Financial incentives significantly increase invoice retention, and VAT-registered suppliers respond by reporting more to the tax authority. However, firms with stronger supplier relationships are substantially less responsive to these incentives, consistent with relational contracts sustaining informal tax evasion. The third essay, co-authored with Daniela Scur and Morten Bennedsen, examines how management practices affect employee mental health in Denmark. Linking management survey data to prescription drug records, we show that better management reduces mental health medication use—by 14% in the public sector and 10% in the private sector. Effects are heterogeneous: high-ability workers in the public sector are disproportionately sensitive to management quality, while private sector workers benefit uniformly. Better management compensates for wage compression in the public sector, making it especially valuable for high-ability workers who accept pay penalties for mission-driven work.