Research
Research
Abstract: This study compares the conventional one-flock tournament in broiler production contracts with an emerging T-week tournament which relies on a rolling window of multiple flocks/batches to evaluate growers' performance. Using the principal-agent model with heterogeneous agents, we show that equilibrium effort in the new tournament depends on the intensity of the incentives but also the growers' discount factor. Our theoretical results show that the one-flock tournament filters common shocks but introduces league composition risk, the T-week tournament does exactly the opposite. Counterfactual simulations demonstrate that the T-week tournament reduces grower income risk and enhances overall welfare, it comes at a small cost to the integrator, which makes the moving average payment scheme a possible viable alternative to replace the controversial conventional tournament.
Abstract: Despite theoretical predictions that optimal tournament incentive slopes should flexibly adapt to economic primitives, real-world broiler contracts universally fix the slope at unity. We resolve this empirical puzzle using a principal-agent model with grower-side moral hazard and structural simulations across five production complexes. Fine-tuning slope to its theoretical optimum yields negligible profit gains for integrators—less than 1% of wholesale revenue—while severely eroding grower welfare (up to a 38.4% drop in complex-wide average certainty equivalence) when slope is less than one. Because optimization gains are virtually nonexistent and transaction costs from grower unrest are substantial, maintaining the slope as 1 is a strategically rational equilibrium. Consequently, regulatory attempts to cap tournament payments may inadvertently harm grower welfare.
Abstract: The broiler production literature has long relied on tournament models assuming that production scale is fixed at unity. This paper relaxes this standard unit-scale assumption by developing a structural principal-agent framework with stochastic output scale and heterogeneous, risk-averse growers under a two-part piece-rate tournament. Theoretical results demonstrate that when output is variable, an agent's effort directly interacts with quantity risk—raising expected yield while simultaneously amplifying exposure to output volatility. Consequently, equilibrium effort decreases relative to the conventional unit-scale benchmark and varies systematically with farm capacity scale and grower ability. To evaluate these theoretical mechanisms empirically, this study constructs a two-stage structural GMM estimation strategy using flock-level panel data across five contract types. Overall, this research highlights the methodological caveats of output normalization and examines whether abstracting from quantity risk introduces nontrivial bias in empirical tournament evaluations.