Cost-Informed Contract Design for Municipal Solid Waste Services

Municipal solid waste contracts must provide budget predictability without making continued service financially unsustainable when wages, fuel prices, or facility charges change. This paper links cost measurement to the design of economic price adjustment clauses for collection, transportation, transfer, and disposal services. A bottom-up life-cycle cost model for Phoenix, Arizona supplies auditable component weights, while a separate comparison with Phoenix staffing and fleet disclosures identifies a calibration problem: modeled labor requirements are 2.3 to 2.6 times the city's reported field-services headcount. A ten-city panel contains 51 annual cost observations, including 26 strict-Actual cost-and-tonnage observations from seven cities. City-and-year-effects estimates are not statistically significant under exact small-cluster inference, and the strict-Actual tonnage elasticity changes from 0.658 in nominal cost to 0.180 after CPI-U deflation and 0.003 after waste-collection PPI deflation. The contractual weights therefore come from the Phoenix engineering model, which provides a direct-cost benchmark of $125.02 per ton, allocated among labor (33.75 percent), fuel (8.92 percent), disposal (35.91 percent), and other fixed costs (21.43 percent). Component-specific indices also fit the modeled exposures better than a single broad index: Arizona collector wages correlate negatively with CPI-U (-0.42), whereas regional diesel prices correlate strongly with it (0.87). Contract simulations compare fixed pricing, full contemporaneous indexation, and bounded component indexation across six cost paths. Fixed pricing removes payment volatility but leaves average uncompensated modeled cost of $13.1 million; full indexation eliminates that shortfall but permits a maximum annual budget increase of 26.8 percent. With immediate adjustment, an equal-weight screening rule selects a 5 percent deadband and a 15 percent cap. Imposing a three-month verification lag lowers minimum coverage to about 80 percent in the combined extreme scenario and causes the exceptional reopener to activate in four of six scenarios. The evidence favors a layered design in which competition determines the initial price, a symmetric component formula handles ordinary price movements, and a documented reopener addresses persistent tail events.