Small Datasets Skew Government Costing

Federal agencies lean on parametric cost models trained on limited and often obsolete or unavailable data—especially in space and defense where costs are classified or proprietary. These models are then used (and sometimes misused) to set budgets for novel programs, leading to persistent mispricing and waste versus using actuals, similarity, or expert judgment. The result is a systematic estimation error built into procurement. — If core budgeting tools are structurally unreliable, procurement reform and state capacity must fix estimation methods or keep bleeding money on flagship projects.