Three moves: (1) Government spending monotonically slows economic growth — the data shows no sweet spot at 15–25% of GDP, just a steadily declining power law. (2) If government is structurally a brake, the question isn't 'how much?' but 'on what?' — brakes are useful precisely because they slow things down. (3) Aim it by an objective rule: brake an activity only if it imposes a net wealth loss on external parties AND the brake is cost-effective. Three of four cells in the decision matrix tell government to do nothing.