KERALA RISING · FROM EXPORTING PEOPLE TO EXPORTING IDEAS 18. Risks, and what we expect to get wrong A framework that claims it will not fail anywhere is not credible. The honest position is to name the failure modes and show the control for each, and to publish the assessment either way. Some risks are specific to a program and were named in its chapter. Others are structural: they threaten the machine itself and apply to every program. The serious ones, stated plainly, with the mitigation that is built in rather than hoped for: STRUCTURAL RISK WH Y I T I S SE RI OUS Scope creep dilutes focus Seven programs invite a thin spread that proves nothing. Strict sequencing; no second program activates before the first passes its independent assessment. If citizens stop trusting the audit, the loop breaks. Per-program fraud controls; random human audit; the corpus is public, so gaming is visible and correctable. Verification is gamed TH E M I TI GA T I ON A partner An official can department feels quietly kill a threatened program. 266 The political-economy rule: the partner chairs activation and