Tampering with Evidence: A Cost-Benefit Analysis