Even if lots of people were faking disability, that would not justify a failure to adequately support those who are not. If it was determined that $X/month is needed by disabled persons, and that Y% of recipients of that benefit are committing fraud, then the humane and correct response is to pay $X to those who are legitimate claimants, and to detect, investigate and ultimately jail, those who can be shown to be committing fraud. Paying $(Y% of X)/month, or indeed any amount $<X/month, fails to adequately help the needy, while still rewarding the fraudsters, and is a lose-lose solution.
Reducing the funding for disability benefits is not capable of solving (or even of beginning to address) the issue of fraud. Evidence of fraud might be motivation to spend more on enforcement - but cannot be logically defended as motivation to spend less per claimant on benefits to those known not to be fraudsters, in some misguided attempt to punish those who are getting away with fraud.
Of course, first you need actual evidence that fraud is currently sufficiently common as to be worth the cost of additional enforcement measures.
The whole 'reducing spending on suspicion' approach is ugly as fuck, and dumb as dogshit, and people like Rand Paul who suggest it need a fucking smack.
