Ron Paul is inventing numbers for personal gain, but significant levels of fraud is not at all implausible. Bear in mind that even just $1.5 k per month is $18k per year, which is more than a full time min-wage job. Sadly, that's plenty of incentive for millions of Americans to "fake it".
Also, while initial hoops can be arduous, once in the system it gets much easier to perpetuate, especially for "disabilities" that have no reliable cure or causes (such as most mental disorders and many "pain" disorders).
I've known several people who have faked it, some for over 20 years using mental disabilities by faking depression and suicidal thoughts (they readily admitted to me it was a scam and they have no major mental issues).
The cited official 1% faker number is utterly meaningless. That is the number they catch, and much like with cockroaches, there are thousands more for every one you see. So, we have no good direct empirical basis to estimate the % that are fraud's. Also, it is important to note that "fraud" doesn't just include people completely inventing a disability that don't or never had. It is likely mostly made up of people exaggerating either the severity and/or longevity of the symptoms and exaggerating how it impacts their ability to function on a job (something very hard to objectively assess with most psychological and "pain" disorders).
So, the real question is how feasible is it to fake a disorder? For some disorders, its very easy.
Easy to fake disorders are those where diagnosis relies mostly upon self-report and statements by patients and less upon directly observable biological factors (e.g. smashed pelvis). Many/most of the mental disorders, especially "anxiety" fall into this category as do physical disabilities where the focal disabling feature is "pain" (as opposed to observable tissue/bone damage that may also come with pain).
Part of the problem is that fraud by recipients is aided by highly "soft" science, and incompetence fraud by for-profit medical practitioners. Many patients only see practitioners when the insurance covers it, and the insurance often only covers diagnosed disorders. So, the same "disorder" diagnosis that gets the person a disability check gets the therapist a regular patient with a chronic condition for which there is no reliable remedy. Its a win-win motivation. Given the nonsense that so many therapists perform, there is near zero chance of them losing their license because the American Psychological Association is highly incompetent. Also, the SSA opens the door even further to such fraud by allowing forms of "evidence" most popular among frauds and widely viewed as unscientific and invalid by scientifically based psychological researchers.
The social security administration explicitly allows such bogus test as a form of evidence to support disability claims, including "projective tests", including Rorschach and the Thematic Apperception Test where people say what they think or see in ambiguous inkblots or pics, then the therapist interprets it without anything close to a well validated, reliable, and standardized criteria. Even the more standardized diagnostic procedures in the DSM are so loose that it is quite easy to qualify as "disordered".
Studies with representative samples estimate that
26% of Americans would qualify in any given year as having one or more DSM-IV assessed disorder, and about
half of Americans will meet the criteria sometime in their life. Given that only a portion of the unemployed have disorders, that means that most people with such diagnosable disorders have been working and successfully employed in the past, meaning that all but a small % of people with mental "disabilities" are quite capable of working now and in the future. Determining if a mental disorder exists is loose and not very rigorous as it is. Determining whether a person with a disorder can person their or any job is even more subjective.