Thanks, that's interesting information, some of it quite new to me. I hope you'll forgive me for taking some of it with a grain of salt—not saying I distrust it, at least not any more than I distrust anything, just that I hold it as one possibility, and this one with more actual data than those I've seen to date 
I encourage a large grain of salt. What I offered was simplistic, and much of it quite contentious. Obviously, I think the better argument is for "my" side (which, for the curious, would be market monetarism)– but there are certainly well-reasoned opinions on the other side, to which I cannot always offer convincing rebuttals.
I share that sentiment. I am aware that extra taxes and government spending result in lower economic growth. These things don't come for free, they're not just redistributing an already present wealth. They have an actual cost.
And sometimes, that cost is worth paying. Sometimes it isn't. It's up to each society to decide which costs they are willing to pay.
As a famous republican-appointed supreme court justice once said, "I like paying taxes. With them I buy civilization."
I also concur. I just happen to think that we are rather near the threshold where the cost is exceeding the benefit.
I have never seen any study that showed any significant amount of welfare fraud. I'm well aware that it goes on, but all the data I've ever seen (which I'll admit is not a lot, and not recent, but I have seen some) strongly indicates that welfare fraud, like voter fraud, is primarily used as a boogeyman to scare people into accepting new rules that make it harder for people legitimately on welfare.
This is correct but perhaps misunderstands the issue.
Much of what conservtives call "fraud" is not strictly fraud. For example, if I intentionally abstain from getting a job in order to maintain a certain level of benefits (yes, "welfare cliffs" of this kind where rising wages can cause lower incomes do exist), it isn't fraud. It's still an abuse of the system. But much harder to identify. A welfare-economics course I took last year had a central debate being, "What constitutes appropriate use of welfare?" It's actually a convoluted question– conservatives tend to view nearly any adaptive response to welfare as abuse. Consider an example: if I receive $100 in foodstamps, I will buy less food. If I buy less food, I may buy more cigarettes. If food stamps result in an increase in my cigarette consumption, it is essentially food stamps buying cigarettes (even if it isn't fraud). The question is: are these changes acceptable? What degree of "welfare behavioralism" is appropriate, as opposed to "welfare paternalism"?
My textbook suggested that, if our otherwise identical welfare system existed, but was redesigned to have zero "welfare cliffs" (which are the starkest but not only example of problems in welfare), we would see a 5-10% reduction in caseload, and a rise in incomes and employment (note: that number dated from 2007, using a dataset running from 1983 to 2004– today may be different).
Regarding actual "fraud" rates– the food stamp program is the easiest program to defraud probably given the fungibility of EBT cards. As of June 2012, it had a fraud rate of around 1% of it's total funding, if I remember correctly. That is a large amount of money, but, again, it's an exceptionally easy program to defraud (and most fraud was not by recipients per se, but by places of business doing illegal cash back– which is economically interesting for completely different reasons, but I'm already off topic, so I'll stop here).
