I think that one of the places he gets into trouble is when he starts discussing the "social safety net", which he thinks that liberals see as a good thing and conservatives see as a bad thing. In my usual way, I'm not sure it's not both.
Let's pick an example. Some time ago, I was having a conversation with my friend, Sally, who has actually spent time working with unemployed single mothers on welfare. And she made the point that it's not an easy life. I had to agree with her. And simultaneously I was saying, "But I'm not sure that, despite that, it isn't too easy."
Now I don't think that the vast majority of unemployed single mothers on welfare say to themselves, "I'm going to become a single mother and collect some of that easy money." I think they mostly find themselves in that position by accident. But I also believe that the existence of the social safety net makes people more willing to make bad choices, because the consequences of the bad choices aren't as negative as they would be in the absence of the safety net. When you start playing the game of "What's the worst that can happen?" and the worst is, well, tolerable, then you're more likely to make the decision that produces immediate gratification in the hope that the worst won't arrive and the knowledge that the situation will be tolerable if it does.
Endpoint games are interesting. Let's go to economics and look at the Laffer Curve, which attempts to determine the tax rate that would maximize tax revenue to the government. If your tax rate is 0%, you collect no money. If your tax rate is 100%, you also collect no money, because no sane person would work under those circumstances. (Well, they might work, but I might be in my recording studio instead of producing computer software and you might...) What's hard to determine is what the shape of that curve is in the middle. But -- mathematically -- there's got to be one or more points along the curve of taxation where you maximize your revenue. (And please note that I'm not arguing here that taxes are too high or too low, just that local optima exist.)
So what are the endpoint games for our social safety net? If there's no safety net (and assume no private charity for the purpose of this discussion), then you'll do anything that you can in order to survive, which might include work or might include a life of crime. The latter, of course, is hard to figure as a benefit to society. And there are some people who will not survive in such a situation, which I believe is also a bad thing.
But what if the social safety net is really, really good? What if life on the dole was so comfortable that there was no real incentive to work? There'd be some people who would work, just because they love what they do. But how many folks out there really do love what they do? And how many people who do love what they do would love doing something else even more, except for the fact that they actually like eating regularly?
Now, how good does the social safety net have to be before it's better than a bad job? And what happens if you have multiple generations of a family that have become dependent on the social safety net, because it is better than the bad jobs that are available to them? Nothing good, I suspect.
I don't claim that there are easy answers here. If easy answers existed, we'd be able to solve all the problems of the world before lunch.
But I just don't think that it's as simple as "Our social safety net is a good thing" or "Our social safety net is a bad thing". It's a thing and it has both good and bad consequences.
I'd like it to work better.