History shows that he is right.
In the 1960s we set up the great society which created the safety net for poor people. In the years after it was established poverty dropped.
But then poverty stopped dropping. That was because the government can only do so much before cultural issues take over.
We can provide people with food, education and a place to live, but we can't instill them with a desire to work and succeed.
Thus the cultures that value hard work and education succeed regardless of where the people live and what obstacles they face, while the cultures that lack those values fail.
Therefore the best form of government would be one that gives people a positive environment where the people who want to succeed can as opposed to a government that focuses on artificial outcomes based on notions of 'fairness.'
It is a pretty conservative argument.
I think we may be arguing at cross purposes here. Culture will effect outcome, however he uses Sweden as a metric and I understand that. He then uses examples of other cohorts to say that policy is irrelevant. My point is that he cannot do this because stating that like groups cannot be influenced by policy decisions isn't a valid argument, because "policy" is an ambiguous term.
Here's an example.
Given two groups of culturally similar people on welfare:
Group A is given relatively unfettered discretion to spend funds received in any way possible, and has no mandate to work if they are able.
Group B has restrictions placed on how they can spend their money, are required to do some work suited to their abilities and are further required to improve their education with carrots and sticks for incentives.
Now since the second hasn't been tried I cannot say for certain that Group B will universally decide to get off of medicaid when possible, however I CAN state that Group A has individuals who do not know what it is to work at all. Given training and experience it may occur to some that doing the same work below market value is not a good idea and that with job placement assistance may find a viable way out of poverty. To say "get a job" to someone who lives where no jobs are to be had, with no education and no work experience is tantamount to "let them eat cake"
It would therefore seem that the best approach would be to enact policies
with the goal of positively influencing problematic behaviors within a given culture.
Yes to some degree this is social engineering, but so is telling your kids to get an education and work hard and smart.
The problem with past efforts was the assumption that sending a check every month would be sufficient to bootstrap the impoverished into prosperity and that the inherent nature of people will preclude substantial abuse. That has clearly been shown to be incorrect.
It isn't about enacting handouts (I'm told that's a code word, but whatever) but providing motivation and the tools necessary to achieve.
In the worst case policy can say "if you refuse to comply you have lost any assistance". That's guaranteed incentive.