First of all, I'm genuinely glad to see Paul Ryan talking at length about poverty, as he did in a speech yesterday. I'm guessing that makes him second only to John Edwards in terms of how much verbage a recent presidential candidate has given the issue. And in order to remember the poor with our actions and policies, we have to literally remember that they exist. Also, unlike Charles Blow, I won't go so far as to suggest that the whole thing is fraudulent. There's no question that Ryan's speech was a calculated political play, but that doesn't mean that's all it was. I assume this is entirely sincere:

Mitt Romney and I are running because we believe that Americans are better off in a dynamic, free-enterprise-based economy that fosters economic growth, opportunity and upward mobility instead of a stagnant, government-directed economy that stifles job creation and fosters government dependency.

Of course, sincere doesn't mean correct. The real question isn't whether super-awesome, extremely effective supply-side economics are superior to a lousy, statist alternative. It's whether supply-side dogma is in fact awesome and effective, and whether the alternative is in fact lousy statism. And Ryan, sincere though he may be, didn't really make the case yesterday.

Instead he praised the 1996 welfare reform law, which he characterized as coming about because the problem of top-down solutions that create dependency on government "was so obvious" that both parties got behind reform. (I remember it having something to do with a neoliberal president who would rather do anything than nothing, but what do I know?) The result? "Instead of seeing increases in hunger and poverty, we saw welfare enrollment drop dramatically." Instead of A, we saw something that is not the opposite of A.

Lower enrollment was of course the explicit goal of welfare reform, and it succeeded at that. Even when the recession hit in 2008, welfare-to-work caseloads only went up very slightly.

But effectively keeping up with need is exactly the point of an aid program. At least, it is if the goal is a functioning safety net. If the goal is simply to spend less money on a program, then making the program smaller is a pretty much foolproof way to do so. In 1995, the old welfare system assisted 75 percent of poor Americans. In 2011, welfare-to-work assisted 25 percent. Welfare reform may have helped lift some people out of poverty, but many were simply left behind without benefits.

You can make a case for stricter requirements for aid eligibility. You can make a case for less centralized control of aid programs. But if a reform is set up primarily to serve fewer people, then that's what it's primarily going to do—and that's where it's going to find most of its savings. Ryan argued yesterday for welfare-reform-a-tizing the whole safety net:

Here’s the problem: The welfare-reform mindset hasn’t been applied with equal vigor across the spectrum of anti-poverty programs. In most of these programs, especially in recent years, we’re still trying to measure compassion by how much government spends, not by how many people we help escape from poverty.

Well, the government measures both. But it's clear that Ryan's priority here is how much the government spends. If we were to enact his proposals for other safety-net programs, the government would save money—by serving fewer people, leaving a gap the private sector couldn't hope to fill.

Ryan's rhetoric here is about local control: "If the question is what’s best for low-income Ohioans, shouldn’t we let Ohioans make that call?" But converting a federal entitlement to a state-run program doesn't just give states control over how they tackle a program; it also gives them permission to simply tackle it less, which many are glad to do. (States already have the power to determine how many people to cover under Medicaid. As a result, a working mother of two in Arkansas is ineligible if she earns $3,200 a year—less than a tenth of the eligibility cap in Minnesota.) And if the federal money doesn't keep up, states don't even have a choice.

Ryan's speech makes it clear that the role of government, not poverty, is its real subject. "The short of it," he said, "is that there has to be a balance—allowing government to act for the common good, while leaving private groups free to do the work that only they can do." At face value, I agree. That's why I like the charitable giving deduction, because the government really can't do it all.

But this immediately followed Ryan's suggestion that voluntary giving is categorically superior to coercive taxation. As I wrote a while back, if fighting poverty is truly the primary goal, this argument is a nonstarter. There's more than enough antipoverty work to go around, and public and private dollars aren't competing in a zero-sum game. (Even those tax increases that can hurt economic growth seem to do so only when taxes are already relatively high, which right now they very much are not.) The role of government may be a worthy subject, but it's not the same subject as reducing poverty. While supply-siders like Ryan may believe that B follows A, the evidence says otherwise.

At one point in the speech, Ryan calculated that federal and state governments spend more than a trillion dollars on antipoverty programs. "For that amount of money," he said, "you could give every poor American a check for $22,000." That's actually not a bad idea. But it's not what Ryan's proposing. He's long pushed for less government spending on antipoverty programs. And when it comes to results for the poor, "how much" matters at least as much as "how."