To his credit, Rep. Paul Ryan is trying to start a conversation about poverty. The test of his seriousness will be whether he ends up talking about spending.

Ryan, chairman of the House Budget Committee, issued a lengthy report Monday evaluating the welfare state. To no one's surprise, the Wisconsin Republican found it lacking. "The federal government's anti-poverty programs are duplicative and complex," it reads, then proceeds to detail them for 204 pages. The report combines descriptions of poverty programs with poverty data, and finds the former has had too little impact on the latter.

What's not in the report is also important. For starters, it contains none of the hyperbolic rhetoric that marked Ryan's 2011 Path to Prosperity budget (that's the one, remember, that claimed the U.S. was "on the brink of bankruptcy"). The current report is the product of a more mature politician. Also missing, however, are policies to address the shortcomings Ryan cites. That, presumably, will have to wait until he produces his budget blueprint, customarily released after the president's, which was released Tuesday.

For Ryan, federal anti-poverty programs suffer from two main defects. The normative problem is that too many programs "penalize families for getting ahead." Financial aid is withdrawn, sometimes abruptly, as family income rises. The "sudden drop-off in benefits create extraordinarily high effective marginal rates," the report notes, "which reduce the incentive to work." Practically speaking, there is a confusing maze of programs, some of which overlap, with too little evaluation of their effectiveness.

Ryan's critique is no doubt valid, especially when it points to the lack of evidence. Yet his own analysis sometimes suffers from a similar lapse. In the mid-1990s, for example, welfare reform was very successful at moving poor, unemployed Americans into the workforce, in part by imposing a limit on benefits. Yet when the economy soured, many of the welfare-to- work graduates ended up receiving other government subsidies. It's less clear what strategies would work in the current economy, which features drastically higher unemployment, significantly lower growth and greater challenges from technology than the economy of 1996.

Ryan is a strong supporter of the earned income tax credit, a proven success that provides about $60 billion in cash assistance to working families. And he duly notes the program's stinginess toward single workers without children.

If Ryan wants to expand that credit, drawing money from less successful programs, then his forthcoming budget proposal will be a leading indicator. Will it call for a smarter reallocation of federal spending on poverty prevention, or a sharp reduction?

Ryan paints the past five decades of fighting poverty as a lost cause. He is mostly wrong: Presidents from Lyndon Johnson to Ronald Reagan to Bill Clinton all made their contributions, and evidence shows federal programs reduce the poverty rate by almost half. At the same time, it would be a mistake to call the welfare state an unmitigated success. Its programs must be constantly re-evaluated, and its failures terminated. The trick is to figure out what policies can jump-start a chronically weak labor market and create opportunity for mobility and advancement in a stratified economy. If Ryan is prepared to realistically grapple with the problem of poor people and poor jobs, he has the talent and political standing — especially among conservatives — to do a world of good.