Does our $927-billion-per-year welfare state discourage the poor from seeking work? Does it cause them to shun pay hikes? Does it, in short, create dependency?

This may be the most intractable problem afflicting the welfare state. With a renewed interest in the Clinton-era welfare reform and President Obama’s recent announcement that he would ignore its famously successful work requirements, understanding the relationship between welfare and work is more important than ever.

Our 80 or so federal welfare programs are designed so that benefits phase out as incomes rise. Fair enough. After all, the government’s generosity must end somewhere. Neither Warren Buffet nor the typical middle-income family should receive even one dollar of food stamps or other welfare assistance, right?

But this feature of welfare — the necessity of phasing out benefits as incomes rise — brings a serious moral hazard. In many cases, economists have calculated, welfare recipients who enter the work force or receive pay raises lose a dollar or more of benefits for each additional dollar they earn. The system makes fools of those who work hard.

Recently the chairmen of two important subcommittees on Capitol Hill convened a hearing on this issue. The hearing elicited some revealing testimony from one of the chairmen’s congressional colleagues.

“The more benefits the government provides, the stronger the disincentive to work,” Representative Geoff Davis (R., Ky.) pointed out. The great irony, he added, is that although federal welfare programs “are designed to alleviate poverty while promoting work,” collectively they have “an unintended side effect of discouraging harder work and higher earnings.”

Less work and lower earnings, in turn, translate into greater dependency on the government — and zero or even downward social and economic mobility for those mired in poverty.

Davis invited Great Britain’s chief welfare reformer, Sir Ian Duncan Smith, and a panel of esteemed economists and social scientists to testify on the extent of this problem. But it was another witness, Representative Gwen Moore (D., Wisc.), whose testimony, which raised some eyebrows, stood out.

Moore is a product of inner-city Milwaukee. She persevered through hard times to graduate from Marquette University and win election first to the Wisconsin assembly and then to the state senate, where she rose to be senate president. She is the first African American to represent Wisconsin in Congress.

In her written statement, Moore challenged the premise of the hearing, describing it as “predicated on a series of false assumptions about our social-safety net.” She went on:

I can tell you that I didn’t sit down after a long day of work to calculate the “marginal tax rate” of government benefits to determine my choices. I think that the panelists here today — who make it sound like low-income women get home every night and pore over a spreadsheet to figure out how much they’ll be taxed on the next dollar they earn, or their next dollar in benefits — need a bit of a reality check.

But her oral testimony sent an entirely different message. She drew from her own compelling personal experience as a young mother on welfare to illustrate why we need to turn our welfare state upside down and start over.

“I once had a job,” Moore acknowledged, “and begged my supervisor not to give me a 50-cents-an-hour raise lest I lose Title 20 day care.” The same work disincentive arose when she contemplated the health coverage she received through Medicaid. “I would want to work if in fact I didn’t risk losing Medicaid.”

Low-wage workers may not sit in front of spreadsheets at night calculating their marginal tax rates to determine whether or when they should take a job or agree to a pay raise. But some providers of welfare benefits, it seems, perform that function for them.

Representative Moore explained:

My day-care provider told me that I was . . . earning $17,000 a year with three kids, I was still poor, that I had in fact hit that marginal tax rate, and that if I earned any more, I was still poor, that . . . so, when . . . January came around, and the automatic increases . . . in Title 20 occurred, the inflationary increases, then I could take . . . the 50-cents-an-hour raise.

So, with a little “welfare math,” welfare recipients are instructed in how to time their entry into the work force, their departure from it, and any salary increases. Should that raise come before or after the next scheduled cost-of-living increase in a welfare program? Should she accept that promising entry-level position without first calculating whether the new source of income would mean a reduction in the welfare benefits she receives?

Every time these considerations, which are based on rational economic self-interest, compete with what should be no-brainer decisions — take the job! accept the pay raise, and work even harder for the next one! — the aspiring poor lose.

Let’s fix the system by injecting some much-needed moral clarity into the rewards and benefits it offers.

— Michael G. Franc is vice president of government studies at the Heritage Foundation.