In Edwardian Poverty, Tim Worstall explains one reason why many anti-poverty programs — like those suggested by Senator John Edwards — are guaranteed not to reduce poverty:
So how can all of these things be true? That there are people in poverty, we spend a lot to help them out of it and even when we propose (as some of the above plans are) entirely sensible things, they have no effect on the number of people in poverty? No, this isn’t an effect of moral hazard, or “welfare doesn’t work” or anything so contentious. It’s a simple artefact of the way in which poverty is defined. From the Census:
Poverty definition [...] The official poverty definition uses money income before taxes and does not include capital gains or noncash benefits (such as public housing, Medicaid, and food stamps).
The EITC has, as cited above, made a real difference. But not to the number of people counted as being in poverty; because we count only pre-tax income: something which does not include tax credits, just as it does not include taxes paid. Housing vouchers make no difference because they are not included — i.e. they are non-cash transfers. Food stamps do not count because … well, you get the picture. We could in fact print up housing vouchers like there was no tomorrow, spend on food stamps like Ted Stevens does on bridges, double, triple, even quadruple the EITC and do you know what? There would still be 37 million people defined as being below the poverty line in the USA. All that spending would have absolutely no effect whatsoever on that number.