It is also delusional. The poverty rate has gone from 55 to 22 because the GDP has expanded from 2319BB in 1959 to 9214BB in 2001. (1996 real dollars.) If the poverty rate was measured the same in 2001 as in 1959, poverty would be in the low single digits, if it made a whole digit.

Which part is delusional, Phelps? This article from the US Bureau of the Census explains how poverty has been defined since 1948. As you will observe, the definition has been updated once over the period 1959-2001, in 1964. The threshhold of poverty is extremely parsimonious in view of gigantic increases in health care costs, residential costs, and transportation costs.

(GDP counts the increased amount of transportation consumed by urban commuters as a gain; so if P6 has to travel twice as far to get to work than he used to, and it costs him twice as much, GDP statistics count that as an increase in his consumption. I could bore you into a coma with examples such as these.)

That said, we could observe that the gigantic reduction in poverty of the 1960's came at the same time the threshhold of poverty rose. A good practical earmark of what poverty means, and what enables it to be compared across countries, is "entitlement sustainablity" (WARNING: economic term of art). "Entitlement" here refers to one's legitimate command of income--e.g., wages, interest, persions, rent or state assistance. If my income is reduced to a point where I cannot continue to commute to work, live in the city where I need to work, or maintain my health so that I am able to work, then I am in poverty.

The notion that GDP growth necessarily reduces poverty is wholly erroneous. If one measures across states or municipalities, for example, one can readily see that growth in per capita gross state product (GSP) is not linked to reduction in poverty. In Latin America, GDP growth of the mid-1990's accompanied increased poverty. That GDP growth trickles down to all sectors of society automatically is an assumption which is not borne out by the facts.

I keep it real:

if I can't afford to support myself, I'm broke and probably poor (which I phrase that way because one of my more senior associates says being broke is about money and being poor is about attitude).

A lot of noise has been made about folks living with their parents until their 30s being a new social thing. I think (and reading between the lines for the interviews in such articles supports my contention) the reason is simpler. They can't afford their own place.

I was going to post a big long response, but in looking at the data and spending a half an hour charting it to make sense of it, I realised that the source of the data used here is lacking metric data in half of the 60s. I'm not going to get into (another) argument here about intuition.

Sorry, I have to admit my best sources are offline--On Well-Being & Destitution by Partha Dasgupta and econ textbooks. But the argument is really very simple: either the threshhold of poverty is going up, it is remaining the same, or it is going down. If that threshhold is declining, then the reduction in the proportion of American/African American households in poverty is caused by lowering the standards.* If so Phelps is correct and I am delusional.

If it is staying the same then possibly the reduction is caused by an evenly distributed growth in GDP--which certainly doesn't defy the laws of physics, but anecdotal evidence of urban gentrification suggests that GDP growth can often make it far more expensive for the very poor to live near a source of income, and of course income distribution has become decidedly more concentrated.

If it is growing then the achievement is more impressive still. It should be pointed out that for years after the Great Depression major industrial firms favored the social welfare programs because they feared another implosion of aggregate demand. From their point of view, the raising of this floor of consumption prevented a repreat of the 1929-1933 calamity.

-----------------------------------
* Another possibility would be that the rate of households in poverty declined as African Americans moved in with each other and mean household size increased. Since the combined poverty threshhold income of two households of 2 people each is higher than that of 1 household with 4 people each, poverty rates could be reduced by making people live together. In fact, the chart shows in 2001 mean AA household size was 1.27, whereas in 1959 it was 1.08. That would explain part of it, I suppose.

If it is staying the same then possibly the reduction is caused by an evenly distributed growth in GDP--which certainly doesn't defy the laws of physics, but anecdotal evidence of urban gentrification suggests that GDP growth can often make it far more expensive for the very poor to live near a source of income, and of course income distribution has become decidedly more concentrated.

This assumes some sort of zero-sum game. That isn't the case. As the GDP goes up, it can be unevenly distributed and still raise people out of poverty. If the rich quadruple their wealth, and the poor double thier wealth, the poor have still doubled thier wealth.

I've seen what poverty was in the early 50s. (I'm not trying to be facitous here -- documentary and anecdotal, not my own eyes.) That poverty was much worse than what we call poverty now.

This assumes some sort of zero-sum game. That isn't the case

No, it does not assume a zero-sum game. I'm the one arguing that spending a minor share of GDP on poverty reduction improved everyone's welfare. It enriched the non-poor by increasing the level of aggregate demand and stimulating a virtuous cycle. This, incidently, is especially excellent because once aggregate demand and public goods were raised to a higher level, people were less poor because their incomes were higher, not because of income transfers.

In cases where public goods are inadequate, income redistribution becomes a crutch employed to placate the poor and angry. If public goods and aggregate demand are elevated, income redistribution becomes superfluous. There's no longer an urgent political obligation to stop riots by handouts. And middle income quintiles are richer too. This is important because younger cohorts of workers face a future of probable promotions which supports their expected lifestyle.

If the rich quadruple their wealth, and the poor double thier wealth, the poor have still doubled thier wealth.

I have no problem with this, except that if living requires twice as much wealth because the rich are consuming more things for which supply is inelastic--like land--or things for which there are negative spillovers, like leaded gasoline, then the poor will not have doubled their welfare. P6 put it best:

if I can't afford to support myself, I'm broke and probably poor

Which is a problem even if P6 is drinking clean water instead of that cholera-laced crud he used to. I mean, I'm really really happy he's got clean tap water--way to go, P6!--but if a gigantic section of the populace is in his shoes, the political establishment has a lot of explaining to do.