I enjoyed you article on statistcal measures of the US economy in the latest Monthly Review. In there, you contrast two measures of poverty, (1) absolute (which is used for the "official" rate), and (2) relative. As I'm sure you know, both are deeply flawed. As I explain below I think relative and absolute measures can be combined to make a better measure than either, and, in fact, the BLS used to publish such a measure.
The relative measure--typically 50% of median income--that you suggest many researchers prefer is simply arbitrary; it lacks any vestige of the substance of the term. If I argued poverty should be defined as 62.41% of the median, what would be your argument against me? Where did 50% come from? Moreover, during 1929 to 1933 per capita income dropped 46.3%, per capita consumption dropped 42.5 % (some dissaving), and real per capita consumption declined 27.3% (after deflation). Using the relative measure, a person with who was poor in '29 could, with the same income, be nonpoor, perhaps even "rich" in '33. The substance of the concept of poverty--whether a person can meet his basic social and physical needs--is lost using a *purely* relative standard like some proportion of median income.
The "official" measure combines an absolute measure of spending needs with inflation adjustments, but, as it is done, it is even worse. Contrary to what most assume, however, it is *not* based on a typical consumption basket. Rather, poverty lines are calculated by estimating family food spending only, and that number is multiplied by three to reflect the rest of a family budget. This is obviously inferior to using data on all consumption because over time spending proportions change. But worse than that, the "official" number, as introduced by Johnson in '64, is an entirely political concoction. Johnson's people chose to base poverty calculations on the low, emergency food budget designed by the BLS to sustain someone for two weeks, on an emergency basis. It was specifically not intended as a longer term budget. The BLS, in fact, called this budget the emergency food budget. It's not that they didn't have better measures. The BLS had prepared several budgets at that time, including one, called modest but adequate, that was clearly suited for poverty measures (read anything by Mollie Orshansky of the BLS in the early 60s to get this story). Johnson chose the emergency food budget because it reduced the count of the poor, of course, and also because it resulted in a nice round number he could publicize. A family of four was poor in '64 if you made less than $3,000. Needless to say this political choice has resulted in a substantial undercount of the poor ever since.
The absurdity of the official poverty numbers is clear when compared to the budgets the BLS used to publish based on all consumption spending, starting since at least 1909. After WWII upon request of Congress, the BLS designed the City Workers' Family Budget (later called the Urban Family Budget Index, I think), specifically to estimate what it cost the unskilled manual laborer's family to live in large cities. The BLS concluded that budget should "represent the necessary minimum with respect to items included and their quantities as determined by prevailing standards of what is needed for health, efficiency, nurture of children, social participation, and the maintenance of self-respect and the respect of others." To do this the BLS compiled a budget from standards of calories and vitamins, from housing standards based on needs, and "from standards that are revealed by the ways in which people actually spend their money". The consumption basket underlying the budgets was to be adjusted as spending patterns changed. All of this involves some judgenent of course, but when you look at specific items they came up with, they are reasonable and basic to say the least. "Necessary minimum" describes their budget well; so does bland.
But these budgets were more than twice as big as the one Johnson chose. In a flick of the pen, he eliminated a large number of poor people! Bet he slept well that night.
The point is the BLS family budgets captured the essence of the meaning of poverty. They combined relative measures (based on social needs) with absolute ones and were adjusted over time to reflect changes in consumption. They are a far superior measure of poverty than either measure you discuss. Using them today would probably result in the poor totalling more than either of the measures for the mid 90s you show on p. 131--11.7% (absolute) and 19.1% (relative).
More important for me, the BLS budgets captured precisely what Marx meant by the concept of the social subsistence of labor that is used to measure the reproduction cost of productive labor. They are useful analytically.
I am using the past tense when referring to the budgets because soon after seizing power in '81 the Reaganites killed them. They said they were too expensive to do. But we know the real story: they were afraid of marxism sweeping the country as the real nature of the immiseration became more clear. OK, a joke. But seriously, the killing of the budgets was one action in a series where the Reagonites eliminated important information in an attempt to make it more difficult for people like you and me to both understand what is going on, and cricitcize their policies. Their elimination of data on gross product of foreign affiliates of US companies is another example. You could simply add that data to gross domestic product to see the total production of US based capital, as well as factor shares, and particularly the distribution of profits, foreign and domestic (foreign profit shares were outstripping domestic).
As to the profit rates you use in the article, first a question. Does the profit rate for nonfinanciial corps you use in the table on p. 122 contain interest payments (they are not mentioned)? They should be included. Interest payments by nonfinancial corps are part of capital's share siphoned off by finance capital. An increasingly important share. As corp. debt has risen, more capital is being returned in the form of interest payments and relatively less in the form of dividends.
Why do you use pretax profits (I see it often in other work too, including marxian stuff)? Only after tax profits are capital's share, the actual reward to suppliers of capital. They are what is available to accumulate (invest) or pay out in dividends. Taxes on both corp. and personal income are part of surplus value that is redistributed. This matters because corp. profit taxes have been reduced several times since the 50s, so that pretax rates understate the trend. Plus, if you were able to include the profit rates from foreign affiliates of US based capital, as I mentioned above, the profit rate trend would be raised some more.
All three factors I have mentioned would adjust profit rates in the same direction--upward. If you accounted for these things I think you would find that the 70s and 80s were a lot more like the 50s than you thought, and, perhaps, profitability in parts of the 90s may have reached or exceeded the postwar peak of the mid 60s.