Question:

Suppose Medicare doubles the price it pays for a knee surgery. Has the Medicare beneficiary become better off?

Your answer to that seemingly technical question will determine a great deal of where you stand on the great inequality debate.

Here's Jim Pethokoukis on the American Enterprise Institute blog today:

President Barack Obama has a theory of the case, yes he does. For the past 30 years, the living standards of middle-class Americans have gone nowhere even as the overall U.S. economy has grown markedly. The Obama explanation: Wealthier Americans grabbed all the money. Time to raise their taxes for the sake of “fairness.” ....

Underlying Obama’s entire thesis is the work of two economists, Thomas Piketty and Emmanuel Saez. According to them, median American incomes rose just 3.2% from 1979 through 2007. (All figures are inflation adjusted.) ...

But it’s just not true, according to a new study in National Tax Journal from researchers at Cornell University. The academics, led by economist Richard Burkhauser, don’t say the findings of Piketty and Saez are wrong — just incredibly, massively incomplete. According to the Cornell study, median household income—properly measured—rose 36.7%, not 3.2% like Piketty and Saez argue. That’s a big miss.

That is a big difference. How is it generated?

A big part of the difference between the two figures comes from this decision: do you attribute as income the value of government benefits, not only cash benefits like unemployment insurance and Social Security, but in-kind benefits, like health care?

The cost of health benefits has rather notoriously exploded since 1979. And if you treat that cost as a form of imputed income to the middle class, then yes, things look rather better than they would if you look at cash income only.

Question: is that decision realistic?

I'd like to say, you make your decision on that threshold question, and it yields your answer.

But I rather fear that the truth is the opposite: You decide the answer you want, and thus decide the threshold question.