The NYT had a good story on the falling wage share of output and the growing concentration of wage income among high wage earners (e.g. doctors, lawyers, CEOs). While the basic story is accurate, there are a couple of points that should be treated with more care.
The article notes that the sharp drop in the wage share of GDP over the last three decades, from 53.6 percent in the first quarter of 1970 to 45 percent in the first quarter of 2006. While most of this drop is attributable to an increase in profits at the expense of wages, part of it is attributable to an increase in depreciation (the share of GDP that goes to replace warn out equipment and obsolete capital goods). The share of GDP going to depreciation has increased by almost 3 full percentage points over this period. This would imply a fall in the wage share of GDP, even if there was no redistribution to profits. The moral is just use national income or net national product (NNP) as the denominator.
The point about depreciation matters little to this story, but the growing share of depreciation does inflate standard growth comparisons between the post-1980 period (when depreciation shares first started rising) and the pre-1980 period. It also complicates international comparisons since other countries measure capital goods and depreciation somewhat differently.
The other point concerns the comparison between real wage and productivity growth over the period from 2000 to 2005. While real wages have tracked productivity reasonably well in the pre-1980 period, it is no longer reasonable to expect an exact tracking, both because of the rising depreciation share in GDP and the different price deflators used in the two measures.
Productivity is measured against the GDP deflator, while real wages are measured against the consumer price index. The GDP deflator has risen by an average of 2.0 percent over these five years, while the CPI has increased by an average of 2.5 percent. This means that even if there had no redistribution from wages to profits (or an increase in the depreciation share of output), real wage growth would be 0.5 percentage points slower than productivity growth.