EPI Greatly Exaggerates Wage/Productivity Divergence
I don't know how they made their analysis, but I do know that the numbers are extremely misleading. If their numbers had been correct, the percentage share of national income going to compensation of employees would have fallen by more than a fourth.
In reality, the share of gross domestic income going to compensation of employees fell from 57.7% in 1989 to 56.1% in 2009 (2010 numbers aren't available yet). The drop was even lower in relation to net domestic income (which excludes capital consumption), from 65.6% to 64.8%.
The share of income going to labor tends to be counter cyclical (decrease during booms, increase during slumps) and 2009 was a year when the economy was still in effect in a slump while 1989 represented the peak of the Reagan boom, but even in 2006, it had only fallen to 54.9% of gross domestic income and 62.5% of net domestic income.
Thus, while wages and other forms of labor compensation have fallen behind productivity somewhat, the divergence is much smaller than the EPI claims. I am again not sure just how they have come up with their misleading numbers, but most likely they have ignored some forms of compensation of labor and used different price deflators for productivity and labor compensation.