Labor's share of national income

I'm not sure if there's some common underlying motivation, but both David Frum and Stephen Gordon posted recently on a famous "stylized fact" of macroeconomics - the relatively constant share of national income going to labor. Frum's is here and Gordon's is here. Labor's share of national income has been falling for the last couple decades, which is a trend that a lot of people have been speculating on. One of the better known treatments has been this IMF study (cited by Gordon) which attempts to identify the source of the decline for several different groups of countries.

I've worked a little on this issue with Wayne Vroman (recent recipient of an award from NASI!). One of the things we looked at was the change in the components of what's put into "labor share". There's wage and salary, total compensation (which is wage and salary plus benefits), and some portion of proprietor's income (usually set at something like two-thirds). Not unsurprisingly, the major source of decline has been in wages and salary. Non-salary compensation (i.e. - health and retirement benefits) have gone up over the last several decades, so that the total labor share has declined, but not as much as wages and salaries. Wayne and I are specifically looking at the implications for the unemployment insurance program, which is paid for with payroll taxes. As wages and salaries go down (while other benefits go up) the tax base for unemployment insurance is shrinking, even as the importance of a job for supporting retirement and health insurance is growing. In other words - losing a job is getting more costly at the same time that insurance against losing your job is getting harder to finance.

Right now we just have a lot of data, literature, and outlining that I've collected on the trends and their implications for unemployment insurance. We're hoping to write something up this summer on it.