Mitchell’s Musings 8-22-16: A Clue from the Wages Fund

20 Aug 2016 12:18 PM | Daniel Mitchell (Administrator)

Mitchell’s Musings 8-22-16: A Clue from the Wages Fund


Daniel J.B. Mitchell


An important element in the current presidential campaign is wage stagnation and the potential role of trade competition in retarding wage growth. So an interesting question is how much higher real wages might be if policy on trade or whatever you might think was depressing wages was changed. The question is a bit like one we asked about manufacturing jobs we reviewed in prior musings. In that case, we asked how much larger that sector could be if the trade deficit were eliminated.


The answer we found in the manufacturing case might be described as “somewhat,” but not a dramatic return to the kind of manufacturing share of jobs that existed in, say, the 1950s. Note that the answer being only “somewhat” (and not “huge”) is not a reason to do nothing. Indeed, I have urged that there should be a policy in place to get to balanced international trade. But what the “somewhat” answer means is that there are limits to what the effect of a policy shift might be.


What about real wages? It might not surprise you that a back-of-the-envelope calculation also suggests a “somewhat” type answer. But that the “somewhat” in the wage case comes from a variant of the “wages fund” doctrine of the 19th century (and even earlier) might be a surprise. The old wages fund doctrine relates to a supposed constancy of labor’s “share” – the dollars going to labor in the form of wages and benefits – as a proportion of national income.


Despite the wages fund doctrine, there really isn’t a theoretical reason why the share of labor has to be constant. And there is some cyclical variation in the share. Empirically, the share doesn’t literally stay constant, even adjusting for the business cycle. But it does change slowly over time. The table below shows the share and the ratio of employment to population, both in percentage terms.[1] Years shown on the table roughly are business cycle peaks. (The year 2015 – the last full year available – was not a peak; we, of course, don’t know when the next peak will occur.)


      Labor’s     Employment-

      Share of    to-

      National    Population

      Income      Ratio

_____________________________


1949     60.2%         55.4%

1959     62.3          56.0

1969     65.1          58.0

1979     65.9          59.9

1990     66.4          62.8

2000     65.8          64.4

2007     64.1          63.0

2015     61.9          59.3

_____________________________


Generally, labor’s share rose until 2000 and declined thereafter. Although there is not a simple, mechanical link evident in the data, the working population (represented by the employment-to-population ratio) also rose until 2000 and then fell. Possibly, there is some connection between those two trends. Perhaps growth in the labor force was a factor in enlarging the share.


If you think that, say, international trade was squeezing the share of labor compensation, and if you think you have a policy that could offset that squeeze effect, how much would wages rise with an implementation of your policy?[2] A simple answer might be developed from the observation that at its peak, labor’s relative share was about 66% of national income, and now (2015) it’s about 62%. If you were to raise the share back up to 66% - that is, push it up by 4 percentage points – and if the number of workers receiving the share remained unchanged, dollars per worker would rise by something like 4/62 or 6.5%.[3] In magnitude, that’s a “somewhat” answer.


If you think the proportion of the population working has something to do with the size of the share, you might note that the last time the employment-to-population ratio was in its current range of around 59+ percent was the 1970s. And back then, labor’s share was about 66%. So, even standardizing for employment, you’ll still get 4/62 or 6.5%.


Of course, even if the average wage/employee rose by something like 6-7%, that gain says nothing about how the dollars would be distributed, demographically or occupationally, within the workforce. That is, some groups might be benefited by an average wage increase more than others. Overall, however, the effect is modest. Still, no one would turn down a pay raise, even if it isn’t huge.


There are other candidates than international trade for being the cause of wage stagnation such as the decline of private-sector unionization, slippage in the minimum wage, and/or “technology.”


If you think the rise in wages would pull more folks into the workforce, i.e., that the labor supply curve has a positive slope, the gain in wages might be reduced a bit. Or if you think there is a backward-bending curve (with a negative slope), the gain might be a bit more. But for back-of-the-envelope purposes, what you assume about supply is not going to matter much.

---

[1] Labor’s share data are from the U.S. Bureau of Economic Analysis national income account Table 1.12. Employment data are from the U.S. Bureau of Labor Statistics.

[2] There are other candidates than international trade for being the cause of wage stagnation such as the decline of private-sector unionization, slippage in the minimum wage, and/or “technology.” 

[3] If you think the rise in wages would pull more folks into the workforce, i.e., that the labor supply curve has a positive slope, the gain in wages might be reduced a bit. Or if you think there is a backward-bending curve (with a negative slope), the gain might be a bit more. But for back-of-the-envelope purposes, what you assume about supply is not going to matter much.


Employment Policy Research Network (A member-driven project of the Labor and Employment Relations Association)

121 Labor and Employment Relations Bldg.

 

121 LER Building

504 East Armory Ave.

Champaign, IL 61820

 

The EPRN began with generous grants from the Rockefeller, Russell Sage, and Ewing Marion Kauffman Foundations

 

Powered by Wild Apricot. Try our all-in-one platform for easy membership management