Inequality and rising profitability in the United States, 1947–2012

Abstract

The last 40 years has seen slow growing earnings and income for the middle class, as well as rising overall inequality. In contrast, the early postwar period witnessed rapid gains in wages and family income for the middle class and a moderate fall in inequality. The ‘booming’ 1990s and the first decade of the 2000s did not bring much relief to the middle class, with median income growing by only 2% (in total) between 1989 and 2012. The stagnation of middle class living standards since 1973 or so is attributable to the slow growth in earnings. While average earnings almost doubled between 1947 and 1973, it advanced by only 22% from 1973 to 2012. The main reason for the stagnation of labor earnings derives from a clear shift in national income away from labor towards capital, with overall profitability rising either back to previous postwar highs or to new highs by 2012. Based on regression analysis, a positive and significant connection is found between top income shares and the profit share. The unionization rate, computer investment per worker, the minimum wage, and the unemployment rate are also significantly related to top income shares.

Keywords:

• profitability income
• earnings
• wage share
• profit share

Notes

1. The source is: http://www.nber.org/cycles/cyclesmain.html.

2. The source is the US Bureau of Labor Statistics at: http://data.bls.gov/timeseries/LNS14000000.

3. It should be noted that individual states also set minimum wages for employees in their states, and as of 2014 a majority of states have minimum wages that are above the federal level. Details of state minimum wages are available at http://www.dol.gov/esa/minwage/america.htm (accessed April 10, 2014).

4. The data source for all income statistics, including the Gini coefficient and the poverty rate, is the US Bureau of the Census, Current Population Survey, available at http://www.census.gov/hhes/www/income/histinc/ (accessed April 10, 2014). Income figures are in 2011 dollars unless otherwise indicated. It would actually be preferable to use household income rather than family income. Unfortunately, official US Bureau of the Census series on household income begins only in 1967, whereas family income data are available from 1947 onward. I also use the Census Bureau recommended consumer price index (CPI-U-RS) to deflate incomes over time.

5. These figures are based on the US Bureau of Labor Statistics (BLS) hourly wage series for production and non-supervisory workers in private, non-agricultural industries. The source is: US Council of Economic Advisers, Economic Report of the President, 2013. This is the most widely used wage series. The BLS converts nominal wage figures to constant dollars on the basis of the Consumer Price Index (CPI-U).

6. These two are the National Income and Product Accounts wages and salaries per full-time equivalent employee (FTEE) and employee compensation (the sum of wages and salaries and employee benefits) per FTEE. Both series are deflated to constant dollars using the CPI-U price index. A third, not shown here, employee compensation plus half of proprietors’ income per person engaged in production (PEIP), shows very similar time trends. The data source is: US Bureau of Economic Analysis, National Income and Product Accounts, Internet, at http://www.bea.gov/iTable/iTable.cfm?ReqID=9&step=1#reqid=9&step=1&isuri=1 (accessed April 10, 2014).

7. The source for these data is the US Bureau of the Census, Statistical Abstract, 2009, Table 576.

8. The source is: US Bureau of labor Statistics, Women in the Labor Force: A Databook, February 2013, Table 4, available at: http://www.bls.gov/cps/wlf-databook-2012.pdf. Since wages were generally stagnant from 1980 to 2000, the increase in real median family income over these years was likely due to the fact that households worked more hours. This latter could be the result of more workers per family, more hours per worker, or a combination of the two. According to the Economic Report of the President, 2013, Table B47, average weekly hours in the total private sector actually declined from 35.2 in 1980 to 34.3 in 2000. The data source is: http://www.gpo.gov/fdsys/pkg/ERP-2013/pdf/ERP-2013-table47.pdf (accessed June 22, 2014).

9. The data source for series is the US Bureau of the Census, ‘Detailed historical income and poverty tables from the March Current Population Survey 1947–2007’ (see footnote 1). These figures are based on unadjusted data.

10. The source is the World Top Incomes Database, available at: http://topincomes.parisschoolofeconomics.eu/ (accessed April 10, 2014).

11. The rates quoted here are for married couples, filing jointly. The data source is: http://www.irs.gov/ (accessed April 10, 2014).

12. The 39.6% rate is effective for AGI above $400,000 for singles and above $450,000 for married couples.

13. In the case of an economy characterized by competitive input markets and constant returns to scale, it follows that wages and labor productivity should grow at exactly the same rate. In particular, w = ϑX/ϑL = ε_L X/L where w is the wage rate, X is total output, L is total employment, and ε_L is output elasticity of labor, which equals the wage share in this special case.

14. Results are shown for employee compensation per FTEE. Results are almost identical for employee compensation plus half of proprietors’ income per Persons Employed in Production PEP. The data source is: US Bureau of Economic Analysis, National Income and Product Accounts, Internet, at http://www.bea.gov/iTable/iTable.cfm?ReqID=9&step=1#reqid=9&step=1&isuri=1 (accessed April 10, 2014).

15. This definition of capital income excludes the pay of CEOs and other top management, bonuses, stock options, and the like, which are counted as labor compensation. If these components were included in capital income instead, the rise in the capital share since 1980 or so would be even greater than reported below. Also, as noted, the CCA is excluded from the definition of net profits. There is evidence that depreciation has risen as a share of the capital stock over recent decades. This trend is likely due to a higher share of technology goods in capital. The effect of a rising share of depreciation in the capital stock would be to lower the rate of increase of the net profit share in national income. The data source is: US Bureau of Economic Analysis, National Income and Product Accounts, Internet, at http://www.bea.gov/iTable/iTable.cfm?ReqID=9&step=1#reqid=9&step=1&isuri=1 (accessed April 10, 2014). Also see Giovannoni (2014) for a discussion of alternative definitions of the labor and profit share.

16. Several other explanatory variables were included in the original model, including the annual rate of labor productivity growth, the annual rate of total factor productivity growth, research and development expenditures as a fraction of GDP, equipment investment per worker, the ratio of exports to GDP, the ratio of imports to GDP, and foreign-born workers as a percentage of the labor force. These variables proved to be statistically insignificant and, as a result, are not used here.

17. Regression results are almost identical when OCA per FTEE is used instead of OCA per PEIP.

18. The actual regressions are estimated using AR(1) in order to correct for autocorrelation (the Durbin-Watson statistic is typically about 1.00).

19. Ideally, it would be better to have a measure of investment in computers and Information Technology alone but this variable is not available for the full 1947 to 2012 period while investment in OCA is available.

20. The source is http://topincomes.parisschoolofeconomics.eu/ (accessed April 10, 2014).

21. It is likely that the corporate profit rate would be more highly correlated with the inequality measures than the corporate profit share. However, a data series for capital stock in the corporate sector is not currently available.