Innovation and Income Inequality in the USA: Ceremonial versus Institutional Changes

Abstract

This article investigates the distributive effects of technological progress in the United States during the last four decades. The result of our econometric analysis reveals that the shift in R&D investment from the public to the private sector was associated with an increase in income share of the richer classes at the expense of the poorer income classes. Taking an institutionalist perspective, these findings can be explained by ceremonial encapsulation of innovation by corporate capital that slows the pace of social progress. In this context, diffusion of innovation may be treated as a progressive institutional change.

Keywords:

• innovation
• income inequality
• ceremonial changes
• institutional changes

JEL Classification Codes:

• B52
• O31
• O33

Notes

1 If innovators are paid by capital owners to do research, they are not able to retain control over their ideas. That capital owners hold the patent rights to inventions of employees is not limited to the corporate-funded research. According to the Bayh-Dole Act from 1980, ownership of inventions made with federally-funded research is also given to colleges and universities, including the right to license the patent to private sector for commercialization. Furthermore, most universities in the United States also have patent rights on inventions created by undergraduates, using a significant amount of their resources

2 Moreover, many startups see acquisition as an exit strategy in the sense that their funders are more interested in selling their business to big companies than building their own companies.

3 Government support for business R&D in the United States relies on a combination of direct funding and tax relief. According to the latest available data, direct government funding and tax support for business R&D in the United States accounts for approximately 0.23% GDP in 2016 (OECD n.d.).

4 In this context, crowding out appear if R&D investment of a subsidy recipient is smaller than the investment it would have made if it had not received a subsidy (Busom 2000).

5 We prefer three-year averages to annual data for two reasons. First, income shares are changing slowly over time. As a result, the changes in the explanatory variables do not have the strong impact on income distribution on annual basis. Second, three-year averages reduce the effects of economic cycles, which allow us to concentrate on structural relationships. In order to address the potential problem of endogeneity and delayed impact of the explanatory variables on the dependent variables, the explanatory variables are included in the equations as measured at the start of each three-year interval.

6 In this way, we control the correlation among error terms across individual equations and gain efficiency from combining information on different equations. The result of the Breusch-Pagan Lagrange multiplier test for error independence indicates statistically significant correlation between the errors in the equations (chi2(3)=26.67; p= 0.000).

7 For example, Xavier Gabaix and Augustin Landier (2008) show that the six-fold increase of CEO pay between 1980 and 2003 in the United States can be fully attributed to the six-fold increase in market capitalization of large companies during the same period.

8 The result of this test is not reported here due to the page limit but is available upon request.

Additional information

Notes on contributors

Kosta Josifidis

Kosta Josifidis is at the University of Novi Sad in Serbia.

Novica Supic

Novica Supic is on the Faculty of Economics at the University of Novi Sad in Serbia.