Surplus Value Production and Realization in Marxian Theory – Applications to the U.S., 1990–2015

ABSTRACT

Marx's Capital shows that surplus value can be produced in one industry, yet realized as profit (and possibly revenue) by other industries over the course of circulation. This paper highlights the separation between surplus value production and realization in Marx's work, and develops a new method for estimating surplus value production at the industry level to trace out transfers of surplus value across industries. The framework is based on the ‘New Interpretation’ and links money value added to surplus value production at the industry level. Data on value added by industry for the U.S. are used to estimate surplus value production by industry. The analysis allows comparison of surplus value production and realization in each industry. The pattern of differentials between surplus value creation and realization across industries sheds light on the processes of capitalist competition and points to a source of instability for capitalist economies.

KEYWORDS:

• Circuit of capital
• labor theory of value
• productive and unproductive labor
• surplus value
• transfers of surplus value

JEL CODES:

• B5
• B51
• E11

Acknowledgements

The author thanks Duncan Foley, Roberto Veneziani, and three anonymous referees for invaluable comments and suggestions on earlier versions. The usual disclaimer applies.

Disclosure statement

No potential conflict of interest was reported by the author.

ORCID

Jonathan F. Cogliano http://orcid.org/0000-0002-5292-7727

Notes

1 For further exposition of the NI see Foley (1982, 1986, 2000) and Mohun (1994, 2004). For a critique of the NI see Shaikh and Tonak (1994) and Shaikh (2016).

2 The finer points of the transformation problem are beyond the scope of this paper. Helpful summaries can be found in Foley (2000), Moseley (2011), and Mohun and Veneziani (2017).

3 Thanks are owed to an anonymous referee for pointing out this added detail of the circuit of capital, which is often abbreviated as $M\to C\dots P\dots C^{\mathrm{\prime }}\to M^{\mathrm{\prime }}$.

4 For Marx, the competition between industrial and financial capital also forms the interest rate (Marx 1981, p. 493). The Marxian theory of interest is beyond the scope of this paper, but for detailed analyses of the interest rate in Marxian theory, see Panico (1980, 1988).

5 To further explain the assumption of a uniform rate of surplus value Marx (1981, p. 241) refers to Smith’s (2000) discussion of the equalization of the advantages and disadvantages of labor. For further discussion of Marx's equalized rate of surplus value assumption see Cogliano (2011, 2013) and Cogliano et al. (2018).

6 Assuming a uniform $e$ may be a strong assumption to make given the institutional barriers to labor mobility that exist in modern economies. One possible view in light of the labor market segmentation literature is that persistent wage differentials within firms and/or lines of work are the result of barriers to mobility. However, as Gordon, Edwards, and Reich (1982, p. 192) note, studies of segmentation have shown mixed results and do not fully confirm the existence of labor market segmentation. Further, Gordon, Edwards, and Reich’s (1982, p. 193) study of labor market segmentation is limited by available data (industry-level data only) and unable to examine the existence of internal labor markets within firms or the ways in which different firms within industries may fit into a story of segmentation, thus the analysis is only approximate. Studies of labor segmentation are also conducted using data on money value added and inherently focus on magnitudes after the realization of surplus value has taken place. Thus, labor segmentation analysis to date is only partially applicable to Marxian analyses and does not necessarily indicate that some equalization of $e$ is not operating, although certain insights from the labor segmentation literature would be informative for more direct Marxian studies of labor mobility.

7 Any empirical estimation of labor values implicitly makes some assumption about the distribution of $e_i$ in order to derive estimates for labor values from available price data. Morishima and Seton (1961, p. 206) were the first to recognize the need for some assumption about the rate of surplus value to estimate labor values.

8 Marx's assumption of a uniform $e$ implies that the conditions and experience of exploitation are uniform across workers. There are open questions as to the coherence of this approach and the labor theory of value as the basis for a theory of exploitation. These questions have generated a large literature on exploitation theory—largely based on Roemer (1982)—with recent examples found in Cogliano, Veneziani, and Yoshihara (2016), Skillman (1995, 2014, 2017), Veneziani (2007, 2013), Veneziani and Yoshihara (2012, 2013, 2015, 2017), and Yoshihara (2010). It is worth noting that Veneziani and Yoshihara (2013) show how a coherent set of definitions of exploitation and classes is consistent with the NI. A comprehensive survey of developments in exploitation theory is provided by Yoshihara (2017).

9 This can be shown in the standard $n$-commodity linear economy. These results are available from the author upon request.

10 This type of transfer could be viewed as a transfer of value within the circuit rather just a transfer of surplus value. In fact, some parts of the FIRE industry's income could be perceived as transfers of value from households as well.

11 The spreadsheets used to organize the data and the code to perform the surplus value imputation and to generate the charts in Section 4 are available from the author upon request. The imputation and charts are done in Mathematica version 11. Additional details of the data are provided in Appendix A.

12 The shaded regions in all figures denote NBER recessions rounded to the nearest year.

13 The imputation method of this paper does not make adjustments for the surplus value that could be transferred from the rest of the world to the U.S. or vice versa via international trade. If prices in every country were proportional to labor values yet countries could have different $e$, then there will be transfers of surplus value between countries engaging in international trade. If prices are not proportional to labor values, which is the general case, then there will inevitably be transfers of surplus value between countries through international trade.

14 Thanks are owed to an anonymous referee for clarifying this detail.

15 Results using these alternative assumptions are available from the author upon request.

16 Each set of bars respectively sum to one.

17 Thanks are owed to an anonymous referee for highlighting this detail.

18 Some earlier empirical investigations into profit rate equalization find results consistent with Duménil and Lévy’s (2002) later study. For example, Glick (1985) and Semmler (1981) find empirical support for a long-run tendency of profit rates to equalize across industries. Glick (1985) shows that from 1948-1979 inter-industry profit rate differentials across many of the same manufacturing industries studied by Duménil and Lévy (2002) are small, which can be taken as an indication that, over time, profit rate fluctuations take place around some center of gravity for the most part. Semmler (1984) also finds that there is evidence of increased capital mobility over time and that pricing practices are consistent with profit rate equalizing prices of production in the long run.

19 Earlier studies by Glick (1985) and Semmler (1981, 1984) also exclude financial and other unproductive industries.

20 Private goods-producing industries are defined by the BEA as Agriculture, Mining, Construction, and Manufacturing.