Capitalizing Time

For a general discussion of topics relating broadly to power and political economy (e.g., capital-as-power, Marxism, neo-classical economics, institutionalism).

Moderator: sanha926

Capitalizing Time

Postby blairfix » Fri May 02, 2014 3:31 pm

Fifty years ago, the average American male could support a (large) family entirely from his own income. Today, only the upper income bracket earners could afford to do so.

Over the last half decade, the size of the US labor force, as a portion of the total population, has steadily crept upwards. Can we explain this using CasP?

Here's an attempt. Capitalist share of income is well correlated with the relative size of the paid workforce (employees/total population).
Capitalist Income & Paid Work Linear.png
Capitalist Income & Paid Work Linear.png (172.5 KiB) Viewed 3221 times

Higher capitalist income = More Paid Time

Let's start with Veblen's hypothesis that humans have an innate desire to work and that most adults spend much of the day doing some sort of activity broadly called "work". This work can be paid or unpaid, but my hypothesis is that monetization of "work" shouldn't really affect how much human time is devoted to such industrious activity (although it will wildly alter what this "work" looks like).

Capitalist income depends on the monetization of "work". Unpaid workers don't have any money to give to capitalists.

From this it follows that increases in capital's share of income requires more "work" to be monetized.
Posts: 91
Joined: Fri Dec 20, 2013 3:45 pm

Re: Capitalizing Time

Postby jmc » Thu May 15, 2014 3:44 pm

Hey Blair,

Thank you for repeatedly sharing your graphs on the forum. So far, each one has been very interesting.

For this particular one, I have a technical question and a follow up.

1) What is the empirical definition for the measure "% of Population in Paid Work"? Is there a difference between this measure and the inversion of unemployment rate?

2) Is there a potential opposition between this figure and Nitzan and Bichler's claim that Share of Capitalist Income and Unemployment are positively correlated? If I recall correctly, they are critical of underconsumption theories of national income distribution. This is not a simple defense of their claim, but rather something that you could respond to in the future.
Posts: 40
Joined: Mon Oct 19, 2009 2:33 pm

Re: Capitalizing Time

Postby blairfix » Thu May 15, 2014 7:35 pm

To answer your question:

We can define 3 categories of people:

Employed: Have jobs
Unemployed: don't have jobs, but want one
Non-Employed: don't have jobs, don't want one

Employed + Unemployed + Non-Employed = Total Population

To construct the time series, I used Employed/Total Population

This is not the inversion of the unemployed population. It is the inversion of Unemployed + Non-Employed.

So the Employed/Total Population ratio can be affected by both a change in unemployment and a change in non-employment. Only during the Great Depression, was unemployment a significant factor in this ratio.

Regarding the share of capitalist income and unemployment, I find it helpful to disaggregate interest and profit. If we do this, we find that profit's share is negatively correlated with unemployment, and interest's share is positively correlated.

profit.jpg (18.43 KiB) Viewed 3180 times

Interest.jpg (17.28 KiB) Viewed 3180 times

I have thought a great deal about Nitzan & Bichler's findings on capitalist income and unemployment. The 3 year delay is what concerns me most. Without it, I get a negative correlation between the two:
Capital.jpg (16.13 KiB) Viewed 3180 times
Posts: 91
Joined: Fri Dec 20, 2013 3:45 pm

Re: Capitalizing Time

Postby Jonathan Nitzan » Tue May 20, 2014 8:45 am

Nonlinearities of the Sabotage-Redistribution Process

The exchange here suggests a possible confusion regarding our claims, so a clarification is in order. Over the years, we have argued that the relationship between sabotage and distribution tends to be nonlinear. Up to a point, sabotage redistributes income in favour of those who impose it; but after that point, sabotage becomes ‘excessive’ and the effect inverts. One illustration of this nonlinearity is given by the relationship between unemployment and the capital share of income.

Blair plots this relationship with the income share of capitalists on the vertical axis and the rate of unemployment on the horizontal axis. However, the low-pixel graphics of the chart are too crude to reveal the nonlinearity. Figure 1 corrects this shortcoming. It shows the same relationship, but with finer graphics that make the nonlinearity visible (the definitions and sources for all figures are given in the Appendix). Note that, unlike Blair, we use the capital share of domestic income rather than of national income. The reason is that the latter measure includes foreign profit and interest, which are unaffected by domestic unemployment. In practice, though, the two sets of data yield similar results.

fig01_large.jpg (137.06 KiB) Viewed 3122 times

Now, if we treat the entire 1929-2013 period as representing a single pattern, the relationship is negative. But we can also think of this history as representing two very different regimes, separated by what econometricians call ‘structural change’: (1) the prewar period (16 years), when sabotage was excessive and unemployment undermined the capitalist share of income (regression slope = –0.35); and (2) the postwar era (70 years), when, following a structural change, sabotage has become strategic and unemployment has boosted the share of capital (regression slope = +0.3).

To see the nonlinearity more clearly, Figure 2 smoothes the two variables as 5-year moving averages. The difference between the two regimes is now easier to discern. The negative prewar relationship is almost linear, while the positive postwar relationship is tighter than the one shown with the unsmoothed data. This relationship, using national income data, was first plotted in our paper ‘Capital Accumulation: Breaking the Dualism of “Economics” and “Politics”’ (Nitzan and Bichler 2000: Figure 5.2, p. 80) and later updated in various publications.

fig02_large.jpg (108.83 KiB) Viewed 3122 times

Figure 3 takes the analysis a step further by showing the relationship between the income share of capital and the rate of unemployment three year earlier (with both series still expressed as 5-year moving averages). The same relationship – though without the prewar data – was shown in Figures 15 and 16 of our paper ‘Can Capitalists Afford Recovery: Economic Policy When Capital is Power’ (Bichler and Nitzan 2013).

fig03_large.jpg (117.39 KiB) Viewed 3122 times

All three figures indicate a nonlinearity; this nonlinearity becomes clearer as we smooth the data; and the positive effect of strategic sabotage in the postwar period grows much tighter when we use unemployment with a three-year lag. These regularities suggest that strategic sabotage takes time to creorder the distribution of income. In principle, one can estimate this process with a distributed-lag regression, with the capital income share as the variable of interest and lagged values of unemployment as carriers; unfortunately, though, the high multicollinearity of the carriers will likely prevent us from assessing their distinct impacts.

Figure 4 shows Blair’s original relationship between the capital share of income (which, here too, we measure as a proportion of domestic rather than national income) and the per cent of the population engaged in paid work. To enable comparison with the previous three charts, we switch Blair’s axes, putting the capital share of income on the vertical axis and the proportion of paid workers on the horizontal one.

fig04_large.jpg (95.66 KiB) Viewed 3122 times

Our interpretation of this relationship is that the proportion of the population in paid work reflects the ability of capitalists to force people into the capitalization process, and that this process is a manifestation of capitalist power. But the effect of this power is nonlinear as well: beyond a certain point the underlying sabotage becomes ‘excessive’, the relationship experiences a ‘structural change’ and the impact on the capitalist share of income inverts. In the United States, this inversion appears to have happened after 1984: as the proportion of the population in paid work rose beyond 45 per cent, the income share of capital started to drop. Figure 5 smoothes both variables as 5-year moving averages, yielding a sharper picture of this nonlinearity.

fig05_large.jpg (95.23 KiB) Viewed 3122 times

Appendix: Definitions and Sources

Domestic income. Source: U.S. Bureau of Economic Analysis through Global Insight (series codes: GDY).

Domestic net interest. Source: U.S. Bureau of Economic Analysis through Global Insight (series codes: INTNETDBUS).

Domestic profit. Reported pretax and includes capital consumption adjustment (CCAdj) and inventory valuation adjustment (IVA). Source: U.S. Bureau of Economic Analysis through Global Insight (series codes: ZBECOND).

Population. Source: Historical Statistics of the United States: Earliest Times to the Present, Millennial Edition (online) (series code: Aa7 [till 1929]); U.S. Bureau of the Census through Global Insight (series code: N@US [from 1930 onward]).

Unemployment. Expressed as a share of the labour force. Source: Historical Statistics of the United States, Earliest Times to the Present: Millennial Edition (online) (series code: Unem-ployed_AsPercentageOf_CivilianLaborForce_Ba475_Percent [till 1947]); U.S. Bureau of Labor Statistics through Global Insight (series code: RUC, computed as annual averages of monthly data [1948 onward]).


Bichler, Shimshon, and Jonathan Nitzan. 2013. Can Capitalists Afford Recovery? Economic Policy When Capital is Power. Working Papers on Capital as Power (2013/01, October): 1-36.

Nitzan, Jonathan, and Shimshon Bichler. 2000. Capital Accumulation: Breaking the Dualism of "Economics" and "Politics". In Global Political Economy: Contemporary Theories, edited by R. Palan. New York and London: Routledge, pp. 67-88.
Jonathan Nitzan
Posts: 45
Joined: Sat Dec 06, 2008 2:39 pm

Re: Capitalizing Time

Postby blairfix » Tue May 20, 2014 8:45 pm

Great charts, Jonathan.

I think a really interesting CasP research agenda would be to look at the role of profit vs. interest. By aggregating capitalist income, we can get really interesting results, but we can get similarly interesting results by disaggregating.

For instance, the ratio of interest to profit shows a long-term correlation with unemployment, and no structural shifts are evident. Why is low profit and high interest worse than high profit and low interest? Both flow to capitalists ... what is the difference?

Interest Profit.png
Interest Profit.png (162.1 KiB) Viewed 3091 times

A note on this chart. If we include the years 1932 & 1933, but shift to the absolute value of the interest:profit ratio, we get an even better fit.

ABSt Interest Profit.png
ABSt Interest Profit.png (131.9 KiB) Viewed 3091 times
Posts: 91
Joined: Fri Dec 20, 2013 3:45 pm

Re: Capitalizing Time

Postby Jonathan Nitzan » Tue May 20, 2014 9:59 pm

For the empirical relationship between unemployment and the interest/profit ratio and what this relationship may signify, see:

1. Differential Accumulation: Toward a New Political Economy of Capital (1998), pp. 191-6.

2. Capital as Power: A Study of Order and Creorder (2009), pp. 256-8.
Jonathan Nitzan
Posts: 45
Joined: Sat Dec 06, 2008 2:39 pm

Re: Capitalizing Time

Postby blairfix » Wed May 21, 2014 10:55 am

Here's a way to analyse how the relationship between capitalist income and unemployment changes over time.

We plot the ten year moving correlation between unemployment and capitalist income. This gives a cyclical pattern that is mostly negative, but does become positive 3 times: After WWII, during the early 80s, and the early 2000s.

This cyclical correlation patter shows significant (inverted) similarities to Francis's buy-to-build index (after transforming the latter by removing the exponential growth trend and then taking the logarithm).

The unemployment rate is positively correlated with capitalist share of national income during lulls in mergers & acquisitions, and negatively correlated during booms in mergers & acquisitions.

Buy to Build UN_K.png
Buy to Build UN_K.png (164.3 KiB) Viewed 3067 times
Posts: 91
Joined: Fri Dec 20, 2013 3:45 pm

Return to Political Economy

Who is online

Users browsing this forum: No registered users and 1 guest