“All that frenetic switching between left and right means something. It’s not random noise. It’s as if you’re flipping a coin. Of course you expect as many heads as tails, but instead you get twice as many heads as tails. So you conclude that the coin is loaded, in our case, that the polarization modulation isn’t random; it has content….”
—“Ellie Arroway,” Contact, by Carl Sagan.
Recent studies have found that the distributions of a number of important economic variables follow well-known patterns or distributional forms. These include changes in stock prices, corporate rates of growth, enterprise profitability, and individual income. These persistent distributional forms are a real boon for economic analysis: they allow us to take a “macroscopic” and observational approach to economic analysis. Instead of formulating economic theories at the micro level, based on detailed descriptions of individual characteristics and interactions we can never hope to observe (as much contemporary economics does), we can take these persistent patterns as the observable outcome of social interactions that economic theory needs to explain.
Many observed economic distributions — such as those involving profitability, income, or the valuations of corporate securities — have a very specific mathematical signature. Technically speaking, they are distributions that maximize entropy over phase spaces that are at least partly described by constraints on the first moments of the distributions they support. In other words, those distributions suggest that the economic processes shaping individual measures of those variables are zero-sum games. They are processes in which any increase in the value of the variable in question for one individual is necessarily accompanied by an equal total decrease in the value of the variable for all other individuals.
This finding poses a puzzle for economic theory. It suggests that these variables are subject to “conservation principles,” that is, that at least over some time horizons, total quantities of them across the economy are fixed. But we know this is not true. In fact, one of the chief reasons economics emerged as a discipline was the need to explain the ceaseless variations in individual and aggregate measures of variables like these.
Why, then, are so many important economic variables distributed as if they were subject to a conservation principle? As I suggest in a recent paper, they likely do so because competition in capitalist economies often gives rise to processes of social scaling.
Social scaling can be explained with a simple thought exercise. Suppose that in a community all individuals evaluate their well-being in line with the idea of “keeping up with the Joneses.” That is, they do so according to how their level of consumption compares to the average measure of consumption in their community. It should be easy to see how this creates complex, dynamic interdependences between individual measures of well-being. If, for example, the Joneses increase their consumption not only will their well-being increase, but that of their neighbors will decrease, as their consumption relative to the Joneses falls. This kind of valuation can result in zero-sum interdependences that reflect not the presence of a “conservation principle,” but the irreducibly social content of economic competition.
According to this principle of social scaling, many complex competitive interactions in capital, labor, and product markets boil down at least partly to rather simple, zero-sum interactions of this kind. Social scaling can help account for observed distributions of “Tobin’s q” (a measure of the market valuation of a corporation’s liabilities relative to their assets), as well as for observed distributions of income.
The distribution of income offers a good example of the principle of social scaling. Because capital can move with relative ease, there is a strong tendency for realized rates of profit to equalize across enterprises. Labor, however, cannot move so easily. Skills and experience — not to mention discrimination based on race, gender, immigration status, etc. — limit the mobility of labor across jobs and across levels of pay. As a result, wage earners find themselves largely cordoned off into specific, smaller segments of the labor market.
The relative bargaining power of wage earners and employers shapes the division of value added into wages and profits in each of these labor-market segments. A wide range of factors shapes the bargaining power of either group, including the size of potential labor supply relative to demand, the extent and effectiveness of trade union organization, the confidence and broader social standing of wage earners in that segment, the political and regulatory climate, etc.
An increase in the bargaining capacity of a worker or group of workers in a particular segment of the labor-market will result in an increase in their distributional share at the expense of the profitability of their employers. But competition in capital markets ensures that the loss in profitability tends to be spread across all enterprises. Facing lower returns, all enterprises push back on wages across all segments of the labor market. Workers who have not experienced improvements in their bargaining capacity will suffer a reduction in their own share in value added as a result. An individual’s wage income is shaped by a social scaling of their bargaining power by an average or social measure of bargaining power across all wage earners.
As a result, the distribution of wage income is shaped by thorny zero-sum interdependences among wage earners that have been largely ignored by economics and most strains of political economy. Yet those interdependences have profound implications for our understanding of income inequality, and of the kinds of interventions that may succeed in reducing it. Most broadly, they suggest that without concerted, collective action that boosts bargaining power across all segments of the labor market, distributional gains by any group of workers will at least partly come at the expense of workers whose economic, social, or political conditions place them in weaker bargaining positions.
Social scaling can help account for the observed distributions of several economic variables along these lines. In doing so, it reinforces a central point from the classical political economy of Adam Smith and Karl Marx: behind complex patterns of economic interaction between large numbers of people there often lies emergent social content. The principle of social scaling identifies a tell-tale, observable signature that suggests that some outcomes of market competition embody a rather simple social normalization or scaling of certain individual characteristics. This draws attention to social and systemic features of decentralized capitalist economies that are not often considered by economic analysis.
Paulo L. dos Santos is an Assistant Professor of Economics at The New School for Social Research.