A theory of ethically-biased technological change (toward an empirical test of the Frankfurt School)

The Puzzle

There is a widespread consensus in the economics literature that the rise of information technology over the past several decades has been a “skill-biased technological change.” A skill-biased technological change increases the productivity of skilled workers more than it improves the productivity of unskilled workers, therefore increasing the demand for skilled workers relative to unskilled workers. This increased demand for skilled workers, ceteris paribus, increases their income relative to unskilled workers. It is widely agreed that this is a key explanation for at least some of the rapid increase in income inequality since the 1980s.1 One key observation consistent with this theory is that it is only after the appearance of micro-computers that income inequality begins to rise in the early 1980s.2 A second key observation is that more educated workers are more likely to use a computer on the job3.

Despite this growing interest in the economic implications of technological change, social scientists have failed to take seriously a surprisingly similar line of thought advanced most notably in twentieth-century German philosophy. Heidegger and, later, the Frankfurt School theorists developed critical theories of technology which argued, in different ways, that technological changes shape ethical attitudes toward the world. In particular, a specific line of thought revolves around the relationship between technological development and instrumental rationality, or in other words formal analytical rationality which renounces consideration of substantive goals and concerns itself only with calculating the optimal means toward given ends.

For Heidegger, from a phenomenological perspective, technology represents a “clearing” of Being which generates a relation to the world in which nature and other humans are merely “standing-reserve” waiting to be exploited as means to whatever ends. For Adorno, Horkheimer, and especially Marcuse, modern technological advancements were “organizing and perpetuating (or changing) social relationships, a manifestation of prevalent thought and behavior patterns, an instrument for control and domination.”4

While positivist social scientists have long perceived these types of claims from continential social theory to be hopelessly speculative and empirically intractable, this is puzzling given that the core claim emerging from this particular current in the critical theory of technology is relatively straightforward: technological development increases the prevalence of instrumentalism among other possible ethical attitudes toward the world. Moreover, the theory of skill-biased technological change is a powerful heuristic for modeling this argument given that it is a well-established model of how technology can generate unequal distributions of power without any reference to Marxian bugaboos which have likely caused more confusion than clarity by now.5

However, it is perhaps more fair to observe that critical theorists indeed sometimes lack what contemporary social scientists would call “mechanisms,” or specifiable and testable triggers which effectively do the work of some variable X supposedly causing some variable Y. While this particular critique of technological development is not delivered with any particular “mechanism” which would be intelligible to many empirical social scientists working today, critical theory does supply many useful hints in this direction. Moreover, the basic argument is surprisingly consistent with well-known economic models, however disparate might be the vocabularies, thus permitting the elaboration of a critical theory of technological change empirically testable and consistent with a rational-choice framework.

A Theory

To better specify the mechanism whereby technological change may plausibly generate a greater prevalence of instrumental ethics, I propose a theory of ethically-biased technological change. I argue that certain technological changes can increase the income paid to certain ethical orientations relative to others. As in the theory of skill-biased technological change, certain technological changes increase the productivity of certain ethical orientations relative to others, increasing their relative demand, and therefore increasing their relative income. This increase in relative income represents a premium paid to those with the ethical resources required by the new technology, just as skill-biased technological change generates a premium paid to those with the technical resources required by the new technology.6

More specifically, a technological change which increases productivity but transgresses any particular ethical value will increase the productivity of individuals with the weakest commitment to that ethical value relative to those more firmly committed to that ethical value, thus increasing demand for them relative to the more ethical individuals, and thereby increasing their relative income. Ethically transgressive technological innovations increase the productivity of the least ethical because ethical compunctions are effectively a psychological tax on a worker. It is hard to take a job, and then demoralizing to work it, to the degree one feels ethical compunction for doing so. While an employer would have to pay very high wages to attract highly ethical workers to an ethically transgressive job, the wages they have to pay to attract relatively unethical workers will tend to be lower. The firmer one’s ethical commitment, the greater is the tax one pays to work an ethically transgressive job, the less worthwhile the wage, and the less likely is one to take the job; the weaker one’s ethical commitment, the lower is the tax, the more worthwhile is the wage, and the more likely one is to take the job. The overall implication is that technological change rewards individuals with weak ethical convictions relative to individuals with strong ethical convictions, and therefore exerts a direct, material pressure on the evolution of ethics within society.

There are three key, testable implications we would expect to observe if this model indeed captures important aspects of the historical record. First, during times of ethically transgressive technological innovation, the first workers will be individuals who before the technological innovation were already less ethically scrupulous on the relevant ethical dimension. Second, workers of the new technology will be paid a premium which would be, in principle, distinct from any skill premium, although of course these can be correlated. Third, if we add the assumption of modern capitalist institutions, wherein wage workers have no common land and no immediate means of production, then the price of integrity for committedly ethical individuals literally becomes death. This last implication means that, if technological change is ethically biased as I hypothesize, the steady-state to which it must tend in the long-run is the absolute evacuation of integrity from all participants in the labor market. Note that this can come about through two different causal pathways: ethical individuals can learn to sacrifice their integrity to compete with the unethical and survive, and/or ethical individuals can literally die off while unethical individuals thrive and replace them (evolutionary change). It should be noted explicitly that while observations 1 and 2 show how ethically-biased technological change reward certain ethical orientations differently and unequally,

Anecdotally, a general acquaintenace with the contours of modern history seems to suggest some striking evidence for this model of ethically-biased technology change. Consider, for instance, that while a certain amount of philosophical liberalism was no doubt necessary for the industrial revolution to emerge, and this perhaps explains why it began in England, it seems equally undeniable that the industrial revolution in turn generated an extraordinary instrumental turn and valorization of self-interest wherever it went. It will obviously require more work to test this expectation more rigorously, but it seems highly probable.


In summary, building from insights of the Frankfurt School, I provide a specific, empirically testable mechanism which accounts for the large-scale historical variations in ethical outlook they described to the dismissal of the mainstream social scientists of the time. Working heuristically from rational-choice economic theory, my stylized theoretical model is consistent with but more general than the rational-choice framework because the model here in some sense explains what rational-choice models merely assume (the prevalance of instrumentally rational dispositions).

If correct, the long-run, world-historical implication of this model could not be more staggering for understanding the ethical experience of living through modern, let alone contemporary, capitalism: if this model is correct, then it would mean that today we have inherited a world in which ethical integrity has been subjected to a roughly 300-500 year forced evacuation by labor market institutions.

  1. See Card, David, and John E DiNardo. 2002. “Skill Biased Technological Change and Rising Wage Inequality: Some Problems and Puzzles” and Guvenen, Fatih, and Burhanettin Kuruscu. 2010. “A Quantitative Analysis of the Evolution of the U.S. Wage Distribution, 1970–2000.” NBER Macroeconomics Annual 24(1): 227–76.

  2. Katz, Lawrence F. 2002. “Technological Change, Computerization, and the Wage Structure.” In Understanding the Digital Economy Data, Tools, and Research, Cambridge: MIT Press.

  3. Autor, David H, Lawrence F Katz, and Alan B Krueger. 1997. “Computing Inequality: Have Computers Changed the Labor Market?”

  4. Marcuse, H. 1941. “Some Social Implications of Modern Technology.” Studies in Philosophy and Social Science, Vol. IX.

  5. I have in mind such perennial bad habits as lazily but self-certainly referring to Capital as a conspiratorial agent, an argument which has never boded very well for Marxism.

  6. Indeed, the analogy between skill-biased and ethically-biased technological change is all the more apt if we, with Foucault, conceptualize the practice of ethos as itself a kind of technical game; if ethical resources are themselves understood as technologies of the self. Key examples in this context are those most bourgeois of ethical resources: the various techniques of repressing ethical compunctions and moral anxieties, and the various techniques of guilt management consequently required by this ethos. In this register, ethically-biased technological change would be essentially identical to skill-biased technological with the exception that the “skills” in question are the skills of self-management rather than economic production. The reason we must avoid this reductionism at the outset is that it would beg the most important question, namely, the question regarding how exactly did the contemporary human become, quite peculiar in the long-run of world history, so instrumentalist that the only activities remaining in contemporary life are exploiting and suffering exploitation?

Cite this post: RIS Citation BibTeX Entry

Murphy, Justin. 2014. "A theory of ethically-biased technological change (toward an empirical test of the Frankfurt School)," http://jmrphy.net/blog/2014/06/21/a-theory-of-ethically-biased-technological-change-toward-an-empirical-test-of-the-frankfurt-school/ (April 24, 2017).