Defending the economics discipline and arguing it is ready for a paradigm shift
Review of Cogs and Monsters: What Economics Is, and What It Should Be by Diane Coyle
I have always felt ambivalent toward economic “science”. It is highly mathematical, constructing abstruse models often impossible to connect to real life (I mean, “aggregating individual utility maximization”?) and its proponents act as if they are gurus while flatly contradicting each other. I struggled with it at 3 universities. In spite of my skepticism, there is no denying how useful economics was to me professionally and indeed, how it fascinates me. This brilliant book presents a defense and critique of it, arguing that a fundamental overhaul is required if its relevance is to continue.
The images of the title are well chosen. “Cogs” refer to the traditional economic assumption of individuals interacting as independent agents in an essentially static context. A result (its “equilibrium”) is straightforwardly linear, individual preferences are fixed, their influence on the aggregate negligible. A part in a machine.
In contrast, “monsters” are of the digital age, which is social, even collective; their interactions can be exponential in impact, snowballing suddenly into unpredictable and disproportionate results – distinctly non-linear, with fluid preferences and an elusive end state. If there are equilibria, they are unstable, dynamic, unfixed, multiple. Chaos.
Coyle begins with a look at what the market can determine. The free market, she explains, has indisputably made people better off. What it does is provide information on, and to some degree the means of satisfying, individual preferences. “What markets do brilliantly,” she writes, “is coordinate the use of resources in a process of discovery and challenge.” Price, where supply and demand meet, serves as the mechanism, a rationing device worked out between buyers and sellers. Notably, this can be fruitfully used in cost-benefit analysis, a systematic calculation of pros and cons and “rational framework” for public policy decisions. This was what formed the bulk of my graduate training.
However, she notes, the market ignores or neglects many important questions, from externalities (pollution, global warming) to fairness and other civic values. Economists, she argues, must take account of this because they are both participants in and shapers of society – what they advocate not only changes things in a self-fulfilling manner, but also reflects their personal biases, of which many remain unaware, in particular the neo-liberal variety. Moreover, according to Coyle, many of the assumptions about homo economicus (the cog) are demonstrably inaccurate. For example, individuals rarely make choices as a function of rational or selfish calculation. Instead, they operate from rules of thumb, from gut feeling, usually with badly incomplete information. The advertising industry, like propaganda, capitalizes on the distortion of information in pursuit of irrational purpose. Finally, in attempting to portray the economy as a clockwork mechanism, economists neglect narrative and historical development, which can have a decisive impact on both personal choice and the overall direction of society. This fatally narrows their vision. They pay no attention to the roles of power and politics – they can be fired if bosses don’t like their views – or the issues of implementation, especially how unpredictable outcomes and behavior can be.
Coyle then shifts to the details. Though to please their political masters many economists prefer to claim objectivity and certainty, she says, they understand how tentative many of their claims are; econometrics – statistics analysis applied to real-world data sets – can show correlation, but cannot prove causation. And yet, she writes, econometricians know “the temptation to prefer inaccurate precision to accurate imprecision that would more properly characterize noisy data….eager to claim an empirical silver bullet.”
The remedy to this, she argues, is that they should rely less on the determinism of mechanistic models and more on narratives (the ad-hoc-ery that academic economists traditionally despise). This would require modesty, the acknowledgment of unforeseen contingencies, and awareness that their prescriptions may not lead to the desired results. Because the profession can’t or won’t, she concludes, economists confuse description with normative judgments that can have severe consequences in the real world. A prime example of this is the advocacy for austerity policies in places like Greece during its currency crisis, which was an economic and political disaster.
She also goes into depth about the evolving challenges to economic measuring. The information age, she argues, is rendering the machine metaphor obsolete. First, information products benefit from increasingreturns to scale – once software is written, it costs nothing to install and wears out only due to human intervention. As such, growth is self-reinforcing, like a snowball gathering mass as it rolls downhill. This stands in stark contrast to industrial goods, the fixed costs of which rise when called upon to provide higher levels of production unless investments are made to increase manufacturing facilities.
Second, as with the butterfly in Brazil whose beating wings might cause a hurricane in Florida, small changes can lead to huge differences with unprecedented rapidity, also a trait that conventional economics neglects.
Third, given network effects, the information economy leads to unusual aggregates, in which the whole can be far greater than the sum of individual parts. Among other things, this means that the more users there are, say on a platform like facebook, the more value accrues to users, rendering competition more difficult, if not impossible, to mount. Again, this is the opposite of mass production.
Because we do not currently have a way to measure these differences, we may understand the information economy far less than many economists assume. Coyle’s exposition got me to seriously reconsider my longstanding dismissals of the info economy’s mysteriously low scores in productivity enhancement in the workplace. For example, we need to learn how to gauge the value of data flows, cloud computing, and applied AI.
Finally, Coyle directly attacks the aggregation of individual utility curves, in particular how rising GDP or consumption often fails to translate into enhanced social welfare. This is, she argues, the default position of the many economists that continue to think conventionally, in linear algebraic algorithms (e.g. price x quantity = revenue, which is demonstrably untrue in many instances today). What we need, she suggests, are computer models that take into account the new types of feedback loops, interdependence, increasing returns to scale, and the like.
More radically, Coyle argues that the welfare of society must be put into the heart of economic thought rather than merely an individual’s utility. We must study how economic policies impact institutions and society, which is a call to return to “political economy”. In this schema, the government must play a larger role, not just as a regulator or neutral arbiter, but as an active agent that intervenes in the service of opportunities (or “where the economy needs to go”, as she puts it). Bottom line: the strict laissez faire of neo-liberalism is obsolete, which an increasing number of economists is recognizing.
This was a tremendously interesting and useful book for me, renewing my faith in economics. It is beautifully written in the form of dense, evocative essays that assume an undergraduate knowledge of economics at a minimum. It re-opened my mind. Can’t get a better recommendation than that.
Related review: