The Frontal Cortex

Empiricism, Economics and Mystery

I finally got around to reading Paul Krugman’s takedown of modern economics, which is a lucid dissection of his own field. His core argument is that economists made the old Keatsian error, mistaking a beautiful theory for the truth:

As I see it, the economics profession went astray because economists, as a group, mistook beauty, clad in impressive-looking mathematics, for truth. Until the Great Depression, most economists clung to a vision of capitalism as a perfect or nearly perfect system. That vision wasn’t sustainable in the face of mass unemployment, but as memories of the Depression faded, economists fell back in love with the old, idealized vision of an economy in which rational individuals interact in perfect markets, this time gussied up with fancy equations. The renewed romance with the idealized market was, to be sure, partly a response to shifting political winds, partly a response to financial incentives. But while sabbaticals at the Hoover Institution and job opportunities on Wall Street are nothing to sneeze at, the central cause of the profession’s failure was the desire for an all-encompassing, intellectually elegant approach that also gave economists a chance to show off their mathematical prowess.

That’s a lovely explanation – one might even call it pretty – but it strikes me as a little too neat. The larger problem, as I see it, is a false certainty in knowledge, which goes well beyond mere aesthetics. (The beautiful models make it easier for us to slip into certainty, but they are merely one of many factors. The history of science, after all, is filled with ugly models that persisted for centuries – just look at epicycles.) This certainty, I think, stems from our optimistic narratives of scientific progress, in which human knowledge, through the miracle of experimentation and peer-review, travels in a straight line from utter mystery to a unifying theory. Here’s how I summarized this naive belief a few years ago in a Seed article:

The history of science is supposed to obey a simple equation: Time plus data equals understanding. One day, we believe, science will solve everything.

But the trajectory of science has proven to be a little more complicated. The more we know about reality–about its quantum mechanics and neural origins–the more palpable its paradoxes become. As Vladimir Nabokov, the novelist and lepidopterist, once put it, “The greater one’s science, the deeper the sense of mystery.”

Consider, for example, the history of physics. Once upon a time, and more than once, physicists thought they had the universe solved. Some obscure details remained, but the basic structure of the cosmos was understood. Out of this naïveté, relativity theory emerged, fundamentally altering classical notions about the relationship of time and space. Then came Heisenberg’s uncertainty principle and the surreal revelations of quantum physics. String theorists, in their attempts to reconcile ever widening theoretical gaps, started talking about eleven dimensions. Dark matter still makes no sense. Modern physics knows so much more about the universe, but there is still so much it doesn’t understand. For the first time, some scientists are openly wondering if we, in fact, are incapable of figuring out the cosmos.

Or look at neuroscience. Only a few decades ago, scientists were putting forth confident conjectures about “the bridging principle,” the neural event that would explain how the activity of our brain cells creates the subjective experience of consciousness. All sorts of bridges were proposed, from 40 Hz oscillations in the cerebral cortex to quantum coherence in microtubules. These were the biological processes that supposedly turned the water of the brain into the wine of the mind.

But scientists don’t talk about these kinds of bridging principles these days. While neuroscience continues to make astonishing progress in learning about the details of the brain–we are a strange loop of kinase enzymes and synaptic chemistry–these details only highlight our enduring enigma, which is that we don’t experience these cellular details. It is ironic, but true: The one reality science cannot reduce is the only reality we will ever know.

The fundamental point is that modern science has made little progress toward any unified understanding of everything. Our unknowns have not dramatically receded. In many instances, the opposite has happened, so that our most fundamental sciences are bracketed by utter mystery. It’s not that we don’t have all the answers. It’s that we don’t even know the question.

What does this have to do with modern economics? Given the financial turmoil of recent years, I think it’s plausible to argue that economists in 2009 are in a similar position to physicists in 1909: they’ve just realized that their elegant models obscure an awful lot of ignorance. The question, of course, is what the new model will look like. Will neuroeconomics amend the rational agent model, just as Einstein amended the Newtonian conception of space-time? Or does economics need the equivalent of quantum mechanics, a radical new theory that will unravel the most basic tenets of the field? Who knows? I surely don’t. But the real lesson is that this isn’t just a problem of economic theory, or pretty theories that look too pretty to be true – this is the ancient problem of knowledge, which bedevils every empirical field again and again.* The world has humbled us once again.

*And I think we’re especially sensitive to the failures of economics, since they so directly concern our pocketbooks. String theory could be falsified tomorrow and I have a feeling most people wouldn’t notice.

Comments

  1. #1 Michael F. Martin
    September 21, 2009

    I agree that Krugman’s “seduced by aesthetics” principle is too pat as an explanation. When the data is there to falsify a beautiful theory, it will be discarded — which points to the real problem, which is that there hasn’t been enough data to test economic theory.

    Economies, like many complex systems, has been measured only in two basic modes — at the granular scale of individual behavior or at the grand scale of societal movements. We can watch measure a single molecule’s diffusion in solution or take the temperature of the solution, but we can’t watch every molecule within the solution.

    Except that now we can when it comes to economics. Now we have data sets generated by consumer web traffic that are both granular enough to characterize individual behavior and grand enough to encompass millions of such individuals. Whatever new theories of economics will guide us in the future, I think it safe to say that these will be drawn from the new data.

    The same was true of physics.

  2. #2 David
    September 21, 2009

    To Michael F. Martin: Predicting consumer behavior (70 percent of the economy) at a certain level of confidence only gets you so far. Who didn’t already know, web data or not, that consumers were spending far beyond their means? Who couldn’t have predicted that consumers would cut back when unemployment rose and credit dried up? Not to mention that not all consumer behavior occurs on the web.

    As a thought experiment, try making a long list of other factors that affect economies and tell me where you’re going to get your data. Not to mention what your algorithms are going to look like.

  3. #3 Burt
    September 21, 2009

    Burt’s conjecture:

    If economics were a science, then there would be many wealthy economists. Unfortunately the usual way economists make money is to write books and articles touting their expertise in economics – the same way as hucksters do.

  4. #4 Anahita
    September 21, 2009

    Perhaps a *bit* off topic but as it relates to neuro-economics and because Dan Ariely was in this interview, thought I’d share from one of my favorite podcasts (All in the Mind):

    http://www.abc.net.au/rn/allinthemind/stories/2009/2435763.htm

    On a more related note – thank you for writing about the paradox of our ever-growing realization that what we don’t know seems always to exponentially outgrow what we DO know. Humbled, indeed.

  5. #5 Gray Gaffer
    September 21, 2009

    He describes a beautiful theory about why beautiful theories don’t work.

    The flow of money is not determined by mob rule. It is determined by the incentive structure of the banking and stock market systems. This time the failure mode was grossly overly-optimistic leverage vs exposure estimates. We failed to allow the dumbasses who made those calls to fail. It will happen again.

    And, of course, the above is also a beautiful theory.

  6. #6 Anahita
    September 22, 2009

    “He describes a beautiful theory about why beautiful theories don’t work.”

    That might be one of my new favorite quotes ;)

  7. #7 DR
    September 22, 2009

    Like politics, economics is a social construct. Finance is a mathematical one. Through the alleged power of its mathematical models (together with lavish monetary consideration), Wall Street found it rather easy to persuade policy makers that economics was and ought to be outside their purview. The New Deal yielded to Reagonomics under the spell of some mystic algorithms. Deregulation is the result, the goal, in fact, of mathematical determinism.

  8. #8 oldfuzz
    September 23, 2009

    The probability of economists predicting economic behavior is confounded by the never to be known joys and fears of the human variable… all six billion of them.

    However, economic predictions are probably more reliable than neuroscience predictions and weather forecasts.

    All three are forecasts based on the empirical data gathered from non-deterministic–both non-linear and stochastic–events.

The site is undergoing maintenance presently. Commenting has been disabled. Please check back later!