The Omitted Variable Bias We Call "Being Human"

©Smith Patel via Unsplash

I. The Illusion of Control: Why Models Seem Too Theoretical 

People are clearly embedded in the foundations of all academic and interdisciplinary curricula, ranging from econometrics to the philosophy of economics. Their interactions, their behaviour, and ultimately their decisions. Some subjects explore the rational aspects of such behaviour in greater detail, as they expect to uncover tangible truths or create consistency in a world full of intangible variables. In econometrics, these are called random variables. In traditional Marshallian economics, these factors are held constant; the classic "ceteris paribus" is invoked because otherwise the analysis becomes "too messy" or, in economic terms, "undetermined" and complicated, leading to conflicting solutions.

But consider what happens when a market crashes. On September 15, 2008, Lehman Brothers collapsed, forcing every model that had comfortably held its variables constant to let go. The random variables rushed in. The error terms, as econometrics calls them, turned out to have names: panic, herd behaviour, and a fund manager who hadn't slept in three days making a decision worth hundreds of millions with shaking hands. Ceteris paribus had always been a polite fiction. A crash is the moment it stops being polite.

The crash is what the models never fully account for: the uncertainty of humans and human behavior. They make reckless decisions when random variables and emotions, which, needless to say, overpower the independent variable of thought and affect the dependent variable, the consequent decision, in an unprecedented manner. Unfortunately, for those who view people as subjects and their actions as data, humans do not possess a fixed mandate with rules applicable to specific situations. Even if such a mandate existed, the possibility of omitted-variable bias would always persist, as unpredictability is the prevailing law, and few laws endure the relentless passage of time.

II. The Real Variables: Beyond Emotions

A potential fallacy here is that emotions are not the only random variable. In this practical world we live in, many factors influence its functioning: experiences, habits, desires, and, most importantly, circumstances. In 2008, the housing market was built on borrowed confidence. The habit involved trusting the models. The desire was another quarter of growth. None of these is exactly an emotion. But none of them is rational, either.

Humans, the most evolved and distinct members of the animal kingdom, are left at the mercy of mere situations in their lives, as in these situations, their most formidable faculties, their heart and mind, collide.

A trader in freefall is not irrational. They are too rational, in too many directions at once, paralysed by the collision of what they know and what they feel.

Something else is worth investigating. We live in a world where technology has made transactions, communication, and even forms of human expression, such as emoticons, more digital. 

But what's more subtle and remarkable is how fluently we've embedded human psychology in every subject we interact with. In mathematics, we make certain assumptions and derive solutions based on them. But how often do we pause to ask what these solutions help solve? These complicated methods—elimination, simultaneous equations, and the models that helped price the very mortgage securities that collapsed in 2008—lead us to clean numbers, but they take a rather confusing route to arrive there. Today, esteemed professors and institutions learn these methods and pass them to future generations. But why? What is so important in these equations, in economic laws and psychological frameworks like cognitive dissonance, that the most brilliant minds of our time dedicate their energy to them?

III. Why We Still Build and Believe in Models: The Eternal Search for Comfort

The answer is straightforward and less complex than the means to the end. We are finding ways to make sense of the world. Through different mechanisms, we are seeking comfort in what we know so that we are prepared against any irregularities, any random errors, as econometrics would say. In astrophysics, they are called chance alignments. They are referred to as "black swans" in the financial industry. The name changes. The fear underneath does not.

Being “open” or “exposing ourselves to risk without any sort of control measure or backup plans” puts our innate understanding—our instinct to fight, flee, or freeze—in jeopardy. When exposed to such situations, the human brain is more likely to make mistakes and to fail to draw reasonable conclusions from experience. Therefore, some derivation of predictable outcomes helps us feel at ease because we believe we have hoarded enough knowledge to protect ourselves in case of discontinuities or the unexpected. A market crash makes the inconsistency visible. It is the moment the model fails that the human race must decide, alone, without the comfort of the equation, what to do next.

Perhaps this moment is the pure innocence of being human. We give everything we have to tame something, whether it is our experience or our nature, which is by definition untamable and not under our control. The market, it turns out, was never a machine. It was always a mirror. And what it reflects, in the cold light of a crash, is not a broken algorithm. It is us, in all our terrifying uncertainty. How we approach this complicated truth is diverse: equations, philosophies, econometric models, and the occasional prayer. But at the end, our aim converges to one thing: feeling less lonely and somewhat more in control in an unpredictable and irrational world.