When we make important decisions, we are often not as rational or objective as we’d like to believe. The base rate fallacy is the tendency to misjudge the probability of a situation by not accounting for all relevant information. This cognitive bias affects everything from first impressions to voting preferences to broad market behavior.
Listen to the Podcast
Imagine walking onto a subway in New York and taking your seat. Glancing around for several moments and taking in your surroundings, you notice a woman seated across from you wearing a scarf and nice slacks. She is thumbing through a copy of the New York Times, reading diligently with glasses, a novel spilling from her purse. Which of the following choices is most likely about this woman?
She has a PhD.
She does not have a college degree.
This narrative is adopted from psychologist Daniel Kahneman in Thinking, Fast and Slow. What is your first impression?
Impressions Matter
Many investors claim a “gut feel” for a sound investment, an almost magical or primal know-how to gauge success. Wall Street and Main Street alike teem with individuals who claim an almost superhuman knack of foresight, the ability to predict the future performance of businesses, industries, and markets. Leadership, charisma, and even the gift of gab are proxies for future success.
Elizabeth Holmes was (and probably still is) by all accounts an extremely skilled saleswoman. In Bad Blood, the story of Holmes and the rise and fall of the biotech firm Theranos, investigative reporter John Carreyrou quotes a former Theranos employee as saying Holmes could sell ice to an Eskimo. In Carreyrou’s telling, Holmes wanted badly to be a billionaire tech visionary. She even went as far as to literally fashion herself after Steve Jobs, wearing black turtlenecks in all public appearances. Holmes and her business partner and lover Ramesh “Sunny” Balwani deceived many powerful business leaders, politicians, and even military generals. She convinced them to invest in, support, and later commercialize a product that essentially never existed.
Holmes, adept at portraying a simple story of profound positive change, built a world of falsehoods. She kept investors and supporters from reasoning with the many visible cracks forming in the business, even when the signals were flashing bright and red. Kahneman wrote in Thinking, Fast and Slow:
“He had an impression, but some of his impressions are illusions.”
Behavioral Economics and the Two Systems of the Mind
Kahneman and his research partner Amos Tversky helped to pioneer behavioral economics. This field of study that examines how humans are prone to errors in judgment from biases and heuristics. As opposed to neoclassical economics, Kahneman, Tversky and others, notably Richard Thaler, argue that we do not act entirely rationally or objectively in most cases, even though we believe we do. Two systems of the mind drive the way we think and the decisions we make:
System 1 is snappy and emotional, a bit of a knee-jerk reactionary prone to jumping to conclusions in the absence of evidence.
System 2 is much slower, more logical, and far more adept at reason.
The problem is that System 1 likes to do a lot of the driving.
Those who find a baseball player handsome and athletic are also more likely to rate him better at throwing the ball.
The Power of Representativeness
Representativeness, or appearances or vibes, is rather enticing to System 1 and can overwhelm our ability to reason our way through a good decision. If given a tour of a business that is sleek and modern, it’s tempting to equate these impressions with success. Impressions matter greatly to System 1. Concerns such as debt or poor revenue expectations might not appear immediately obvious.
Thorough assessments would involve the lazy System 2 and a study of tedious financial data. And these trends aren’t limited to business. As Kahneman notes, those who find a baseball player handsome and athletic are also more likely to rate him better at throwing the ball.
The Base Rate Fallacy: The Woman on the Subway
Again, impressions matter. The scarf-shrouded woman on the subway reading the New York Times met the representative vibe of an educated rider. It seemed highly plausible, compared to the alternative of her being non-college-educated, that she probably held a PhD. But if our analytical System 2 kicked in, we might reason that PhDs are likely volumetrically minor in a population of New York subway riders. In fact, only about 2% of Americans hold PhDs, while nearly 52% of Americans hold no college degree. In New York specifically, those figures are 1.5% and 58%, respectively.
Despite this woman’s appearance and behavior, highly alluring to the reactionary System 1, the probability of this woman holding a PhD on a New York City subway reflects the population as a whole, hovering close to 0%. The chances that she holds no college degree, however, is above 50%, better than the odds of a coin toss. If you were to wager money on this bet, you’d do best to bet on her having no college degree.
When assessing the educational attainment of the woman on the subway, the mind focused on specific and incomplete information—the clothing and the reading material (System 1)—while ignoring the probability, or the base rate information (System 2). This cognitive bias is known as the base rate fallacy, or the tendency to misjudge probabilities by not accounting for all relevant data.
Anchor to the Base Rate
In the absence of compelling information we should anchor to the base rate. In the subway example, the base rate is the education demographics of the population as a whole, or New York City more specifically. A reason to deviate from the base rate might be a change of venue. For instance, moving the setting from a subway to an academic conference, where PhDs are plentiful, would justify deviation from the base rate.
The Halo Effect: Flashy Versus the Pragmatic
Two associates approach you with two different business propositions. Associate A wants to open a gym. The business plan is to compete with high-end brands, with decorative and attractive décor and coffee and beers on tap. The kind of gym you enjoy. Associate B wants to open a laundromat. The design is simple and decidedly unattractive, everything needed to launder clothes and nothing more. Which sounds like the best investment?
Regarding base rates, you find that 81% of fitness studios fail or close within the first year. You are even more surprised to learn that 95% of laundromats are still operating after five years. With the base rate information considered, now which investment will you choose?
The idea that a flashy and trendy gym could be a complete failure and a dingy laundromat a success violates the simple and compelling narratives we create. Consider the statement “Hitler loved dogs and little children.” As Kahneman notes, that statement will probably bulge your eyes every time you read it. The idea that someone so evil could also exhibit kindness goes against the grain of simple narrative and representativeness.
The halo effect is a cognitive bias in which our impressions of a person (or brand, business, etc.) sway our evaluations of other specific traits. Most carry the impression that Hitler was an evil man, so it’s hard to imagine him rubbing puppy bellies and giving piggyback rides. For the same reason, we are likely to favor the policies of a politician if we find them entertaining.
Base Rates and Stock Market Performance
In this post I noted that broad-based stock market returns have always been positive over any twenty-year period. Put another way, someone holding an index fund approximating the broad market performance for at least twenty years has never lost money on that investment. And the longer you hold it, the less downside risk.
We can view this past performance of the stock market as the base rate. While it’s tempting to view short-term volatility or pessimistic punditry as a harbinger of poor future performance, highly alluring to System 1, we would be participating in base rate neglect. But if we engage System 2 and consult the data, we find no precedent for losses over long enough periods. Not to say that the pessimistic case is impossible, but nearly far less probable than initial impressions might suggest.
Alternatively, the consideration of an individual stock, company, or even industry investment is far less optimistic. Barely 20% of individual stocks survive and outperform the market over twenty years. The likelihood of picking the next Apple or Google in a sea of letdowns is highly improbable. Put another way, when picking individual stocks, the probability of picking a stock that underperforms the market over twenty years is about 80%. Ouch.
What You See Is All There Is
When in doubt, consider the mantra repeated throughout Thinking, Fast and Slow: What you see is all there is. In other words, we often make judgments not rationally or empirically, but according to the often-incomplete information we have available. This information is cut with our backgrounds, beliefs, and values.
What to do?
If approached with a business proposition, investment, voting suggestion, or even medical intervention, take great lengths to quantify appropriate base rates. Key questions you might ask:
What are the typical revenue expectations?
What is the failure rate of this type of business or intervention?
Do the quantitative indicators pair with the narrative (useful in a voting year)?
What are the typical outcomes of an intervention (A success rate of 10% is appropriate probably only for those desperate and hopeless)?
Given the high recurrence of failures, it’s easy to understand the common advice to limit one-off or angel investments to no greater than 5% of net worth. And even that might be a stretch. The probability of loss when investing in single entities is high. Venture capitalists are successful because of their ability to spread risk among many investments. One or two extraordinary successes greatly outweigh the far more numerous and typical losses. When in doubt of the quality of the evidence, anchor assessments of probability close to the base rates.
When meeting new people, listening to the news, or even assessing a new boss, take some additional time to assess initial reactions. Impressions can often be illusions.
Affiliate links are used on this page. You will incur no extra charges if you purchase a linked product, but we will receive a tiny-baby portion of the sale. Those minimal proceeds help us keep the digital lights on around here. We wouldn’t link to a product we wouldn’t buy ourselves.