In small, tribal, and pre-modern societies, I care about my offspring, my family, and my tribe — roughly in that order. Over evolutionary time, such a limited extension of empathy and compassion was well-suited for the survival of the group; over millennia, that narrow moral compass served us well. Dunbar’s number, the upper cognitive limit of relationships that our brains can handle, is a strict outcome of this.
The fascinating tale of human civilization, beginning with the earliest cities, is the gradual development of cultures, moral codes, and societal institutions to overcome this evolutionary limitation. This perennial story has involved getting complete strangers to not only tolerate one another’s presence but actively cooperate with one another. For the world until yesterday, anthropologically speaking, that was mostly unthinkable, the potential for mutually beneficial trades wiped out by violent raids.
Psychologically and biologically, we haven’t quite caught up with the explosion of wealth and the massive expansion of connectivity to other humans with which the modern world has blessed us: “It is a cruel fact of human nature,” writes former National Review editor Jonah Goldberg in his Suicide of the West, “that evolution makes us biased towards our own kin in ways the rational mind cannot always accept or explain.”
In small societies or extended families, this bias usually works in our favor — or at least does very little harm. We could, with reasonable ease, observe what was happening and what problems demanded our attention. If we didn’t have personal experience of a certain event, we were never far away from somebody who did and we could simply acquire information from them directly, listening to their desires to exchange one state of affairs for another.
The 21st century, in which most people see more human faces in a day than our predecessors did over a lifetime, is entirely different. Our challenge is no longer to have biologically induced empathy direct us toward beneficial actions, but to merely get along with strangers. Once we go beyond the small-size societies in which humans evolutionarily flourished, our ability to directly observe outcomes, hear experiences, or understand what is going on rapidly decreases. Something other than the tools of empathy, kinship, and gossip is required for us to grasp the world, however well-suited they were to our evolutionary past.
Enter our solution: stats. Contrary to what most people think, condensing reality to statistical relationships or human suffering to numbers is not cruel and callous. On the contrary, statistics is how we understand, deal with, and approach our world morally. Statistics is how we productively employ our empathy in ways that actually help rather than feeling like helping — doing good rather than feeling good. In Enlightenment Now, Steven Pinker writes:
“A quantitative mindset, despite its nerdy aura, is in fact the morally enlightened one, because it treats every human life as having equal value rather than privileging the people who are closest to us or most photogenic.”
To paraphrase a resounding argument from his earlier book The Better Angels of Our Nature: narratives, life stories, and personal experiences without recourse to statistics are blind alleys.
There’s a lot of knee-jerk objection to the use of statistics in most camps: numbers are invading sectors that allegedly are too sacred to be reduced to mere numbers; statistics can’t capture everything — or even the most important — aspects of human life (in a sublunar world, could anything?); even my esteemed AIER colleague Peter Earle has been known to join the anti-numbers chorus on occasion.
In undergrad, my economic history professor would always smirk at claims that we live in a world obsessed with numbers and a manic desire to measure everything. That’s hardly unique; the Victorians, including eminent statisticians like Francis Galton, Karl Pearson, and Florence Nightingale, were similar numbers freaks, and I’m sure we can point to many other societies at different times and places that also put strong emphasis on numerical data recording.
Let me offer a defense of the unfairly accused as follows: Statistics is how we understand and make sense of any human society exceeding Dunbar’s number. While there are many ways to go wrong with statistics (one of my favorite topics to write about), it remains the only practical way to make sense of our world and — like Pinker explains — the only morally apt way to approach modern societies.
The Austrian Connection
Empirical knowledge, the end goal of all statistical inquiries, has long occupied a somewhat despised place in the Austrian tradition. Holding that economic knowledge stems solely from a deductive process from first principles (the “action axiom”), the role of empirical investigations has always taken a backseat to praxeological theorizing. This is in stark contrast to the tendency among most other economists, for whom rich databases, programming, and statistical software have greatly expanded the scope of empirical work, ushering in what we call the “Credibility Revolution.”
To place my pro-stats argument in this light, think of it as analogous to Mises’s calculation argument. Socialism, the economic system where capital goods are collectively owned and untraded — meaning that market prices for them do not exist — usually works fine between household members or close-knit communities. Since family members and friends usually both care for one another and are able to directly observe ills, scarcities, and needs, small units — as in Gerald Cohen’s famous camping analogy in Why Not Socialism? — don’t really suffer many drawbacks from operating without market prices for capital goods.
Ramping up the size of a society creates different conditions, introduces different challenges, and calls for different tools. More is different. Just as the lack of market prices makes a socialist central planner “groping about in the dark,” in Mises’s enduring phrase, people in a large-scale society refusing to use statistics are making decisions blindly, without the aid to the human condition provided by statistical knowledge.
Where Empathy Leads Us Astray, Statistics Corrects
Empathy is the instinctive feeling we developed to nurture our own tribe — that little group of mostly kinsmen that our bodies still believe we surround ourselves with. That empathy is mostly unsuited to a civilized society that counts fellow humans in billions rather than dozens and that holds their moral equality sacrosanct.
Yale psychologist Paul Bloom suggests in his Against Empathy: The Case for Rational Compassion that empathy makes us hone in on individual stories, missing the forest for the trees. Empathy is not merely neutral with regard to the ills of the world, but positively evil, prompting us to favor “the one over the many” for no other reason than proximity, familiarity, or relatability: “Empathy,” he writes, “is particularly insensitive to consequences that apply statistically rather than to specific individuals.” To appreciate utilitarian notions like the idea that it would be better if a specific and visible child dies than an anonymous larger number of children elsewhere, we must check the impulses of empathy with other capacities — particularly reason:
“[Empathy] is biased and parochial; it focuses you on certain people at the expense of others; and it is innumerate, so it distorts out moral and policy decisions in ways that cause suffering instead of relieving it.”
Reflecting on Bloom’s work, the neuroscientist and podcaster Sam Harris writes: “To be moved to action merely by empathy is to lurch blindly toward who knows what.” Other statistically minded intellectuals like Nobel laureate Angus Deaton have pointed to similar moral pitfalls that our primitive minds are tempted to make: my family over yours; my tribe over yours; my lifestyle over yours.
The key to large-scale peaceful human interaction is to properly balance this guttural, intuitive, and archaic drive. Doing what seems to be the empathically right thing often backfires, to which the entire industry of development aid is one long testimony. In that context, Deaton laments “how little we can say without [measurement] and how important it is to get it right.”
In fact, it’s a fool’s errand to reject statistics as the primary tool for understanding what the world looks like. It is virtually impossible to say anything about reality without recourse to numbers. Pinker points out that every reference to “more,” “increasing,” fewer,” or other similar words is a hidden use of numbers — and a consistent anti-statistics proponent would have to banish them from their vocabulary. What would they be left with? A bunch of unconnected, ungeneralizable stories, feelings, and individual experiences that essentially collapse to “anything goes.”
What to Do?
We feel the pain of another person’s suffering, but our brains are not equipped to understand the suffering of millions. The Swedish professor of international health Hans Rosling — most well-known from his entertaining public lectures illustrating world progress using bubble charts — often noted that although people attending his lectures enthusiastically followed along in his endeavor to update our outdated information about the world, the minute they stepped outside they reverted back to their old pessimist selves. The world is a dire place; Africans are perennially starving; our unsustainable world has a runaway population problem.
In the book published after his death, Rosling speculated that perhaps information wouldn’t do, that updating incorrect knowledge wasn’t enough; he had to change the way people felt about the world. He realized that it wasn’t the brain leading us astray — it was the heart. We simply felt that the world was bad.
Empathy’s claim to fame relies on the same mechanism. Feel — don’t think. Alleviate the other’s immediate pain — don’t step back to consider what’s the best way to help. That’s uncomfortable. The cynical explanation for why so many people refuse to embrace the world statistically is offered by Deaton, who, like me, stands perplexed before those who proudly display their innumerate values:
“The need to do something tends to trump the need to understand what needs to be done. And without data, anyone who does anything is free to claim success.”
Our emotional toolbox for handling problems comes unequipped for dealing with problems larger than those occurring in societies of Dunbar-number size. Our “natural sentiments … are innumerate,” concludes Bloom. In 21st-century societies consisting of millions and billions of strangers, stats become the invaluable navigational tool for doing so.
Yes, statistics can deceive — think Mark Twain’s apocryphal claim about lies and damned lies — but statistics can also give nuance to our worldviews. Like the British economist and Financial Times journalist Tim Harford, I am “wedded to the idea … that gathering statistical information might help us understand and improve our world.”
You should be too.