The Wisdom of the Butterflies: Critical Theory, Again
Where is the affirmative evidence on "gender-affirming" interventions?
The Butterfly Effect is not what we popularly understand it to be. It’s not about how small perturbations in complex systems can induce spectacularly large effects but rather about how we might never be able to know enough about a given state of a complex system, even a deterministic system, to be able to avoid spectacularly large errors in our predictions of future states of the system.
As popularly understood, the “Butterfly Effect” seems to run something to the effect of: arbitrarily small perturbations in a complex, dynamic system can induce spectacularly large effects. So, if a butterfly in Mexico flaps its wings, it may set the system off on some trajectory that leads to a typhoon in the Philippine Sea.
The real butterfly effect comes to us from Edward Lorenz, “The Predictability of Flow which Possesses Many Scales of Motion,” Tellus 21 (1969), pp. 298-307. Lorenz actually characterizes it as more of a “Seagull effect,” but my very coarse, formative understanding is that it is one thing to know everything about a deterministic system—that is, to know everything about how a particular system evolves over time as well as to know everything about the state of a complex system at a particular point in time. Knowing all that enables one to figure out the state of system at any point in the future. And, so long as we restrict our attention to reversible processes, we could also work backwards and figure out the state of a system at any time in the past. But, we may not have the luxury of knowing everything about the state of a complex system at any given time, and it may be case that we can never acquire information that is fine enough to enable ourselves to fully ascertain a given state. We may yet miss arbitrarily small aspects of the system’s trajectory – the flapping of a sea gull’s wings, say – that could yet inform our prediction of that same trajectory. There may be a finite limit to what we can know about a system. And that can matter. Or not.
Ken Arrow – Nobel Prize Weatherman
In my mind Ken Arrow is the Richard Feynman of economics: a smart guy who knew he was smart, didn’t require anyone to tell him he was smart, and had a lot of interesting things to say on a broad range of policy-relevant topics. On top of all of that, he was a gracious fellow. I’d suggest that he was an exemplar of the “great oak tree” in the Taoist parable of the “great oak tree”. See my earlier essay on “The Revenge of the Nerds: Critical Theory.”
I had the privilege of listening to Ken Arrow speak, in passing at least, about the hazards of forecasting. His comments were informed by his own experience as a weatherman during the Second World War. (He went on to become a Nobel laureate in economics in 1972—hence, “Nobel Prize Weatherman.”) Weather forecasts, he suggested, are only good as far as three days out or so. That was true in the 1940’s. It remains true now, never mind the fact that we sit on top of so much more computing capacity and so much more capacity to measure weather phenomena using satellite data. Never mind those butterflies and sea gulls.
Ken Arrow’s larger point was that economic forecasting is not obviously easier than weather forecasting, and it is not obvious that we are situated to fine-tune macroeconomic performance. I cannot remember if he said anything specifically about the Federal Reserve and central banking, but one can wonder a few things about the Federal Reserve. The Fed may maintain the conceit that it does maintain a lot of control over the economy; it does maintain the right model of how the world works; it does have access to the right data to monitor what is going on in that world; it does have its hand on some number of control variables and is thus able to nudge the economy in one direction or another. It may believe in its own conceit. But, it would be nice to see some compelling evidence that the Fed really does exercise some affirmative influence over economic performance and is not merely chasing its tail or, worse, is distorting economic performance. But, sea gulls and butterflies: even if the Fed does view the world through the “right” model, even if the Fed does have its hand on control variables—two big “If’s”—are its forecasts really any good?
Let me add the “McNamara Fallacy” to this business of forecasting: The McNamara Fallacy overlaps with “Management by the Numbers.” Here the idea is that data are important, but part of the fallacy is that, one may or may not have access to useful data, but one does become fixated on certain, affirmatively bad metrics. Robert McNamara is infamous, as Secretary of Defense, for getting fixated on “body counts” in the Vietnam War. The American military leadership in Vietnam then became fixated on body counts: increase the enemy’s fatalities; decrease your side’s fatalities; ignore larger, strategic considerations; victory will yet obtain. Add to that private incentives to fudge the numbers. Were the reported data reliable?
In The Best and the Brightest (1972), David Halberstam identifies McNamara as one of the original “Best and Brightest” technocratic elites of the Kennedy and Johnson administrations in the 1960’s.
The British and the French laid siege to Sevastopol on the Crimea in 1854. Leo Tolstoy, meanwhile, had been serving as a junior officer in the Russian Army and had the privilege of observing the siege from the inside. The siege dragged on far into 1855. As it dragged out, Tolstoy sent out dispatches for publication in the Russian papers. An uncle or someone had suggested that his dispatches were really very good; the young Tolstoy should think about keeping up with his avocation as a writer.
It is hard not to guess that Tolstoy’s experience in the Crimean War informed his detailed and entertaining sketches of war councils and military engagements in War and Peace. The dominant theme in the many sketches involves the unpredictability of complex, chaotic processes—and the delusional conceit of the leadership that it could scientifically and deterministically manage just such processes. Tolstoy makes a big point of ridiculing the Self-anointed Best-and-brightest and ridiculing their scientistic certainty. These things are much the subject of Gary Saul Morson, “The Greatest of All Novels,” The New Criterion 37 (2019), https://newcriterion.com/issues/2019/3/the-greatest-of-all-novels.
I made contact with Morson’s essay as well as with one of Tolstoy’s vignettes in my own essay, “Our ‘Democratic Imperialism’ versus Their Old School Russian Imperialism.” I make contact again with such material with reference to recent, still-unfolding process in Ukraine … and to the failure of the Self-anointed Best-and-brightest to responsibly manage the coronavirus phenomenon, turning it into something more deadly and more destructive than it had to be. And right here I make contact with such ideas with reference to the “trans” phenomenon.
It may have been difficult not long ago to imagine that we would now find ourselves debating the merits of subjecting children entering adolescence to hormone treatments. The grand theory is that people are endowed at conception with a sexual orientation distributed somewhere on “the binary” between male and female. Some people may be endowed with XY chromosomes, but they may identify as people who would otherwise have been endowed with XX chromosomes. That is, some “biological males” may perceive that they are more female than male. Similarly, some biological females may perceive that they are more male than female.
This idea of people being distributed on a continuum between male and female might not have made for more than a speculative matter except that more than a few public authorities have urged young people to contemplate taking on irreversible, and potentially harmful, therapies so that their physical manifestations might more closely align with their perceptions of their places on the continuum. Young people are encouraged to suppress their bodies’ natural hormone processes so that they might channel their adolescent bodies into something that aligns more closely with their sexual gender identities. Hormone treatments ultimately set up young people to take on aggressive “gender affirming surgeries” once they come of age.
A contrary approach would be to let nature take its course. People will become whatever they will become. What’s wrong with that? (Again, see the Taoist parable of the “great oak tree.”) The proponents of aggressive intervention would argue that allowing nature to takes its course would leave a lot of young people physically stuck in the wrong body. These people need to be “affirmed,” not denied.
Where to start? The only concern I will advance here pertains to the scientism underlying the aggressive “gender affirming” regime. Like Tolstoy’s generals or Lorenz’s weathermen, are the proponents of aggressive interventions really situated to know what they are doing? What contingencies, both unforeseen and unfortunate, could yet obtain? Do we have data on this kind of thing? Are we situated to intervene effectively, or are we playing with dynamite? Ultimately, do these interventions tend to leave people better off or worse off?
This last question might be more susceptible to analysis than the question of how to intervene in complex biological processes. There is some understanding out there that people who “transition” do not end up happier than people who desire to transition but have yet to transition. Admittedly, waking up in the morning, every morning, and thinking that one is in the wrong body makes for a rough place to start the day, every day. We are not talking about the happiest people here. But, one might wonder whether the underlying comparison between people who desire to transition and those who have actually pulled the trigger and made the transition has controlled for self-selection. Specifically, comparing people who have transitioned to people who have not transitioned may not amount to an apples-to-apples comparison in that the people who have transitioned have demonstrated a willingness to go through irreversible processes. Such people may have perceived more desire, more demand to go through such processes. They may be the people most likely to perceive harm from not transitioning. Such people may or may not end up being very happy after transitioning, but it could be just such people who could benefit to some degree from making the transition. They might be less unhappy. Or they might feel liberated and ecstatic. They’d be better off. That would be the hypothesis.
To examine this hypothesis, it would be nice to get a sample of people very eager to transition and to randomly assign some number of such people to “gender affirming” surgery. How do the treated perform relative to the un-treated going forward?
Where could one secure such data? Might there be some quasi-natural experiment to take advantage of? Might it be the case, for example, that people in some jurisdictions have easier access (financially or otherwise) to gender-affirming surgeries than in other jurisdictions? It would be nice to know the results of such quasi-experiments rather than merely asserting that gender affirming surgeries are good things.
The analysis would look much like the analysis applied to “school choice.” Evidence that students in charter schools perform better than students in public schools is not obviously evidence that any one randomly selected student would tend to perform better in a charter school. The reason, again, would pertain to self-selection. Conceivably students going to charter schools may come from families that are more serious about education in the first place. They may be growing up in environments that are more conducive to success. True or not, researchers have accommodated the prospect of important self-selection in the data-generating process by taking advantage of certain quasi-experiments. Some charter schools have found themselves over-subscribed. They have then rationed access by selecting students by random lotteries. So, researchers can examine the performance of students who self-selected into the lotteries. Some portion of them will have subsequently attended the charter schools. Others will have defaulted to the public schools. How did these two pools of students perform?
Ideally, researchers would not have to exploit the charter school lottery to resolve the self-selection problem. Instead, they would take a pool of students and randomly assign them to charter schools and public schools. But, researchers can’t go around infringing the civil rights of others. But, the data from charter school lotteries did provide a compelling alternative. Are there no such data on gender affirming interventions—data that control for the potential self-selection problem?