Task Failed Successfully

The Trump years are ending – at least for now. Granted, its somewhat problematic here to even use the past tense. Trump is still President, and he may run again in 2024. Which makes the underlying problem of how to think about the question of “what was/is Trumpism” all the more important. I have talked about this puzzle in the past, and this post both updates and continues many of the themes I’ve discussed this year about the Trump enigma.

The Trump presidency is in one sense a collection of ill-understood near-misses, like confrontations on the Korean peninsula in 2017-2018, that could very well have resulted in catastrophic outcomes but for various debatable reasons did not. In another sense, many of the worst-case fears have been validated or shown to be overly conservative. Both COVID-19 and the Trump administration’s willingness to sit on its hands while bodies piled up show the perils of ignoring low-probability but nonetheless high-impact risks.

However, even before Trump leaves office, arguments are already erupting over whether political scientists, pundits, and other observers exaggerated, underrated, or appropriately delineated the nature of the threat Trump posed and the damage his presidency inflicted.

Some argue that warnings were legitimate given available information and helped spur the action necessary to stop democratic backsliding. That worst-case fears did not pan out is simply another case of the so-called “paradox of preparation”. Others argue (see some of the links collected here) that the first group of people are trying to avoid accountability for failed predictions by setting up a condition in which it is impossible for them to be wrong.

There’s some truth to that, more than I’d like to admit. At a very minimum, even the “paradox of preparation” argument feels somewhat “self-serving” because it flatters the sensibilities of resisters. If it were not for me, fascism would reign supreme!

But as people argue about who is right or wrong, it is difficult not to think about another supposed failure of prediction. On the eve of the Iranian Revolution, the Central Intelligence Agency (CIA) infamously declared that the Shah could hold onto power. This intelligence assessment has been roundly mocked, but there is some reason to believe it was accurate. Very little of the explanations for why it was wrong hold up over time.

The frustrating “anti-explanation” is that the revolution was inconceivable until it actually happened. In the chaos and confusion of the struggle for power, the regime’s aura of invincibility unraveled. If viewed simply as a binary outcome, the CIA failed to forecast it. But obviously this is insufficient in accounting for the perplexing nature of the forecast/outcome mismatch. Cases like this should give pause to the various analysts arguing about whether the 2020 election shows that warnings about President Donald J. Trump’s threat to democratic institutions were justified or ill-considered.

The Trump era is a bundle of contradictions that, while not totally unprecedented, will set the stage for inconclusive ruminations for years to come. Frustratingly, the Trump presidency has been both simultaneously obvious and mysterious. Let’s start with “obvious.” There is simply very little depth to the President. Trump is a psychological void that has to be filled. He is not constrained by internal consistency or external reality; he is drawn to whatever will give him power and deference. Accordingly, there is very little evidence that Trump has a coherent goal beyond keeping himself in play – no matter the cost to the overall game itself.

This often makes speculation over Trump’s motives and goals pointless. Even if you injected the President with truth serum, he might not be able to tell you. Similarly, Trump has a set of well-honed instincts and tricks rather than any particularly grand strategy. This is not to diminish the outsized effectiveness of many of these tricks and the harm that the President has inflicted, but to point out that they are not particularly sophisticated. Adding too much subtlety and sophistication is often a means of evading reckoning with the horrific nature of the Trump era.

Trump has repeatedly behaved in an abnormal and destructive fashion, and observers’ unwillingness to deal with the gravity of these transgressions leads them to play a rather silly game. Every time the President signals his intent to do something panic-inducing, critics must then explain what the President’s goal is and predict his chances of success. Both of these demands tend to shift the focus away from the illegitimacy of the behavior and its corrosive effect on the US political system.

It also doesn’t diminish the behaviors to point out that the normalization cycle would be ridiculous in any other context. If, in the middle of a night out at the bar, a bar patron attacks you should the first order of concern be explaining his motives and estimating the probability he will land a hit on you? Anyone that believes that writing a 3,000-word thinkpiece is the superior option has never been attacked in a bar! For sure your response does depend at least minimally on the legitimacy of your guesses about motives and probability of success, but too much consideration inhibits any kind of response.

And on this score the dismal record of the Trump presidency more than justifies fears of critics. Look at this year alone.

By June 2020, over 100,000 Americans were dead from COVID and the President, facing riots and protests, was threatening to invoke the Insurrection Act. Analysts responded by once again shifting the goalposts, debating how responsible the chief executive was for failed COVID response and blaming street brawlers alone for the breakdown of order. As Daniel Nexon argued in mid-summer, much of this revealed how the United States depended on “informal norms and formal – but fundamentally toothless guidelines” to prevent abuse of executive power.

Trump’s personal character flaws combined with Republican complicity set the stage for abuses and manipulations of both routine and emergency functions as well as America’s catastrophic response to the COVID-19 pandemic. Enumerating all of these abuses and manipulations is important but also can be futile if analysts respond by merely revising their baseline of how much deviancy to accept.

People predict that Trump will do X, Trump does X, and then the response is to ask the people who correctly predicted X to explain why X is different from other presidents, how likely X is to succeed, what Trump’s goal in doing X is, and if X is really that bad after all and worth all of the fuss and commotion. Once or twice this is understandable. But after a while it gets tiresome and patience runs out.

However, there is also a deep mystery to the Trump presidency. This mystery becomes apparent whenever its major scandals are even trivially examined. RussiaGate and most of the other scandals of the administration were not hoaxes. However, ultimate understanding of them will likely remain elusive. Why? Consider the following description of Arnold “Big Bankroll” Rothstein, one of the key figures behind the fixing of the 1919 World Series:

Yet most of the book’s new claims and speculations about the fixing of the World Series are not well documented, unlike the rest of the biography, which includes over 60 pages of endnotes and a bibliography. This is not to criticize Pietrusza for dereliction of duty. The point, rather, is that some of the truth is beyond us, lost forever in the shadows of the past. We will never know, for example, to what extent Abe Attell, a colorful, duplicitous small-time gambler and former prizefighter known as “the Little Champ,” worked on his own to fix the Series and to what extent he was working for Rothstein. The fact that Rothstein “spent a lot of time and money shielding Attell from prosecution” does not prove that he was buying Attell’s silence about Rothstein’s own involvement in the fix, whatever it was. There could have been other reasons. Rothstein’s intentions were almost always self-interested but were rarely transparent.

The broad parameters of the administration’s scandals reveal them to be far more Thomas Pynchon than Robert Ludlum thriller, a collection of strange semi-related but never quite intersecting schemes that never cohere into anything resembling a master plan. Due to their ambiguous, sub-rosa nature and their population by Rothstein-like and Attell-like figures, investigating them and summarizing them is inherently difficult if not impossible. After all, if we still cannot tell the full story of something like the 1919 World Series fixing even a century later, what hope do we have of bringing to light far messier, bizarre, and perverse political scandals?

Consider the Michael Flynn saga as representative example, which begins with a National Security Advisor freelancing for various public and private entities and ends with oaths of allegiance to the Qanon cult. But the opaqueness and mystery goes beyond merely getting to the bottom of murky scandals. Describing Rothstein as “self-interested but rarely transparent” is actually the most important part of the paragraph. How could someone be simultaneously self-interested and rarely transparent?

Look at what happens whenever Trump tweets. Donald J. Trump vomits up an angry and vaguely threatening comment. What does he mean? Is he serious? Does it show he is weak or strong? Is he mentally unwell, or has he always been like this? Is it even Trump that is tweeting, or is someone impersonating him? All questions that could be asked – and have been asked – at any arbitrary time from 2015 to 2020. They are not easy to answer.

As noted earlier, even Donald J. Trump doesn’t know what explains the baffling behavior of Donald J. Trump. The Trumpian tendency is to incessantly, irregularly, illogically, inconsistently, and above all else incoherently bluster. Due to Trump’s status, personality, and lack of transparency, the bluster is not easily dismissed even if obsessing over it can often be a waste of time.

The strangest aspect about the Trump presidency therefore precisely lies in its combination of publicity and opacity. Trump lacks even minimal self-restraint, rarely if ever leaving a crude opinion unvocalized (or untweeted). His motives are almost comically faithless and malign. Yet there is always something uncertain or mysterious about the circumstances in which Trumpian phenomena plays out. Despite the way in which, at surface level, no such mystery ought to exist!

Trump has been in the public eye for decades. He is not a eloquent Shakespeare villain, he is a vulgar and cruel would-be dictator. And yet, if it was so easy to explain, predict, and control his behavior, American politics from 2015-2020 would look much, much different. Trump’s tweeting, by nature, forces the analyst to theorize about observed inputs and outputs without understanding about his inner workings or the underlying reliability of any singular interpretation of his behavior.

Like many similar historical figures, Trump’s intentions are clearly self-interested in theory but never so transparent as to make them trivial to interpret in practice. Thus, inner understanding of how Trump behaves is remarkably elusive in spite of his crude and instinctual behavior. And this precise lack of inner understanding seems to generate paralysis by analysis.

There are many different plausible “theories” that can explain the sparse and ambiguous “data,” each of which is seemingly as good as any other. These micro-level competing explanations dovetail with numerous incompatible but nonetheless seemingly equally plausible macro-scale attempts to explain “how we got Trump” among the commetariat. As will be observed later, these macro-scale theories often are both overwhelmingly numerous and analytically underwhelming.

Understanding basic aspects of Trump and Trumpism have always meant dealing with problems of observational equivalence, the way in which all of our theories of how an ill-understood system work depend on a sparse and noisy collection of observable behaviors. Is Trump suffering from the effects of COVID and/or COVID treatments? Perhaps. But all of the external evidence to support this inference – his abnormal, erratic behavior and the secrecy and obfuscation surrounding him – are hard to distinguish from every other day of his administration.

This kind of basic observational problem occurs virtually every time we try to understand, explain, or predict habitually unruly Trump administration behavior. And this is simply the issue with explaining observable decision-making, not Trumpism in and of itself. Every attempt to explain Trumpism suffers from another similar contradiction. No one ever says “Trump is how we got Trump” because of the seeming contradiction between our knowledge of Trump as an empty shell and the inability of an empty shell to cause Trumpism. But yet, here we are. Something is happening here, Mr. Jones.

Hence the search for some kind of structural explanation for Trump and representative base of support for Trump. Neither effort has yielded any conclusive answers and – as is often the case in social inquiry – often generates more difficult questions than clear answers. So even in late November 2020 analysts still wring their hands about who the mysterious Trump voters are and why polling failed to register them. We find our intrepid scribblers falling back on the oldest of social science cliches – its the social alienation, stupid!

Trump himself, by abdicating any kind of responsibility for his speech and behavior, makes this basic explanatory problem worse. By making himself a (distorted) mirror, he asks us to see our own reflections in his arrogance, incompetence, and cruelty. But we ought to recognize this as a deflection, especially because every attempt we have made to interpret what exactly this mirror reflects back at us has come up short. The temptation to conclude that its all just a fluke is understandable, if just as pernicious.

It is probably incorrect to say that Trumpism is merely a fluke, but most explanations tend to unduly neglect contingency. Given the recurring importance of contingent events in the Trump saga – such as FBI director James Comey’s October 28 intervention in the 2016 election – any number of things can potentially tip the scales. Even forecasts about electoral results have some small yet nontrivial impact on electoral results. We lack good intuition for how to avoid neglecting the role of random processes in the Trump drama, though perhaps this is part of the challenges we face in accounting for any kind of complex historical event.

Trumpism itself has far more contingency to it than structural explanations like to acknowledge. A charismatic would-be Bonapartist figure will, by nature, hoover up all manner of malcontents that otherwise do not have much in common. Such figures tend to have free reign when institutional constraints are weak and cannot suppress the impact of individual personalities. But if this is the case, then it doesn’t make sense to talk about the next threat as a “more competent Trump.” Why not a thousand little Trumplets below the executive level? Why even a charismatic populist? The same set of broad institutional weaknesses – provided we can agree on what they are – can be attacked a number of different ways.

All of these considerations are a prelude to the most critical question of all – if Trump ultimately failed to secure re-election by hook and crook, does the failure validate or indict American democracy? And, whatever our answers to this question, how can we know if we are right or wrong? Will any answers we have today hold up when observers with more critical distance examine our period?

Unsurprisingly the reader will find mystery and contingency here as well. Thomas Pepinsky observed right away the problem with answering the question – explaining how Trump was elected President and failed to secure re-election could mean using the same explanation (weak parties and strong partisanship) to account for both outcomes. Moreover, would we be as confident about the outcome if the “margin of contestation” allowed by the depth and breadth of Joe Biden’s electoral victory were much smaller?

What isn’t up for dispute is that Trump tried to discredit the results of an election he lost and did so largely unopposed by his party. So in trying to explain how warnings and predictions of disaster were robust in spite of the dog refusing to bark, analysts have often relied on a similar set of analogies and metaphors. The plane crash landed but was not totally destroyed, but the plane still crashed. Concluding that the plane is safe or there was no risk is a category error. Others made references to famous if often misunderstood studies of World War II aircraft survival or the infamous toleration of defective O-Rings prior to the Space Shuttle Challenger disaster.

So what of it? These metaphors and analogies correctly ask us to be wary of survivorship bias and post-hoc recalibration of how much barbarism we ought to accept. Just because something bad fails to happen does not mean it will fail again, and just because the worst-case scenario did not occur does not mean something intolerable wasn’t tolerated. But there’s also something wrong with the form of these claims. I am sympathetic to the normative bent of these metaphors and analogies but think that their content is ill-served by their form. If only it were so easy!

Failure analysis involving transportation systems or large physical systems in general is a well-structured and orderly process. Experts with universally acknowledged credibility assemble to pour over the ruined system and trace the mechanical and organizational steps that led to its breakdown, issuing a conclusive report that diagnoses the causes of the failure. Certainly this is not always the case, but it works well enough to give people in most Western countries sufficient confidence in the safety and reliability of their airplanes and other similar complex systems.

Confidence that verges on irrational faith because the operation of the system is disembedded from the layman’s personal experience. Nothing of the sort obtains for diagnosis of democratic failure, which is both inescapably subjective and concerns a system that is far less objectively controllable and predictable. We lack “ground truth” understanding of how the system is supposed to work to use as a basis for comparison.

Instead the more unsettling metaphor/analogy should be a nonsensical error message like “task failed successfully.” The humor embedded in the message is a function of the gap between expert and layman understandings of its semantics. For the expert, it means that the computer task failed but in a way that nonetheless suggests the system is behaving normally. For the non-expert, it is gibberish.

Because the nature and operation of the system is often opaque to the user, it is questionable how useful the error message really is. For all the user knows, it could be misleading in the way most interface abstractions since the invention of the Graphical User Interface (GUI) often have been. As Neal Stephenson lamented about the concept of “documents” on computers, interfaces can often be terribly misleading.

Consider only one word: “document.” When we document something in the real world, we make fixed, permanent, immutable records of it. But computer documents are volatile, ephemeral constellations of data. Sometimes (as when you’ve just opened or saved them) the document as portrayed in the window is identical to what is stored, under the same name, in a file on the disk, but other times (as when you have made changes without saving them) it is completely different. In any case, every time you hit “Save” you annihilate the previous version of the “document” and replace it with whatever happens to be in the window at the moment. So even the word “save” is being used in a sense that is grotesquely misleading—“destroy one version, save another” would be more accurate.

Anyone who uses a word processor for very long inevitably has the experience of putting hours of work into a long document and then losing it because the computer crashes or the power goes out. Until the moment that it disappears from the screen, the document seems every bit as solid and real as if it had been typed out in ink on paper. But in the next moment, without warning, it is completely and irretrievably gone, as if it had never existed. The user is left with a feeling of disorientation (to say nothing of annoyance) stemming from a kind of metaphor shear–you realize that you’ve been living and thinking inside of a metaphor that is essentially bogus.

This underlying feeling of disorientation is a better fit for the nature of the Trump presidency and its “successful failure” than all of the plane/shuttle/etc crash metaphors/analogies I have seen so far. Stephenson further observes that users are forced to repeatedly learn and re-learn different misleading interface abstractions. Which takes us to the further realization that what might be a successful failure for a system vendor and/or operator is not necessarily good for the user – and vice versa.

Take the matter of system bugs. Users can adapt very readily to bugs in the system, developing folk theories about why they occur and molding their behavior to the needs of the system instead of the system serving user behavior. However, users can often exploit bugs to their benefit even if the original designer of the system seeks to patch or otherwise do away with them. Bugs can become features. The diverging social interpretations of technical facts is fascinating in the abstract but often personally frustrating.

The contradictory nature of “task failed successfully” has to be semantically resolved in a way that is far from neutral or impartial. The worst that can happen if one loses work on a Microsoft Word document is, well, losing valuable data. We can tolerate these kinds of failures, successful or otherwise, in a way that we do not and should not tolerate political failures. Whether or not the system properly handles the failure by its own implicit or explicit specifications says little about the objective and subjective impact of the failure to the user.

All of this is suggestive of a basic asymmetry in attributions of system failure, and this asymmetry should enter into the way we think about the “successful failure” of the Trump presidency. Trump indeed “failed successfully” but successfully for whom? Was it a success for the American system, showing that even Trump’s despotic impulses could be held in check? Or was it a success for Trump and would-be Trumpists more broadly, corrupting the system even further and demonstrating a critical and exploitable vulnerability? As I have implicitly and explicitly suggested, the issue is also whether or not the task failed or succeeded.

After all, while we know that Trump would very much like to have won the election his underlying desire to see himself as a winner regardless of the outcome makes him extremely flexible about both goals and the means by which they are achieved. Win or lose, he gains as long as he is able to inflict long-term damage to our system for his personal benefit. For those concerned with the future of American liberty, it has always been a question of bad vs worse outcomes. This underlies the explanatory burden we face in thinking about whether the supposed failure validates or indicts the system.

It is easy to handwave away the question by pleading for more time to answer it. Sufficient distance perhaps can give us a more objective perspective, but I strongly doubt this will be the case. Consider again the earlier discussion of the 1919 World Series. Is there a solution to its riddle a century later? Not really. And the passing of time can also hinder explanation by depriving observers of the necessary tacit perspective and experience required to appreciate what is at stake. This is undoubtedly the most important part of the Trump explanatory conundrum.

Let us return to the question raised at the beginning of the post. Were observers right to fear Trump or did they exaggerate the risk? Who is right? But, more importantly, how much does accuracy matter when juxtaposed with questions of responsibility? Ambiguity about the nature of observation does not remove the responsibility of the observer if the observer has some stake in the outcome. And most Americans cannot be impartial and distant observers in this particular case.

This post is about the puzzle of Trump and Trumpism but it is also ultimately about the conflict between ambiguity and responsibility. American observers are citizens of a free country. They have a moral stake in the outcome and are not disinterested spectators looking on from a comfortable distance. Their responsibility – our responsibility given that me and most of my readers are American – is complicated by the actual puzzle of understanding Trump and Trumpism writ large.

This complication makes the responsibility more burdensome. Any kind of inference or action risks the wrath of some notional unsympathetic future audience that will pitilessly judge behavior with the benefit of hindsight. Anxiety about judgment motivates the intensity of debates about who was right or wrong about Trump, because observers are often desperate to pre-emptively decide the conditions under which they will be judged by this future, abstract jury of their peers.

But no such jury really exists. It is a figment of our imaginations. Or, at the very minimum, no such earthly jury exists, because it is a secularization of a theological concept. The lack of an ultimate human authority that will bless or damn our choices makes us ultimately cursed by our freedom. And we cannot escape this curse, no matter how hard we try.

Facile historical comparisons have always been an easy way for analysts to escape it, but they will find no refuge in such cheap intellectual tricks. We are not Germans in the 1930s or people that populate any other tired historical comparison. There is some justification to the mockery of what one might call “Hannah Arendt cosplay”; far too much of the Trump years consisted of people yearning to be heroes in a black-and-white moral drama.

Still, there is something generally suspect about people that spent the Trump era largely mocking even far-fetched concerns by people with little power and underplaying or even apologizing for dangerous behavior by the chief executive of the most powerful nation on Earth. What, if anything, would have convinced them that Trump was dangerous? Too often, the answer was “nothing.”

Either way, we live and will die in a particular historical and material time and place of our own. Nothing that we are capable of doing will transmit the circumstances of our lives to future generations without severe information loss. We have to ultimately justify our attributions and choices on our own terms. Ambiguity is a burden that we must face in meeting our responsibilities, but it does not discharge them.

I have strong opinions about what the morally correct attributions and choices of the Trump era were, but I’m not confident history will validate them. Nonetheless, I don’t care. I have no way to predict how history will be written, and who writes history often surprises us. And most importantly its besides the point.

Trump was and still is a malignant cancer on the American body politic, and time will tell about the patient’s ultimate prognosis. I am relieved to see, one step at a time, his power diminish. I side with the people who – imperfectly and clumsily – worked to excise him over those who pooh-poohed said efforts. I do not care about history’s verdict, I only care about what I know today, when it actually mattered.

I come out of the last four years both relieved and anxious. I am relieved that, at least for now (we shall see in 2024), we are spared four more years of Trump and all of the further damage he could inflict on us. I am anxious because the experience has exposed the weaknesses of so many critical institutions and also underlined the importance of contingency. There is much to be thankful for today, but we will still roll the dice again tomorrow.