If you prefer to listen rather than read, this blog is available as a podcast here. Or if you want to listen to just this post:

Or download the MP3


I recently read an article titled How I Almost Destroyed a £50 million War Plane and The Normalisation of Deviance. The article opens with the story of a multinational military exercise, which the author, Tim Davies, participated in back in the mid 2000s. As part of this exercise, everyone was assigned to a specific jet, and if your jet was having problems you couldn’t switch. This is, unfortunately, exactly what happened to the Davies.

Our jet had a problem with the undercarriage or landing gear – it wouldn’t lock up under normal flight conditions; the wheels couldn’t be stowed away.

The engineers had found significant and unfixable wear to the mechanical uplock. It would only lock up under 0g and this would mean that I would have to bunt the aircraft, nose down, towards the ground whilst selecting the gear ‘up’.

(For those unfamiliar with the term, a bunt is half an outside loop, meaning you start out straight and level, then dive in a curve, eventually ending upside down.)

With this “solution” in hand and not wanting to be grounded for the remainder of the exercise Davies (and his Weapons Systems Officer) decided to try it out. And lo and behold, it worked. They’d have to do it every time they wanted to fly, right after takeoff, at the point when they had the maximum amount of fuel (five tons), but they had figured out a way to keep flying. So that’s what they decided to do. And this, if you haven’t guessed, was a deviation from the normal guidelines for safely flying a £50 million war plane.

Despite comments from other pilots, and the concerns of their Programme Director, who Davies managed to avoid, they continued performing this maneuver, and everything went great until the last day of the exercise when the weather was worse than expected. It was then, while performing the maneuver, that they entered a cloud, and when they finally emerged from that cloud Davies realized he was in a very bad place.

We were low on energy and the nose was rising too slowly to recover the aircraft before we would hit the ground.

The Ground Proximity Warning System sounded.

‘WOOP, WOOP – PULL UP, PULL UP!’

‘7, 6, 5 – that’s 400 ft Tim!’, called my WSO.

The jet was shuddering against my demands, it just didn’t have the performance to pull out of the dive.

The cockpit was silent. To make things worse, due to our high rate of descent, we were well outside of any ejection option.

I quickly selected full flap and slats to increase the lift over the wing.

The sudden increase in lift meant that the nose started to pitch faster towards the horizon.

A bad picture was starting to look better.

Eventually I levelled the jet at around 2-300 ft above the ground and gradually I climbed us back up into cloud.

The gear had never locked up. It was going to be a long, and a very quiet, journey home.

Why had all this happened? How had it come to pass that in addition to almost destroying a £50 million war plane, he had almost killed himself and his Weapons Systems Officer? It happened because they had taken that initial deviance and normalized it.

I was an experienced pilot but in the bracket where my over-confidence could well have been my downfall. The longer we’d continued performing the manoeuvre the more confident we’d become at doing it.

We had convinced ourselves that the rule breaking was for the benefit of the exercise and that what we were doing was essential.

But I’d almost destroyed a £50 million aircraft.

My actions in performing a zero ‘g’ bunt after take-off, in order to secure the gear, as outside of the rules as it was, had become the normal way to get airborne – I thought that what I was doing was right.

But I was wrong.

Knowing what happens it’s obvious he was wrong, but it’s also easy to see where it might not have been quite that obvious the first time he tried it. And it’s equally obvious where this problem might not be limited to flying. Life is full of very important rules for how things should be done, but it is also full of situations where it would be convenient and seemingly harmless to violate those rules. The initial violation almost always appears to be minor and in any event it will obviously be only temporary, but once we’ve done it the first time it becomes even easier to do it again and again and again… Until, before you know it, we’ve “normalized the deviance”. 

That article offers this formal definition for the normalization of deviance, from Diane Vaughan:

Social normalization of deviance means that people within the organization become so much accustomed to a deviant behaviour that they don’t consider it as deviant, despite the fact that they far exceed their own rules for elementary safety.

Vaughan coined the term “normalization of deviance” in the course of reviewing the organizational and management failures which lead to the Challenger and Columbia disasters. This is yet another example where it’s easy to identify which choices were wrong in hindsight, but apparently more difficult to avoid making those ultimately fatal decisions in the absence of such foreknowledge. And while there is some utility to identifying the problem after it’s happened it’s vastly preferable to identify it before the disaster. With that in mind I’m going to attempt some identification in advance. Also while the examples I’ve offered thus far, and most of the examples you’ll find, deal with small scale “deviance normalization”, as you might imagine I’m far more interested in whether we have any deviance normalization going on at a societal level.

Let’s take the last point first, what would it look like to normalize deviance at the level of a whole society? 

Let’s start by dipping back into the article. One of the things Davies mentions is how it can be very difficult to define what deviant behavior is at the extreme ends of things. While he admits that his take off maneuver was very obviously deviant, what if you’re trying to perform that evaluation on one of the military’s flight demonstration squadrons? In America we have the Blue Angels and the Thunderbirds, in the UK they have the Red Arrows, and Davies was involved in an inquiry after two of his friends were killed. As part of that he spoke to other people tasked with assessing the standards of the Red Arrows, one of them:

…told me that, when assessing a Red Arrows pilot, he found himself upside down at 100 ft over RAF Scampton’s runway in formation with two other jets only a couple of feet away.

How on earth was he supposed to know if this was normal?

How indeed? Davies doesn’t say except to point out that no organization is ever so specialized that it’s beyond needing external assessment. But the question of evaluation at the extremes is an interesting one, since that appears to be where modern society has ended up. In the same way that there’s not a lot of flight data for going hundreds of miles an hour, 100 ft off the ground with other aircraft only a couple of feet away, there’s not a lot of historical data about being able to form ideological echo chambers with anyone on Earth, with the ability to instantly communicate to any of those people, all while having less and less ability to know if that communication is actually truthful. That seems kind of abnormal or deviant, but how do we know?

To take another more recent, and perhaps more concrete example. News recently broke that Facebook paid contractors to transcribe it’s user’s audio chats. (Most of the other tech companies have been similarly accused.) Is this just a cool thing we can now do which well help Facebook deliver content people will appreciate more? Or is it a horrible invasion of privacy? Regardless of your answer to that question, it shouldn’t change when we get to the point where the transcription doesn’t require contractors; when there’s an AI that can do it. But I think it will. Perhaps more importantly, to get back to Davies’ point about needing external assessment, what do you think Facebook’s answer would be to these questions? (Spoiler: They think it’s awesome, particularly if they could get an AI to do it.) Pulling all this together, my first stab at spotting a normalization of deviance before it happens it to point out that technology is going to create a lot of “deviance” and that it’s going to be difficult to recognize, particularly if we don’t demand external assessment.

From there let’s move on to politics. To start, one imagines that deviations from the norm accumulate the longer an organization is around, and that this would apply at least as much to governments as it would to corporations, and probably much more so. As I have pointed out in the past, the United States and its government is older, relative to other nations, than most people think. All of this means that there should be plenty of examples of accumulated political deviations which have been normalized, and indeed I can think of several. I intend to provide a couple of examples of what I mean, but before I do, it’s important to point out that each could easily be the subject of its own post, and that by necessity, I am going to be leaving a lot of things out. Also, I’m sure that whether you view something as a deviation which has been normalized depends on your core political leanings.

With those caveats aside let’s start by talking about criminal justice. There are lots of things I could talk about in that space, but I’m going to focus on plea bargaining. Currently over 95% percent of federal cases end in a plea deal rather than a trial, and it’s not much better at local level. My sense, when I first thought of this as a potential deviation that has been normalized, is that it had only gotten this high recently, but when I looked into it I discovered that as long ago as 1945 it stood at 70%, and that it actually dipped to 63% in 1982 before starting a steady rise to where it is now. 

(I just gave two links to what appear to be the same statistic from the same governmental report, but while they agree on the general trend there’s a lot of variation in the statistics. For example, the Washington Post has it at 85% in 2000 while Albany University has it at 95% in 2000. Strange, but it doesn’t matter very much to the point I’m trying to make.)

Despite the fact that I was wrong about the increase in plea bargaining being a recent phenomenon, it was nevertheless definitely not a part of normal jurisprudence at the time of the Constitution. Once you start to dig into the history of it, it turns out that it wasn’t practiced with any frequency until “well into the nineteenth century“ and it didn’t come to the “attention of the public” until the 1920s. When “the general reaction-of scholars, of the press, and of the crime commissions themselves [which had publicized the practice]-was disapproval.” On top of this, apparently as late as 1958 it looked like the Supreme Court might declare the practice to be illegal, and while it didn’t, it didn’t formally sanction the practice until 1970. And while it seems normal now, this is all a deviation. The original guidelines for “safe jurisprudence” (similar to the rules for safe flying) included lots of rules about trial by jury, how the jury should work, the rights of the accused, what was and wasn’t permissible evidence, etc. But, at some point, after the system had been working well for possibly as long as a century, someone came along and said, “Wouldn’t it be easier if we skipped the trial and you just agreed to plead guilty to X, and in exchange we’ll make sure that your punishment is only Y?” And it was easier that once. In fact it was probably eminently sensible. But now, a century or more later, the original concept of trial by jury is used, at most 5% of the time (here’s yet another set of numbers from the NYT saying it’s 3% of federal cases and 6% of state cases) and the deviation has been made into the norm.

Why hasn’t this deviation been corrected? Probably because it only harms (or is perceived to harm) the powerless. Insofar as plea deals (and the associated practice of charge stacking) are bad they’re only bad for potential criminals. Not necessarily a coalition which is essential to anyone staying in power (see my review of The Dictator’s Handbook) and possibly a coalition whose support you would actively avoid.

The question of who the deviation harms is an important one, and comes up again when discussing my other example, though in a more complicated way. What is this example, you ask? It’s the current and growing practice of ignoring immigration laws. As with plea bargaining it’s somewhat difficult to tell exactly when or how this deviation started, but it’s easy enough to imagine why. Immigration enforcement is difficult, with lots of areas of questionable morality, and hard choices that have to be made. Still the current state has not existed for all that long. While the first sanctuary city was Berkeley in 1971, and a few other cities adopted that designation in the 1980s, most cities and states didn’t get serious about it until the 2010s. Meanwhile in the 90s there was serious concern about the state of US immigration policy. From Wikipedia:

The U.S. Commission on Immigration Reform, led by former Rep. Barbara Jordan, ran from 1990 to 1997. The Commission covered many facets of immigration policy, but started from the perception that the “credibility of immigration policy can be measured by a simple yardstick: people who should get in, do get in; people who should not get in, are kept out; and people who are judged deportable are required to leave”. From there, in a series of four reports, the commission looked at all aspects of immigration policy. In the first, it found that enforcement was lax and needed improvement on the border and internally. For internal enforcement, it recommended that an automated employment verification system be created to enable employers to distinguish between legal and illegal workers. The second report discussed legal immigration issues and suggested that immediate family members and skilled workers receive priority. The third report covered refugee and asylum issues. Finally, the fourth report reiterated the major points of the previous reports and the need for a new immigration policy. Few of these suggestions were implemented.

The yardstick mentioned in the article could be used in a Trump campaign ad, and indeed last year Trump put out a presidential message honoring Barbara Jordan, which upset numerous people since Jordan was a black female Democrat. It would certainly be hard to imagine someone similarly situated today being a Trump supporter, or even making similar recommendations. Indeed these days, many people consider it inappropriate to even use the word illegal. But, beyond what I’ve said so far, I don’t think it’s worth going into a deep dive on how this is a deviation which has been normalized, since I suspect you either already entirely agree with me or are never going to agree, but I would like to look at who it harms.

I said earlier that the harms of this normalization are more complicated. In particular there are a lot of fairly powerless people who are helped, and indeed that’s a good argument for the continuance of the practice. But beyond that there are also a lot of powerful people who benefit as well, and that, more than the powerless people it helps, is why it continues. As I pointed out in a previous post, other than Trump and a handful of other politicians, the lax enforcement of immigration is something which is supported by nearly every member in congress despite a majority of actual voters being against it.  The harm, or perceived harm, all falls, once again, on a group of people who have largely been without power, that is until Trump came along. Which is to say even if you don’t see any other harms from this particular normalization of deviance, it probably pushed Trump over the top in the last election…

Finally when we’re talking about deviations being normalized it’s hard not to turn our minds towards behaviors formerly classified as deviant. The entire culture war revolves around this process, and obviously there are quite a few people who believe that quite a few activities should not have been normalized. That “progress” is just another word for the greatest “normalization of deviance” of all. There are of course an equal if not larger number of people (depending on the country) who think this is ridiculous. It would be nice if sheer numbers could decide the issue, but I don’t think they can. All of these issues remain contentious, but, for the moment, let’s assume that in addition to being (at one point) labeled as deviations, that they are actual deviations. What would this mean? In the story I started with, that deviation almost led to a fatal crash. Is that also what we should be worried about here? Perhaps, perhaps not. The modern world is very different from the world of even 50 years ago, accordingly I would never claim that the normalization of these particular deviations will inevitably result in a “crash”. They may in fact be desirable in our current situation. Still, as I have repeatedly pointed out, there just might have been a reason for declaring these behaviors “deviant” beyond just massive historical bigotry. 

If we were to systematize all of this, you could imagine that things might operate in a cycle. Some “bad thing” happens, and as a result rules are put into place to ensure that particular “bad thing” doesn’t happen again. Initially, when the memory of the “bad thing” is still very fresh, those preventative rules would carry a great deal of weight, people would be eager to follow them, and they certainly wouldn’t be viewed as a burden. Gradually, however, the connection between the rules and the “bad thing” would fade in the minds of those forced to follow the rules. These rules would start to appear more onerous and less necessary. As this process continues, eventually rules start being broken. Initially this rule breaking wouldn’t cause any harm, and the longer things went without any harm the more the process of rule-breaking accelerates. All of this would continue until eventually, the “bad thing” the rules were trying to prevent, happened again. Naturally the rules would be reimplemented (and perhaps strengthened) and the cycle would begin anew. 

I’m obviously not the first nor the last to suggest that history, behaviors, and events might be cyclical, but my particular suggestion would be that while this is certainly true, it is also horribly complicated. Yes, history does move in cycles, many, many cycles which overlap, feed on one another and are weak or strong at various times and places. For the next couple of posts I’m going to examine a couple of other cycles and look at which might be strong or weak in our own day and age. As a teaser, in my next post I’m going to talk about how a nation would successfully implement communism. 


There might be some who argue this entire blog is a normalization of deviance! 

*Something*

*Something* 

Donate?