Combating organisational inertia

Why is it that we can't get our act together even in the face of a pandemic?

I

First, a worry and a question. Who should we blame for the fact that we suck at doing the bare minimum to get through a raging pandemic?

Let's start with the people first, at the lowest level of aggregation, before we go trying to look at the more institutional levels.

Scott Adams, before he became a reality denier, had coined a Law of Slow Moving Disasters. It said that when we see problems coming for a long enough while, even when things look inevitably bleak, we manage to find ways to solve them. It's a 'we'll figure it out, we always do' sort of a law. Whether it's through an innate technological optimism or a selective lens with which to view history, I've always had a soft spot for it.

Well 2020 managed to make that particular law look like a right idiot. We stood and watched as the economy and the healthcare slowly collapsed, even as there were clear paths laid out by Taiwan and South Korea and Singapore on how to handle the situation.

The Covid crisis clearly demonstrated, and still demonstrates, that we (the western countries collectively) have a phenomenal capacity for self inflicted punishment and the ability to affect absolute nonchalance in the face of danger. So we did the only logical thing - we bickered as this annoyingly realistic scenario depicted by Randall Munroe shows.

In fact enough people scoff in this manner than even with the supposedly Draconian restrictions that've been put in place, which have been too little too late at the best of times, the deaths keep rising and the hospitals keep filling up.

I can't help think about why this is. As the death toll globally has crossed a million, and that's with severe restrictions placed in a large number of places already, with very limited travel and recreation and socialising, things are this bad. Why is it so hard to convince some folks that this is so exceptionally dangerous?

I wanted to figure out if there's a way to understand what's happening without just calling everyone idiots. That seems too simplistic and too naive a viewpoint to be useful.


There is a persistent issue shown in psychological literature about availability heuristic. It determines how we get worried about things that have highly visible consequences where there's easy recall. It's why we are more worried about terrorists than car accidents, and why we believe crime is rising when it's falling. The affect heuristic meanwhile makes the recipient wonder why you're fearmongering when the issue clearly isn't that severe.

But I think there's a corollary to this. The emotional flashpoint heuristic. It seems like a supercharged version of the availability bias, but it's not just the fact that we can recall certain things better, but the fact that things which spark much higher emotional levels tend to stick our more in our mind.

Tversky and Kahneman argued that it's the frequency with which people recall examples from memory that seems to suggest to people the frequency with which the events occur. But with that justification, people would have to be making a distinction between people they've known to get Covid and have had a bad reaction, vs all other scenarios. And that's a low enough percentage due to whatever success we've managed to get.

The other argument on availability heuristic was argued by Schwarz, where the argument rested on ease of memory retrieval. This might come closer to the reality where people can clearly recall that there's a pandemic, they just don't think it's that bad for them. If the infection rate was half but the death rate was 5x less people would've died but we might have cared more.

There's no single emotional flashpoint here like 9-11 to spark outrage. There's no visible cases of horror to break through the barrier. There's just quiet deaths by the thousands that happen far enough away to be invisible, while millions cry out because their lives have been disrupted.

It could also be because nobody is ashamed to be associated with worries that have high emotional resonance. There's no stigma associated with being worried about terrorism when it happens. And that's because the impact rate of those actions are very high. If you are in a place that gets hit by terrorists, you're at high risk. And compared to that the 1-2% risk seems small. This is heavy speculation of course, but at least anecdotally it seems to confirm what I've heard.

That's why we saw countless examples of arguments about how this "is just like a flu", or "car accidents kill tens of thousands each year but we don't shut down the economy" or "this only kills old people, why are we all freaking out". That's why the arguments hit a core inside many people even as the facts were clear.

And it's not because the media didn't play up the pandemic either. There's been little else that we've (rightfully) paid attention to. The problem wasn't the top of the funnel awareness, but rather the conversion of that into actual action-intent.

The natural resistance that some folks have to government doesn't help. I reckon it's very hard to reconcile "they suck at everything they do" to "now I need to help them do their best by doing <insert outrage here>". That's a cognitive dissonance that has to be dissolved over decades, calcified as it is by propaganda over decades. You can't wipe that away so easily. And so of course reactance becomes absolutely a major factor when you identify some of the talking heads as a clear outsider.

In a clear case of tragedy like terrorism or even illnesses like Ebola, this threat doesn't exist. We all know what to be afraid of. And once we know that the rest, even if completely unnecessary, follows. Here we have the opposite problem. An illness that kills without much fanfare and therefore gets the questioning eyebrow.

It's worth recollecting that we're telling people to change their lifestyle for the foreseeable future, while suffering major economic hits, without a clear path out or an explanation as to an alternate path. For a problem they're not personally feeling, or worse they've personally felt but don't think was that bad.

And if you like more biases in your analyses, there's the illusion of control that allows us to think that we can control our exposure and environment far more than we can. As an anecdote I've thought to myself, when I've stepped out of the house forgetting to put a mask on, whether it's worth the trouble to go back and get one. I did, but I can't pretend there wasn't a decision point in there. And once you agree there is a decision point then at the very least we can't be surprised that some people take the other fork.

Moreover I can't help think that the countries that have done well in handling the pandemic have all done one thing well - they have had extraordinary clarity in their policies and have carried them out as communicated. If there was resistance on the part of their populace this is clearly one way to have that become a non-issue. By breaking through their immediate mental models. Initially I thought that there are advantages to doing this in more disciplined and authoritarian regimes like Singapore or even China. But New Zealand and Taiwan should make us think twice about this knee-jerk conclusion too. Seems like clarity isn't antithesis to democracy.

But equally importantly something I've come to believe is that clarity is still underrated. Without that I'm not sure if I can even easily blame parts of the population on their decision to take things lightly. They have the combination of biases working with their own pattern recognition and answers are coming up in opposition to reality. If ignoring everything is the only way their cognitive dissonance can make the world make sense, then calling them idiots isn't helping change the situation for the better.

Instead we had extraordinary levels of dysfunction across politics and media. For me, it's made it impossible to watch a movie about a pandemic ever again. Even Hollywood treated mask wearing and government lockdowns as fait accompli, the enemy was the virus. But through a combination of wilful blindness and active stupidity we've basically reinforced the negative belief in the populace.

Mostly though, it still seems unkind to blame individuals in this instance. It was the government's job to have clarity in communication and decision making so that our inherent biases didn't feed in on itself. It wouldn't have solved everything, but making perfect be the enemy of good is also how we got to this point. Vacillation kills.

Onwards to step two.

II

So if our dysfunction isn't a function of individuals acting crazy, but something more systemic, what could it be? And is this a symptom of a particular malfunction in our circuits, or is this something more pervasive? Let’s have a look at the organisations themselves.

There's a fact that's so well known that mentioning it seems banal sometimes. It's that companies ossify in their evolution and become uninspiring, uninnovative and just plain stagnant. Sure there exists research on how there's still major value being created within the largest companies, even in non sexy industries, but it doesn't feel true. The greatest culprit of course is the government, where everyone on the left and right and every other political position besides is convinced that it barely functions.

Francis Fukuyama, he of the end of history fame, had an argument also about low trust and high trust societies which purported to show why certain countries seemed to be more innovative. The theory was that they had higher sociability, which meant groups could form faster, and so you could do more things.

The argument is that us in the West clearly are part of the high trust camp and therefore are far more sociable and able to just achieve things that others can't. In some ways this is clear too. Would you rather start a company in Delaware or in Abuja? With nothing against Abuja, even Nigerian entrepreneurs were happy to sign on in bulk to Stripe's Atlas because they saw the value.

It's also clear in other ways. Despite the stupendous amount of graft, would you rather move your successful giant company to the US or away from it? If there's a lesson in Ant Financial’s IPO troubles it's that societal trust in its governing institutions isn't to be taken lightly.

And yet, with all of those benefits in hand, we've been spectacularly bad at some things which seemed like they should be all too easy. For instance, just in the Covid timeframe:

  • We haven't been able to distribute the Covid vaccines once we've gotten the damn things invented, manufactured and pre-ordered. They're piling up in warehouses and in some places, inexplicably, ready to be tossed. Scott Aaronson eloquently argued against this idiocy to, as far as I can see, little avail.

  • We haven't been able to manufacture enough masks or PPE. Others have written far more eloquently on this failing, but by any measure this is spectacular. In the sense that it's an insane spectacle.

  • We haven't been able to put together a half decent government policy to just tell people what they should do. The policies in the UK are constantly on a seesaw, as if we're arguing over knicker prices in a wholesale market, and wont to move in any direction on a whim. Singapore sent its residents a message daily with the latest news and updates, continuing to demonstrate why it should run everything. Bear in mind that I am not asking for good policies here, just reasonably clear communication.

  • We couldn't get a test and trace working, and still can barely get people to use one. There was one by Google and Apple, there was one developed by various governments, there were development issues, there were scaling issues and there definitely were communication issues.

Is this just a particular issue with the government being uniquely useless? I know everyone likes to think of the Good Ol' Days of great government, which usually includes WWII (80 years ago, mind) and Moon Landing (50 years ago). But surely there has to be a solution. Even Operation Warp Speed, which is rightly hailed as a success, required ridiculously low levels of actual capability and primarily succeeded through being speedier funding and pre-purchase mechanism.

So let's turn to the libertarian utopia that are companies that are supposed to be much more efficient in their operations. Are they though?

In one sense, absolutely. Amazon dealt with a once-in-a-lifetime deluge of requests by everyone around the world beautifully, delivering toilet paper and essentials and care packages with enviable verve. Netflix made streaming entertainment the mainstay for people. Doordash delivered food and went public with a bang. Elon Musk ... well comments are insufficient here.

But these were examples of companies that did what they were doing before, just better under difficult circumstances. Where are the equivalents of us repurposing Detroit car factories to make armoured cars and fighter planes? There were a few - US factories retooled from hoodies to face masks, Honeywell ramped production of N95 masks and moved chemical plants to make sanitiser.

Let's compare this to China, where they exported 150 billion masks and several hundred thousand ventilators. Blah blah centralised planned economy blah blah. In the middle of a pandemic, when they needed to get their act together in short order and do stuff, they did and we didn't. And that's a black stain no matter what. And it's not because the Chinese products were "shoddily made", but because they were made on time.

So while we should give a couple of high fives to Honeywell and co, it seems woefully inadequate when compared to what it could've been like. Seems it should be instructive to actually understand why this is the case.

III

Perhaps it's worth asking another simple question. Why is it that we've collectively become this culture where to do anything seems impossible. There’s something in the water that seems to make smart, rational, motivated people to collectively do stupid things once they’re inside the strictures of an organisation.

One place to start is probably by having a look at how organisations look at themselves. The 1998 book Seeing Like a State is a decent place to start. James Scott blames our failure on high modernism, a culture that aims to transform our entire institutional apparatus into becoming more scientific in its essence, and fails. Our obsession with gathering quantifiable epistemic knowledge leads us to develop huge blind spots everywhere else and lose our metis.

The argument here, in a tiny nutshell, is that we try to make societies too legible to laws, and therefore arranged in a fashion that's easy for us (humans) to analyse and control and troubleshoot, and that leads to suboptimal outcomes.

Despite the fact that metis by itself, standing in for "local knowledge", more or less is a component of how complex systems function (chaotically, through intense self-reference, and slowly groping their way towards fulfilling a need while under ecological pressure) it's a useful proxy to use to think about the case of what-might-have-been.

Examples abound: Brasilia was constructed as a bureaucratic utopian capital city, and resulted in the people who were forced to move there (guess where they worked) getting diagnosed with a disease called Brasilitis. That's humanity 1, Le Corbusier 0 for those keeping count. Pol Pot's Cambodia and Maoist China had giant 'millions of dead people' sized gaps in their conception of society. And they failed too. The issue with Balinese agriculture being messed about by outsiders is an example too. So is scientific forestry (which I can't wait to see as a graduate course now) suggested cutting down all trees in Prussia and planting Norwegian spruce, which screwed up the entire ecosystem.

Needless to say, there are more. And also needless to say, finding examples of cases where "I think I can plan this complicated system to work better" falls flat on its face are incredibly easy to find. And it's not just a quasi-Austrian take on collectivist planning. Saying "I think I can design a market to make this work better" also falls flat through excessive standardisation.

In of itself, this doesn't seem like that much of a story these days. In some ways the explanation seems about a decade out of date. But finding a cohesive argument of why people keep thinking this is. It's because, as Scott argues, we live in a culture that tries to figure out the most efficient ways of doing things. And worshipping at the altar of efficiency means we end up causing a whole lot of harm.

Organisations often deal with uncertainty by focusing on things they can control. Leaders ask for information, and getting information involves measurement. And taking measurement means you have to know what you're measuring and how to measure it. And also that the things you're measuring don't go all Schrodinger's Cat on you and change as you're measuring them. (I am the middle manager that did so well that you got all of 17 reports on it, which you didn't read, which means I must be even better then you thought).

And so everyone starts the management by measurement process. And the process starts the fad of measuring things, which in a helpful enough way, changes what you're measuring to make it less useful.

It's worth saying that occasionally, when it works, things work like magic. The financial system is an incredible success in how it has distributed credit, for instance, and helped make time travel real - we invest money now to create awesome stuff to create future profits. It was only by moving away from its artisanal roots that this was made possible at a global scale.

The downfall of 08 is also instructive here in that it came because our over-reliance on the models, and quantisation made us forget about the actual link to reality. It was a rude awakening that the map is not the territory.

But let's check in with our Eastern friends. How's China getting around this problem? By making sure that the top level goal that the Party sets has supremacy. It's a single threaded decision making engine and the ultimate goal to satisfy is clear.

The decisions are not often made by people who have a credible belief, whether through fear or proactive aim, or seeing multiple outcomes and maximising their probabilistic winning. They live in the realm of binary outcomes in this multidimensional world.

The second big error that creeps up in organisations is a fear of ambiguity. The Ellsberg Paradox documented how there's an aversion to ambiguity in business owners, who showed ambiguity aversion whenever they had fear of an outcome. The literature on the topic at least suggests that ambiguity is a stallion to be mounted and bridled, through regulation, through back-to-basics rules, through rational thinking.

The Ellsberg Paradox or its Keynesian equivalent aren't just examples of how individual risk aversion differs from homo economicus conceptualisation. (And that it was named after someone who leaked the Pentagon Papers is just delicious). Similarly another explanation called the "other-evaluation" seems to also show an effect where a decision maker, when faced with a choice, anticipates that others will evaluate her decision and makes one that's most justifiable. There's even a model showing how, in policy implementations, ambiguity aversion is normatively justifiable and rational!

In whichever conceptualisation, it shows that there's incentives that act within the participants in any organisation that simultaneously makes them act to manage things legibly while also ensuring they need to reduce ambiguity w.r.t their actions.

Together this creates a wonderful combination of high legibility and increased defensibility for all decisions made. And thus starts the problem.

It's easy to see how this also shows up in multiple modern organisations. To go back to vaccine distribution, you need the leaders to demonstrate to their constituents and colleagues that they've done this successfully. They have to make legible policies that are understood by their bosses, the public, their underlings and the press. So naturally they want to cover all bases, with the theory that the best order is one that is mighty comprehensive, since that's what they're being graded on.

The same issue happens in healthcare. Incentives for doctors are skewed towards faster resolutions, better treatment metrics (sometimes) and reduced administrative burden. This strikes against the giant apparatus of hospitals to maximise care provision, of insurers to minimise payouts, and of the entire administrative staff to make a way for this to happen under the Byzantine system we've set up.

Does this also affect institutions like the FDA? You bet. Where's the individual credit for taking the risk and saying yes to a new drug, when the downside is so massive? They want to make policies that are legible and orderly in hindsight and defensible to everyone regardless of the individual binary outcome.

And so in crafting and implementing regulations we have similar tradeoffs. The regulators want to only do things that will satisfy their constituencies. Yes there are practical issues of lobbying and corporate interests to deal with in the crafting of the laws. But the implementation leaves a tremendous amount to be desired. The individual departments have widely varying levels of incentives to do anything. For those who are interested in upward mobility, their ability to do so revolve around them mostly not screwing up and looking good.

So when they're asked to do things that haven't been part of their job descriptions at all, then naturally their incentive would be to not take any seemingly unnecessary risks and do whatever they can. Taking risks is a currency that's only spent very carefully.

The theory isn't that we should do whatever we can or want either. There are no easy answers except at the extremes, and answers at the extremes aren't particularly useful.

If chaos is allowed to rule in a somewhat more dynamic fashion you'd see companies that are like cities, ungoverned chaos zones with unpredictable spikes of genius emergent at regular intervals. One could wish that more companies like this existed, or that the cities would have better governance, but either of those are in competition with their internal incentives and unlikely to come about.

IV

We have this cultural conception that thinks of companies or organisations as a machine. Something with fixed inputs and desired outputs and which cranks to make that happen. But companies are actually weird entities. They're assumed to have non-human values and non-human needs which it exploits just like a bacteria does that's stumbled upon some juicy sugar. But in reality they contain true human values all through its core.

The CEO has dreams of succeeding amongst all CEOs, buying an island and kicking Larry Ellison off it. The CTO has dreams of being recognised as a genius who singlehandedly pushed the technological frontier forward. The CFO has dreams of ... actually I don't know what CFOs dream of. Perfectly balanced books probably.

People who sell things like to understand and know the people they're selling things to so that they can sell them more things. There's a sort of dance that happens in understanding the need, in product selection, even in customer service that makes the whole thing run like an extremely human endeavour.

We could remove the "friction" by removing the people. Maybe there's a CEO AI that can give the right decisions to be made. Maybe there's a selling AI that helps figure out best ways to sell the product to the customer.

But this quickly becomes a divergent series. You're not going to be able to enable an AI to follow the human passions that guide product design or user need research beyond a minimal goal before it quickly runs aground that much of what it's aiming to satisfy is itself a moving target.

The whole reason for this house of cards that we call an economy is not to maximise some growth as a Platonic ideal, it's because we all want different things. And in our extreme human need to get those things we have built this giant machine that gives it to us. And in order to make that happen it turns out someone needs to smelt iron ore and someone needs to write software and someone needs to mine palladium so that we can drive pretty cars around.

There will be no non-human economy run entirely by AIs like described by Nick Land. Why should there be? It seems an extraordinarily inefficient system that would destroy the GDP, for one thing.

(Parenthetically this is also a reason I've always found the brain-in-a-vat argument specious. Sure, it's possible, like anything's possible. But if I were a brain in a vat that's capable to dreaming up scenarios, this one seems a little too filled with body-regret to make the hypothesis worth considering!)

But in a human built system there will always be inefficiencies and, well, humanity to deal with.

We won't be able to easily get away from our biases necessarily. But we can set up systems that don't quite so obviously suffer from them.

It doesn't even require something extraordinary like a pandemic. Other countries are managing it. Other companies are managing it. And we used to manage this too, and much more, in the days gone past.

What's needed is a sword to cut through the biases and make those in the middle able to actually act rather than wonder about all the issues that they might have to deal with at a later date. The ambiguity aversion can be overcome by pushing away from fear of punishment to one of hope.

That's a management shift. Yes the blame game is reinforced by the media and the culture and other elites and Twitter. But if there's one thing the Trump era has taught us it's that sometimes it's worth pushing ahead on your agenda and damn the noise. It works more often than one might think, especially when we're so far on one side of this weird Laffer-ish curve.

I've started thinking of it like ferromagnetism. You need the atoms to look in the same direction in response to a force. And you won't do that just because you're forced to. In any organisation beyond, say, 100 people, it's incredibly easy to hide for long enough that it becomes a problem. There will always be committees and material shortages and division of labour decisions that hurt someone. The way you overcome is by knowing you're pulling for the same team in a more ineffable sense.

The key success stories within organisational strictures all revolve around something elusive and intangible yet critical. It's the fact that the structure that is able to do things fast have two characteristics - 1) they're motivated by a desire to do something, and 2) they're fractal in their motivation - i.e., they show similar motivational acts at all levels of the organisation.

This is one of the advantages most ascribed to startups. They're often founder led and have a clear mission. And they're young enough that the natural process of motivation diffusion (my phrase, unsure if I've stolen it from somewhere) hasn't kicked in yet.

So when Zuckerberg tells Facebook employees to move fast and break things he's explaining something crucial about the culture that the individuals in the middle can understand. They're being given permission to veer away from fear driven accountability to being empowered to try things.

When engineers think of joining a startup where they might have the chance to "make a difference", that behaviour gets made real because there's a clear line of sight from vision to execution, since the size of the company is small, and because they're part of a tribe rather than someone just seeking a paycheck to do some specified tasks.

In a sane world you could also tie it all down to incentives. But as someone who's designed a fair few performance management pyramids, it's well nigh impossible to design a system so well that there's no value leakage. There's no substitute for intrinsic motivation and alignment towards a goal where people feel actually empowered to actually make the decisions they're supposed to do. ESOPs can help, but if you think there aren't crazy political fights inside smaller organisations you haven't seen enough.

Share

V

So where do we end up? Do we just throw all institutions out and start over?

It's not clear if something that drastic is required. This feels closer to the act of removing barnacles from the ship rather than building a new ship.

This is partially the equivalent of Theseus' ship where through replacement of 'almost good enough but not perfect' parts we get a ship that doesn't work well enough.

But mostly this is the result of multiple attempts at optimisation at the decision making process stage vs actually trying to test things by experimenting and implementing.

The balance between how much planning is appropriate vs experimentation seems widely out of whack. And as time goes on the internal processes and incentives together push towards increased planning and forecasting effort, as we saw above.

Sure the knowledge that organisations face methodological drift isn't new. The fact that this is mainly driven by the individual motives of the middle layers that work in them is slightly new but still understandable from the point of view of individual biases that are flared up by the organisation itself.

The key here is that the systemic incentives as it's set up today irresistibly leads to motivation diffusion and this organisational drift. It feels like an organisational law what we should be aware of and work around.

And what would be an example of where this type of drift is invisible?

The first batch of counterexamples are the list of companies that have survived for centuries. The ones that have shown remarkably little organisational drift.

To quote the Oracle:

According to a report published by the Bank of Korea in 2008 that looked at 41 countries, there were 5,586 companies older than 200 years. Of these, 3,146 (56%) are in Japan, 837 (15%) in Germany, 222 (4%) in the Netherlands, and 196 (3%) in France.[1] Of the companies with more than 100 years of history, most of them (89%) employ fewer than 300 people.

And why is it that Japan seems so over-represented?

In Japan, very old companies, called shinise, are particularly prestigious. A nationwide Japanese survey counted more than 21,000 companies older than 100 years as of September 30, 2009.

It's not conclusive, but damn if it isn't suggestive!

It should also be said that most of these companies aren't solving some new need that emerged in recent years, but solving deep seated needs that barely shift in centuries. People have to eat, people have to go to hot springs, people need funeral services. Until Aubrey de Grey achieves his goal that's not going to change.

The second curious thing about the companies is the sheer narrowness of their ambition. For instance, Tsusen Tea, started in 1160 AD, has a 38 year old proprietor who says:

We’ve focused on tea and haven’t expanded the business too much. That’s why we’re surviving.

Spitting in the face of Milton Friedman seems to be their entire raison d'etre; and to make just enough to live and survive on. Growth isn't even in the vocabulary. Neither is competition. It's an acceptance of stasis as a way to live. And unlike YouTubers who've also figured out a chill way to make a living, the work here seems substantially (often physically) harder.

Another example of survival is in the eschatological realm. The Catholic church has been trundling along and has stayed relatively steady through thick (crusades?) and thin (sacking)! It seems a great case for something that's survived an incredible number of twists and turns in its lifetime and considering the variety of corner cases it's survived would make a great candidate for an organisation that would survive another 1000 years.

Their strategy also involved an incredibly limited amount of movement in their core proposition. They still believe in God and Jesus and the general shape of the society needed to meet them. They've moved on topics like occasional warring and rights for women and acceptance of homosexuality, even if extremely slowly, by bringing them into the fold rather than redefining the fold.

The interesting aspect of either one of these types of organisations is how little change is demanded of those who are part of the middle management. The smaller organisations deal with that issue by just not having anyone in the middle. And the church deal with it by being explicitly taking away any need to have an incentive to change.

Did their presence slow down change? You bet! But if staying in a static stage is the whole point, then ossification is an advantage. There's no incentive to grow, only to move within the ranks. Is this how a GM at Uber thinks about her role? Decidedly not.

VI

Stepping back inside the organisations we have cause to discuss it's the combination of growth expectations from outside and the pressure to nor shift from inside that causes the problems.

Either you do away with the entire cohorts of managers, or we redefine ossification to be the operating principle. And when we do the latter then we lose any ability to be flexible or responsive to the situation.

And doesn't that sound like the actual problems we hear from the government? Or even older and more boring companies for that matter?

The government seems pretty decent at the basics. It runs elections, collects taxes and makes the economy work. It runs the military. It gives healthcare to a large proportion. It employs millions.

Meanwhile Oracle, to take an example of an ossified tech company that's thought to be constitutionally incapable of innovation, they do similar things too. Thousands of companies run on their ERP. More on their databases. Everything works. They just can't do much that's new. And that'll remain true even if they end up buying Tiktok.

What about insurance companies or banks? Same story. They're also old, and set in their ways. Famously so in fact. But we still take money out of Barclays or JPMorgan, and take policies out with Allianz. It works.

It's just that none of them can change direction. Management theory meets a ferromagnetic alignment challenge. There are similes about turning oil tankers as if that explains anything.

If this were true then you'd see problems cropping up wherever you ask one of these organisations to shift their ways or adapt in some way. And isn't that what we're seeing?

  • The church adapting to a more liberal world through pronouncements on climate change, on gay marriage, on capitalism, on being good samaritans, and more.

  • The FDA trying desperately to claw out of its accrued layers of bureaucracy to move faster to approve drugs, to do cost benefit analyses on vaccines, to not stifle testing shops that have sprung up around the country during Covid.

  • Microsoft seemingly imitating a phoenix as it becomes a cloud computing company

  • Apple actually doing a phoenix and creating consumer electronics manna

The pain happens when there are extremely good reasons for those inside the organisations to drift in their mission from the overarching one to their individualised risk minimised versions.

When the FDA tries to tell people they shouldn't test so much, or put out guidelines that mask wearing isn't that great after all, it isn't because they're trying to be malicious. Or even act stupid. It's because somewhere inside its Byzantine bureaucracy exists multiple parties who have high degrees of risk aversion that tell them to not allow others to test in case they get things wrong. Or to not mandate mask wearing unless it's been proven conclusively that it's effective.

They're stuck inside an organisational Ellsberg loop, and there aren't easy exits. "Couldn't they act faster?" Sure, in a perfect world. But that would require a dismantling of what people think their jobs should be. And historic patterns of behaviour get ossified atop each other so much that they simply cannot actually listen to the truth happening all around them and make decisions.

Again, this isn't because individuals in there are stupid, or malicious, or misinformed. On average they're, if anything, highly educated, motivated and often highly motivated. It's an institutional failure. "Could we blame the leadership for this?" Sure we can. Why not. But their failure is not in the output (push through the decisions faster) but in the process (unblock the system while not destroying its original function). And the process is hard to edit. Unintended consequences are everywhere. Trying to avoid them causes delay, and delay causes further analysis paralysis.

Dealing with change is never easy. There's always maintenance to be done as an organisation grows, and that creates technical, political and organisational debt. There's always cost trade-offs in how information spreads through an existing hierarchy.

If we want systems to respond to circumstances and be flexible in response to the environment, we have to build that in. There has to be a willingness to accept a cost of rebuilding occasionally that's implicit in our calculations. Every now and then there needs to be a clean-sheet design. Costs of experimentation have to be made lower. Prices of success has to be made higher for those who are not raising Silicon Valley money to start companies.

Without that we can moan as much as we want, we'll still be stuck in the same endless loop. While complaining about it is fun, I’d prefer some understanding, and some action.