Unpredictable Patterns #94: Making threats
The art of making effective threats, threats and regulation, threats and the economy, thinking in threats and throffers
Dear reader,
This week’s note will be about something slightly unsettling: the rising importance of threats. The increased geopolitical tensions and weakening economy is likely to turn up the temperature on how we interact, and understanding how threats are made and what makes them work is going to be a key skill we need to develop more. Here are some pointers on what that will take.
Coercion and the human condition
Whether we like it or not, we live in a world that sometimes requires coercion. It can be the hard coercion of the law in the case of criminal activity, or the soft coercion of parenting where we try to convince our children to eat their greens or go to bed, but coercion is a part of the human condition. It is not a nice part, or an especially pleasurable part of our existence, but understanding coercion is important to understand everything from relationships to war.
Let’s say at the outset that coercion often is wrong, and that it is morally reprehensible to coerce someone into doing something they do not want to do. There is nothing admirable about coercion, and the reason we study it is not to recommend it - but to figure out how to analyse and deal with coercion as a fact of life.
Coercion is at the heart of the most primitive concepts of power. The primal image is that of someone grabbing another person, forcing them to go this way or that, leading them to a prison cell, say. This is power at its most primitive - physical coercion, the imposition of one’s will on another. Coercion negates the subject’s autonomy and free will and forces them to act in a way that the attacker determines.
But physical coercion is not the most important form of coercion. If we want to understand coercion’s role in society and politics we need to look at something far more complex - psychological coercion. Here, we force someone to act in a certain way but setting up a situation in such a way that the choices to do something else is either not available or very costly.
Psychological coercion depends on making threats, and the art of making threats is an important area of study for any student of politics or economy overall.1
Threats have come to the fore again in international politics, with the war in Ukraine. After the Cold War, threats diminished in importance in international politics, but we now are living through an unfortunate renaissance for threats as a means of political action. And it is not only in the Ukraine war that threats have become more common - we also see threats made in the conflict around Taiwan and in other political conflicts.
Understanding threats, then, is yet again becoming a key task for anyone who wants to engage in foresight work or political analysis. In this note we will look at three aspects of threats - first we will ask what a well made threat looks like, then how threats are designed into policy and legislation and finally we will look at threat modelling as an analytical practice.
A well-made threat
How do you make a good threat? It may seem a crazy question - when we have come to a point where we are threatening each-other there seems to be no good outcomes, right? So why should we pretend there is such a thing as a good threat? Well, what we are looking for here is not a beneficial threat as much as one that is functional and effective - and the reason we want to understand that is that we are actually better off if actors in a conflict are really good at making threats.
A badly made threat is likely to worsen a situation and lead to an escalation of the conflict, so we do actually want people to be able to make really effective threats.
The key component to an effective threat is the credibility of the threat.2 Or more precisely, the credibility of you taking the action that you propose to take if the other party does not comply with your demands. This means that a well-design threat needs to outline an action that you have the capability to undertake, and that will have consequences that are worse for the other party than for you - it needs to cost more for them than you.
This is hard. Designing a threat that hurts the other party more than you is not as easy as it sounds, and in some cases the threat is designed in a different way: the action we takes means that no-one wins. This is the threat at the heart of nuclear deterrence: if you deploy nuclear weapons, I do too and then the world ends. There is no more world to fight over, the game is over for all of us.
Such threats depend on the other party seeing that if they were to take the action we want to stop them from taking, there is no next move in the larger game. It ends there. That was always a dangerous move, since it presumes that we assign the same value to the continuation of the game. A dictator who realises that he will likely be shot or hanged if he gives up, may well enter into that apocalyptic mindset where he declares that if he cannot live, then the world doesn’t matter.3
In this extreme case, the threat does not work because one party to the conflict simply attaches no value to the next move. Such threats should be avoided at all times, since mutually assured destruction (as this threat paradigm is known) is a badly designed threat when we have desperate players.
It is much better if we can find some action to take that would impose more costs on our opponent than on ourselves, and this asymmetry of costs lies at the heart of a well-designed threat.4
The simplest case is the person with a gun who asks you to hand over your wallet. A robbery is a despicable act, but as an act it is based on an asymmetric threat that can work. Now, the act of shooting someone is not costless in any way shape or form - but given the asymmetry of the threat it is unlikely to come to that - but it needs to be considered.
The robbery scenario, then, highlights the other aspect of a well-designed threat - you have to make it credible that you are ready to absorb the cost that the action will impose on you. Even if you have found an action that is asymmetric, there will always be some cost to you - and so you have show that it is likely that you are willing to carry that cost.
When that is not the case, the asymmetry does not matter. This is familiar to us from numerous movie scenes in which one character points a gun at another, and the threatened person calmly asks“What are you going to do? Shoot me?” and then proceeds to take the gun from the person pointing it at them. In variations of this scene we also find ways in which people can make a threat credible: the person with the gun may fire it in the air (to show that they can fire the gun) or even shoot the other person in the leg or in some other non-fatal way.
Making a credible threat is something that requires some kind of evidence or commitment mechanism at play.
Evidence can be provided by reputation. If you are ever threatened with a gun by someone who you know has shot several other people, that reputation will make the threat instantly credible.
The other version - the commitment - is more subtle. This was at the heart of what Thomas Schelling explored in his writings about the strategy of conflicts. Being able to pre-commit to a course of action creates credibility as efficiently as - or even more efficiently than - a reputation does.5
The creative design of commitments is a key part of any strategic threat - the automatic firing systems for nuclear weapons, so-called dead hand systems6, ensure that even if you have been able to hit my command centres, the weapons will be fired. The commitment is built into the architecture of the threat at the technological level.
In the robbery situation or the hostage situation, credibility can be created through the shooting of another person than the one being threatened - another popular movie trope. This also creates a commitment, because once there is a one person dead, the person shooting is facing legal consequences no matter what - there is no walking away from the situation.
Schelling also suggested that threats can be designed with a random component to them. Again, we find this in the world of movies - the aggressor that spins the chamber of a revolver and declares that there is one bullet in there - and then subjects their victim to an involuntary round of Russian roulette. The act of pulling the trigger shows credible commitment, even if the result is just a click.
A well-designed threat does not need to be precise. In fact, creating uncertainty about how you react - if combined with the real availability of a number of known retaliatory actions - can also create a functional threat. Anyone who manages to project a certain amount of insanity is likely to be able to present a credible threat to a rational actor, since we cannot predict what they will do. If the crazy person also has demonstrated a range of destructive capabilities, we will be reluctant to cross them and their threats carry definite credibility.
It is very hard to threaten the truly insane, but they can be very threatening.
A well-designed threat then requires thinking through both costs and credibility, and modelling the other parties reaction - to the extent that is possible. But it also requires something else - and that is the willingness to follow through on a threat.
This is what all parents are told: never threaten your children with something you are unlikely to do. If you say that you will send them to bed without dessert, you have to have the strength to do that even when they cry and you feel lousy about it - otherwise you teach them that your threats are empty.
Commitment and credibility ultimately depend on follow-through. Establishing a consistent pattern of acting on your threats is key to being taken seriously when you make a new threat.
Policy as a game of threats
Explicit threats may be uncommon - and that is a good thing - but there is value in looking at more peaceful situations as threats as well. In one sense, then, a legislative or regulatory process can be modelled as a threat on several different levels.
The first is the obvious one, where the policy maker says that an industry needs to clean up their act or else there will be legislation. In this case, the threat is not even implicit. This requires, however, that the policy maker in question can make it credible that the legislation will be more costly or punitive for the companies than the current situation. Such threats also need to be credible in the sense that the legislation envisioned - the threatened action - should be possible to get through the legislative process.
A parliament caught in gridlock severely reduces the threat-making capability of policy makers. In fact, legislative capability is one of the key indicators of political power in any space. The unencumbered ability to write, release, negotiated and decide on legislation is a necessary pre-requisite for any political threats of legislation.
The second model is also fairly open, and that is where policy makers build a threat-mechanism into the law. This is usually done with fines - costs and remedies that the legislator can hand over to a regulator to impose if the companies do not act in a way that complies with the intention of the legislative body.
Fines have become more and more common in European legislation, and the design of the fines clearly indicate that they are envisioned as components in a well-designed threat: they are set at levels that fall just short of existential threats to companies at percentages of global revenue that make the difference between a company being profitable or running a loss.
It is a lesson in threat design to study how this has been applied historically in, for example, competition legislation. Here we see that even if eye-watering fines have been imposed on companies, they have not led to any specific reactions from the companies in question beyond adapting to avoid future fines. If the risk of the fines had been even greater - if they had bordered on existential harm - companies would have had to think about if they should withdraw from the European market. This clearly would be an unwanted outcome for Europe, and so the fines have been designed to deliver a sting without forcing an exit.
The other aspect of this is of course that the policy maker is counting on the economics to keep companies in the European market, and so can threaten fines to the extent companies are dependent on European revenues or hopeful about European revenue growth.
In this case, the threat is not the legislation as such, it is the enforcement of the legislation. And again, there is a lesson here - there is significant uncertainty built into larger European regulatory acts that will make companies hesitant to push the boundaries in the European market. The model of legislation and enforcement as a threat highlights the surprisingly aggressive dynamic that is at play here.
The third way in which it is interesting to think about legislation and regulation as a threat, is in the relationship not between governments and companies, but between the governments themselves. In many ways, legislation that targets primarily one country’s companies can be modelled as a complex mutual threat: your companies are powerful in our markets, but we will regulate them as to ensure we have control over them and benefit from their economic growth.
The conflict here is one where regulation are used to try to keep some kind of power balance between foreign economic power and European regulatory power. But is this a sustainable model? The European regulator model is sometimes said to aim at making Europe a regulatory superpower - suggesting that regulation is an act that credibly imposes more cost on those regulated than the regulator. But is this true? This mutual balance falls apart if it turns out that the cost of the regulation actually also is imposed on European companies and reduces economic growth in Europe more than it restricts the economic power of other countries. If the means Europe uses to threaten other countries imposes more harm on Europe than on those countries in the long run, it risks falling behind even further.
This should worry European regulators far more than it seems to do, and at the very minimum those interested in European economic growth should set up a set of “kill criteria” for the idea that regulation is a credible threat that can create economic balance. This could be done by defining a basket of metrics that must be met by 2026 - and if they are not met the EU should shift path and shift position and try to find other ways to establish the desired economic balance.
If we continue to see reports of large drops in venture capital investment in Europe, as we saw following the GDPR, this would seem to indicate that regulation is a less well-designed threat than many European policy makers have thought.7 Even if there is an increase in compliance related start-ups, it does not help balance out the drop either - and compliance based start ups are less dynamic than innovation-based start ups, creating slower growth down the line.
Thinking in threats
Which threats can you credibly make as an organisation? As an individual? Have you thought through if there are any threats that you should design and have available if needed? Probably not - there is something slightly distasteful about thinking in threats, but maybe this is the time to think more about threats: the weakening economy, the increased geopolitical tension and the harsher political climate seems to suggest that we will see more, not less, threats in the future.
Designing threats is not necessarily something we do because we plan to deploy those same threats. We do it because it allows us to make explicit the power dynamic between different actors in the environment we are in. Consider it a form of war gaming: looking at what options for making threats you have and what others would do in response if you put those threats in place. Threats are options!
You may also want to look at a softer approach, and model the throffers you can make. Throffers are mixes of threats and offers - almost along the lines of offers you cannot refuse - and they allow you to think about power dynamics in the overall negotiation of positions that we are constantly engaged in.8
And you can move beyond intentional threats, and look at what threats you are facing that are purely environmental. What are the trends and events in the world around you that can be perceived as threats? Where do you face threats to your business model or technological position? Such threats are games with nature, but they can still usefully be modelled as threats, since they often combine with other threats in our environment.
And finally - maybe it would be a good idea to think through when you would actually commit to making a well-designed threat. In what situations would you consider withdrawing from a market? When would you contemplate breaking off a business relationship? At what point do you share those possibilities as threats? Implicitly or explicitly?
It is exceeding rare to see companies do this - and one reason is that often conflict costs more than it delivers. The very fact that you have threatened something is often turned against you - it is seen as a form of blackmail. But the well-designed and thoughtfully delivered threat is a legitimate move in both the game of business and politics. It always was - and even if our rhetoric has changed, it remains a key element in our human condition.
Understanding and, when necessary deploying, threats is sometimes key to successful strategy.
So what
The underlying assumption in this note is that threats are becoming more important in international politics and business. Understanding the mechanics of threats, then, becomes really important. Whether or not you choose to deploy threats of your own, you should understand how threats work. We can, from this, make a few predictions.
First, that we should expect to see a few geopolitical threats be tested, and that the way those tests turn out will be pivotal. When someone is caught in not making a credible threat, their power is severely reduced. If, for example, Russia were to deploy nuclear weapons in Ukraine, and NATO does nothing to react, then the power dynamics in Europe have changed dramatically for decades to come.
Second, we should see more threats in business, as the economy becomes tighter. These threats are likely to be threats around competition and perhaps also around key talent. Understanding and responding to such threats is going to be key, and organisations that play out threat scenarios will be better equipped to deal with them than those that shun the idea of threatening others.
Third, regulation as a threat is going to be tested by the markets. Will the European economy remain strong and regulation work as a deterrence against the economic power of US and China, or will it reduce European economic growth and the startup activity in the EU? If this test shows that the European stance hurts Europe more than it balances power, Europe will be forced to some painful self-examination.
As always, thank you for reading!
Nicklas
Threats represent a peculiar and understudied form of speech act, as pointed out by H I Schiller, here.
The mandatory reading here is essentially everything Thomas Schelling has written. See eg. Schelling TC. The Strategy of Conflict. [Rev. ed.] ed. Cambridge: Harvard University Press; 19801960.
Would Hitler have detonated nuclear bombs if he had them even if it meant the end of the world? We do not know - but it is not hard to imagine the dictator that said that his people deserved to perish also decide that the rest of the world could die with them.
However, this is only true in zero-sum situations, as pointed out by Thomas Schelling:
“The proposition that the enemy’s costs in meeting some threat that we pose, or in responding to some measure we take, should be taken into account in deciding whether the measure is worthwhile, is not hard to defend in a zero-sum context; and one can set Scherer straight by saying that the zero-sum context is all anyone had in mind. But though one can defend the letter of the proposition, it is not so easy to defend the spirit. Should we confine ourselves to the zero-sum context, where the answers are comparatively easy, evading the more interesting contexts where the answers may be harder? Should we defend the cost criterion in a way that increases our vulnerability to Scherer’s charge that quantitative military analysis too readily assumes a zero-sum relation, even while our discussions of strategy are almost exclusively nonzero-sum?”
See Schelling, T. (1985) Choice and Consequence. Harvard University Press.
See Ibid.
See https://en.wikipedia.org/wiki/Dead_Hand
See eg. this study.
There is a rich body of work on throffers. The Wikipedia page offers a good starting point for the interested: https://en.wikipedia.org/wiki/Throffer
The robbery example brought back a memory from college that's always felt like a bonus question in a game theory exam. My rather drunk friend was walking home at night in Berkeley and was held up at gunpoint. He (stupidly) decided it was an empty threat and started to walk away on principle. His robber beat him up with the gun handle - very painfully but not seriously - and then discovered his wallet was empty to begin with and ran away. ¯\_(ツ)_/¯