Thinking about deterrence
Deterrence is becoming a less viable strategy. The reasons for this are as follows:
- It is becoming harder to trace attacks back to their origins.
- Non-state actors are becoming capable of conducting attacks of extreme lethality.
- Some non-state actors are not deterred by the thought of losing their lives.
What are the causes of these changes?
- The world is becoming more multi-polar. There are more centers of power. The two-sided division of the Cold War is being replaced by something a lot more complicated.
- Political actors motivated by reasons of aside from nationalism are becoming significant players
- Technological advances are making it easier for smaller states and non-state actors to pursue the development of weapons of mass destruction.
- The distances separating the regions of the world are shrinking. Communications and transportation advances are creating an environment in which people from each region see themselves as more affected by what were formerly very distant governments and cultures.
The idea behind deterrence is that if your enemy strikes you then you will strike back with such ferocity that your enemy, knowing what is in store, will not choose to strike in the first place. In order for deterrence to work the following must hold true:
- You must be capable of striking back with devastating power.
- Your enemy must believe that you are able to strike back with devastating power.
- You must be able to identify who struck in the first place.
- You must be willing to strke back.
- A potential attacker must be pursuing Clausewitzian goals.
The model is pretty simple when there are just two sides, neither side wants to die, retaliatory capability can be protected (which is safe to assume is the case against all of America's current enemies - eg Saddam or Al Qaeda can't track Trident subs) and when one can assume that an attack on one side must have originated with the other side. But in real world there are many sides. The origins of an attack may be hard to determine (who mailed the anthrax?). The types of conflicts that some factions are fighting would leave Clausewitz confused. Also, deterrence does not involve just the two parties facing off with WMD. . For these and other reasons deterrence can fail in many ways. So lets explore why.
Suppose Saddam manages to develop nuclear weapons.
Underlying the debate about Iraq is a debate about deterrence versus preemption. Can we rely on our ability to launch a massive counterstrike to deter the likes of Saddam Hussein from attacking us at some future point with weapons of mass destruction?
The argument for deterrence sounds suspiciously like an argument for fighting a previous war. All too often in history Generals have responded to the threat of a new war by preparing to fight the last war. The last war in this case is the Cold War. It is important to distinguish how the last war differs from the present one. So lets contrast the present conflict with the Cold War.
Characteristics of the Cold War:
- There were few separate enemies and the enemies were (with the exception of traitors in our ranks) easily identified.
- Many of the lesser enemy states were clients of larger enemy states and were constrained by the larger enemy states. So, for instance, Cuba and North Korea had limits placed on their actions by the USSR and China.
- The Soviet side was under a single very firm hierarchical command authority.
- Major attacks were easily traceable to their point of origin. Even if a city had been nuked by a smuggled in bomb the USSR was clearly candidate number one for the origin of the attack. Enemy stategists saw little advantage to using that style of attack.
- Both sides fought for classical Clausewitzian political goals.
Characteristics of the present conflict:
- There are multiple states (some even lacking the characteristics of nation-states) which contribute to the terrorist threat that are run quite independently of each other.
- Not only are client states not well restrained by their sponsors (eg Pakistan didn't do a good job of restraining the Taliban) but even within the client states the lines of command are not easily identified or faithfully obeyed. The ISI in Pakistan may well be operating in ways that conflict with Musharraf's orders to it. Bin Laden was not under the firm control of the Taliban. Saudi Arabia has factions that are working at cross purposes with no strong central figure to restrain them.
- Since some of the states have agencies that are not under firm control leaders of those states the leaders can claim a lack of responsibility for what those agencies do and can try to do that even when the agencies are operating at the behest of the leaders. This is a source of plausible deniability that is hard to confirm or disprove.
- There ar many more private figures and organizations operating outside of the control of nation-states.
- Attacks will not always be traceable to their point of origin.
- An attack may not even be clearly identifiable as an attack. For example, suppose mad cow disease broke out in Wisconsin. We can not assume that we'd be able to determine whether the outbreak was accidental or intentional.
- Some factions of our enemies do not fight for classical Clausewitzian goals.
It is extremely important to emphasise one point: Attacks will not always be traceable to their point of origin. Suppose you are leader of country A. You hate both country B and county C and yet B and C do not like or trust each other (Iraq, the US, and Iran fit A, B, and C very nicely though there are other real world combinations that do as well). Well, how can you win double bonus points in this Hobbesian world? Attack A using a method that makes it seem as if the attack comes from C. A then retaliates against C and you sit back watching your enemies do grievous damage to each other. This is one of the nightmare scenarios of deterrence.
The deterrence strategy requires the following:
- That an attacker will know that he will be clearly identifed as the source of an attack.
- That the attacker is pursuing Clausewitzian goals.
- That we are prepared to accept the potential losses should deterrence fail.
- That we are willing to kill large numbers (even millions) of innocents in a counter-attack in the event that deterrence fails.
During the Cold War we did not have a practical alternative to the deterrence strategy. Direct conventional warfare was pointless because neither side would allow itself to lose a conventional war. Each could escalate to a nuclear level with an enormous arsenal of nuclear weapons and delivery vehicles. Both sides faced the prospect of staggering losses in event of conflict. It was easy to conclude that deterrence was the logical strategy to pursue.
The debate about whether to attack Iraq has become a debate about whether to use preemption or deterrence as the fundamental strategy. Supporters of the deterrence strategy cite over 50 years of Cold War history as a precedent for choosing deterrence as the prudent proven strategy. But an appeal to historical precedent only makes sense if the key historical conditions that made a previous strategy wise are matched under the current condtions.
In order for Iraq to be considered a normally deterrable nation it must be argued that:
Cold War Deterrence Did Not Work Perfectly.
- The Soviet Union supported terrorist groups that carried out attacks against Americans and American interests. A notable example of this is Yasser Arafat. He was in Romania enjoying the protection of the Romanian government (itself a client of the USSR) while his group carried out an attack in Sudan in which they killed the US ambassador and his assistant.
Problems with the deterrence strategy:
- Misunderstandings about intentions and actions can happen.
- If it fails millions of Americans or those of other nations could die in an attack.
- We may not be able to identify where to direct a retaliatory strike toward.
- Retaliation would require the killing of millions of people who had no say in the decisions that led to the attack against us.
- The retaliation would certainly not endear the rest of the world to us.
A reliance on the deterrence strategy opens the world up to an especially nasty sort of threat. Country A's leaders could arrange for a strike against Country B (eg smuggle in a nuke to take out a major city or a bioweapons attack against agriculture or against humans) and do it in such a way that the evidence points toward country C. Well, country B might then unfairly retaliate against country C while unknowingly allowing country A to get away with their attack.
Plausible Deniability Is Becoming Easier.
Technological advances are making plausible deniability of attacks easier. The reason is simple: technological advances make it easier to develop and deliver weapons of mass destruction (WMD).
The number of people, amount of money, and average skill level of the people needed to develop WMD has been and will continue to decline as technologyl advances across a large number of fields. Not only are fewer resources required but fewer tools that are unique to developing WMD are required. So fewer flags are raised and fewer tracable trails are left by suspcicious purchases. The fewer resources and people required for development then the smaller the footprint left by the effort and the easier to hide a development effort.
At the same time, the sheer quantity of people and goods traversing the globe and the speed with which this is done makes it harder to track who is doing what. There are more actors on the scene and fewer spend all their lives just within the area around a single village.
Topping it all off the development of smaller nukes and bioweapons makes it easier to conceal and smuggle weapons around the world. The huge illicit trade in narcotics and other smuggled goods helps by providing an infrastructure thru which WMD can more easily be smuggled.
Fewer Resources Needed to develop WMD
- Fewer people.
- On average less skill needed per weapons development team member.
- Less money.
- Rising affluence means countries and groups have the needed money and people.
- Fewer single purpose tools needed.