Abstract
The last decade has seen several advances in the study of deterrence. These advances have sparked some strong disagreements regarding interpretation of the models and what their contributions signify. This paper appraises the discussions from a model theoretic perspective. It is argued that when comparing rivalling models three aspects; (i) target, (ii) criteria and (iii) type of purpose should be taken into account in order to make a proper appraisal. Informed by these aspects it is evident that the three deterrence models analysed address different aspects, in different ways and to different ends. From this perspective, the so-called Perfect Deterrence model must be recognised as a clear advancement in the research field. Model comparison will always be context relative and a plurality of models should be viewed favourably.
Notes
1. For matters of convenience, I adopt the standard of identifying the Classical Deterrence Theory models as CDM, with appropriate index and the models associated with the Perfect Deterrence Theory as PDM.
2. These contributions are part of a debate, which seems to reappear with regular intervals, for the more central contributions see Stephen Walt’s 1999 article “Rigour and Rigour Mortis – Rational Choice and Security Studies” and Frank Zagare’s reply in the same year “All Rigour No Mortis” ,Stephen Quackenbush’s discussion from 2011 entitled “Deterrence Theory – Where Do We Stand?”, Fred Lawson’s critique from 2013 of Quackenbush’s article entitled “Back to the Future in the Study of Deterrence,” and Frank Zagare’s reply to Lawson in Citation1999, Citation2013 “Deterrence Then and Now – There Is No Going Back”.
3. Clarke and Primo discusses five types of purposes a model can have; foundational, structural, generative, explicative and predictive (Clarke and Primo Citation2007). The motivation for examining different models’ intended purpose is that many social scientist only values prediction and tend to value a model on after its ability to predict, when in fact models can have a variety of different types of purposes. While not considered here, the clearest case of one model being superior to another is the predictive model. For example, two models predicting the infection rate of Ebola might have the same target (the rate of how Ebola spreads), and the same criteria (e.g. Simplicity and A-general) and type of purpose, (to predict the spread of Ebola). In this context, it can make sense to discard one model because of its inferiority to predict compared to its rival (Kupferschmidt Citation2014).
4. See, Lawson, Fred. Citation2013. “Back to the Future in the Study of Deterrence” St Anthony’s International Review Vol. 9 No. 1:144–156.
5. A contemporary example of this is the Greek debt negation with the EU, which was described in the media as a Game of Chicken (Guardian 20 Feb. Citation2015).
6. See for instance Paul Milgrom and John Roberts article “Predation, Reputation and Entry Deterrence” (Milgrom and Roberts Citation1982) and David Kreps and Roberts Wilson’s article “Sequential Equilibria” (Kreps and Wilson Citation1982) who further discusses Selten’s idea of Entry Deterrence.