Robert Northcott

What kind of epidemiological modeling works well? This is determined by the nature of the target: the relevant causal relations are unstable across contexts. I look at two influential examples of modeling from the Covid pandemic. The first is the paper from Imperial College London, which, in March 2020, was influential in persuading the UK government to impose a lockdown. Because it assumes stability, this first example of modeling fails. A different modeling strategy is required, one less ambitious but more effective. This is illustrated by a second paper from Imperial College London, which, in December 2020, first estimated the transmissibility of the Alpha variant.

Reflexivity is, roughly, when studying or theorising about a target itself influences that target. Fragility is, roughly, when causal or other relations are hard to predict, holding only intermittently or fleetingly. Which is more important, methodologically? By going systematically through cases that do and do not feature each of them, I conclude that it is fragility that matters, not reflexivity. In this light, I interpret and extend the claims made about reflexivity in a recent paper by Jessica Laimann. I finish by assessing the benefits and costs of a focus on reflexivity.

It is often claimed that only experiments can support strong causal inferences and therefore they should be privileged in the behavioral sciences. We disagree. Overvaluing experiments results in their overuse both by researchers and decision-makers, and in an underappreciation of their shortcomings. Neglecting other methods often follows. Experiments can suggest whether X causes Y in a specific experimental setting; however, they often fail to elucidate either the mechanisms responsible for an effect, or the strength of an effect in everyday natural settings. In this paper, we consider two overarching issues. First, experiments have important limitations. We highlight problems with: external, construct, statistical conclusion, and internal validity; replicability; and with conceptual issues associated with simple X-causes-Y thinking. Second, quasi-experimental and non-experimental methods are absolutely essential. As well as themselves estimating causal effects, these other methods can provide information and understanding that goes beyond that provided by experiments. A research program progresses best when experiments are not treated as privileged but instead are combined with these other methods.

To succeed, political science usually requires either prediction or contextual historical work. Both of these methods favor explanations that are narrow-scope, applying to only one or a few cases. Because of the difficulty of prediction, the main focus of political science should often be contextual historical work. These epistemological conclusions follow from the ubiquity of causal fragility, under-determination, and noise. They tell against several practices that are widespread in the discipline: wide-scope retrospective testing, such as much large-n statistical work; lack of emphasis on prediction; and resources devoted to ‘pure theory’ divorced from frequent empirical application. I illustrate, via Donatella della Porta’s work on political violence, the important role that is still left for theory. I conclude by assessing the scope for political science to offer policy advice.

Has the rise of data-intensive science, or ‘big data’, revolutionized our ability to predict? Does it imply a new priority for prediction over causal understanding, and a diminished role for theory and human experts? I examine four important cases where prediction is desirable: political elections, the weather, GDP, and the results of interventions suggested by economic experiments. These cases suggest caution. Although big data methods are indeed very useful sometimes, in this paper’s cases they improve predictions either limitedly or not at all, and their prospects of doing so in the future are limited too.

Should we insist on prediction, i.e. on correctly forecasting the future? Or can we rest content with accommodation, i.e. empirical success only with respect to the past? I apply general considerations about this issue to the case of economics. In particular, I examine various ways in which mere accommodation can be sufficient, in order to see whether those ways apply to economics. Two conclusions result. First, an entanglement thesis: the need for prediction is entangled with the methodological role of orthodox economic theory. Second, a conditional predictivism: if we are not committed to orthodox economic theory, then we should demand prediction rather than accommodation – against most current practice.

Pre-emption cases have been taken by almost everyone to imply the unviability of the simple counterfactual theory of causation. Yet there is ample motivation from scientific practice to endorse a simple version of the theory if we can. There is a way in which a simple counterfactual theory, at least if understood contrastively, can be supported even while acknowledging that intuition goes firmly against it in pre-emption cases—or rather, only in some of those cases. For I present several new pre-emption cases in which causal intuition does not go against the counterfactual theory, a fact that has been verified experimentally. I suggest an account of framing effects that can square the circle. Crucially, this account offers hope of theoretical salvation—but only to the counterfactual theory of causation, not to others. Again, there is (admittedly only preliminary) experimental support for this account.

We propose a novel account of the distinction between innate and acquired biological traits: biological traits are innate to the degree that they are caused by factors intrinsic to the organism at the time of its origin; they are acquired to the degree that they are caused by factors extrinsic to the organism. This account borrows from recent work on causation in order to make rigorous the notion of quantitative contributions to traits by different factors in development. We avoid the pitfalls of previous accounts and argue that the distinction between innate and acquired traits is scientifically useful. We therefore address not only previous accounts of innateness but also skeptics about any account. The two are linked, in that a better account of innateness also enables us better to address the skeptics.

Much recent work in neuroscience aims to shed light on whether we have free will. Can it? Can any science? To answer, we need to disentangle different notions of free will, and clarify what we mean by ‘empirical’ and ‘testable’. That done, my main conclusion is, duly interpreted: that free will is not a testable hypothesis. In particular, it is neither verifiable nor falsifiable by empirical evidence. The arguments for this are not a priori but rather are based on a posteriori consideration of the relevant neuroscientific investigations, as well as on standard philosophy of science work on the notion of testability.

Much philosophical attention has been devoted to whether economic models explain, and more generally to how scientific models represent. Yet there is an issue more practically important to economics than either of these, which I label the efficiency question: regardless of how exactly models represent, or of whether their role is explanatory or something else, is current modeling practice an efficient way to achieve these goals – or should research efforts be redirected? In addition to showing how the efficiency question has been relatively neglected, I give two examples of the kind of analysis it requires.

We propose a novel account of the distinction between innate and acquired biological traits: biological traits are innate to the degree that they are caused by factors intrinsic to the organism at the time of its origin; they are acquired to the degree that they are caused by factors extrinsic to the organism. This account borrows from recent work on causation in order to make rigorous the notion of quantitative contributions to traits by different factors in development. We avoid the pitfalls of previous accounts and argue that the distinction between innate and acquired traits is scientifically useful. We therefore address not only previous accounts of innateness but also skeptics about any account. The two are linked, in that a better account of innateness also enables us better to address the skeptics.

Can purely predictive models be useful in investigating causal systems? I argue ‘yes’. Moreover, in many cases not only are they useful, they are essential. The alternative is to stick to models or mechanisms drawn from well-understood theory. But a necessary condition for explanation is empirical success, and in many cases in social and field sciences such success can only be achieved by purely predictive models, not by ones drawn from theory. Alas, the attempt to use theory to achieve explanation or insight without empirical success therefore fails, leaving us with the worst of both worlds – neither prediction nor explanation. Best go with empirical success by any means necessary. I support these methodological claims via case studies of two impressive feats of predictive modelling: opinion polling of political elections, and weather forecasting.

Comparing different causes’ importance, and apportioning responsibility between them, requires making good sense of the notion of partial explanation, that is, of degree of explanation. How much is this subjective, how much objective? If the causes in question are probabilistic, how much is the outcome due to them and how much to simple chance? I formulate the notion of degree of causation, or effect size, relating it to influential recent work in the literature on causation. I examine to what extent mainstream social science methods--both quantitative and qualitative--succeed in establishing effect sizes so understood. The answer turns out to be, roughly: only to some extent. Next, the standard understanding of effect size, even though widespread, still has several underappreciated consequences. I detail some of those. Finally, I discuss the separate issue of explanandum-dependence, which is essential to assessing any cause’s explanatory importance and yet which has been comparatively neglected.

Partial explanations are everywhere. That is, explanations citing causes that explain some but not all of an effect are ubiquitous across science, and these in turn rely on the notion of degree of explanation. I argue that current accounts are seriously deficient. In particular, they do not incorporate adequately the way in which a cause’s explanatory importance varies with choice of explanandum. Using influential recent contrastive theories, I develop quantitative definitions that remedy this lacuna, and relate it to existing measures of degree of causation. Among other things, this reveals the precise role here of chance, as well as bearing on the relation between causal explanation and causation itself

, whereby some causes are deemed more important than others, are ubiquitous in historical studies. Drawing from influential recent work on causation, I develop a definition of causal-explanatory strength. This makes clear exactly which aspects of explanatory weighting are subjective and which objective. It also sheds new light on several traditional issues, showing for instance that: underlying causes need not be more important than proximate ones; several different causes can each be responsible for most of an effect; small causes need not be less important than big ones; and non-additive interactive effects between causes present no particular difficulty. Key Words: causation • explanation • history • interaction • proximate • underlying.

This is a chapter written for a popular audience, in which I use poker as a convenient illustration of probability, determinism and counterfactuals. More originally, I also discuss the roles of rationality versus psychological hunches, and explain why even in principle game theory cannot provide us the panacea of a perfect winning srategy. (N.B. The document I have uploaded here is slightly longer than the abbreviated version that appears in the book, and also differs in a few other minor details.)

We make the case that the Prisoner’s Dilemma, notwithstanding its fame and the quantity of intellectual resources devoted to it, has largely failed to explain any phenomena of social scientific or biological interest. In the heart of the paper we examine in detail a famous purported example of Prisoner’s Dilemma empirical success, namely Axelrod’s analysis of WWI trench warfare, and argue that this success is greatly overstated. Further, we explain why this negative verdict is likely true generally and not just in our case study. We also address some possible defenses of the Prisoner’s Dilemma.

I propose an analysis of harm in terms of causation: harm is when a subject is caused to be worse off. The pay-off from this lies in the details. In particular, importing influential recent work from the causation literature yields a contrastive-counterfactual account. This enables us to incorporate harm's multiple senses into a unified scheme, and to provide that scheme with theoretical ballast. It also enables us to respond effectively to previous criticisms of counterfactual accounts, as well as to sharpen criticisms of rival views

I present a new case in which the Doomsday Argument runs afoul of epistemic intuition much more strongly than before. This leads to a dilemma: in the new case either DA is committed to unacceptable counterintuitiveness and belief in miracles, or else it is irrelevant. I then explore under what conditions DA can escape this dilemma. The discussion turns on several issues that have not been much emphasised in previous work on DA: a concern that I label trumping; the degree of uncertainty about relevant probability estimates; and the exact sequence in which we integrate DA and empirical concerns. I conclude that only given a particular configuration of these factors might DA still be of interest.

Denis Walsh has written a striking new defense in this journal of the statisticalist (i.e., noncausalist) position regarding the forces of evolution. I defend the causalist view against his new objections. I argue that the heart of the issue lies in the nature of nonadditive causation. Detailed consideration of that turns out to defuse Walsh’s ‘description‐dependence’ critique of causalism. Nevertheless, the critique does suggest a basis for reconciliation between the two competing views. *Received December 2009; revised December 2009. †To contact the author, please write to: Department of Philosophy, 599 Lucas Hall, One University Boulevard, University of Missouri, St. Louis, MO 63121; e‐mail: [email protected].

Jonathan Schaffer (2004 ) proposes an ingenious amendment to David Lewis's semantics for counterfactuals. This amendment explicitly invokes the notion of causal independence, thus giving up Lewis's ambitions for a reductive counterfactual account of causation. But in return, it rescues Lewis's semantics from extant counterexamples. I present a new counterexample that defeats even Schaffer's amendment. Further, I argue that a better approach would be to follow the causal modelling literature and evaluate counterfactuals via an explicit postulated causal structure. This alternative approach easily resolves the new counterexample, as well as all the previous ones. Up to now, its perceived drawback relative to Lewis's scheme has been its non-reductiveness. But since the same drawback applies equally to Schaffer's amended scheme, this becomes no longer a point of comparative disadvantage.

Standard statistical measures of strength of association, although pioneered by Pearson deliberately to be acausal, nowadays are routinely used to measure causal efficacy. But their acausal origins have left them ill suited to this latter purpose. I distinguish between two different conceptions of causal efficacy, and argue that: 1) Both conceptions can be useful 2) The statistical measures only attempt to capture the first of them 3) They are not fully successful even at this 4) An alternative definition more squarely based on causal thinking not only captures the second conception, it can also capture the first one better too.