Effects of Causes and

Causes of Effects
A. Philip Dawid1 and Monica Musio2
1
Statistical Laboratory, University of Cambridge, UK; email:
[email protected]
2
Dipartimento di Matematica ed Informatica, Università degli Studi di Cagliari,
arXiv:2104.00119v1 [math.ST] 31 Mar 2021

Italy; email: [email protected]

Xxxx. Xxx. Xxx. Xxx. YYYY. AA:1–28 Keywords

https://doi.org/10.1146/((please add causal Bayesian network, counterfactual, decision theoretic causality,
article doi))
directed acyclic graph, instrumental variable, interval of ambiguity,
Copyright © YYYY by Annual Reviews. potential outcome, probability of causation, statistical causality,
All rights reserved
stochastic causal model, structural causal model, structural equation
model, twin network

Abstract

We describe and contrast two distinct problem areas for statistical

causality: studying the likely effects of an intervention (“effects of
causes”), and studying whether there is a causal link between the
observed exposure and outcome in an individual case (“causes of ef-
fects”). For each of these, we introduce and compare various for-
mal frameworks that have been proposed for that purpose, including
the decision-theoretic approach, structural equations, structural and
stochastic causal models, and potential outcomes. It is argued that
counterfactual concepts are unnecessary for studying effects of causes,
but are needed for analysing causes of effects. They are however subject
to a degree of arbitrariness, which can be reduced, though not in general
eliminated, by taking account of additional structure in the problem.

1
 INTRODUCTION
1. Overview
The enterprise of “statistical causality” has seen much activity in recent years, both in
its foundational and theoretical aspects, and in applications. However it remains rare to
draw the distinction (recognised by Mill (1843)) between two different problem areas within
it: assessing (in individual cases, or in general) the likely effects of applied or considered
interventions—the problem of “effects of causes”, EoC; and assessing, in an individual case,
whether or not an observed outcome was caused by an earlier intervention or exposure—
the problem of “causes of effects”, CoE. Where this distinction is made, it is typically
assumed that both problems can be represented and addressed using a common theoretical
framework, such as the structural causal model of Pearl (2009).
The purpose of the current article is to emphasise the important logical and technical
differences between EoC and CoE problems, and to explore and compare the various ways
in which problems of each kind can be and have been formulated. In particular it is argued
that different tools are appropriate for the two different purposes.
In § 2 we introduce the variety of concerns to be addressed, in the context of a specific
law suit. In Part I we introduce and compare a variety of formalisms that have been
proposed to address “Effects of Causes”. Section 4 briefly summarises some philosophical
and implementational issues. Section 5 introduces, with examples, the problem of inference
in the presence of an instrumental variable, which is then used throughout Part I as a hook
on which to hang the general discussion. Section 6 describes purely probabilistic aspects.
Then § 7 introduces the decision-theoretic approach to EoC, § 8 an approach based on linear
models, § 9 a nonparametric generalisation of that, and § 10 the approach based on potential
outcomes.
We turn to address “Causes of Effects”in Part II, for problems similar to those of § 2.
Section 11 points to the need for counterfactual inference, which can not however totally
resolve the ambiguities inherent in such problems. Two ways of conducting counterfactual
modelling are described in § 12, which can both be subsumed in the potential outcome
approach of § 13. In § 14 we consider how empirical data can be used to inform CoE analysis,
but can not totally resolve the inherent ambiguities. In § 15 we apply this to address the
legal CoE issues of § 2, showing how the basic ambiguity, expressed by interval bounds
on the “probability of causation”, can be refined when we can observe other variables in
the problem. Section 16 indicates just how limited our CoE analyses have been, and what
difficulties might attend further extension.
Our concluding remarks summarise some of the lessons to learned from this review of
the different approaches to EoC and CoE.

2. A law suit
In 2014 a class action (“multidistrict litigation”, MDL) was brought in the United States by
more than three thousand women who sued the pharmaecutical company Pfizer, claiming
that they developed (type 2) diabetes as a result of taking its drug Lipitor (Atorvastatin
Calcium) (Case Report 2015). The plaintiffs identified two “bellwether cases” of women
making such a claim for closer attention.
In order to succeed in such a suit, the plaintiffs would have to demonstrate, in succession,

2 A. P. Dawid and M. Musio

to the Law’s satisfaction, two points (Dawid et al. 2014):

General causation: Can Lipitor cause diabetes?

Specific causation: In the individual cases, did Lipitor cause their diabetes?

The eventual judgment was in favour of Pfizer. It was judged that general causation had
not been established for doses 10mg, 20mg and 40mg of the drug, but could be considered
for the 80mg dose. And with regard to the bellwether cases, it was judged that specific
causation could not be established.
The distinction the court made between the two varieties of causal question, general
and specific, is fundamental, and occurs in many contexts. It has various descriptions.
Philosophers talk of “type” and “token” causation. Legal scholars talk of “group” and
“individual” causation, and have coined the expression G2i (“Group to individual”) for the
task of arguing from one to the other (Faigman et al. 2014). In statistical contexts we may
talk of inference about “the effects of causes” (EoC), and “the causes of effects” (CoE),
which are the designations we mostly use here.

3. Statistical and causal questions

Questions about individual cases can usefully be organised in a fourfold classification. We
exemplify these for the bellwether case of Juanita, who is 55 years old, has a total cholesterol
of 250 mg/dL, LDL of 175 mg/dL, HDL of 46 mg/dL, triglycerides of 142 mg/dL, weighs
176 lbs, and has a body mass index (BMI) of 26.37.

Forecasting: Juanita has started taking a 80mg dose of Lipitor daily. Is she likely to
develop diabetes?
Backcasting: Juanita has developed diabetes. Did she take Lipitor, and if so in what dose
and for how long?
Decision: Juanita is considering whether to take Lipitor, but is worried about developing
diabetes. What should she do?
Attribution: Juanita took Lipitor 80mg daily for 3 years, and developed diabetes. Was
that because she took Lipitor?

While Forecasting and Backcasting are fundamentally purely statistical exercises,

Decision and Attribution can be classified as “causal” questions—the former addressing
“Effects of Causes” (EoC), and the latter, “Causes of Effects” (CoE).

Forecasting. This is an apparently straightforward statistical task, at least conceptually:

we gather high quality data on individuals sufficiently like Juanita, taking the same
treatment, and observe the proportion going on to develop diabetes. In practice this simple
recipe will be complicated by non-random sampling of cases, differences in background
characteristics, difficulties associated with long-term follow-up, censoring by death, etc.,
etc. Handling such complications has been a prime focus of statistical research over many
decades, and though the issues raised are very far from trivial, they raise no new issues of
principle. But we would also need to argue that the proportion, estimated from the data,
of individuals developing diabetes can be identified with Juanita’s “individual risk”. While
this does raise some subtle philosophical issues (Dawid 2017), they can largely be ignored
for practical purposes.

www.annualreviews.org • Effects of Causes and Causes of Effects 3

 Backcasting. This refers to the task of “predicting” uncertain past events on the basis of
later observations. In a statistical context, this is most typically performed by application
of Bayes’s theorem. Suppose we do not know whether or not Juanita took the Lipitor, but,
as above, have estimated the two “forward” forecast probabilities, under each scenario. We
would also need to assign a prior probability to the event that she did, in fact, take the
drug. Bayes’s theorem supplies the machinery for combining these ingredients to produce
the required “backward” probability that she indeed took Lipitor, on the basis of her having
developed diabetes. Although such Bayesian inferences have, from the very beginning, often
been described as estimating the “probabilities of causes”, use of the term “cause” here is
not really appropriate, since even if we can conclude that Juanita had taken Lipitor, that
might not have been the cause of her diabetes.
Applying Bayes’s theorem is not the only way to conduct backcasting. More straight-
forwardly, we could simply collect a sample of individuals sufficiently like Juanita, confine
attention to those who develop diabetes, and use the proportion of these who had taken
Lipitor to estimate the desired probability for Juanita. Indeed, there are circumstances
where this simple approach may be preferable to the Bayesian route (Dawid 1976).

Decision. Forecasting is of fundamental importance in decision analysis. Suppose Juanita

has not yet started taking Lipitor, and is considering whether or not to do so. One of her
concerns is whether she will develop diabetes. She should thus consider, and compare, how
probable this event is under two possible scenarios: that she does, or that she does not,
take the drug. This would requires two separate forecasting exercises, and correspondingly
data from two different sets of individuals, according as they do or do not take Lipitor.
But new difficulties now arise in gathering and using such data. In particular, the
very treatment desired by such an individual might be related to her overall health status,
and thus affect her risk of developing diabetes—even were she not to receive that desired
treatment. In such as case it becomes problematic to disentangle the effects of desire
for treatment and of application of treatment. This is an example of the problem of
“confounding”, which requires careful attention in such cases.

Attribution. Questions of forecasting, backcasting and decision, although beset with

many practical difficulties, can all, in principle at least, be answered directly by means of
probabilities attached to unknown events of interest, probabilities that can be estimated
given suitable data. However, a question of attribution—such as “did taking Lipitor cause
Juanita’s diabetes”—is not so readily resolved. For what is it now that is unknown? We
know that Juanita took Lipitor, and we know that she developed diabetes. There is no
unknown event about which we require inference. Rather, it the relationship between these
events that is uncertain—was it causal, or not? Even to understand what we might mean
by such a question is problematic.
We shall consider how to formalise such questions, and explore just what can be con-
cluded from data about them, in Part II below.

4 A. P. Dawid and M. Musio

Part I
Effects of Causes
4. Introduction
4.1. Causality and agency
Philosophers have debated causality for millennia, and have propounded a large variety
of conceptions and approaches. Statisticians, on the other hand, had traditionally been
reluctant to imbue their inferences with causal meaning. But in recent years much more
attention has been given to what we can now term “statistical causality”. Particularly
influential have been the contributions of Rubin (1974), who promoted a formulation based
on “potential outcomes”, and of Pearl (2009), based on graphical representations.
Implicit in both these approaches is the idea of a cause as an intervention applied to
a system, in line with the “agency” interpretation of causality (Reichenbach 1956, Price
1991, Hausman 1998, Woodward 2003, 2016). A main task for statistical causality is to
make inference about the effects of such interventions—that is, understanding the “effects
of causes”—on the basis of data. When making use of data, it is important to distinguish
between data generated through experimentation and purely observational data.

4.2. Experiment
In an experiment, interventions are made on experimental units according to some known
protocol, often involving randomisation, and their responses measured. To the extent that
the experimental units and interventions can be regarded as representative of future inter-
ventions on new units, it is in principle straightforward to infer what effects those interven-
tions will have in future. “Design and analysis of experiments” is a major enterprise within
modern statistics, involving many subtle and technical considerations, but no special issues
of principle arise.

4.3. Observation
Things are not so straightforward when the data available are purely observational, and the
process whereby treatment interventions were applied to units is not known. For example,
when choosing between two treatments, a doctor may have given one preferentially to those
patients he considers sicker. Then a simple comparison of the outcomes in the two treatment
groups will be misleading, since even if there is no difference between the treatments, a
difference in outcomes may be seen because of the difference in general health of the two
treatment groups. This is the problem of “confounding”, which prevents us from taking the
observational data at face value. In such a case it may or may not be possible to assess,
by more sophisticated means, genuine causal effects, depending on what is observed and
what assumptions can reasonably be made. If we know or can reasonably assume how the
doctor behaved, and have data on the patient characteristics that the doctor used, then we
can make meaningful comparisons and extract causal conclusions; but—in the absence of
further structure or assumptions—this will not be the case if either of the conditions fails.
Much of the modern enterprise of statistical causality is focused on addressing this issue
of extracting causal conclusions from observational data. In order to do so, it will invariably
be necessary to make assumptions, generally untestable in practice, about the relationship

www.annualreviews.org • 5
 between the behaviours of the “idle” observational system, which generates the observed
data, and the same system under a specified intervention—which is what it wanted, but
is not directly observed. Such assumptions are sometimes made explicit, and so open to
reasoned scrutiny and debate, but sometimes they remain implicit and hidden, being taken
for granted without critical examination. The kind of relationships required can typically be
expressed, explicitly or implicitly, as asserting the equality of certain ingredients in both idle
and interventional circumstances. While such invariance properties have sometimes been
taken as the very definition of causality (Bühlmann 2020), they can be applied without any
such philosophical commitment. (Our own philosophical standpoint remains that based on
agency.)
The “do-calculus” (Pearl (2009, §3.4); see also Dawid (2015, §9.7)) applies to problems
that can be modelled by means of a directed acyclic graph, representing both assumed
conditional independence properties of the observational regime, and assumed relationships
between the observational and interventional regimes. For such a case it supplies a com-
plete method for determining whether a causal estimand of interest can be identified from
observational data, and if so how.

5. Instrumental variable
Below we shall introduce, compare and contrast some of the different statistical formalisms
that have been used to model effects of causes. To be concrete, we shall consider, for each
formalism, how it might model an instrumental variable problem (Bowden & Turkington
1984). This involves, in addition to the treatment variable X and response variable Y , a
further observed variable Z (the instrument), and an unobserved variable U —all defined
for individuals in a study or larger population. Typically Z is binary, X and Y are binary
or continuous, and U is multivariate. Note that in this problem it is not possible, without
imposing still further structure, to identify the causal effect of X on Y from observational
data.
We suppose:
(a). U is a set of pre-existing characteristics of the individual
(b). Z is associated with X, but not with U
(c). While X could in principle be assigned externally, in the study it was not
(d). “Exclusion restriction”: Given X, and the individual characteristics U , the response
Y is unaffected by Z. (This vague requirement will be clarified further below).
Example 1 Encouragement trial
In an encouragement trial (Holland 1988), students are randomly assigned to receive encour-
agement to study. However, a student may not or may not respond to the encouragement.
Here Z is a binary assignment indicator, taking value 1 for encouragement, 0 for no encour-
agement; X is the number of hours the student actually studies; Y is the student’s score
in the final test; and U comprises individual characteristics of the student, that may affect
both X and Y ; because of randomisation, Z is independent of U . We are interested in how
a student’s choice of study hours affects their test score. 
Example 2 Incomplete compliance
In a medical trial, each patient is randomly assigned to take either active treatment (Z = 1)
or placebo (Z = 0). However, the patient may not comply with the assignment, so that
the treatment actually taken, X = 1 or 0, may differ from Z. Finally we observe whether

6
the patient recovers (Y = 1) or not. We allow for possible dependence of both X and Y
on further unobserved patient characteristics, U . Again, randomisation ensures that Z is
independent of U . We are principally interested in the effect of taking the treatment on
recovery. 

Example 3 Availability trial

In a variation of Example 2, Z = 1 means the active treatment is made available to the
patient, Z = 0 that it is not; X = 1 if the treatment is taken, X = 0 if not. It is supposed
that if the treatment is unavailable (Z = 0) it can not be taken (X = 0) (though it need
not be taken when it is available). 

Example 4 Mendelian randomisation (Katan 1986)

Low serum cholesterol level (X = 1) is thought to be a risk factor for cancer (Y = 1). Both
serum cholesterol and cancer may be affected by indicators of lifestyle (U ). Possession of the
E2 allele (Z = 1) of the apolipoprotein E (APOE) gene is known to be associated with low
serum cholesterol level: this relationship need not be causal, but may arise because APOE
is in linkage disequilibrium with the actual causative gene. Since “Nature” randomises
the APOE allele at birth, and its level is not thought to affect lifestyle, U should not be
associated with Z. We are interested in whether intervening to raise serum cholesterol could
lower the risk of cancer. 

There are a number of questions we could ask (but not necessarily be able to answer) in
such examples. In Example 2, these might include:

(i). “What is the probability of recovery for a patient who is assigned to active treatment
(irrespective of the treatment actually taken)?”
(ii). “What is probability of recovery for a patient who (irrespective of assigned treatment)
in fact took active treatment?”
(iii). “What is the probability that a patient who recovered complied with the assignment?”
(iv). “What is the effect on recovery of assignment to treatment?”
(v). “What is the causal effect of taking the treatment on recovery?”

Questions (i), (ii) and (iii) inhabit the lowest rung, “seeing”, of Pearl’s “ladder of causation”
(Pearl & Mackenzie 2018), the first two being instances of “forecasting”, and the last of
“backcasting”. Questions (iv) and (v) are on the second rung of the ladder, “doing”, being
instances of “decision”. In the sequel we shall mainly be interested in (v). Since Z has
been randomised, we could argue that the “intention to treat” question (iv) is essentially
the same as (i), which can be straightforwardly addressed from the study data on (Z, Y ).
However, (v) can not readily be answered in the same way as (ii) (an “as treated” analysis),
since X has not been randomised, and any observed association between X and Y might
be due to their common dependence on U .

6. Seeing: Conditional independence

We first consider how to express, formally, purely probabilistic properties of the observa-
tional joint distribution of (X, Y, Z, U ). This is all that is required to address forecasting
and backcasting questions such as (i)–(iii). However, it will not be possible to formulate,
let alone solve, causal queries such as (v) in this setting: these live on the second rung,
“doing”.

www.annualreviews.org • 7
 Specifically, (b) implies that Z is independent of U : using the notation of Dawid (1979),
we write this as
Z⊥⊥ U. 1.

Here (d) is interpreted as asserting the probabilistic independence of Y from Z, conditional

on X and U :
Y⊥ ⊥ Z | (X, U ). 2.

6.1. Graphical representation

Z U

Y
Figure 1
Instrumental variable: seeing

It is often convenient to display such conditional independence properties by means of a

directed acyclic graph (DAG). Each node in the DAG represents a variable in the problem,
and missing arrows represent assumed properties of conditional independence in their joint
distribution—see Dawid (2015, Section 6) for full details. The DAG is a partial description,
displaying only qualitative aspects, of the joint distribution.
The DAG representing (1) and (2) looks like Figure 1 (the dotted outline of U is non-
essential, merely a reminder that U is unobserved). The absence of an arrow between Z
and U represents their independence, (1), while the missing arrow from Z to Y represents
their conditional independence, given the “parents” of Y , namely X and U , (2). In general
in a DAG representation, any variable is conditionally independent of its non-descendants,
given its parents. Further conditional independence properties implied by these can be
read off the DAG, using the d-separation (Verma & Pearl 1990) or equivalent moralisation
(Lauritzen et al. 1990) criteria, as described in Dawid (2015).
The qualitative DAG representation of a joint distribution can be expanded to a full
quantitative description by specifying, for each variable, its conditional distribution given
its parents in the DAG. (This would be required to encode the condition in (b) that X is
not independent of Z). Elegant algorithms exist, taking advantage of the DAG structure,
for streamlining quantitative computation of joint and conditional probabilities (Cowell
et al. 1999). Such probabilities are what is needed to address questions of forecasting and
backcasting.

7. Doing: Decision-theoretic causality

The decision-theoretic (DT) approach to causality has been described in this journal (Dawid
2015); its foundational underpinnings are examined in Dawid (2021a).

8
 We have several regimes of interest. For each possible value x of X we have an inter-
ventional regime, where treatment value x is forced on an individual (that is, X is “set”
to x, which we notate as X ← x). We also have an “idle” regime, in which the treatment
X is merely observed, and any value may occur. It is helpful to introduce a non-stochastic
“regime indicator” variable FX , where FX = x labels the interventional regime with X ← x,
and FX = ∅ labels the idle regime. The response variable Y may have different distribu-
tions in the different regimes. The object of causal inference will usually be some contrast
between the response distributions in the various interventional regimes—this is what is
required to address the decision problem of choosing which value to set X to. For exam-
ple, when X is binary, interest typically centres on the difference in the expected response
between the two interventional regimes, E(Y | FX = 1) − E(Y | FX = 0), which is termed
the “average causal effect”, ACE.
But in the cases to be considered there will be no data available directly relevant to
the interventional settings of interest, and we shall want to make use of observational data
collected under the idle regime, FX = ∅, to make inferences about what would happen in
interventional settings. This may or may not be possible. At the least, it will be necessary
to make, and justify, relationships between the idle and the interventional settings. DT
studies when and how such relationships can be used to support causal inference from
observational data.
For example, we might be willing to assume, in addition to (1) and (2), that, no matter
whether X is merely observed (FX = ∅), or is set by external intervention (FX = 0 or 1),
the following ingredients will be the same:

(e). the distribution of Z

(f). the distribution of U , with U independent of Z
(g). the conditional distribution of Y given (Z, X, U ) (which would then in all cases depend
only on (X, U ), since this is so under regime FX = ∅, by (d).)
These properties do not follow logically from (1) and (2) (Dawid 2021b), and if they are to
be applied they need additional argument, such as described in Dawid (2021a).
We can interpret assumptions (e)–(g) as conditional independence properties:

Z ⊥
⊥ FX 3.
U ⊥
⊥ (FX , Z) 4.
Y ⊥
⊥ (FX , Z) | (X, U ). 5.

Even though FX is a non-stochastic indicator of regime (observational/interventional), these

intuitively meaningful “extended conditional independence” expressions can be manipulated
essentially just as if FX were a random variable (Constantinou & Dawid 2017).
In this approach, the conditions (e)–(g), or equivalently (3)–(5), which relate to be-
haviour under possible intervention at X, are the full “causal ingredients” of our model.

7.1. Graphical representation

We can augment Figure 1 with an additional node for FX (square to indicate it is non-
stochastic). We obtain Figure 2. This DAG represents (in exactly the same way as before)
the assumed conditional independence assumptions (1)–(5), which fully embody our causal
assumptions. (The dashed arrows from Z and U to X are there to indicate that they
are absent under an interventional regime FX = x, since then we have X = x, trivially

www.annualreviews.org • 9
 Z U

FX X

Y
Figure 2
Instrumental variable: doing

independent of (Z, U )). Note in particular that the arrow from Z to X in Figure 2 does
not encode a causal effect of Z on X, since (e)–(g) are fully consistent with cases, such as
Example 4, where Z and X are merely associated (Dawid 2010, §10).

7.2. Causal Bayesian Network

Pearl (2009) uses the same causal semantics as described above to construct what he terms
a causal Bayesian network (CBN). The difference is that he would normally consider the
possibility of intervention on every observable variable—which in our case would mean
adding further intervention indicator nodes FZ , FY to Figure 2, parents, respectively, of Z
and Y . In such a case the presence of all the intervention nodes is usually taken for granted,
and omitted from the augmented DAG—so rendering it visually indistinguishable from an
unaugmented DAG, here Figure 1. However there are clear advantages to retaining explicit
intervention nodes in the figure:

(i). This eliminates the possibility of confusion between rung 1 (seeing) and rung 2 (doing)
interpretations of apparently identical DAGs.
(ii). The “causal” links assumed between regimes are fully represented by d-separation
properties of the augmented DAG.
(iii). It will (as above) often be appropriate to consider interventions on only some of the
variables. In particular, there will then be no need to impose the additional cross-
regime causal constraints associated with further, inessential, intervention indicators.

7.3. Estimation
Even after assuming links, as above, between the observational and interventional regimes,
it does not follow that we have enough structure to enable us to use the observational data
to estimate, say, the causal effect, ACE, of X on Y . And indeed, in this example, further
structure must be imposed to support such causal inference. For instance, in Example 1 we
might require that Y has a linear regression on (X, U ) (this being the same in all regimes,
by (5)):
E(Y | X, U, FX ) = W + βX 6.

where W is a function of U . Since then E(Y | FX = x) = w0 + βx, where w0 = E(W ),

β has a clear causal interpretation. Also, restricting attention to the observational regime
FX = ∅, we obtain E(Y | Z) = E{E(Y | X, U, Z) | Z} = E{E(Y | X, U ) | Z}, by (5),

10
= w0 + βE(X | Z), by (4). This implies that we can estimate β, from the observational
data, as the ratio of the coefficients of Z in the sample linear regressions of Y on Z and of
X on Z.
In cases such as Example 2 with binary X, (6) is equivalent to

SCE(u) = β 7.

a constant for all u, where SCE(u) = E(Y | U = u, FX = 1) − E(Y | U = u, FX = 0) is the

specific causal effect of X on Y , relevant to the subpopulation having U = u. That is to
say, the specific causal effect is required to be non-random, the same in all subpopulations.
Then ACE = E{SCE(U )} = β also, and so is estimable as above. Alternatively, when all
variables are binary, without making any modelling assumptions we can determine bounds
on ACE from the data (Balke & Pearl 1997, Dawid 2003).

8. Linear structural equation model (SEM)

Linear structural equation modelling, closely related to path analysis (Wright 1921), is
perhaps the earliest approach to instrumental variable problems—and much else besides.
It can be considered as an extension of linear regression modelling. In the context of the
encouragement trial of Example 1, we might express the relationship between Z, X, Y by
the pair of regression-like equations:

X = α0 + α1 Z + UX 8.
Y = β0 + β1 X + UY . 9.

(Such a system would often be completed with a further equation for Z, which here would
simply be Z = UZ . However we omit this on account of its triviality). Here UX , UY are
zero-mean “residual error” terms. In this problem it would be assumed that UX and UY are
uncorrelated with Z, but not necessarily with each other. The absence of Z in (9) embodies
the “exclusion restriction”.
This model can be rendered graphically as in Figure 3. This may be compared with

UX UY

Z X Y
Figure 3
Structural equation graph

Figure 1, identifying U = (UX , UY ).

As discussed by Pearl (2009, §5.1.2), the intended interpretation—in particular, the
causal interpretation—of SEM has often been unclear. Pearl’s suggestion is as follows.
(i). In the system (8)–(9), X is functionally determined by (Z, UX ), and Y is functionally
determined by (X, UY ). So we can solve for (X, Y ) in terms of (Z, UX , UY ). If
we have a joint distribution for (Z, UX , UY ), this determines a joint distribution for
(Z, X, Y )—which can be regarded as representing the undisturbed system.

www.annualreviews.org • 11
 (ii). If, alternatively, we intervene to set X to x, it is assumed that we can replace (8)
by X = x, but retain (9) essentially as is, so that Y = β0 + β1 x + UY , with the
distribution of UY unchanged.
This approach gives a causal semantics to a SEM, relating the observational regime
with possible interventional regimes. As with any such assumed relationship, it is not to be
taken for granted, but argued for in the context of each particular problem. When we can
assume (ii), β1 has a clear causal interpretation, being the rate of change of E(Y | X ← x)
with respect to the value set, x. However, since UY is correlated with UX , and hence
in general with X, β will not be the coefficient of X in the observational regression of
Y on X in (i), so can not be identified from that. Instead we can argue that, since UX
is uncorrelated with Z, E(X | Z) = α0 + α1 Z; and, since UY is uncorrelated with Z,
E(Y | Z) = β0 + β1 E(X | Z) = β0 + α0 β1 + α1 β1 Z. It again follows, as in § 7.3, that β1 can
be identified as the ratio of the coefficients of Z in the observational regressions of Y on Z
and X on Z.

9. Structural Causal Model (SCM)

We can generalise the system (8)–(9) by dropping the linearity requirement, yielding

X = fX (Z, UX ) 10.
Y = fY (X, UY ) 11.

where fX , fY are specified general functions of their arguments, and (UX , UY ) have a
specified joint distribution, typically not being independent of each other, but being jointly
independent of Z. The absence of Z as an argument of fY embodies the exclusion restriction.
Pearl & Mackenzie (2018) refer to such a nonparametric structural equation model as a
Structural Causal Model (SCM), and we use this designation in the sequel.
Again, this system can be represented graphically by Figure 3. But with no loss of
generality we can use U instead of the pair (Ux , UY ), and write the system as

X = fX (Z, U ) 12.
Y = fY (X, U ) 13.

with U independent of Z. This system again determines a joint observational distribution for
(X, Y, Z), which is represented by Figure 1. We might again imbue this structural equation
system with causal semantics (whose relevance in a real life context will need justification):
assume that, under an intervention X ← x, we can replace (12) by X = x, and (13)
by Y = fY (x, U ), where U is supposed to retain its original distribution. The extended
structure is then again represented by Figure 2. In particular, Y will be independent of Z in
an interventional regime, where the dotted arrows in Figure 2 are absent. In contrast, in the
observational regime, the distribution of Y , given X = x, Z = z, is that of Y = fY (x, U )
given fX (z, U ) = x; because of this conditioning, the value of Z will typically make a
difference to the distribution of Y .
In using Figure 1 (and, implicitly or explicitly, Figure 2) as representations of the SCM
system (12)–(13), we are supplying these figures with yet another semantic interpretation,
where the dependence of X and Y on their parents is taken as deterministic, not stochastic.
This is to be contrasted with the CBN interpretation of § 7.2, in which all relationships are
allowed to be stochastic.

12
 It can be be shown that, by suitable choice for the distribution of its U (which distri-
bution is, however, not uniquely determined), the SCM model can fully reproduce the joint
distribution of (X, Y, Z), in all regimes, implied by a given fully stochastic DT model. This
property holds in general for any problem represented by a DAG. For identifying effects
of causes we gain nothing by replacing a stochastic DT model with a deterministic SCM
model.1 In particular, we again can not identify the causal effect of X on Y without further
assumptions, such as linearity in (13), as for (9) (with UY some function of U ).

9.1. A comment
If, in a SEM or SCM, U is regarded as a persistent attribute of an individual, the assumed
determinism would mean that we would get the same output each time we applied the
same intervention to that individual. That would be an unreasonable assumption in most
contexts. Consequently we should normally consider U as also incorporating information
specific to the occasion of application (including, perhaps, “random error”), varying from
occasion to occasion. Nevertheless, assuming the distribution of U does not change, average
causal effects will still be constant, and so be meaningful, across occasions.

10. Potential outcomes

In the potential outcomes (PO) approach to statistical causality (Rubin 1974), for each
possible value x of the treatment X, we conceive of a version Y (x) of the outcome variable
Y , all of these versions co-existing, even before application of treatment. It is supposed
that Y (x) (or, to be more explicit, Y (X = x)) is the outcome that would be observed in
the interventional regime FX = x. Typically it is further assumed (“consistency”) that in
the idle regime FX = ∅ also, whenever X = x, the outcome will be Y = Y (x) (which is why
we don’t distinguish between Y (X = x) and Y (X ← x)). Consistency is required to relate
the observational and interventional regimes.
In the special but common and important case of binary X,2 intervening to set X to
1 would reveal the value of Y (1), while Y (0) would remain unobserved; and similarly on
interchanging 1 and 0. In this approach, the single response Y is replaced by a bivariate
quantity Y = (Y (0), Y (1)), which must thus be endowed with a bivariate distribution. The
fundamental causal contrast, comparing the effects of the two interventions, is considered
to be the individual causal effect, ICE = Y (1) − Y (0). However, direct inference about
ICE is complicated by the fact (termed “the fundamental problem of causal inference” by
Holland (1986)) that, because it is logically impossible to intervene on the same individual
in two mutually exclusive ways simultaneously, we can never observe ICE, or estimate its
distribution. For this reason, it is customary to divert attention to the expected individual
causal effect, E(ICE). By linearity of expectation, this is E{Y (1)} − E{Y (0)}, each term of
which involves only one intervention. This then is the PO version of ACE, as introduced in
§ 7, with essentially the same interpretation. Note however that there is no analogue of ICE
in the DT approach; neither is there any DT analogue of, say, var(ICE), which involves the

1 See however § 10.1, and especially § 10.1.1, for further discussion of this point. Also, while

Balke & Pearl (1997) make essential use of the deterministic functional relationships in the SCM
to derive estimable bounds on ACE in the case of binary variables, Dawid (2003) showed how the
same bounds can be obtained from the purely stochastic DT model.
2 Of course, similar considerations apply more generally.

www.annualreviews.org • 13
 correlation between Y (0) and Y (1)—a correlation that can never be estimated, on account
of the fundamental problem of causal inference.
If we start with a SCM representation of a system, we can use it to construct associated
potential outcomes. For example, starting from (13), just define, for each z, X(Z = z) =
fX (z, U ), and for each x, Y (X = x) = fY (x, U ). Under the SCM causal semantics, X(Z =
z) = fX (z, U ) is assumed to supply the value of Y , when X = x, whether or not there are
interventions at X or anywhere else in the system (except at Y itselvf); this corresponds
to the PO consistency property. This construction of POs makes them all functions of U ,
whose distribution thus generates a joint distribution for all POs.
Typically, however, a PO analysis would not make explicit use of an exogenous variable
such as U , and might not want to require that there exist any real-world variable or set of
variables U with the properties assumed in § 9 (for example, that Y is fully determined by
X and U ). Instead, one starts by introducing, as primitives, jointly distributed stochastic
potential outcomes, X(Z = z), Y (X = x), Y (Z = z), for all possible values of x and z,
and work directly with them. The exclusion restriction now becomes Y (Z = z) = Y {X =
X(Z = z)}.
Introduce now a new variable V , which is simply the collection of all X(Z = z)’s and
Y (X = x)’s, as z and x vary. Then X is fully determined by (Z, V ): when Z = z, we
simply select the relevant element X(Z = z) of V (this being valid in all regimes, by
consistency); similarly Y is determined by (X, V ). We thus recover a formal3 identity with
the SCM of (12) and (13), with V substituting for U —so long as we have V independent
of Z. If we were starting from a SCM, as above, the independence of U and Z, and thus
of V and Z, would be easy to justify, since U represents pre-existing characteristics of the
individual, and Z is randomised. To make a similar argument for V when taking POs as
primitive is more problematic, since V does not correspond to any real-world quantity (in
particular, on account of “the fundamental problem of causal inference”, certain elements
of V , e.g. X(Z = 1) and X(Z = 2), are not simultaneously observable). Nevertheless, the
typical assumption is that it is indeed meaningful to consider the collection V of all possible
potential responses as a pre-existing (albeit unobservable) characteristic of the individual,
and thus argue that V is independent of the randomised variable Z. In this case, we recover
a purely formal identity with an SCM model. Now the linearity condition (6) is equivalent
to Y (X = x) − Y (X = x0 ) = β(x − x0 )—which is thus being required to be non-random.

10.1. A variation
The above specifications can be considered simply as more detailed ways of realising the
CBN structure of § 7, which is more general since in a CBN we need not assume the existence
of potential outcomes, which can not be derived from its stochastic form. And indeed, for
estimating the average causal effect, the extra structure imposed beyond that of a CBN
does not offer any improvement. But a SCM or PO approach does allow us to formulate,
and purports to solve, other causal questions. We consider one such in the context of
Example 2, where Z, X, Y , are all binary (Imbens & Angrist 1994, Angrist et al. 1996).
Let X denote the pair (X(Z = 0), X(Z = 1)), and Y the pair (Y (X = 0), (Y (X = 1)).

3 In a genuine SCM, U is regarded as a set of unobserved real-world background variables, with

an appropriate, in principle knowable, distribution, that, together with X, would determine Y . But
is is hard to conceive of such a real-world interpretation of V .

14
We shall assume consistency, the exclusion restriction Y (Z = z) = Y {X = X(Z = z)}, and
Z⊥⊥ (X, Y). It is easy to see that

Y (Z = z) = Y (X = 1)X(Z = z) + Y (X = 0){1 − X(Z = z)}. 14.

We can define the following “individual causal effects”:

ICEZ→X = X(Z = 1) − X(Z = 0) 15.

ICEZ→Y = Y (Z = 1) − Y (Z = 0) 16.
ICEX→Y = Y (X = 1) − Y (X = 0) 17.

and deduce from (14) that

ICEZ→Y = ICEZ→X × ICEX→Y . 18.

We note that, since Z is randomised, ACEZ→X = E(ICEZ→X ) = E(X | Z = 1) − E(X |

Z = 0) is readily estimable from the observational data, and so likewise is ACEZ→Y =
E(Y | Z = 1) − E(Y | Z = 0). However, there is no immediate parallel for ACEX→Y , since
X has not been randomised.
If we could replace each ICE term in (18) by its expectation ACE, we would have

ACEX→Y = ACEZ→Y /ACEZ→X 19.

where the right hand-side of (19) is estimable from the observational data (it is assumed
that Z has a causal effect on X, so that ACEZ→X 6= 0).
When we can assume (7), ICEX→Y = β is non-random, we can take expectations in
(18), and (19) does indeed hold, allowing estimation of β. But more generally X and Y are
not independent of each other (when constructed from a SCM, they both involve the same
variable U ), and so neither are ICEZ→X and ICEX→Y . So we can not just take expectations
of all terms in (18), and (19) is typically not valid.
To make further progress, other assumptions must be imposed, in particular, mono-
tonicity:
X(Z = 1) ≥ X(Z = 0). 20.

That is to say, we do not have any “defiers”, for which both X(Z = 0) = 1 (treatment
would be taken when not assigned) and X(Z = 1) = 0 (treatment would not be taken when
assigned).
Even monotonicity is not sufficient to allow estimation of ACEX→Y . However, it does
allow a new interpretation of the right-hand side of (19). For it implies that the individual
causal effect ICEZ→X of (15) is either 1 or 0. Thus, from (18),

ACEZ→Y = E(ICEX→Y | ICEZ→X = 1) × Pr(ICEZ→X = 1)

= E(ICEX→Y | ICEZ→X = 1) × E(ICEZ→X ).

It follows that
ACEZ→Y /ACEZ→X = E(ICEX→Y | ICEZ→X = 1). 21.

The right-hand side of (21) is termed the “local average treatment effect”, LATE. Under
monotonicity, LATE is estimable from the data, since the left-hand side of (21) is.

www.annualreviews.org • 15
 10.1.1. Critical comments.

(i). Considerations similar to § 9.1 suggest that it would typically be appropriate to regard
potential outcomes, and so individual causal effects, as varying from one occasion to
another, only their expectations remaining constant.
(ii). In general the monotonicity assumption is untestable, since (under the assumptions of
(i) above) X(Z = 1) and X(Z = 0) can not both be observed on the same occasion.
However, it must hold in the case of an availability trial as in Example 3, where
necessarily X(Z = 0) = 0. Another extreme case where it can be inferred is in the
presence of a variable W , a complete mediator between Z and X (so that X(Z) =
X{W (Z)}), where we have empirical evidence that, with probability 1, W (Z = 1) = 1
and X(W = 0) = 0. If X(Z = 0) = 1, then we deduce W (Z = 0) = 1 = W (Z = 1),
and so X(Z = 1) = X(Z = 0) = 1, and we have no defiers.
(iii). LATE is an average causal effect in a subgroup of the population: those for whom
both X(Z = 0) = 0 and X(Z = 1) = 1. These are termed “compliers”, since they
would take the treatment if assigned to do so, and not take it if not assigned (in an
availability setting, they are those who would take the treatment if assigned to do so).
However it is impossible to tell who belongs to this subpopulation by knowing only
what treatment was assigned and what treatment was taken (in an availability trial,
an individual who was assigned treatment and took it must be a complier, but we
still can not tell the status of an individual who was not assigned treatment). Indeed,
assuming as in (i) that X will vary from occasion to occasion, so too will the group
of compliers. So the relevance of LATE in practice is debatable. Even its definition,
relying as it does on a “cross-world” comparison of potential outcomes under both
Z = 1 and Z = 0, can be criticised as essentially metaphysical and unscientific (Dawid
& Didelez 2012).
(iv). In cases such as Example 4 where Z is not directly causal for X, the notation X(Z =
z) is meaningless, and the above analysis can not even get started.

Part II
CAUSES OF EFFECTS
11. Introduction
Let us consider again the initial attribution example: Juanita took Lipitor 80mg daily for
3 years and developed diabetes. Was that because she took Lipitor?
One way of formulating this CoE question is through what the courts sometimes refer to
as the “but for” test: is it the case that, but for her having taking Lipitor, the diabetes would
not have developed? This immediately plunges us into counterfactual considerations. We
know that, in the actual world, the Lipitor was taken and diabetes developed, and are asked
to contrast this with the outcome in a counterfactual world, in which (counter to the known
facts) the Lipitor had not been taken. The problem of course is that the counterfactual
world is, by definition, unobservable, and even its existence—certainly its uniqueness—are
questionable.
Even in deciding on the exact question, choices have to be made. Juanita took Lipitor
80mg daily for 3 years. Did she develop diabetes because she took the 80mg dose (the

16
only one for which the court accepted general causation), rather than 40mg? Did it develop
because she took it for 3 years, rather than 2 years? Each such choice conjures up a different
counterfactual world for comparison with this one. We also have choice over what was the
observed response: that she developed diabetes at some point?; that she developed diabetes
within 1 year of stopping Lipitor? Detailed specification is obviously important in cases
where the response is death: since death is certain, even in a counterfactual world, we can
never say that an individual would never have died, but for some exposure.4
Under the “but for” criterion, “causation” is understood as the case that, in the ap-
propriate counterfactual world, where Juanita did not take the Lipitor (in the same way
that she in fact did), she did not develop diabetes (in the relevant time-frame). This is
appropriate when the response is all or nothing. We can also consider cases with a con-
tinuous response, such as time to death, but then it is not so clear what the focus of our
attention should be. We might ask, for example, does death occur later, in the relevant
counterfactual world, than it actually did in this world (Greenland 1999)?
Even when our variables have been carefully specified, and the relevant counterfactual
question formulated, it remains unclear just how to conceive of and structure the counter-
factual world of interest. Lewis (1973) develops an approach based on the “closest possible
world” to this one, save only for the change to the exposure; but this only shifts, not solves,
the problem. There appears to be an unresolvable ambiguity about our counterfactual
contrast.
Clearly there are deep philosophical problems, as well as technical specification issues,
besetting any approach to formulating a CoE problem. In the sequel we deal only with the
case of binary exposure and outcome variables, denoted by X and Y respectively, assuming
the above specification problems have been addressed. But there will still remain some
ambiguity about the relevant counterfactual world, which will be reflected in ambiguity
about the answer to the CoE question.
In § 12 we introduce two approaches to relating the actual and counterfactual worlds:
SCM and CIM. The former is essentially deterministic, while the latter allows some stochas-
tic elements. However both make assumptions that might be regarded as over-strong, lead-
ing to misleadingly precise answers to the CoE question. In § 13 we show how each of these
models can be reformulated in terms of “potential outcomes”. In § 14 we explain how, tak-
ing full account of real-world data on exposure and outcome, this approach can handle and
quantify the remaining ambiguities, by supplying an appropriate “interval of ambiguity”
for the probability of causation, PC.
We can narrow the interval of ambiguity for an individual case by deeper understanding
of the mechanisms and processes involved (Beyea & Greenland 1999)—even when we can’t
access the specific details of these for the individual case at hand. We develop this theme
in the remaining sections, showing how information about additional variables can tighten
the bounds on PC.

4 Even in EoC cases, specification of the outcome matters. In March 2021, a few cases of blood

clots among individuals who had received the Astra-Zeneca vaccine against Covid-19 were observed,
and concerns were raised about a possible causal connexion, leading to a pause in the vaccine’s roll-
out in some countries. When it was pointed out that the rate of such clots was in fact lower than
in the general population, attention turned to the few cases of a specific rare presentation of the
blood clot, cerebral venous sinus thrombosis, and whether the vaccine could cause that. (For this
outcome too, there was no evidence for causation).

www.annualreviews.org • 17
 12. Counterfactual constructions
12.1. SCM
The approach of Pearl (2009) (see also Pearl (2015), Dawid et al. (2015)) to CoE is based
on SCMs. In the case of Juanita, this would involve the introduction of an unobserved
exogenous “background” variable U , and the assumption that Juanita’s diabetes status Y
is fully determined by her Lipitor status X and U : Y = fY (X, U ) (this requires a conception
of U as comprising all other pre-existing quantities that, together with X, would totally
determine Y —a collection that may not be easy to comprehend, let alone specify). In some
contexts it might be appropriate (“ignorability”) to regard X as independent of U in the
observational regime, as would happen, if, for example, X is generated by a randomising
device. We do not impose this throughout, and will specify where we do assume it.
The same functional relationship Y = fY (X, U ) is assumed to hold (“consistency”)
whether or not X is imposed by external intervention. To this invariance requirement,
familiar from “effects of causes” analysis, we add another, specific to “causes of effects”:
that the value of the background variable U be the same in both the factual world, and in
the counterfactual world that we wish to contrast with this one.
To start with, we assume that the function fY and the joint distribution of (X, U ) are
known. These unrealistic requirements are removed in § 14 below.
In the factual world we have observed X = 1 and Y = 1, i.e. fY (1, U ) = 1. We
can express the resulting uncertainty about the value of U by means of its conditional
distribution, given X = 1, fY (1, U ) = 1 (or, under ignorability, given only fY (1, U ) = 1).
We now turn to consider the counterfactual world. Although U is supposed to be the same
in both worlds (and thus endowed with the above conditional distribution), X and Y need
not be. We introduce “mirror variables”, X 0 and Y 0 , as their counterfactual counterparts.
We retain the general structure across worlds, so that Y 0 = fY (X 0 , U ), both in observational
and in interventional counterfactual regimes.
In the counterfactual world, we now consider the effect of an intervention X 0 ← 0. The
value of Y 0 will be fY (0, U ). Using the previous conditional distribution of U , we obtain
the counterfactual distribution for Y 0 , given the hypothetical intervention X 0 ← 0 and the
factual knowledge X = Y = 1. We can thus evaluate the “probability of causation” as the
probability that, in this distribution, Y 0 = 0. That is,

PC = Pr(Y 0 = 0 | X = 1, Y = 1, X 0 ← 0). 22.

12.2. Stochastic Causal Model (StCM)

A generalisation of the above model, which does not require deterministic functional rela-
tionships, was suggested by Dawid (2000, §12), although it has not been developed in detail.
We again assume that the variable U retains its identity across the parallel worlds, and in-
troduce the mirror variables X 0 , Y 0 . But we now allow the dependence of Y on (X, U ) to be
given by a known conditional probability distribution: this allows a more liberal attitude to
the nature of U , which can be a perfectly normal, in principle observable (though typically
not observed), variable.
We assume that the same stochastic relationship also governs the dependence of Y 0 on
0
(X , U ). We again require consistency to relate observational and interventional regimes:
the value of Y (resp., Y 0 ) would be the same, whether X (resp., X 0 ) arose by intervention
or not. In order to complete the specification, we regard Y 0 and Y as conditionally indepen-

18
dent, given (X, X 0 , U ) (for example, we might consider them as involving “random noise”,
operating independently across worlds).
Having now a joint distribution for all variables in all worlds, we can again compute
PC = Pr(Y 0 = 0 | X = 1, Y = 1, X 0 ← 0). Specifically, by conditional independence,

Pr(Y 0 = 0 | X = 1, Y = 1, X 0 ← 0, U ) = Pr(Y 0 = 0 | X 0 ← 0, U )
= Pr(Y 0 = 0 | X 0 = 0, U )
= Pr(Y = 0 | X = 0, U ).

Then PC is the expectation of this, in the conditional distribution of U given X = 1, Y =

1, X 0 ← 0. But setting X 0 does not affect the joint distribution of (X, Y, U ), so we just
compute, from Bayes’s theorem, the posterior density

p(u | X = 1, Y = 1) ∝ Pr(X = 1 | u) Pr(Y = 1 | X = 1, u) p(u),

where p(u) is the prior density of U (and the first term on the right-hand side can be omitted
under ignorability). Finally,
R
Pr(Y = 0 | X = 0, u) Pr(X = 1 | u) Pr(Y = 1 | X = 1, u) p(u) du
PC = R . 23.
Pr(X = 1 | u) Pr(Y = 1 | X = 1, u) p(u) du

12.3. Twin network

In both the SCM and StCM approaches, the required computation can be automated by
building a “twin network” representation of the problem (Pearl 2009, §7.1.4), as in Figure 4,
and making use of probability propogation algorithms (Cowell et al. 1999) as implemented
in software systems such as Hugin.5 This is useful in more complex problems with more
variables.

FX’ X’ Y’

X Y
Figure 4
Twin network. Under ignorability, the arrow from U to X can be removed. That from U to X 0 is
absent because we are considering X 0 as set by external intervention, taking no account of U .

The factual information X = 1, Y = 1 and the counterfactual intervention FX 0 = 0

(implying X 0 = 0) are entered at the relevant nodes, and propagated through the network
to obtain the appropriate conditional distribution for Y 0 .

5 https://www.hugin.com

www.annualreviews.org • 19
 12.4. More general StCMs
The StCM approach involves nominating some of the variables in a problem as shared
across worlds, while the others are allowed to differ. The associated twin network will
have a single copy of the shared variables and mirror copies of the others, with the original
DAG replicated and stitched together through the shared variables. As discussed in Dawid
(2000), the choice of which variables are to be regarded as shared6 is a matter of imagination
rather than science, and should relate to the specific problem of interest—there can be no
context-free right answer. For example, there have been law suits by various states against
tobacco companies, claiming that if they had publicised their knowledge of the dangers
of smoking when they first knew of them, many lives could have been saved. Damages
are sought for the additional costs placed on health services—meaning the excess cost in
the actual world, over that of an imagined world in which they had made their knowledge
public. But how should we imagine that world? One could reasonably argue that, in such a
world, by giving up smoking, people would have lived longer than they actually did. Then
the actual (non-)actions of the tobacco companies might well have increased the cost to the
health services. But what seems to be required for the case at hand is to imagine a world
where people had the same lifetimes, but were healthier, i.e., to regard lifetimes as shared
across parallel worlds—and this even though lifetimes can be considered an effect of the
companies’ decisions. It is not clear how such considerations could be accommodated in a
SCM.

13. Potential outcomes

As observed in § 10, the SCM approach produces implied potential outcomes, Y (1) =
fY (1, U ), Y (0) = fY (0, U ). The pair Y = (Y (1), Y (0)) is a function of U , with a bi-
variate distribution induced by that of U . And in fact Y is all that needs to be retained of
U to fully describe the problem: we can replace U by Y, with the functional dependence of
Y on (X, Y) given simply by Y = Y (X). The problem can thus be more concisely expressed
in terms of (X, Y), these having a joint distribution. Then (22) becomes

PC = Pr(Y (0) = 0 | X = 1, Y (1) = 1). 24.

Note that under ignorability X is independent of Y (this is indeed the very definition of
ignorability in the PO framework), and then the conditioning on X = 1 in (24) can be
removed.
For the StCM approach we proceed as follows. The stochastic dependence of Y on (X, U )
can be modelled by introducing a further unobserved “noise” variable V , independent of
(X, U ), and representing Y = fY (X, U, V ) for a suitable function fY . This can be done
in many ways. One possible way uses the probability integral transformation: if Y is a
univariate variable whose conditional distribution function Fx,u (y), given X = x, U = u, is
−1
strictly increasing, take V to be uniform on [0, 1], and fY (x, u, v) = Fx,u (v).
There will be a counterfactual mirror V of V , with Y = fY (X, U, V 0 ). We now define
0 0

potential outcomes Y (1) = fY (1, U, V ), Y (0) = fY (0, U, V 0 ), having a joint distribution

induced by that of (U, V, V 0 ). Although the variables so constructed will depend on the

6 Mackie (1980) refers to these as the “causal field”, and notes the rôle of our own choice in its

specification.

20
specific choices made for the noise variable V and the function fY , it is easy to see that
their joint distribution will in all cases be that of (Y, Y 0 ), given interventions X ← 1,
X 0 ← 0. And since X = 1 ⇒ Y = Y (1), etc., (24) again holds.
We thus see that, in all cases, we can ignore the finer details, and represent the problem
by means of a joint distribution for (X, Y), with PC given by (24).

14. Empirical information

We have so far supposed that the full probabilistic structure of the model, with its variables
U, X, Y , is known. In a StCM, we can take U to be a specified potentially observable vari-
able, and then this assumption is not unreasonable. However it is typically implausible for
a SCM, where, in order to achieve the required deterministic dependence of Y on (X, U ),
we would have to conceive of a fantastically rich U . Alternatively, we can re-express the
problem in terms of the pair Y of potential responses, with a joint distribution for (X, Y).
Without making further assumptions on the originating SCM or StCM, there are no con-
straints on this joint distribution (other than independence of X and Y under ignorability).
We can however gather empirical data to constrain it, and thus hope to estimate PC by
(24).
In the sequel we proceed on this basis, and consider what can indeed be estimated. We
initially assume that we can only observe X and Y , in interventional and/or observational
circumstances. We can estimate Pr{Y (x) = 1} = Pr(Y = 1 | X ← x) from interventional
studies. When we can not assume ignorability, we can also estimate, from observational
data, Pr{Y (1) = 1 | X = 1} = Pr(Y = 1 | X = 1) (by consistency), and Pr{Y (0) = 1 |
X = 0} = Pr(Y = 1 | X = 0), as well as the marginal distribution of X. For this general
case it might initially seem problematic to estimate, say, Pr{Y (1) = 1 | X = 0}, since this
involves non-coexisting worlds, one with X = 1 and the other with X = 0; but we can in
fact solve for it, using Pr{Y (1) = 1} = Pr{Y (1) = 1 | X = 1} Pr(X = 1) + Pr{Y (1) = 1 |
X = 0} Pr(X = 0), where all other terms are estimable.
We can thus estimate the bivariate distribution of (X, Y (1)), and likewise that of
(X, Y (0)). However, the full trivariate distribution of (X, Y (1), Y (0)) is not estimable:
since we can never observe both Y (0) and Y (1) simultaneously, no data can tell us directly
about the dependence between Y (0) and Y (1), either marginally or conditionally on X.
And since (24) requires such information, typically PC is not identifiable from empirical
data.
Nevertheless, the estimable bivariate distributions do impose constraints on the possible
values of PC. Moreover, these constraints can often be tightened still further when we
can observe other, related, variables in the problem. We now turn to investigate such
constraints, in a number of contexts.

15. Analysis of the probability of causation

Let us consider the initial attribution example: Juanita took Lipitor 80mg daily for 3 years
(X = 1) and developed diabetes (Y = 1). Was that because she took Lipitor? We wish to
address this question and assess the probability of causation, PC, for Juanita’s case, using
data collected on other individuals. To this end we assume
Exchangeability Juanita is similar to the population from which probabilities have been
computed, so that those probabilities apply to her.

www.annualreviews.org • 21
 Y (1) = 0 Y (1) = 1
1 1 1
Y (0) = 0 2
(1 − ρ − ξ) 2
(ξ + τ) 2
(1 + τ − ρ)
1 1 1
Y (0) = 1 2
(ξ − τ ) 2
(1 + ρ − ξ) 2
(1 − τ + ρ)
1 1
2
(1 − τ − ρ) 2
(1 +τ + ρ) 1
Table 1 Joint distribution of Y (0) and Y (1)

Exchangeability may require restriction of the data considered to individuals deemed suffi-
ciently like Juanita.
Except where relaxed in § 15.3 below, we shall also assume

Ignorability The fact that Juanita chose to take the drug is not informative about her
response to it, either factually or counterfactually. Formally, we require independence,
X⊥ ⊥ Y, between X and the pair of potential responses Y = (Y (1), Y (0)).

Ignorability is a strong assumption, and will often be inappropriate. When it can be as-
sumed, we can use data from either experimental or observational studies; otherwise we
need data from both of these.
Under ignorability, the target (24) becomes

PC = Pr(Y (0) = 0 | Y (1) = 1). 25.

We proceed to assess this using the general potential outcome framework of § 13 and § 14,
where no assumptions are imposed on the joint distribution of Y (0) and Y (1) beyond those
that can be informed by the empirical data. Further details may be found in Dawid et al.
(2017), Dawid & Musio (2021).

15.1. Basic inequalities

Suppose we have access to (observational or experimental) data, supplying values for

Pr{Y (x) = y} = Pr(Y = y | X = x) (x = 0, 1 y = 0, 1). 26.

Define

τ := Pr(Y = 1 | X = 1) − Pr(Y = 1 | X = 0)
ρ := Pr(Y = 1 | X = 1) − Pr(Y = 0 | X = 0).

The joint distribution of (Y (0), Y (1)) must have the form of Table 1, where the marginal
probabilities are given by (26), re-expressed in terms of τ and ρ, and where the unidentified
“slack” quantity ξ embodies the residual ambiguity in the distribution. For all the entries
of Table 1 to be non-negative, we require

|τ | ≤ ξ ≤ 1 − |ρ|. 27.

The probability of causation (25) is

ξ+τ
PC = . 28.
1+τ +ρ
On using (27), we obtain the following interval bounds for PC:
   
2τ 1+τ −ρ
l := max 0, ≤ PC ≤ min 1, =: u, 29.
1+τ +ρ 1+τ +ρ

22
or equivalently
   
1 Pr(Y = 0 | X = 0)
l = max 0, 1 − ≤ PC ≤ min 1, = u, 30.
RR Pr(Y = 1 | X = 1)

where
Pr(Y = 1 | X = 1)
RR = 31.
Pr(Y = 1 | X = 0)
is the risk ratio (Robins & Greenland 1989).
In the absence of additional information or assumptions, these bounds constitute the
best available inference regarding PC. In particular, RR > 2, “doubling the risk”, implies
that PC > 0.5. In a civil legal case, causality might then be concluded “on the balance of
probabilities”. However, because of the remaining ambiguity, expressed by the inequalities
in (30), finding that RR falls short of 2 does not imply that PC < 0.5.

15.2. Refining the inequalities: Covariates

When we have additional information we may be able to refine our inferences about PC
(Kuroki & Cai 2011).
Thus suppose that we also have information on a covariate S, a pretreatment individual
characteristic that can vary from person to person and can have an effect on both X and
Y . The relevant potential responses are now X := (X(s) : s ∈ S), Y := (Y (s, x) : s ∈
S, x = 0 or 1) and the relationship between potential and actual responses is X = X(S),
Y = Y (S, X). We again assume exchangeability and ignorability, the latter now being
formalised as mutual independence between S, X and Y.
For simplicity we suppose that S is discrete and that we can estimate from the data the
full joint distribution of (S, X, Y ).
In the case that we are also able to measure S for Juanita, say S = s, we can simply
restrict the experimental subjects to those having the same covariate value (who are thus
like Juanita in all relevant respects). The probability of causation is now

PC(s) = Pr (Y (s, 0) = 0 | Y (s, 1) = 1) .

We can bound this just as in (30), but with all probabilities now conditioned on S = s,
obtaining l(s) ≤ PC(s) ≤ u(s).
More interesting is the case in which we don’t observe S for Juanita. We have to
consider what would have been the response if, counterfactually, Juanita’s exposure had
been X = 0. We assume that this is the minimal change made between the factual and the
counterfactual worlds, so that, in particular, there is no change to the value or distribution
of S.
The probability of causation is now:

PC = Pr {Y (S, 0) = 0 | X(S) = 1, Y (S, 1) = 1}

X
= PC(s) × Pr(S = s | X = 1, Y = 1). 32.
s

There are no logical relationships between the distributions of (Y (s, 0), Y (s, 1)) for dif-
ferent values of S. So by independently varying the values taken by the slack variables in
the joint distribution of these potential responses, all the lower bounds l(s) for PC(s) can

www.annualreviews.org • 23
 be achieved simultaneously. This leads to an achievable lower bound for PC:
X
PC ≥ L = l(s) × Pr(S = s | X = 1, Y = 1). 33.
s

We can express

1
L =
Pr(Y = 1 | X = 1)
X
× max {0, Pr(Y = 1 | X = 1, S = s) − Pr(Y = 1 | X = 0, S = s)} × Pr(S = s | X = 1).
s
34.

Similarly we obtain upper bound

1
U = 1−
Pr(R = 1 | E = 1)
X
× max{0, Pr(Y = 1 | X = 1, S = s) − Pr(Y = 0 | X = 0, S = s)} × Pr(S = s | X = 1).
s
35.

We can not compare these bounds directly with those of (30), since when we don’t take
account of S the relation between X and Y is generally non-ignorable: Pr(Y = y | X ←
x) 6= Pr(Y = y | X = x).

15.3. Non-ignorability
Tian & Pearl (2000) analysed this non-ignorable case where we don’t observe S, either for
Juanita or in the external data (still considered exchangeable with Juanita). We now need
both observational and experimental data on X and Y . Tian & Pearl (2000) develop the
following lower bound for PC = Pr{Y (0) = 0 | X = 1, Y (1) = 1}:
 
Pr(Y = 1) − Pr(Y = 1 | X ← 0)
L0 = max 0, . 36.
Pr(X = 1, Y = 1)

Dawid & Musio (2021) show that this can also be derived as a special case of our
expression (34), if we substitute for S the binary variable D = “desired exposure” (Corradi
& Musio 2020). D will be identical with X in an observational context, but need not be so
in an experimental setting, where D may not be observable.
In our case, with access to information on S, we could compute Pr(Y = 1 | X ← 0) by
the “back-door formula”:
X
Pr(Y = 1 | X ← 0) = Pr(Y = 1 | X = 0, S = s) × Pr(S = s), 37.
s

and thus compute L0 of (36). It can be shown (Dawid & Musio 2021) that L0 ≤ L of (34),
with equality if and only if all the conditional risk ratios

Pr(Y = 1 | X = 1, S = s)
(s ∈ S)
Pr(Y = 1 | X = 0, S = s)

24
lie on the same side of 1: knowing, and using, the information about S is at least as good
as ignoring it. Similarly we can show that the upper bound U of (35) does not exceed the
upper bound U 0 derived by Tian & Pearl (2000),
 
0 Pr(Y = 0 | X ← 0) − Pr(X = 0, Y = 0)
U = min 1, ,
Pr(X = 1, Y = 1)

with equality if and only if all the ratios

Pr(Y = 1 | X = 1, S = s)
(s ∈ S)
Pr(Y = 0 | X = 0, S = s)

lie on the same side of 1.

15.4. Mediators
We now consider the case that a third variable M acts a complete mediator in the causal
pathway X → M → Y between the exposure X and the response Y . Again we restrict
to the case that all variables are binary. We introduce the potential value M (x) of M for
X = x, and Y (m), the potential value of Y for M = m, and define M = (M (0), M (1)), Y =
(Y (0), Y (1)). We observe X, M = M (X) and Y = Y (M ). We assume the exchangeability
and the ignorability conditions, the latter expressed as mutual independence between X,
M and Y. This implies the observational conditional independence

Y⊥
⊥X | M 38.

which is a testable implication of our assumptions. We assume we have data supplying

values for Pr(M = m | X = x) and Pr(Y = y | M = m), and compute, by (38),
X
Pr(Y = y | X = x) = Pr(Y = y | M = m) Pr(M = m | X = x), 39.
m

For the case that M is observed in the experimental data, but not for Juanita, Dawid
et al. (2016) showed that this additional information does not change the lower bound l on
PC in (30), but does lower the upper bound, u. Dawid & Musio (2021) extend this analysis
to cases with additional covariates, while Dawid et al. (2019) deal with the case that we
have a complete mediation sequence X = M0 → M1 → · · · → Mn−1 → Mn = Y , and know
the probabilistic structure of each link in the chain; with all, some, or none of the M s being
observed for Juanita.

16. Further CoE problems

We have only dealt here with the case of a single putative cause, understanding causation
in terms of the “but for” criterion. There are many more complex problems that can not be
handled in this way: in particular, the whole field of legal causation has to handle a wide
variety of problems involving multiple competing causes and other concepts of causality
(Hart & Honore 1985, Goldberg 2011). While there have been some interesting statistical
treatments of specific problems, e.g., Cox (I984), it seems fair to say that general philosoph-
ical understandings of causality in such problems have not reached maturity. To the extent
that problems are modelled in formal terms, this often involves a purely deterministic un-
derstanding of causality, which is not easily translated into a stochastic framework. Halpern

www.annualreviews.org • 25
 (2016) makes an interesting attempt to pin down the concept of “the actual cause” using
the SCM framework, but admits that he is unable to reach a fully satisfying conclusion.
There is clearly much ground remaining to be covered in understanding “causes of
effects”, but it is perhaps premature to attempt more detailed statistical treatment before
clearer general principles have emerged.

CONCLUSION
We have presented a thorough account of a number of ways in which the statistical problems
of effects of causes (EoC) and of causes of effects (CoE) have been formulated. Although
most treatments of statistical causality use essentially identical tools to address both these
problem areas, we consider that this is inappropriate. Popular formalisms, such as potential
outcomes and structural causal models, involve deterministic relations, and allow formal
statements concerning two or more parallel worlds simultaneously. While something of this
nature appears unavoidable for CoE considerations, it is unnecessary for EoC analyses,
which can proceed using stochastic models and statistical decision theory. Furthermore,
the use of an inappropriate formal framework brings with it the danger that we treat any
mathematically well-formed formula (such as those describing “individual causal effect” and
“local average treatment effect”) as meaningful, when it may not be.

DISCLOSURE STATEMENT
The authors are not aware of any affiliations, memberships, funding, or financial holdings
that might be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS
The second author was partially supported by the project STAGE of Fondazione di
Sardegna.

LITERATURE CITED
Angrist J, Imbens G, Rubin DB. 1996. Identification of causal effects using instrumental variables.
Journal of the American Statistical Association 91:444–455
Balke AA, Pearl J. 1997. Bounds on treatment effects from studies with imperfect compliance.
Journal of the American Statistical Association 92:1172–6
Beyea J, Greenland S. 1999. The importance of specifying the underlying biologic model in esti-
mating the probability of causation. Health Physics 76:269–274
Bowden RJ, Turkington DA. 1984. Instrumental Variables. Cambridge University Press
Bühlmann P. 2020. Invariance, causality and robustness (with Discussion). Statistical Science
35:404–436
Case Report. 2015. Hempstead v. Pfizer, Inc. (In re Lipitor (Atorvastatin Calcium) Mktg., Sales
Practices & Prods. Liab. Litig. 150 F. Supp. 3d 644 (D.S.C. 2015)
Constantinou P, Dawid AP. 2017. Extended conditional independence and applications in causal
inference. Annals of Statistics 45:2618–2653
Corradi F, Musio M. 2020. Causes of effects via a Bayesian model selection procedure. Journal of
the Royal Statistical Society, Series A 183:1777–1792

26
Cowell RG, Dawid AP, Lauritzen SL, Spiegelhalter DJ. 1999. Probabilistic Networks and Expert
Systems. New York: Springer
Cox LA. I984. Probability of causation and the attributable proportion of risk. Risk Analysis 4:221–
230
Dawid AP. 1976. Properties of diagnostic data distributions. Biometrics 32:647–658
Dawid AP. 1979. Conditional independence in statistical theory (with Discussion). Journal of the
Royal Statistical Society, Series B 41:1–31
Dawid AP. 2000. Causal inference without counterfactuals (with Discussion). Journal of the Amer-
ican Statistical Association 95:407–448
Dawid AP. 2003. Causal inference using influence diagrams: The problem of partial compliance (with
Discussion), In Highly Structured Stochastic Systems, eds. PJ Green, NL Hjort, S Richardson,
pp. 45–81, Oxford University Press
Dawid AP. 2010. Beware of the DAG!, In Proceedings of the NIPS 2008 Workshop on Causality,
eds. I Guyon, D Janzing, B Schölkopf, vol. 6 of Journal of Machine Learning Research Workshop
and Conference Proceedings, pp. 59–86. http://tinyurl.com/33va7tm
Dawid AP. 2015. Statistical causality from a decision-theoretic perspective. Annual Review of Statis-
tics and its Application 2:273–303
Dawid AP. 2017. On individual risk. Synthese 194:3445–3474
Dawid AP. 2021a. Decision-theoretic foundations for statistical causality (with Discussion). Journal
of Causal Inference (in Press) arXiv:2004.12493
Dawid AP. 2021b. The tale wags the DAG. In Probabilistic and Causal Inference: The Works of
Judea Pearl, eds. R Dechter, H Geffner, J Halpern. Association for Computing Machinery. (In
Press).
Dawid AP, Didelez V. 2012. “Imagine a can opener” — The magic of principal stratum analysis.
International Journal of Biostatistics 8. DOI: 10.1515/1557-4679.1391
Dawid AP, Faigman DL, Fienberg SE. 2014. Fitting science into legal contexts: Assessing effects
of causes or causes of effects? (with Discussion and authors’ rejoinder). Sociological Methods and
Research 43:359–421
Dawid AP, Faigman DL, Fienberg SE. 2015. On the causes of effects: Response to Pearl. Sociological
Methods and Research 44:165–174
Dawid AP, Humphreys M, Musio M. 2019. Bounding causes of effects with mediators.
arXiv:1907.00399
Dawid AP, Murtas R, Musio M. 2016. Bounding the probability of causation in mediation analysis,
In Topics on Methodological and Applied Statistical Inference, eds. TD Battista, E Moreno,
W Racugno, pp. 75–84, Springer
Dawid AP, Musio M. 2021. What can group level data tell us about individual causality? In
Statistics in the Public Interest, eds. A Carriquiry, W Eddy, J Tanur. Springer. In Press
Dawid AP, Musio M, Murtas R. 2017. The probability of causation. Law, Probability and Risk
16:163–179
Faigman DL, Monahan J, Slobogin C. 2014. Group to individual (G2i) inference in scientific expert
testimony. University of Chicago Law Review 81:417–480
Goldberg R, ed. 2011. Perspectives on Causation, Oxford. Hart Publishing
Greenland S. 1999. Relation of probability of causation to relative risk and doubling dose: A method-
ologic error that has become a social problem. American Journal of Public Health 89:1166–1169
Halpern JY. 2016. Actual Causality. Cambridge, Massachusetts: MIT Press
Hart HLA, Honore AM. 1985. Causation in the Law. Oxford: Clarendon Press
Hausman D. 1998. Causal Asymmetries. Cambridge: Cambridge University Press
Holland PW. 1986. Statistics and causal inference (with Discussion). Journal of the American
Statistical Association 81:945–970.
Holland PW. 1988. Causal inference, path analysis, and recursive structural equations models.
Sociological Methodology 18:449–484

www.annualreviews.org • 27
 Imbens GW, Angrist J. 1994. Identification and estimation of local average treatment effects. Econo-
metrica 62:467–476
Katan MB. 1986. Apolipoprotein E isoforms, serum cholesterol, and cancer. The Lancet 327:507–508
Kuroki M, Cai Z. 2011. Statistical analysis of ‘probabilities of causation’ using co-variate informa-
tion. Scandinavian Journal of Statistics 38:564–577
Lauritzen SL, Dawid AP, Larsen BN, Leimer HG. 1990. Independence properties of directed Markov
fields. Networks 20:491–505
Lewis DK. 1973. Counterfactuals. Oxford: Blackwell
Mackie JL. 1980. The Cement of the Universe: A Study of Causation. Oxford University Press
Mill JS. 1843. A system of logic, ratiocinative and inductive: Being a connected view of the principles
of evidence, and methods of scientific investigation. London: John W. Harper
Pearl J. 2009. Causality: Models, Reasoning and Inference. Cambridge: Cambridge University
Press, Second ed.
Pearl J. 2015. Causes of effects and effects of causes. Sociological Methods and Research 44:149–164
Pearl J, Mackenzie D. 2018. The Book of Why. New York: Basic Books
Price H. 1991. Agency and probabilistic causality. British Journal for the Philosophy of Science
42:157–176
Reichenbach H. 1956. The Direction of Time. Berkeley: University of Los Angeles Press
Robins JM, Greenland S. 1989. The probability of causation under a stochastic model for individual
risk. Biometrics 45:1125–1138
Rubin DB. 1974. Estimating causal effects of treatments in randomized and nonrandomized studies.
Journal of Educational Psychology 66:688–701
Tian J, Pearl J. 2000. Probabilities of causation: Bounds and identification. Annals of Mathematics
and Artificial Intelligence 28:287–313
Verma T, Pearl J. 1990. Causal networks: Semantics and expressiveness, In Uncertainty in Artificial
Intelligence 4, eds. RD Shachter, TS Levitt, LN Kanal, JF Lemmer, pp. 69–76, Amsterdam:
North-Holland
Woodward J. 2003. Making Things Happen: A Theory of Causal Explanation. Oxford: Oxford
University Press
Woodward J. 2016. Causation and manipulability. The Stanford Encyclopedia of Philosophy, Ed-
ward N. Zalta (ed.).
https://plato.stanford.edu/entries/causation-mani/
Wright SS. 1921. Correlation and causation. Journal of Agricultural Research 20:557—-585

28