THE EMERGENCE OF A “FLEXIBLE” CONDITIO SINE QUA NON TEST TO FACTUAL CAUSATION?

For many years the factual causation inquiry has remained relatively uncontentious as the determination of a factual nexus has not posed many problems for the courts. The courts establish a factual nexus on the evidence and probabilities of each case and employ the well-known conditio sine qua non test to affirm their findings on factual causation (we use the terms conditio sine qua non test and “but for” test interchangeably) (Visser “Gedagtes oor Feitelike Kousaliteit in die Delikte Reg” 2006 TSAR 581 583). However, in a recent judgment of the Constitutional Court, Lee v Minister of Correctional Services (2013 (2) SA 144 (CC)), the Court was tasked to determine a factual nexus in a multifaceted set of circumstances which challenged the traditional application of the conditio sine qua non test. Specifically, the Court had to determine the existence of a factual nexus between a systemic state omission in protecting prisoners from contracting pulmonary tuberculosis (TB) and the subsequent harm that ensued. In response to this scenario, the Court articulated a so-called “flexible” conditio sine qua non test which purports to be more adaptable than the traditional formulation of this test.


Introduction
For many years the factual causation inquiry has remained relatively uncontentious as the determination of a factual nexus has not posed many problems for the courts. The courts establish a factual nexus on the evidence and probabilities of each case and employ the well-known conditio sine qua non test to affirm their findings on factual causation (we use the terms conditio sine qua non test and "but for" test interchangeably) (Visser "Gedagtes oor Feitelike Kousaliteit in die Delikte Reg" 2006 TSAR 581 583). However, in a recent judgment of the Constitutional Court, Lee v Minister of Correctional Services (2013 (2) SA 144 (CC)), the Court was tasked to determine a factual nexus in a multifaceted set of circumstances which challenged the traditional application of the conditio sine qua non test. Specifically, the Court had to determine the existence of a factual nexus between a systemic state omission in protecting prisoners from contracting pulmonary tuberculosis (TB) and the subsequent harm that ensued. In response to this scenario, the Court articulated a so-called "flexible" conditio sine qua non test which purports to be more adaptable than the traditional formulation of this test.
This paper mainly seeks to interrogate whether the Constitutional Court indeed created a novel approach to factual causation by expanding on the traditional conditio sine qua non test. In doing so, we will examine existing law on the various approaches to establish a factual nexus in an attempt to provide context to this "flexible" conditio sine qua non test. In addition, we shall also consider two schools of thought regarding the appropriate approach to be utilized to determine factual causation, which schools draw upon the co-existence of these various approaches, in a further attempt to highlight the possible flexibility in the factual causation inquiry. Thereafter, consideration will be given to the statements made by the Constitutional Court in the Lee case, and the preceding lower-court judgments, in relation to the issue of factual causation in order to gain greater clarity on the substance of the "flexible" conditio sine qua non test. Lastly, we shall provide an analysis of the court's judgment to ascertain the importance of flexibility in the factual causation inquiry, and then we shall evaluate whether the Court succeeded in creating a new approach to factual causation and whether the judgment enabled further legal development in this area of the law.
In determining factual causation, courts traditionally employ the wellknown conditio sine qua non test as the point of departure for establishing a factual nexus (Visser 2006 TSAR 583-585 and 587; see also Du Plessis 2010 THRHR 533-535; Neethling et al Law of and Loubser et al The Law of Delict in SA 284). What the conditio sine qua non test involves, according to the SCA in International Shipping Co (Pty) Ltd v Bentley (supra), is to employ a process of hypothetical deduction to establish whether the offending act is a necessary condition for the harm to occur (a condtio sine qua non) and not merely a pre-existing antecedent (700F-H; and see also Van der Walt and Midgley Principles of Delict 3ed (2005) 200). Specifically, an act will be considered to be a necessary condition if the act cannot be removed hypothetically from the prevailing factual matrix without the harm also disappearing. Depending on whether the act is in the form of a commission or omission, the process will either involve hypothetical elimination of the offending act (for a commission) or hypothetical substitution of lawful conduct (for an omission) to determine whether the consequence of the act might also simultaneously disappear. Accordingly, if the hypothetical deduction or substitution exercise removes the harm in question, the act will be considered to be a necessary condition for the harm to occur (for a detailed explanation of the substance of this test, see the discussion of the Constitutional Court in Lee v Minister of Correctional Services below; see also Van der Merwe and Olivier Die Onregmatige Daad in die Suid-Afrikaanse Reg (1989)

OBITER 2015
Despite the fact that the conditio sine qua non test has received widespread acceptance by the courts, they have also recognized that this is not the only test to determine factual causation. In circumstances where the application of the conditio sine qua non test may lead to an unjust and/or illogical result, the courts have recognized the need for exceptions and accordingly utilized other tests as alternatives (Kakamas Bestuursraad v Louw 1960 (2)  An example of one such an alternative is the so-called "common-sense approach of the man in the street" as put forward in the case of Portword v Svamvur (supra). According to the Supreme Court of Appeal, this approach gives express recognition to the notion that a factual nexus can be established in terms of human knowledge and experience by establishing whether an act is a "probable cause" of harm and not in fact a necessary condition ( Another test making a brief appearance as an approach to determine factual causation is the material contribution test. Originally articulated by the Appellate Division in Minister of Police v Skosana (supra) as an addition to the conditio sine qua non test, the material contribution test did not subsequently find widespread acceptance and the utility of this test has not received full judicial attention (Minister of Police v Skosana (supra) 35; see also Midgley in Glover (ed) Essays in Honour of AJ Kerr 296-297;and Kakamas Bestuursraad v Louw;Humphreys NO v Barnes 2004 (2) SA 577 (C)). Nonetheless, as noted by Midgley, this test is particularly useful in circumstances where the defendant's conduct, together with other preexisting conditions, contributed towards the plaintiff's harm, but the defendant's conduct cannot be established as a necessary condition due to the substantial concurrence of other factors (Midgley in Glover (ed) Essays in Honour of AJ Kerr 295 and 298). A similar line of reasoning has been adopted by the House of Lords in McGhee v National Coal Board ([1972] 3 All E.R. 1008 (H.L.)) and the Canadian Supreme Court of Appeal in Snell v. Farrell ([1990] 2 S.C.R. 311). The material contribution test dictates that an act is the factual cause of harm if the act, on a balance of probabilities, in any way contributed towards or increased the risk of the plaintiff's harm, regardless of the concurrent contribution of other pre-existing conditions (which in contrast to the conditio sine qua non test denotes a lesser standard for establishing a factual nexus) ( Ultimately, the co-existence of these approaches and tests to determine factual causation raises the question of how factual causation should be established. In an attempt to answer this question, reference can be made to two converging schools of thought. First, according to Neethling and Potgieter, there can be no universal test to determine factual causation in all factual matrixes and the probabilities and facts of each case must be examined to simply determine whether the plaintiff's harm arose out of the defendant's conduct (Neethling et al The Law of Delict 185-187). Secondly, Midgley proposes that a value judgment must be made to determine the most appropriate test to be utilized in a particular set of facts to ensure a just result. Specifically, this process entails that common sense, as an underlying value rather than as a distinct test, should be employed to determine whether one should follow the conditio sine qua non test or deviate from this test and follow another test (and, if so, which test) (Midgley in Glover (ed) Essays in Honour of AJ Kerr 300-302). Therefore, Midgley essentially argues that the factual causation inquiry is not devoid of any normative content as the determination of a factual nexus entails a value judgment based on society's notions of common sense and justice (Midgley in Glover (ed) Essays in Honour of AJ Kerr 300-302). Importantly, this assertion challenges the traditional view that factual causation is a purely factual concept which is not influenced by normative considerations.
Despite the different appearance of each of these schools of thought, closer examination reveals that they both conceal the underlying notion that various factors and considerations must be taken into account in order to determine factual causation. It can be argued that this notion illustrates a flexible approach, or at least a contextual one, to factual causation as the determination of a factual nexus should not rely on the formalistic application of a single test (a similar line of reasoning is adopted for the other delictual elements, see

3
Lee v The Minister of Correctional Servicesthe case

1 The facts
In November 1999 Mr Dudley Lee (the plaintiff) was arrested for, amongst others, fraud, counterfeiting and money laundering. He was incarcerated in Pollsmoor maximum security prison until his acquittal and release in September 2004 (except for a short period of about two months during 2000 when he was released on bail). The plaintiff appeared in Court no less than 70 times during his incarceration (Constitutional Court judgment (CC) par 6; Supreme Court of Appeal judgment (SCA) par 1; and High Court judgment (HC) par 2).
The plaintiff was 54 years old when he was arrested and was in reasonable health. For a majority of the time he spent at Pollsmoor he was housed in a cell designed for occupation by one person, but he shared the cell with two other inmates. At a point, he was detained in a communal cell with about 25 other inmates. Prisoners were confined to their cells, together with those they shared a cell with, for up to 23 hours a day (CC par 7 and 10; and SCA par 2).
After about three years of being incarcerated, the plaintiff contracted TB. Despite this diagnosis, and the possibility of his disease being infectious, he was returned to his cell which he shared with at least one other person at all times, and he attended court, which involved his travelling in a police van in close-confined proximity to others (the Constitutional Court uses the phrase "stuffed into vans like sardines") and spending the day in a "jam-packed" court cell. The plaintiff received treatment for about 6 months before being declared healthy (CC par 6 and 10; and SCA par 2, 15, 29, 30 and 31).
After his release from prison, the plaintiff instituted action against the Minister of Correctional Services for delictual damages. The basis of his claim was that the prison authorities had failed to implement adequate precautions to protect him from contracting TB; that he had contracted TB due to their failure; and that the failure by the prison authorities amounted to a violation of, amongst others, his right to protection of his physical integrity, which right existed under common law, the Correctional Services Act 8 of 1959, and the Constitution of the Republic of South Africa, 1996 (CC par 10 and 13; SCA par 3; and HC par 2). Only the question of liability was dealt with at the trial (CC par 11; SCA par 4; and HC par 2).

2 The High Court decision: Western Cape High Court, Cape Town
The High Court, after detailed consideration and analysis of the evidence, considered the legal position in relation to delictual claims. The Court correctly stated that "[i]n order to establish a claim in delict, a plaintiff has to prove that the defendant negligently committed an act which was unlawful and that the act so complained of was causally related to the harm which ensued" (par 199). The Court was required to reach a conclusion on unlawfulness, fault and causation (par 211). Given that the focus of this piece is factual causation, only those parts of the judgment relevant to the discussion will be dealt with.
The High Court briefly sets out the law relating to causation and confirms that there has to be a causal nexus between the conduct complained of and the harm which ensued. The enquiry into causation involves two distinct enquiries, namely factual causation and legal causation. Factual causation involves the use of what has become known as the "but for" test, which determines "whether or not the postulated cause can be identified as the sine qua non of the loss in question". If the test for factual causation is satisfied, "it must be determined whether the wrongful act is linked sufficiently closely to the loss concerned for liability to ensue" in order to determine legal causation. No liability will result where it is found that the damage is too remote (par 211).
De Swardt AJ formulates the issues relating to causation as follows (par 212): "(1) Whether or not the prevailing conditions in the maximum security prison at Pollsmoor, during the period November 1999 to June 2003, were such that the spread of TB was facilitated thereby; if the answer to this issue is in the affirmative.
(2) Whether it is more probable than not, that the plaintiff's illness with TB was occasioned by, or resulted from, the prevailing conditions in the maximum security prison at Pollsmoor during his incarceration." In applying the law to the facts to determine issue 1 above, the court finds that, based on the evidence presented, TB was already prevalent in the prison at time the plaintiff was first taken to Pollsmoor, and TB remained a problem throughout the time of his incarceration (par 213). Effective control over the disease depended upon proper screening of incoming prisoners, isolating those infected, and administering medication properly, all of which depended upon there being a sufficient number of nursing staff available who were suitably qualified (par 214). The Standing Correction Orders, which were compiled in order to give effect to the provisions of the Correctional Services Act 8 of 1959 (this legislation was found to be applicable as the Correctional Services Act 111 of 1998 only came into operation on 31 July 2004, after the plaintiff become ill with TB) and the Correctional Services Act 111 of 1998, were said to impose strict obligations in regard to the medical procedures to be followed when admitting persons to the prison. The reason for these strict obligations is two-fold, "[f]irstly, prisoners who were ill or injured had to receive medical attention" and, "[s]econdly, prisoners who posed, or could reasonably pose, a health risk to others had to be identified in order that the necessary steps might be taken to prevent other inmates from becoming ill" (par 215). In addition, the Standing Orders contained express provisions relating to communicable and contagious diseases and procedures to be followed in regard to prisoners so infected (par 215).
It became apparent from the evidence that incoming prisoners were not screened for TB, or any other disease, by a medical practitioner or registered nurse when arriving at the prison (par 216). Prisoners were screened only the following morning and the screening by nurses did not involve a physical examination of the prisoners, but "merely noted whether a prisoner provided a positive or negative answer to the question as to whether he had any medical complaints". Only those who gave a positive answer were referred for examination by a doctor (par 217). Many prisoners, frequently from lower economic classes who smoked, would have coughed, but such cough would not have been regarded as pathological. Thus, they OBITER 2015 would have provided a negative response to the question and would not have been examined, despite possibly already being infected with TB, which a chest examination would have revealed (par 218).
After admission, authorities relied on a self-reporting system to detect infectious diseases which required prisoners to make it known if they were ill or required medical attentionno further screening procedures were employed (par 220). The failure to screen incoming prisoners adequately was said to be a contravention of the Standing Orders which allowed persons who were ill with an infectious disease to mingle with other prisoners (at least overnight). Those who remained in the general prison population while infected (as they did not request medical assistance) expelled the TB bacteria into the overcrowded cells when sneezing, coughing or spitting (par 221). Other factors which contributed to the transmission of TB included "overcrowding, a lack of free flowing air, [a] lack of isolation facilities", the inadequate administration of medication, and a shortage of nursing staff (par 222-225 and 227-228). Based on the evidence presented, the nature of the disease and the manner in which it is transmitted, the Court found that the factors considered led one to the inescapable conclusion that prevalent conditions in the prison facilitated the spread of TB (par 229).
In deciding whether "it was more probable than not that the plaintiff's illness with TB was occasioned by, or resulted from, the prevailing conditions" in the prison, the Court considered the plaintiff's medical history, particularly the fact that he had not been ill with TB prior to his incarceration, and that the plaintiff had become infected with the TB disease approximately three years after his admission to prison (par 230 and 232). De Swardt AJ found that these factors, together with the manner in which the disease is spread, resulted in one coming to the conclusion that, but for his incarceration in Pollsmoor, "the plaintiff would probably not have become ill with TB", and thus factual causation is satisfied (par 234). The Court was satisfied, on a consideration of all the evidence before it, that it was "more probable than not that the plaintiff contracted TB as a result of his incarceration" in prison (par 236).

The Supreme Court of Appeal
The Minister of Correctional Services took the decision of the High Court on appeal to the Supreme Court of Appeal. In considering the matter, Nugent JA held that "[t]he three elements of a delictual claim that is founded on negligence are well establisheda legal duty in the circumstances to conform to the standard of the reasonable person; conduct that falls short of that standard, and loss consequent upon that conduct" (par 33). In regard to wrongfulness, the SCA agreed with the High Court and found in favour of categorizing the failure on the part of the authorities as wrongful (par 34-42). After a very brief consideration of the fault element, Nugent JA also agreed with the High Court and concluded that "the prison authorities failed to maintain an adequate system for management of … [TB] and in that respect they were negligent" (par 44; see par 43-44 for the discussion on fault).
The element of causation was classified as the "problematic" element by the SCA, and an in-depth analysis of causation was embarked upon (the correctness or otherwise of this analysis is dealt with below, see paragraph 4 below; see par 45-64). The Court found that for causation to be present, "a plaintiff must establish that it is probable that the negligent conduct caused the harm" (par 46). The test to be applied to make a determination in this regard is the "but for" test (the Court proceeded to embark on a consideration of the "but for" test and the hypothetical enquiry to be made during its application; par 46-49).
The Supreme Court of Appeal holds that the Court a quo, in determining causation, erroneously considered whether the plaintiff had been infected with TB while incarcerated and, after finding that this was probably so, considered the causation enquiry to be complete (par 50 and 55). According to Nugent JA, "[t]he question was not whether the incarceration caused the harm, but whether it was caused by the negligent omission. Whether or not he was infected while incarcerated was a necessary but not an exhaustive step in that enquiry" (par 55). It was said that, if the plaintiff was not infected while in prison, the enquiry would end, but if it was found that he was infected during his imprisonment, the question was whether he would have been infected if the prison authorities had measures in place to reasonably manage the spread of TB. Bringing terminology in from the negligence enquiry, Nugent JA took the view that "[p]roof alone that reasonable precautions were not taken to avoid foreseeable harm, and that the harm occurred, does not establish that the former caused the latter" (par 55).
To determine whether the plaintiff would have been infected with TB had reasonable measures been in place, one is required to undertake a twostage enquiry, firstly, one has to consider what "a reasonable person in the position of the defendant [would] have done to avoid the occurrence of harm"; and secondly, one needs to determine "whether the harm would have been avoided had that been done" (par 56). A reasonable person is only required to take reasonable steps to avoid foreseeable harm; the law does not expect a defendant to guarantee that foreseeable harm will not occur (par 56). The first part of the enquiry, what ought to have been done, is usually determined during the negligence enquiry, but according to the Court, this need not always be the case (however, the correctness hereof falls beyond the scope of this piece; we shall not deal with it here; par 57). A finding that the authorities failed to adhere to the standard of what ought reasonably to have been done is sufficient for a finding of negligence, but not for causation. For purposes of causation one needs to go further and establish "what the prison authorities ought to have done: only from there can one proceed to the enquiry whether that would have prevented [the plaintiff] being infected" (this was dealt with by the High Court to some extent in the enquiry into negligence; par 57 and 58).
In determining what one might reasonably expect in a large prison, various factors have to be considered and balanced against one another, including "the security demands of the prison; the financial resources that are available to the prison authorities; generally accepted practice amongst prison authorities; the extent to which trained personnel are available; the OBITER 2015 space available for isolation; [and] the incidence of the disease" (par 60). The enquiry is said to be a "substantial and complex systematic enquiry" (par 60). Despite all this, the Court found that whatever enquiry is conducted, one insurmountable hurdle remains for the plaintiff, namely whatever management strategies were in place, the risk of contagion will always exist as intervention must be preceded by diagnosis, and diagnosis often takes place only after the prisoner has become contagious. It would be unreasonable to expect prison authorities to examine about "4000 prisoners with such regularity and thoroughness that [TB] will always be detected before the prisoner becomes contagious" (par 61). The plaintiff's difficulty was that he was not aware of the source of his infection, forcing him to rely on a systematic omission. The SCA found that "in the absence of proof that reasonable systematic adequacy would have altogether eliminated the risk of contagion … it cannot be found that but for the systematic omission he probably would not have contracted the disease" (par 64). It was on this ground that the claim failed and the appeal was upheld (par 64 and 70).

4 The Constitutional Court
The focus of this discussion will be limited to the manner in which the Constitutional Court dealt with the element of causation and the approach adopted by the Supreme Court of Appeal (the judgment also deals with other issues, such as leave to appeal, the introduction of new evidence, and the amendment of pleadings to introduce a new claim for constitutional damages). The question in relation to causation to be answered by the Constitutional Court was "whether the causation aspect of the common-law test for delictual liability was established and, if not, whether the common law needs to be developed to prevent an unjust outcome" (par 2; and see also par 29).
The causation debate was the focus of the judgment. In this regard, Nkabinde J, writing for the majority, held that "[t]he point of departure is to have clarity on what causation is" (par 38). The Court confirmed that the causation element involves two distinct enquiries, namely the factual enquiry to establish factual causation and a juridical enquiry to establish legal causation (par 38). Factual causation is a question of fact for which there are different theories to be applied, but the one most frequently used by the courts is the "but for" test or conditio sine qua non test (par 39-40). As stated in International Shipping Co (Pty) Ltd v Bentley (supra 700H-F), and quoted by Nkabinde J (par 40), this test: "may involve the mental elimination of the wrongful conduct and the substitution of a hypothetical course of lawful conduct and the posing of the question as to whether upon such an hypothesis plaintiff's loss would have ensued or not. If it would in any event have ensued, then the wrongful conduct was not a cause of the plaintiff's loss; [otherwise] it would not so have ensued. If the wrongful act is shown in this way not to be a causa sine qua non of the loss suffered, then no legal liability can arise." When one is dealing with a positive act, the conduct is mentally removed from the factual situation in order to determine whether the consequence would still have ensued. When the conduct in question is an omission, a hypothetically positive act must be inserted into the factual matrix to determine whether such reasonable act would remove the consequence in question. This is, however, not an inflexible rule for a number of reasons, first, in certain instances a strict adherence to this would result in injustice, and second, it is not always easy to draw a distinction between positive conduct and omissions (par 41).
Nkabinde J disagrees with the approach adopted by the Supreme Court of Appeal in the causation enquiry for two reasons. One, "it was not necessary for the substitution of reasonable alternative measures to determine factual causation because our law allows for a more flexible approach"; and two, "even if the use of a reasonable alternative substitution was necessary in the circumstances, our law does not require evidentiary proof of the alternative, but merely substitution of a notional and hypothetically lawful, non-negligent alternative" (par 43). Factual causation is established, according to the Constitutional Court, if the substitution exercise involved the "but for" test to be done in this way (par 43). The substitution exercise conducted in determining factual causation is to be applied flexibly; the law does not require an inflexible application of the test (par 45 A further reason for exercising caution when adopting the substitution test to determine factual causation, the Court said, is that it "involves an evaluation of normative considerations" (par 51). Social and policy considerations should not be depended upon in the determination of questions of fact, although it is an evaluative exercise. It was found that "[e]ven though the purpose of using the normatively determined lawful conduct as an alternative is not primarily aimed at making an 'is' question an 'ought' question, it seems to me that it inevitably makes it at least a missed question of fact and law. The distinction between factual and legal causation made in our law becomes unnecessarily less clear" (par 51). Nkabinde J moves on to consider the development of wrongfulness as the element used to determine the boundaries of delictual liability. Wrongfulness is today determined by considerations of public and legal policy, consistent with the norms embodied in the Constitution. The general criterion of "reasonableness" is used in relation to both the elements of wrongfulness and negligence. In regard to wrongfulness, one enquires whether or not it is reasonable to impose liability on the defendant, while for purposes of negligence, the enquiry relates to the reasonableness of the defendant's actual conduct. Therefore, factual causation should not be determined with reference to "these kinds of normative considerations based on social and policy considerations" (par 53).
Based on the fact that it is not always easy to draw a distinction between a positive act and an omission, the Constitutional Court found that the there is nothing in the law which prevented the High Court "from approaching the question of causation simply by asking whether the factual conditions of [the plaintiff's] incarceration were a more probable cause of his [TB], than that which would have been the case had he not been incarcerated in those conditions" (par 55). Therefore there was no reason for the Court to interfere with what the High Court found based on the law (par 55). However, the Court did continue in its analysis and provided that even if the substitution exercise is to be preferred for the factual causation enquiry, it is not required of the plaintiff to adduce further evidence to prove "what the lawful, nonnegligent conduct of the defendant should have been", on a balance of probabilities (par 56). Only a substitution exercise is to be carried out, inserting hypothetically lawful conduct into the factual scenario, and questioning whether the plaintiff's loss would still have ensued. No actual proof of the conduct is required (par 56). One has to determine what ought to have been done by the defendant in order to prevent the damage, and ask whether such conduct would have had a better chance of preventing the loss than that conduct which was actually adopted by the defendant (par 58).
As a non-negligent system would have reduced the risk of contagion among the prison population, the Constitutional Court found that factual causation has been established (par 60-67). The next enquiry to become relevant was therefore the question of legal causation, which was said to involve the question "whether the defendant should be held liable" (par 68; for a discussion on this see par 68-70). In concluding the enquiry into causation, the Court held that "there is a probable chain of causation between the negligent omissions by the responsible authorities and [the plaintiff's] infection with TB" (par 71). The Court therefore upheld the plaintiff's claim (par 71, 75 and 77).
Nkabinde J did not regard it necessary to develop the common law relating to causation as the "but for" test should not be applied inflexibly (par 72-75). This was, however, not the stance adopted by the minority. The minority took the view that the common law did in fact need to be developed and the matter should be remitted to the trial court for a consideration of how the common law ought to be developed (par 78-116).

Discussion
In our discussion of the Constitutional Court's judgment, we are firstly going to consider the response of the Constitutional Court to the judgment of the Supreme Court of Appeal regarding the issue of factual causation. Thereafter, we shall set out a few points of critique against the approach adopted by the Constitutional Court in establishing a factual nexus.
In dealing with the judgment of the Supreme Court of Appeal, the Constitutional Court highlighted two important considerations underlying the factual causation inquiry, namely, the need for employing flexibility in the factual causation inquiry, and secondly, the possible existence of normative considerations in establishing a factual nexus. Arguably, these considerations influenced the Court when it reached its decision on factual causation.
Nkabinde J, in writing for the majority, points out numerous times that one has to adopt a measure of flexibility when applying the "but for" test, as the determination of a factual nexus or the application of a particular test is not "inflexible" (par 45-50). According to Nkabinde J, this measure of flexibility is needed in order to avoid the injustice that may be created by the inflexible application of a particular rule of factual causation, as identified and demonstrated by previous case law (par 49). In particular, such flexibility is most suitable in circumstances where courts have to explore the potentially porous boundaries between a positive act or an omissionthe interrogation in these circumstances should aim to produce a just result instead of reducing the inquiry into either an inflexible elimination (for positive acts) or substitution (for omissions) exercise (par 49).
Moving on to the next consideration, the Court further highlighted that the factual causation inquiry involves "an evaluation of normative considerations" (par 51). Essentially, the Court asserted that the finding of a factual nexus is an "evaluative exercise" combining the questions of fact and law (par 51). Therefore, the inquiry into a factual nexus, to some extent involves a value judgment based on the relevant facts to postulate a cause of the plaintiff's harm (par 51). However, the mentioned normative considerations are not of a social or policy nature and do not, for example, require an investigation into the criterion of reasonableness, which primarily dealt with the elements of wrongfulness and fault (par 51). Ultimately, the Court's assertion in this regard seems to resemble Midgley's proposition that the determination of a factual nexus also involves the exercise of a value judgment (see discussion above).
It would appear that the Constitutional Court attempted to transcend the Supreme Court of Appeal's somewhat formalistic conception of factual causation and application of the conditio sine qua non test. In doing so, the Court purportedly advanced its notion of a "flexible" conditio sine qua non test to the particular set of facts (par 51-58).
In turning to the actual approach adopted by the Court, it becomes clear that the main consideration for the Court was whether the State's conduct was a "probable cause" of the plaintiff's harm (par 55). Although the Court attempted a hypothetically substitution exercise, almost as an afterthought, which reflects the methodology of the conditio sine qua non test, the Court only came to the conclusion that such hypothetical conduct would have decreased the chances of the plaintiff contracting TBbut the Court could not establish on the facts that such conduct would have conclusively prevented the plaintiff's infection of TB (par 55 and 58-60). Framed differently, the Court established only that the State's conduct contributed towards the plaintiff's harm in the sense that the State's omission could not be considered to be a necessary condition for the plaintiff's infection to occur (which would have been indicative of a conditio sine qua non), but rather played a significant role in the plaintiff's infection of TB (which is indicative of a material contribution). However, the Court did not expressly articulate its findings as such and simply stated that it was probable that the plaintiff contracted TB due to the State's omission (as explained earlier, this line of reasoning reflects the common-sense approach).
Accordingly, the Court's formulation and application of its so-called "flexible" conditio sine qua non test is not immune to critique. In the discussion that follows, we shall raise three points of criticism against the Court's approach to factual causation.

OBITER 2015
First, the Court incorrectly described its approach as a "flexible" conditio sine qua non test. Essentially, the Court found that common sense should dictate that there existed a factual nexus as the State's omission contributed towards the plaintiff's infection (despite the fact that the State's conduct could not be considered as the main cause of harm) (par 58 and 60-61). Therefore, this test formulated by the Court is not a proper or even a variant formulation of the condtio sine qua non test, as the supposed flexible application of this test did not highlight a necessary condition (but did highlight the Supreme Court of Appeal's formalistic application of this testsee par 62-67).
Secondly, the Court failed to distinguish properly between the various tests to factual causation it considered in finding a factual nexus. Arguably, such an "amorphous" treatment of the factual causation inquiry hindered any further contemplation on the utility of the different tests to factual causation. For instance, the Constitutional Court could have expressly recognized the material contribution test as a viable alternative to the conditio sine qua non test. In particular, this test provides a plausible solution to circumstances where a defendant's conduct, although not a necessary condition, contributed together with other factors to the plaintiff's harm (which was arguably the case before the Court). Similarly, the Court could have interrogated the concept of "probable cause" under the common-sense approach even further. Although it is true that the probabilities of each case should be considered, and that a determination of factual causation should not offend common-sense standards. It could have been questioned whether common sense should function as a distinct test for factual causation. Specifically, Midgley argues that due to the vagueness and capriciousness of this concept, common sense cannot properly function as a distinct test for factual causation (Midgley in Glover (ed) Essays in Honour of AJ Kerr 200-202). However, common sense can be applied as an underlying value to any test for factual causation to ensure that a just and logical result has been achieved (which found some buy-in by the Court although it considered common sense indirectly as a distinct test. See Midgley in Glover (ed) Essays in Honour of AJ Kerr 293-294).
Thirdly, the Court's amorphous treatment of the factual causation inquiry also fell short of providing a systematic analysis and application of the law. In doing so, the Court missed an opportunity of clarifying how these different approaches can coincide in the factual causation inquiry. Importantly, a systematic analysis of the different approaches to factual causation (and how these interrelate with one another) is not an elusive exercise. In this regard, mention can be made of Midgley's proposal on how these different approaches can coincide (as discussed above). However, Midgley's approach draws from existing law on factual causation and he simply articulates the ideas in a conceptually coherent manner and attempts to explain the connections between the different approaches by adopting a flexible approach to factual causation (and not limiting the discussion to the flexible application of a particular test, in contrast to the approach adopted by the Court).
Ultimately, it is questionable whether the Court developed a new approach as it draws from various values and tests to factual causation but fails to explain how these can be utilized collectively to produce a just result. Arguably, the Court attempted to broaden the scope of the conditio sine qua non test by employing a "flexible" version of it to find liability for a systemic state omission in circumstances where it was difficult to establish a necessary condition. However, existent law sufficiently recognizes that formalism in the factual causation inquiry can lead to unjust results, and that the conditio sine qua non test is not the exclusive test for factual causation (see discussion above). Furthermore, deviation from the conditio sine qua non test is allowed in circumstances where this approach cannot produce a satisfactory result according to society's notion of justice (even if this test is not applied rigidlysee discussion above). Accordingly, all that was needed was to articulate and apply the law in a systematic manner which could have provided a satisfactory solution to the question of liability for a systemic state omission, or could have initiated the debate of whether the common law should be developed. However, by focussing on the flexibility of a particular test and not on the possible flexibility of the factual causation inquiry itself (as various tests and underlying values to the factual causation inquiry are available to establish a factual nexus), the Court missed an opportunity to contribute meaningfully towards the law on factual causation.

Conclusion
The law regarding factual causation has been developed over a number of years with various approaches emerging through the times. All these approaches are available and find support in different cases. What was necessary was an explanation of how the different approaches should be applied, together with a systematic application of the law to illustrate this. Despite the Constitutional Court having the perfect opportunity to clearly and coherently set out the law relating to factual causation for a systemic omission, and provide guidance on which test should be applied when, the Court opted to rather adopt the so-called "flexible conditio sine qua non test".
Although it is greatly welcomed that the Constitutional Court affirmed liability for a systemic state omission, the Court failed to make use of this opportunity to provide clarity and rather added to the confusion by introducing another possible approach. Arguably, legal development regarding factual causation therefore continues to be lost in the process due to insufficient explanations and incorrect terminology.

CJ Visser and Christin Kennedy-Good
School of Law, University of the Witwatersrand, Johannesburg