SUPREME COURT OF THE UNITED KINGDOM
LORD HOPE, Deputy President
26 JANUARY 2011
(with whom Lord Hope and Lord Walker agree)
The issue in this case is what is meant by the word “violence” in section 177(1) of the Housing Act 1996. Is it limited to physical contact or does it include other forms of violent conduct? The Court of Appeal, as it was bound to do by the earlier case of Danesh v Kensington and Chelsea Royal London Borough Council  EWCA Civ 1404,  1 WLR 69, held that it was limited to physical contact:  EWCA Civ 1543. The appellant contends that it is not. As the appellant is a woman, and the majority of victims of all forms of domestic violence are women, I shall refer to the victim as “she” throughout. But of course I realise that men can be victims too.
The evolution of the statutory scheme
The modern scheme of local housing authorities’ powers and duties towards homeless people dates back to the Housing (Homeless Persons) Act 1977. That Act provided that a person was homeless if there was no accommodation which she (together with other members of her family) was entitled to occupy. Even if there was such accommodation, a person was also homeless if “it is probable that occupation of it will lead to violence from some other person residing in it or to threats of violence from some other person residing in it and likely to carry out the threats”: 1977 Act, section 1(2)(b). That provision was repeated when the 1977 Act was consolidated with other housing legislation in the Housing Act 1985: see section 58(3)(b).
Then came the case of R v Hillingdon London Borough Council, Ex p Puhlhofer  AC 484, where the House of Lords held that a person was not homeless even if it was not reasonable for her to have to continue to occupy the accommodation to which she was entitled. In response to this, the Housing and Planning Act 1986 inserted two new subsections into section 58 of the 1985 Act. Subsection (2A) provided that “A person shall not be treated as having accommodation unless it is accommodation which it would be reasonable for him to continue to occupy”; but subsection (2B) permitted the local housing authority, when deciding whether it would be reasonable to continue to occupy, to have regard to “the general circumstances prevailing in relation to housing in the district”. No change was made to the basic definition in section 58(3), under which a person was automatically homeless if there was a risk of violence from another person living in the accommodation which she was entitled to occupy. Neither in 1977 nor in 1985 did the subsection specify who had to be the victim of such violence: it may have been assumed that it had to be the person claiming to be homeless or it may have been assumed that it would also cover the people living with her, in particular her children.
The scheme was recast in Part VII of the Housing Act 1996, although retaining its basic shape. The definition of homelessness, now contained in section 175 of the 1996 Act, remained the same as it had been in the 1985 Act as amended in 1986, but section 58(3)(b) dealing with violence and section 58(2B) dealing with local housing conditions were removed into section 177 (see para 5). The former “reasonable to continue to occupy” requirement in section 58(2A) is now contained in section 175(3):
A person shall not be treated as having accommodation unless it is accommodation which it would be reasonable for him to continue to occupy.
The former section 58(3)(b) and (2B) have been translated into the new section 177, which is headed “Whether it is reasonable to continue to occupy accommodation”. The former section 58(2B), dealing with local housing conditions, is now contained in section 177(2), which reads as follows:
|In determining whether it would be, or would have been, reasonable for a person to continue to occupy accommodation, regard may be had to the general circumstances prevailing in relation to housing in the district of the local housing authority to whom he has applied for accommodation or for assistance in obtaining accommodation.|
Although there was some debate about it before us, the phrase used is the “general circumstances in relation to housing” and not “the general condition of the housing stock in the area”. This strongly suggests that regard may be had, not only to the quality of housing available locally, but also to the quantity.
The former section 58(3)(b), dealing with the risk of violence, was recast as section 177(1) of the 1996 Act. In its original form, it read as follows:
It is not reasonable for a person to continue to occupy accommodation if it is probable that this will lead to domestic violence against him, or against –
For this purpose ‘domestic violence’, in relation to a person, means violence from a person with whom he is associated, or threats of violence from such a person which are likely to be carried out.
This made two changes of substance from the old law.
First, it expressly encompassed violence against other members of the homeless person’s household: a mother for example, could not reasonably be expected to occupy accommodation where her children were at risk of domestic violence.
Second, it was no longer limited to violence from someone living in the same accommodation but covered violence from an associated person, whether or not living in the same household. Section 178 spells out the “Meaning of associated person” in detail, but of course it includes spouses and former spouses, cohabitants and former cohabitants, and (since 2005) civil partners and former civil partners.
But these changes did not change the underlying purpose of section 177(1). It has variously been called a “deeming” or a “pass-porting” provision. The effect is, as it has been since 1977, that a person who is at risk of the violence to which it applies is automatically homeless, even though she has every right to remain in the accommodation concerned and however reasonable it might in other respects be for her to do so. Questions of local housing conditions or shortages do not come into it.
There was, however, another important consequence of the particular drafting technique employed in section 177. This was new to the 1996 Act and was not referred to in the argument before us. As it is automatically not reasonable for a person to continue to occupy accommodation where she is at risk of violence, she cannot be treated as intentionally homeless if she leaves. Section 191 defines when a person becomes homeless intentionally as follows:
A person becomes homeless intentionally if he deliberately does or fails to do anything in consequence of which he ceases to occupy accommodation which is available for his occupation and which it would have been reasonable for him to continue to occupy.
This result follows a recommendation of the Home Affairs Committee in their 1993 Report on Domestic Violence, to which I shall return in paragraph 21.
Section 177(1) was amended, and a new section 177(1A) introduced, by the Homelessness Act 2002. These now read as follows:
It is not reasonable for a person to continue to occupy accommodation if it is probable that this will lead to domestic violence or other violence against him, or against –
For this purpose ‘violence’ means –
and violence is ‘domestic violence’ if it is from a person who is associated with the victim.
Once the prospect of “other violence” was introduced into this pass-porting provision, it is not easy to see why the specific reference to “domestic violence”
(together with the complex definition of associated persons in section 178) was retained, unless perhaps it was thought that “domestic violence” had a special meaning. But this is quite hard to reconcile with the phrase “violence is domestic violence”. I return to this question in paragraph 31.
One reason may be that the phrase “domestic violence” has been in the scheme throughout, even though it was not originally used in the definition of homelessness in section 1 of the 1977 Act (see para 2 above). Section 5 of the 1977 Act dealt with responsibility for housing homeless people as between different local housing authorities. The authority first approached could in effect transfer responsibility to another housing authority if the applicant, or other members of her household, had no local connection with their area, but did have a local connection with another area, and “neither the person who so applied nor any person who might reasonably be expected to reside with him will run the risk of domestic violence in that housing authority’s area”: see section 5(1)(iii). The risk of domestic violence or threats of domestic violence was defined in terms of a risk “from any person with whom, but for the risk of violence, he might reasonably be expected to reside or from any person with whom he formerly resided”: see section 5(11). These provisions were consolidated in the 1985 Act as section 67(2)(c) and (3).
The same principles were carried through into section 198(2)(c) and (3) of the Housing Act 1996 in virtually identical form, save that the risk had now to come from a person with whom he “is associated”. With the introduction of “other violence” into section 177(1) by the 2002 Act, changes were also made to section 198. Section 198(2) remains in its original form, but a new section 198(2A) has been introduced and section 198(3) replaced. These now read as follows:
But the conditions for referral mentioned in subsection (2) are not met if –
For the purposes of subsections (2) and (2A) ‘violence’ means -
and violence is ‘domestic violence’ if it is from a person who is associated with the victim.
As with section 177, it is not easy to see why the distinction between domestic and other violence was retained, as the consequence is the same, unless there was thought to be some difference between them.
There is one further provision in the homelessness scheme to which I must refer. The 1996 Act introduced a new provision in section 177(3):
The Secretary of State may by order specify –
There is no equivalent power in section 198. Thus, in theory, the Secretary of State could expand the categories of people who are automatically homeless by reference to some other risk, but they could then be sent back to a district where they would face exactly that same risk.
Danesh v Kensington and Chelsea Royal London Borough Council  EWCA Civ 1404,  1 WLR 69, concerned the meaning of non-domestic “violence” in section 198. The applicant and his family were asylum seekers who had been living for just over a year in Swansea when they were granted indefinite leave to remain and thus became eligible under Part VII of the 1996 Act. They applied to Kensington which referred them to Swansea. They complained of trouble from local youths in Swansea, shouting abuse and making insulting gestures, racist abuse on a bus, and two specific incidents of assault outside a community centre and in the city centre. The local authority took the view that the two assaults were random incidents of crime which might happen anywhere to anyone and were not part of a course of harassment against the applicant or his family. The verbal abuse did not amount to a threat of violence and accordingly there was no reason to believe that it was more likely than not that violence would result if they returned to Swansea.
The Court of Appeal held that in this context, “violence” involved some sort of physical contact: Neuberger LJ accepted the council’s contention that “In section 198 ‘violence’ means physical violence, and the word ‘violence’ on its own does not include threats of violence or acts or gestures, which lead someone to fear physical violence”: see para 14. He went on to give five reasons for this, to which I shall return.
Finally, it is worth noting another innovation made by the 1996 Act. Sections 145 and 149 amended the 1985 Act and the Housing Act 1988 by introducing for secure and assured tenancies a new ground for obtaining possession of a dwelling let to a married or cohabiting couple by, respectively, a local authority on a secure tenancy and a registered social landlord or charitable housing trust on an assured tenancy, where one partner has left because of violence or threats of violence towards that partner or a member of the family living with her and is unlikely to return. This was in response to a recommendation of a Department of the Environment Homelessness Policy Division Working Party Report on Relationship Breakdown and Secure Local Authority Tenants (December 1993).
The facts of this case
The appellant is a married woman with two young children, a girl who is now aged eight and a boy who is now aged two. They were aged respectively six and eight months in August 2008 when she left the matrimonial home in which she lived with her husband, taking the children with her, and (having nowhere else to go) sought the help of the local housing authority. The matrimonial home was rented in her husband’s sole name. In her two interviews with the housing officers, she complained that “her husband hates her and [she] suspects that he is seeing another woman. [She] is scared that if she confronts him he may hit her. [However her] husband has never actually threatened to hit her.” She went on to complain of his shouting in front of the children, so that she retreated to her bedroom with them, not treating her “like a human”, not giving her any money for housekeeping, being scared that he would take the children away from her and say that she was not able to cope with them, and that he would hit her if she returned home. The officers decided that she was not homeless as her husband had never actually hit her or threatened to do so.
She consulted solicitors who applied for a review which was unsuccessful. The panel noted that “your root cause of homelessness is not that you fled after a domestic incident, but it was your decision to leave the matrimonial home because you felt that your husband did not love you any more and was not close to you, in addition to suspecting that he was seeing another woman”. They believed that “the probability of domestic violence is low” and found her fear that her husband would take the children away from her to be contradictory, as she had also said that he took no interest in the children. Hence they concluded that it was reasonable for her to continue to occupy the matrimonial home while taking action to secure a transfer under the Family Law Act 1996 or alternatively seeking accommodation in the private sector.
Mr Richard Drabble QC, who appears for the local authority, accepts that the housing officers and review panel applied the Danesh meaning when they decided that the appellant was not homeless within the meaning of the Housing Act 1996. If this Court decides that there is a wider meaning, the case will have to be considered afresh. There is no need, therefore, to make any further comment on the facts or upon the reasoning in the decision and review letters.
The meaning of “violence”
In Danesh the first, and principal, reason given was that “physical violence” is the natural meaning of the word “violence”: para 15. I can readily accept that this is a natural meaning of the word. It is, for example, the first of the meanings given in the Shorter Oxford English Dictionary. But I do not accept that it is the only natural meaning of the word. It is common place to speak of the violence of a person’s language or of a person’s feelings. Thus the revised 3rd Edition, published in 1973, also included “vehemence of personal feeling or action; great, excessive, or extreme ardour or fervour; .... passion, fury”; and the 4th (1993), 5th (2002) and 6th (2006) Editions all include “strength or intensity of emotion; fervour, passion”. When used as an adjective it can refer to a range of behaviours falling short of physical contact with the person: see, for example, section 8 of the Public Order Act 1986. The question is what it means in the 1996 Act.
The 1996 Act was originally concerned only with “domestic violence”, that is violence between people who are or were connected with one another in an intimate or familial way. By that date, it is clear that both international and national governmental understanding of the term had developed beyond physical contact. The Court is grateful to the diligence of both interveners, the Secretary of State for Communities and Local Government and the Women’s Aid Federation of England, for gathering so many of the references together. Internationally, in 1992 the United Nations Committee, which monitors the Convention on the Elimination of all Forms of Discrimination against Women (CEDAW), adopted General Recommendation 19, which included in its definition of discrimination in relation to gender based violence “acts that inflict physical, mental or sexual harm or suffering, threats of such acts, coercion and other deprivations of liberty”. In 1993, the General Assembly adopted the Declaration on the Elimination of Violence against Women, defined for this purpose as “any act of gender-based violence that results in, or is likely to result in, physical, sexual or psychological harm or suffering to women ....”
Nationally, in 1993 the House of Commons Home Affairs Committee in its Report on Domestic Violence adopted the definition “any form of physical, sexual or emotional abuse which takes place within the context of a close relationship” (Session 1992-93, Third Report, HC 245-I, para 5). The Home Affairs Committee report used two reports as the basis for its inquiry: the Report on Domestic Violence of a national inter-agency working party convened by Victim Support (1992) and the Report of the Law Commission on Domestic Violence and Occupation of the Family Home (1992, Law Com No 207). The Law Commission gave this explanation of domestic violence, at para 2.3:
|The term ‘violence’ itself is often used in two senses. In its narrower meaning it describes the use or threat of physical force against a victim in the form of an assault or battery. But in the context of the family, there is also a wider meaning which extends to abuse beyond the more typical instances of physical assaults to include any form of physical, sexual or psychological molestation or harassment which has a serious detrimental effect upon the health and well-being of the victim.|
The recommendations made in the Law Commission’s Report were embodied in the Domestic Violence and Occupation of the Family Home Bill which passed through most of its Parliamentary stages in the session 1994 – 1995 before falling at the last hurdle. The same clauses were reintroduced, with immaterial amendments, in the Family Law Bill 1995 – 1996 and became Part IV of the Family Law Act 1996.
It cannot be a coincidence that the definition of an associated person in section 178 of the Housing Act 1996 bears a very close resemblance to the definition of an associated person for the purpose of occupation and non-molestation orders under the Family Law Act 1996, in section 62(3) to (6) of that Act. It will be recalled that the Housing Act 1996 had shifted the focus, away from the presence of the perpetrator in the same accommodation as the victim, to the nature of the relationship between them. These are strong indications of joined up thinking on the part of the legislators. The Home Affairs Committee had also made the link between the criminal and family law remedies, with which it was concerned, and the housing law remedies, which were then the concern of the Department of the Environment; thus, it recommended that local authorities “put an end to the nonsense where a victim fleeing domestic violence is deemed to have made herself intentionally homeless” and that “appropriate priority be given to rehousing victims of domestic violence” (para 131). In fact, the Department of the Environment had already gone some way towards meeting the first point, as the 1991 version of the Code of Guidance for Local Authorities on Homelessness had stated (para 7.11) that authorities should not automatically treat an applicant as intentionally homeless because she had failed to use legal remedies to protect herself from domestic violence. The Department of the Environment’s Relationship Breakdown Working Party (see para 15 above) was well aware of the Law Commission’s Report: not only was the Law Commission represented upon it but the Working Party recommended implementation of the Commission’s two most relevant recommendations.
All of this indicates a consciousness in 1996 of the need to align housing, homelessness and family law remedies for victims of domestic violence, so that they could have a genuine choice between whether to stay and whether to go and the local authority or social landlord would not be obliged to continue to provide family sized accommodation to the perpetrator. There was also an explicit acknowledgement in the report which led to the Family Law Act 1996 and by the Home Affairs Committee that “violence” could have a wider meaning than physical contact.
In my view, therefore, whatever may have been the original meaning in 1977 (and, for that matter, in the Domestic Proceedings and Magistrates’ Courts Act 1978), by the time of the 1996 Act the understanding of domestic violence had moved on from a narrow focus upon battered wives and physical contact. But if I am wrong about that, there is no doubt that it has moved on now. In March 2005, the Home Office published Domestic Violence: A National Report, in which it was stated at para 10:
|To support delivery across government and its agencies through a common understanding of domestic violence, we now have a common definition. This follows the definition already used by the Association of Chief Police Officers, and is: ‘Any incident of threatening behaviour, violence or abuse (psychological, physical, sexual, financial or emotional) between adults who are or have been intimate partners or family members, regardless of gender or sexuality.’|
That definition, or something very close to it, has been adopted by many official and governmental bodies, including the Association of Chief Police Officers: Guidance on Investigating Domestic Abuse (2008); the Crown Prosecution Service Policy for Prosecuting Cases of Domestic Violence (2010); the Ministry of Justice, in Domestic Violence: A Guide to Civil Remedies and Criminal Sanctions (February 2003, updated March 2007); and the UK Border Agency, in Victims of Domestic Violence: Requirements for Settlement Applications. Indeed, it is cited in Hounslow’s own leaflet, Domestic Violence: What it is and how you can get help (2009), which goes on to explain:
|It is rarely a one off incident and it is not only about being physically or sexually abused, you may be subject to more subtle attacks, such as constant breaking of trust, isolation, psychological games and harassment. Emotional abuse is just as serious and damaging; many survivors will carry the emotional scars long after the physical injuries have healed.|
The 2006 version of the Homelessness Code of Guidance for Local Authorities is explicit at para 8.21:
|The Secretary of State considers that the term ‘violence’ should not be given a restrictive meaning, and that ‘domestic violence’ should be understood to include threatening behaviour, violence or abuse (psychological, physical, sexual, financial or emotional) between persons who are, or have been, intimate partners, family members or members of the same household, regardless of gender or sexuality.|
This was new to the 2006 Code. The fourth reason given by the Court of Appeal in Danesh, at para 18, was that various passages in the previous, 2002, Code had given a different impression, for example by comparing “severe harassment” with “actual violence” (para 8.32).
However, it is not for government and official bodies to interpret the meaning of the words which Parliament has used. That role lies with the courts. And the courts recognise that, where Parliament uses a word such as “violence”, the factual circumstances to which it applies can develop and change over the years. There are, as Lord Steyn pointed out in R v Ireland  AC 147, at p 158, statutes where the correct approach is to construe them “as if one were interpreting it the day after it was passed”. The House went on in that case to construe “bodily harm” in the Offences Against the Person Act 1861 in the light of our current understanding of psychological as well as physical harm. The third reason given by the Court of Appeal in Danesh was that it was impermissible to construe the meaning of one phrase by reference to the meaning of another. This I accept.
However, as Lord Clyde observed in Fitzpatrick v Sterling Housing Association Ltd  1 AC 27, at p 49, which was concerned with whether same sex partners could be members of one another’s “family” for the purpose of succession to Rent Act tenancies, it is a “relatively rare category of cases where Parliament intended the language to be fixed at the time when the original Act was passed”. In other cases, as Lord Slynn of Hadley explained at p 35:
|It is not an answer to the problem to assume .... that if in 1920 people had been asked whether one person was a member of another same-sex person’s family the answer would have been ‘No’. That is not the right question. The first question is what were the characteristics of a family in the 1920 Act and the second whether two same-sex partners can satisfy those characteristics so as today to fall within the word ‘family’. An alternative question is whether the word ‘family’ in the 1920 Act has to be updated so as to be capable of including persons who today would be regarded as being of each other’s family, whatever might have been said in 1920: see R v Ireland  AC 147, 158, per Lord Steyn; Bennion, Statutory Interpretation, 3rd ed (1997), p 686 and Halsbury’s Laws of England, 4th ed reissue, vol 44(1) (1995), p 904, para 1473.|
“Violence” is a word very similar to the word “family”. It is not a term of art. It is capable of bearing several meanings and applying to many different types of behaviour. These can change and develop over time. There is no comprehensive definition of the kind of conduct which it involves in the Housing Act 1996: the definition is directed towards the people involved. The essential question, as it was in Fitzpatrick, is whether an updated meaning is consistent with the statutory purpose – in that case providing a secure home for those who share their lives together. In this case the purpose is to ensure that a person is not obliged to remain living in a home where she, her children or other members of her household are at risk of harm. A further purpose is that the victim of domestic violence has a real choice between remaining in her home and seeking protection from the criminal or civil law and leaving to begin a new life elsewhere.
That being the case, it seems clear to me that, whatever may have been the position in 1977, the general understanding of the harm which intimate partners or other family members may do to one another has moved on. The purpose of the legislation would be achieved if the term “domestic violence” were interpreted in the same sense in which it is used by the President of the Family Division, in his Practice Direction (Residence and Contact Orders: Domestic Violence) (No 2)  1 WLR 251, para 2, suitably adapted to the forward-looking context of sections 177(1) and 198(2) of the Housing Act 1996:
|‘Domestic violence’ includes physical violence, threatening or intimidating behaviour and any other form of abuse which, directly or indirectly, may give rise to the risk of harm.|
That conclusion is consistent with the decision of the Court of Appeal in AN (Pakistan) v Secretary of State for the Home Department  EWCA Civ 757. This was concerned with the meaning of “domestic violence” in para 289A of the Immigration Rules, which stipulates the requirements to be met by a person admitted as the spouse or civil partner of a person present or settled here who is the victim of domestic violence which has caused the relationship permanently to break down and who is seeking indefinite leave to remain in the United Kingdom. Richards LJ quoted the definitions in the 1993 Home Affairs Committee Report, the 2005 National Report (repeated in a more recent Report of the Home Affairs Committee, Domestic Violence, Forced Marriage and ‘Honour’-Based Violence, 2007-08, 6th Report, para 4), the guidance given by the UK Border Agency, and the President’s Practice Direction. He pointed out that “The general thrust of all those definitions is much the same” (para 23) and accepted that the term was not limited to physical violence, although “it must reach some minimum level of seriousness, which will depend upon context and particular circumstances” (para 24).
It remains to be discussed whether giving the words the meaning given them by the President of the Family Division would be inconsistent with anything in the statutory language or purpose.
The statutory language
The second reason given in Danesh for preferring a narrow construction was that, in both section 177(1) and section 198(3), violence is defined as violence or threats of violence which are likely to be carried out: para 16. If the concept of violence already included conduct which puts a person in fear of physical violence there would be no need to refer to threats at all. I am not convinced of this. For one thing, there are some forms of conduct which undoubtedly put a person in fear of violence but which would not necessarily be described as threats. Silent phone calls, heavy breathing, the sorts of stalking behaviours which were the subject matter of Bond v Leicester City Council  EWCA Civ 1544,  HLR 158 and R v Ireland  AC 147, can all put the victim in very real (and justified) fear of violence in the narrow sense. They should be covered by the concept of violence.
More importantly, if the concept of violence includes other sorts of harmful or abusive behaviour, then the reference to threats is not redundant. Locking a person (including a child) within the home, or depriving a person of food or of the money to buy food, are not uncommon examples of the sort of abusive behaviour which is now recognised as domestic violence. There is nothing redundant in a provision which refers to threats of such behaviour which are likely to be carried out.
In this Court, Mr Drabble urged an alternative solution upon us: that if there were forms of ill-treatment falling short of physical violence which ought to be included within the pass-porting provision in section 177(1), the Secretary of State could use the power in section 177(3)(a) to include them. Mr Maurici, on behalf of the Secretary of State, explained that the Secretary of State has not done so because in his view the concept of “violence” already bears the wider meaning for which the appellant contends. There is the further objection to this solution, that there is no equivalent power in section 198, so that a person might be accepted as homeless under section 177(1) but could then be referred to a district where she would face exactly the same risks.
There may also be a concern that an expanded definition is setting the threshold too low. The advantage of the definition adopted by the President of the Family Division is that it deals separately with actual physical violence, putting a person in fear of such violence, and other types of harmful behaviour. It has been recognised for a long time now that it is dangerous to ignore what may appear to some to be relatively trivial forms of physical violence. In the domestic context it is common for assaults to escalate from what seems trivial at first. Once over the hurdle of striking the first blow, apologising and making up, some people find it much easier to strike the second, and the third, and go on and on. But of course, that is not every case. Isolated or minor acts of physical violence in the past will not necessarily give rise to a probability of their happening again in the future. This is the limiting factor. Sections 177 and 198 are concerned with future risk, not with the past.
The introduction in 2002 of “other” violence into a statute which was previously concerned only with domestic violence also raises questions. They are readily answered, if I am right that the concept of domestic violence in 1996 was already wider than physical contact. As Miss Nathalie Lieven QC for the appellant points out, the introduction of “other” violence in 2002 cannot possibly have been intended to cut down the meaning which the statute already had. However, if the understanding of the conduct to which the word applies has moved on, the question of whether this also applies to “other violence” does not arise on the facts of this case, and so it is unnecessary for us to express a concluded view. Reading the statute as it now stands, there are arguments on either side. On the one hand, if “violence” has the same meaning in both “domestic violence” and “other violence”, there was no need to retain the separate concept of domestic violence, together with the complicated definition of associated persons in section 178. A person who was at risk of any violence if she stayed in or returned to the property or the locality would be protected. Retaining them as separate concepts suggests that “domestic violence” is limited by the relationship between the victim and the perpetrator, rather than by the nature of the conduct involved. “Other violence”, having no such limitation and lacking the connotations of an intimate or familial relationship, might relate to a narrower set of behaviours. On the other hand, providing in sections 177(1A) and 198(3) that “violence is ‘domestic violence’” suggests that “violence” has a constant meaning. Hence, I would incline towards the view that it does. Nor would that be surprising. People who are at risk of intimidating or harmful behaviour from their near neighbours are equally worthy of protection as are those who run the same risk from their relations. But it may be less likely that they will suffer harm as a result of the abusive behaviour of their neighbours than it is in the domestic context. In practice, the threshold of seriousness may be higher.
As the housing officers and review panel adopted a narrow view of domestic violence in this case, it is agreed that it must be remitted to the authority to be decided again. I accept that these are not easy decisions and will involve officers in some difficult judgments. But these are no more intrinsically difficult than many of the other judgments that they have to make: for example, as to the circumstances in which it is reasonable to continue to occupy the accommodation; as to whether a person has rendered herself intentionally homeless; and as to the suitability of accommodation provided by the local authority. Was this, in reality, simply a case of marriage breakdown in which the appellant was not genuinely in fear of her husband; or was it a classic case of domestic abuse, in which one spouse puts the other in fear through the constant denial of freedom and of money for essentials, through the denigration of her personality, such that she genuinely fears that he may take her children away from her however unrealistic this may appear to an objective outsider? This is not to apply a subjective test (pace the fifth reason given in Danesh). The test is always the view of the objective outsider but applied to the particular facts, circumstances and personalities of the people involved.
I would therefore allow this appeal and remit the case to be decided by the local housing authority.
I agree that the appeal should be allowed for the reasons given by Lady Hale. I add a few comments of my own, since the point is not free from difficulty and we are differing from two decisions of the Court of Appeal.
The term “domestic violence” rose to prominence in the 1970s in connexion with “battered wives” – women who, whether married or not, suffered violence at the hands of their husband or partner. One reaction was to set up refuges. Another was public pressure for the law to be reformed to give such women greater protection.
Of course, it was known that physical violence was not the only form of abuse which women suffered. For example, in 1974 Dr Elizabeth Wilson referred to a case where the husband’s constant abuse in the form of offensive and cruel denigratory remarks had already damaged his wife’s psyche “possibly in a more irreparable way than if he had broken her nose ....”: “Battered wives: why they are the born victims of domestic violence”, The Times 4 September 1974, p 13. But, understandably, the predicament of women who were the victims of physical violence was at the forefront of demands for the law to be reformed.
It is therefore not surprising that the term “domestic violence” first entered English law in the short title of the Domestic Violence and Matrimonial Proceedings Act 1976 (“the 1976 Act”) which derived from the Private Member’s Bill promoted by Miss Josephine Richardson MP. There can be no doubt that the main aim of Parliament in passing the legislation was to give some additional protection, by way of injunctions in the county court – and the possibility of including a power of arrest in certain cases - to women, whether married or cohabiting, who were likely to suffer physical violence at the hands of their husband or partner. Section 2 did indeed refer to the other party to the relationship “using violence”. But the Act was not confined to such cases. As Lord Scarman noted in Davis v Johnson  AC 264, 348C-E, the mischief at which section 1 of the Act was aimed (“molesting”) went beyond physical violence and included “conduct which makes it impossible or intolerable .... for the other partner, or the children, to remain at home.”
When, the following year, Parliament enacted the Housing (Homeless Persons) Act 1977 (“the 1977 Act”), it included provisions that were designed to provide additional help to victims of violence in the home. On this occasion it did not refer to cases where the woman was “molested”. Parliament therefore seems to have been concentrating on the paradigm case of battered wives, women who feared physical violence – understandably enough, since the new Act was imposing novel obligations on local authorities.
More than 30 years have passed. The legislation has become a familiar part of the legal landscape and has been re-enacted in the Housing Act 1996 (“the 1996 Act”). The question before the Court is whether the word “violence” in section 177(1) and (1A) of the 1996 Act is confined to physical violence.
At first sight it is curious that Parliament has maintained the special term “domestic violence”. Section 177(1) now applies to cases where it is probable that continuing to occupy accommodation will lead to “domestic or other violence” - “other” violence being violence from people, such as neighbours, who are not associated with the victim. Subsection (1A) then says that violence is “domestic violence” if it is from a person who is associated with the victim. In my view, there is no doubt that violence means the same, whether it comes from a person associated with the victim or from a third party. The form of the provision may simply reflect the way that the provision has evolved. More likely, however, the retention of the term “domestic violence” is intended to serve a purpose. The aim, it seems to me, may well be to ensure that the same standard is applied to violence within the home as to other violence and so to counter any suggestion that violence within the home is to be treated as being somehow of less significance than violence outside the home. Subsection (1A) makes it clear that any conduct that would count as violence outside the home counts as violence if it occurs within the home: the law does not give a discount to the perpetrator because of the domestic setting.
In 1974 Dr Wilson saw that the husband’s constant denigration of his wife had damaged her psyche – possibly irreparably. The Court has not been referred to any case where a court had to consider whether such conduct would have counted as “violence” for the purposes of section 1(2)(b) of the 1977 Act. I have already made the point that cases of that kind were not the focus of Parliament’s attention in enacting that provision. But it is common place for courts to have to consider whether circumstances, beyond those at the forefront of Parliament’s consideration, may properly be held to be within the scope of a provision, having regard to its purpose.
Similarly, cases of physical violence surely remain the main focus of section 177(1) of the 1996 Act. And, similarly, the question remains: does deliberate non-physical abuse which harms the other party fall within the scope of “violence” in that subsection, having regard to its purpose? Parliament has provided that it is not reasonable for someone to continue to occupy accommodation if it is probable that this will lead to her being subjected to violence in the form of deliberate conduct, or threats of deliberate conduct, that may cause her physical harm. So the person at risk is automatically homeless for the purposes of the 1996 Act. I can see no reason why Parliament would have intended the position to be any different where someone will be subjected to deliberate conduct, or threats of such conduct, that may cause her psychological harm. I would therefore interpret “violence” as including such conduct and the subsection as applying in such cases. To conclude otherwise would be to play down the serious nature of psychological harm.
A necessary precondition of a right to be rehoused under the homelessness legislation is that the applicant is without accommodation. Section 175(3) of the Housing Act 1996 as amended (the 1996 Act) provides:
|A person shall not be treated as having accommodation unless it is accommodation which it would be reasonable for him to continue to occupy.|
Section 177(1) of the Act provides:
|It is not reasonable for a person to continue to occupy accommodation if it is probable that this will lead to domestic violence or other violence against him ....|
The issue identified by the parties for the Court’s determination on this appeal is:
|Is the concept of ‘domestic violence’ in section 177(1) of the Act limited to actual physical violence or is it capable of extending to abusive psychological behaviour which could reasonably be described as ‘violence’?|
It has long been known that psychological abuse within a domestic context can cause at least as much long-term harm to the victim (most commonly the woman) as physical abuse. Certainly no one who has read the extensive material put before us by the Women’s Aid Federation of England could fail to appreciate that fact. But I have nonetheless found this a much more difficult case than other members of the Court appear to have done and I cannot hide my profound doubt as to whether at any stage of their legislative history the “domestic violence” provisions with which we are here concerned – now enacted as sections 177 and 198 of the 1996 Act - were intended to extend beyond the limits of physical violence.
A number of indicators to my mind point to this being Parliament’s limited intention. One of these is the primary meaning ordinarily given to the word “violence” as connoting physical violence (in contrast, in the present context, to other forms of domestic abuse). A second pointer is the very definition of “violence” and “domestic violence” contained in both section 177 (1A) and section 198 (3) of the 1996 Act:
For this purpose ‘violence’ means –
violence is ‘domestic violence’ if it is from a person who is associated with the victim.
Psychological abuse would plainly encompass threats whether or not they are likely to be carried out: it is the threats themselves which are intrinsically abusive and harmful. It is not generally apt to speak of a threat to carry out psychological abuse. Even if one postulates a threat, say, to lock someone up in their room or deprive them of all funds, the statutory definition stipulates that it is only if the threat is likely to be carried out that it constitutes violence: the threat itself, however hurtful and humiliating, unless likely to be carried out, is excluded from the definition.
Another pointer to Parliament’s intention is the fact that “violence” falls to be construed in the same way irrespective of whether the perpetrator is “a person associated with the victim” (sections 177(1A) and 178) or some other person. If, of course, the perpetrator is associated with the applicant, the question arising under section 177(1) is whether the applicant’s continued occupation of the accommodation would probably lead to domestic violence; the question arising under section 198(2)(c) being whether, if referred to another local housing authority for re-housing in their district, the applicant would then run the risk of domestic violence in that district. If, however, the perpetrator is not associated with the applicant, the question under 177(1) is whether continued occupation of the accommodation would probably lead to violence by that person; the question under 198(2A) being whether the applicant (whom the housing authority contemplates referring to another authority) has in the past suffered (non-domestic) violence in that other authority’s district and would probably suffer violence of a similar kind if returned there. I do not say that psychological abuse (as opposed to actual or threatened physical violence) at the hands of a non-associated perpetrator is literally incapable of being described as “violence” and of justifying respectively
deemed homelessness leading to a section 193 duty to re-house or
non-referral back to the district whence the applicant came.
I do say, however, that Parliament is unlikely to have contemplated or intended these consequences.
Fourthly, it must be recognised that when the homelessness legislation was first introduced (by the Housing (Homeless Persons) Act 1977 - homelessness by section 1(2)(b) of the Act being deemed to exist in the case of those whose occupation of accommodation would probably “lead to violence from some other person residing in it or to threats of violence from some other person residing in it and likely to carry out the threats”, described as “the risk of domestic violence” in sections 5(1)(iii) and 5(11) of the Act, the equivalent provisions to those now in section 198 of the 1996 Act) – the public’s concern as to domestic violence was essentially about battered women (for whom, one recollects, Ms Erin Pizzey was starting to provide refuges).
This view, moreover, that in the homelessness context domestic violence meant physical violence, was reflected in successive statutory Codes of Guidance issued by the Secretary of State (under provisions similar to what is now section 182(1) of the 1996 Act), certainly up until the 2006 Code. The 1978 Code, for example, referred to “fear of violence” and to “battered women .... at risk of violent pursuit or, if they return home, at risk of further violence” (paras 2.10(b) and 2.12(c)(iii)). The 2002 Code (issued following the 2002 amendments to the 1996 Act) refers (at para 6.18) to the required assessment of the likelihood of a threat of violence being carried out not being based “solely on whether there has been actual violence in the past” (emphasis added) and (at para 8.26) to “the safety of the applicant .... [being] of paramount concern” (emphasis added). A little later, not in the context of deemed homelessness but rather of priority need for accommodation because of vulnerability for some “other special reason” (section 189(1)(c) of the 1996 Act), the 2002 Code (at para 8.32) says:
|People fleeing harassment. In some cases severe harassment may fall short of actual violence or threats of violence likely to be carried out. Housing authorities should consider carefully whether applicants who have fled their home because of non-violent forms of harassment, for example verbal or psychological abuse or damage to property, are vulnerable as a result.|
There, it can readily be seen, “verbal or psychological abuse” is mentioned as an example of “non-violent forms of harassment” and contrasted with “actual violence”.
True it is that from 1991 onwards the successive codes from time to time refer to violence or threats of violence including, for example, “racial harassment or attacks”, “sexual abuse or harassment”, and “harassment on the grounds of religious creed”. Invariably, however, until 2006, this was in the context not of deemed homelessness under section 177(1), but rather of whether it was reasonable for the applicant to continue to occupy his (or more generally her) accommodation, the question now arising under section 175(3) of the 1996 Act. This is the basic question which has arisen ever since 1986 (when the Housing Act 1985 was amended to overturn the effect of R v Hillingdon London Borough Council, Ex p Puhlhofer  AC 484) in every case save when section 177(1) deems continued occupation not reasonable.
Only in the 2006 Code (at para 8.21) did the Secretary of State first indicate his support for a wider interpretation of section 177(1):
|The Secretary of State considers that the term ‘violence’ should not be given a restrictive meaning, and that ‘domestic violence’ should be understood to include threatening behaviour, violence or abuse (psychological, physical, sexual, financial or emotional) between persons who are, or have been, intimate partners, family members or members of the same household, regardless of gender or sexuality.|
It is not, of course, suggested that this notable change in the Secretary of State’s Code of Guidance could directly affect the true construction of the statute: such guidance can be at most persuasive of the meaning to be given to legislative provisions. It is, after all, for the courts not the executive to interpret legislation. But it is suggested that, consistently with the “living instrument”, “always speaking” approach to statutory construction, and following the decision of the House of Lords in Fitzpatrick v Sterling Housing Association Ltd  1 AC 27, the developing perception and understanding of domestic violence now enables, indeed requires, the interpretation of the relevant sections in line with the Secretary of State’s present views, reflecting as these do modern thinking on the question. By the same token that the majority of the Court in Fitzpatrick had regard to changes in social habits and opinions to interpret the phrase “tenant’s family” as being capable of encompassing a same-sex partner, so too, the appellant argues, nowadays it would be wrong to continue construing “domestic violence” (or, indeed, this being a necessary part of the appellant’s case, “violence” outside the domestic context) as meaning physical violence only.
Tempting though it is to accept this argument – one does not, after all, like to appear old-fashioned – I confess to doubts and hesitation here too. If one considers just why it is that domestic violence (indeed, violence generally), in contradistinction to all other circumstances, has been thought to justify a deeming provision – a provision, that is, which deems it unreasonable that a probable victim of future such violence should continue to occupy his or her present accommodation, the explanation would seem to me to lie partly in the obvious need for the speedy re-housing of those identified as being at risk of violence in order to safeguard their physical safety, and partly in the comparative ease with which this particular class of prospective victims can be identified. With the best will in the world I find it difficult to accept that there is quite the same obvious urgency in re-housing those subject to psychological abuse, let alone that it will be possible to identify this substantially wider class of prospective victims, however precisely they may be defined, with anything like the same ease. Confining the deeming provision to the victims and potential victims of physical abuse does not, of course, remove all other victims from protection. Rather it leaves their cases to be assessed under section 175(3). If, then, an applicant does come to be assessed as a victim of sufficiently severe psychological abuse to satisfy the section 175(3) test for homelessness (a process which I accept would be likely to take rather longer than a section 177(1) judgment in respect of physical abuse), then obviously he or she would have to be re-housed just as if they had been deemed homeless under section 177(1).
It is, of course, true that, in section 175(3) cases generally but not in deemed cases, the housing officer is empowered by section 177(2) to have regard to “the general circumstances prevailing in relation to housing in the district”, so that theoretically, on the present understanding and application of the statute, a victim of psychological abuse, in contradistinction to a victim of physical abuse, could be subject to an adverse decision on homelessness by reference to the limited stock of housing available to an authority for re-housing purposes. Realistically, however, I see this as only a theoretical possibility since it seems to me that section 177(2) exists essentially to deal with complaints about the quality of an applicant’s existing housing: the housing officer may on occasion have to decide that an applicant’s present accommodation, however un-ideal, must suffice given the quality and quantity of the authority’s stock generally.
I had at one time thought that the solution to the problem raised by this case – if problem there is – lay in the Secretary of State’s order-making power under section 177(3)(a) of the 1996 Act. I recognise, however, that there are difficulties in the use of this power: first, that, given the Secretary of State’s view that the victims of psychological abuse are already covered by section 177(1), he cannot properly specify their needs as arising in “other circumstances”; secondly, that the use of this power could not in any event affect the proper approach to section 198 so that the problem would not be entirely solved. There would remain the possibility of someone being returned for re-housing to an area where, although not cohabiting with an abuser, he or she might be at risk of future psychological abuse from a non-cohabiting family member (essentially the position in Bond v Leicester City Council  HLR 158, although that case was in fact concerned with intentional homelessness and appears to have been argued and decided on the assumption that section 177(1) dealt with physical violence only) or, indeed, a neighbour.
Certainly, I no longer see section 177(3) as the solution to this case. Rather the Court has no alternative but to decide whether it is indeed now right, pursuant to the Fitzpatrick principle, to give to the terms “domestic violence” and “violence” the wider meaning contended for by the appellant and both interveners. In taking this course we would, of course, be overturning two clear and unanimous decisions of the Court of Appeal: respectively of Mummery, Jacob and Neuberger LJJ in Danesh v Kensington and Chelsea Royal London Borough Council  1 WLR 69 and of Waller, Laws and Etherton LJJ in the present case. I have already indicated my very real doubts about doing so. At the end of the day, however, I do not feel sufficiently strongly as to the proper outcome of the appeal to carry these doubts to the point of dissent. I am content that the views of the majority should prevail and that the appeal should be allowed.
Nathalie Lieven QC & Martin Hodgson (instructed by Scully and Sowerbutts) for appellant.
Richard Drabble QC & Matthew Feldman (instructed by London Borough of Hounslow Legal Services) for respondent.
James Maurici Stephen Knafler QC (instructed by Treasury Solicitor) for Sec of State for Communities & Local Government, intervener.
Neil Jeffs (instructed by Sternberg Reed) for Women's Aid Federation of England, intervener.
all rights reserved