Richard Ekins: Yemshaw and “updating” statutes

Should judges update the meaning of statutes?  Consider Yemshaw v London Borough of Hounslow [2011] UKSC 3, in which the Supreme Court updated a statute.

The question in Yemshaw was the meaning of ‘violence’ in s 177(1) of the Housing Act 1996. The precursor to s 177 was s 1(2)(b) of the Housing (Homeless Persons) Act 1977, which said a person is deemed to be homeless if ‘it is probable that occupation of [his residence] 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’.  There was no use of the term ‘domestic violence’ in this section, although the term does feature in s 5(1)(iii).  The 1977 Act was consolidated in 1985 and this in turn was recast in the 1996 Act, s 177(1), which used the term ‘domestic violence’ but defined it, just as in the 1977 precursor, to mean ‘violence from a person with whom he is associated, or threats of violence from such a person which are likely to be carried out.’  The section was amended by the Homelessness Act 2002, which inserted a new s 177(1A):

 (1) 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 – (a) a person who normally resides with him as a member of his family, or (b) any other person who might reasonably be expected to reside with him.

 (1A) For this purpose ‘violence’ means – (a) violence from another person; or (b) threats of violence from another person which are likely to be carried out; and violence is ‘domestic violence’ if it is from a person who is associated with the victim.

The new formulation in 1996 made clear it extends to violence against one’s family members.  The 2002 amendment includes ‘other violence’, extending protection to those at risk of ‘violence’ from non-associated persons; but the definition of violence is the same.

Lady Hale gave the lead judgment, arguing that s 177(1) now extends to harmful or abusive action at large.  She said physical violence is only one of the natural meanings of ‘violence’ (another is intensity of feeling and passion).  By 1996, when the term ‘domestic violence’ is used, Lady Hale notes, there is a consensus amongst national and international governing bodies that ‘domestic violence’ is more than just ‘physical violence’.  She argues that:

 ‘…whatever may have been the original meaning in 1977 … 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.’ [24]

Her main support for this claim is a major Home Office report in 2005.  This change in understanding, Lady Hale says, is relevant because ‘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.’  She relies on Lord Steyn’s opinion in R v Ireland [1998] AC 147 for this proposition.  She then refers to Lord Clyde and Lord Slynn’s opinions in Fitzpatrick v. Sterling Housing Association Ltd [1999] 3 WLR 1113 , the latter citing Bennion in support.  Lady Hale notes that the Fitzpatrick decision involved changes in relation to the word ‘family’.  She then argues that ‘violence’ is similar to family: it is not technical and its meaning may change over time.  The essential question, Lady Hale says, is whether an updated meaning is consistent with the statute’s purpose.  She concludes:

‘…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 [to include] ‘physical violence, threatening or intimidating behaviour and any other form of abuse which, directly or indirectly, may give rise to the risk of harm.’ [28]

Lady Hale recognises some problems with this interpretation, which I note below. Interestingly, Lord Brown too, in his quasi-dissent, refers to the ‘always speaking’ approach or the Fitzpatrick principle and says the question is whether the court should apply it here.  He suggests various reasons why ‘violence’ means ‘physical violence’ yet does not dissent.

Why object to updating?  The problem with updating is that it is inconsistent with legislative authority. The statute has been chosen by the legislature and may not be amended by any other. Updating the meaning of the statute, in the sense I have discussed, is to amend it. It is irrelevant that some legislators prefer updating, for it is inconsistent with authority. The updating doctrine is an illicit Henry VIII clause of uncertain application. This means the content of the law changes when word meaning changes, which is arbitrary. This is a very poor way to decide when or if to change the law for there is no careful consideration, by relevant authority, of whether this change is warranted. Instead, changes turn on (are mediated through) arbitrary word manipulation and then are applied retrospectively.  It is uncertain when or if the content of the law is liable to updating or will be changed. And it is very difficult for the legislature to avoid its decisions being amended in this way.

Let me return to Yemshaw.  I take it the legislature’s intended meaning in defining ‘violence’ as it did was to convey the type or class ‘physical violence’, the deliberate infliction of force.  The legislature’s choice of language in 1977, 1996 and 2002 make this the obvious inference. If ‘violence’ includes psychological abuse, the term should include threats whether likely to be carried out or not.  Yet the statute limits threats to those likely to be carried out. Further, it is awkward to speak of threats to carry out psychological abuse. Moreover, ‘violence’ is constant whether the violent person is associated or not, yet it seems implausible for Parliament to proscribe other harmful action from non-associated. Also, as Lord Brown says, that this is an emergency/deeming provision supports this reading.  Updating the meaning by reference to understandings of ‘domestic violence’ at large and post-enactment is to amend this statute to make it conform to those understandings. That is, it ignores the choice the legislature made and instead takes advantage of word choice.

The court says that the statutory purpose constrains. However, the problem here is that the statutory purpose is being picked out quite arbitrarily. The court says the purpose is to protect people from harm. And this is true to the extent that violent action is a subset of harmful action. But one might think the purpose was to protect people who are at risk of physical violence. Lord Rodger says that to fail to extend this would be to downplay psychological abuse, but this is an argument for extension to harm at large (say, drug use, criminal associates). And indeed, it looks very much like extending the statute by analogy. I note also that Lady Hale answers the argument that her reading extends to abusive conduct from non-associated persons by saying there is a threshold of seriousness. This introduces into the provision an otherwise absent, qualitative standard. This confirms, I think, that the interpretation in this case was illicit judicial amendment.  There is no good reason to update statutes in the sense I’ve discussed.  Indeed, updating statutes is unconstitutional and the courts should stop doing it.


This post is extracted from a lecture given to the Statute Law Society, 9 May 2011. 


Richard Ekins is a Senior Lecturer at the Faculty of Law, The University of Auckland.