JH v Secretary of State for Work and Pensions and LH (CSM): [2016] UKUT 440 (AAC) [2016] UKUT 0440 (AAC)

UPPER TRIBUNAL
ADMINISTRATIVE APPEALS CHAMBER
[2016] UKUT 0440 (AAC)
Between
JHAppellantSecretary of State for Work and Pensions and LHRespondent
Before
. TheDate 6 October 2016Category: Child supportSubcategory: Child support other
[1]This appeal is about the bright line rule that applies to shared care in the child support system. The present appeal is, in many ways, an exemplar of what many critics say is wrong with the statutory child support scheme. Two obviously intelligent and articulate parents are arguing over a matter of a night or two of shared care over the course of a year. Be that as it may, the parents are entitled to a judicial determination of their respective rights and liabilities according to the law.[2]In the family courts, decisions about the residence of children and their financial support (insofar as the courts still have the power to decide such matters) are made by judges applying a broad set of principles. Those various considerations are typically described as “guidelines not tramlines”, indicating the general direction of travel but not prescribing a particular destination. Such a discretion-based scheme obviously has both advantages and disadvantages.[3]The statutory child support scheme operates under a very different framework. The formula means the formula, with a narrow scope for variation and only a restricted role for discretion-based decisions. Thus the child support regime is, by and large, a system of tramlines, with very limited scope to jump the tracks and head in a different direction. Changing the metaphor somewhat, the child support scheme operates on a system of ‘bright line’ rules with minimal flexibility to depart from prescribed outcomes. This approach also has both (different) advantages and disadvantages. JH v(1) Secretary of State for Work and Pensions(2) LH (CSM) [2016] UKUT 0440 (AAC) CCS/227/2016 2 The Upper Tribunal’s decision on this appeal in summary[4]In summary, the father’s appeal to the Upper Tribunal is allowed. The decision of the Leeds First-tier Tribunal on 29 September 2015 involves an error on a point of law. The Tribunal’s decision is therefore set aside. None of the parties has requested an oral hearing before the Upper Tribunal. I am satisfied that it is fair and just to determine the case on the papers. I proceed to re-make the original decision under appeal myself. My final decision is to the same effect as that of the First-tier Tribunal decision that has been set aside, albeit I have arrived at it by a different route. The relevant legal provisions central to this appeal[5]As noted above, this appeal concerns the proper application of the rules on shared care in child support cases. In particular, it concerns the proper application of regulation 7 of the Child Support (Maintenance and Special Cases) Regulations 2000 (SI 2001/155; “the 2000 Regulations”).[6]Under the Child Support Act 1991 (“the 1991 Act”) – or at least the version that applies to the current case – the amount of child support payable is reduced where there is shared care, providing the minimum threshold of at least 52 nights a year is met. Paragraph 7(4) of Schedule 1 to the 1991 Act (as amended by section 1 of, and Schedule 1 to, the Child Support, Pensions and Social Security Act 2000) provides as follows:
“(4) The amount of that decrease for one child is set out in the following Table – Number of nights Fraction to subtract 52 to 103 One-seventh 104 to 155 Two-sevenths 156 to 174 Three-sevenths 175 or more One-half”
[7]These various adjustments are known as “bands”. So where the child stays with the non-resident parent for between 156 and 174 nights a year inclusive, the case falls within band 3. Where the child stays over for 175 nights or more, the case becomes band 4.[8]So then the question is the period over which such shared care is measured. Paragraph 7(3) of Schedule 1 to the 1991 Act (as amended) provides that “there is to be a decrease according to the number of such nights which the Secretary of State determines there to have been, or expects there to be, or both during a prescribed twelve-month period.” Thus the Act contemplates that the decision-maker may look backwards in time, or forwards, or a combination of backwards and forwards.[9]Regulation 7(3) of the 2000 Regulations then stipulates as follows:
“(3) Subject to paragraph (4), in determining the number of nights for the purposes of shared care, the Secretary of State shall consider the 12 month period ending with the relevant week and for this purpose ‘relevant week’ has the same meaning as in the definition of day to day care in regulation 1 of these Regulations.”
[10]So the default position under the 2000 Regulations is that the decision-maker looks backwards in time. This rule, however, is subject to the qualification in regulation 7(4) of the 2000 Regulations: JH v(1) Secretary of State for Work and Pensions(2) LH (CSM) [2016] UKUT 0440 (AAC) CCS/227/2016 3 “(4) The circumstances in which the Secretary of State may have regard to a number of nights over less than a 12 month period are where there has been no pattern for the frequency with which the non-resident parent looks after the qualifying child for the 12 months preceding the relevant week, or the Secretary of State is aware that a change in that frequency is intended, and in that case he shall have regard to such lesser period as may seem to him to be appropriate, and the Table in paragraph 7(4) and the period in paragraph 8(2) of Schedule 1 to the Act shall have effect subject to the adjustment described in paragraph (5).”[11]It follows that the starting point must always be the degree of shared care over the period of 12 months before the “relevant week” (on which see regulation 1(2) of the 2000 Regulations) – see regulation 7(3). However, in the two types of case specified in regulation 7(4) a lesser period than 12 months may be taken, ending with the relevant week – see further the careful and comprehensive analysis by Judge Jacobs in Child Support Commissioner’s decision CCS/2885/2005 (at paragraph 20). The practical significance of the decision on shared care[12]In the present case the dispute centred on whether the case of the couple’s two daughters fell within band 3 or band 4. The parent with care (“the mother”) said that band 3 applied (156-174 nights). The non-resident parent (“the father”) said that band 4 applied (175 nights or more). The Child Support Agency and the First-tier Tribunal (“the Tribunal”) both decided this was a band 3 case. The background to the child support appeal[13]The father received child benefit for the older daughter and the mother received child benefit for the younger daughter. I note the mother’s observation that the father has that award notwithstanding that he is a higher rate tax payer. She assumes “he has done this to either simply stop me receiving it, to use it as a way to ‘back up’ his CSA claims, or both”. I do not need to and do not make any finding on that allegation, but I return later to the child benefit issue.[14]On 18 May 2011 the parents agreed a shared residence consent order in the county court. The consent order spelt out in extensive detail over two sides of A4 the practical arrangements for shared residence. This included a four-weekly cycle or rota of shared care in term time followed by separate provision for the Easter, summer and Christmas school holidays and again separate rules for the various school half-terms.[15]By way of background, in an earlier appeal on 25 June 2013 a First-tier Tribunal decided that both parents were ‘persons with care’ under the terms of the Child Support Act. However, that Tribunal also found that the father “provided care (as defined) to a lesser extent than [the mother] and is therefore a non-resident parent (as defined)”. It did not determine precisely how many nights of shared care were attributable to the father. That Tribunal decision was not appealed to the Upper Tribunal. The Child Support Agency’s decision under appeal to the First-tier Tribunal[16]On 7 November 2014 the Child Support Agency made a decision that the father was liable to pay £101.14 a week in child support maintenance. This calculation included a band 3 shared care allowance (i.e. 156-174 nights). The father had submitted a detailed log to the CSA showing that the children had stayed overnight with him for 176 nights in the year to 15 April 2013. The mother’s own log showed that the children had stayed with their father for 169 nights (not adjusting for holidays) or for 172 nights (having adjusted for holidays) in the same 12 month period. Both JH v(1) Secretary of State for Work and Pensions(2) LH (CSM) [2016] UKUT 0440 (AAC) CCS/227/2016 4 parents insisted that their calculation was in accordance with the terms of the consent order.[17]The CSA decision maker clearly recognised that s/he had a difficult decision to make:
“The shared care contact order has not changed, and depending on when Easter falls, and which 52 weeks we consider, the number of nights NRP has the QCs [qualifying children] could tip over the 175 nights. However, the intention of the contact order is the same every year and it is a cycle that repeats itself, the only major change in the yearly schedule is that Easter falls at a different time each year. The effective date of 15th April does not make this decision straightforward.”
[18]The decision-maker concluded that the shared care in the relevant 12 month period amounted to 169 nights and so fell within band 3. The explanation given on the file was as follows:
“Refuse to revise the shared care bandwidth from Band 3 to Band 4, given the period considered as 52 weeks prior to the effective date, which could be slightly higher as Easter came early in 2013. The contact order, that both parties agree to adhere to, has not changed since the last appeal when he stated he has 172 nights shared care and [Tribunal] directed [mother] was the PWC. As NRP has [child benefit] for [older daughter] this would close the case for her, whilst next year this may change back to PWC having more care. There are too many determining factors on a contact order that borders on the cusp of equal shared care.”
[19]I simply observe here that the father denies that he ever told the previous tribunal on 25 June 2013 that the children stayed with him for 172 nights a year. He said his evidence was that they had stayed for 177 nights for the 12 months then in issue (this difference would not have affected the outcome in that appeal, given the mother would still be providing more care over 365 days). The only document from those earlier proceedings on the present appeal file is the tribunal’s decision notice, which unsurprisingly does not deal with that point. I will assume for present purposes he did not make that statement attributed to him by the Agency. The First-tier Tribunal hearing[20]The Tribunal held an initial hearing on 18 May 2015 but this was adjourned to allow the parties to address further evidence that had been produced by the mother. The Tribunal reconvened on 29 September 2015. The District Tribunal Judge explored the possibility of an agreed resolution to the appeal (entirely in keeping with the spirit of rule 3 of the Tribunal Procedure (First-tier Tribunal) (Social Entitlement Chamber) Rules 2008 (SI 2008/2685)). In the colloquial language of lawyers, this process is sometimes described as “the judge banging heads together”. The Judge suggested simply adopting the court order as the basis for the resolution of the dispute. The CSA’s presenting officer canvassed another possible compromise, i.e. agreeing band 3 for one child and band 4 for the other. There was a short break for the parties to consider these options but none could be agreed.[21]The hearing accordingly went ahead, and continued for very nearly a further hour. There was detailed discussion about the terms of the consent order and the parties agreed at each step that those terms were complied with. The Agency’s presenting officer stated that the 12 month shared care assessment period ran from JH v(1) Secretary of State for Work and Pensions(2) LH (CSM) [2016] UKUT 0440 (AAC) CCS/227/2016 5 16 April 2012 to 15 April 2013 and the decision maker had concluded that the extent of the father’s shared care amounted to 169 nights. He added that if the Tribunal was satisfied the court order was followed and reflected the shared care in practice then it was a reasonable basis on which to proceed.[22]The mother argued that the court order was followed and that was a fair way on which to deal with the case. She added that the father’s shared care for each of the years from 2012 to 2016 inclusive had ranged from 173 (2012 and 2016) through 174 (2013 and 2015) to 175 (2014). The father responded by arguing it was wrong to use calendar years.[23]The father stated that in principle he did not object to following the court order. However, he made this subject to a qualification (according to the record of proceedings):
“… but because of when holidays start in relation to days of the week it produces different results. The reason for the difference is that I take the holidays to start on a Friday night and this results in me getting more Fridays than the order would suggest. Some years get extra days, others don’t, but don’t lose days.”
[24]Working on a shared care assessment period from 16 April 2012 to 15 April 2013, the father argued at the hearing that he had shared care for 176 nights. The mother’s case at the hearing, on reflection, was that his shared care amounted to 174 nights over that same period. The First-tier Tribunal’s decision on the father’s appeal[25]The Tribunal’s decision notice, issued on the day, dismissed the father’s appeal and provided the following summary of reasons:
“There is a court order which both parties agree is followed. The order, made by consent, provides a fair basis for the calculation of shared care. That order provides for a 4 weekly cycle. In this cycle [the father] has overnight care for 13 out of 28 nights. Holidays are treated differently with the intention of equal shared care during most of the holiday periods with the exception of May half term and 2 weeks during the summer holiday. These periods are included in the 4 weekly cycle. At Easter [the father] has one extra night. The shared period represents 10 weeks which gives [the father] 36 nights. The remaining 42 weeks are part of the 4 weekly cycle. There are 10.5 cycles giving [the father] either 136 or 137 nights. The total is therefore either 172 or 173 nights and this places shared care in band 3.”
[26]So whereas the father had argued all along for 176 nights, and the mother had eventually conceded 174 nights, the Judge came down on the mother’s side of the bright line at 172 or 173 nights.[27]The Judge subsequently issued a statement of reasons. This set out, in a rather less compressed fashion than the decision notice, his findings of fact (at paragraph [6]) as to how he had arrived at a figure of 172 or 173 nights. “a. There is a consent order (p.31). This is set out in a four weekly cycle in respect of term times and is, in general, followed by both [parents]. b. During term times, excluding half terms, the four weekly cycle gives [the father] two overnight contacts on each of weeks 1 and 3, four overnight contacts JH v(1) Secretary of State for Work and Pensions(2) LH (CSM) [2016] UKUT 0440 (AAC) CCS/227/2016 6 on week 2, and 5 overnight contacts on week 4. This gives a total of 13 nights over a 28 night cycle. c. Holidays and half terms are treated differently in the consent order, except for two weeks in the Summer holidays and the May half term which are both part of the term time four weekly cycle. d. There are 13 weeks of holiday or half-term. 6 weeks in the summer, 2 weeks at Christmas and Easter and 3 half terms. Of these 13 weeks, 3 are included in the four weekly cycle. This gives an overall period of 42 weeks in the year, which is governed by the four week cycle. This is 10.5 cycles. This gives [the father] in one year 136 nights in this period and in the following year 137 nights. The 10 ‘holiday’ weeks are shared equally. This is 70 nights and would give [the father] 36 nights. e. The total, under the Consent Order, is either 172 or 173 overnight contacts for [the father] in each year. Band 3 is set out as 156-174 nights. [The father’s] overnight contact with his children puts him in Band 3.”[28]The Judge’s reasons were as follows: “7. [The father] does not agree with this method of calculation. In his letter of appeal he states that he has his children to stay overnight on more than 175 nights and therefore should be assessed on the basis of being in band 4. He provided a spreadsheet (pp.12-19) which he states shows that he had overnight contact of 176 nights. There are well over 100 pages of diaries, spreadsheets and other evidence as to the pattern of shared care submitted by both parents which give differing levels of shared care. From this it is clear that both parents follow the consent order in terms of shared care.[8]The starting point for the assessment of shared care is the 12 month period ending with the relevant week. In this case there is a pattern, set out in the Consent Order and which both parties agree is followed. That pattern when calculated places the father in Band 3.” The application to the First-tier Tribunal for permission to appeal to the Upper Tribunal[29]The father lodged an application for permission to appeal to the Upper Tribunal. His main point was that the Judge had, as he put it, “used a theoretical calculation rather than the effective 52 week period to calculate the number of overnight contacts as prescribed in the law.” He reasserted his argument that he had 176 nights of shared care in the relevant period and so should have been placed in band 4.[30]The First-tier Tribunal gave the father permission to appeal. The Judge (the same Judge who conducted the hearing) added this comment to his ruling giving permission:
“The pattern of shared care changes depending on the date taken as the effective date for this calculation. To avoid a situation where repeated applications could be made to vary the shared care band, Regulation 7(4) has been applied and this is possible because the level of shared care varies. It is common ground that the number of nights varies between 172 and 176. The tribunal considered that the correct method was to reflect the consent order. This provides for consistency and certainty.”
JH v (1) Secretary of State for Work and Pensions (2) LH (CSM) [2016] UKUT 0440 (AAC) CCS/227/2016 7[31]I simply observe that there was no reference in either the Tribunal’s decision notice or its statement of reasons to regulation 7(4). Nor was there any mention in either document of any variation in the number of nights varying across different years, although this had undoubtedly been discussed at the hearing and in the prior correspondence. The Upper Tribunal’s analysis of the appeal Introduction[32]The Secretary of State’s representative has at my invitation made two submissions on this appeal; both parents have each had and taken the opportunity to comment in turn. I have considered all their comments. In this decision I focus only on those matters which have been directly relevant to resolving this appeal. What was the correct 12 month period for the purposes of regulation 7(3)?[33]The shared care assessment period is “the 12 month period ending with the relevant week” under regulation 7(3). There was some confusion in the papers as to whether that 12 month period ended on April 14, 2013 or April 15, 2013. I agree with the submission by Mr O’Kane, the Secretary of State’s representative, that as the effective date of the decision was April 15, 2013, the relevant 12 month period runs from April 15, 2012 to April 14, 2013. In the event, however, irrespective of whose diaries or calendars are used, this one day makes no difference to the outcome of the calculation. Can the Upper Tribunal have regard to the further reasons given by the First-tier Tribunal in the permission ruling?[34]The short answer is no. It is always easy to be wise after the event. There may be some circumstances in which it may be permissible to supplement the reasons originally given by a tribunal (see Barke v SEETEC Business Technology Centre Ltd [2005] EWCA Civ 578). However, the general rule is “best leave well alone”. Mr O’Kane has helpfully reminded me of the observations of the Court of Appeal in Brewer v Mann [2012] EWCA Civ 246 at paragraph [31]:
“…However, we feel that we can provisionally state that where a judge has received no request from the parties to reconsider his judgment or add to his reasons, and has not demonstrated the need in conscience to revisit his judgment, but on the contrary has received grounds of appeal and an application for permission to appeal on the basis of the alleged inadequacies of his judgment, then it would be most unwise for him to rewrite his judgment (other than purely editorially) and it would take the most extraordinary reasons, if any, to justify such a course on his part. It is also plain to us that this was not the case of a short judgment on a straightforward issue where an appeal might be avoided if the judge supplied further reasoning which had been requested of him.”
[35]There was no request in this case for additional explanatory reasons. There was, on the contrary, an application for permission to appeal. I therefore agree with Mr O’Kane that I cannot have regard to the Judge’s further discussion of regulation 7(4) in the permission ruling in deciding whether the original substantive decision involves any error of law. Understandably neither parent has made any submissions on this rather technical legal issue. The parties’ submissions on the appeal to the Upper Tribunal JH v(1) Secretary of State for Work and Pensions(2) LH (CSM) [2016] UKUT 0440 (AAC) CCS/227/2016 8[36]I think I can fairly summarise the parties’ respective arguments as follows.[37]The father’s position as set out in his original application for permission to appeal is summarised at paragraph 29 above. In his reply to the other parties’ responses, he reiterated his argument that the Tribunal erred by not looking at the actual number of nights of shared care between April 15, 2012 and April 14, 2013, which he contended was a representative period. He reasserted his case that he had 176 nights of shared care in that prescribed 12 month period.[38]The mother’s position is that the Judge had correctly calculated the extent of shared care by reference to the consent order, so reaching a decision based on a document that had been followed by both parties for more than five years. She argued that this was the only way to arrive at a correct assessment and to achieve some finality to the dispute.[39]The Secretary of State’s position has changed. Mr O’Kane’s first written submission was to the effect that the Tribunal was entitled to reach the decision it had, having established that both parents had agreed that they abided by the terms of the consent order. However, in my subsequent directions on the appeal, I posed the following question for the Secretary of State (and for the other parties) to comment on (with emphasis as in the original):
“The Tribunal relied on the terms of the consent order. Assuming for the present that there was no error in the arithmetic, was that on any basis consistent with the requirement in reg.7(3)? After all, was not the Tribunal adopting a January-December view of the consent order, and not an assessment of how the consent order played out on an April to April basis for the specific period in question (which may be more affected by e.g. timing of Easter)? Put another way, was it incumbent on the Tribunal to make a specific ruling as to which parent’s account of the actual number of nights of shared care in the year to April (14 or 15) 2013 that it accepted?”
[40]In his second submission Mr O’Kane resiled from (i.e. abandoned) the stance he had adopted in his first response. Now he accepted that it was arguable the Tribunal had erred in law by failing to make adequate findings of fact as to the actual number of nights of shared care there were for the relevant year, assessed on an April to April basis. Does the First-tier Tribunal’s substantive decision involve any error of law?[41]With considerable regret I have come to the conclusion that the Tribunal’s decision does indeed involve an error of law. Having ascertained that both parties said that they followed the terms of the consent order, the Tribunal then used the consent order to calculate the extent of the shared care. That may seem eminently sensible. The problem, however, was that while both parents said that they followed the consent order, they did not agree with each other as to precisely what that order required. There was, for example, an ongoing disagreement as to what the terms of the consent order mandated over the Christmas period. That disagreement (albeit at the margins) was perhaps understandable given the complexity of the consent order. However, there was clearly a factual dispute between the parents as to how many nights of overnight contact the father enjoyed in the particular 12-month assessment period from April 15, 2012 to April 14, 2013. The Tribunal should have resolved that dispute, one way or the other, rather than resorted to the hypothetical calendar year assessment under the terms of the consent order. JH v(1) Secretary of State for Work and Pensions(2) LH (CSM) [2016] UKUT 0440 (AAC) CCS/227/2016 9[42]There is also, of course, ample authority for the proposition that when deciding issues of day to day care and shared care decision-makers must have regard to what actually happens on the ground, and not simply rely on the terms of a court[43]Having found there to be an error of law in the Tribunal’s decision, I may – but do not have to – set aside that decision (Tribunals, Courts and Enforcement Act 2007, section 12(2)(a)). However, tempting though it may be in certain respects, I think it would be an abdication of judicial responsibility in the circumstances of this case simply to leave the Tribunal’s decision intact. I therefore set aside the Tribunal’s decision. I must then do one of two things.[44]One option is to remit the matter to a new First-tier Tribunal for its reconsideration (see Tribunals, Courts and Enforcement Act 2007, section 12(2)(b)(i)). There is understandably no enthusiasm from the parties for me to adopt that course of action. It may not be possible to achieve finality, but that course of action would certainly do nothing to assist in that process.[45]The alternative option is for me to re-decide the case (see Tribunals, Courts and Enforcement Act 2007, section 12(2)(b)(ii)). I have concluded that is appropriate here and entirely consistent with the overriding objective. There is extensive written evidence and I do not believe that anything new would emerge were the matter remitted to the First-tier Tribunal for re-hearing. With the best will in the world, any new oral evidence now is likely to be self-serving. No party has sought an oral hearing before the Upper Tribunal and I do not consider that one is necessary. I bear in mind that I can make any decision the First-tier Tribunal could have made and may also make such findings of fact as I consider appropriate (see Tribunals, Courts and Enforcement Act 2007, section 12(4)). I therefore proceed to re-decide the case afresh. The Upper Tribunal’s re-making of the decision under appeal Introduction[46]The re-making of the decision essentially proceeds in two stages. First, what is the outcome under regulation 7(3)? Second, is any different outcome appropriate in the light of regulation 7(4)? The application of regulation 7(3)[47]It is agreed on all sides that the relevant assessment period is April 15, 2012 to April 14, 2013. The father’s case is that he had shared care for 176 nights over that 12 month period. The mother’s case (at the outset) was that he had shared care for 172 nights over that period, although by the Tribunal hearing she had revised this upwards slightly to 174 nights.[48]The parents’ respective calendars showed that there was agreement on the actual number of nights concerned in eight of these months (April-July 2012, September-November 2012 and February 2013).[49]In the course of correspondence before the First-tier Tribunal hearing the father reconciled the original disagreement between the parents as follows (letter dated 27 January 2015 at p.113). He identified 9 days (or, strictly, nights) when the mother had JH v(1) Secretary of State for Work and Pensions(2) LH (CSM) [2016] UKUT 0440 (AAC) CCS/227/2016 10 claimed the children were with her when in fact he argued they were staying with him. These 9 nights were 1 and 2 January 2013 (evidenced by e-mails) and 31 March-6 April 2013 (as per the consent order). He also referred to 5 nights when the mother’s calendar said the children were with him when in fact they were with her. The net difference of 4 nights accordingly accounts for the overall disagreement between 176 (he says) and 172 nights (she says).[50]As a starting point, I am inclined to accept the father’s attribution of the disputed nights as set out in the previous paragraph. I do that because the mother accepts that when her phone broke she lost most of her historic records for 2012/13 and so lost access to her relevant diary entries and e-mails. Her own account for 2012/13 is accordingly in part a reconstruction of events, as best she can. I can certainly see no reason not to accept the father’s account as to the true status of the 5 nights in respect of which he concedes that the mother had incorrectly attributed those nights of shared care to him.[51]I also accept the father’s statement that the children were with him overnight for the 7 nights from 31 March to 6 April 2013. In 2013 Easter Sunday fell early on 31 March, with school terms running right up to the day before Good Friday. The consent order specified that “from break up of school on the Friday until 12 noon on the following Sunday a week later with the father” (clause 1(b)(ii)(a)). At the hearing the mother accepted that the consent order had the effect of providing for the children to have an unbroken run of 9 nights with father and then 8 nights with her over the Easter break. Accordingly those 7 nights should properly be attributed to the father.[52]This then takes us to the disputed nights of 1 and 2 January 2013. On balance I also accept that these shared care nights were nights the children spent with the father, given that the mother’s calendar cannot be supported by the lost text messages and emails. In addition it appears from the record of the proceedings that at the hearing the mother accepted “these January changes” when making her final adjustment to the figures.[53]This analysis would suggest that the father is correct in arguing that for the relevant 12 month period under regulation 7(3) he had shared care for 176 nights.[54]However, this is not the end of the matter. There is also a disagreement over a date in August 2012. The father states that the children were with him on the night of August 25, 2012. The mother disagrees. The consent order provides as regards summer holidays that the children spend the first 15 days with the father commencing on the first Saturday and then “for 15 days from the Saturday of week 4 with the mother” (clause 1(b)(i)(b)). The father’s calendar has the children staying with him for an unbroken period of 15 nights and subsequently with the mother for an unbroken period of only 14 nights. Given both parties are adamant they comply with the terms of the consent order, I find on the balance of probabilities that the children were in fact with the mother on the night of August 25, 2012. Accordingly the father’s shared care over the relevant 12 months under regulation 7(3) amounts to 175 nights, not 176 nights.[55]Thus applying regulation 7(3) in isolation, the children stayed with the father for 175 nights in the 12 month period ending with the relevant week. On the face of it that is enough (just) to bring him within band 4, which applies where the number of nights of shared care is “175 or more” (1991 Act, Schedule 1, paragraph 7(4)). JH v(1) Secretary of State for Work and Pensions(2) LH (CSM) [2016] UKUT 0440 (AAC) CCS/227/2016 11 The application of regulation 7(4)[56]Regulation 7(3) is subject to regulation 7(4), and its two exceptional cases. As regards the first special category, this is not a case in which there was “no pattern for the frequency with which the non-resident parent looks after the qualifying child for the 12 months preceding the relevant week”. There was a pattern, albeit a highly complex and intricate pattern provided for by the consent order and notwithstanding that there were some differences between the parents at the margins as to how that order should be interpreted and applied. So the first exception did not apply.[57]The second exception to regulation 7(3) applies where “a change in that frequency is intended”. There is a learned analysis of this provision in Child Support Commissioner’s decision CCS/2885/2005 at paragraphs 17-27 (per Judge Jacobs), although those observations were not all strictly necessary for the purposes of deciding that particular appeal. It seems to me the point can be taken fairly shortly here. The question for these purposes is whether the parties intended a change in the frequency of shared care, not in its pattern. As a matter of the ordinary meaning of the statutory language, the frequency is the number of times that something happens within a particular period. In the present case, the parties certainly intended a (marginal) change in the frequency of shared care. The parents’ respective calculations demonstrate beyond question that one matter on which they agree is that the consent order has the effect of producing a different frequency of shared care depending on the particular year or assessment period chosen. They agree it is not always the same figure in the terms of the precise number of nights. The father argues that it is always at least 175 nights, irrespective of the 12 month period chosen. The mother’s contention is that it is sometimes 175 nights or more but more often than not it is less than 175 nights a year.[58]It follows that I am satisfied that the second exception in regulation 7(4) applies.[59]That being so, it follows I “may have regard to a number of nights over less than a 12 month period”. That lesser period must still end with the same relevant week. But which is the lesser period? According to regulation 7(4), I must “have regard to such lesser period as may seem … to be appropriate”. There is, therefore, a double discretion to be exercised – the opening words of regulation 7(4) provide that I may have regard to a shorter period rather than applying regulation 7(3) in isolation and in doing so I must have regard to such lesser period as may seem appropriate. I also remind myself that in “considering the exercise of any discretionary power conferred by this Act” I must “have regard to the welfare of any child likely to be affected” by the decision (1991 Act, section 2). The only two children so affected are the parties’ two daughters.[60]I have given this matter considerable thought. As to the first discretion, I conclude that it is not appropriate to rely solely on the 175 nights as assessed under regulation 7(3). I say that for a number of reasons, not least that 175 nights is on the very cusp of two shared care bands and the parties are agreed that the total number of nights per year varies depending on which 12 month assessment period is selected. I deal with other relevant factors in the exercise of this discretion below.[61]As to the second discretion, I consider that it is appropriate to exclude the Easter and Christmas holiday periods. I exclude Easter – even though the consent order appears to provide for equal shared care – as the date at which Easter falls is notoriously variable and may have a random impact depending on the 12 months selected. I also exclude Christmas because the terms of the consent order provide for the dates of this shared care over this holiday period to vary from year to year. In addition, there is clearly a bitter and perennial disagreement between the parents as JH v(1) Secretary of State for Work and Pensions(2) LH (CSM) [2016] UKUT 0440 (AAC) CCS/227/2016 12 to what the consent order actually means by “the remainder of the holiday being shared equally” (clause 1(b)(iii)(b)) at Christmas. I also recognise that the Easter and Christmas holidays are by definition atypical periods of the year taken as a whole.[62]The father’s calendars state that there were 17 days of school holidays at Christmas 2012 and a further 17 days of holidays at Easter 2013 (i.e. 34 days in total). Excluding all those days brings the 365 days for the 12 month period of April 15, 2012 to April 14, 2013 down to a figure of 331 days. The father had a total of 17 days shared care aggregated over those two periods (8 days at Christmas and 9 days at Easter). Accordingly he had 158 nights of shared care (175 nights less 17 nights) over a revised assessment period of 331 nights (365 nights less 34 nights).[63]Regulation 7(4) then provides that the relevant calculation is “subject to the adjustment described in paragraph (5).” Regulation 7(5) provides that in such a case one must “adjust the number of nights in that lesser period by applying to that number the ratio which the period of 12 months bears to that lesser period”. As a ratio of the lesser period of 331 days, 365 days is (to three decimal points) the figure of 1.103. Adjusting the number of nights (158) by that ratio results in a figure of 174.3 (rounded up to one decimal point).[64]On that basis the adjusted calculation under regulation 7(4) produces a figure that lies at the very top of band 3 (156-174 nights) and does not quite reach band 4 (175 nights or more).[65]I return to the issue of the discretion to be exercised. The father repeatedly makes two points in his various submissions which he argues mean that any discretion should be exercised in his favour.[66]First, and foremost, he argues that as the couple each have child benefit for one child it follows that they have equal shared care and there should be no child support award to either parent. The obvious difficulty with this argument is that child benefit and child support have different objectives and are governed by different statutory schemes. In particular, the child support scheme assesses shared care by reference to the number of nights the child in question stays with a parent. In contrast the child benefit legislation applies a test based on ‘responsibility’ for a child (not confined to nights) and also has a system of priorities between claimants, with HMRC having a residual discretion (Social Security Contributions and Benefits Act 1992, section 143 and Schedule 10). It may well be reasonable for HMRC in the exercise of that discretion to pay child benefit for one child to one parent and to the other parent for another child (see e.g. R (on the application of Ford) v Board of Inland Revenue [2005] EWHC Admin 1109). But there is no direct read across from one scheme to the other.[67]Second, the father argues that the children are already well provided for as the mother has more liquid assets, the benefit of a mortgage-free house and her own income from employment as well as child tax credits. This is also not a persuasive argument. The capital settlement on divorce deals with the position as between the parents; the children are not party to that agreement and the child support liability survives divorce. In addition, one of the principles of the post-2003 child support scheme is that the children (wherever they live and whoever they live with at any given time) are indirectly entitled through child support to a share in each parent’s income throughout their minority. I also bear in mind (though it is not determinative) that any financial provision agreed between the parties and/or ordered by the courts in ancillary relief proceedings is likely to have been made on the basis that the father had a child support liability for both girls. JH v(1) Secretary of State for Work and Pensions(2) LH (CSM) [2016] UKUT 0440 (AAC) CCS/227/2016 13[68]It follows that I do not regard either of the father’s discretion-based arguments as carrying any particular weight. It is well established that in exercising any discretion under the Act it is important to have regard to all the circumstances (see Brookes v Secretary of State for Work and Pensions and CMEC [2010] 2 FKLR 1038). I must also not lose sight of the underlying objective of the child support scheme that nonresident parents should pay maintenance calculated in accordance with the child support scheme (reported decision R(CS) 4/96, at paragraph 8). In exercising my discretion (and in particular as regards the first discretion in regulation 7(4)) I also bear in mind that if the father were to be found to have equal shared care, then the effect would be to classify the mother (who on any reckoning has more nights per year of care) as a non-resident parent by virtue of regulation 8(2)(b)(i) of the 2000 Regulations. That in turn would lead to the closure of her child support claim for the older child, in respect of whom the father has child benefit. That would cause obvious inconvenience if the situation were to change again in the following year when the mother might have more nights of shared care. Taking all these matters into account, in my view it is appropriate(a) to depart from the formulaic result produced by the proper application on regulation 7(3) and(b) to adopt the 12 month period before the relevant week, but disregarding the Christmas and Easter holiday periods. Conclusion[69]For the reasons explained above, the Upper Tribunal allows the father’s appeal. The decision of the First-tier Tribunal is set aside. However, also for the reasons set out above, I re-make the original decision under appeal to the same effect as the First-tier Tribunal. Signed on the original Nicholas Wikeley on 06 October 2016 Judge of the Upper Tribunal

order

see e.g. Child-Villiers v Secretary of State for Work and Pensions [2002] EWCA Civ 1854 at paragraphs [27] per Potter LJ and [34] per Chadwick LJ, the decision of then Chief Commissioner for Northern Ireland in CSC 4/98 at paragraph 20, CCS/2885/2005 at paragraph 9 and PB v CMEC [2009] UKUT 262 (AAC) at paragraph 24. What then should the Upper Tribunal do?