SUEN MO JOEL v. DIRECTOR OF SOCIAL WELFARE

HCAL 117/2012

IN THE HIGH COURT OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

COURT OF FIRST INSTANCE

CONSTITUTIONAL AND ADMINISTRATIVE LAW LIST

NO 117 OF 2012

------------------------

-----------------------

J U D G M E N T

-----------------------

A. INTRODUCTION

1.  The applicant is a recipient of Comprehensive Social Security Assistance (“CSSA”) living in private housing.  He is also a recipient of rental allowance under the CSSA Scheme.  The CSSA Scheme is managed and administered by the Social Welfare Department (“SWD”), which is headed by the Director of Social Welfare (“the Director”).

2.  The CSSA Scheme provides cash assistance to families which cannot support themselves financially to help them to meet basic needs.  Rent allowance is payable as a standard special grant under the scheme to CSSA recipients to meet the cost of accommodation.  The amount of the allowance is the actual rent paid or the prescribed maximum level (“MRA”) by household size, whichever is less.  The MRA is adjusted annually.

3.  It is the applicant’s case that in 1996, the Government adopted a policy on MRA (“the 1996 Policy”) to ensure that the MRA should reflect the actual rent currently paid by 90^th percentile of the CSSA rent paying households living in private housing.  For convenience, I would refer these households collectively as “private rent paying CSSA recipients”, and the objective to cover 90^th percentile of the private rent paying CSSA recipients as “the 90^th percentile objective”.

4.  This is the applicant’s judicial review against the Director’s decisions in 2011 and 2012 to only grant him rental allowances in an amount which did not cover the actual rent paid by him.  The principal basis of the challenge is that these decisions were made not in accordance with the 1996 Policy.

5.  The applicant therefore seeks to quash the Director’s said decisions and a declaration that it is the Government’s policy that the annual adjustment of the MRA should reflect the 90^th percentile objective.

6.  The Director opposes this application.  As would be discussed in detail later, the Director’s main ground of opposition is that as a matter of fact, there was never the 1996 Policy as alleged by the applicant.

B. BACKGROUND

7.  The following background is not in dispute.

B1. CSSA Scheme and rent allowance

8.  The CSSA Scheme is a non-contributory scheme administered by the SWD.  It is to provide financial assistance to bring the income of the needy individuals and families up to a prescribed level to meet their basic needs such as food and clothing.  The CSSA Scheme receives heavy subsidies from the Government in order to carry out these objectives.

9.  There are three types of payments under the CSSA Scheme: the standard rate, supplements and special grants.  Standard rates are provided to help CSSA recipients to meet their basic needs.  Supplements are provided to meet the needs of special groups such as families having members who are old, disabled or medically certified to be of ill health, or single parents.  Recurrent and one-off special grants can be made to CSSA recipients on a need basis so as to tailor assistance to the varied needs of individuals and families.  At present, there are various special grants payable to eligible CSSA recipients to cover rent, water and sewage charge, schooling expenses, childcare centre fees and burial expenses, etc.  They are classified as standard special grants.

10.  Rent allowance is one of these standard special grants payable to eligible CSSA recipients to meet the cost of accommodation. As mentioned above, the amount of the allowance is the actual rent paid or the MRA, whichever is the less.

B2. The 1996 CSSA Review Report and the annual adjustment of the MRA

11.  Prior to 1996, the MRA were set at 150% of the highest public housing rents according to the size of the households.  The actual rent paid by CSSA recipients living in public housing estates or subvented residential institutions were thus fully covered by the rent allowance.

12.  In 1994, the Government decided to carry out a detailed examination of the social security arrangements in order to examine how effectively both the social security system and the benefits it provided were meeting the needs of the recipients.

13.  In March 1995, a Steering Group chaired by the then Director was established to review the CSSA Scheme.  The members of the Steering Group also comprised of senior representatives from the SWD, the Health and Welfare Branch (“HWB”)[1], the Finance Branch (“FB”)[2] and the Census and Statistics Department[3].

14.  In March 1996, the Steering Group published its Report on Review of the CSSA Scheme (“the 1996 Report”).  Chapter 9 of the 1996 Report dealt with the review on rent allowance.

15.  At paragraph 9.7, it is stated that the rent allowance was set with the intention of covering the rent paid by households living in private housing up to a “reasonable level”.  A ceiling based on household size would ensure that the amount of rent paid was reasonable.  It further stated that the maximum levels of rent allowance (ie, the MRA) should be able to cover the actual rents paid by the great majority of CSSA households living in appropriately sized accommodation.

16.  Then, at paragraph 9.10, it was recommended that:

(1) The current MRA should be increased to reflect the actual rent now paid by 90^th percentile of the private rent paying CSSA recipients; and

(2) These levels would be reviewed annually to assess whether the MRA were still achieving the objective of covering 90% of CSSA rent paying households in private housing.  It is also stated that this new formula was more effective in achieving the objective set out at paragraph 9.7 and was relatively simple.

17.  The MRA at 1 April 1996 was then set at an amount reflecting the actual rent then paid by 90^th percentile of the private rent paying CSSA recipients after being approved by the Finance Committee of the LegCo (“FC”).

18.  In April 1997, the FC approved the increase of MRA for the 1997-98 financial year by 8% pursuant to the movement of the Consumer Price Index (A) rent index for private housing (“CPI(A) rent index”) as sought by the Government.

19.  In April 1998, the FC also approved (as sought by the Government):

(1) The increase of the MRA for the financial year 1998-99 at the rate pursuant to the CPI(A) rent index;

(2) The delegation of authority to the Secretary for the Treasury (“ST”) to revise annually the MRA in future in accordance with the movement of the CPI(A) rent index.

20.  Since then, the MRA has been consistently reviewed annually in accordance with the movement of the CPI(A) rent index.

21.  The adjustments of the MRA by reference to the CPI(A) rent index for the financial years in 1996/97, 1997/98 and 1998/98 did cover the rent then paid by 90^th percentile of the private rent paying CSSA recipients.  However, this was no longer the case in recent years.

B3. The applicant

22.  The applicant had worked in the construction industry, but became unemployed in August 2009 after suffering a work-related injury.

23.  He began to receive CSSA in February 2011 and rent allowance in August 2011.  The monthly rent he had to pay then was $1,600.  He received the MRA for a single person household which was $1,265. The applicant therefore had to apply part of the standard rate under CSSA ($1,890 in 2011) he received to cover the shortfall between the rent of $1,600 and the MRA of $1,265.

24.  In February 2012, he moved to another rented accommodation after being required to vacate his former place.  The monthly rent of the new room was $1,700, which the relevant MRA then was $1,335. Again, the applicant had to apply part of the CSSA standard rate he received to pay the shortfall between the actual monthly rent and the received MRA.

25.  Dissatisfied that he had to use part of the CSSA standard rate to cover the rent, he lodged an appeal to the Social Security Appeal Board in March 2012.  The Board wrote to him on 22 March stating that it was not authorized to review the policy related to the setting of the amount of the rent allowance.  It therefore transferred the applicant’s letter to the Lai Chi Kok Social Security Field Unit under the department for follow-up and provision of appropriate assistance.

26.  The applicant then contacted an NGO after receiving the Board’s abovementioned letter.  He was advised to apply for legal aid, which he did on 29 March 2012.

27.  Legal aid was granted to the applicant on 6 August 2012.  On 7 September 2012, the applicant applied for leave to apply for judicial review.  Leave was granted by this court on 2 July 2013 after hearing submissions from both sides.

C. THIS JUDICIAL REVIEW

C1. Grounds of judicial review

28.  The applicant raises a few grounds in support of this judicial review.

29.  The first two grounds are based on the applicant’s principal arguments that the Director’s relevant decisions on the MRA granted to the applicant were made not in accordance with the 1996 Policy.

30.  Decisions made not in accordance with Government policy is subject to judicial review and the court will make appropriate orders, including an order of mandamus, to hold the decision-maker to the relevant Government policy (see: R v Secretary of State for the Home Department, ex p Urmaza [1996] COD 479; Cathay Pacific Airways Flight Attendants Union v Director-General of Civil Aviation (unreported, HCAL 19/2005, Hartmann J, 25 August 2005) at paragraph 15).

31.  Insofar as the 1996 Policy is concerned, the applicant seeks to run two alternative submissions (and thus two separate grounds) as to what this policy is:

(1) First, the 1996 Policy is that the objective of the MRA is to cover “the great majority of CSSA rent paying households in private housing”.  Counsel for the applicant describes this as the “qualitative policy”.  See paragraphs 13‑16 of the applicant’s skeleton submissions.

(2) Second and alternatively, as I mentioned above, the policy is that the MRA should reflect the 90^th percentile objective to cover rent paid by 90^th percentile of the private rent paying CSSA recipients.  Counsel refers to this as the “quantitative policy”.  See paragraphs 17-28 of applicant’s skeleton submissions.

32.  Under these grounds, the applicant says:

(1) In the relevant years of 2011 and 2012, the MRAs did not cover the actual rent paid by 90^th percentile of private rent paying CSSA recipients.

(2) Thus, the Director’s relevant decisions of paying the relevant rent allowances to the applicant in an amount that did not cover the rent paid by him are decisions made not in accordance with the 1996 Policy (whether it is the qualitative or quantitative one).

(3) In the premises, the Director’s decisions are unlawful and should be quashed.

33.  The third ground is only sought to be added to the Form 86 by way of the applicant’s application for amendments made just before the hearing.  This ground relates to the Director’s position as revealed in her evidence that the Government adopted a policy in 1998 to adjust the MRA annually in accordance with the movement of the CPI(A) rent index.  The applicant says if this is shown, it represents a change of policy from the 1996 Policy, which is unlawful for lack of fairness, transparency and certainty in the change (see paragraphs 29 to 34 of the applicant’s submissions).

34.  The fourth ground relates to a constitutional challenge to any change of policy from the 1996 Policy.  This challenge is based on the Government’s accountability to the LegCo and the LegCo’s function to approve public expenditure respectively under Articles 64 and 73(3) of the Basic Law (see paragraphs 35 to 41 of the applicant’s skeleton).

35.  At the hearing, I have indicated to Mr Lo (counsel for the applicant) that I would be prepared to hear the first two grounds and the third ground as provided in the draft Amended Form 86.  However, I would not permit the applicant to rely on the fourth ground (as also objected to by Mr Mok SC for the Director) since this ground was not even pleaded in the draft Amended Form 86 and was only raised for the very first time in the skeleton.  This ground purports to raise serious debates on the scope of the relevant provisions in the Basic Law and may also require the filing of relevant evidence by the Director to meet the accusations.  This is clearly unfair to the Director and should not be condoned[4].

36.  I would therefore in this judgment only deal with the first three grounds of judicial review as follows.

C2. The parties’ positions in gist

37.  All the three grounds of the judicial review are fundamentally premised on the existence of the 1996 Policy.  It is Mr Lo’s submissions that the existence of the 1996 Policy is clearly supported by the contemporaneous documents.

38.  On the other hand, the Director says that the Government has never adopted the 1996 Policy.  Instead, it is the Director’s position that it was only by February 1998 that the Government had confirmed and formulated a policy to adjust the MRA annually by reference to the movement of an inflation-linked rental index, being the CPI(A) rent index. There is thus also no question of any change of policy.

39.  Mr Mok similarly submits that this position is supported by all the contemporaneous documents read together and as whole.

40.  It is therefore clear from the above that the success of the grounds of judicial review depends on whether it can be shown that there is in existence the 1996 Policy.  That, in light of the parties’ arguments, rests on the reading of all the contemporaneous documents put before this court and relied on by the parties.

41.  I would now turn to examine this question.

C3. Whether there is the 1996 Policy

42.  Mr Lo for the applicant submits that the existence of the 1996 Policy is supported and demonstrated by the following documentary evidence.

43.  First, in the 1996 Report, the Steering Group expressly stated at paragraphs 9.9 and 9.10 as follows:

“Recommendations

9.9 There is a case for determining more precisely what percentage of those paying private housing rentals should be fully covered by the maximum rent allowances. Our analysis of rents currently paid by CSSA customers living in private housing shows that for each household size, about 90% of customers pay rent in a narrow range but that about 10% pay markedly higher rents for a variety of exceptional reasons (eg a recently unemployed person who is still accommodated in higher standard housing). This profile of rents paid is shown in Annexes 9.1 to 9.6. This analysis shows that the 90^th percentile of actual rents paid could be taken as a reasonable benchmark for determining the maximum rent allowances.

9.10 It is recommended that the current maximum levels of rent allowance by household size should be increased to reflect the actual rent now paid by the 90^th percentile of the CSSA rent paying households in private housing.  These levels will be reviewed annually to assess whether the maximum levels are still achieving the objective of covering 90% of CSSA rent paying households in private housing.  This new formula is more effective in achieving the objective set out in paragraph 9.7 above and is relatively simple.”

44.  These, Mr Lo submits, unequivocally show that by March 1996, it was the considered view of the Steering Group (with the presence of the representative from the FB) that the MRA should be set at a level by reference to the 90^th percentile objective.

45.  This objective had then been (says Mr Lo) adopted by the Government when the HWB applied to the FC for funding for the 1996-97 financial year to improve the CSSA Scheme with effect from 1 April 1996.  It is demonstrated (Mr Lo further submits) by the relevant FC paper (“the 96-97 FC Paper”)[5] submitted by the HWB, which provided relevantly as follows:

“Members are invited-

a. to approve a package of improvements to the Comprehensive Social Security Assistance Scheme with retrospective effect from 1 April 1996;

…

…

PROBLEM

We need to introduce improvements to the Comprehensive Social Security Assistance (CSSA) Scheme following a comprehensive Review [ie, the 1996 Review] completed recently.

PROPOSAL

2. We propose the following improvements to the CSSA Scheme with retrospective effect from 1 April 1996-

a…

…

d. to increase the levels of maximum rent allowance;

…

…

Maximum rent Allowance

11. The Review concluded that the method of setting the maximum rent allowance at levels equivalent to 150% of the highest public housing rents, previously endorsed by Members, was no longer appropriate for CSSA recipients living in private housing. In order to ensure that the rent allowance would cover the actual rents paid by the great majority of CSSA households living in private housing, the maximum levels should reflect the actual rents paid by the 90^th percentile of these households. To achieve this objective and based on an analysis of rents currently paid by CSSA recipients living in private housing, we propose to increase the levels of maximum rent allowance as follows-

…

BACKGROUND INFORMATION

…

22. The Steering Group completed the Review of the CSSA Scheme in February 1996. Based on the detailed findings and recommendations of the Review, the Financial Secretary announced in his Budget Speech on 6 March 1996 additional improvements to the CSSA Scheme which would cost an extra $200 million a year (including the additional staff required). We published the Report on the Review of CSSA Scheme on 8 March 1996 and distributed it to Members on the same day.” (emphasis added)

46.  Mr Lo says the 96-96 FC Paper shows that the Government (through the HWB) had asked the FC for funding to increase the MRA by reference to and relying on the above recommendations and objective set out in the 1996 Report.  The Government had thus accepted and adopted the Steering Group’s above recommendations on the MRA based on the 90^th percentile objective.

47.  Similarly, when the HWB sought approval for funding for the 1997/98 financial year with effect from 1 April 1997, Mr Lo points out that it was again made on the premise that the increase in MRA sought was to cover the rent paid by 90^th percentile of the private rent paying CSSA recipients (see FC paper FCR (96-97) 118 at paragraphs 2, 11 and paragraphs 2 and 5 of the Supplementary Note incorporated in that paper).

48.  Mr Lo also refers me to a number of internal memos[6] between the ST on the one hand and the then Director or the Secretary for Health and Welfare (“SHW”) on the other hand.  He says these internal memos also show that the ST clearly recognised the existence of the 1996 Policy at the relevant time:

(1) In a memo dated 6 October 1995 (“the 6 Oct 1995 Memo”) from the ST to the Director in relation to the review to be carried out by the Steering Group, the ST said at paragraph 2 that “…In search of a more rational mechanism which would more or less achieve the same original objective of ensuring majority coverage and reflecting actual needs, I think we need to be assured that the level of assistance would not turn out to be driven by the clients’ [meaning the CSSA recipients] aspirations to improve their accommodation…” (emphasis added)  This (Mr Lo says) shows that the ST recognised the objective of the rent allowance is to cover the majority of the actual rent paid by the CSSA recipients.

(2) In a memo from the ST to both the SHW and the Director dated 23 January 1997 (“the 23 Jan 97 Memo”) relating to the annual review of MRA for 1997/98, it was stated at paragraph 2 that “…We also agree that an annual review exercise should be conducted to ensure that the level of [MRA] would continue to achieveyour objective of covering 90% of CSSA rent paying households living in private housing…” (emphasis added)  Mr Lo submits that, again, this shows the ST’s clear recognition of the existence of the 1996 Policy to adopt the 90^th percentile objective.

(3) In another memo dated 19 February 1998 (“the 19 Feb 98 Memo”) from the ST to the Director relating to the MRA, the ST said this “… That said, we continue to have reservations on maintainingthe policy to cover 90^th percentile of CSSA recipients in private housing by MRA.  Notwithstanding your policy review …” (emphasis added)  The ST similarly (Mr Lo submits) recognised the existence of the 1996 Policy although she might have reservations about it.

(4) In a memo dated 24 February 1998 from the ST to the Director (“the 24 Feb 98 Memo”), the ST again said this: “…It will be a non-starter if your review, commenced to study the continued validity of the 90^th percentile policy, ends up recommending changing the inflation adjustment mechanism with effect from 98/99.” (emphasis added)  Similarly, notwithstanding the ST’s reservation about it, she clearly recognised (emphasised Mr Lo) in this memo the existence of the 1996 Policy.

(5) In response to the 24 Feb 98 Memo, the Director in a memo to the ST dated 27 February 1998 (“the 27 Feb 98 Memo”) also referred to (Mr Lo says) the 1996 Policy at paragraph 2, which stated: “Paragraphs 9.9 and 9.10 of the [1996 Report] would suggest to any normal reader that the 90^th percentile policy including annual review and, if necessary, revision by household size is indeed established Government Policy…” (emphasis added)

(6) Finally, in the notes of a meeting to discuss the MRA held on 19 February 1998 at the SWD headquarters (“the 19 Feb 98 Meeting Notes”), it was recorded at paragraph 2 that: “The meeting also discussed the arguments for and against maintaining the ‘90^th percentile’ policy for MRA, and the implications arising from FB’s stand that it would only support an annual inflation-related rental adjustments while funds to support the ‘90^th percentile’ policy were subject to RAE bids.” (emphasis added)  Mr Lo says this similarly shows the existence of the 1996 Policy even though the correctness of that policy was questioned by the ST by then.

49.  Mr Lo therefore submits that the above public and internal documents show and prove the existence of the 1996 Policy (whether qualitative or quantitative).

50.  Mr Lo’s submissions on these documents are initially attractive.  However, after reading all the documents as a whole and in context, I accept Mr Mok’s submissions that they in fact show that the Government had never adopted the 1996 Policy.  I would explain why.

51.  It must be noted that the applicant’s case is that the Government has adopted the 1996 Policy.  Thus, one must examine these documents to see whether the Government has done so, but not whether say the Director or the SHW had in 1996 harboured such an objective or policy in how to administer the rent allowance.

52.  This distinction is important.  As submitted by Mr Mok (and not disputed by Mr Lo), before a proposed policy which has an impact on the public expenditures and thus the Government’s financial position can be confirmed and adopted as a Government policy, the said proposed policy has to have the consensus and agreement of all the relevant and participating policy bureaux, including the FB, after thorough discussions.  This is so as in order to have a policy which is workable and sustainable, it has to have the backing and consensus of the FB to provide a workable financial support to the relevant policy.  The composition of the Steering Group to review the CSSA Scheme as mentioned above also illustrates this position.

53.  For the present purpose, when one is looking at whether a policy has been formulated and adopted by the Government regarding the rent allowance, one has to examine it under the context of whether and if so, when the relevant policy bureaux (ie, the HWB headed by the SHW) and the FB (headed by the ST) had come to an agreement and consensus on this.  In that respect, it is also pertinent to note that the SWD (headed by the Director) is not a policy bureau but the administrative arm of the HWB.

54.  Once looked at in that context, I accept it is clear from reading the following documents that before February 1998, there was no consensus between the SHW and the ST on the adoption of the 1996 Policy.

55.  First, from the internal memos exchanged before the publication of the 1996 Report, they show that the Steering Group had formulated two alternatives objectives on how the MRA should be measured and adjusted annually:

(1) To peg the maximum level of the actual rents paid by the 90^th percentile of private rent paying CSSA recipients from year to year (“the 90^th Percentile Formula”); or

(2) Alternatively, to reflect the actual rents currently paid by the 90^th percentile CSSA recipients to form the “base rate”, and thereafter to adjust them annually according to a rental index so as to maintain the real level of the benefits of those in need (“the Alternative Mechanisms”).

56.  The first of these memos is the 6 Oct 95 Memo, which the ST stated relevantly as follows:

“ Following on discussions at yesterday’s Steering Group meeting, I have given some further thought to the subject of rent allowance under the CSSA.

2. The meeting has reached a consensus that the present mechanism of pegging the maximum level of rent allowance to the highest public housing rent is no longer appropriate. In search of a more rational mechanism which would more or less achieve the same original objectives of ensuring majority coverage and reflecting actual needs, I think we need to be assured that the level of assistance would not turn out to be driven by the clients’ aspirations to improve their accommodation. I am worried that such would be the case if we go simply for a formula which pegs the maximum level to the actual rents paid by the 90^th percentile of the CSSA rent-paying households in private housing from year to year.

3. An alternative mechanism to ensure that we are in better financial control is to revise the maximum levels on a newly devised basis, say reflecting the actual rents currently paid by the 90^th percentile of CSSA recipients and thereafter to adjust them annually according to a rental index so as to maintain the real value of the benefits for those in need. The question then is of course whether we have a rental index that best reflects the type of private property tenements applicable to CSSA recipients. I suggest you seek the advice of CC&S and CR&V on this.” (emphasis added)

57.  The second is a memo dated 13 February 1996 from the ST to the Director, which also says as follows:

“ I wish to respond to the Commissioner for Census & Statistics’ comments on the mechanism for setting and adjusting the maximum levels of rent allowance vide his memo of 12 February.

2. I appreciate the Commissioner’s concern about the precision of the CPI(A) rent index for the said purpose. That’s why in the course of discussion we have also examined rental movements by type of tenements as recorded by CR&V. We eventually settled for the CPI(A) rent index which we feel would be a reasonable basis for adjusting for inflation.

3. The Commissioner’s proposed alternative to set the levels annually based on the actual rents paid by CSSA clients as revealed by the Annual Study of CSSA recipients would not address my very initial concern as expressed in my [6 Oct 95 Memo]. To quote, I said therein that ‘in search of a more rational mechanism which would more or less achieve the same original objectives of ensuring majority coverage …. we need to be assured that the level of assistance would not turn out to be driven by the clients’ aspirations to improve their accommodation … such would be the case if we go simply for a formula which pegs the maximum level to the actual rents paid by the 90^th percentile of the CSSA rent-paying households in private housing from year to year.’ I should just add that the mechanism to set the base rates at an appropriate level and then to adjust it annually for inflation is common. This, of course, does not preclude us from reviewing the base rates in the light of changing circumstances from time to time. This arrangement would also ensure that we are in better financial control and CSSA clients have greater certainty in what rental allowances they would receive.

4. For clarity, I think the agreed arrangement is to introduce the new maximum levels of rent allowance from 1 April 1996 (incidentally, I just notice that the new rates were not set out anywhere in the draft Report) and have these adjusted a year later in accordance with movements in the CPI(A) rent index over the last year. In other words, there is no need to resort to forecast increases in rent for the purpose of inflation adjustment.

5. I am sorry for dwelling on this at some length but rent allowance is a significant selling point in the package of CSSA improvements and is likely to come under close scrutiny by the concern groups.” (emphasis added)

58.  From these memos, it can been seen that the ST was concerned that if the 90^th Percentile Formula was adopted, it would result in a situation where the level of assistance under the MRA would be driven by the CSSA recipients’ “aspiration to improve their accommodation”, which in her opinion was undesirable.

59.  Notwithstanding the above, it appears that the 1996 Report only stated a recommendation on the 90^th Percentile Formula at its paragraphs 9.9 and 9.10 (see paragraphs 43 and 44 above).

60.  This apparent effect of the 1996 Report’s recommendations was in fact noted by the ST just before the 1996 Report was published.  Thus, she wrote a memo dated 1 March 1996 alerting the then Director (who was the chairman of the Steering Group) this omission of the draft 1996 Report to refer also to the Alternative Mechanisms.  The memo says as follows:

“ I have just received by hand a copy of the final version of the Executive Summary and the Report on the Review. I wish to register my concern about the treatment of future adjustments to the rent allowances in this final version, paragraph 45 in the Executive Summary and paragraph 9.10 in the Report refer. My views on the subject have been conveyed to the Steering Group on several occasions, the most recent vide my memorandum of 13 February 1996.

2. This final version has omitted the reference to the agreed annual inflation-related adjustment based on CPI(A) rent index and referred instead to ‘these levels (viz the new levels of rent allowances) will be reviewed annually to assess whether the maximum levels are still achieving the objective of covering 90% of CSSA rent paying households in private housing’. I am concerned that it gives the impression that we would automatically adjust the levels every year according to the 90^th percentile of actual rents paid by CSSA rent paying households in private housing as surveyed. I am afraid this is not the approach we have agreed to fund the future adjustments. What I have agreed, and as recorded in the notes of the 22^nd meeting of the SG, is that future inflation‑related adjustments on the agreed mechanism would not require separate bidding of funds. That is to say, if the CPI(A) rent index shows, say a 10% increase, an across‑the‑board 10% increase to the maximum levels of rent allowance would not require RAE. However, if the annual review suggested indicates that the levels would need to be raised by 20% in order to cover 90% of actual rents paid, additional funds for this purpose would have to be secured.

3. I was told that you intend to arrange separate follow-up discussion on the subject. In the meantime, I should be grateful if SG members who need to explain the Report could take note of the above when asked specifically on this question of future adjustments.” (emphasis added)

61.  The ST reiterated in this memo that the 90^th Percentile Formula was not the approach the FB had agreed to fund the future adjustments.  If the Director wanted to adjust the MRA in the future over and above an inflation-linked adjustment (under the Alternative Mechanisms), the Director would have to bid for any such additional funds in the annual Resource Allocation Exercise (“RAE”).

62.  RAE is an internal Government exercise whereby different departments bid competitively for the available resources for services and projects promoted by those departments.  As such resources are limited and would be exceeded by the bids made by the departments, it would not be possible to fund some of the proposed services projects.  Moreover, in all reality, the prospect of a bid becoming successful is likely to be substantially diminished if it does not receive at least the support of the ST (now the Secretary for Financial Services and the Treasury)[7].

63.  The fact that the recommendations on MRA made in the 1996 Report were an omitted version of what was agreed in the Steering Group by the FB on this matter was also recognised by the then Director in her 27 Feb 98 Memo, where she stated this at paragraph 2 thereof:

“2. Paragraphs 9.9 and 9.10 of the March 1996 Report on Review of Comprehensive Social Security Assistance Scheme (attached at Annex A for easy reference) would suggest to any normal reader that the 90^th percentile policy including annual review and, if necessary, revision by household size is indeed established Government policy. The unfortunate reality, however, is that there never was before the March 1996 publication of that Report or, subsequently, an appropriate and adequate level of agreement within the Administration and specifically between, on the one side, FB and, on the other side, HWB and SWD to justify that assumption. Your proposal has always been that if we do wish to cover the 90^th percentile as a matter of agreed policy then we should make an RAE bid. The public crunch has never yet come because one way or another we have always been able to say, or suggest, that the proposed MRA increases are covering the 90^th percentile anyway. Nonetheless, this is a deeply unsatisfactory situation and one that we should resolve as soon as possible.” (emphasis added)

64.  When all these documents are read together and in proper context, I do not accept Mr Lo’s submissions that the Steering Group had reached a consensus and agreement to adopt the 90^th Percentile Formula as the objective of setting the MRA annually.  To the contrary, I accept that the evidence shows that even by the time of the publication of the 1996 Report:

(1) The 90^th Percentile Formula and the Alternative Mechanisms were both discussed at the Steering Group.

(2) The ST had maintained her views that she favoured and supported the Alternative Mechanisms and did not support the 90^th Percentile Formula.

(3) If the Director wanted to have the MRA adjusted in line with the 90^th Percentile Formula, and if that would result in requiring funding above the inflation-linked adjustment as proposed under the Alternative Mechanisms, the Director would have to bid for such additional funding each year in the RAE.

(4) There was therefore by then no consensus and agreement between the HWB and the FB to recommend to only adopt the 90^th Percentile Formula.

65.  Second, from March 1996 to February 1998, the internal memos exchanged between the ST and the Director or the SHW again show that, although the Director and the SHW wanted to adopt and pursue a 90^th percentile objective or policy in setting the MRA, this was strenuously and consistently opposed by the ST.  The ST instead insisted that the FB would only be prepared to support an annual adjustment of the MRA by reference to the movement of an inflation-linked index (where the CPI(A) rent index was adopted), and if the Director or the SHW wanted to pursue their 90^th percentile objective, they would have to find the necessary additional funding through the RAE on a yearly basis, and the ST would not support the bid.  These memos are as follows.

66.  In a memo dated 23 January 1997 to the Director relating to the annual review of MRA for 1997/98, the ST stated as follows:

“ I would like to set out our position on the subject from the funding point of view.

2. We agree that the maximum rent allowances, following the improved maximum rent allowance introduced in April 1996, should be adjusted in accordance with projected inflation based on CPI(A) rent index for private housing.  Inflation-related adjustments based on such a index would not require separate bidding of funds under the RAE.  We also agree that an annual review exercise should be conducted to ensure that the levels of maximum rent allowance would continue to achieve your objective of covering 90% of CSSA rent paying households living in private housing.  Any improvement sought under this annual review exercise however, will involve securing of additional funds under the RAE.  We reiterated our position on funding in March 1996, as we were concerned about the way the Report on the Review of the CSSA Scheme was drafted.  I enclose a copy of our memo at Annex for your ease of reference. [that memo was the memo dated 1 March 1996 quoted at paragraph 60 above]

3. In the light of your observations about the inadequacy of some (not all) groups of family-size households, it is perhaps more productive if we agree on the annual review exercise to justify and support your bidding of funds in the 1997 RAE.  As for revised rate for maximum rent allowance to be effective as from 1 April 1997, we would advise that we do it on the basis of the CPI(A) rent index.  The previous analysis produced by SWD in December 1996 would be the agreeable starting point for such a purpose.”

67.  Then, in a follow-up memo dated 3 February 1997 to the Director, the ST again emphasised her position as follows:

“ While I appreciate your effort in exploring alternatives to adjusting the maximum levels of rent allowance, I am afraid that from the funding perspective, I am not able to accede to your request for rates of increase above the CPI(A) rent index. You are fully aware of my stance as explained in my earlier memo dated 23 January and at the Steering Committee of SSA review on 24 January. We have particular difficulties with the idea of institutionalizing the annual adjustment with actual rent paid by CSSA households in private housing. Even if your proposal B constitutes your RAE bid, we will challenge the basis of the actual rent data and the principle of pegging adjustments with actual rent paid.” (emphasis added)

68.  In response, the SWH in a memo dated 4 February 1997 to the ST agreed to adopt the CPI(A) rent index as a mechanism to adjust the MRA for the purpose of the 1997/98 funding application.  She also agreed that the issue on the MRA adjustment mechanism required a re‑consideration and examination.  However, she suggested that for the temporary purpose of seeking funding for that year, they would adopt the CPI(A) rent index movement as a basis for adjusting the MRA:

“ Thank you for your memos of 30 January and 3 February 1997.

2. We agree that time is not on our side. For the purpose of the current exercise, we would go for Option A as set out in DSW’s memo of 30 January 1997, ie 1996/97 MRA levels adjusted by 8% in line with the projected increase in the CPI(A) rent index (March 97 over March 96), and process the FCai on the revision of rates for the CSSA and SSA on this basis.

3. The current exercise has brought out an unusual phenomenon which warrants a re-examination/re-consideration of the issue – the adjustment mechanism as well as the funding arrangement. May I therefore suggest that whilst proceeding with the revision of rates for 1997/98, we commence work on the MRA review starting perhaps with the research proposed by S for Tsy in his memo of 3 February and inviting the views of Commissioner of Rating & Valuation and Commissioner for Census & Statistics on DSW’s findings in respect of the actual rent paid by our CSSA recipients as suggested in my memo of 21 January? Your views on how else we should proceed are most welcome.” (emphasis added)

69.  After the FC had approved the 97/98 funding for the MRA increase in accordance with the CPI(A) rent index, the ST wrote to the SHW in a memo dated 1 April 1997 relating to the question of MRA.  She repeated her position that the FB would only agree to provide funding support without RAE bidding to cater for the annual inflation-related MRA adjustment by using the CPI(A) rent index.  She also emphasised that if the SHW wished to improve the MRA to meet the SHW’s favoured 90^th percentile objective, the SHW would have to secure that through the RAE.  The ST further expressed her reservation about that objective and encouraged the SHW to review that policy objective.

70.  The SHW thereafter indeed sought to review the mechanism and funding arrangement for adjusting the MRA.  The Director by a memo dated 10 July 1997 sought the expert advice from the Government Economist on the “need to compile a new series of rental index to be used as reference for annual inflation adjustment”.  In this memo, the Director stated that in setting the 1997/98 MRA, they applied the estimated rate of increase under the CPI(A) rent index.

71.  The Government Economist replied on 18 July 1997 to the Director saying that there seemed to be no apparent need for a separate rent index other than the CPI(A) rent index to be adopted for CSSA recipients living in private households to measure the movement of rents.

72.  At the same time, in the above review on the mechanism and funding arrangement, a longitudinal study of the rent paid by CSSA recipients in private households showed that an adjustment proposed to be made for the 98/99 MRA in accordance with the CPI(A) rent index should cover 96% of the CSSA households living in private tenements.  This was relayed to the ST by the Director to show that the 90^th percentile objective could in fact be generally achieved by adjusting the MRA annually in line with the movement of the CPI(A) rent index.

73.  In response to this study, the ST in her memo dated 19 February 1998 to the Director stated the following:

“ Thank you for copying your longitudinal study of the rent paid by the same batch of CSSA recipients in private tenements to compare their rate of change with the CPI(A) rent index. The findings in Table 1 showing that the proposed MRA for 98/99, after adjustment for CPI(A) rent index by 8.9%, should cover 96% of the CSSA households in the same private housing residence with the same number of eligible members is useful. That said, we continue to have reservations on maintaining the policy to cover 90^th percentile of CSSA recipients in private housing by MRA. Notwithstanding the outcome of your policy review in the light of the longitudinal study and other considerations, the position on funding remains; we would only provide funding without RAE bidding to cater for the annual inflation-related rental adjustment. In the context of the on‑going CSSA review, you will appreciate that any request for further real increase in the MRA rates in the RAE exercise can hardly be justified. I floated to both the bureau and the department earlier that we should seek FC’s approval to delegate the authority to S for Tsy for future revisions to the MRA in accordance with the CPI(A) rent index in the FCai on CSSA scheduled for 20 March 1998. Our assessment is that the proposed delegation is timely and appropriate, and is not likely to be controversial in the FC. Grateful for your agreement to include this proposal in the FCai which you are drafting.” (emphasis added)

74.  Thus, notwithstanding the result of the longitudinal study, the ST restated in this memo her position that she would not support any mechanism for funding the annual adjustment of the MRA other than by an inflated-related rental adjustment.  She further suggested that the Government should seek the FC’s approval to delegate the authority to the ST for future revisions to the MRA in accordance with the CPI(A) rent index in the coming FC meeting on CSSA.

75.  The FB’s above position that it would only support an inflation-linked adjustment mechanism for MRA but not a mechanism to carry out a 90^th percentile policy was fully recognised by the HWB and the Director herself, as reflected in 19 Feb Meeting Notes.  The notes provided relevantly as follows:

“ The Chairperson [i.e., the Director] stated that the purpose of this meeting was to have a general discussion on the method and magnitude of adjustment to the levels of Maximum Rent Allowance (MRA) under the CSSA Scheme, having regard to the data produced by Sr Stat (SW) based on the preliminary results of the 1997 Study of CSSA Recipients. It was noted that the revision of MRA for 1998/99 would be considered in the coming FC meeting scheduled for 20 March 1998.

2. The meeting also discussed the arguments for and against maintaining the ‘90^th percentile’ policy for MRA, and the implications arising from FB’s stand that it would only support an annual inflation-related rental adjustment while funds to support the ‘90^th percentile’ policy were subject to RAE bids.

3. After discussion, the meeting agreed that the draft FCai on the MRA issue should be prepared along the following lines:

· A recently conducted analysis of the rent paid by CSSA households in private housing shows that the 1997/98 MRA levels can broadly cover 90% of such households.

· Based on the past years’ experience with the review of the MRA levels, it has been found that the movement of CPI(A) rent index for private housing is a good indicator of the changes in rent level of CSSA households in private housing. In fact, both the Government Economist and C&SD have been consulted. The advice given is that the movements of rent paid by CSSA households are also reflected by the CPI(A) rent index and there seems to be no apparent need for a separate rent index for CSSA households.

· Unlike the standard rate payments where inflationary adjustment is projected to cover the remainder of the financial year, the rate of MRA adjustment is projected only up the price level at the beginning of the financial year. To improve the mechanism, proposal is to be made to increase the MRA levels for 1998/99 to bring them up to the projected average price level of the whole financial year.

· Proposal will also be made to seek FC’s approval to delegate the authority to S for Tsy for future revisions of MRA to be made in accordance with the movement of CPI(A) rent index for private housing.” (emphasis added)

76.  Further, in another memo dated 24 February 1998 to the Director, the ST again reiterated that the FB would not support any adjustment mechanism different from an inflation-linked mechanism to support the validity of a 90^th percentile policy.  She said as follows:

“ I refer to your memo to Government Economist yesterday.

2. The annual inflation adjustment mechanism for MRA is different from that for standard CSSA payment. As you described in your paragraph 2, the MRA, adjusted for CPI(A) rent index, does not provide for forecast increase in inflation. I therefore would like to state from the start that I cannot accept changing the inflation adjustment mechanism which will have an impact of providing for inflation adjustment for more than 12 months; there is no allowance for such in our funding provision. In the overall context of the CSSA review, I also do not see any justification to raise the MRA further, albeit by changing the inflation adjustment mechanism. It will be a non‑starter if your review, commenced to study the continued validity of the 90^th percentile policy, ends up recommending changing the inflation adjustment mechanism with effect from 98/99.” (emphasis added)

77.  Finally, by a memo dated 27 February 1998, the Director realised that without the support of the FB, it would be unrealistic to seek to adopt the 90 percentile policy to adjust the MRA annually.  She therefore agreed that the HWB and the SWD would not pursue a 90^th percentile policy and would accept an inflation-linked annual adjustment for the MRA.  The relevant parts of the memo stated as follows:

“ Let us come clean about the real nature of the problem before us.

2. Paragraphs 9.9 and 9.10 of the March 1996 Report on Review of Comprehensive Social Security Assistance Scheme (attached at Annex A for easy reference) would suggest to any normal reader that the 90^th percentile policy including annual review and, if necessary, revision by household size is indeed established Government policy. The unfortunate reality, however, is that there never was before the March 1996 publication of that Report or, subsequently, an appropriate and adequate level of agreement within the Administration and specifically between, on the one side, FB and, on the other side, HWB and SWD to justify that assumption. Your proposal has always been that if we do wish to cover the 90^th percentile as a matter of agreed policy then we should make an RAE bid. The public crunch has never yet come because one way or another we have always been able to say, or suggest, that the proposed MRA increases are covering the 90^th percentile anyway. Nonetheless, this is a deeply unsatisfactory situation and one that we should resolve as soon as possible.

3. Other salient facts are these –

(a) we must take the next round of inflation adjustment proposals, including MRA, to 20 March Finance Committee;

(b) because of (a), an RAE bid in time for the current inflation adjustment round is not of course feasible;

(c) it is, anyway, quite difficult to see what form an RAE bid would take. It would clearly be very odd if, in any individual year in which the 90^th percentile policy was not achieved we proposed an increase in MRA for the household size(s) affected subject to availability of funds and put in a subsequent RAE bid. It would be almost as odd if we were to try to put a price tag on an indefinite commitment to stand by the 90^th percentile policy. Anyway, we accept that even if were to put forward such an MRA bid it would not have a high chance of success in the near future given other competing priorities;

(d) the CSSA caseload is a mixed one but MRA is a policy that applies to all. Within the caseload are certain groups, particularly the able-bodied unemployed, whom we are coming to regard with some disfavour, while there are others like the elderly, and disabled whom we see as justifiable beneficiaries of society’s generosity. We must therefore beware of making policy ‘on the wing’ if the policy change is one that will have broadly the same effect on these very different groups;

(e) the CPI(A) rent index has, in fact, proved quite a reasonable indicator of movements in rental paid by our CSSA recipients.

4. Faced with this really rather difficult situation, a meeting chaired by the undersigned and held on 19 February (notes attached at Annex B) decided that the best way forward was, in effect, to turn our backs on the 90^th percentile policy in favour of adjustment by reference to the CPI(A) rent index as requested in your memo in FIN CR 2/4821/80(96) III of 19 February (My memo in this series of 19 February also refers).

…

7. Having got this far in our thought processes we finally felt that it would be altogether logical that if possible when we adopted this new basis then we should also make the inflation adjustment for rent allowance thoroughly consistent with our other inflation adjustments and therefore make the reference periods the same. From this came the impetus for my (28) in this series of 23 February to Government Economist.” (emphasis added)

78.  At the end, and based on this consensus, the HWB prepared the FC paper (FCR(98-99)10) submitted to the FC for discussion on 3 April 1998 seeking the funding for the CSSA Scheme, including the MRA for the 1998/99 financial year.

79.  In this paper, the HWB adopted the relevant movement of the CPI(A) rent index (8.8%) as the reference for the increase in MRA for that year.  It further asked the FC to delegate to the ST the authority to revise annually thereafter the MRA in accordance with the movement of the CPI(A) rent index.

80.  It is common ground that the FC resolved and approved what were proposed in this paper, and since 1998, the adjustment of the MRA has been made in accordance with the CPI(A) rent index by the ST under the authority delegated to her by the FC.

81.  In my view, these internal memos and FC papers when read together show objectively that from March 1996 to February 1998:

(1) Although the SHW and the Director had wanted and proposed to adopt a 90^th percentile policy for an annual adjustment of the MRA, this was consistently opposed by the ST for the FB.

(2) The ST had repeatedly made it abundantly clear that the FB would only support an inflation-linked mechanism to adjust the MRA annually based on the CPI(A) rent index.  If the HWB wanted to pursue a 90^th percentile policy for adjustment, they needed to seek to obtain any additional funding to achieve that through the yearly RAE.  However, the ST had made it clear that she would not support the HWB’s bid in the RAE.

(3) In between March 1996 and February 1998, when discussions were ongoing between the ST and the SHW and the Director, there was the holding operation in each of those financial years whereby the HWB applied to the FC for funding for adjusting the MRA based on the CPI(A) rent index, which happened to also cover the 90^th percentile criteria.

(4) At the end, by February 1998, the Director accepted that it would be unrealistic to pursue a 90^th percentile policy if it could not have the support of the FB.  She therefore agreed that the mechanism for adjusting the MRA should be one based on CPI(A) rent index as suggested by the ST instead of a 90^th percentile policy.

(5) Based on this agreement and consensus, the HWB submitted the 98/99 paper to the FC seeking (a) the funding for the increase of the MRA for 98/99 financial year based on the CPI(A) rent index, and (b) the FC’s approval to delegate in the future the authority to the ST to annually adjust the MRA by reference to the CPI(A) rent index.  There were approved by the FC.

(6) Since March 1998, the MRA has been adjusted annually by reference to the CPI(A) rent index by the ST under the said delegated authority from the FC.

82.  In light of the above, I conclude that the Government had never adopted the 1996 Policy based on the 90^th percentile objective.  It was also only by February 1998 that the relevant policy bureaux (being the FB and the HWB) had reached an agreement and consensus to use, and thus the Government had adopted, an inflation‑linked mechanism (by reference to the CPI(A) rent index) as the basis to adjust the MRA annually[8].

83.  In relation to the applicant’s alternative contention that the 1996 Policy is the qualitative one, in light of the evidence I have discussed above, I also do not think it can be shown that the Government had adopted such a policy in 1996.  All the above evidence shows that the relevant bureaux were engaged in only discussing whether to adopt a policy based on the 90^th percentile.  Moreover, as Mr Mok rightly pointed out, the fact that there was not the qualitative policy is clearly recognised by the applicant himself: as mentioned above, even in the relief sought, the applicant is only asking the court to declare the policy to be one with the 90^th percentile objective and to mandate to Director to follow that.  The applicant has not asked the court to declare alternatively that the policy is the qualitative one.

84.  In the premises, the applicant has failed to satisfy me that there was the 1996 Policy which had been adopted by the Government.  All the grounds in support of this judicial review must therefore fail.

D. CONCLUSION

85.  For the above reasons, I would dismiss the judicial review.

86.  There are no reasons why costs should not follow the event.  I therefore further order (on a nisi basis) that costs of this application, including any reserved costs, be to the Director, to be taxed if not agreed.  The applicant’s own costs should also be taxed in accordance with legal aid regulations.  Unless any of the parties applies to vary it by summons, this costs order shall become absolute 14 days from today.

87.  Lastly, I thank counsel for their assistance.