Scottish Court of Session Decisions

You are here: BAILII >> Databases >> Scottish Court of Session Decisions >> SCHOOL AND NURSERY MILK ALLIANCE LTD FOR JUDICIAL REVIEW [2022] ScotCS CSOH_11 (28 January 2022)
URL: http://www.bailii.org/scot/cases/ScotCS/2022/2022_CSOH_11.html
Cite as: [2022] CSOH 11, 2022 SLT 262, 2022 GWD 4-52, [2022] ScotCS CSOH_11

[New search] [Printable PDF version] [Help]

OUTER HOUSE, COURT OF SESSION

[2022] CSOH 11

P583/21

OPINION OF LORD BRAID

In the Petition

SCHOOL AND NURSERY MILK ALLIANCE LIMITED

Petitioner

for

Judicial review of the funding rates set in the Milk and Healthy Snack Scheme (Scotland)

Regulations 2021 (as amended) and the guidance issued by the Scottish Ministers in relation

to the Scottish Milk and Healthy Snack Scheme

Petitioner: MacGregor QC; Balfour + Manson LLP

Respondent: Crawford QC, McKinlay; Scottish Government

28 January 2022

Introduction

[1]

For many years, pre-school children in Scotland have received a publicly funded

189 millilitre (

of a pint) serving of milk on each day they attend nursery school or other

childcare setting (settings). Before 1 August 2021, funding was provided under the Nursery

Milk Scheme, a UK scheme which reimbursed settings the actual costs of providing milk.

On that date, the respondent, Scottish Ministers, replaced the old scheme with a new

scheme the Milk and Healthy Snack Scheme by virtue of The Milk and Healthy Snack

Scheme (Scotland) Regulations 2021 (as amended, in particular, by The Milk and Healthy

Snack Scheme (Scotland) Amendment (No 2) Regulations 2021 (the Amendment

Regulations)). The new scheme is wider than the old, in that it allows for the provision of

(i) a non-dairy alternative to children who cannot take milk for health, religious or ethical

reasons; and (ii) a healthy snack. However, settings' costs are no longer reimbursed.

Instead, funding is provided by periodical payments made in advance by local authorities,

based upon a local serving rate (LSR) set by the respondent for each local authority area. It

is the LSR which has proved to be controversial and which is the subject of this petition for

judicial review.

[2]

The petitioner is a membership organisation representing the dairy, health and

education sectors in relation to the provision of milk and non-dairy alternatives. Its

members specialise in providing such products to young children. It objects to the funding

aspect of the new scheme. It asserts that the scheme has resulted in funding cuts for many

settings, which has in turn had a detrimental impact on the petitioner's members, by

reducing their ability to compete in the market, so reducing competition. The petitioner

challenges the rates at which the LSRs have been fixed, and the guidance about them. It

seeks various orders, including reduction of the decision to fix the LSRs, and of the relevant

parts of the Regulations. It has four grounds of challenge: (1) failures in relation to the

consultation exercise which the respondent undertook; (2) failure to make proper inquiry

and/or to take relevant considerations into account; (3) failure to consult and/or to take

relevant considerations into account before amending the Regulations; and (4) irrationality

in the methodology and information used by the respondent in calculating the LSRs.

The hearing

General

[3]

The substantive hearing called before me. In addition to detailed submissions by

senior counsel for both parties, I was referred to various items of correspondence and other

documents lodged in process, including affidavits: two sworn by the petitioner's chairman

Jonathan Thornes,; and one each by Christopher Hogg, managing director of Cool Milk, a

milk supplier or agent; Calum McQueen of McQueen's Dairies; and Donald Laird, of

Fairfield Dairies. The respondent did not dispute the factual accuracy of the affidavits,

which largely spoke to the effect of the new scheme and the impact it has had on milk

suppliers and some settings. The productions also included a number of emails written on

behalf of settings. The respondent, although not formally admitting that the contents were

accurate, did not challenge the provenance of these.

Mr Adlard's report

[4]

The petitioner also relied upon expert evidence, in the form of a report by

Jon Adlard, an economist, commenting on the methodology used in calculating the LSRs

and the effect on competition. Mr Adlard's expertise was not in dispute, although the

respondent did take issue with the use which could be made of his report.

[5]

The extent to which expert evidence might be used in a judicial review in English

proceedings was considered in R (Law Society) v Lord Chancellor [2019] 1 WLR 1649, from

paragraph 36. Senior counsel for the respondent submitted that Mr Adlard's report did not

satisfy "the test" in that case. I do not agree that the case laid down a test of universal

application to Scottish judicial review proceedings. There are essentially two questions: is

evidence relevant at all; and, if so, may the court have regard to expert evidence? While the

nature of judicial review is such that evidence directed to showing whether the decision

under challenge is right or wrong is not relevant, evidence about the factual background

may often be considered by the court, for example, to inform the context in which a decision

was taken or to explain its consequences. In the present case, it is permissible to look at the

affidavits for those purposes, and it was not suggested otherwise. That being so, I see no

reason why the court should not also have regard to expert evidence if it is admissible under

the test set out in Kennedy v Cordia (Services) LLP 2016 (UKSC) 59. In his report, Mr Adlard

gives helpful evidence, founded upon his expertise as an economist, by explaining the

process by which the LSRs were calculated. He also gives evidence of the impact a change

such as that introduced by the respondent is likely to have on competition. I find all of that

evidence helpful to my understanding of the issues. I do not place any more reliance on

Mr Adlard's report than that. Insofar as he expresses disagreement with the respondent's

approach, I have not had regard to that (although the mere fact that he does put forward a

contrary view is relevant when considering the fairness of the consultation, as we will see).

Background

The previous scheme

[6]

As already noted, the old scheme was reimbursement-based. Settings could source

and pay for milk themselves, and claim back the costs. In practice, many settings in the

private sector chose to buy milk from an "agent", that is a company specialising in

supplying milk to nurseries and schools. One such agent was Cool Milk, whose managing

director, Mr Hogg, described in his affidavit how the system operated. Cool Milk, and other

agents, would collate orders from settings, arrange delivery from a wide range of

subcontracted dairies and other suppliers of milk, and submit claims to the Nursery Milk

Reimbursement Unit (NMRU) on behalf of the settings. Not only did settings receive milk

which met their (and the children's) needs when and where they needed it, they were

relieved of the administrative burden of collating receipts and submitting claims. Agents

could also, and did, supply settings with facilities for storing and serving milk (such as free

fridges on loan, beakers and cups). Settings thus had flexibility to source milk in a way that

best met their needs and maximised the benefits to the children. The reimbursement model

meant that relatively small settings located in rural areas, which tend to face higher costs for

delivered milk, were not left out of pocket.

[7]

The old scheme incidentally benefitted milk agents and dairies who supplied milk to

settings. Cool Milk, for example, supplied milk to over 700 settings, including nurseries and

childminders. Because orders were collated by the milk agents, suppliers would receive a

consolidated order from each agent. Their administrative costs were reduced meaning that

they could supply milk to locations which might otherwise not be viable.

The New Scheme

[8]

The policy behind the new scheme is to promote healthy eating habits in children

from an early age, by increasing the consumption of milk and a healthy snack (fruit and

vegetables) by pre-school children. One of the respondent's aims in promoting the new

scheme was to increase access to the scheme by reaching more children than the previous

scheme. An incidental policy aim was to reduce the administrative burden on settings.

From the outset, the respondent made clear that under the new scheme, funding would be

provided by local authorities. However, it was not an aim of the scheme to reduce the cost

of providing milk to children. Fairly read as a whole, the emphasis in the consultation

document was that the provision of benefits under the scheme would be wholly funded by

the Scottish government; that the change was about how payment was to be made, rather

than how much was to be paid. Crucially, as developed below, the consultation document

was silent as to how payments to settings would be calculated.

The LSRs

[9]

Since an understanding of the concept of the LSR and its importance to the scheme is

crucial to an understanding of the issues raised by the petition, I will give a brief overview

of what it is; how the rates were calculated; and the use made of them. A more detailed

description is at paragraphs [56] to [58] below. Having considered and rejected other

options, the respondent used Scotland Excel data on milk supply contracts to derive an

average Scottish milk serving rate per child. Scotland Excel is Scotland's procurement body

for local authorities. The resultant average figure was then weighted for each local authority

area, having regard to the number of eligible children in that area, and to the degree of

rurality, to arrive at a local serving rate for each area. (Mr Adlard, in his report, gave a

detailed explanation as to how this exercise was carried out, but it is sufficient for present

purposes simply to note that a rurality adjustment was made.) Local authorities must use

the LSR for its area as the basis for calculating the payments due to each setting in its area,

regardless of the location within the area of that setting, and regardless of the type of setting

and whether that setting is, as a matter of fact, able to procure milk at the LSR rate. The

amount each setting receives in respect of each payment period is the LSR multiplied by the

anticipated number of servings of milk to eligible children in that period. (Each LSR is the

sum of two amounts: (i) a milk LSR, intended to represent the cost to a setting in that local

authority area of providing a child with a serving of milk or a non-dairy alternative; and

(ii) a healthy snack LSR. No issue arises in this petition regarding the calculation of the

healthy snack LSR. It is the milk LSR, and the calculation of the rates, to which the

petitioner takes exception, and in the remainder of this opinion, references to the LSR are to

the milk LSR.)

The impact of the new scheme

[10]

The petitioner asserts that the new scheme, in particular the use of LSRs, and the

rates at which they have been set, has had a detrimental impact on the market for the supply

of milk, as spoken to in emails from settings lodged in process, and in the affidavits. Having

regard to the manner in which the LSRs have been calculated namely, an average figure

across each local authority area, applied to all settings within the area it was perhaps

inevitable that within each area there would be some settings whose costs would not be

covered in full by the LSR; that is in the nature of an average. Indeed, it was perhaps

inevitable that some local authorities would fare better than others. So it has proved. A

number of settings have complained to the petitioner that the cost of acquiring milk from

agents is no longer covered, and that they are faced with a choice between covering the

shortfall themselves, or purchasing milk from other suppliers (such as supermarkets),

diverting staff resources to carrying out that task. There has been a corresponding sharp

decline in the number of orders placed with milk agents. Mr Hogg, for example, states that

the number of settings continuing to buy from Cool Milk is now just 30. Mr McQueen states

that McQueens Dairies has lost several hundred settings. The petitioner asserts that the

scheme has had the opposite effect from that intended, in that it reduces the incentive to buy

from local suppliers. Mr McQueen believes that some settings, in particular child minders,

have simply chosen to cease supplying milk altogether, such is the burden of the new

scheme; again, the opposite of what was intended.

[11]

Although the emails appear to have been orchestrated by Cool Milk for example,

notwithstanding some variation in wording, each of the emails comprising 6/69 to 6/75 of

process begins "Under the UK Nursery Milk Scheme ..."; ends "Are we misunderstanding

the new scheme? Can you help us?"; and all refer to Cool Milk there was no suggestion

that the emails did not honestly reflect the position faced by the settings in question, namely,

that the amount of funding under the new scheme does not cover Cool Milk's prices.

[12]

Mr Adlard explains in his report why this might be. The Scotland Excel data which

was used is not representative of the costs incurred by all settings since only a subset of

child care providers local authorities are able to access the Scotland Excel framework for

the procurement of milk. Further, the data reflects only the cost of milk, not of non-dairy

alternatives (although that does not seem to have been a concern of the nurseries whose

emails have been lodged in process). Mr Adlard also concluded that the LSRs, being based

on a statistical analysis, are a relatively poor predictor of the actual costs incurred by

settings, even those which can procure milk through the Scotland Excel framework. This is

illustrated by taking two examples. Dundee City's LSR is £0.182 whereas data from

Scotland Excel suggests that the price of milk under the framework is £0.28. Argyll and

Bute's LSR is £0.227, but the price of milk there under the framework is £0.36.

[13]

As regards impact on competition, Mr Adlard's evidence, which I accept, is that

where the LSR is below the cost that a setting actually incurs when providing milk, the

setting may respond in a number of different ways. One of those is, in theory, to pass on the

costs to parents, but as senior counsel for the respondent pointed out, the terms of the

scheme prohibit that. That leaves settings with essentially two options (as the emails from

nurseries suggest): either absorbing the additional costs themselves, or changing supplier,

so as to acquire milk more cheaply. The latter solution, where adopted, will have a

corresponding effect on the pattern of demand for milk by settings, by decreasing the

demand for milk from certain suppliers and increasing demand for milk from other

suppliers. This may affect the ability of some suppliers to compete in the relevant market,

and could potentially prompt suppliers to exit the market. Mr Adlard's opinion evidence is

consistent with the factual evidence in the affidavits. It is not hard to accept that if settings

are unable to recover the cost of supplying milk to children, that is likely to affect the way in

which they buy milk, which in turn is likely to impact upon the market.

The Regulations

[14]

The primary legislation under which the 2021 Regulations were made is the Social

Security Act 1988, section 13 (1)(c) of which provides for the establishment of a scheme or

schemes to provide benefits for prescribed descriptions of children. S ection 13(4)(a)(ii)

provides that a scheme may include provision for a benefit to consist of food of a prescribed

description being provided by a person providing a service such as day care.

Subsection (4)(d) provides for the payment by the Secretary of State of sums to registered

day care providers, in respect of things provided or done by them in accordance with the

scheme.

[15]

There being no dispute that the power to do so is a devolved power, the Scottish

Ministers made the 2021 Regulations, which came into force on 21 March 2021. Regulation 2

defines "benefit" as the food benefit prescribed in regulation 5, which provides for the

entitlement of an eligible child to milk or a non-dairy alternative and a healthy snack on

each day the child receives child care for two hours or more. An eligible child is defined by

regulation 3 as one who is pre-school, and who receives a service from a childcare provider.

By regulation 4, a childcare provider is the provider of day care or child minding.

Regulation 6 is the counterpart to regulation 5, and imposes an obligation on childcare

providers to provide the benefit to which the child is entitled. Regulation 7 provides for the

payment to childcare providers by Scottish Ministers, under the scheme. That power,

among others, is delegated to local authorities by virtue of regulation 15.

[16]

The critical provisions for present purposes are regulations 9, 10 and 11. Since the

petitioner's challenge is to the 2021 Regulations as originally enacted, and as they were

amended by the Amendment Regulations, I will set out regulations 9 and 10 showing the

original deleted text in strike through format, with the new text introduced by the

Amendment (No 2) Regulations in bold, as follows:

"9.--Determining the amount payable

(1)

The payment made by the Scottish Ministers to a childcare provider in

accordance with regulation 7, in respect of a payment period, is to be determined by

reference to--

(a)

the number of eligible children enrolled with that childcare provider;

(b)

the benefits provided (or reasonably expected to be provided) by that

childcare provider under regulation 6,

(c)

the market price of the benefit provided (or reasonably expected to be

provided) by that childcare provider under regulation 6, the local serving

rate specified in the row of column 2 of schedule 3 which corresponds to

the local authority area in which the registered childcare provider provides

the benefit, and

(d)

any adjustment made under regulation 10.

(2) In determining the payment under paragraph (1), the Scottish Ministers may

also take into account any other relevant matter."

As indicated in that regulation, a table of LSRs appears in schedule 3. Several comments

may be made at this stage. First, the function of making payment is one which is delegated

to local authorities under regulation 15. Second, local authorities have no discretion as to

the amount to be paid: it "is" to be determined by reference to the number of eligible

children, the benefits provided and to the LSR. Third, the regulation as originally enacted

provided for the amount paid to be determined by reference to the "market price" of the

benefit provided, although no definition was given of that term. The amendment to LSR,

rather than market price, was made shortly before the scheme was due to come into effect,

apparently to bring the Regulations in line with the guidance referred to below. Although

the petitioner complains that no inquiry was made into the market price of milk, it would

have no quibble with a funding regime based on market price.

[17]

Regulation 10 then provides:

"10.--Adjustment to the amount payable

(1)

The Scottish Ministers may make an adjustment to the amount payable under

regulation 9 in respect of a payment period as a result of a significant change, in

that period, to--

(a)

the number of eligible children enrolled with that childcare provider,

the benefits provided by that childcare provider under regulation 6,

and

(c)

the market price cost of the benefit provided by that childcare

provider under regulation 6.

(2)

The Scottish Ministers must make any adjustment to the amount payable in

respect of a payment period by the end of the immediately following payment

period."

Again, several comments may be made. It is the local authority, to whom the power has

been delegated, which may make an adjustment. Before it may do so, there must have been

a significant change in, among other things, the cost of the benefit. The change must have

occurred in the period in respect of which the adjustment is made, but it is unclear against

what the change is to be measured. It is also unclear how any change in the cost of the

benefit is to be measured, when the benefit is not itself quantified by reference to cost. It is

also unclear whether "and" is to be read conjunctively or disjunctively, although since a

change in the number of children will also result in a change to the cost of the benefit, it may

not matter which. Finally, although the guidance has passing references in paragraph 10.5

and 10.8 to reconciliation, it is not obvious from the Regulations themselves that there will

be a reconciliation, at least of the amounts paid to settings, in every case. Regulation 10 does

not provide for that, partly because of the requirement for there to have been a "significant"

change, and partly because the power to adjust the amount paid is discretionary.

[18]

I need not set out regulation 11 in full, but in brief, it provides that a childcare

provider must inform the local authority of any "significant change" to the number of

eligible children enrolled with the provider and to the milk and snacks provided by that

provider. The term "significant change" ties in to the same wording in regulation 10,

although it may be noted that settings have no obligation to inform the Scottish Ministers of

any change to the cost of providing the benefit (even if the cost is significantly less).

The statutory guidance

[19]

Regulation 17 provides that a local authority must have regard to any guidance

issued by the respondent when exercising any function under the scheme. Statutory

guidance was issued on 4 June 2021. It confirms the policy behind the scheme, stating that it

will support the respondent's intention to improve children's health and wellbeing by

establishing healthy eating habits, and that its aim is to ensure that as many pre-school

children in day care as possible will receive a daily helping of milk and a healthy snack to

support the development of healthy eating habits for later life. The policy overview, at

annexe A, states that settings:

"will use the funding provided by the local authority to purchase the necessary

volume of milk and healthy snacks ... Settings are encouraged to source milk ...

locally as part of our ambition to improve access to, and understanding of, the

benefits of healthy local foods."

Paragraph 2.5 prohibits settings from charging parents for the provision funded under the

scheme. Paragraph 4.1 provides that the respondent will provide funding for the scheme in

line with the agreement "to be reached" with COSLA. 4.2 requires local authorities to make

upfront payments from 1 August 2021 to settings that have registered for the scheme, those

payments to be determined by the local authority.

[20]

The key provisions regarding LSRs appear from paragraph 10.1. Insofar as

material, they are set out as follows (retaining the guidance's uncertainty as to whether to

use title case or not when referring to a local authority):

"10.1 Local authorities are responsible for determining and making payments to

settings, having regard to guidance provided by Scottish Ministers. These payments

will be based upon the anticipated benefit (the number of servings of milk and

healthy snack) that settings expect to provide ...

...

10.3

Local authorities will receive their allocations from Scottish Ministers

calculated in line with the agreement with COSLA, and will determine payments to

settings in line with the provisions of the [Regulations].

10.4

It is anticipated that Local Authorities will set a rate of payment per serving

of milk and healthy snack (one serving = one 189 or 200ml portion of cow's milk or

specified alternative or 189ml first infant formula for children under 12 months and

one healthy snack where children are age 6 months or over).

10.5

Scottish Ministers recommend that the rate of payment per serving should be

set at the Local Serving Rate confirmed to the Local Authority by Scottish Ministers.

This is the rate at which Scottish Ministers will calculate end-year reconciliation

payments.

10.6

Where a local authority wishes to set a rate in excess of the Local Serving

Rate, they may do so in line with the provisions of the regulations, but the excess cost

will only be met by Scottish Ministers where this has been agreed by exception on

the basis of evidence and in advance of introduction of this rate ...

10.7

Local authorities have agreed to use the Local Serving Rate for Scheme year 1,

without prejudice to future decisions.

10.8

The determination of ... payments to registered day care settings should be

calculated by multiplying Provision (the anticipated (and at reconciliation, actual)

number of servings) for the payment period x by (sic) Rate of payment per serving

(Local Serving Rate except where local authorities determine otherwise)."

It will be seen that there is a clear tension between paragraphs 10.5, 10.6 and 10.7 as to the

extent to which local authorities have a discretion not to use the LSRs, but 10.7 makes clear

that in year 1, local authorities have no autonomy to set a different rate. Paragraphs 10.5

and 10.8 imply that a reconciliation based upon actual servings will occur, and those

paragraphs also give the impression that local authorities do have discretion to set a rate

other than the LSR. Even if the reconciliation referred to is as between the Scottish

government and local authorities, rather than a reconciliation with settings, it is unclear how

that part of the guidance is consistent with the regulations themselves, which do not impose

on settings any obligation to inform the respondent of the actual number of settings served

merely to inform the respondent of any significant change. Finally, the meaning and effect

of clause 10.6 is opaque to say the least, insofar as it refers to agreement by exception on the

basis of evidence "and in advance of introduction of this rate". The words "by exception"

are otiose and add nothing; and the requirement to have reached agreement in advance of

introduction of the LSR would appear to deprive the provision of any practical content

which it might otherwise have had in future years.

The COSLA Agreement

[21]

The statutory guidance refers to an agreement between the respondent and COSLA.

On 19 May 2021, the Settlement Distribution Group (SDG), a joint group attended by the

respondent, COSLA and local authority officials approved recommendations in relation to

the funding of the new scheme. On 28 May 2021 COSLA leaders approved the

recommendations. That appears to be the agreement in question.

The petitioner's grounds of challenge to the scheme

[22]

I have summarised above the petitioner's complaints, and the four grounds of

challenge. In more detail, the petitioner seeks reduction of: (i) the statutory guidance and/or

the rates table; (ii) the respondent's decision to determine that the funding formula of the

new scheme should be represented by the LSRs; (iii) regulation 2 of the Amendment

Regulations (which amended regulations 9 and 10 in the manner shown above); and (iv) the

agreement between the respondent and COSLA.

[23]

Although there are four grounds of challenge, in reality both parties accepted that

the challenge to the Amendment Regulations will stand or fall with the challenge to the

Regulations and the guidance. Further, for the reasons set out below, the challenges

founded upon, respectively, a lack of proper inquiry by the respondent, and irrationality

overlap to such a substantial degree that I propose to deal with them together. Accordingly,

in the following sections I will consider, first, the petitioner's case that the decisions,

guidance and regulations are all flawed because of failures in the consultation exercise; and,

second, the irrationality ground of challenge. I then comment briefly on the third ground.

The failure to consult case

The consultation exercise

[24]

The consultation document was headed "Welfare Foods a consultation on meeting

the needs of children and families in Scotland". After asking consultees in question 8:

"What do you think about the proposal to offer milk as part of the free meal offer for all

children in ELC ...?", the document stated:

"To deliver this we propose removing the administrative burden of the UK NMS by

providing milk to children as part of their funded ELC offer, removing the need for a

reimbursement scheme ... The provision of milk would be available to children

accessing funded ELC provision, whether this is offered in a local authority, private,

third sector or childminder setting. We would discuss and agree funding with local

authorities."

[25]

The document then referred to the possibility of offering a healthy snack (question 9)

and of delivering milk and a snack to children outwith early funded learning. Questions 10

and 11 sought views about how the government could best support children outwith ELC

entitlement. Question 11 sought views on the provision of a healthy snack. The thrust of the

document was about removing administrative barriers and widening access so that children

who needed milk and healthy snacks could access them.

[26]

The petitioner lodged a response to the consultation. Among other comments, it

expressed a concern about the proposed change to the funding structure, "as it may present

a further regulatory burden upon the agricultural sector."

[27]

On 28 February 2020, Philip Canavan, who introduced himself as having recently

been appointed to lead on the delivery of the Scottish Nursery Milk and Healthy Snack

Scheme, sent an email to Jordan Newfield acting for the petitioner, in which he stated:

"I want to reassure you that I intend to engage with all interested parties and

provide an opportunity for you to share your views on the delivery process."

[28]

The Scottish Government hosted a meeting on 5 March 2020, attended by a

representative of Cool Milk, a member of the petitioner, who expressed concerns over the

lack of information and consultation. The note of the meeting records that:

"The new Scottish scheme will remove any need for claiming as each individual local

authority will be responsible for allocating a lump sum of funds to each Early Years

setting.

Each local authority will have autonomy over how they will implement distribution

of the funds".

(As already seen, that last statement turned out to be incorrect, at least for year 1).

Paragraph 4 of the minute records that at that time there was to be no funding for non-dairy

alternatives to milk, but as the minute also goes on to state: "Nothing is as yet set in stone."

[29]

On 3 April 2020 Joe Fitzpatrick, Minister for Public Health, Sport and Wellbeing and

Maree Todd, Minister for Children and Young People, sent a letter to the chairman of the

petitioner, stating in paragraph 2:

"We appreciate you taking the time to raise your concerns with us about the

introduction of the new scheme. However, hopefully we can provide some

reassurance that the scheme should in no way result in a disruption to the milk

supply to childcare settings ... Our proposal, as you know, is to move away from a

reimbursement scheme to one where funding is provided in advance to local

authorities, for onward distribution to childcare settings ... Scottish Government

have ... been engaged in discussions with CoSLA and local authority representatives

to ensure we take into account all the practical issues for implementation ... To be

clear, the new scheme is not proposing to disrupt the current relationship between

milk suppliers, agents, and care settings. Contractual relationship for the delivery

and provision of milk can still stand. The key difference is that settings will pay suppliers

directly for this service, rather than being reimbursed for it via the NMRU" [emphasis

added].

[30]

On 26 May 2020, Marie Todd wrote again to the chairman of the petitioner, making

clear that discussion would take place:

"Please let me make clear that continued discussion with the sector as the design of

the scheme develops has always our intention, (sic) and that no final decisions on the

scheme have yet been taken. This will only be done when we have taken into

consideration all the feedback we receive from all stakeholders as part of our

ongoing engagement plans...

... To reiterate, however, no final design has been agreed with CoSLA at this time.

Our intention remains that once we have finalised a proposal, we will engage with

stakeholders, including the SNMA, as previously envisaged. This engagement will

also include consideration of a number of impact assessments, including a Business

and Regulatory Impact Assessment ... However let me be clear that we do not anticipate

that this will prove to be a cheaper alternative for Government to fund, [emphasis added]

and cost is not a driver behind the proposed changes ... Moreover, engagement with

the sector to date has suggested that changing the way that funding is provided

could drive an increase in the consumption of milk, especially in smaller settings

where people may not currently submit claims.

... it is our expectation that the new scheme should not disrupt the current

relationship between milk suppliers, agents, and care settings. The key difference is

that settings will pay suppliers directly for this service, rather than being reimbursed

for it via the Nursery Milk Reimbursement Unit. I note your concerns that settings

will switch from contracts with dairies to supermarket purchasing, and my officials

will be happy to discuss these further with you ... we will ensure that the SNMA has

the opportunity to participate in the next stages of engagement and to raise any

further concerns."

[31]

On 19 November 2020, Marie Todd again wrote to the petitioner:

"As previously stated, we share your view that consultation on the new scheme is

important and indeed we consulted on the broad principles for a new scheme within

the [2018 consultation]. The proposed scheme ... remains based upon the key

features set out in that consultation. As we have indicated previously, we will

consult further on the outstanding elements of the scheme and notably the impact on

business ...

... Once we have an agreed proposal for discussion, we plan to undergo an intensive

engagement process with all our key stakeholders including the SNMA ... [the]

window for engagement is short, however, I am confident that it will provide you

with the necessary opportunity to engage to help us put in place a robust mechanism

for delivering milk to the early years sector, and to contribute to the development of

the impact assessments ...

The Scottish Government's position remains that it will be for individual settings to

determine their supply arrangements as is the case at the present time. Where the

service provided via existing contracts is good and represents value for money, there

should be no driver for settings to move from these arrangements."

The Business and Regulatory Impact Assessment

[32]

A Business and Regulatory Impact Assessment (BRIA) was issued on 17 February

2021. It stated at page 2 that:

"Since funding will be delivered via local authorities, Scottish Government have

been working collaboratively with COSLA and the relevant governance groups to

develop proposals on the design and delivery of the Scheme".

[33]

A section headed "Contractors and Milk Agents" included the following:

"The key focus of the Scheme is to route the funding for [the purchase of milk and

snack] directly to the childcare setting, as opposed to requiring them to claim back

costs ... indicative costings have used NMRU claims data, which include the

administrative costs charged by milk agents or others contracted by settings to

provide the milk."

[34]

The BRIA then recorded that NMRU data suggested that 1,391 childcare settings

used milk agents, representing 16% of the total of 8,754 settings, or 38% of non-childminder

settings. It went on to state that the new scheme need not preclude a continuing

procurement relationship between local authorities/settings and contractors and milk

agents, nor would it prevent new relationships being established, but it would change the

way payment was made. Contractors and agents would no longer make claims on behalf of

settings and settings would need to make arrangements to pay contractors.

[35]

At page 11, the following statement was made:

"The overall quantum will be based on costing assumptions to be agreed with

COSLA. These are derived from the averages of current claims and characteristics of

Scottish settings to the UK NMS and cost projections based on the Scotland Excel

framework. The cost estimates incorporate adjustments to account for the addition

of a healthy snack".

[36]

At page 12, the BRIA referred to the costs reclaimed from Scottish settings both in

respect of milk supplied by milk agents (where the cost ranged from £0.15 to £0.29

per 189mls) and local authority agent claims, where the average cost ranged from £0.15

to £0.17. It also referred to an additional uplift to account for a range of variables, including

an increase in the number of children attending and a fluctuation in market prices. At

page 13 the BRIA stated that the Scottish Government had estimated the maximum cost of

the new scheme based on (among other things) a milk unit cost per child, per annum

derived from claims data informed by the NMRU and cost projections based on the Scotland

Excel costings, to which an uplift had been applied to take account of uncertainties such as

the impact of the expansion to 1140 hours of funded provision and inclusion of appropriate

alternatives to dairy where required. (As will be seen from paragraph [56] and [57] below,

these statements were not strictly accurate. No uplift had been applied to the Scotland Excel

costings, nor was it obvious that any uplift applied to the NMRU costings was to account for

the inclusion of non-dairy alternatives.) The BRIA then stated that the variables may be

adjusted further following discussion with key stakeholders such as COSLA though

feedback suggest that in-principle assumptions above were reasonable.

[37]

At page 16 the BRIA stated that:

"The Scottish Government will be developing guidance to support delivery of the

Scheme, and will involve stakeholders in this process, and ensure the points raised

through the engagement process to date are considered as this is developed."

[38]

In the immediately following section under the heading Competition Assessment,

the BRIA stated:

"The Scheme will have no competitive impacts; therefore the Scottish Government

does not expect the proposals to have an impact on competition as they will not:

Limit the number or range of suppliers

Limit the ability of supplier (sic) to compete

Limit supplier's (sic) incentives to compete vigorously

Limit the choice and information available to consumers".

Events after the BRIA

[39]

Notwithstanding the assurances previously given, and the content of the BRIA,

Mr Canavan wrote to the petitioner on 25 February 2021, disclosing a change of position:

"With respect to your continued concern about our purported obligation to publicly

consult on the Scheme and regulations establishing it, our position remains

unchanged the regulations establishing the Scheme and the terms of the Scheme

itself are not the subject of a public consultation obligation".

[40]

Draft guidance was sent to the petitioner on 7 May 2021. Paragraph 4 of the draft

stated that the Scottish Government "will provide funding for the Scheme in line with the

agreement to be reached with COSLA". Paragraph 4.2 stated that local authorities would

make upfront payments from 1 August 2021 to settings that had registered for the scheme,

these payments to be determined by the local authority. A so-called placeholder then

appeared, signifying a section yet to be inserted. Paragraph 10.1 stated that payments

would be based on the anticipated number of servings of milk and snack. 10.3 had a further

placeholder to the effect that further guidance would be added following conclusion of the

agreement with COSLA. The draft therefore provided no meaningful information about

how the funding would be calculated.

[41]

The final guidance was issued on 4 June 2021 some 8 weeks before the scheme went

live. The key provisions appear in section 10.3 onwards and are set out above. This was the

first time that any consultee other than COSLA received information about how the scheme

would work and how it would be funded. Insofar as the Regulations as originally draft ed

were inconsistent with the guidance, since they referred to market price, that inconsistency

was later removed by the Amendment (No 2) Regulations.

The duty to consult

[42]

The law regarding the duty to consult is not in dispute. The following cases were

referred to in submissions: Regina v North and East Devon Health Authority, ex parte

Coughlan [2001] QB 213; R (Law Society) v Lord Chancellor, above; R (Eisai) v National Institute

for Health and Clinical Excellence [2008] EWCA Civ 438; R (London Criminal Courts Solicitors'

Association) v Lord Chancellor [2014] EWHC 3020; R (Milton Keynes Council) v Secretary of

State 2011 EWCA Civ 1575; R (Medway Council) v Secretary of State for Transport

[2002] EWHC 2516; R (Help Refugees Ltd) v Secretary of State for the Home Department

[2018] 4 WLR 168.

[43]

From these, I derive the following propositions.

Even if no duty to consult exists, where consultation is carried out, it must be

carried out properly and fairly: Coughlan, paragraph 108 ; Eisai, paragraph 24.

For a consultation to be carried out properly and fairly, the following

principles, all derived from Coughlan, must be observed:

consultation requires to be undertaken at a time when proposals are at

a formative stage;

ii.

sufficient reasons for a proposal must be given, in sufficiently clear

terms to allow consultees to give intelligent consideration to it, and to make

an intelligent response (see also R (Help Refugees Ltd), paragraph 90);

iii.

adequate time must be given for a response;

iv.

the product of the consultation must be conscientiously taken into

account;

These are all aspects of the same fundamental question: has there been procedural

fairness.

For a consultation to be fair, there may be a need to disclose facts or

information: R (Law Society) v Lord Chancellor. In that case, in deciding whether non-

disclosure of information was fair, four factors were said to be relevant

(paragraph 73):

the nature and potential impact of the proposal put out for

consultation;

ii.

the importance of the information to the justification for the proposal;

iii.

whether there was a good reason for not disclosing it;

iv.

whether consultees were prejudiced by non-disclosure.

The decision maker cannot pick and choose in deciding whom to consult. A

fair consultation requires fairness in deciding whom to consult as well as fairness in

deciding the subject matter and timing of the consultation: R (Milton Keynes Council),

paragraph 32.

The impact of a decision is a material factor in deciding what fairness

demands in any particular case: R (London Criminal Courts Solicitors Association)

paragraph 5.

Each case is fact specific. This is stated in many of the cases but see, eg, Eisai,

paragraph 27.

Legitimate expectation

[44]

Again, the law is not in dispute. An authority may not depart from a clear and

unambiguous undertaking previously given, unless it is fair for it to do so, the court being

the arbiter of fairness: Re Finucane's Application for Judicial Review [2019] UKSC 7,

paragraph 62 (that was said in the context of a substantive legitimate expectation but, as

senior counsel for the petitioner submitted without contradiction, is equally applicable to a

procedural one). A legitimate expectation of consultation can also arise "from an interest

which is held to be sufficient to found such an expectation": R (on the application of Moseley)

v Haringey London Borough Council [2014] UKSC 56; 2014 1 WLR 3947, per Lord Reed JSC

at [35].

The petitioner's case

[45]

The essence of the petitioner's case of inadequate consultation is that it was not told

of the proposal to base payments on LSRs, or the rates themselves, until it was too late for it

to make any meaningful representations. The petitioner was therefore deprived of the

opportunity to give intelligent consideration, and to make an intelligent response, to the

proposal. Fairness required it to be consulted when that proposal was at its formative stage.

The four factors set out in R (Law Society) v Lord Chancellor, when applied to the present case,

all pointed to the non-disclosure of the proposed LSRs being unfair: milk agents and

suppliers would require to exit the market; funding was key to the operation of the scheme,

hence the information about LSRs was clearly important; there was no good reason as to

why it could not have been disclosed; and prejudice to suppliers (and to some settings) was

well borne out. The livelihood of suppliers was at stake: they ought to have been consulted

on a matter so fundamentally important, particularly after they had been led to believe by

the respondent that there would be no change in the amounts paid under the scheme. The

petitioner also had a legitimate expectation, based upon the consultation document and the

correspondence referred to above, that it would be consulted on the workings of the scheme.

The respondent's response

[46]

The respondent's response was that there had been a full consultation exercise on the

principles of the scheme, including that funding was to be taken over by local authorities.

Consultees, including the petitioner, had been asked to, and did, comment on that. As an

analysis of the responses showed, some had been taken into account, and others

disregarded. The respondent did not require to consult the petitioner further. None of the

four factors in Law Society v Lord Chancellor favoured the petitioner. There was no evidence

that the use of LSRs would impact upon or cause prejudice to those whom the scheme was

designed to benefit: children. The information about LSRs had little importance to the

proposal to adopt up-front funding information which the petitioner or others might have

been able to provide was not material to the proposal. The respondent had a good reason

for not disclosing the information. The 2018 consultation was not about funding of the

proposal: funding was a matter as between the respondent and local authorities. The latter

had an interest in knowing that there was sufficient funding for children in their areas. That

was why COSLA was consulted and not others, such as the petitioner or settings. Finally,

no prejudice had been caused to the petitioner or others. The LSRs were not the basis for the

scheme, simply the method to arrive at how much was paid. As for legitimate expectation,

the scheme was not intended to benefit suppliers, but children. Suppliers could therefore

have no legitimate expectation that they would be consulted about how funding would be

calculated.

Decision on adequacy of the consultation exercise

[47]

The question as to whether there has been proper and fair consultation is one of

fairness, and is fact-sensitive. While senior counsel for the respondent sought to distinguish

a number of the cases on their facts, the question is simply whether what happened was

procedurally fair in the particular circumstances of this case. In considering that, it is

important to have in mind the assertion by senior counsel for the respondent, which she

emphasised throughout the hearing, that it was never the respondent's intention that the

LSRs should represent the actual cost to settings of providing milk and non-dairy

alternatives. When the fairness of the consultation is considered through that prism, it is

easily seen that the process was unfair, since at no stage prior to finalisation of the proposal

and publication of the LSRs was that intention made clear, or even hinted at, whether in the

consultation document or in subsequent correspondence; with the result that the petitioner

did not have a fair opportunity to comment on the funding of the scheme.

[48]

Dealing with the specific arguments which were advanced, the first point to dispose

of is the respondent's submission that the respondent owed no duty to consult with the

petitioner because it was not a beneficiary under the scheme. That submission is not

supported by the case law, which makes clear that the impact of a decision is a material

factor in deciding what fairness requires see paragraph [43] above. A decision may clearly

and foreseeably impact on a person who is not a direct beneficiary of a scheme. As senior

counsel for the petitioner submitted, Law Society v Lord Chancellor is the paradigm example

of fairness requiring consultation with a body (in that case, the Law Society) which was not

directly benefitted by the scheme in question (in that case, the legal aid scheme, the

beneficiaries of which were accused persons).

[49]

Developing that point, the change from a scheme which reimbursed costs to one

which does not even attempt to pay settings the actual cost of supplying milk, is one which

will inevitably impact upon the business of suppliers and milk agents. That this is so as a

matter of fact is spoken to in the affidavits, and by Mr Adlard. As regards the four factors in

the Law Society case, the respondent's arguments are unconvincing. Those arguments are

underpinned by senior counsel's assertion that the LSRs were "not the basis for the scheme,

simply the method to arrive at how much was paid". Such an encapsulation of the role of

the LSR significantly underplays its importance. The LSR is the very cornerstone of how the

scheme will operate in practice. It is the LSR which determines how much funding a setting

will receive. Since the amount which a setting will receive will in turn impact on its ability

to purchase milk, or indeed its willingness to register in the scheme at all, it is supremely

important to the operation of the scheme as a whole.

[50]

Further, it seems to me that the switch to local authority funding has, whether

consciously or unwittingly, served to conflate three related but distinct issues. The first is

the overall cost of the scheme to the respondent the total cost of providing milk to all pre-

school children in Scotland. The second is how that overall cost is divided up among local

authorities. The third is how each local authority then distributes its own allocation to

settings. While COSLA has a clear and perhaps of all stakeholders, unique interest in the

first and second of these, the same cannot be said of the third, where settings and suppliers

have at least an equal interest as COSLA. It is therefore no answer to the question, was there

a good reason not to disclose information about the proposed LSRs to the petitioner, to say,

as the respondent does, that COSLA had an interest in knowing that there was sufficient

funding for children in their area. That has no bearing on the distinct question of how local

authority allocations should be allocated amongst settings. It was unfair for the respondent

only to consult with COSLA on that crucial question. To do so was to pick and choose in an

unfair way.

[51]

The key issue lies in the revelation that the respondent had never sought to calculate

the actual cost to each setting of providing milk. If that was truly its position from the outset

(senior counsel for the petitioner doubted that, describing the assertion as "rewriting

history"), that was not made clear in the consultation document; nor in the correspondence

from Ministers. The ministers' statement in the letter of 3 April 2020 that the "key difference

is that settings will pay suppliers directly ... rather than be reimbursed for it ..." was either

misleading, or was departed from. That was a difference, but could not fairly be said to be

the key difference, where settings were no longer to receive, either by reimbursement or in

advance, a sum which represented the actual cost of obtaining milk from suppliers.

Whether that was always the respondent's intention, or the scheme evolved as it was being

developed, either way, fairness required that the petitioner be told of that proposal, so that it

had an opportunity to make an informed and intelligent contribution to its development.

From the information which it did receive, the petitioner had no means of gleaning that an

average figure was to be used, far less how the LSRs were to be calculated. The draft

guidance was of no assistance, with its inclusion of placeholders where information about

LSRs would eventually appear in the final version. Nor did the regulations as originally

enacted give any clue, since they referred to market price. If LSRs were to be used as the

method of calculation, fairness also required that the petitioner should have the opportunity

of commenting on the data which lay behind the setting of the rates, as COSLA did. It is

nothing to the point that the petitioner was able to, and did, respond to other aspects of the

consultation, when the information it was provided about funding was to all intents and

purposes, non-existent.

[52]

Senior counsel for the respondent submitted that no regard should be had to

Mr Adlard's report because it simply set out an opposing point of view to that of the

respondent, but, in a way, that is the point here. If nothing else, the report illustrates the

type of points which might have been made by the petitioner had it been given the

opportunity to participate; it might have been able to persuade the respondent of a different

approach. It was deprived of that opportunity. It is instructive to consider the extent to

which COSLA's representations held sway. A Scottish Government paper dated 2 February

2021, shared with COSLA, discloses that following representations made by COSLA in

response to the respondent's original proposal (proposal A) a new proposal, containing

higher figures in relation both to milk and healthy snack, was put forward (proposal B). The

new approach was to apply "local authority costs under existing contracts to all settings

including childminders, implying a per pint cost of around £0.59." That no consideration at

all seems to have been given to the feasibility of non -local authority settings purchasing milk

at that price simply emphasises the unfairness of formulating the proposal without

consulting with a broader range of stakeholders, including the petitioner. COSLA's

proposal made the point that the Scotland Excel data, from which the proposal B figure was

derived, did not take into account providers who did not obtain milk under these contracts,

but that was not followed up by the respondent. This is precisely the sort of area where

consultation with the petitioner might have made a difference to the final shape of the

scheme.

[53]

The second strand to the petitioner's argument is that it had a legitimate expectation

that it would be consulted, derived from clear and unambiguous statements in the

correspondence and in the BRIA, and that it was unfairly deprived of that opportunity. I

accept the petitioner's submission that the statements made were such as to give the

petitioner a legitimate expectation that it would be consulted on the "delivery of the

scheme", as the BRIA had stated. For example, the email of 28 February 2020 gave

reassurance that all interested parties would be involved in the process. At the meeting of

5 March 2020 it was represented that local authorities would have autonomy in distributing

their allocation. The ministers' letter of 3 April 2020 gave reassurance that existing contracts

with milk suppliers would not be disrupted and made reference to the supposed key

difference being the mode of payment, referred to above. The letter of 26 May 2020

provided reassurance to similar effect and stated that the government would "ensure" that

the petitioner had the opportunity to participate in the next stage of engagement. The letter

of 19 November 2020 provided still further reassurance, and a statement that there would be

an "intensive engagement process" with, among others, the petitioner, and that where

existing contracts represented value for money, there should be no driver for change. Not

only were these statements clear and unambiguous the respondent could not have used a

stronger term than "ensure" - no mention was made in any of that correspondence of an

average-based scheme. Given the statements consistently made about the content and likely

impact of the scheme, and the assurances given that the petitioner would be afforded the

opportunity to participate as the scheme was developed, it was not fair for the respondent to

depart from those assurances, and remove from the petitioner the opportunity to make

informed and intelligent comment about the proposal, particularly where what had been

said in the correspondence about how the scheme would operate was either not the full

picture, or was to be departed from.

[54]

For all of these reasons, I have concluded that the consultation process, at least as the

LSRs were developed, was unfair. The petitioner's challenge to the regulations and

guidance, insofar as it is based upon a lack of proper consultation, must therefore succeed.

The irrationality challenges

Introduction

[55]

The second and fourth grounds of challenge are, respectively, a failure to undertake

proper inquiry before setting the levels at which the LSRs were to be fixed, and irrationality

in setting those levels. The essence of the complaint of failure to undertake enquiry is that if

the respondent wished to lay down a formula by which the amount of funding to be

provided to a setting was to be ascertained, it was incumbent on it to make proper inquiry to

ensure that the formula adopted reflected the costs to settings of providing the benefits. The

essence of the irrationality challenge is that having failed to make proper inquiry, the

approach then adopted in laying down a formula was irrational since it left out of account

matters which ought to have been taken into account. Those two grounds are to that extent

a mirror image of each other, as the grounds of each challenge reflect. For that reason, and

since a failure to make proper inquiry has been said to be a particular aspect of irrationality,

I propose to deal with both grounds of challenge together. Before doing that, it is instructive

to consider the respondent's reasoning and approach in fixing the LSRs.

The respondent's reasoning

[56]

The respondent avers that three methods were considered. The first, on which the

original, March 2020, costing was based, drew on data derived from the NMRU, which was

suitably weighted by the numbers in each type of setting, and an "uncertainty" uplift of 20%

was then applied to it to account for uncertainties such as general variations in the milk

price and additional costs incurred due to the inclusion of non-dairy alternatives. The

20% uplift was based on "assumptions of reasonableness" rather than any precise

calculation. Those averments should be contrasted with the "in confidence paper" shared

with COSLA in February 2021, which stated:

"By using Care Inspectorate data on the average number of child registrations in

settings of the type described by the NMRU, we were able to estimate the number of

children registered in settings claiming for milk. The average cost of milk provision

across all provider types ... was £20.55 [per annum].

We expect this cost to increase in the coming years:

(i)

The expansion to 1140 hours may lead to children attending ELC for more

days a year. However, this effect will be limited as it only applies to 3, 4 and

some 2 and 5 year olds. It may also only displace paid-for `wraparound'

hours and thus not increase the total level of ELC attendance for these

children.

(ii)

The delivery mechanism of the new scheme may lead to settings providing

more milk to those children that attend.

As such [emphasis added], we have applied a 20% uplift to this figure, bringing the

per child per annum cost to £24.66."

That paper is therefore very clear that the 20% uplift was to cater for only two possibilities:

children attending on more days a year, than previously; and more milk being provided for

children who did attend. No mention was made of possible variations in the price of milk;

and, specifically, no mention was made of the cost of providing non-dairy alternatives.

[57]

When one then turns to the SDG paper 24 of 2021, issued a short time later, it too

describes how the figure of £24.66 was arrived at, but gives a subtly different explanation:

"The March 2020 costing drew on data from the NMRU to develop the milk costings,

and applied a 20% % (sic) uplift to this figure to take account of potential fluctuations

and the potential inclusion of approved alternatives to milk which had not been

confirmed at March 2020."

If that was intended to be a summary of the reasons previously given for the 20% uplift, it

was incorrect.

[58]

Be all that as it may, the second method, the one adopted, was to use the Scotland

Excel data on milk supply contracts for local authorities purchasing milk through the

Scotland Excel framework, with the NMRU figure to all intents and purposes being

discarded. This gave a weighted average annual figure of £34.27, or 19.6 pence per 189ml.

The respondent adopts this figure, as it was higher than the estimate obtained by the first

method after the 20% uplift. For the purpose of calculating the anticipated maximum

which I take to be the maximum Scotland-wide cost it was assumed that all eligible

children currently registered in childcare settings were claiming milk and an uplift was

applied to account for the potential for uptake to increase due to changes in the operation of

the new scheme but no uplift was applied to the 19.6 pence itself. The respondent has

further averments explaining how the base figure of 19.6 pence was adjusted to take account

of rurality only. As Mr Adlard explains in his report, depending on the degree of rurality, in

some local authorities the LSR is more than 19.6 pence; in others it is less.

[59]

For completeness, the respondent avers that it considered a third method, which

used a milk serving cost, weighted to assume the NMRU estimate for non-local authority

settings and the Scotland Excel average for local authority settings. It was discarded, as

different rates might be paid to local authority and private settings which the respondent

avers would be inconsistent with the intentions of the new scheme.

The petitioner's case

[60]

By the very nature of the grounds of challenge, and as senior counsel for the

petitioner acknowledged, the petitioner's submissions overlapped to a degree. The reasons

advanced by the petitioner as to why the setting of the rates, and producing the guidance, is

irrational can be distilled down to the following.

(i)

The respondent failed to make proper inquiry into, and therefore to take

account of, relevant factors including: the cost of non-dairy alternatives; the market

price of milk for all settings, not simply those which could access the Scotland Excel

framework; the differences between settings of different types; and the indirect costs

of providing milk such as delivery costs, refrigeration and beakers.

(ii)

The respondent acted on the basis of two material errors in fact: that prices

paid by local authorities were also representative of the prices paid by private settings;

and that there would be no impact on competition. The respondent had an inadequate

empirical basis for either conclusion. In particular, when the BRIA was formulated,

the LSRs had yet to be determined, and the agreement that they must be used had yet

to be reached. The respondent did not have factual material entitling it to conclude

that there would be no impact on competition.

(iii)

The respondent's approach to setting the quantum of the LSRs was irrational

insofar as it took no account of the cost of non-dairy alternatives; and it based the LSRs

solely on the Scotland Excel framework. The 20% uplift which had been applied to the

NMRU figures was arbitrary.

(iv)

The rates fixed were irrational, in that private settings could not afford them,

and in some instances, such as Dundee City and Argyll and Bute, they were below

even the rates at which local authority settings could purchase milk through the

Scotland Excel framework. This was compounded by the absence of any mechanism

allowing individual settings to claim to be paid at a rate higher than the LSR.

[61]

The petitioner further argued that now that it had become evident that the rates were

in some instances too low, the respondent's failure to amend the scheme was itself irrational.

[62]

Distilling the foregoing down still further, the common themes of the petitioner's

submission were: irrationality in relation to the 20% uplift; lack of inquiry into, and failure

to take account of, the cost of non-dairy alternatives; a failure to look beyond the Scotland

Excel data; the application of an erroneous proxy for the market price of the benefit

provided by ignoring costs of inputs such as storage costs (refrigeration) and distribution

(cups and beakers); and a lack of any proper evidential basis for the conclusion in the BRIA

about the impact on competition.

The respondent's response

[63]

The respondent's over-arching position is that the decision to pay settings by

reference to LSRs, and the method by which those LSRs was calculated was a political one,

on a matter of socio-economic policy, and that the respondent enjoyed a large margin of

appreciation in which the court should not interfere. The petitioner's complaints simply

arose from the consequences of those political decisions, and not from any irrationality. The

question was not whether a better scheme might have been devised, but whether the scheme

was unlawful. Dealing specifically with each of the petitioner's arguments:

(i)

The market prices paid by private settings were taken into account since the

respondent had regard to NRMU data, giving details of historic claims. Even with a

20% uplift applied, that brought out a lower figure than the Scotland Excel weighted

figure. Indirect costs such as refrigeration and beakers did not have to be taken into

account, as the benefit was the milk itself, but were in fact included in the NRMU

data.

(ii)

The BRIA showed that other options had been considered. Responses from

stakeholders had been taken into account. The respondent had been entitled to

conclude that there would be no impact on competition.

(iii)

It had not been irrational to base the LSRs on the Scotland Excel framework

prices, when they were higher than the NMRU figures with an uplift. The 20% uplift

was irrelevant since it had been applied to NRMU data which were not used, but in

any event it was not arbitrary.

(iv)

The court had no evidence that settings could not afford to purchase milk for

the LSRs. The respondent was entitled to fix a policy with "bright lines". There

would always be hard cases, and some settings which might lose out. That did not

render the scheme as a whole irrational, or detract from the purpose of the scheme,

which was to benefit children. Any anomalies would be looked at as part of the

review process. There was no need to revisit the scheme at the present time.

The law

[64]

The starting point is to recognise, as senior counsel for the respondent submitted,

that in matters of social and economic policy, the court must accord appropriate respect to

the choices made by the legislature and by government, albeit there will come a point at

which the justification for a policy is so weak that even with the broad margin of

appreciation accorded to the state, the court will conclude that the policy is unjustifiable: R

(SC) v Secretary of State for Work and Pensions [2021] 3 WLR 428, per Lord Reed at

paragraph 144. Another way of expressing this is to say that the degree of intensity with

which the court will review the reasonableness of a public law or act or decision (including a

provision of secondary legislation) varies according to the nature of the decision in question,

the degree of intensity being less in areas of social and economic policy: R (Pantellerisco) v

Secretary of State for Work and Pensions [2021] EWCA (Civ) 1454, Underhill LJ at

paragraphs 56 and 57. The idea that there is a sliding scale of margins of appreciation, and

consequently in the degree of intensity of review, was also explored by Green J in R (British

Academy of Songwriters, Composers and Authors) v Secretary of State for Business, Innovation and

Skills [2015] Bus LR 1435, particularly at paragraphs 143 and 144, where he drew a

distinction between macro and micro issues of policy. It has been said in the Court of

Appeal that the threshold for establishing irrationality is very high but is not insuperable:

R (Johnson) v Secretary of State for Work and Pensions 2020 PTSR 1872, Rose LJ at

paragraph 107. I acknowledge this, but how high the threshold is in any particular case

must to an extent depend on the degree of intensity of the review.

[65]

With these comments in mind, I turn to what the authorities have to say on

irrationality. While in Pantellerisco Underhill LJ hinted that a court might one day need to

undertake a wide ranging analysis of the cases on irrationality, he was broadly content (at

paragraph 55) to adopt the general formulation set out in R (Law Society) v Lord Chancellor,

above, at paragraph 92 where the court identified two aspects of the law of irrationality as a

basis for judicial review. The first is Wednesbury unreasonableness (also formulated as a

decision outwith the range of reasonable decisions open to the decision maker); the second is

a demonstrable flaw in the process of reaching the decision in question, such as reliance

being placed on an irrelevant consideration, lack of eviden ce to support an important step in

the reasoning or a serious logical or methodological error.

[66]

This second aspect is closely linked to the petitioner's second ground of challenge a

failure to make proper inquiry so as to fully inform the decision maker on all relevant

matters before making a decision which is either a subset of the flawed process aspect, or a

third aspect of irrationality in its own right. R (Campaign Against Arms Trade) v Secretary of

State for International Trade [2019] 1 WLR 5765 contains guidance on the duty of inquiry on a

decision maker. The facts were far removed from the present the Secretary of State had

failed to follow published criteria in granting an export licences for the sale of arms to Saudi

Arabia, the Court of Appeal holding that he had acted irrationally, and unlawfully, by

failing to make proper inquiry into past violations when assessing future risk but the case

nonetheless contains useful statements of principle on the approach to be taken by the court

where irrationality is claimed, and provides a useful analysis of what the court should or

should not do in relation to matters which are within the expertise of the government. In

particular, the wisdom of government policy is not a matter for the court: paragraph 54. On

the other hand, if the government has erred, it is the court's role to say so: paragraph 56.

[67]

At paragraph 58 the court referred to the "Tameside duty", emanating from Secretary

of State for Education and Science v Tameside Metropolitan Borough Council [1977] AC 1014, 1065,

observing that this was a "specific application of the doctrine of irrationality", being

"the duty which falls upon a decision maker to `take reasonable steps to acquaint

himself with the relevant information' in order to enable him (sic) to answer the

question which he (sic) has to answer."

At paragraph 59 the Court set out the principles derived from Tameside, as formulated by the

Court of Appeal in R (Balajigari) v Secretary of State for the Home Department [2019] 1

WLR 4647, paragraph 70. These can be summarised as follows. First, the decision maker

need only take reasonable steps to inform itself. Second, it is for the decision maker, not the

court, to decide upon the manner and intensity of the inquiry (subject to Wednesbury). Third,

the court should intervene only if no reasonable decision maker could have been satisfied on

the basis of the inquiries made that it possessed the information necessary for its decision ;

the court's own view as to the adequacy of the inquiries is irrelevant. Fourth, the court

should establish what material was before the decision maker and should strike down a

decision not to make further inquiries only if no reasonable decision maker could conclude

that the inquiries made were sufficient. Fifth, the principle that the decision maker must call

its own attention to considerations relevant to the decision, a duty which may require it to

consult outside bodies with a particular knowledge or involvement in the case, does not

spring from a duty of procedural fairness to such bodies, but from the decision maker's duty

so to inform itself as to arrive at a rational conclusion. Sixth, the wider the discretion

conferred on the decision maker, the more important it is that it has all the relevant material

to enable that discretion to be properly exercised.

[68]

Finally, even if a scheme is not in itself irrational, a refusal to rectify anomalies which

have arisen may be found to be so. This was the position in Johnson, above, where the court

held that the Secretary of State's refusal to modify the Universal Credit Regulations to solve

problems created by the so-called "non-banking day salary shift", which had caused great

hardship to the claimants, was irrational. In reaching that view the court weighed the

disadvantages of not "fine-tuning" the regulations against the disadvantages of adopting a

solution, and considered whether a solution would be consistent or inconsistent with the

universal credit regime.

Decision on irrationality

[69]

I observe at the outset that this part of the petitioner's challenge to the regulations

and the guidance is not so much to the concept of the LSR itself although doubtless it

would have had a great deal to say about that, had it been consulted but to the rates

ultimately fixed, and the process by which they were arrived at. The petitioner does not

challenge the rationality or reasonableness of the policy decision to replace a system of

reimbursement of actual costs with one whereby a formula is adopted in terms of which

settings are paid a notional cost in advance. Such a challenge would have been unlikely to

succeed, since that decision undeniably lies within the sphere of the state, being a matter of

social or economic policy. Equally, had it been the respondent's avowed intention to reduce

the overall cost of the scheme, or to pass part of the cost of providing milk or a non-dairy

alternative on to settings, that would also be a matter of political judgment which would

enjoy a wide margin of appreciation or at least the domestic equivalent. But that is not the

respondent's policy aim: rather the aim is to increase access to milk by pre-school children

in child care settings and incidentally to ease the administrative burden on settings, without

reducing the overall cost. What is under consideration is the respondent's means of

implementing its policy. I consider that this allows for a higher degree of intensity of review

than if the policy itself were under challenge: cf R (British Academy of Songwriters, Composers

and Authors) v Secretary of State for Business Innovation and Skills, above. There, as here, the

decision about the means of implementing the policy in question was not a high level

political judgment. The scope of the court's role is conveniently summarised in

paragraph 145: it is not to substitute its own decision, but is to ensure that the decision

maker -

"has collected sufficient relevant evidence to meet the needs of the decision to be

taken; ensure that the inferences that are drawn from the evidence are rational and

reasonable and within the boundaries of that which may properly be inferred; and

assess whether errors that are made are material to the final decision."

In the following paragraph Green J went on to say that an error may be fundamental, even if

far from obvious; it will be manifest if it goes to the heart of the impugned measure.

[70]

Adopting that approach, I propose to consider the five main themes highlighted in

the petitioner's submissions, as summarised above in paragraph [62]. In doing so I will

examine whether the respondent had sufficient evidence for the fixing of the LSRs at the

rates selected; and whether the respondent's reasoning stands up to scrutiny when viewed

in light of the policy aim being pursued.

The 20% uplift

[71]

In paragraphs [56] and [57] I have set out the competing explanations given by the

respondent at different times for the 20% uplift having been applied to the NMRU data.

While the second explanation (but not the first) is supportive of the position averred in the

answers, the competing explanations cannot both be correct. If the earlier one was an

accurate reflection of government thinking at the time and why would it not be? it

cannot also be correct that the 20% uplift was to take account of either price variation or the

cost of non-dairy alternatives. The respondent appears to have been unsure as to what the

20% uplift represented, or was cavalier as to what explanations it gave, but on any view,

the 20% was never represented as being an estimated uplift solely to take into account the

cost of non-dairy alternatives. At best for the respondent, if a 20% uplift was applied, some

unspecified and arbitrary part of that was to cover the cost of non-dairy alternatives.

[72]

Senior counsel for the respondent submitted that what the 20% represented is

irrelevant, since it was not applied to the figure of 19.6 pence obtained by the second

method. It followed that criticism of the 20% uplift and any suggestion that it was arbitrary

and irrational was of no consequence as it had not been applied to the Scotland Excel figure.

The respondent's position is that because a higher figure was adopted than one brought out

by the NMRU data with an uplift, that in itself provides sufficient justification for the figure.

I do not see how the fact that no 20% uplift was applied to the Scotland Excel figure assists

the respondent. If anything it does the opposite. The Scotland Excel data may have brought

out a higher figure but it was nonetheless a figure based on local authority costs under

existing contracts for the supply of milk; in other words, it was based on historical data and

by definition it made no allowance for the cost of non-dairy alternatives. It was simply a

higher base figure for milk than the one brought out by NMRU data. If it was appropriate to

increase the latter by a percentage, whether to take account of the possibility of children

taking more milk, or variations in price, or the provision of non-dairy alternatives, then it

was as a matter of logic equally appropriate (and necessary) to increase the Scotland Excel

figure also. The failure to do so is illogical, and irrational.

Failure to take into account non-dairy alternatives

[73]

It follows from the foregoing and from the absence of any evidence or even assertion

to the contrary, that at no stage was any inquiry made by the respondent into the cost of

non-dairy alternatives, or what proportion of eligible children might be likely to require

them (bearing in mind the reference not only to medical, but to religious or ethical reasons).

At best for the respondent, a percentage uplift, which may or may not have been intended to

cater to an unknown extent for the provision of non-dairy alternatives to some children, was

applied to a figure which was not ultimately used in developing the scheme. However,

there is no escaping the fact that the data which was used was data which captured only the

price of milk. The respondent made no attempt to capture data either as to how many pre-

school children would require non-dairy alternatives to milk, whether for health, ethical, or

religious reasons, or the cost of providing such alternatives. Had such enquiries been

carried out it is likely, on the affidavit evidence of Mr Thornes and Mr Hogg, that the

respondent would have ascertained that the cost of non-dairy products can be twice as

expensive, certainly significantly more expensive, than cow's milk. The consequence of the

respondent's omission is that a scheme which is intended to include the cost of providing

non-dairy alternatives to milk to some children contains no provision whatsoever for the

funding of such alternatives. Whatever degree of intensity is applied, that is a manifest

error. It is a decision which no reasonable minister could have reached. When the Tameside

principles are applied, it can be said that no reasonable minister could have been satisfied on

the basis of the inquiries made (because no inquiries were made) that it possessed the

information necessary for its decision. Even if the reasoning which was apparently

utilised that the base figure of 19.6 pence need not be increased, because it was higher than

the NMRU figure with a 20% uplift was not flawed (which it is, for the reasons given in the

previous paragraph), the decision still could not stand because that 20% was itself not based

upon any empirical evidence of any sort. Whatever the ultimate methodology used, it

should be self-evident that in calculating the funding required for a fully-funded scheme

which includes a non-dairy alternative to milk, some attempt must be made to ingather and

evaluate evidence as to the overall costs which provision of that alternative will incur (as

was done for milk, and snacks). For these reasons, the new scheme is irrational insofar as it

fails to take account of the cost of non-dairy alternatives to milk.

The use of Scotland Excel data

[74]

Although the petitioner complains that the respondent failed to make any inquiry

into prices for milk paid by private settings, that is not entirely correct, as the BRIA discloses

that the respondent was aware of the prices being charged by milk agents. The petitioner

was also aware of the historical NMRU data, which covered all settings, and of the number

of private settings. This complaint is therefore not so much one of failure to make inquiry,

as of a failure to take into account information which the respondent had in its possession. I

have come to the view that the respondent's approach here, too, is irrational both in the

sense that no reasonable minister could have adopted it, and in the methodology used. This

is not a criticism of the respondent's high level approach of basing the LSR on an average

figure; rather, at a more micro-level, it is a criticism of the means adopted of going about

that task. An average is essentially a figure arrived at based on a range of data. Just as it is

not permissible to pick and choose in deciding whom to consult, one cannot pick and choose

in deciding the data upon which to base an average; at least, not without giving an

explanation as to why certain data has been excluded. If, for example, the respondent had

decided that the figure charged by (say) McQueens Dairies was an outlier, because it was

uncompetitive, it may have been open to it to have left that (and higher figures) out of

account if there was a rational basis for doing so. What the respondent was not entitled to

do was to select a subset of all settings (namely, settings which could access the Scotland

Excel framework), leaving out of account prices paid by all other settings, in the knowledge

that the latter were higher. Such an approach is illogical, irrational and a manifest error.

The respondent is not saved by saying, as its senior counsel did, that the majority of settings

are local authority settings. That is an observation of fact, rather than a justification. It is not

a legitimate reason for excluding the minority when calculating the average. If the number

of private settings is small, the average figure arrived at will be closer to the Scotland Excel

figure than to an average of the private settings figures, but it would nonetheless be higher

than a figure which excludes the private settings altogether. That the respondent's approach

is irrational is further confirmed when viewed through the prism of what it was trying to

achieve greater access to free milk across Scotland including private settings, without

cutting costs and with no impact upon competition, and freedom on the part of settings to

use existing suppliers. Accordingly, I find that the scheme is also irrational insofar as the

LSRs were based only on Scotland Excel data.

Erroneous proxies failure to take into account other costs

[75]

The respondent did not attempt to argue that it had taken such costs into account

(other than to the extent they were included in the NMRU data). Rather, its position was

that the only cost which had to be considered was that of the milk itself, since that was the

benefit being provided. I consider that is too narrow an interpretation of the legislation.

Section 13 of the 1988 Act refers to a scheme including provision for a benefit to consist of

food being provided [emphasis added] by a childcare provider. The benefit is not simply

the food, but its provision. Regulation 6 imposes an obligation on settings to provide the

benefit of the milk. It is an error in law to completely exclude from consideration the other

costs of provision, including storage, and the means by which the child may drink the milk

(cups or beakers). The scheme is also irrational insofar as the calculation of the LSRs did no t

take such costs into account.

The BRIA

[76]

The crux of the petitioner's argument about the BRIA is that the respondent simply

did not have the data which justified its somewhat bald assertion that there would be no

impact upon competition. It had conducted insufficient inquiry to entitle it to make that

assertion. Moreover, since the funding method was not then known, and the rates of the

LRSs had not been set, it could not be in a position to state with confidence that there would

be no impact upon competition. The respondent states in response that the BRIA shows that

it had considered other options. It could not be said to be plainly wrong. If certain players

were driven out of the market, that would only be known with the benefit of hindsight.

[77]

The short point here is whether by stating emphatically (to the extent of underlining

the word) that there would be no impact on competition, the respondent proceeded upon an

error of fact. The difficulty the respondent faces is that it nailed its colours so emphatically

to the mast, at a time when it could not have known what the impact would be, because the

funding of the new scheme had not been determined, let alone the rates themselves. It is no

answer to the petitioner's challenge to say that the impact is known only with the benefit of

hindsight. The position might have been different had the statement about the impact on

competition been qualified in some way, but it was not. The unqualified statement which

was made, without any empirical basis for making it, did result in the respondent

proceeding upon the basis of a material error.

[78]

That said, in some ways the argument about the BRIA is the flip side of those about

non-dairy alternatives and the exclusive use of Scotland Excel data. If higher rates had been

(or are in the future) adopted, the statement in the BRIA that there will be no impact on

competition may yet turn out to be correct. The argument to an extent becomes circular.

However, in light of the decision to fix LSRs at the rates selected, and the fact that we are

told that it was never the respondent's intention to pay settings a sum which represented the

cost of providing milk, it can be said that the respondent did proceed under a material error

in concluding that there would be no impact on competition. The scheme is to that extent

irrational and unlawful.

Refusal to amend the scheme

[79]

The final issue is whether the respondent's refusal to amend the scheme,

notwithstanding the difficulties with the rates that have been drawn to its attention, is itself

irrational and unlawful. The respondent has two broad answers to this. The first is to refer

to the scope in regulation 10 for payments to settings to be adjusted. The second is to say

that any difficulties will be ironed out in year two. In saying the latter, I take it to be the

respondent's position that it acknowledges that the scheme has imperfections.

[80]

Dealing with each of these in turn, I have already discussed regulation 10 and the

lack of clarity surrounding its meaning. Whatever else it may or may not allow, it does not

provide for the payment of an amount to any setting in excess of the LSR on the basis that

the LSR is insufficient to meet that settings costs. As senior counsel for the petitioner

submitted, it is not a panacea for the scheme's ills. Indeed, the rigidity of the scheme, and

the absence of any mechanism for adjusting the amounts payable in individual cases

(coupled with the agreement in year one that local authorities must apply the LSRs) is

arguably a further aspect of the scheme's irrationality, when assessed against the

respondent's aim of increasing access across Scotland. As for the decision not to review the

scheme in the first year of operation, that is more finely nuanced. This issue only has

relevance if I am wrong in holding that the scheme itself is irrational (because it must also be

irrational to refuse to amend an irrational scheme). The question is, are the problems which

have in fact emerged such that a refusal to amend the scheme is irrational? This involves

considering whether the scheme is achieving its policy aims, and conducting a balancing

exercise of the respective advantages and disadvantages of removing any anomalies:

Johnson, above.

[81]

Against a background where the aim was to increase access to free milk across

Scotland, in all settings including private nurseries and childminders, without impacting

competition, the scheme (with the benefit of hindsight) has thrown up anomalies, all

previously referred to: the surprising difference between the LSR and Scotland Excel data in

at least two local authority areas; the slump in the use of milk agents; and the inability of

some private settings to afford to buy milk at the LSR rates. The respondent is aware of

these difficulties, which have been flagged at Implementation Group meetings since the

scheme came into effect (and in this petition). Where the scheme is Scotland-wide, it will not

do to say that there will be winners and losers: that does appear to run counter to the stated

underlying policy; the more so if an entire local authority area is a loser. The factual

position in Johnson was much more complex, where it was argued that the underlying

computer software could not easily be modified to accommodate any change (which

argument the court rejected). Here, it is not suggested by the respondent that there would

be any great practical difficulty in implementing modifications to the rates or to the scheme.

From what was said, that is likely to occur in any event, in year two. The petitioner submits

that by then it may be too late: certain suppliers may already have exited the market.

[82]

For these reasons, I consider that the disadvantages of not reviewing the scheme now

significantly outweigh the disadvantages of leaving matters as they are until the year two

review. I have concluded that even if I am wrong in holding that the scheme was irrational,

the refusal to consider amending it is itself irrational.

Third ground of challenge

[83]

The third ground of challenge was to the Amendment Regulations. As I have

already recorded, both parties accepted that this challenge stood or fell with the broader

challenges to the scheme. While that is undoubtedly true insofar as the failure to consult is

concerned the failure to consult cannot be cured by the passing of the amendment

regulations the position is slightly different when it comes to irrationality, since it was the

amendment regulations themselves which brought into law, as opposed to guidance, the use

of the LSRs to fix rates; so all that I have said about irrationality applies with even greater

force to the Amendment Regulations than to the regulations as originally enacted. This was

recognised by senior counsel for the petitioner at the end of the hearing when he surmised

that one remedy might be to reduce only the Amendment Regulations, so that the scheme

would revert to one based on market price.

Disposal

[84]

In summary, I have decided that the new scheme is unlawful in that (a) the

respondent did not undertake a proper consultation on a key aspect of it, the LSRs, and

(b) the fixing of the LSRs was irrational. The petitioner's senior counsel was at pains to

stress that the petitioner has no desire to interrupt the supply of free milk (and a snack) to

pre-school children. As agreed, I shall put the case out by order to discuss what orders

ought to be made in light of this opinion.