Wiltshire County Council at fault for using the Matrix Assessment Tool to reduce the of level respite care and transport provisions

Decision Date: 12th April 2018

What Happened

Mrs N complained on behalf of her adult son, Mr P.

Mr P was cared for at home by his parents. He had complex needs, including severe learning difficulties and epilepsy.

Mr P’s father, Mr N, had recently become disabled, which meant that Mrs N was carer for both. Mr P qualified for a Motability vehicle.

Since leaving school, Mr P had attended a day care facility, which also provided his residential respite care. This centre was 10 miles away from his family home, until they all moved to a location 25 miles away in late 2015. The Council provided transport for Mr P to the day care placement.

The Council used its ‘Matrix Assessment Tool’ to allocate the amount of support to Mr P and his family.

The Matrix Assessment Tool

The Council adopted its ‘Matrix Assessment Tool’ (‘MAT’) in 2007. Its stated purpose was to ‘ensure limited resources could be fairly allocated to families based on need’. There are four stages:

Stage one: assess applicant’s need for respite care. The MAT allocates a score in various categories to reflect the applicant’s level of disability or need.

Stage two: assess the ‘current provision of service’. This provides a second score which must be subtracted from the ‘need’ score.

Stage three: assess ‘the carer’s situation’. The ‘carer’s score’ is then added to the previous total.

Stage four: The Council assesses those requiring respite care and lists applicants according to their scores. The Council then calculates the number of respite beds available. It then divides the number of beds available by the number of beds required. This calculation creates the ‘coefficient’. The Council then multiplies the score of each service user by the coefficient. This produces a number which will be the number of nights’ respite s/he will receive that year.

For example; the Council has 1400 bed nights available and the apparent demand is for 2750 bed nights. The Council has half the number of beds it needs which produces a coefficient of 0.5. With this coefficient in place, an applicant with a score of 160 would receive 80 nights’ respite care (160 multiplied by 0.5).

Until 2016, the Council provided Mr P with 104 days of respite care per year (4 consecutive nights every other week).

The Council accepted that Mr P’s needs had not changed since 2013.

When the family moved house at the end of 2015, Mrs N wanted Mr P to keep his existing care package, so before they bought the property, she had emailed the Council; ‘Good morning, my son uses adult care services in the [local town] area, receiving transport to and from [Mr P’s] day care centre. Could someone make contact with me please as we have now sold our house and need to know if the house we are about to offer on is still [in] the catchment area’.

A few days later, the Council wrote back saying the house was ‘in our catchment area’. Mrs N wrote again saying ‘we have tried to get some guide lines as to where we can move to whilst keeping [Mr P’s] all-important care package and understood, as long as we didn’t move to [local village] which we understood was the boundary, we should be fine’. There was no record that the Council ever responded, and Mrs and Mr N went ahead with buying the new property.

In June 2016 the Council told Mrs N it was cutting Mr P’s transport funding, and asked her to provide 2 journeys a week to the day care centre. The alternative was that she could pay the Council £30 per trip.

The Council also told her it was cutting his respite care from 104 nights to 68, using the MAT.

Mrs N complained about these decisions. She was paying the Council for transport, because she highlighted that she still had to care for her husband, so the three hour round trip to Mr P’s day care would leave Mr N uncared for at home. The Council said Mr N did not need constant care. There was nothing mentioned in the report as to whether Mr N had a care package in place or if he had been assessed.  

A September 2016 annual review of Mr P’s needs concluded that there was no change in Mr P’s needs, and that they were all being met.

Mrs N complained about the outcome of this assessment. She said that the respite care was not enough, and complained about the transport. 

The Council maintained that he should only receive 68 days respite a year, as per his MAT score. It said Mrs N had been receiving a level of respite care above that which it gave in similar cases.

After several meetings and lengthy email exchanges, the Council confirmed that it would be imposing the changes to the transport and respite care allocation.

The Council said that that the changes would be introduced ‘gradually’ and fully in place by April 2017, but actually introduced the changes immediately.

The Council wanted Mr P to attend a day care facility closer to their new home, and highlighted that it would be cheaper.

Mrs N said this was ‘non-negotiable’. She had specifically checked with the Council that they could move to their new house and Mr P would be able to continue with his care package.  

She also highlighted that her need for respite had increased right at the time the council cut it, because of Mr N’s needs.

The LGO asked the Council to restore respite care whilst it investigated Mrs N’s complaints. The Council responded ‘…The allocation of support being provided is at the top level…therefore I am not able to agree funding a person above the amount the council deems appropriate’. The report did not state who from the Council replied.

What was found

Overall, the LGO found that the Council was at fault for the way in which it reduced both the level of respite care and the transport provision.

Funding limits

The Council cannot set maximum budget levels. Eligible needs (including for respite) must be met no matter what the cost (if there is only one way of meeting them). The Council said Mr P’s disability fell into a certain band and therefore his funding could not exceed a certain level. The LGO emphasised that this approach does not accord with the Care Act. The Council may use bandings as a guide but, as the Care and Support Statutory Guidance states, such systems are unlikely to work in complex cases like Mr P’s.


The Council’s decision to ask Mrs N to provide transport/additional funding seemed to be based on a general withdrawal of provisions/cost cutting exercise. This was not in line with the Care Act, as it was not based on assessments of needs, and therefore amounted to fault.

The Council’s guidance on transport says ‘The test of eligibility is ‘Would the failure of the Council to provide transport result in an eligible need for services going unmet?’ In this case it would, because asking Mrs N to provide transport would have resulted in Mr N’s needs being unmet (he would have been alone for 6 hours).

The Council also has a duty to consider the wishes of those involved; service users and carers. It cannot force an unwilling carer to provide care she reasonably states she cannot give. There was no evidence that the Council considered whether Mrs N would be able to provide the transport.

The Council was at fault for the way it reduced funding for transport. It failed to follow its own guidance, and comply with the Care Act.  This resulted in Mrs N being required to pay £60 per week towards the cost of transport when she should not have been.

The Council agreed to repay Mrs N, but it argued the amount should be reduced by the Council’s weekly transport contribution (payable by all users) multiplied by the number of weeks reduction.

The Council charges all transport users £3.70 a day (£18.50 a week). The LGO accepted this as fair, so the Council agreed to repay Mrs N £747.50.

The Matrix Assessment Tool

The MAT predated the Care Act. Its purpose was to ration available resources.

The Care Act requires councils to meet eligible needs. It does not allow rationing for any reason. If a council cannot meet an eligible need, it must pay someone else to meet it.

The MAT is, therefore, incompatible with the Care Act. The Council was at fault for continuing to use it.

The Council was also at fault for introducing the reduction in care as quickly as it did. In practice, the reduction was made almost immediately.

The LGO recommended that the Council:

  • Apologise to Mrs N
  • Restore the previous level of respite care pending a re-assessment compliant with the Care Act 2014
  • Confirm it will offer her 24 days respite care to be taken at a time of her choosing in recognition of the respite care wrongly withdrawn
  • Pay Mrs N £747.50 in recognition of money she paid the Council for transport
  • Pay Mrs N £500 in recognition of distress and time and trouble
  • Review its policy and procedure on respite care to reflect the requirements of the Care Act 2014
  • Review other files for evidence of use of the MAT. It should write promptly to anyone similarly affected and review their cases
  • Review the files of anyone whose transport was cut to ensure these cuts were compliant with the Care Act
  • Inform the Ombudsman of the numbers of people involved and undertake to review all cases within a further three months
  • Ensure all staff receive training in the requirements of the Care Act and the relevant guidance
  • Review all relevant documents to ensure they reflect the current law.

Points for the Public, Councils, service users, advocates, commercial providers of RAS or other rationing tools etc.

We do expect this report to cause a certain amount of choking in social services management circles, and we salute the ombudsman for saying it like it is.

Effectively, the LGO is saying that if a tool cuts the sums that have been plugged INTO the calculation to levels that are arbitrary and renders the council apparently unwilling to meet assessed eligible unmet needs, then that tool is done for as non-compliant with the Care Act.

That will cause consternation amongst providers of such tools who have made money by dint of allowing councils to implement budget cuts on THEM by turning down the allocation of the pounds to points ratio, in the resource allocation tool. Of course it’s legal to do that if there’s an evidence base for so doing, like the cost of the services going down, but how likely is THAT, these days, we would ask?

Here’s the legal reasoning behind the LGO’s stance: ever since Savva and KM, the public law principle in play (for a RAS or other tool to be a lawful starting point for care planning) has been that there must be a rational nexus between needs and points; the points and the pounds allocated to the points – and between the pounds and the local market rate for quality provision. Without that, it’s not possible for a council to say that the personal budget has been rationally calculated to be sufficient to meet needs, and thus able be signed off as “the cost to the local authority”.

When one applies the law to the context of respite, there’s a lot more to be said, too.

Respite is the service user’s need, not the carer’s. A carer can say NO at any point. They don’t have to prove that they are unable to care. The Guidance and the caselaw make it totally clear that there is no obligation in the UK for an adult to care for any other adult; it’s the State’s responsibility, unless someone is willing. Respite therefore is not something that can ever be rationed. A person’s willingness to carry on caring informally is only relevant to the council’s overall discretion whether to support a person to stay in their own home, or elects to make an offer to meet the person’s need in a different adequate alternative setting in which the council can make a placement contract. If a person says they’re not caring on a given date, the council has to meet the needs, if they have been logged as needs that were eligible. Flat rate offers of x weekends or y days of ‘respite’ are just a starting position, not a lawful discharge of the care planning function for the client.

“There are many variations of systems used to arrive at personal budget amounts, ranging from complex algorithmic-based resource allocation systems (RAS), to more ‘ready-reckoner’ approaches. Complex RAS models of allocation may not work for all client groups, especially where people have multiple complex needs, or where needs are comparatively costly to meet, such as in the case of deaf-blind people. It is important that these factors are taken into account, and that a ‘one size fits all’ approach to resource allocation is not taken. If a RAS model is being used, local authorities should consider alternative approaches where the process may be more suitable to particular client groups to ensure that the personal budget is an appropriate amount to meet needs” (Care and Support Statutory Guidance 2014, 11.23)

Thus no council can say before it spent thousands on a RAS, that it wasn’t warned of the risks.

One might suggestion that social work discretion, social work supervision and forums that develop social work values driven consensus as to how much of anything is enough for meeting needs, would be a better way forwards.

Transport is a bit different. Transport will be an essential part of a service offer, but transport is often subsidised by a person’s district council, for those on benefits related to disability, and transport within a care package can be charged at full cost under the Charging regulations, so that it’s often cheaper for a person to sort their own transport solutions out.

If someone has a Motability vehicle and someone willing and able to drive it, then whilst their needs may be eligible, but won’t need to be met under s18.

Therefore we can’t say we think that the council was wrong to set itself the question ‘would a need be unmet?’ The question was no doubt designed to make carers really struggle to have the moral fortitude to be seen to say no, they wouldn’t be doing that journey. Mrs N could have and should have just said no.

Since she didn’t appear to say no at any point, one can understand why a council would feel entitled to exploit that ongoing willingness, but recent case law makes that no longer possible. Cases like Redbridge require a council to go on, if choosing to treat acquiescence as good enough, specifically to address evidence about ability, and make a reasonable decision about that.

We think that the LGO got the answer right but for the wrong reason, therefore. We do not think it was correct of the LGO to make the Wiltshire internal question stretch to cover the question whether Mr N’s father would then have unmet needs! If he did not need to be supervised at all times, the LGO could not legitimately decide that he did; and therefore the LGO should just have said Mrs N has been conveying that she is unwilling.

If you want help, please consider seeking advice from CASCAIDr via our referral form on the top bar menu of the site.

The full Local Government Ombudsman report of Wiltshire County Council’s actions can be found here by clicking on the PDF download at the right-hand side.