Data and methods

Background

This analysis uses data from the 2025-26 Health and Care Experience (HACE) survey. We selected all respondents who reported that they provide unpaid care. HACE defines unpaid caring as providing regular help or support to look after someone who is living with a disability, a physical or mental health condition, or has problems related to old age outside of paid employment or formal volunteering.

HACE was designed to measure patient and client experiences of a range of health and care services. Originally, the survey only measured experience of GP services. However beginning in 2013 a range of other services were added to the survey. This analysis substantially expands on the initial purpose of the survey, but can still provide some meaningful insights on reported need and unmet need for support for caring. We have outlined both the strengths and the limitations of this analysis below.

Note on response rates

In 2025-26, HACE received 94,026 total responses – a 19% response rate. 15,655 responses were from people who reported providing unpaid care. 98% of unpaid carers responded when asked about their need for and receipt of support for their caring roles.

Completion rates to individual questions in the survey varied, as did response rates between different groups of respondents, as detailed in the HACE 2025-26 Technical Report. For the specific questions we looked at in this analysis, question completion rates amongst unpaid carers varied between 78% and 99%. The question around whether respondents felt they had “a say in how services are provided to the person(s) I care for) attracted the lowest number of question completions amongst unpaid carers (78%). The total number of responses included at each stage of the analysis is included in the supplementary tables.

Definitions of carers support

HACE asks unpaid carers whether they had received “services or support for me personally to help me have breaks from caring”, and/or “services provided to the person I care for such as overnight or day services to allow me to have a break” in the last 12 months.

We count carers who selected either of these options as having received “breaks-related formal support”. We count carers who did not select these options but did select “help from Carer Centre/ local organisation” as having received “non-breaks-related formal support”.

Given the question wording, it may be possible that some unpaid carers who reported receiving “services or support for me personally to help take a break from caring” may be referring to informal support from family and friends, however given the reference to “services” we have classified this response option as formal support for the purposes of this analysis.

53% of carers who had received support reported receiving only support that was informal in nature (i.e. support from family, friends or neighbours) (Figure 9). 

36% indicated that they had received some formal support as part or all of their support. This was made up of 13% of carers who had received formal support which included some kind of help to take a break from caring, and 23% who had received other types of formal support that did not include support for breaks.

For some other groups (i.e. carers who told us they received only “other support”  (7%), or “other support” in combination with support from family, friends and neighbours (4%)) it was not possible to discern whether the help they got for their caring roles was formal or informal in nature. These 11% of respondents were excluded from the analysis presented in Section 3 and Section 4.

Figure 9. Based on respondents’ answers to the survey questions, we estimate that just over half (53%) of respondents received only informal support and just over a third (36%) received formal support.

Bar chart showing the percentage of unpaid carers receiving each type of support.

Strengths of this analysis

1. HACE has a number of advantages over other data sources for analysing carers needs and unmet needs for support.

There are limited data available to understand support for unpaid carers in Scotland. HACE has a number of advantages for this kind of analysis over other data sources because it addresses a large number of carers, and it asks specifically about unmet needs and carers experiences of caring. It is therefore well placed to provide more granular insights on carers experiences of needing and receiving support.

The Scottish Health Survey (SHeS) also has a large sample size and asks carers whether they receive support, but it does not ask them whether they have an unmet need for support.

SHeS does include some questions on the impacts of caring – for example, on employment – however does not explicitly address carer’s personal experiences of delivering care, or receiving support.

The Carers Census provides more granular insights into type and experience of carers support, however it only collects information about carers who are in contact with local services, including Local Authorities and Carers Centres. As such, it addresses a much smaller number of carers than HACE, and does not include carers who are not known to local services, such as those receiving no support or only informal support from family, friends and neighbours. Therefore, inferences about the caring population as a whole cannot be drawn from the Carers Census data.

For more information on the insights regarding unpaid carers support available in SHeS and the Carers Census see Annex A.

2. This analysis provides first in kind estimates which add more granular insight into carer’s experiences

The HACE national report for 2025-26 reports on unpaid carers responses to standalone questions in the survey. This means it does not seek to group responses together into broader categories (e.g. “received formal support”, “received informal support only” etc.) or, to explore how responses to individual questions varied by different demographic characteristics (e.g. hours of care provided per week, age, sex, deprivation etc.). As such, the estimates presented in this report offer new and more granular insights on the experiences of unpaid carers.

3. This analysis provides estimates of direct relevance to the following legal statutes and Scottish Government policies regarding unpaid carers

The following statutes and policies are relevant to metrics analysed in this report:

• The Carers (Scotland) Act (2016) established a legal right for carers to be involved in assessing the needs of the person they care for and in decisions about discharging them from hospital.
• Ensuring carers’ voices are heard and their views and experiences are taken into account in decisions which affect them was named as a priority strategic outcome by the National Carers’ Strategy (2022). 
• The Care Reform (Scotland) Act (2025) establishes a legal right to breaks from caring for those providing unpaid care.

Limitations of this analysis

1. May over-represent the experiences of carers who provide the highest number of caring hours per week.

A higher proportion of unpaid carers who responded to the HACE survey were providing the highest amount of caring hours per week (50+ hours), when compared to findings from Chapter 2 of the 2024 edition of the Scottish Health Survey (SHeS) – another official statistical collection that tracks unpaid carers in Scotland.

This trend was also apparent when we looked at the distribution of unpaid caring hours provided per week by age across HACE and SHeS. At each age bracket, a higher proportion of HACE respondents reported providing 50+ hours of unpaid care per week, compared to SHeS. This difference was most pronounced in the oldest age brackets (65-74 and 75+).

Despite this, we believe that the number of respondents to HACE who provided fewer than 50 hours of care per week is sufficiently large to allow for robust analysis. Where we believe that findings related to other variables – for example findings around the belief that “caring has not had a negative impact on my health and wellbeing” – are being affected by the presence of a high proportion of carers providing a high number of caring hours per week, we have noted this in the report.

2. No specific insight on young carers

The HACE survey is only sent to people who are aged 17 or over at the time of sampling. This means that there are no specific insights on carers aged under 17 available in HACE.

The Scottish Government has identified supporting young carers as a priority in the National Carers Strategy however, given the complexity and self-completion of the HACE questionnaire it may not be suitable to extend the sample to younger respondents.  Other sources of information are available to provide insights on young carers – for a summary of the key sources of information see the Age and Sex chapter of Scotland’s Carers, 2026.

3. There is a risk that non-response to the survey is affecting our understanding of need and unmet need.

Not everyone who was sent the HACE survey responded to it. If the people who do not respond differ systematically from those who do respond in ways that relate to the subject of the survey, this can bias the results. For example, if non-respondents were more likely to have a very high need for support, we may underestimate the true level of need. This is known as “non-response bias”. As far as possible, the survey weights seek to adjust for these general risks, but we have noted that we did not use custom weights for this analysis. (Please see the part on Survey Weights in this section, for more information on survey weighting).

4. We cannot tell from HACE what kind of support carers with unmet need require.

The HACE survey asks unpaid carers whether they have received any support to help with their caring role in the last 12 months. Respondents can select from a list of support types or select “No support or help, but I felt that I needed some” or “No support or help, but I do not need any”. We cannot meaningfully estimate what kind of support for caring these unpaid carers need from the data available in HACE. As such we refer unmet need for “support” generally, throughout this report.

Weightings

As with the main HACE 2025-26 publication, throughout this report the survey results are presented as weighted percentages. Different proportions of people were selected from each General Practice in Scotland to receive the survey, and the response rate varied by age and sex. Weighting is used to produce results which are more representative of the population eligible to complete the survey.

The details of how the weights for the main HACE 2025-26 analysis were calculated can be found in the technical report of the 2025-26 survey.

Confidence intervals

Confidence intervals provide a way of quantifying uncertainty in the survey results. A 95% confidence interval means that, over many repeats of a survey under the same conditions, one would expect that the confidence interval would contain the true population value 95 times out of 100.

For example, given a result of 80% and a confidence interval of +/-3%, there is 95% confidence that the true result is between 77% and 83%.

There is more information on confidence intervals (including how they are calculated) in the technical report of the 2025-26 survey.

The confidence intervals associated with the results presented for this analysis are available in the supplementary tables.

Where confidence intervals of two central estimates overlap, and contain each central point estimate, the differences observed are not significant. Where the confidence intervals do not overlap the differences observed are significant. All other cases require significance testing.

Significance testing

All comparisons which are reported as “higher” or “lower” that are discussed in the report are statistically significant at the 95% level.

Where results for different groups are described as “similar”, differences are not statistically significant at the 95% level.

The approach to significance testing in this report remains consistent with the main HACE publication. Details of this can be found in the technical report of the 2025-26 survey.

Rounding

The estimates we have presented in this report are rounded to the nearest whole number. The results to two decimal places are available in the supplementary tables, along with the 95% confidence intervals for each percentage estimate.

We have calculated percentages that describe the sum of two or more categories from the original (unrounded) estimates and then rounded them to the nearest whole number for this text. This means that percentages in this text that refer to a total across two or more categories may differ from what you would calculate by adding together the individual rounded estimates from each category.

For example, we would present a result of 2.80% as 3% (rounded), and a result of 5.60% as 6% (rounded). However, we would calculate the total (sum) of the two numbers as 8.40%, which we would present as 8% when rounded to the nearest whole number.