Equality Data Improvement Programme project board - equality data audit and protected characteristics overview: February 2022

Background

For the 15 December 2021 Equality Data Improvement Programme (EDIP) Project Board meeting, the Equality Analysis team provided attendees with a paper that summarised the RAG ratings provided by analysts in each analytical area in relation to the collection and publication of data on race (paper 2). As requested by the Project Board, this paper provides an overview of RAG ratings for all nine protected characteristics.

For ease of reference, background information on the Equality Data Audit provided in the December 2021 paper is duplicated in the following section.

Equality Data Audit

Equality Data Audit returns were received from all 10 analytical areas across the Scottish Government in October 2021. Analysts were required to provide RAG ratings, for each dataset, on the collection and publication of data on each of the nine protected characteristics. Additional information was also requested about each dataset on any barriers to equality data improvement, pressures for improvement, and any current work underway or plans to make equality data improvements. 

Analysts were asked to include in their return all datasets used by their analytical area to:

• produce official or national statistics
• inform National Performance Framework (NPF) indicator progress
• produce statistics that have informed Scottish minister's decision-making

For each dataset, analysts selected from a dropdown list to indicate whether data on each protected characteristic is (a) collected and (b) published. Definitions of each RAG rating are as follows:

Red

Collected: no data collected on this protected characteristic.

Published: no published breakdowns for this protected characteristic.

Amber

Collected: data collected but not robust enough to produce reliable statistics for this protected characteristic (e.g. due to small sample size).

Published: breakdowns for this protected characteristic are not proactively published but are available on request.

Green

Collected: data collected that is robust enough to produce reliable statistics for this protected characteristic.

Published: breakdowns for this protected characteristic are proactively published (e.g. when new data becomes available or according to a publication schedule).

Overview across the nine protected characteristics

A total of 199 datasets were included in returns across the 10 analytical areas. Annex A provides a list of analytical areas and corresponding acronyms. As shown in table one, of these:

• age and sex (or gender) were collected in over 6 in 10 datasets, and breakdowns by these characteristics published from around 5 in 10 datasets
• race (or ethnicity) and disability were collected in just under 3 in 10 datasets, and breakdowns by these characteristics published from around 2 in 10 datasets
• marriage and civil partnership and religion or belief were collected in just over 1 in 10 datasets, and breakdowns by these characteristics published from less than 1 in 10 datasets
• sexual orientation was collected in 1 in 10 datasets, and breakdowns by this characteristic published from less than 1 in 10 datasets
• pregnancy and maternity and gender reassignment were collected and breakdowns by these characteristics published from less 1 in 10 datasets

Table 1: All datasets by RAG rating for the collection and publication of data on each protected characteristic. Note that due to rounding percentages may not sum to 100.

Protected Characteristic	Collected				Published
	R (red)	A (amber)	G (green)	No rating	R (red)	A (amber)	G (green)	No rating
Age-	29%	2%	65%	5%	34%	6%	54%	6%
Sex (or gender)	31%	2%	62%	5%	38%	7%	50%	6%
Race (or ethnicity)	58%	9%	29%	5%	62%	9%	22%	8%
Disability	62%	5%	28%	5%	64%	8%	20%	8%
Marriage and civil partnership	76%	3%	15%	7%	78%	8%	8%	7%
Religion or belief	75%	6%	13%	6%	77%	7%	8%	8%
Sexual orientation	79%	6%	10%	6%	80%	9%	4%	8%
Pregnancy and maternity	86%	4%	6%	5%	86%	3%	4%	8%
Gender reassignment	89%	3%	3%	5%	90%	1%	3%	7%

Collection of protected characteristic data by analytical area

This section provides breakdowns of the RAG ratings for the collection of data on each protected characteristic in each of the ten analytical areas across the Scottish Government.

The overall pattern of RAG ratings for the publication of breakdowns by each protected characteristic closely follows the pattern for collection so this information is not presented.

Please note that some datasets included in the Audit returns do not include information about individuals or households, but about businesses, sectors, local authorities or funding allocations. These datasets are included in the figures presented below as, whilst often more challenging, equality breakdowns may still be possible. Nevertheless, we highlight that, even in an ideal world, it would unlikely be possible to collect and publish data on the protected characteristics from all datasets marked as ‘red’. Some datasets do not have a submitted rating for all protected characteristics – the Equality Analysis team is following up with analysts to ensure these gaps are filled.

Due to substantial variability in the numbers of datasets reported, figures should not be directly compared between analytical areas.

Age

The RAG ratings for the collection of data on age across analytical areas shows:

• of the 27 datasets in the Communities Analysis Division (CAD) return, almost 6 in 10 collect robust data on age
• of the 18 datasets in the Constitution and External Affairs Analysis (CEAA) return, almost 3 in 10 collect robust data on age
• of the 23 datasets in the Education Analytical Services (EAS) return, over 9 in 10 collect robust data on age
• of the 63 datasets in the Health and Social Care Analysis (HSCA) return, almost 7 in 10 collect robust data on age
• of the 14 datasets in the Justice Analysis Services (JAS) return, over 9 in 10 collect robust data on age
• of the 11 datasets in the Local Government Analytical Services (LG) return, less than 1 in 10 collect robust data on age
• of the 17 datasets in the National Records of Scotland (NRS) return, almost 9 in 10 collect robust data on age
• of the 19 datasets in the Office of the Chief Economic Adviser (OCEA) return, almost 6 in 10 collect robust data on age
• of the 5 datasets in the Rural and Environmental Science Analytical Services (RESAS) return, 6 in 10 collect robust data on age
• of the 2 datasets in the Transport Scotland Strategy & Analysis (TS) return, all collect robust data on age

Sex (or gender)

The RAG ratings for the collection of data on sex (or gender) across analytical areas shows:

• of the 27 datasets in the CAD return, almost 6 in 10 collect robust data on sex
• of the 18 datasets in the CEAA return, almost 3 in 10 collect robust data on sex
• of the 23 datasets in the EAS return, over 8 in 10 collect robust data on sex
• of the 63 datasets in the HSCA return, over 6 in 10 collect robust data on sex
• of the 14 datasets in the JAS return, over 9 in 10 collect robust data on sex
• of the 11 datasets in the LG return, less than 1 in 10 collect robust data on sex
• of the 17 datasets in the NRS return, almost 9 in 10 collect robust data on sex
• of the 19 datasets in the OCEA return, almost 6 in 10 collect robust data on sex
• of the 5 datasets in the RESAS return, 8 in 10 collect robust data on sex
• of the 2 datasets in the TS return, all collect robust data on sex

Race (or ethnicity)

Note that the figures for race in some analytical areas are slightly different to those provided to the Project Board in December 2021 as analysts have subsequently provided RAG ratings for datasets where no rating was previously provided.

The RAG ratings for the collection of data on race (or ethnicity) across analytical areas shows:

• of the 27 datasets in the CAD return, almost 4 in 10 collect robust data on race
• of the 18 datasets in the CEAA return, almost 3 in 10 collect robust data on race
• of the 23 datasets in the EAS return, just over 6 in 10 collect robust data on race
• of the 63 datasets in the HSCA return, over 2 in 10 collect robust data on race
• of the 14 datasets in the JAS return, almost 4 in 10 collect robust data on race
• of the 11 datasets in the LG return, none collect robust data on race
• of the 17 datasets in the NRS return, almost 2 in 10 collect robust data on race
• of the 19 datasets in the OCEA return, almost 3 in 10 collect robust data on race
• of the 5 datasets in the RESAS return, none collect robust data on race
• of the 2 datasets in the TS return, half collect robust data on race

Disability

The RAG ratings for the collection of data on sex across analytical areas shows:

• of the 27 datasets in the CAD return, over 4 in 10 collect robust data on disability
• of the 18 datasets in the CEAA return, just over 2 in 10 collect robust data on disability
• of the 23 datasets in the EAS return, 7 in 10 collect robust data on disability
• of the 63 datasets in the HSCA return, almost 2 in 10 collect robust data on disability
• of the 14 datasets in the JAS return, just over 2 in 10 collect robust data on disability
• of the 11 datasets in the LG return, none collect robust data on disability
• of the 17 datasets in the NRS return, almost 2 in 10 collect robust data on disability
• of the 19 datasets in the OCEA return, almost 3 in 10 collect robust data on disability
• of the 5 datasets in the RESAS return, none collect robust data on disability
• of the 2 datasets in the TS return, half collect robust data on disability

Marriage and civil partnership

The RAG ratings for the collection of data on

marriage and civil partnership across analytical areas shows:

• of the 27 datasets in the CAD return, just over 2 in 10 collect robust data on marriage and civil partnership
• of the 18 datasets in the CEAA return, under 1 in 10 collect robust data on marriage and civil partnership
• of the 23 datasets in the EAS return under 1 in 10 collect robust data on marriage and civil partnership
• of the 63 datasets in the HSCA return, under 1 in 10 collect robust data on marriage and civil partnership
• of the 14 datasets in the JAS return, almost 4 in 10 collect robust data on marriage and civil partnership
• of the 11 datasets in the LG return, none collect robust data on marriage and civil partnership
• of the 17 datasets in the NRS return, just over 4 in 10 collect robust data on marriage and civil partnership
• of the 19 datasets in the OCEA return, just over 2 in 10 collect robust data on marriage and civil partnership
• of the 5 datasets in the RESAS return, 2 in 10 collect robust data on marriage and civil partnership
• of the 2 datasets in the TS return, half collect robust data on marriage and civil partnership

Religion or belief

The RAG ratings for the collection of data on religion or belief across analytical areas shows:

• of the 27 datasets in the CAD return, just over 2 in 10 collect robust data on religion or belief
• of the 18 datasets in the CEAA return, just over 2 in 10 collect robust data on religion or belief
• of the 23 datasets in the EAS return, just over 2 in 10 collect robust data on religion or belief
• of the 63 datasets in the HSCA return, under 1 in 10 collect robust data on religion or belief
• of the 14 datasets in the JAS return, over 1 in 10 collect robust data on religion or belief
• of the 11 datasets in the LG return, none collect robust data on religion or belief
• of the 17 datasets in the NRS return, almost 2 in 10 collect robust data on religion or belief
• of the 19 datasets in the OCEA return, just over 1 in 10 collect robust data on religion or belief
• of the 5 datasets in the RESAS return, none collect robust data on religion or belief
• of the 2 datasets in the TS return, half collect robust data on religion or belief

​​​​​​​Sexual orientation

The RAG ratings for the collection of data on sexual orientation across analytical areas shows:

• of the 27 datasets in the CAD return, just over 2 in 10 collect robust data on sexual orientation
• of the 18 datasets in the CEAA return, just under 2 in 10 collect robust data on sexual orientation
• of the 23 datasets in the EAS return, under 1 in 10 collect robust data on sexual orientation
• of the 63 datasets in the HSCA return, under 1 in 10 collect robust data on sexual orientation
• of the 14 datasets in the JAS return, just over 2 in 10 collect robust data on sexual orientation
• of the 11 datasets in the LG return, none collect robust data on sexual orientation
• of the 17 datasets in the NRS return, none collect robust data on sexual orientation
• of the 19 datasets in the OCEA return, just over 1 in 10 collect robust data on sexual orientation
• of the 5 datasets in the RESAS return, none collect robust data on sexual orientation
• of the 2 datasets in the TS return, half collect robust data on sexual orientation

​​​​​​​Pregnancy and maternity

The RAG ratings for the collection of data on pregnancy and maternity across analytical areas shows:

• of the 27 datasets in the CAD return, under 2 in 10 collect robust data on pregnancy and maternity
• of the 18 datasets in the CEAA return, none collect robust data on pregnancy and maternity
• of the 23 datasets in the EAS return, under 1 in 10 collect robust data on pregnancy and maternity
• of the 63 datasets in the HSCA return, 1 in 10 collect robust data on pregnancy and maternity
• of the 14 datasets in the JAS return, none collect robust data on pregnancy and maternity
• of the 11 datasets in the LG return, none collect robust data on pregnancy and maternity
• of the 17 datasets in the NRS return, none collect robust data on pregnancy and maternity
• of the 19 datasets in the OCEA return, none collect robust data on pregnancy and maternity
• of the 5 datasets in the RESAS return, none collect robust data on pregnancy and maternity
• of the 2 datasets in the TS return, none collect robust data on pregnancy and maternity

​​​​​​​Gender reassignment

The RAG ratings for the collection of data on gender reassignment across analytical areas shows:

• of the 27 datasets in the CAD return, less than 1 in 10 collect robust data on gender reassignment
• of the 18 datasets in the CEAA return, just over 1 in 10 collect robust data on gender reassignment
• of the 23 datasets in the EAS return, none collect robust data on gender reassignment
• of the 63 datasets in the HSCA return, under 1 in 10 collect robust data on gender reassignment
• of the 14 datasets in the JAS return, just under 1 in 10 collect robust data on gender reassignment
• of the 11 datasets in the LG return, none collect robust data on gender reassignment
• of the 17 datasets in the NRS return, none collect robust data on gender reassignment
• of the 19 datasets in the OCEA return, none collect robust data on gender reassignment
• of the 5 datasets in the RESAS return, none collect robust data on gender reassignment
• of the 2 datasets in the TS return, none collect robust data on gender reassignment​​​​​​​