Table 4 Data validation and development summary per indicator.

From: An inclusive economy dataset for wards in Great Britain using administrative and synthetic data sources

ID

Descriptor

Validation summary (missingness for 2020 unless otherwise stated)**

Years of input data available; years computed/presented + rolling rules for imputation (if any)

1A*

Participation in paid employment

Missingness: 0.59%.

1991–2021 (dvage), 2009–2021 (employ); 2019–2021

2A*

Involuntary exclusion from the labour market

Missingness: 0.59%.

1991–2021 (jbstat); 2019–2021

3A

Wealth inequality

Missingness: 0.59%. Uses LSOA 2011 to ward 2022 map.

1995–2021 (house price data); 2019–2021

4A*

Earnings inequality

Missingness: 0.59%.

1991–2021 (fimnlabgrs_dv); 2019–2021

5A

Poverty

Missingness: 0.26% (City of London and other sparsely inhabited areas). Uses ward 2021 to ward 2022 map.

2015–2022 (relative poverty data); 2019–2021

6A*

Decent pay

Missingness: 0.59%.

1991–2021 (fimnlabgrs_dv), 1991–2021 (jbhrs), 1991–2021 (jbstat); 2019–2021

7A*

Job security

Missingness: 0.59%.

1999–2021 (jbterm1); 2019–2021

1B*

Skills and qualifications

Missingness: 0.59%.

1991–2021 (hiqual_dv); 2019–2021

2B

Digital connectivity

Missingness: 0.59%. Uses LSOA 2011 to ward 2022 map.

2018 only (IUC data); 2018 values carried forward to 2019–2021

3B

Physical connectivity

Missingness: 0.59%. Uses LSOA 2011 to ward 2022 map.

2017 (bus accessibility data for England), 2017 and 2019 (Scotland), 2019 and 2022 (Wales); most recent data for England and Scotland carried forward, Wales data for 2022 rolled backward

4B*

Housing affordability

For house price data only (as earnings data are from UKHLS): Missingness: 0.63% (27.3% before filling). Uses ward 2020 to ward 2022 change dictionary for predecessor filling.

1995–2022 (house price data), 1991–2021 (fimnlabgrs_dv); 2019–2021

5B*

Cost of living

Missingness: 0.59%.

2009–2020 (hheat); 2019–2021

6B

Inclusion in decision-making

Missingness: 1.55% (3.94% before predecessor filling, then 2.88% before filling with manually added data). Wards matched by name, as errors present in ward codes. Uses 2016 to 2022 ward change dictionary for predecessor filling. Also 200 + wards with missing data, some of which found online and added manually. Others missing as councillors were elected unopposed.

2016–2022, except 2020 as no local elections held due to Covid pandemic, but data missing for many areas and years, as local elections not held in every nation every year; 2020 data for England mostly available, but rolled forward from previous years when necessary, 2017 data for Wales and Scotland rolled forward to 2019–2021

  1. Missingness reflects the level of missing information after application of rolling rules, and before the usage of multiple imputation. UKHLS = Understanding Society, EC = Electoral Commission.
  2. *UKHLS variables used partly (Indicator 4B) or wholly for these indicators, i.e. SIPHER Synthetic Population merged with UKHLS.
  3. **Missingness of 0.59% observed for multiple indicators. This is result of 47 of 7,973 wards being excluded as lacking population centroid, or due to disclosivity; see “Methods” section.
Back to article page