IMD Ward Level Summaries for Sheffield

View source

1 Method

IMD (Indices of Multiple Deprivation) statistics are produced at the LSOA (Lower Super Output Area) level. Sheffield has 345 LSOAs and 28 wards.

To produce ward level summary measures for Sheffield from 2019 IMD we’ve followed the methodology set out in the ward level summaries document on the London Datastore Indices of Deprivation web page (which in turn is derived from the ONS Local Authority level IMD summaries):

“The Average of LSOA Scores measure describes the ward as a whole and is a population weighted average for the ward. This is calculated by first multiplying the IMD score for each LSOA by the population estimate for that LSOA. Where an LSOA is split between wards, a proportion of the population equivalent to the proportion of the LSOA’s addresses falling within that ward is multiplied by the IMD score for that LSOA. These figures are then summed across all LSOAs and part LSOAs within the ward and the result is divided by the total population for that ward (the sum of the LSOA population estimates). This measure takes into account the full range of scores within the ward and the number of people affected by the level of deprivation in each LSOA.”

“The Average of LSOA Ranks measure again takes into account the full range of deprivation across each ward but using the ranks of LSOAs rather than the scores.”

2 Data

IMD data downloaded from MHCLG at:
www.gov.uk/government/statistics/english-indices-of-deprivation-2019

Population and boundary data for Sheffield is via the popsheff and boundsheff R packages.

Residential properties were extracted from the Sheffield LLPG (Local Land Property Gazetteer) on Portal via ArcGIS Pro.

NB: IMD data is 2019. Population and boundaries data are 2020. Residential property data is 2022.

3 Average of LSOA Ranks

Code

# Read residential points
sf_resi <- st_read(
  dsn = "data/Residential.gdb", 
  layer = "ResiAddress", 
  quiet = TRUE
) |> 
  st_cast("POINT") |> # cast from multi-point to point
  st_transform(crs = 4326) # WSG84 coordinates

# Add ward & LSOA variables to residential points
sf_resi <- sf_resi |> 
  st_join(sf_ward, join = st_within) |> 
  filter(!is.na(WD20NM)) |> 
  st_join(sf_lsoa, join = st_within) |> 
  filter(!is.na(LSOA11NM))

# Proportion of residences by LSOA & ward
resi <- sf_resi |> 
  st_drop_geometry() |> 
  count(LSOA11NM, WD20NM, name = "resi") |> 
  add_count(LSOA11NM, wt = resi, name  ="lsoa_resi" ) |> 
  mutate(resi_prop = resi / lsoa_resi)

# Population by LSOA & ward
pop_lsoa <- pop_lsoa_age_band |> 
  filter(`Age band` == "All Ages") |> 
  select(-`Age band`) |> 
  rename(lsoa_pop = Population) |> 
  right_join(resi, by = c("LSOA Name" = "LSOA11NM")) |> 
  mutate(population = round(resi_prop * lsoa_pop)) |> 
  clean_names() |> 
  select(-lsoa_pop, -contains("resi")) |> 
  rename(ward_name = wd20nm)

# Add IMD rank
imd_lsoa <- read_xlsx(
  "data/File_1_-_IMD2019_Index_of_Multiple_Deprivation.xlsx", 
  sheet = "IMD2019"
) |> 
  clean_names() |> 
  filter(local_authority_district_name_2019 == "Sheffield") |> 
  select(
    lsoa_code = lsoa_code_2011,
    imd_rank = index_of_multiple_deprivation_imd_rank
  ) |> 
  right_join(pop_lsoa, by = "lsoa_code") |> 
  relocate(imd_rank, .after = last_col()) |> 
  mutate(imd_rank_pop_multiple = population * imd_rank)

# Sum multiplied figures by ward
imd_ward <- imd_lsoa |>   
  add_count(ward_name, wt = population, name = "ward_pop") |> 
  count(
    ward_name, 
    ward_pop, 
    wt = imd_rank_pop_multiple, 
    name = "imd_rank_pop_multiple"
  ) |> 
  mutate(
    imd_avg_rank = round(imd_rank_pop_multiple / ward_pop, 1), 
    .keep = "unused"
  ) |> 
  mutate(sheff_imd_rank = dense_rank(imd_avg_rank)) |> 
  arrange(sheff_imd_rank)

# Display table
imd_ward |> 
  gt() |> 
  gt_theme_538()

ward_name	imd_avg_rank	sheff_imd_rank
Burngreave	2253.5	1
Firth Park	2349.2	2
Southey	3641.7	3
Manor Castle	3940.1	4
Shiregreen and Brightside	4886.9	5
Darnall	5084.8	6
Park and Arbourthorne	6040.9	7
Gleadless Valley	8345.5	8
Richmond	10028.8	9
Woodhouse	10513.3	10
Beauchief and Greenhill	12331.0	11
City	13285.7	12
Birley	13366.4	13
Nether Edge and Sharrow	13737.7	14
Walkley	14427.4	15
Hillsborough	15745.1	16
East Ecclesfield	16661.4	17
Stocksbridge and Upper Don	16813.6	18
Mosborough	16925.4	19
Beighton	16972.1	20
West Ecclesfield	18729.1	21
Stannington	20632.0	22
Graves Park	21100.6	23
Broomhill and Sharrow Vale	21934.8	24
Dore and Totley	27180.3	25
Crookes and Crosspool	28078.9	26
Fulwood	29822.9	27
Ecclesall	30618.1	28

4 Map

Code

# Display choropleth
sf_ward |> 
  left_join(imd_ward, by = c("WD20NM" = "ward_name")) |> 
  ggplot(aes(fill = sheff_imd_rank)) +
    geom_sf(alpha = 0.8, colour = 'white', size = 0.3) +
    theme_void() +
    theme(
      legend.position = "top",
      legend.title = element_text(size = 9),
      legend.text = element_text(size = 8)
    ) +
    scale_fill_viridis(
      discrete = FALSE,
      name = "Sheffield IMD Average Ward Rank",
      direction = 1,
      guide = guide_colourbar(
        direction = "horizontal",
        barheight = unit(2, units = "mm"),
        barwidth = unit(50, units = "mm"),
        draw.ulim = F,
        title.position = 'top',
        title.hjust = 0.5,
        label.hjust = 0.5
      )
    ) +
    geom_text_repel(
      size = 2.5, 
      force = 1, 
      bg.color = "white", 
      bg.r = 0.15,
      aes(LONG, LAT, label = WD20NM)
    )

5 Export

5.1 Spreadsheet

Code

write_xlsx(imd_ward, "imd-ward-sheff.xlsx")

scc-pi.github.io/imd-ward-sheff/imd-ward-sheff.xlsx

5.2 RDS file

For use with R.

Code

saveRDS(imd_ward, "imd-ward-sheff.rds")

scc-pi.github.io/imd-ward-sheff/imd-ward-sheff.rds

6 Further development

Further development could include “Average of LSOA Scores” and other measures detailed in the ward level summary measures for London.