Appendix B — Using the 1988 tables in analysis

Every table in this book is held as data in the census repository, in data/reference/survey_1988/. An analysis chapter that compares 2026 with 1988 should read the figure from there, not retype it, and link to the table in this book so that a reader can see it in context.

The files

tables.csv lists the tables, one row each.

Column Meaning
table_id The number printed in the report, such as 9.1. Tables set in the text without a number use a letter: 3.a is the first in Chapter 3
chapter Chapter number
numbered True if the report numbered the table
title Title as printed; for unnumbered tables, a short description
source_page Page of the original report
row_header Heading of the row labels
spanner Heading printed above the value columns
layout classifications where the original set several classifications side by side
note The note printed under the table
chapter_file Page of this book that shows the table

cells.csv holds the figures, one row per printed cell.

Column Meaning
table_id As above
row, col Position in the printed table, counting from 1
row_group Block the row belongs to, such as Region or Settlement; blank if none
row_label Row label as printed
row_type data; base for the counts a table rests on (such as Number of Towers); or total
col_group Group of the column, such as Tower Captains in Table 10.8; blank if none
col_label Column heading as printed
value The cell as printed. * means no data was collected or the combination is impossible; too few means the report suppressed the figure
value_num The value as a number; blank for * and too few

Most figures are percentages, but not all: the table’s note, or its title, says which. Table 4.3 holds averages, Tables 9.3 and 9.5 are rates per bell or per tower, and the base rows are counts.

Reading a table in R

The pure helpers in R/survey1988.R turn the long file into the printed shape and build links into this book.

source(here::here("R", "survey1988.R"))
cells  <- readr::read_csv(here::here("data", "reference", "survey_1988", "cells.csv"))
tables <- readr::read_csv(here::here("data", "reference", "survey_1988", "tables.csv"))

widen_1988(cells, "9.2", numeric = TRUE)   # Table 9.2, one column per printed column
url_1988(tables, "9.2")                     # link to Table 9.2 in this book

For example, Table 9.2 comes back as:

row_group row_label row_type Per cent
Region Southwest data 86
Region Southeast data 83
Region West Midlands data 82
Region East Midlands data 56
Region East Anglia data 68
Region North data 77
Region Rest of B.I. data 88
Settlement Inner City data 63
Settlement Suburban data 92
Settlement Large City data 92
Settlement Town data 92
Settlement Village data 70
Number of Bells 5 data 55
Number of Bells 6 data 86
Number of Bells 8 data 92
Number of Bells 10 data 90
Number of Bells 12 data 100

Linking to a table

Each numbered table has its own address: the chapter’s page, then #tbl- and the number with a hyphen. Table 9.2 is at

https://cccbr.github.io/census2026/survey-1988/09-geography.html#tbl-9-2

Tables set in the text have no number, so link to their chapter instead.

Two cautions for comparisons

  • The 1988 bases are small. Only 300 tower captains and 1,052 ringers replied, and many cells rest on a handful of towers: eight tower captains at twelve-bell towers, perhaps five in inner cities. Check the base rows before comparing a single cell with 2026.
  • The 1988 settlement types were assigned by hand, and the report does not give the rule. A 2026 place classification can be mapped onto the five 1988 types only approximately.