County-level Statistics from the Agricultural Census

Table ID:

AGCEN_CNTY     (1246384)

Contents:

County-level Statistics from the Agricultural Census

Approx. number of rows:

331,536

Table type:

Raw Data

Documentation Author:

Humphrey Southall

Chronology:

The data cover the period 1866 to 1971.
Dates and times are identified by:
   Year

Sources:

1. The data are all drawn from the published reports of the Agricultural Census. The table numbers and titles vary over time. Usually the national totals for England, Wales, Scotland, Great Britain, Ireland, and the United Kingdom are in a separate table which comes before the main county table. After 1911 Scotland figures are given in a separate report to England and Wales. Totals for the Isle of Man, Jersey and Guernsey are sometimes given in the national totals table, sometimes at the end of the counties table or sometimes in a separate table. For example, in 1877 the tables are in the report entitled 'The agricultural returns of Great Britain with abstract returns for the United Kingdom 1877 (HMSO: London, 1877)'.
   The figures come from the following tables:
   • Table 1 'Total area and acreage under each kind of crop, bare fallow, and grass; and number of horses, cattle, sheep, and pigs, as returned upon the 4th June 1877, and 25th June in 1876, in each division of Great Britain, with similar particulars for Ireland, and with Total for United Kingdom' pp14-15,
   • Table 2 'Total area and acreage under each kind of crop, bare fallow, and grass; and number of horses, cattle, sheep, and pigs, as returned upon the 4th June 1877, and 25th June in 1876, in each county of England/Wales/Scotland/in the Isle of Man, and in the Channel Islands' pp16-45,
   • Table 3 'Acreage under orchards, market gardens, nursey grounds, and woods, in each county in England, Wales, and Scotland, in the year 1877' pp46-47.
   In 1910 the tables are in the report entitled 'Agricultural statistics, 1910. Volume XLV. Part I. Acreage and live stock returns of Great Britain with summaries for the United Kingdom (HMSO: London, 1911)'.
   The figures came from the following tables:
   • Table 2 'Acreage under crops and grass; and number of horses, cattle, sheep, and pigs, as returned upon the 4th June 1910 and 1909, in England, Wales, and Scotland' p27
   • Table 4 'Acreage under crops and grass; and number of horses, cattle, sheep, and pigs, as returned upon the 4th June 1910 and 1909, in each county of Great Britain' pp32-60,
   • Table 5 'Extent of land under crops and grass occupied by tenants and owners respectively, in 1910, in each county of Great Britain' p61,
   • Table 6 'Area of land and water, mountain and heath land using for grazing, permanent grass, and arable land, in 1910, and woods and plantations in 1905, in each county of Great Britain' pp62-63,
   • Table 7 'Acreage under clover, sainfoin and grasses under rotation, and under permanent grass, as returned upon the 4th June 1910, in each county of Great Britain, distinguishing that reserved for hay' pp64-65,
   • Table 8 'Acreage under carrots, onions, buckwheat, flax, and other crops (not separately distinguished in Tables 1-4); as returned on the 4th June 1910, in each county of Great Britain' pp66-67,
   • Table 9 'Acreage under different kinds of small fruit, as returned on the 4th June 1910, in each county of Great Britain' pp68-69,
   • Table 10 'Acreage under orchards, distinguishing the kind of fruit, and the acreage also accounted for under small fruit and permanent grass, as returned on the 4th June 1910, in each county of Great Britain' pp70-71,
   • Table 12 'Number of Agricultural holdings in each of the under-mentioned classes, distinguishing 'owned or mainly owned' and 'rented or mainly rented' as returned on the 4th June 1910, in each administrative county of Great Britain' pp74-75,
   • Table 15 'Acreage under crops and grass; and number of live stock, as returned upon the 4th June 1910 and 1909, in Great Britain, Isle of Man, channel Islands, Ireland and the United Kingdom' pp80-81.
2. This table combines data transcribed from those reports by several different agencies or projects, each following different strategies. The 'Transcriber' column identifies who did the transcription.
3. Transcriber='DEFRA': These are complete transcriptions of the data for English counties from the reports for years ending in zero, from 1900 to 1980. The 1980 data are not currently included in this table because we have been unable to relate them to the published reports and have therefore been unable to check the data or include the original report labels.
4. Transcriber='DGATLEY': These are fairly complete transcriptions of the reports for 1871 and 1901.
5. Transcriber='RSCHWARTZ': These are complete transcriptions of the data for English and Welsh counties from the reports for 1869 and 1885, created by Robert Schwartz of Mount Holyoke College. He also created much more limited transcriptions for 1875, 1895 and 1915, but these have not been used.
6. Transcriber='GBHGIS': The transcriptions described above have been extensively supplemented by the GB Historical GIS Project, to fill gaps in the sequence of 19th century reports, to include more data for Census years to link to the Census of Population, and to provided some coverage of Scotland and Ireland. Limited data for 1921, 1941, 1951 and 1971 were transcribed by Paula Aucott in March and June 2006 and include just total acreage, and the areas of arable, permanent grass and rough grazings. Much more complete transcriptions for 1931 and 1961 were made by Ben Jakobek in January 2008, which together with David Gatley's transcriptions form a 30-year sequence 1871-1901-1931-1961. Scottish data were added for those years, and also for 1870, 1877, 1884 and 1885. Harold Price added the missing data to make complete transcriptions for England, Wales and Scotland for all revelant rows and columns (including national totals) for 1869-71, 1877, 1884-85, 1891, 1900-01 and 1910 in Winter 2021. He added new values for Great Britain in 1866-68, 1872-76, 1878-83, 1886-90, 1892-99, 1902-09, 1911-20 in Spring 2022. The GBHGIS transcriber ID also appears for certain rows within transcriptions made by others, where we added columns needed to compute checksums.
7. The dataset currently holds full transcriptions of data for England, Wales and Scotland for the following years: 1866-1920, 1931, 1961.
   These years include as many recurrent figures as were possible to identify in the tables. Much more limited data in included for England (1930, 1940, 1950, 1960, 1970) and Ireland (1871). However, these other dates have not been re-assessed to determine if any more columns would be of interest for inclusion in the final dataset. Data entry for this table is still in progress and more will be added as it becomes available.
8. Irish data are currently held for 1871 only here. For more Irish data see the separate table for Ireland.

---

Notes:

1. Most of the data for England and Wales in the twentieth century cover Administrative Counties as defined at the relevant date, but there are a number of exceptions:
   In 1931-61 inclusive, London is combined with Middlesex;
   In 1971, Greater London is covered by two separate units, 'Greater London (Eastern)' and 'Greater London (South Eastern)', which may mean that the GLC area was divided between two regions of England;
   In 1961 and 1971 'Suffolk' was listed as a single unit;
   In 1941-71, the Scilly Isles are listed separately from Cornwall.
2. The table includes totals for the Isle of Man, Jersey and Guernsey in all completed years before 1921. At times these units are combined. The table includes national totals for England, Wales, England and Wales, Scotland, Great Britain, Ireland and the United Kingdom as far as possible before 1921. In some cases the row values are slightly different as they combine variables listed individually in the county data.
3. This table is currently a work in progress, with substantial enhancements to the original transcriptions. In particular, the data have been checked back against the original tables and re-arranged to place variables in the order they were originally listed. All 'report_labels' for crops, animals etc are those used in the original reports. The raw data has an additional temporary 'transcribe_label' column. For data digitised elsewhere this gives the transcribed label. For GBHGIS digitised data this column includes additional information to help distinguish columns. For example, where the columns are just labelled 'Total' in the original report for different tables in the same year these column labels have had additional information added to distinguish between the meaning of the different Total values. This helps identify content for the cellref column. This temporary column is removed before the final table is built.
4. The various units have been systematically linked to the AUO. The following points should be noted:
   • The rows for the 'CROFTING COUNTIES' in the Scottish reports lack g_unit values.
   • The Isle of Man, Jersey and Guernsey are indentified separately in most instances. However, their total population figures are combined in 1868 ( for Jersey and Guernsey) and in 1869, 70 & 72 (for all three). These rows lack g_unit values.
   None of these combined units currently exist in the AUO.
   • The reports for England and Wales include data for the Scilly Isles. Where these islands are reported separately they have been given the unit ID for Cornwall, which this means that data need to be grouped by g_unit.
   • The 19th century reports for England and Wales are broadly by Ancient County but include data for the three Ridings of Yorkshire. These have been given g_unit values as for the Divisions of the Ancient County, and we have added derived values for Yorkshire as a whole, so that the data cover all the Ancient Counties.
5. As yet most rows do not have a "CellRef" value as many additional categories have been added to this table and the DDS has not yet been updated.

---

Checking:

1. The geographical units have been cross-checked against the GBHGIS administrative unit gazetteer to ensure that all units are correctly identified. Each unit has a unique unit ID assigned to it and this can be used to cross-reference against other datasets.
2. The data for individual counties for each variable were checked against the overall totals for England, for Wales, for Scotland, and for Ireland. Wherever possible, we also checked numbers for individual counties against available checksums, such as comparing the sum of the acreages of individual crops against the total arable area. In a small number of cases this included correcting printing errors in the original reports by replacing values that led to checksum errors with the number that met both row and column checksums.
3. In some cases a combined total is given rather than individual values. Like the total population example given for Jersey, Guernsey and the Isle of Man above. While 'Cabbage, Kohl.-Rabi, and Rape' are sometimes combined for national totals, but given individually for counties.
4. Numerical errors identified but where a note in the original report provides explanation:
   • 1877 - the UK total figure for 'Total acreage under crops, bare fallow & grass', 'Green crops: Cabbage, Kohl.-Rabi, and Rape' and 'Total acreage under Green Crops' are all out by 221. A note on page 63 states 'These are corrected figures received from Jersey after the Returns were completed, and the 221 acres previously entered in excess are still included in the figures for the United Kingdom in Table I. and elsewhere, it being too late to alter them'.
   • 1891 - the England total for 'Total acreage under crops, bare fallow & grass' and 'Hops' are both out by 3 acres. A note on page 2 states 'Since the returns were completed. information has been received to the effect that 3 acres of Hops (included in the above figures) were returned for the county of Cambridge in error'.
     - the UK check on the 'Extent of land occupied: total acreage of land occupied' is out by 11,669. This is caused by Guernsey which has a total figure but no breakdown. A note on page 59 states 'not separately distinguished'.
   • 1901 - the UK check on 'Total area of land and water' is out by 128,717. This is caused by Jersey which has a total figure but no breakdown. A note on page 35 states 'not separately distinguished'.
5. Numerical errors where the numbers have been double-checked against the original but remain unresolved:
   • 1869 - Total area for Scotland national total is 30 more than the summed total.
   • 1870 - Total area for Scotland national total is 30 more than the summed total.
     - Total area for England national total is 90 more than the summed total.
     - Total population for Scotland 17,356 more than the summed total.
     - Total population for Great Britain and the United Kingdom both 220,556 more than the summed total.

---

Columns within table:

Column	Type	Contents
county	Text string (max.len.=64).	The name of the county, or nation (see note above).
transcriber	Text string (max.len.=20).	Value identifying who did the transcription (see notes above).
year	Integer number.	Year in which the data were gathered. The farm census has always been carried out in June.
report_table	Text string (max.len.=24).	Identifier for the table within the report. This is a string, not a number, such as 'Table 2'. Where the information comes from mulitple tables, e.g. where the national totals are in a different table the table number is given for the table containing the county data as far as possible.
col_or_row	Integer number.	Column or row occupied by the value in the original table, i.e. the location of the variable in the original report. NB Some tables used different columns for each county, in which case this DB field will hold a row number. Many tables also included values for the previous year, but these have been ignored when computing these column/row numbers. (NB the reason why Scottish data for 1870 and 1884 are included in this table is that they were listed in the tables for the following year and included in the OCR output.)
report_label	Text string (max.len.=154).	The label for the variable as listed in the original report. NB this is not the version supplied by the original transcriber, but text added by the GBHGIS after careful checking against the original reports.
meas_unit	Text string (max.len.=16).	The units in which the variable is measured. For now, the only value is 'Acres'.
value	Floating point number.	The numerical value of the variable for the relevant county and year.
row_type	Text string (max.len.=12).	Abbreviation indicating type of unit.
g_unit	Integer number.	ID number for reporting unit, as defined in the AUO.
g_unit_type	Text string (max.len.=28).	Type of the unit, as defined in the AUO.
cellref	Text string (max.len.=52).	Cell reference identifying what the particular data value measures, as used in the GBH GIS Data Documentation system. These are set using the agcen_variables table and based ona standardised version of report_label, but also reflect the particular year the label appears in.
g_data_status	Text string (max.len.=6).	Used as in the main data table: everything loaded in from the transcriptions is labelled as 'P' for public, while the additional values computed by the script which builds the table have status 'D' for derived.
rec_num	Integer number.	Sequence number added as the data are loaded, to keep them in order.