1841 occupational statistics for counties, selected towns and county residuals.

Table ID:
OCC_1841     (6566297)
1841 occupational statistics for counties, selected towns and county residuals.
Approx. number of rows:
Table type:
Raw Data
Documentation Author:
Humphrey Southall
The data cover the period 1881 to 1911.


  1. The data were taken from Occupational Abstract M.DCCC.XLI. Part I. England and Wales, and Islands in the British Seas and Part II. Scotland, London, 1844. The dataset digitised is actually based on a series of tables, one for each county. The first, for Bedfordshire, is titled 'Occupations of persons enumerated in the County of Bedford, Distinguishing the Borough of Bedford'.
  2. This table was digitised by the Centre for Data Digitisation and Analysis at the Queen's University of Belfast.
  3. The national totals were typed in and added by Humphrey Southall in the spring of 2006, and very extensive data cleaning has also been done by Humphrey Southall.


  1. The 1841 census was the first to gather occupations, and no plans had been made for classifying the results. As a result, this transcription includes over 5,000 different occupational descriptions, some highly specific and others broad aggregates: the list of descriptions used varies between counties. The data are consequently hard to interpret without applying an additional classification. As the same list of occupations is used in all data from a given county, i.e. in both the county total and the data for individual towns, for many combinations of town and occupation there are zero persons; these rows have been deleted from the version of the table held in the database.
  2. While the precise occupations listed are different for each county, the table always ends with seven columns of totals and counts of persons without occupations: 'Total Number of Persons whose Occupations are returned as above', 'Other Persons employed in Trade branch not specified', 'Number of Persons returned as of Independent Means', 'Almspeople, Pensioners, Paupers, and Beggars', 'Other Persons*', 'Residue of Population' and an overall total population. The national totals also include 'Persons travelling on Railways, Canals, &c., on the night of 6th June 1841'. The 'occ_type' column is used to help identify these.
  3. To slightly simplify the table, occupation titles which differ ONLY in punctuation or capitalisation have been standardised to the most common form. However, all variations in spelling have been retained. All occupations listed in the table as loaded here are included in the 'occ_1841_codebook' table.
  4. The urban units for which data were tabulated seem to be the principal towns within each county, defined using whatever administrative areas best approximated to the actual urban area. For example, 'Bristol City with Barton Regis Hundred' and 'Llanbeblig Parish, with the Borough of Carnarvon'. The tables ALWAYS include one town within each county, so some very small towns are included, especially in Wales.
  5. To assist with mapping and other analyses, a special purpose digital boundary data set has been constructed with, broadly, a polygon for each of the urban areas listed in the table, plus a residual area for each county. These areas are also all defined as a specific unit type within our AUO, and we have also included in this data table additional derived values for all the county residual areas, deducting all relevant towns from the county totals. These additional rows have the row type DISTRICT, like the towns, but have data_status set to DERIVED rather than RAW. The following special cases should be noted, which have led to some rows of RAW data being given the row type MISC, because they cannot and should not be linked to our polygon data.
  6. The most complex set of entries was for London, which consists of the following units within the raw data:
    • Kent:
      • Deptford Town
      • Greenwich Town
    • Middlesex:
      • London City within the walls
      • London City without the walls
      • Westminster City
      • Ossulstone Hundred Finsbury Division
      • Ossulstone Hundred Holborn Division
      • Ossulstone Hundred Kensington Division
      • Ossulstone Hundred Tower Division
      • St. Botolph, Bishopsgate Without; Christchurch, Spitalfields; St. Luke's, Old Street; & St. Matthew, Bethnal Green* (simplified to 'London/Spitalfields')
      • Clerkenwell Parish*
      A footnote to the Middlesex table says 'The returns for these parishes [referring to St. Botolph, etc] are included in those for the City of London without the Walls, and in the several divisions of Ossulstone Hundred in which they are situate, as well as being here distinguished; and in like manner, the return for Clerkenwell is also included in that of the Finsbury Division of Ossulstone Hundred.'
    • Surrey:
      • Bermondsey Parish
      • Lambeth Church District
      • Rotherhithe Parish
      • Southwark Borough
      (NB these were all the units distinguished within Surrey, so if they are merged into 'London' there are no Surrey towns left).
    The double-counting of districts within Middlesex was handled as follows. The data for the Spitalfields parishes is divided between other areas in an unknown way, so all the Spitalfields data have been given the row type MISC so they are usually excluded. Conversely, Clerkenwell forms an identifiable part of Finsbury Division, so we include Clerkenwell as one of our Districts, hold as another District the residuals for Finsbury after the Clerkenwell data are deducted (with status DERIVED), and hold the original Finsbury data with row type MISC.
  7. Another special case is Glasgow, where the census report includes data for the City of Glasgow, Glasgow Barony and Glasgow Gorbals, all of which are defined as units in the AUO and have polygons, but also overall totals for 'GLASGOW CITY and SUBURBS', which we have not defined in the AUO and whose simplified name here is simply 'GLASGOW'. This last area is again given the row type MISC.
  8. Finally, the published tables treat the Yorkshire Ridings as counties, treat the Ainsty of York as a fourth Riding with the City of York within it, and do not include overall totals for Yorkshire. We have added overall Yorkshire totals with status DERIVED, given the totals for the Ridings the row type RIDING and then linked them to the AUO, and created county residual districts for each of the three Ridings and also for the Ainsty, excluding the City of York. The original reported data for the Ainsty have been retained but given the row type MISC.
  9. The final result is a table in which contains the following data, with row types:
    • COUNTY: totals for all the Ancient Counties of England and Wales, and the Counties of Scotland (84 units in total).
    • RIDING: data for the three parts of Yorkshire, linked to the relevant Divisons of Ancient Counties in the AUO.
    • DISTRICT: data for selected towns or sometimes parts of towns, plus rural residuals, which together should form a complete but non-overlapping coverage of Great Britain (321 units in total). The loader script checks that each district held here has a matching polygon in the special purpose boundary data set, and vice-versa.
    • MISC: Four specific areas which appear in the original report but should generally be excluded from mapping and analytic work: 'Glasgow and Suburbs', the Ainsty of York, the original data for the Finsbury Division of Middlesex and the data for a set of parishes in the Spitalfields area of London. None of these four have been given g_unit values.


  1. The nation- and county-level data were checked by summing the data for the four age/sex groups and comparing them with the published total for the given occupation. These overall totals do not appear in the published data for individual towns, so this check was not possible for towns.
  2. All data were checked by summing the individual occupational groups and comparing this computed total with the total for 'Total Number of Persons whose Occupations are returned as above' as listed at the end of each table. NB at the time of writing this comment (May 2002) a substantial number of areas failed this test, and more cleaning work would be helpful.
  3. The various 'total' and 'residual' categories at the end of the table were checked by summing all rows with occ_type values 'N', 'R' and 'T', and then comparing this computed total with the total for 'Total Population' (code = 'P') as listed at the very end of each table.
  4. NB while considerable checking has been done on the numerical data, there is another possible source of errors. The list of occupations varies from one county to the next, and these lists were not separately input but copied from previous counties and then modified to fit. These have been proof read and corrected as necessary for England and Wales.


IndexTypeColumn(s) indexed
occ_1841_pk Primary key district_name, row_type, occupation
occ_1841_idx_occ   occupation, district_name, row_type


The table has the following associated constraints:

occ_1841_pk Primary Key See details above for primary key index

Columns within table:

row_type Text string (max.len.=12). Flag indicating, broadly, the geographical level. See notes for how it is used but the main values are NATIONAL, COUNTY and DISTRICT.
row_status Text string (max.len.=12). RAW for all data directly computerised from source, and DERIVED for the additional data for county residuals and the data specially inserted as Yorkshire totals and for the Finsbury area deducting Clerkenwell.
district_name Text string (max.len.=204). The name of the area to which the data relate, as stated in the original report. This may be a county or an urban area (see notes).
simpler_name Text string (max.len.=44). A simpler version of the district name, added by us. For example, 'Bristol City with Barton Regis Hundred' becomes simply 'Bristol'.
container Text string (max.len.=104). The name of the higher-level area, usually a county, to which the data relate. These are all either Ancient Counties or Scottish Counties. This column is blank for data covering county and national totals.
container_agr Text string (max.len.=44). For England and Wales this matches container, but it groups Scottish counties into regions. This was used to support certain analyses, notably the maps in the Gilbert and Southall chapter in the Cambridge Urban History of Britain
nation Text string (max.len.=6). Indicates the nation containing the county: 'E'=England, 'S'=Scotland, 'W'=Wales. Empty for the national totals for 'England and Wales' and 'Great Britain'.
occupation Text string (max.len.=84). The occupational group as stated in the census table (see note).
occ_type Text string (max.len.=6). A set of 3 codes to signify 3 different types of data.: T indicates 'Total Number of Persons whose Occupations are returned as above', which NB excludes the residual category. R indicates numbers in residual categories such as 'Other persons employed in trade'. N indicates categories out of employment, e.g. 'Independent persons', 'Almspeople'. P indicates 'Total Population'. In general, rows with non-null values here should be ignored when analysing occupations.
m_occ_20 Integer number. Number of males 20 years and upwards occupied in the specified occupation.
m_occ_19 Integer number. Number of males under 20 years of age occupied in the specified occupation.
f_occ_20 Integer number. Number of females 20 years and upwards occupied in the specified occupation.
f_occ_19 Integer number. Number of females under 20 years of age occupied in the specified occupation.
pop_occ Integer number. Total occupied population for the specified occupation in each county. NB this is raw data for county totals, but is simply computed from the four age/sex categories for towns.
g_unit Integer number. ID as defined in the AUO for the unit the row of data is for. These include Nations, Ancient Counties, Scottish Counties, Divisions of Ancient Counties (for the Yorkshire Ridings), and the special purpose 1841 Occupation reporting areas which link to specially created digital boundary data.
notes Text string (max.len.=104). Notes.
rec_num Integer number. Sequence number added on loading to keep rows in order.