Browse The National Archives' catalogue

Browse by reference Browse home

Inside you will find

  • MAF 4081989-2001Ministry of Agriculture, Fisheries and Food: Statistics (Census and Prices) Division: Agricultural and Horticultural Census Datasets

    The Agricultural and Horticultural Census is one of the main data gathering exercises for agriculture and horticulture in England and Wales. It is a survey of main agricultural holdings in both countries, covering about 99% of the total agricultural area, and is conducted annually on the first working day of June. It provides the Ministry of Agriculture, Fisheries and Food (MAFF) and the National Assembly for Wales with 'the basic physical statistics of farming: areas of land use and crops; numbers of main livestock types; and the numbers and kinds of person working on the farm summarised by parish. To 1994, the base data for the summaries was collected for all agricultural holdings. From 1995, only the larger holdings submitted annual returns, and smaller holdings were sampled on a three-yearly basis.

    Broadly speaking, the 1989-1999 censuses covered the following areas:

    • The extent and type of crops and fallow, land use and land tenure, and the size of holdings.
    • The agricultural labour force: numbers and types of farmers, managers and workers (full-time and part-time, male and female).
    • Livestock: numbers of cattle, pigs, sheep, goats, poultry and other types of farm animals (e.g. horses, farmed deer).
    • Horticulture: extent of vegetables, bulbs, flowers and fruit grown in the open, orchards, nursery stock, glasshouses and covered structures, and glasshouse crops.
    • Changes in the area of a holding (land given up or taken over), and changes in a farmer's name and address.

    Most of the data collected by the 1989-1999 Agricultural and Horticultural Censuses related to the situation of farmers on the census date, the first weekday in June each year. The census was a snapshot of holdings at a time of prime activity rather than a picture of holdings over an entire year. Consequently, it was complemented by surveys taken at other times, although these were not as comprehensive in their coverage.

    These datasets fall into three main sub-series:

    (1) Parish Summaries, in which the data was aggregated by MAFF, before transfer, to the level of the agricultural parish (on agricultural parishes, see Counties and Parishes lookup tables, below). In each dataset, the data drawn from the census is held in a single table in which there are two types of field: those which record the total of the responses in the parish to a question asked in the census, and those which record the number of holdings which answered the question. While a correlation can be drawn between most of the census questions and the fields in the datasets, this is not always the case:

    • Some questions which appear in the census forms do not have corresponding fields in the datasets, because the data was not held in MAFF's system, or because a statistical decision was taken not to use the data.
    • Where the census form required respondents to provide an answer other than a numeric figure (e.g.'Yes/No'), MAFF has not been able to aggregate the data in a meaningful way. For example, question 135 in the 1989 census asked respondents to answer 'Yes' or 'No' to the question 'Do you intend to keep any turkeys on your holding in the next 12 months'. Rather than there being separate fields for the numbers in each parish answering 'Yes', 'No', or not responding, the values representing 'Yes' and 'No' have been summed within a single field to produce a single numeric total (i.e. fields like this have been aggregated in the same way as fields relating to questions where farmers are asked to provide a numeric figure). Some of the fields which are affected by this problem have been excluded by MAFF from the datasets transferred.

    (2) Counties and Parishes lookup tables. The Parish Summaries datasets include codes for agricultural parishes, in which the first one or two digits identify the county, and the remaining digits identify the MAFF/Welsh Office 'parish'. In England, MAFF's 'parishes' generally correspond to the lowest level of local authority, which typically is a parish council, although it may also be a larger administrative area where there are no parishes. The 'parishes' used by the Statistical Directorate of the National Assembly for Wales (and its predecessors in the Welsh Office) relate to units which predate community councils (the equivalent in Wales of English parish councils). In some cases, the names of MAFF 'parishes' are not identical to the names of modern local authority areas; where the names are the same, MAFF has advised that the MAFF 'parish' and the civil parish may not cover the same areas.

    Lookup tables were transferred, by MAFF and by the Welsh Office, which translate the codes used for parishes in the Parish Summaries datasets. MAFF also supplied a table which translates the codes used for counties. As these tables do not relate to any particular sweep of the census, they have been treated as distinct datasets rather than as part of the Parish Summaries datasets.

    (3) A series of datasets of County Summaries held in MAF 410.

    The datasets in this series are available to download. Links to individual datasets can be found at piece level.