Catalogue description General Register Office and successors: Statistics Division and successors: Primary Births Datasets

Hardware:

• 1963-1964, IBM 1401 with 16000 positions of core storage, six magnetic tape units, a card read/punch unit and an output printer.
• 1974-1975, ICL 1900.

Operating System:

• 1963-1964, A FORTRAN Compiler, with punch card fed programs.
• 1974-1975, GEORGE.

Application Software:

• 1963-1964, Application created specifically for this series of data and was used for subsequent series of data. Fortran and COBOL programming languages were used.
• 1974-1975, In-house application. TAU, a tabulation package.

Logical structure and schema: The Primary Births datasets as supplied to NDAD consisted of fixed column text data.

How data was originally captured and validated: Up to 1962 the bulk of the published vital statistics had been produced (since 1911) through a punched card installation. This installation developed from simple sorters and counters using 36- or 45- column cards to printing-counting-sorters, using 65 column cards. The limitations of conventional punched card methods had begun to make themselves felt in the late fifties. A comprehensive review of the Department's statistical work was carried out in 1959-60 to consider moving to automatic data processing. The General Register Office needed to become automated in order to cope with ad hoc enquiries on data and they needed something which without undue programming effort would be able to process from start to finish smaller jobs for which programming time would otherwise be lengthy compared with processing time. They also required some means of counting to form simple tables and of printing out results legibly. The original machine used for the automation of birth statistics would appear to have been an IBM 1401. Actual operations on this machine began in the middle of 1963 and lasted until 1969.

The data was input by means of punched cards from 1963. Extensive checking took place in order to detect any record which contained incorrect or inconsistent information. In addition the numbers of records were checked against the control figures, records for multiple births were matched with one another and any items of information shown by the registrar as 'not stated' or 'not known' were supplied by the computer as described in the Explanatory note 12 of Part II of the Annual Statistical Review. Those records with impossible or unlikely values were printed out for reference back to the source document and the computer could supply a value in the case of such errors or omissions e.g. if the age of the mother or father or the number of children was not stated, these were assigned a value with similar characteristics to be processed. The number of records dealt with in this way was included in the tables produced.

Records which passed these validation checks were written onto magnetic tape in a code known as binary coded decimal. Any corrections made to impossible or unlikely values were later fed back to the magnetic tape files by means of fresh punched cards. The tape files which were corrected were then processed within the computer to produce two further distinct tape files of primary records arranged in order of (a) area of residence (b) age of mother. These files were then used in two ways: their contents were merged with the records of earlier months ensuring that at the end of the year there would be two complete files in area and age order; and that the figures represented would be added to summary tabulations for each area and each age, so that a year's summary tabulations were accumulated month by month.

The summary tabulations for each area and age were designed so that most of the annual or quarterly tables could be compiled without the need for repeated processing of the large primary files. Where a set of tables refers to a particular category of records, e.g. still births, a sub file of records was extracted from one of the main primary files and used to provide the necessary tables in an economical manner. All the records and summary tabulation files for the year were kept permanently on tape for future reference and tabulation.

Tables for publications could not be produced by the computer printer as it was unsuitable in quality and variety of type face for some publications. The tables therefore were produced on the IBM 870 system, which consisted of a set of electric typewriters attached to a card reading machine. The typewriters printed out the information read from the cards produced by the computer, or punched manually.

Punched cards were again the method of data input used in 1974-75. Validation of the data entered was carried out using dual-entry verification and validation programs.

The Annual Primary Births dataset was produced by the Population Statistics Division of the General Register Office (GRO) and the Office for National Statistics (ONS).

The dataset's main purpose is to record statistical details of births occurring in England and Wales. Formal registration of live births and stillbirths began on 1 July 1837 and 1926 respectively and it was not until the Population (Statistics) Act, 1938 came into operation on 1 July 1938 that the confidential particulars - excluding father's date of birth - were ascertained for statistical purposes. It was also possible to distinguish multiple births on a regular basis as from 1 July 1938. The Population (Statistics) Act, 1960 with effect from 1 January 1961, added the question on father's date of birth in the case of legitimate births and illegitimate births where the father's name is entered in the register. Questions relating to father's and mother's place of birth were introduced on 1 April 1969. Prior to the Census Act of 1920, statistical work carried out was due to administrative action and not to legal requirements. The 1938 and 1960 Population (Statistics) Acts provided for the collection of confidential information at the time of registration.

Birth statistics are compiled annually from the information collected at birth registration for entry into the live-birth and stillbirth registers. Stillbirth registration details are also fed into mortality statistics on childhood gathered by ONS. At the time of registration the informant checks that all the particulars written down by the Registrar are correct and the register is signed by both the informant and the Registrar. In this way all the raw material for detailed statistical analysis of births and stillbirths is drawn together at the registration stage. The confidential information and the information which is open to the public is recorded by the local registrar on a form provided for the purpose. The form is then transmitted to the GRO/OPCS now ONS at the end of each month, at which stage the main processing begins and the forms are used for the compilation of statistics.

The Statistical tabulations are based upon the births registered in a calendar year. Prior to 1974 for the purposes of local administration England and Wales were divided into Administrative Counties and County Boroughs with Administrative Counties being divided further into Municipal Boroughs, Urban Districts and Rural Districts, with the exception of London being divided into the city of London and Metropolitan Boroughs.

Local Government in Britain is structured in two contrasting ways. In Scotland, Wales and parts of England, a single tier,"all-purpose council" is responsible for all local authority functions (Unitary, Metropolitan or London Borough). The remainder of England has a two tier system, in which two separate councils divide responsibilities between district and county councils. Since 1974-5 the organisation of British local government divided England and Wales into counties and districts. For vital statistical purposes the Counties with the County Boroughs therein were grouped into main registration districts and these in turn were divided into sub-districts. In the preparation of statistics for all sub-divisions of England and Wales, births and still-births are assigned to the registration district of the parents' residence, or that of the mother in case of separation of the parents. In the case of people with no fixed residence or with a home outside England and Wales the assignment is to the place of occurrence. Information about the place of birth of the parents of children born in England and Wales has been recorded at birth registration since April 1969.