Using Geographies from Multiple Time Periods

Not only are social science data collected and reported for very many different geographical units, but these geographies are subject to extensive change over time. Changes may range from minor realignment of boundaries, through renaming or recoding of areas, to wholesale redesign. Very minor adjustments may not require any reassignment of data between geographical units but it is sometimes the case that changes between time periods are so great that it is necessary to treat two "generations" of the same geography as if they were entirely separate systems. The existence of a geographical unit with a given name or code at two separate times (e.g. "Portswood ward") should never be treated as automatic evidence that the same geographical area is being referenced. Temporal changes are frequently scale-dependent, for example in the case of electoral wards, where a boundary review may result in ward-level boundaries being redesigned to accommodate population changes, within a local authority boundary which remains unchanged at that time.

The major geographical referencing systems in the UK are revised with widely differing frequencies, driven by the primary purposes for which they were created. Four major geography systems with quite different change processes are outlined in the table:

Electoral and health geographies are especially complex and are subject to frequent and substantial revisions, reflecting the needs of electoral representation and health care organization. These can be particularly difficult to trace due to the fact that revisions often take the form of piecemeal boundary adjustments and mergers of local areas without consistent treatment of area names and codes. Further, periodic major reorganizations of the health care system leads to major discontinuities, for example the discontinuation at various dates of district health authorities (DHAs), regional health authorities (RHAs), family health service authorities (FHSAs), primary care trusts (PCTs) and Strategic Health Authorities (SHAs) in England. Again, there area differences in health geography between the countries of the UK.

Changes in the principal administrative geographies of the UK have been recorded in the ONS Postcode Directory (ONSPD) and preceding postcode directories, which permit a good degree of matching in some cases and include matches to more than one recent health geography. Many new fields were added to these directories in the early 2000s and data for different area codes can be matched using database software such as Microsoft Access. The GeoConvert facility will match data between geography codes using postcode directory data from 2006.

Very little information exists about the detailed relationship between small areas in the UK before the 1980s. Where no lookup table information is available to describe the relationship between two geographies, geographical information systems (GIS) software can be used to intersect two sets of boundaries and create records for the new areas which are created - for example by intersecting ward boundaries from two different time periods. This is not a task that would usually be undertaken by a novice user, and digital boundary data do not always exist in these cases where no lookup tables are available. Further, the intersection of two sets of area boundaries can show which areas overlap but does not provide any direct estimate of the numbers of persons or households contained in each intersecting area, which is usually the issue of interest to the user. The question of how best to allocate information between areas - known as areal interpolation - has been the subject of extensive research, but these methods cannot accurately recreate detailed information which was never collected at the time.