Consultation regarding the redraw of Data Zones

10. Methodology for Producing Draft 2011 Data Zones

10.1. Following a consultation in August 2010, the decision was made that Data Zones would be redrawn using 2011 Census Output Areas. The aim of the redraw process was to maintain the link with current Census data, to re-align to Local Authority boundaries, and to account for changes in population while keeping boundaries relatively comparable to those created in 2001. The main criteria (constraints) used to redraw Data Zones for 2011, in approximate order of importance (descending), were:

• That they be built up from 2011 Census Output Areas;
• Maintain approximately equal populations of 500 to 1,000 people, with an absolute minimum of 375 people and a maximum population of 1,125;
• Be a continuous area without multiple extents, unless it contains islands;
• Where possible, changes are made within Intermediate Zone boundaries;
• Maintain a compactness of shape.

10.2. The production of the draft 2011 Data Zones was largely an automated process. (If you require a detailed methodology, please contact us.)

10.3. At the outset, a first cut of boundaries was created by aggregating 2011 Census Output Areas to 2001 Data Zones, based on the location of the COAs population weighted centroid. The resident population was summed for each 'best-fit' Data Zone to assess whether it fell within the required population thresholds.

10.4. Data Zones that have seen a dramatic decrease in population (e.g. as the result of a demolition) were merged with a neighbouring Data Zone (this effected 23 Data Zones). Conversely, Data Zones that have grown significantly in population (e.g. as the result of a new housing development in the area) were split into one or more new Data Zones, this has effected 364 Data Zones. The population thresholds to which either a merge or split was required are 375 and 1,125 respectively.

10.5. Where a merge was required, a Data Zone with a population of less than 375 was merged to a neighbouring Data Zone with the longest shared boundary, maintaining the population criteria where possible. This effected 23 data zones.

10.6. For Data Zones that required splitting (364); Census Output Areas were grouped into 'sub' Data Zones that meet the population requirements. Starting with the most south westerly, Census Output Areas were appended on to create a new Data Zone until the resident population totalled 750 or more. At the point, a new 'sub' Data Zone was created, and the remaining Census Output Areas appended on until the population threshold was reached. The number of new Data Zones that were created from a split Data Zone was calculated by dividing the resident population by 750 and rounding to the nearest whole number.

Fig 1.5: Examples of Data Zones that have been split or merged

10.7. Figure 1.5 shows examples of Data Zones that have been split or merged. Building up from 2011 Census Output Areas, Data Zone S01003303 (left) has a resident population of 2,006 - well above the desired population range of 500-1,000. In the draft set of 2011 Data Zones produced, this area has been split into 3 new Data Zones (labelled DZ460300, DZ460301, and DZ460302).

10.8. Data Zone S01005071 (Figure 1.5, right) currently has a resident population of 190, which is well below the minimum desired population. To correct this in the draft set of 2011 Data Zones, S01005071 has been merged with S01005075 (labelled DZ240110).

10.9. In some cases, where the Census Output Area geography has significantly changed, multi-extent Data Zones were created after aggregating Census Output Areas to 2001 Data Zones. Where this occurred, some additional editing was required to correct this. An example of this is shown in Figure 1.6, where 2001 Data Zone S01000163 is cut across by the 2011 Census Output Area, creating two parts of the Data Zone that do not touch. In this case, Census Output Area S00090409 was removed from Data Zone S01000163 and appended to 2011 Data Zone DZ330151. Changes made as a result of multi-extents affected 319 Data Zones.

Fig 1.6: Example of a Multi-extent Data Zone created by building 2011 Census Output Areas to 2001 Data Zones on a best-fit basis

10.10. The redraw process made changes (i.e. a split, merge, or correction of a multi-extent data zone) to 706 of the original 6,505 Data Zones. The remaining 5,799 have been rebuilt using the 2011 Census Output Areas on a best-fit basis. The total number of 2011 Data Zones is 6,940.

10.11. Figure 1.7 shows a high level summary of methodology described above.

Figure 1.7: High level methodology for draft 2011 Data Zones