1. Providing Access to Census- based Interaction Data in the UK: That’s WICID! John Stillwell School of Geography, University of Leeds Leeds, LS2 9JT, United Kingdom Presentation at the Third International Conference on Computing, Communications and Control Technologies (CCCT '05) Austin, Texas, July 24-27, 2005

2. Presentation UK Census interaction data The WICID system - Building queries in WICID - Output and analysis in WICID Mapping and analysis Conclusions

3. UK Census interaction data Since 2002, Census Interaction Data Service (CIDS) has been providing social science researchers and students in the UK with access to the ‘Origin-Destination Statistics’ collected by the statistical agencies (through decadal censuses). Origin-Destination Statistics, so-called because they involve two geographical areas, comprise: - Special Migration Statistics (SMS): counts of migrants in the 12 months prior to each population census (in 2000-01, 1990-91 and 1980-81) - Special Workplace Statistics (SWS): counts of those commuting to work at the time of the census (in 2001, 1991 and 1981) Sets of tables containing a number of variables or data counts

4. Tables and counts in the 2001 and 1991 interaction data sets for the UK Data setsLevel 1Level 2Level 3 2001 SMS10 tables, 996 counts 5 tables, 96 counts1 table, 12 counts 1991 SMSSet 2: 11 tables, 94 counts Set 1: 2 tables, 12 counts - 2001 SWS7 tables, 936 counts 6 tables, 354 counts1 table, 36 counts 2001 STS7 tables, 1,176 counts 6 tables, 478 counts1 table, 50 counts 1991 SWS*-Set C: 9 tables, 274 counts - * 10% sample

5. Data issues Migration and commuting data sets are very important resource for research and planning since there is no population registration system in the UK and there are no alternative sources that provide data of similar reliability or spatial coverage Large and complex data sets with which the user needs to gain some familiarity before using them with confidence Shortcomings include problems because of the requirement for the census agencies to preserve confidentiality and lower the risk of disclosure In 2001, the approach of ONS to reduce the risk of disclosure was a small cell adjustment method (SCAM), in which values of 1 and 2 were adjusted to either 0 or 3 Another problem is that geographical boundaries change from year to year, thus obscuring the easy comparison of data over time The results of boundary change have prompted the CIDS team to produce estimates of flows from the 1991 and 1981 censuses for 2001 geographical areas

6. WICID Software system used to access and extract data is called Web-based Interface to Census Interaction Data WICID programmed by Oliver Duke-Williams see Stillwell and Duke Williams (2003, 2005) for more details of the system

7. WICID system architecture Server: Apache (http://www.apa che.org)http://www.apa che.org Programming language: PHP: Hypertext Preprocessor (http://www.php. net). DBMS: PostgreSQL (http://www.post gresql.org)http://www.post gresql.org Add-on for spatial data: PostGIS (http://postgis.re fractions.net/)http://postgis.re fractions.net/

8. Building queries in WICID Major objective to create a user-friendly interface so that queries can be constructed and data can be downloaded easily WICID has a guest function that allows anyone to logon and explore the system but those wishing to download data are required: - to obtain an Athens userid and password - to become a registered census user The Census Registration Service facilitates access to the census data resources (http://census.data-archive.ac.uk/)http://census.data-archive.ac.uk/ Currently the system is only available to members of the academic community in the United Kingdom Once registered, access to the CIDS homepage is at http//cids.census.ac.uk /

9. CIDS home page

10. WICID home page

11. Summary of commuting flows to and from the City of London, 1991

12. Age pyramid of commuters from Havering to the City of London, 1991

13. General query interface before selection

14. Map selection window in WICID

15. Example of map selection of City of London as destination area

16. Query to select commuting data to City of London, 2001

17. Screen indicating extraction completed

18. Output and analysis in WICID Output planning screen

19. Analysis indicators available

20. Mapping and analysis: Example How have commuting flows to London boroughs changed between 1991 and 2001? Focus on the changes in flows within London and those arriving from the rest of England and Wales Aggregate statistics from the SWS for 1991 and 2001 suggest an overall increase of 20% in the commuting flows into London boroughs from 2.1 to 2.5 million persons per day Increase in flows from outside London is much smaller (8.6%) than the increase in inflows from other London boroughs (25%); Largest increase in commuting (nearly 60%) during the 1990s was taking place within boroughs Look at variations across London between boroughs

21. London boroughs

22. Percentage shares of commuting inflows to London boroughs, 2001 Inflows from elsewhere in London Inflows from outside London

23. Percentage changes in commuting inflows to London boroughs, 1991-2001 Inflows from elsewhere in London Inflows from outside London

24. Conclusions CIDS has now become established as one of a number of Data Support Units serving social science researchers within the academic community in the UK In 2004, users in 169 academic institutions carried out 3,246 sessions using WICID However, use of the service increased significantly during the last quarter of 2004 and the first quarter of 2005 as the new 2001 interaction data sets have become available. We would expect the increase in usage to continue. Unfortunately the service is not available at the moment to academics or students outside the UK but it is hoped that this situation might be remedied in due course. An adapted version of the system (ACID) has been implemented that allows staff at the University of Queensland to have access to Australian internal migration data

25. Results from 2004 user survey Frequency % Valid % Cumulative Very poor42.4 Poor1.63.0 Average4627.930.9 Good7444.875.8 Very Good4024.2100.0 Total165100.0

26. Conclusions Various facilities in WICID have been developed to facilitate user access to the data before extraction and user analysis of data following extraction: -Map selection tool -Map selection tool constructed in response to feedback from users (particularly students) whose knowledge of the geographies of the UK is limited and who require help in selecting sets of zones -Analysis facilities -Analysis facilities are useful because they add value to the raw counts extracted from the primary and derived data sets though currently only a restricted set of indicators is available Development of WICID over next 12 months will focus on map selection and analysis functions for 2001 data

27. Contact j.c.h.stillwell@leeds.ac.uk Acknowledgement CIDS is funded by the ESRC/JISC under Census Programme Research Grant H507255177 Contact My email: j.c.h.stillwell@leeds.ac.uk Acknowledgement CIDS is funded by the ESRC/JISC under Census Programme Research Grant H507255177 http://cids.census.ac.uk