Summary
The Census Coverage Survey (CCS) was the largest household survey ever run in the United Kingdom. A project of this scale required some careful planning well in advance and continuous monitoring and assessment as it progressed. Overall the survey can be considered a major success, with all occurring challenges overcome. Inevitably there are some lessons to be learned out of these but none of them were prohibitive in ensuring the CCS delivered its aim sufficiently for the One Number Census methodology to be applied. There is nothing significant that should be done differently were the exercise to be repeated. In particular, the team working, field staff structure, communications and training all worked as they should on the whole. Further information on the wider impact of the CCS will be provided in a high level summary which will examine cross-project issues and interfaces. The summary will aim to address the value of the data collected in the context of the One Number Census and the balance between the timing of the CCS and Census follow-up procedures.
Introduction
This report provides a high level overview of how well the practical aspects of the Census Coverage Survey (CCS) have gone. It is written from the perspective of the Survey Director, and draws on information obtained from a fairly wide set of sources. The report was written in isolation from other evaluation reports and as such the content should not be taken out of context.
Many aspects of the CCS were relatively new - certainly when used in conjunction with each other on a survey of the size of the CCS. The purpose of this report is to highlight procedures that are considered to have worked well in running the survey, to make note of those that didn't work so well, and to document the lessons learned from the experience.
Scope
The report is limited in scope to comment on the practical aspects of the survey. It makes no attempt to evaluate the statistical impact of:
How well the survey picked up people missing from the Census - this information is not yet available;
How well the One Number Census procedures have worked.
There is very little in the way of quantitative assessment. Full analyses of response rates and other measures can only be done once the CCS and Census processing is complete. As a result the report concentrates mainly on the qualitative aspects.
The report includes information only from the experiences of the operation in England and Wales. Evaluation of Scotland and Northern Ireland is taking place separately.
Format
Each key area of the CCS project is covered in a subsection of the report. The subsections give:
a brief overview of the area as background,
a list of issues that were effective within the workpackage,
a list of issues that did not work as planned, and
a description of the lessons learned from the experience. The lessons learned concentrate on issues that could be done differently were the CCS to be run again.
The key lessons learned are summarised at the end of the report.
Sources of Information
The following sources of information have been drawn on in the preparation of this document.
Debrief questionnaires.All field staff employed on the survey were given a debrief questionnaire for self-completion at the end of the project, asking for their views on how well different parts of the survey had worked.
Training evaluation questionnaires. These were presented to attendees of the field staff training events and gathered feedback on how well the training had been received.
Management information. This was provided by the Team Reporting And Communication System (TRACS) used throughout the survey period. Field and Team Managers recorded a good deal of management information on this system to enable Headquarters staff (and themselves) to monitor the performance of the survey and to respond accordingly. The TRACS system was also used in encountered.
Advice from Workpackage Managers. The CCS project was broken down into smaller workpackages, covering for example recruitment, training or field procedures. Each of these workpackages was managed by a member of the Headquarters Team who was in the best position to judge what worked and what didn't. Subsequently they were responsible for evaluating their area of expertise and this has been used in compiling this overview document.
Correspondence. Survey Headquarters received a good deal of correspondence throughout the project, from members of the public who had been approached to take part in the interview, from field staff themselves, from telephone calls and messages to the public helpline. This provided valuable insight into how well the survey was received.
Aims and Design
The CCS was designed within the One Number Census methodology with the sole purpose of enabling ONS to identify those people and households missed by the Census. To this end, it had to be entirely independent from the Census, there had to be full geographical coverage, it had to allow population counts to be adjusted for underenumeration down to small area level and it had to allow population estimates to be made for all districts.
The sheer number of Local Authority districts meant that to apply the CCS design to each one would lead to a prohibitively large sample size. Therefore local authorities were grouped together to form "design groups" which formed a broad regional stratification for the UK and had a population of about half a million each. The design was applied to each of these and they were constructed to enable estimates of local authority district underenumeration to be made indirectly.
Information from the 1991 Census was used as a proxy to select 1991 Enumeration Districts (EDs) as a first stage selection. The EDs were stratified according to a "hard to count" index - an indication constructed from Census variables known to be associated with underenumeration. Within each HtC category the EDs were clustered into relatively homogenous groups based on their age-sex distribution. Five postcodes were then chosen at random and the aim of the CCS was to fully enumerate each of these selected postcodes.
Overall, the sample size amounted to around 16,000 postcodes - a number assessed to provide an acceptable level of accuracy.
The postcode based selection with full enumeration of these made workload planning and monitoring relatively straightforward.
The overall multi-stage design kept the sample size manageable (and affordable) while still allowing the estimation of underenumeration to small areas with full geographical coverage.
Difficulties overcome:
At the stage of selection it became apparent that the average number of households in each postcode was higher than previously estimated. Therefore the overall sample size, in terms on the number of households, was larger than previously planned.
The random selection of postcodes meant that very difficult places could occur in the sample. Difficulties included the distance required to travel, logistical difficulties (e.g. Isles of Scilly) and potentially dangerous areas. Each needed careful local management at the time of the survey.
Lessons learned
The sample needs to have a high degree of contingency built in in terms of the number of households expected to be selected.
Field Management Structure
Over 4,000 members of the public were employed to work on the CCS within a hierarchical staff structure consisting of three levels. 3,900 Survey Interviewers worked in teams to 270 Team Managers. Each Team Manager reported to 34 Field Managers who were in direct contact with 5 Regional Managers. The three-tiered management system meant that the enormous task of overseeing a survey of 320,000 households could be broken down into manageable units.
The essential structure of Survey Interviewers and Team Managers were tested in the 1999 CCS Rehearsal, however, Field Managers were added as part of the scaling-up of the exercise in 2001.
Highly effective:
The hierarchical structure enabled the survey to be closely managed at all times for all areas of the country despite the Headquarters Team containing only five Regional Managers.
The structure led to an excellent team spirit and a degree of friendly competition across teams building up. This was proved time and time again to be highly effective in the smooth running of the survey and provided many benefits.
Difficulties overcome:
In some areas it was difficult to get the message across that the Field Manager's role was distinct from the Team Manager's role. Avoidance of perceived interfering was required.
Different team sizes, particularly the number of interviewers reporting to a Team Manager, caused a small amount of friction.
Staff from Headquarters had dual roles as Regional Managers and Workpackage Managers. There were cases where the workpackage was not completed before they had to take up the regional manager duties and this impacted on both sides.
Lessons learned
The staffing structure used was very effective and it is difficult to see how it could be improved upon. Removing the Field Manager layer and only having Team Managers, as was the case in the 1999 Rehearsal, would not have been practicable for a survey of this size and to ensure full geographical coverage. Overall the ratio of Team Managers to Field Managers and the ratio of Interviewers to Team Managers was about right - any more and recruitment and organisational activities would have added significantly to their individual workloads. However, Regional Managers should not be Workpackage Managers as there is insufficient time to perform both functions.
Team Reporting and Communication System (TRACS)
Rapid communication to and from the field were essential so that problems encountered could be acted on as soon as possible. A communication system based on Web technology was commissioned, and all the Field Managers and Team Managers were loaned a PC for connection to this system. The system allowed three way communication with field staff and provided an administrative tool for the staff to keep records of recruitment progress, to print additional stationery and to monitor survey performance.
Highly effective:
The system facilitated good communications between survey managers.
It enabled Headquarters to pass additional stationery (eg letters about Foot and Mouth) into the field quickly, as well as new sets of instructions.
It meant that instantaneous announcements could be made to all managers without relying on a chain of managers.
It supplied all the necessary management information required to steer the survey as it ran live, and also provided a post-survey record of all issues arising.
It was independent of the Census reporting system, which would otherwise have been overloaded.
Difficulties overcome:
From the start field staff expectations of the system were too high.
Development and specification could have started earlier and timetables could have been adhered to more strictly. Due to the time taken to develop and sign off the system it was delivered in a piecemeal fashion for the field staff to use.
The final cost was more than originally planned due to necessary upgrading of equipment to ensure it could be redeployed at the end of the project.
ONS maintained ownership of all the computers in the field. We were therefore running a risk of loss or damage to property, and associated Health and Safety issues.
Lessons learned
Development of an IT system such as this needs to start very early and timetables should be adhered to. The balance between system simplicity and functionality needs to be carefully considered - the requirement for a system that performs all of the functions TRACS did (particularly those early on such as recording sift and interview scores for recruitment) may be offset by the need to get the system fully tested and released in advance of its use. Delivery in a piecemeal way, whilst ensuring parts are properly tested before release, has a negative effect on the way the system is accepted by the field staff.
Emergencies and Changed Procedures
The Foot and Mouth outbreak in the UK was an unforeseen problem that coincided with the timing of the CCS. Emergency procedures had to be compiled and disseminated to field staff quickly to ensure the adverse impact of the disease was kept to a minimum. Additionally a General Election was announced to take place within the survey period. Allowances for this had to be made and relevant instructions issued to field staff.
Highly effective:
At all times Headquarters staff were able to devise alternative interview plans within a very short timeframe and to disseminate them into the field. The reactive capability of the Headquarters team was excellent.
The field staff were highly receptive to the necessary alterations to survey procedures. They provided useful feedback on the state of the situation at all times.
Difficulties overcome:
No pre-planning for contingencies was made for emergency procedures, though by their nature this is very difficult.
Lessons learned
Some allowance must be made in the resource allocation of the survey project for the offchance that something unexpected will occur during the survey period. This includes consideration of funding as well as staff time available. A degree of contingency time for Headquarters staff should be built into the survey plans.
Recruitment
Recruitment involved the sifting and interviewing of over 4,000 staff. This was a huge exercise requiring many weeks of work, initially for Headquarters staff and subsequently for Field Managers and Team Managers. The recruitment adhered to open and fair competition criteria. It was done in a cascading way with Headquarters staff recruiting Field Managers, Headquarters staff and Field Managers recruiting Team Managers, and Team Managers recruiting interviewers.
Highly effective:
The Headquarters team prepared for this phase of work well in advance. There was a good deal of involvement from Human Resources Division (HRD), training the CCS team in how to conduct sifts fairly and how to hold evidence-gathering interviews with candidates. This preparation continued until the first phase of the Field Manager interviewing and as a result all team members felt confident and capable enough to hold the interviews.
Having the CCS team recruit Field Managers directly (rather than alternatives such as using agencies to do the work) meant that control of the quality of successful applicants was maintained and on the whole it ensured that people with the correct skills/competencies were employed.
The choice of sift and interview criteria meant that successful applicants at the Field Manager level had good proven managerial and survey experience.
A decision to allocate a modest advertising budget to Field Managers for placing ads locally was invaluable. It meant the Field Managers felt supported in the recruitment campaign and really helped target resources to some of the employment blackspots (generally areas with negative unemployment).
The use of external employment agencies in the later stages of the recruitment campaign to find suitable interviewers. Sufficient numbers were found though the quality of individuals varied.
Difficulties overcome:
There was a degree of confusion caused when the CCS posts were originally advertised. Adverts were published at the same time as the advertisements for Census Area Managers, sometimes in the same publications. Some potential applicants assumed that both adverts related to the same post.
At the interviewer recruitment stage in particular there was a lack of suitable candidates. This inevitably led to some otherwise unsuitable people being employed, who therefore needed a greater degree of training and monitoring in their day-to-day work.
The Interactive Voice Response System (IVR) helpline number used to request an application form was overloaded and sometimes it was difficult for potential applicants to get through.
Some people employed complained that the workload was higher than they anticipated and that the recruitment campaign should not have mentioned that the work could be done part-time in conjunction with full-time employment.
Lessons learned
The key point here is that despite the statistical requirement to keep the CCS independent from the Census there would be benefit in running a combined recruitment campaign for the two exercises. This would ensure there was no confusion or perception of uncoordination when the adverts were placed, and would help in making sure there were enough applications to fill all the posts. It is easier to run a single recruitment campaign to recruit 70,000 people than a campaign to recruit 66,000 people overlapped with a campaign to recruit 4,000 people.
Form Design
Although the CCS was to be conducted by personal interview with the householder, it was important to have a questionnaire that was clear and easy for the interviewer to complete. The form included a subset of Census questions and was designed and tested in an iterative way. Each rehearsal or pilot exercise carried out during the development of the CCS led to form refinements to make the task as easy and accurate as possible for the interviewer. At the same time the form had to be scannable for data capture. Colours, completion box sizes, layout and content all evolved on this trial basis.
Highly effective:
Prompts were included at the start that the interviewer could read to help the interview to be conducted politely and in a professional manner.
The form was designed such that the interviewer first listed all the people in the household and then collected details on each one in turn. This helped focus the mind of the interviewee on the person in question and ensured that accurate data was obtained.
Reminder guidance notes were included where necessary.
There were a number of "probe" questions included at the end of the form, further ensuring that nobody was being overlooked by the householder.
The form was designed to be used on a clipboard in all weather conditions. An early draft needed redesign to minimise the amount of page turning and manipulation by the interviewer.
Difficulties Overcome
A late decision to merge originally three separate country forms into one common form for use throughout the UK meant that the form had to cater for the differences between the three. Redesign took typesetting, desktop publishing and proof reading timetables up to the limit.
Lessons Learned
The best way to design the survey form is through a trial process, using it as much as possible in pilot exercises beforehand - feedback from those using the form in the field is the most useful information in this process. The data capture requirements are paramount throughout the design process. Late changes to form layout should be avoided wherever possible.
Field Methodology
Most of the field methodology used was developed as a direct result of the findings of the 1999 CCS Rehearsal and other pilot exercises. All the methodology was geared towards achieving a high response rate to a voluntary survey and therefore finding as many people as possible who had been missed by the Census.
Highly effective:
There is real evidence that the interviewers found whole properties that were missing from other address lists during their on-the-ground property listing phase.
Having a public interface in the form of the interviewer meant that any questions members of the public had about why the survey was being run could instantly be answered. Persuasion could be used to keep compliance rates up, and the form-filling burden for the public was non-existent.
The recommended calling strategy (suggesting that interviewers vary their calling times in order to increase the likelihood of finding the householder at home at some point in the survey period) limited the number of wasted visits. Interviewers realised this early on and varied the day and time of day when repeat calls were made.
The postback "last chance" option involving a self-completion questionnaire for all households not contacted by the last day of the survey resulted in a further 2 per cent response.
There is anecdotal evidence that the probe questions on the survey form worked well, with people who would have otherwise been missed being found in households.
In areas of underperforming response, particularly in London, the use of a "flexible field force" in conjunction with an extension to the fieldwork period meant that we were able to improve the final response rate so that it was more in line with the rest of the country. Team Managers were good at moving interviewers from areas with almost-completed workloads to areas which were falling behind.
Difficulties overcome:
There was some ambiguity and misunderstanding on how double workloads should be worked, leading to discrepancies in the way teams implemented this (The term "double workload" was initially intended to address the issue of one pair of interviewers handling two areas. It was subsequently used to describe the sharing of a large area between two pairs of interviewers which resulted in some confusion).
We received a small number of complaints from the public that persuasion was being taken too far and persistence was becoming annoying.
Lessons learned
The field methodology as it was implemented is sound - the final response rates to the survey of around 91 per cent are testament to this. Some clarification on the way sections of it should be implemented could be taken forward but this is more of a training issue.
Maps
Interviewers were provided with street maps for their area. The maps highlighted a rough boundary for the selected postcodes but this was only to be used as a guide. On-the-ground checks by the interviewers were to be used to identify households within the selected postcodes rather than identification from the maps or other means.
Highly effective:
The maps were very clear and easy to follow, even those for inner city areas.
Maps were delivered by hand to Team Managers at training events and so we did not rely on third party delivery companies. There was no scope for things to get lost.
Different levels of maps were provided (overview, detailed) and these helped Team Managers tremendously in planning the work throughout the period.
Difficulties overcome:
Some of the maps were fairly out-of-date leading to queries from Interviewers and Team Managers. This was not a significant problem due to the nature of their property listing duties.
Lessons learned
We need to retain the ability to produce maps on an ad-hoc basis throughout the survey period to replace those accidentally lost or spoiled. Also, compilation of the maps into Team Manager and interviewer packs was a significant administrative task that needed to be done at a time when the team were busy delivering training.
Geography and Workload Planning
This was a complex area that involved manually grouping selected postcodes together into interviewer workloads based on a number of criteria (e.g. expected household count, distance apart). There was a real need to keep them as equitable as possible to avoid conflict between interviewers.
Highly effective:
In general interviewers considered the size of their workloads (in terms of households) to be about right.
The whole workload planning was kept entirely independent from the Census operation and the sample treated with strict confidentiality at all times.
Difficulties overcome:
The geography and interviewer workloads did not scale up as well as was anticipated from the rehearsal exercise. We ended up with far more dispersed workloads than previously tested, resulting in greater travelling times and mileage claims, and higher costs.
Workload planning was very difficult due to the requirements of the survey design. Keeping postcodes within strata together and optimising the numbers of households included in workloads meant that it was not always possible for a pair of interviewers to work in a neat area around where they lived, but sometimes had to travel large distances between postcodes.
The proximity of interviewer workloads was judged on straight-line distances which was not always appropriate. There were cases where postcodes looked close together on a map but required a drive of many miles to get from one to the other due to the road layout and topography.
Although the areas Field Managers were responsible for had clearly defined boundaries, Team Managers had no such boundaries within which to work. Their workloads consisted of a set of discrete and often isolated postcodes with no clear boundary around them. This caused a number of problems such as Field Managers having great difficulty in allocating interviewer application forms to the correct Team Manager for sifting, and Team Managers becoming quite perplexed as to why their areas sometimes crossed other Team Manager areas.
Lessons learned
It is difficult to see how the workload planning could be improved - there are too many constraints imposed by the survey statistical design - but the amount of time required to complete the work should not be underestimated. Greater consideration of road networks could be made but this would increase the planning time required to prohibitive proportions without more people working on it. Team Managers should be allocated specific boundaries in the same way as Field Managers were.
Training
Training was delivered in a three part residential course to Field Managers and passed on in a cascading fashion to Team Managers and Interviewers. All training material was developed by the CCS Headquarters team. The training covered all aspects of the survey including the recruitment of staff, pay issues, conducting an interview, geography, TRACS etc.
Highly effective:
Headquarters staff delivered training directly to Field Managers, and trained Team Managers in conjunction with Field Managers. This kept the message very much undiluted and helped build up real enthusiasm for the project in the field staff.
All training events involved a large amount of team-building throughout and were residential in order for the managers to promote this aspect. The team-working aspect was proven time and time again to have real benefits for the CCS.
A very professional training video was produced showing interviewers how to carry out their job and was very well received by all.
The choice of regional venues was good, particularly for the residential courses.
Interactive elements and exercises were the most beneficial parts of the course, particularly those on recruitment.
Difficulties overcome:
One or two courses had too many people on them which made getting the message across more difficult. Some close management of these courses was necessary to ensure things moved on at the correct pace.
Different attendees had different preferred learning methods and rates of learning. One or two thought that the amount of material delivered was insufficient for the length of the training course, but this was offset by others considering it to be about right.
Lessons learned
Training courses have to be meticulously planned and rehearsed in advance which the CCS ones were. The training video proved to be a great success and was an ideal medium for presenting the interviewer work in a common way for 3,900 staff. Training events provide a good way to build the team spirit, essential for getting the CCS work done on time.
Logistics
A survey operation the size of the CCS required some careful planning of logistics, on both the delivery and the pick-up of supplies in the field. It required close liaison with the General Register Office for Scotland (GROS) and the Northern Ireland Statistics and Research Agency (NISRA) on forms design and delivery.
Highly effective:
All completed survey forms were delivered to the processing centre in time for processing to commence.
Liaison with TNT, the delivery contractor, was solid. Any problems were dealt with swiftly and efficiently.
Difficulties overcome:
Slippage to timetable for a variety of reasons, mainly due to shortage of resources as staff were side-tracked onto regional management roles.
Lessons learned
Careful monitoring of the delivery of all items to and from the field is absolutely essential. Headquarters need to know immediately of any problems arising. The amount of resource required to perform this function should not be underestimated. There should continue to be a good deal of liaison with the Census operation in this area as so many of its functions are common to both projects.
Public Forms
The survey questionnaires were developed and tested during the rehearsal and other pilot exercises. Although the survey was interviewer driven there was still a requirement for clear, neatly laid out forms to make sure that recording mistakes did not occur. The forms had to be manageable in the field, and had to ask questions in such a way as to avoid confusing the householder about who the questions referred to.
Highly effective:
The choice of questions, particularly the inclusion of probing questions, found people within households who would otherwise have been missed. Therefore as well as finding missing households (using the property listing field methodology) we were also finding missing people within households.
Very clear and simple showcards kept the interview time short - multiple choice questions could be displayed instantly to the householder.
Overall, a very simple matrix-based questionnaire was used that meant all questions were asked in a sensible order and information gathered appropriately.
Difficulties overcome:
A late decision to merge originally separate country forms into a single, common UK form for contractual reasons.
Lessons learned
The form worked very well. There are no reports of systematic interviewer or interviewee errors occurring as a result of the form design.
Public Helplines
A helpline number was provided that allowed members of the public to phone in if they had any questions regarding the purpose or legitimacy of the CCS. Unlike the Census helpline number the CCS one was not widely publicised as it was relevant to only the sampled 1.4 per cent of households. Instead it was printed on official explanatory letters and given out by interviewers on request.
Highly effective:
An excellent script was supplied to the company running the helpline that meant almost all queries were answered directly by them. Only 95 calls were passed to CCS Headquarters for answering.
Difficulties overcome:
None.
Lessons learned
The public helpline served a very useful purpose in explaining more fully what the CCS was all about. Demand was pretty much as expected with around 2000 calls received in total, a call rate of less than one per cent.
Use of Social Surveys Directorate Interviewers
About 100 SSD interviewers were employed to work on the CCS in an advisory and assisting role. The concept was that in an ideal world the whole CCS would be conducted by interviewers already employed, and therefore fully trained, by the ONS as occurs in other countries. However, the sheer size of the survey meant that this was impossible and so the approach was adapted to introduce in-house survey taking expertise whilst employing people specifically to carry out the bulk of the work. The SSD interviewers were allocated to the areas known to be most difficult to enumerate - mainly in London and other metropolitan areas.
Highly effective:
Many interviewers found the SSD interviewers a good source of information and advice.
A significant number of conversions from non-compliance was made by the SSD staff, boosting response.
Difficulties overcome:
In some areas the Team Managers were not entirely sure how to use the SSD interviewers effectively. This meant that they were underemployed.
Management of the SSD interviewers was split between the Team Managers and their SSD line management.
Lessons learned
All evidence suggests that having the SSD interviewers working alongside the CCS interviewers was a good approach and helped the efficiency of the data collection. There is a need to ensure roles are well defined and understood, by both the SSD interviewer and the Team Manager.
Handling Correspondence
Throughout the project period Headquarters received a fair number of items of correspondence, from members of the public (typically wanting to find out more about the survey), from unsuccessful field staff applicants (typically wanting to know the reasons why they were not successful), and from the field staff themselves (wanting clarification of instructions or chasing up pay queries).
Highly effective:
All letters were answered or acted on within ten days of receipt.
Difficulties overcome:
There were more items to deal with than originally anticipated, particularly from field staff as a result of the pay situation.
Lessons learned
Resources need to be allocated to ensure that we are able to respond to requests for information, arbitrate on
complaints and reply to queries from the public.
The method used for calculating pay rates based on expected task analysis was basically sound and meant that rates were defensible when questioned. Had hourly rates been used instead then costs would have quickly spiralled out of control.
Despite problems with the pay contractor procedures everyone was eventually paid in full, and advances were made to alleviate problems encountered with delays to payments.
Although the pay rules had to be pretty well fixed there was some degree of flexibility in the additional payment strategy which was used to compensate for a number of unexpected issues.
Difficulties Overcome:
The necessarily very detailed instructions and administrative form-filling requirement for the CCS managers did not go down well. There were complaints of too many forms to complete and too much time being taken up on these duties.
There was some friction between field staff over the differential pay rates being used and the criteria applied in defining the three pay bands, particularly as these differed slightly from the Census definitions.
Despite the prescriptive training field staff still made basic mistakes on their pay forms causing delays to payments.
There were problems associated with meeting the timetable for payment of field staff and a considerable amount of work continued to resolve pay problems following the survey closedown.
Exception limits for Field and Team Manager expenses were set too low. This was one area where scaling up from the rehearsal exercise was underestimated and many travel claims came out higher than previously anticipated.
Lessons learned
The whole pay area is a complex and emotive one that needs some very careful planning. Even with detailed instructions mistakes will always be made by field staff on their forms but good planning and testing will reduce this. A big improvement would be to separate the payment of fee from the payment of expenses so that the fee may be paid even in the case of a bad
expense form being received.
Overall Survey Performance
It is worth recording the overall performance of the CCS because this is a good indicator of the balance between things that worked well and things that did not go entirely to plan. The number of people found who had been missed by the Census is the key measure of its success, and although this will not be known until processing is complete evidence gathered so far indicates that the survey will have met its objectives.
Of particular note:
A national response rate (number of successful interviews achieved as percentage of properties found) of 91 per cent was finally achieved, outstanding for a voluntary survey;
In London and the South East, known to be the most difficult area to enumerate, a response of 84 per cent was achieved;
Of the 101 design groups covered, 58 had a response of over 90 per cent. Only 9 had a response of less than 80 per cent, caused by isolated pockets of low response within the design group.
The postback exercise for non-response was successful, increasing the number of households covered by nearly 2 per cent;
The national refusal rate was only 5 per cent, varying between 4 and 7 per cent by area.
The overall response is certainly considered a success since the size of the CCS was chosen such that it would provide built-in contingency for lower performing areas to borrow strength from others if need be.
The survey came in roughly on budget and within the allocated timescale.
Response Rates and Refusals
The following table provides more detail on the response and refusal rate at the time of survey closedown. Note that these are given as the number of households interviewed compared to the number of properties listed. They are not a direct measure of the level of Census underenumeration found, which will eventually be calculated from the combined Census and CCS One Number Census process.
Region
Percent of properties listed resulting in successful interview
Refusal Rate
North
91
4.2
Wales and West Midlands
91
4.7
East
92
4.1
South West
91
4.8
London and South East
84
6.9
Total*
91
5.0
*includes postback response.
Summary of lessons learned
There are a number of detailed issues arising in this report, but the key overarching lessons learned are:
All of the procedures developed for the CCS were successful - none failed completely though small improvements could be made here and there;
Good and early planning on all aspects is essential;
The team working ethos of the CCS was a real strength throughout the project;
The amount of resources required should not be underestimated. Dual roles are difficult to maintain.
Census News
