Select Committee on Treasury Minutes of Evidence

Examination of Witnesses (Questions 60-79)



  60. There is a case for arguing that one has a right to be counted and if so under the Disability Discrimination Act there is possibly a case for saying that you have to provide the form in a way which is easy for the recipient to fill out. Do you want to comment on that?
  (Mr Cook) We would want to do that. We talked earlier about the bilingual Census form for Wales and I should like to think that we had a bilingual form in Wales not only because it was a legal obligation but because the organisation itself was alert to the importance of being able to communicate to the communities in Britain in the way they want to be communicated with.

  61. What other special needs did you seek to accommodate in the design of the Census form and in the conduct of the Census itself? Were there other aspects which you tried to anticipate?
  (Mr Pullinger) We have covered quite a lot in the discussions we have had to date. A lot of the special needs are around simply finding different groups of people and recognising that in a sense each person is an individual. You have to target your response on them. The special needs of the group called "babies", which I mentioned in passing; an important group to get but you tackle it in a very different way. We tackled that by giving every baby born on Census day a sleep suit which raised the profile of it and hopefully encouraged more people to include their babies on their Census form. A very, very large number of those kinds of initiatives were targeted at particular groups and particular areas. We would always try to find out where there are people who were going to have particular issues; questions around blind and partially sighted people and also deaf people. We had long consultations with the RNID about how we could ensure that those people got to be counted properly, different religious and ethnic groups with different needs and requirements.
  (Mr Cook) For women's refuges, we seek to be as sensitive as we can in how they are enumerated. The ONS sought guidance from the Home Office on those areas where additional language support would be required. We got lists of asylum seekers from the Home Office of where people were so that we could send people to those places who have the capacity to communicate. In all sorts of groups where we are able to build up knowledge in advance, either about exactly where they are or ensure that we are aware that when we have recognised them we ought to have some sensitivity and we seek to have a process for doing that. That has been a real success of the Census. Whether it is legal obligations or just greater awareness by the bureaucracy itself we should become better at it.

  62. Do you feel that you managed to reach the great majority of homeless people, people with no fixed abode?
  (Mr Cook) We put a huge amount of effort into doing that. In London the actual size of the areas which enumerators were working to was half the size of those in the rest of the country. We had special enumerators who had a particular interest in that. In addition we had a lot of support from the local authorities in London. Many of them helped us out with enumerators. Quite a number of them actually had staff in their planning offices who supported the local Census managers in how they operated. I visited Brent where a huge amount of support was provided by the local authority to our Census manager.

Mr Plaskitt

  63. Overall what ultimate response rate were you hoping for?
  (Mr Cook) We estimate that 98 per cent of the forms we sent out, which would have gone to households where people were eligible within the population, have been sent to us. The Census coverage survey will allow us to make that a more exact figure. About 95 per cent of the forms which we sent out in total have come back to us from completed households. Of the rest what we know from past Censuses, particularly the last Census, is roughly the proportion of households in effect not eligible, not on the Census because they are empty or for other reasons.

  64. You are aiming to get to the 1991 response rate for an overall response rate.
  (Mr Cook) We want a response rate in aggregate of around 98 per cent, that sort of order of magnitude. For several reasons. One is that a response rate of that amount will be much less in some groups such as young men, teenage males for example. We are seeking both to have an aggregate response rate, which is around that order of magnitude but the other thing we need to do is to not have a particularly large difference in the response rates across the different population groups.

  65. Across population groups, but what about across geographic groups? Are you anticipating that the response rate will fall well short of that in some hard-to-collect areas like inner city areas?
  (Mr Cook) We expect that to be the case. One of the benefits of the Census coverage survey is that it does give us the ability to get a very strong measure, a very good measure, an independent measure of that differential response rate. We have been able to do that by having such a comprehensive Census coverage survey this time.

  66. Do you have a return rate which you are aiming for within the inner city areas? It clearly is not 98 per cent there. Are you trying to get to something?
  (Mr Cook) We do not have a target but we have worked very hard to achieve that overall target. One of the things we recognised was that it was more difficult to get to that target in London so we halved the areas people had to work in. People actually had smaller loads so they could put more time into areas. A lot of our support activities have been in the more difficult to enumerate areas. Our most significant compensation has been in anticipating that as a problem and putting more resource in to even it out.

  67. So you have an overall target but no geographic sub-targets.
  (Mr Cook) Where we have thought we were falling significantly short of that we have reacted. We have found no significant areas where the things we have done were wrongly targeted in order to compensate for that prior expectation of where that shortfall would be.

  68. I think that is a no, is it not?
  (Mr Cook) Yes.

Mr Tyrie

  69. I still feel astonished that it should take so long to get these numbers out. I just wonder whether some of the information you are going to want to be providing to people is not going to be redundant by the time you have got it out. Are you not getting a lot of pressure for some sort of guidance, from parts of local government for example, for early indications of the numbers.
  (Mr Cook) The office is getting pressure from its own head. We really do want to get the information out earlier. We are set on a processing path which was agreed on with users in terms of the resource base we have and that is what is being done.
  (Mr Pullinger) A significant driver for our timetable was the resource allocation round for local government and that is one of the primary uses. One of the primary uses of the information is to feed into the local authority resource round and that gives us a requirement to produce information in August of each year. In a sense we had the task of producing results either for August 2001 or for August 2002. If we had gone for August 2001, we would have been able to do an Indian style adding-it-up job and the technology we have would have certainly allowed for that. We would in no way have been able to do the adjustment process, which all users came back to us last time and said they really wanted, so we could take account of these differences and give users confidence they were getting information which was comparable between areas. Clearly for resource allocation that is the critical thing. We were allowing for any differential coverage in inner London as opposed to some of the county areas. We were never going to make August 2001, so we targeted August 2002, essentially to get the best deal in cost terms by spreading it out. You only need to employ so many people for a longer period of time, only to buy so much equipment and it works out cheaper to be able to spread that out over the period. We targeted our budget to meet that timetable. We could have done things faster but that would have cost more and the user demand was to hit this August deadline which is what we are now doing.

  70. Give me a feel for how much more. We have spent £254 million. Is it going to be £264 million, £274 million?
  (Mr Pullinger) Our total processing costs are in the region of £50 to £60 million. If we had condensed the amount of effort we were putting into it, if we had halved it, I guess we could be talking about 25 per cent on top of that.

  71. Twenty-five per cent of £50 million.
  (Mr Pullinger) Yes, of £50 million.

  72. So up to £65 or £70 million.
  (Mr Pullinger) Yes. That is an estimate I am making off the top of my head based on the overall numbers. The fact of the matter is that we were not in that position. We had to budget to run the Census. We had some timings and we planned the whole strategy around the budget we had and the timings we were aiming for.

  73. It would be very interesting to see something in a note from you on that trade-off. That seems to be quite crucial. If the taxpayer is going to spend £¼ billion on a census, somewhere or other as a whole economy benefit we want more than £¼ billion back in some form or other. It is not your job directly to find where those benefits are being delivered. It is only something you look at indirectly. You look at it indirectly when you say you are refusing to add questions to the Census if you think the information might be available elsewhere. In a sense that is an on/off switch somewhere in between: yes, there is some value to this question but is there enough value. What we really need to do in order to assess the Census is to have some kind of estimate of what the whole economy return is on the Census. It is perhaps too difficult a question to answer orally, but could you give us some indication in writing of where you think that return comes? My guess is that it comes in the very area we have just been discussing: the needs assessment area and the value of the information, identifying serious need, deprivation, targeting particular problems in certain areas, which you can get from the Census but you cannot get with polling. In which case, it would be very valuable to have some idea how much benefit you are getting and whether you are getting £¼ billion from it.
  (Mr Pullinger) If there were a reliable model people could accept for doing that, that would be a wonderful thing to have because it would help with the decisions we are faced with on a day by day basis.

  74. You could then become world market leaders in this.
  (Mr Pullinger) Absolutely; that would be great and I am sure we could market the methodology.
  (Mr Cook) We could give you a couple of pages on an analysis of the impact on health, the ability to understand differential take-up rates or differential needs, the impact of an ageing population, the impact on the next 20 years of the significant shift in the ageing of the British population, the ageing of the British work force. These are all pieces of information which we would not actually have in sufficient detail to have thoughtful policy responses to.

  75. And knowing the savings we have as a result of being able to make those policy responses, which would not be available from polling evidence, is what is required in order to justify another Census.
  (Mr Cook) We could provide you with an overview of the major public policy areas.

  76. I just added a bit which you did not seem quite so keen to address.
  (Mr Cook) We are happy to do that.


  77. What were the criteria used to decide whether to prosecute for not completing a form?
  (Mr Cook) Several criteria. The first one was demonstrable evidence that someone had in fact broken the law. The second criterion is the ability to mount a prosecution. The third is a broader judgement of the basis of such a prosecution in terms of the commonsense nature of it, which is a more difficult judgement. Basically that is what decides whether we carry out a prosecution.

  78. If broadly two per cent of the forms are not sent back, how many forms is that roughly?
  (Mr Pullinger) Two per cent of forms would be two per cent of 30 million which is going to be 600,000.

  79. You have told us that 86 cases have been referred to the Solicitor's Office. Is that right?
  (Mr Pullinger) Yes, it is. Those numbers are both correct. We know we have no response from people but it is often very, very difficult to get the first criterion Mr Cook mentioned which is demonstrable evidence that someone has actually failed to comply. The evidence the court will require is an interview under caution with someone who has actively refused to complete the form. We have had far more difficulties on this occasion in making contact with people, partly for the reasons I gave, but partly also because with the post-back systems we were relying on only one contact rather than two, so the opportunity to get that evidence was reduced.

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2001
Prepared 26 November 2001