Vol. 428 No. 2
6 Mr. J. Bruton asked the Taoiseach the steps, if any, the Central Statistics Office must take from the collection of new census data to the publication of the data on a district electoral division basis; the extent to which modern information technology is used in each of these steps; and the reason that the use of information technology has not led to any reduction in the time lag between the collection and the publication of the data.
Approximately one million completed census questionnaires are received in the Central Statistics Office from the 3,200 temporary census enumerators about six to eight weeks after census day. When all the returns are accounted for, preliminary figures, down to the level of district electoral division, are compiled on the basis of summaries prepared by the enumerators at the end of the field operation. In 1991 preliminary figures for over 200 urban and rural districts were published within three months of taking the census. These figures have displayed a high degree of accuracy when compared with the final figures as these have become available.
The compilation of these final figures has to await the completion of two intensive processing operations which are undertaken simultaneously. The first is the area coding operation and involves a detailed examination of enumeration summaries, maps and listings with a view to correctly classifying every household and institution geographically. In addition to classifying households to district electoral divisions they are also allocated where appropriate within other legally defined boundaries. International recommendations have to be followed in determining the existence of new population clusters and in extending the suburbs of existing towns. Detailed comparisons are also done with the previous census returns in order to ensure consistency and completeness.
This area coding operation is of necessity extremely detailed and time consuming since it requires the clerical scrutiny of enumerators' maps and other field work documentation. The application of information technology has not had a major impact on it to date.
The second opertion entails the detailed statistical processing of each of the 1 million individual census questionnaires. This is carried out using CSO's highly computerised network environment and for the 1991 census consisted of the detailed coding of the 33 census questions, keying the coded information to computer and subsequently editing or correcting the output generated.
While information technology plays an increasingly important role in the census processing, substantial clerical resources are also needed, particularly for coding and keying the basic data. The major gain from the use of information technology has been the ability to generate a greater variety of analyses than was possible when the results were processed clerically.
The timelag between the taking of a census and the publication of final district electoral division figures for the State essentially reflects: the staff resources devoted to the production of the census results and the strategy adopted in the overall processing of the census, particularly regarding the balance between the finalisation of detailed and summary results.
In response to user demand, the emphasis in the 1986 and 1991 censuses was placed on finalising basis demographic details, that is age, sex, marital status, household size and composition, as early as possible and publishing these in local population reports on a county by county basis as soon as they were available. This strategy has been favourably received by general census users but has led to a longer publication timelag for the volume 1 report containing the full set of final DED figures for the State.
In the case of the 1991 census the timelag was also longer because a census of agriculture was carried out immediately following the fieldwork stage of the population census, using the same field staff. This joint operation, resulted in savings of approximately £1.5 million over the cost of two separate operations but meant that census headquarters staff were involved in the payment of field staff for longer periods than in previous censuses and, therefore, were not available for the processing of census of population results.
A summary statistical release will be issued at the end of this month giving the final 1991 population figures for Dáil constituencies. The final 1991 DED figures have been published already for most counties in the series of local population reports.
Is the Minister aware that the delay between the taking of the census and the publication of the detailed district electoral division breakdown of the census has been longer in the case of the 1991 census than in any census since 1926?
I have answered a question on this matter. I gave detailed statistics, which I do not have at present but which the Deputy may have, which show that on average the timelag has been roughly the same since 1926.
About the same, but worse.
In 1946 there was a 49 week timelag.
This is not a political matter, but will the Minister of State agree that the general public must be bamboozled, particularly younger members of the public and children who are familiar with the use of computers, that despite the availability of computers we are still slower in producing census data than in any years since 1926? The Minister has offered fairly convincing explanations, but they are not totally convincing. Will he agree that we should examine this matter further to see if there is adequate use of computers in this area or if the use of computers is as backward in the Central Statistics Office as it is in reporting the proceedings of the Houses of the Oireachtas?
Deputy O'Sullivan has informed me that once the office is transferred to Cork, data will be produced much faster.
I notice Deputy O'Sullivan is sitting beside the Minister in case he changes his mind. What about Trim?
In relation to the availability of statistics, while the general public have a great interest in getting various figures from the Central Statistics Office in regard to the census, the CSO has tried to respond to the requirements of the users of those statistics rather than appealing to what the population might consider a fitting timespan. Nevertheless, I accept the Deputy's point and I assure him that every available opportunity is used by the statistics board and by the Central Statistics Office to avail of computerisation where possible. In addition to the approach adopted by the CSO to meet user demands for census figures, the details of censuses now are much greater than was the case in 1926 or in the years since then. The more detailed the statistics become the longer it takes to produce them. When the Deputy reads the report of the statistics board he will note that it makes a number of suggestions for speeding up the process and I am open to any suggestions in that regard.
When the Minister has completed his consultative process in regard to statistical requirements will he consider having a short debate in this House in view of the fact that we are all users of statistics and must have accurate and reliable ones? This matter tends to be neglected in the political process and perhaps in the Minister's other capacity he might arrange a debate here on that matter.
Subject to agreement with the other Whips, I see no difficulty in debating the report of the statistics board some Friday.
Will the Minister indicate if he has any knowledge of Deputy Bruton's source for all this valuable material and whether he thinks Deputy Bruton might be claiming confidentiality in regard to the source?
The source is related to somebody on the Opposition benches.
I have made inquiries already and I think I know who the source is.
