Mr. Cormac Halpin:

To respond to the question from the Deputy around the 50 tables, these were in the dedicated thematic release relating to disability from 2016. There were more like 90 tables across the board produced across the full suite of information from 2016 where the disability variable was included. What these tables I am referring to do is to cross-tabulate people with the various categories of disabilities that appeared on the census form in 2016 with other key variables on that form. For example, "general health", "economic status", "social class" and "socio-economic group". Geographic aspects are also included, so they would be disaggregated down to county level, electoral division level and then down to small area level. This is really about bringing out the various different characteristics of persons who reported on the census form that they had a disability and locating where they are in the country. This is the essence of census data.

Turning to the 2027 census, we are open to including any different cross tabulation that users or people might want to include. The only real limits we have on producing census tables is ensuring that we do not disclose information about persons on an individual level. This goes across the board and is possibly the key tenet of census taking. Subject to that constraint, however, we can cross tabulate the disability data with any other data on the census form and produce it to the lowest level of geography possible, which allows this disaggregation that is so important to users. Looking forward to 2027, it is going online. I stress it is early in the development of the 2027 census. Why are we talking about 2027 now? This is, unfortunately, the way the census works. It is an unusual project in that it has an immutable deadline. The census will take place in April 2027 and this cannot be changed. This aspect brings its own dynamics in terms of working backwards when certain things have to be achieved. We must, for example, have all our IT systems ready in 2026. This cannot change, because if that does not happen, then the census cannot take place in 2027.

This means, therefore, that we must start preparing for the next census before the current census data has been disseminated. If we waited until the dissemination programme for this census was over, then we would not be able to have a census in 2027. This would be particularly the case because it is going online, which is bringing up major technical challenges around data security, storage, etc. We must start preparing now, and we must also start preparing the questions. We intend to hold a census pilot next year in September, which will test the new questions and any revisions to existing questions that will be included in the 2027 census to ensure they are comprehended by the public and deliver accessible and good quality data. This is why we must have the consultation now. It will allow us to form a census advisory group which can give us recommendations and allow us to form those subgroups. I take Nem's point that it might not have suited every group. We have had 400 submissions from hundreds of different organisations across the board regarding the census and it is probably difficult to suit everybody. We can probably work offline with Nem, though, to see if we can accommodate additional submissions from the DPO network relating to the disability question. We do not want to be exclusive, rather we want to include everybody to the greatest extent we can.

The comment I made in my opening statement about the organisations we worked with was for a very early pilot test of going online. It was very limited and we will do much more testing as we go. Accessibility is a key consideration for us, so we will continue to work with those organisations we have worked with to date and expand this endeavour to other relevant organisations. A key facet of the census is coverage. The fewer the number of people who complete the census, the less valuable it is. We want to be as universal and accessible as we can. There have been challenges in the past. Part of this has related to the paper form. We know people have had difficulties completing the form, but we are hoping that going online will alleviate some of these difficulties. We will, though, still have the same level of support in place that we had and hopefully we will enhance it going forward, especially with the focus on the website, to make it as accessible as possible. It is in the CSO's interest to be able to say that we got to 100% of the population and did not exclude anybody. If we exclude groups, that casts doubt on the data. We will, therefore, strive to the greatest extent possible. There is much international experience, especially in other English-speaking countries, regarding censuses that have gone online. We have close relations with all those countries and we will continue to liaise with them to bring best practice to the greatest extent possible to make the census 2027 accessible.