The census and future provision of population statistics in England and Wales
The answers below are in response to the ONS consultation document available here
Q1: What are your views of the different census approaches described in this document?
LARIA represents a range of different local researchers: in particular local government, NHS, police, fire and rescue services, private sector and individual freelance and academic researchers. We have 700 individual members and over 119 corporate member organisations. A full list of our corporate members is available on our website. The views expressed in this submission are on behalf of LARIA members. We have engaged with them through inviting them to attend a workshop discussing your consultation (held on 22nd October 2013) and to take part in discussions online. We have also encouraged them to individually respond to your online consultation questionnaire with specific points relevant to their organisation and their own work.
We support a mixed approach: both a future census and the greater use of administrative data. We think that although great progress has been made on the use of administrative data that the new model would have to run alongside at least one census to validate the findings.
There is value that only a census can provide: in particular the ability to gain data at a very local level and to cross reference findings. We do not believe that the ability to cross tabulate data will be in place by 2021. We could therefore lose sight of very small communities and groups of people within local places. This would make it harder to deal with issues of inequality and could leave the most hard to reach further excluded from our society.
We commend the work that the ONS has done to use administrative data in population estimates and look forward to further details. We fully support this initiative moving as quickly as possible but we do identify some risks, and how to address them, later in our submission.
Q2: Please specify any significant uses of population and housing statistics that we have not already identified
Census data is integral to the work of many of our members. In a recent member survey we asked which activities were most important for LARIA members to be involved in over the next 12 months. Over half (53%) said developing the future of the census beyond 2011. This was the top answer.
The data is used by virtually every public body at a local level: local authorities, police, NHS, and Fire and Rescue Services to name just a few. This helps them tailor their services to the needs of their local population. In particular it helps to identify areas of inequality and the most hard to reach individuals. Often they are the heaviest users of public services and therefore accurate estimates of their need are vital. They are also the most likely to not be a stable presence in administrative data making the job of capturing their need all the more difficult.
ONS population and housing statistics are also often used by a whole range of individuals and small local organisations. These include local ward or parish councillors, charities and regeneration areas. These users often contact their local authority for help in accessing ONS data or to access their local analysis: such as providing a local profile. The local authority may draw down the data but this then feeds out into a broad ecosystem of local research needs. An example is that the Big Lottery fund often requires a local profile for its work based on ONS data. It is difficult to put an estimate on the number of local organisations and individuals that access ONS data in this way. However, a conservative estimate would be for at least one request a week. This may not sound huge but with over 400 local authorities over a 52 week period this estimates demand at over 20,000 requests. Changes to the approach in collecting population and housing statistics run the risk of undermining the work of a rich network of local organisations: many of which we suspect have not responded to your consultation as they are not aware of the role of the census in feeding into the analysis they see.
Q3: Please specify any significant additional benefits of population and housing statistics that we have not already identified
See answer to Q2
Q4: What would the impact be if the most detailed statistics for very small geographic areas and small population groups were no longer available?
High. Our members are very concerned at any attempt to remove access to small area data. We require as much granularity to the data as possible. The more that statistics are reliable at a small area the more likely we are to be able to identify and tackle issues of inequality. There are many areas where huge differences exist within a local authority and areas such as wards. Not having small geographic areas and small population groups runs the risk of giving the impression of an increasingly homogenised nation. It is likely this would undermine the credibility of ONS statistics as the evidence on the ground would suggest an increased diversity.
Q5: What would the additional benefits be if more frequent (i.e. annual) statistics about population characteristics were available for areas like Local Authorities and Electoral Wards?
High. However, this depends on the quality. We would like to have data on an annual basis and we do not think it needs to be more frequent than this. The benefit would be to have more consistent faith in the data. Census data does date and has less use after about 6-7 years. This can have a negative impact on decision-making, not just in terms of being based on out-of-date information but decisions being delayed while we wait for the next census data. Anything that could be used to provide greater consistency in the data available on a year by year basis would be much appreciated.
Q6: Please specify any significant uses of census information for historical research that we have not already identified?
Researchers need the ability to access historic data and to be able to replicate the findings.
Q7: What advantages or disadvantages for genealogical or historical research can you see from a move to a solution based on archiving administrative data sources?
We think that archiving separate administrative data sources will appeal to genealogical research. However, we are concerned that the combined dataset will not be available to future social researchers. This dataset would form the basis of the figures that are produced by ONS and it would be vital to future researchers to be able to go back and validate the figures as needed. This would be made more difficult if the data sources were only available separately and a combined dataset had to be assembled each time. Social researchers would only need to have access to anonymised data.
Q8: What are your views of the risks of each census method and how they might be managed?
The risk of the online census option is that we feel that the estimates of the proportion of people who would complete it online are too high. We think you will need to do more face-to-face fieldwork.
1 – This would be a new use of administrative data and the public will have to be engaged in this. Knowing that a particular data source was being shared could drive down registrations and make it harder to deliver particular services.
2 – This would also be a new skill set for ONS. We suggest that the Government establishes within the remit of the ONS or as a separate organisation the Office of Administrative Data. There is a need to ensure central responsibility for the quality and stability of administrative data. This is a different set of skills and responsibilities to those that the ONS currently holds. We do not think that the ONS is able at present to compel access to certain data sources or set standards. There is also no power to ensure that the importance of administrative data in population estimates is recognised and protected. Changes in future policy could alter key data sets. This has already happened due to changes in the welfare system.
3 – We are particularly keen that ONS develops a methodology that allows local areas to provide useful data to fine tune the findings. This may only be available in some areas and not others but could provide a degree of granularity needed in certain areas. We are concerned if the ONS is setting a bar that only enables national datasets to be used in the same way everywhere. There are areas where more detailed datasets would be relevant but would not have to be used or available elsewhere. We would welcome a dialogue with the ONS on how best to reflect local circumstances and to bring in the expertise of our members on their local areas.
Q9: Are there any other issues that you believe we should be taking into account?
We would encourage the ONS and Government to make a quick decision in order that we can discuss the types of data that need to be collected. In particular we are concerned that the new census questionnaire is not being discussed while we debate methodology.
The ONS undertook substantial work in 2011 to engage with local authorities and other local organisations. We feel that this helped to ensure that the effectiveness of the delivery of the census was maximised. Local authorities were at the forefront of this work and committed substantial resources to help the ONS run the census locally. We feel that this level of support is unlikely to be available in 2021 or indeed in the lead up to the census. Local government will be smaller and resources are likely to only be available for the priorities of the organisation. We would therefore encourage ONS to engage more with senior decision makers within local authorities, but also other local organisations such as the NHS. The consultation has been engaging with researchers on the technicalities, but the final delivery has to engage with those leading local organisations
LARIA is keen to help and support the ONS in any way we can. We have a network of local researchers who we think would engage with you to ensure the very best estimates of their local populations in as fine a detail as is possible.
Photo credit: Newport, Wales taken by Andrys Stienstra