Issued: 23 September 2013
The Royal Statistical Society (RSS) welcomes the public consultation by the Office for National Statistics (ONS) on the future options for the Census in England and Wales published on 23 September 2013. The decision on the future of the Census will affect the nature and quality of information available to everyone in the country when making choices about schools, hospitals, transport, shops and many other services for decades to come. The RSS urges all users of these statistics to contribute to the consultation.
It is essential that ONS gather good evidence about the decisions that draw on census-type information. In deciding on the expansion of a school, for example, what information is needed, when and at what level of geography? How accurate does it have to be? What is the economic, social or environmental impact if the information is not available, out of date or not precise enough? This evidence should be gathered first before ONS makes judgements about which option will best meet national data needs.
Once this is done, an analysis of the attributes of the statistics that could be generated to meet each of the user requirements should be made available so that users can see the basis of the case for each of the options. This analysis should include not just those statistics that would be generated directly by the option under consideration but also any modelling or use of alternative data that would be involved. This should be done in a clear way so that users can fully understand the implications for them of different approaches – users should not be required to try and work this out for themselves. Such analysis of each “use case” can then be put alongside the broader value to social science and evidence based policy and practice of the database that will be created under the various options.
To assess options it is also important to understand the assumptions that are being made about future changes in the context for gathering data. Public attitudes, the legal environment (eg for data linking) and changes to technology could result in a very different landscape for the census in 2021 and beyond. These assumptions provide the means to judge the sensitivity and robustness of alternatives and the level of risk each would entail given the uncertainty involved in predicting the future.
In its consultation, ONS recognises that there is still more work to do before a decision is made. In reaching its views on the way forward, the RSS will be looking to explore three essential criteria. First, the robustness of the evidence base gathered by ONS – what information does the nation need? Second, the statistical basis for analysing the alternatives – what quality of information can be collected to meet user needs? Third, the validity of the assumptions being made about the legal, attitudinal and technological context – what are the risks involved in taking each course of action?
Once this information is available, all those with a stake in the decision will be able to see how ONS has reached its assessment – which alternative provides the optimum combination of benefit, cost and risk? The RSS hopes that ONS will be able to come forward with an analysis in this form quite quickly after the consultation closes so that a timely decision can be taken during 2014.