Re-engineering the statistical service
Despite a quiet announcement, a significant change occurred in June to the management of Government statistics as a new semi-independent agency for the UK was launched.
The new National Statistics agency meets a Government manifesto pledge to provide independent national statistics "that command the trust and confidence of the public".
This is just in time. The accuracy and credibility of Government statistics have taken quite a battering.
Accuracy and trust are essential. The Government has committed itself to improving confidence by demonstrating that they are produced to best professional standards and free from political interference.
So, whilst the new agency must deliver a significantly improved service, its integrity will be guaranteed by:
· an independent Statistics Commission;
· a greater scrutiny role for the UK Parliament and devolved legislatures;
· statisticians protected by a professional code of practice;
· a right of access to the Prime Minister or to First Ministers on matters affecting the integrity of government statistics.
These principles are enshrined in a Framework Document that contains two significant features - one good; one bad:
· The agency will not be the exclusive instrument of Government. It is charged with providing a "comprehensive and meaningful description of the UK economy and society" to the UK legislatures and "the wider community". Its work is meant to guide "decision-making and debate" and to offer a "window on the work and performance of government" allowing the impact of government policies and actions to be assessed.
· The agency is tasked with minimising the burdens on business and other data collectors like local government. Whilst smarter IT may help, the Government believes it can minimise these burdens by relying on existing sources and administrative data "wherever cost effective and practicable" and "controlling the use of censuses and statistical surveys to minimise the burden on data suppliers"
Ministers have said the agency represents the biggest overhaul in 30 years. It overturns the "Rayner" principles that have determined the priorities and culture of the Government Statistical Service for almost two decades. Rayner created a secretive and suspicious service that saw central Government as its only legitimate customer. It rarely sought the views of non Governmental users and introduced a pricing regime for data of almost racketeering proportions.
Now that its customers are recognised within government and the wider community, the service needs to focus on what its users actually need. The Framework Document says that Government statistics should help business to be more efficient and furnish citizens, researchers and analysts with the information to judge Government performance.
Although Government is now just one of the customers, it is a very important customer and requires a step-change in the way that it is serviced. National Statistics will centralise all the statistics of Whitehall Departments that previously did their own thing - like Environment, Transport and the Regions, the Health Department and the Home Office.
This should help bring coherence to policy making. "Joined up Government" requires statistical sources that can offer multiple insights into specific policy problems and which help to identify the causes not simply the consequences of economic or social failure.
And the new agency must do it across new - and generally smaller - geographies. The statistical service is still saddled with an outdated business model for Government that is geared to knowledge about the country as a whole. Yet there is scarcely any branch of Government today that intervenes on a UK-wide basis wielding a one-size-fits-all policy basis. Increasingly policy is focussed at the regional, sub-regional and even lower level. We live in a country where intra regional variations are usually more profound than inter regional differences.
So National Statistics must provide data from diverse sources that focus on specific small areas. Policy Action Team reports have particularly identified the need for co-ordinated information where neighbourhoods face multiple disadvantage. Accurate, timely, compatible and coherent data is required to diagnose problems, develop strategy, re-design public services and to measure the effectiveness of intervention. At the very least, we need to know what works and how much it costs.
This is the biggest challenge for the statistical agency.