Smoking and alcohol time series under threat

Plans by the NHS Information Centre to save £300,000 a year by withdrawing its  contribution to the cost of the General Lifestyle Survey could see the end of  a 40-year time series on drinking and smoking.
Andrew Lansley, the Health Secretary, has been told by the UK Statistics Authority that the decision, apparently taken without user consultation, is in breach of the Code of Practice for Official Statistics.
The Office for National Statistics has just completed its business plan for statistics after the cuts, so has no maney available to make up the shortfall, Sir Michael Scholar, chair of UKSA, says in a letter to Mr Lansley. So that means, he says, that the decision by by Information Centre will lead to the immediate discontinuation of long-established national statistics on smoking, drinking, health conditions and use of health services. 
While the decision is up to the health department, Sir Michael’s intervention  makes it plain that continuing with the plan will cause trouble. Not only would it be a breach of the code because no proper consultation has taken place, but it would eliminate a time series that monitors important health behaviours. While other sources may be available, they are likely to be inferior, he says.
A spokeswoman for the Information Centre said: “Like all government bodies, we are reviewing all our areas of spending. We are aware of Sir Michael’s letter and the DH is considering the matter.”
The DH said: "No decisions have been taken and discussions are still ongoing with the Information Centre. The Department will respond to Sir Michael Scholar once these have been completed."
Politically, any decision that to stop measuring alcohol consumption would play badly for Mr Lansley, who is already seen by anti-drink campaigners as being in the industry’s pocket. They have urged him to take stronger steps to discourage drinking, so any suggestion that it was being measured less reliably would imply he was letting alcohol rip while not being prepared to measure the consequences.
But the situation is not straightforward. The ONS is in the middle of a consultation on the future of the GLS (the acronym used by ONS is actually GLF, for reasons too complicated to explain) with the default option being to abandon it in 2012, switching some questions to other surveys. The statistics on income and living conditions required by European law would, for example, be collected in the Family Resources Survey. But the consultation, which opened on 11 February and will close on May 6, has little to say about the future of the health questions.
If the GLF ends with the 2011 survey, ONS “will explore ways in which the other data can be collected in other surveys”, the consultation document says. The intention is to switch these questions to the Omnibus survey (now called Opinions) but this depends (I understand) on the NHSIC continuing to pay its share of the costs.
If this course is adopted, the data would continue to be gathered, but using a different methodology. So critics argue that the continuity of the time series would be lost, even if the NHSIC relents or DH provides it with another £300,000.
The General Lifestyle Survey (formerly the General Household Survey) has run continuously since 1971, with two brief interruptions in 1997-98 when it was reviewed, and 1999-2000 when it was redeveloped. In 2009 more than 8,000 households took part and 15,000 interviews were conducted.
It collects data on

  • demographic information about households, families and people 
  • housing tenure and household accommodation 
  • access to and ownership of consumer durables, including vehicles 
  • employment 
  • education 
  • health and use of health services 
  • smoking 
  • drinking 
  • family information, including marriage, cohabitation and fertility