Worries that the 2011 Census will miss its targets are spelled out in an new assessment report from the UK Statistics Authority.  

This is the first of three the authority intends to publish on the Census, and it lays down several unambiguous requirements. The sharpest is the demand that the Census publish plans for how it intends to raise participation rates from the worryingly low figures achieved in the rehearsals - 42 per cent in Scotland and 35 per cent in England and Wales.
 
The rehearsals were voluntary, while the Census itself is mandatory, which will obviously help. But there is a risk, the authority says, that low participation rates may compromise the quality of the results. Even after follow-up, the response rate in England and Wales was only 74 per cent - 16 percentage points lower than the ONS's target figure of 90 per cent.
 
A second worry is whether there is sufficient money to make proper use of the Census data after it has been collected. In the current financial climate, there is a risk that areas such as data processing and output production will be under-resourced and that quality assurance could be compromised.
 
“We hope that the census offices will be able to secure the required budgets for the later phases of the Census as soon as possible, and we suggest the census offices develop and publish contingency plans in the event of any shortfall in funding” the report says.
 
The authority is concerned about arrangement for delivering the Census forms. In Scotland they will be delivered and collected by hand, but in England and Wales will rely on the post, using address lists that are being produced now. “Relying on these lists carries an inherent risk due to ongoing problems with missing and duplicate addresses” the report says. “The development of robust contingency plans should be a priority.”
 
There is also a technical issue. The Census plans to publish estimates of the resident population of the UK by counting all those who are resident or intending to be resident for at least 12 months, but the Census form will actually ask something different:  whether people are resident or intend to be resident for three months (in England, Wales and Northern Ireland) or six months (Scotland).
 
This will help capture short-term migrants, but raises the issue of how to convert the three- or six-month residency into 12-month residency so that the count is compatible with the official mid-year population estimates. The report says that the Census offices should publish explanations of how this will be done.  
 
The authority has also published a summary of its findings in the first 27 assessments of whether National Statistics are fit to retain their designation. 
 
In all cases it found they were, subject to certain changes being made. The greatest areas of failing identified were in meeting user needs, and in frankness and accessibility.
 
Under the act that set it up the UK Statistics Authority has the task of assessing every set of statistics designated as National Statistics – several hundred of them – and completing the job by 2012. Like painting the Forth Bridge, it’ll probably then be time to start again.
 
The first 27 reports, all available on the authority’s website, cover 80 sets of statistics. None complied completely with the Code of Practice in every respect, but all were conditionally designated subject to conditions, called requirements. By the end of February, ten sets had met these requirements and were duly designated.
 
The assessment team has found little to complain about in terms of integrity, and most of its recommendations relate to frankness and accessibility, meeting user needs, release practices, and impartiality and objectivity. 
 
In some cases, producers knew little about the users of their statistics, or failed to engage them in dialogue. Under frankness and accessibility, the team found a few cases where statistics had been released with no accompanying commentary and many where information, though plentiful, is hard to locate on websites. “The accessibility of ONS’s statistics on its website is particularly problematic” it admits.
 
The quality of the commentary on statistics varies. There are some good examples – the report cites the commentary attached to the statistics on the National Child Measurement Programme by the NHS Information Centre – and some not so good.