Open data is all the rage across government and in IT circles. If data is published in a format that people can use, understand and manipulate (in a good way) it leads to greater transparency.
Critics of the Press Complaints Commission often use data to support their arguments. The most frequent criticisms are that:
- it doesn’t receive enough complaints;
- it resolves too many cases at the expense of adjudications,
- it rarely finds against newspapers
- those infrequent adjudications lack sufficient prominence
The PCC rightly defends itself against criticism which it deems unfair, and has previously criticised Nick Davies, the culture media and sport select committee and the Media Standards Trust for misunderstanding its data.
Now, thanks to PCC Watch, we have greater transparency of the PCC’s data. Rather than having to calculate totals manually and read through individual cases, the site does it for us. But it doesn’t produce transparency because the PCC’s data is too complex.
The table below shows the key data for January – December 2009. It is taken from the three main sources: the list of adjudicated and resolved complaints on pcc.org.uk (which form the data for PCC Watch’s site) the 2009 annual report and an amalgamation of the biannual summaries of 2009 (excluding Jan-March).
None of them agree on how many complaints were investigated, how many were resolved, adjudicated, upheld or how many did not breach the code.
|PCC website||2009 annual report||Apr-Dec summary *|
|Complaints investigated||Not published||1134||1902|
|Adjudication not upheld||20||Not published||17|
|No breach of the code||Not published||993||516|
* Jan-Mar is not included because this data is only available in a Oct-Mar 2008/09 summary
Perhaps the biggest challenge (and the one most easy to resolve) of analysing the PCC’s data is the inconsistency in its terminology. The website simply divides cases between those which are resolved and those which were taken to adjudication. The annual report then subdivides these into those that were resolved to the satisfaction of the complainant (609) and those that were resolved to the satisfaction of the PCC (111). This useful distinction ought to continue but it should also be possible to identify which cases on the website fall into which category.
The website and biannual summaries make clear how many adjudications were upheld, and how many were not. However, the annual report abandons this terminology, instead favouring to cite instances of ‘public censure’ – which is the same as an adjudication. This might be obvious to those familiar to the PCC, but possibly not to the general public. And the figures are different because a number of cases were published in January 2009, but the PCC actually made the decision the previous year – so it was included in the previous annual report.
The complaints investigated data is also unclear. The annual report declares 1134 complaints investigated. Given that 720 were resolved it implies that of the 993 where there was no breach of the code, the vast majority required no investigation to determine that this was the case. In fact, that’s not the case. The PCC secretariat makes an initial assessment about whether there’s a prima facie breach, and this is checked by the commission. That may be an investigation to some and not to others – but the lack of clarity is unhelpful for everyone.
Not all cases are published on the PCC’s website. If the complainant does not wish it to be, it remains unpublished. That makes perfect sense for the complainant but is less satisfactory for the public (or the journalist if the complaint was rejected). For example, it means there is no certain way of telling how frequently someone complains to the commission.
Effective regulation usually focuses on outcomes above process. To that principle, the PCC’s work might be altogether more clear for the public if it drew the following, simple distinctions:
- Cases which were outside the PCC’s jurisdiction because the complaint was not covered by the code
- Cases where the newspaper did nothing wrong (either because there was no prima facie breach or because it was not upheld at adjudication)
- Cases where the newspaper did something wrong but corrected it sufficiently to avoid censure
- Cases where the newspaper did something wrong to such an extent that it required public censure
A spokesperson for the PCC told us that:
“The PCC already publishes a lot of data about its complaints and is always happy to explain this in more detail to anyone with a specific query. We recognise that there is more to do to make our complaints data easier to understand, however, and are actively engaged in improving the ways that we present our statistics.
“The PCC has recently embarked on a major overhaul of its database and website to help us with this. As an organisation, we are committed to openness and accountability, and of course it is important that our data reflects this.
This is welcome. It is unclear that anyone benefits from the lack of clarity – not least the PCC which cannot place a reliable number on how many cases it deals with each year.