Update: Open Misinformation and the Illusion of Transparency
William Waites
government contractor, the Open Knowledge Foundation has released
details of its financial transactions with the UK government:
http://lists.okfn.org/pipermail/okfn-discuss/2011-July/007890.html
A brief look at their account with the Cabinet Office (COI) shows that
not only some 180k GBP [1] is unreported, but the two transactions
that are reported do not match anything in OKF's books. I would tend
to trust OKF's books because they actually have a legal requirement to
keep them...
What does this tell us? Where do these numbers come from? How is it
that even the Cabinet Office apparently does not deliver on the
transparency promise? Is this data of any use at all?
-w
[1] total of invoices > 25k less the 70k that actually is reported by COI
Christopher Gutteridge
On a related note, we published some of our (University of
Southampton's) payments as a dataset, and some information was not
included as it was either strategically sensitive, or subject to data
protection.
When I used this for some visualisations, it gives a false impression of
percentage spend on various things. It would seem like good practice, in
transparency data to have a clear way to indicate amounts in a dataset
deliberately not shown. That would be helpful to see the difference
between accidental and intentional omission. Maybe a few standard
categories, "non itemised payments" for things < �500 (or whatever the
threshold used), "sensitive payments" or "payments not shown due to
dpa". Most (all?) of the DPA related stuff was, I believe, staff
salaries, so maybe that could also be shown as a combined value rather
than individual payments.
Unfortunately there are some bugs in the tool which exports our finance
data, and I'm still waiting on a patch/fix from UNIT 4 who make the
Agresso software we use. On the upside they do have an RDF export, which
will be great when and if they get the kinks out.
--
Christopher Gutteridge -- http://id.ecs.soton.ac.uk/person/1248
/ Lead Developer, EPrints Project, http://eprints.org/
/ Web Projects Manager, ECS, University of Southampton, http://www.ecs.soton.ac.uk/
/ Webmaster, Web Science Trust, http://www.webscience.org/
David Read
> The speculation in the original article is no longer such. The example
> government contractor, the Open Knowledge Foundation has released
> details of its financial transactions with the UK government:
> http://lists.okfn.org/pipermail/okfn-discuss/2011-July/007890.html
>
> A brief look at their account with the Cabinet Office (COI) shows that
Ah, I see where you have misunderstood things - COI (Central Office of
Information) is an intermediary between the Cabinet Office and
providers such as the OKF. i.e. look for the COI transactions in the
CO spending data.
So - no conspiracy after all. But I do think 'armchair auditors'
should ask government sensible questions about the data, to help
understand anomalies. This list or the data.gov.uk forum are
reasonable places to ask these questions.
It's also worth mentioning here that, as a government department, the
Cabinet Office is "only" asked to publish transactions over £25000
(whereas local authorities the bar is £500). Check the PM
announcements last summer for all the facts, if you are interested.
And yes, lobby your MP to open it up even further, should you find the
data thus far valuable.
I was very pleased to see the recent announcement that the spending
data will start to refer to payees with unified URIs.
David
William Waites
> Ah, I see where you have misunderstood things - COI (Central Office of
> Information) is an intermediary between the Cabinet Office and
> providers such as the OKF. i.e. look for the COI transactions in the
> CO spending data.
Actually, I was considering the reporting where the recipient is OKF. Unless
OKF appears in the openspending as something other than "Open Knowledge
Foundation" (and I've checked a few plausible alternatives) this should be all
of them:
If you or anyone else can find the missing data and post a link, I'd be very
interested (and would seriously consider eating my hat).
> It's also worth mentioning here that, as a government department, the
> Cabinet Office is "only" asked to publish transactions over £25000
> (whereas local authorities the bar is £500).
I am aware of this. I am only considering OKF invoices > 25k. I am also
assuming that the payments were't made in chunks of £24999.
> I was very pleased to see the recent announcement that the spending
> data will start to refer to payees with unified URIs.
That would be nice.
Cheers,
-w
Christopher Gutteridge
included in a dataset for a period, where possible broken into reasons
for non-inclusion?
--
Paola Di Maio
what original article?
(sorry have not seen a reference to an article in this thread)
cheers
P
William Waites
> William
>
> what original article?
> (sorry have not seen a reference to an article in this thread)
Was mentioned in a thread of almost the same name the other day:
http://eris.okfn.org/ww/2011/07/misinformation
Sorry I wasn't able to easily join the original with the update because
of the way google groups handles copies to self.
Cheers,
-w
Paola Di Maio
thanks for the update, had not seen the original link
I became curious about funding of CKAN (and other stuff) a while back,
and poste a note to this list, also asked around, but got replies
also put in a foi request, still unanswered
did not talk to you directly because you work for them and did not
want to compromise your confidentiality and loyalty to people who give
you work
However bringing transparency to the IT supply chain goes well beyond
disclosing financials
I meant to add: not only the value of financial transactions (amount)
but the specifications and justification of how decisions were made
(was it public tender, that lead t contracts being awarded, if so can
the documentation - call for tender, bids submitted, documentation of
the selection process etc be accessed? ) should via foi if nothing
else
will keep on stirring the pot, hope everyone will gain from more info,
the contractors, the middle men, the public and the governing bodies
hope the trouble is for the better good
cheers
p
Paola Di Maio
also the public should be able to see (and ideally help to devise)
the requirements and some specs for the work commissioned by public
bodies to be comfortable with the whole process is not casually
undertaken
P
On Thu, Jul 21, 2011 at 9:22 PM, Paola Di Maio <paola....@gmail.com> wrote:
> Will,
>
> thanks for the update, had not seen the original link
>
> I became curious about funding of CKAN (and other stuff) a while back,
> and poste a note to this list, also asked around, but got no replies
William Waites
>
> also the public should be able to see (and ideally help to devise)
> the requirements and some specs for the work commissioned by public
> bodies to be comfortable with the whole process is not casually
> undertaken
Paola, just to clarify. In this instance I have not been looking into any
question of decision-making transparency, though I generally agree
that a more transparent process would be a good thing. Rather, I have
been trying to ascertain something about the quality of the financial
data published by the government. OKFN is only incidentally involved
because it was possible to get their side of the accounts to compare
with the data published by the Cabinet Office. From this perspective,
how OKFN came to be hired and CKAN used is not relevant, the
starting point is that it was hired. I also made a point of asking OKFN
for the relevant parts of their accounts in a public forum so as to be
able to do this analysis from public information without recourse to
any confidential information that I may have been privy to in my work
with them.
The conclusions, that at least for this subset of the data there is much
information missing and what information is published seems to be
wrong is worrisome because it puts the whole of the financial data
published in the £25k spending series in doubt. The situation would
appear to be far worse than the standard line of "the data is known to
contain errors".
That the data is presented by the government, both on the prime
minister's office web site and data.gov.uk as ostensibly accurate
without any obvious statement about errors and omissions carries
implications that I'll refrain from speculating about if, say, a journalist
or policy analyst were to try to use it to draw any kind of conclusions
or indeed do anything at all with it.
-w
Paola Di Maio
(I expressed similar frustrations in the past)
*my* point by contrast is that financial data represents only one part
of an even more interesting set,
and it's not just how much money is exchanging hands, but also exactly
what for? according to what mandate? following what procedure? just
wondering...
and not just the OKFn, or the UK Gov
any organisation handling public contracts in the EU should also be
transparent and accountable, especially in relation to support
initiatives designed to promote openness transparency and
accountability (dogfood type of argument)
guess is a learning curve...
P
6tricky9
On Jul 21, 4:21 pm, Christopher Gutteridge <c...@ecs.soton.ac.uk>
wrote:
> Unfortunately there are some bugs in the tool which exports our finance
> data, and I'm still waiting on a patch/fix from UNIT 4 who make the
> Agresso software we use. On the upside they do have an RDF export, which
> will be great when and if they get the kinks out.
>
Richard