Post at the Guardian

July 16, 2010

Just a quick post to link to a piece I wrote for the Guardian Data Blog. The article explains how The Whole of Government Accounts was excluded from the COINS publication and when I requested the Whole of Government Accounts I was refused.

Advertisements

Interesting rejection of a FOI request

July 13, 2010

The whole of government accounts (WGA) has been running every year, for 10 years, and during this time the public have seen exactly zero results from the exercise.

The WGA gives a very detailed picture of the financial health of every government funded body — so clearly would be really useful to where does my money go.

I was told by the small team at the Communities and Local Government that the WGA 2008/09 is sitting in the Treasury. I was also told that two departments failed the WGA audit (the department of health and the ministry of defense), which led to none of it being published.

So I asked for that report (minus the results for the two departments that failed the audit) in this Freedom of Information Request.

The rejection I got was unlike anything else I have received.

They have the information I requested, but the response went to on say:

“Ministers and officials need space in which to develop policy, including space for the development of policy through an interative process of testing and refining ideas. This process could be weakened if information was released prematurely or when proposals where not finalised, as this could lead to poorer decision-making”.

Any suggestions for how to proceed are most welcome.

I should also add that the WGA is stored in COINS, but none of the WGA was included in the COINS publication.


Long Term Trends

July 9, 2010

I’m liking the long term trends part of the ‘Where Does My Money Go?‘ display for the Country Regional Analysis data set.

The Country Regional Analysis shows where public spending had effect. This is for 2009, but I’m told the 2010 database dump will be published on 16th July.


Meeting Cambridge County Council

July 9, 2010

I met a couple of people from the finance team at Cambridge City Council (CCC) this week.

This meeting makes me want to beam a message out to all councils saying:

if you publish your data in machine readable form, you pretty much don’t have to worry about presenting this data — there is already a community who will do it all for you for FREE. Everyone wins!

The outcomes of the meeting where:

  • They’re going to give me a copy of their accounts in a spreadsheet.
  • They refused to give me an export of their Oracle “Balance Sheet” report.
  • All council budgets are published in excel
  • They’re going to report their £500 + spending before the Jan 2010.
  • When they publish they’re going to make sure their data meets the open data standards defined by Chris Taggart.

Here is some more detail about the meeting for those interested.

Background:
I’ve been asking to meet with the finance team in CCC for months. The first thing I did was write to them to request a meeting to explain ‘where does my money go?‘ and get some idea of their finance data. I didn’t hear back so I asked again. Nothing back. So I sent a freedom of information request for the database type, schema and training notes all of which I duly received.

Then I asked for the data, and one of the councilors at CCC saw my request on What Do They Know? and helped me by giving the exact tables that I needed and he also suggested a report to ask for. I asked for all of this and added that I would like to meet as I appreciate it is a big ask.

I didn’t get the data but I did get a meeting.

The meeting:
I explained the ‘where does my money go?‘ project.

They explained about their current work reporting all spending above £500.

They said that the vast majority of their spending data is below £500, but even so this still is more data than they have every shared before.

They plan to share their data before the Jan 2010 deadline and they are concerned about if the public will be able to interpret it and also how to physically host this large amount data.

I pointed out the open data standards from Chris Taggart and co and how there is a community of people eager to do the work of communicating the spending to the public, and making the data useable will allow them to do this.

They said they would send me a copy of their online accounts in a spreadsheet as this is what they have to convert into a pdf before they put it on their website.

They refused to give any exports of their reports of which there are a number describing in the training notes.

I offered our support for publishing their spending data and they agreed that keeping the lines of communication open with OKFN would be useful to us
both.

They were keen to look up Chris’ blog post and the open data standards and said they would make sure they published their data following those guidelines.


Inputs to the Office of National Statistics

July 9, 2010

I’ve mentioned before that I’m interested in the data the Office of National Statistics (ONS) use for their reports. I mean the raw data they use. Through looking at local authority finances I’ve been able to add more to my picture:

The acronyms stand for:

  • The CSDB (central shared database) does not use an ORACLE database. I have some details about the CSDB, like its size and how frequently it is backed up. I requested the CSDB training materials but that request has been delayed. I’ve also, cheekily, requested a database dump of it. That request has also been delayed!
  • The CORD system (Centralised ONS Repository of Data) used for some parts of the annual production process in ELS does use an ORACLE database. I have the CORD schema, and have requested the CORD training materials, but this request has been delayed. Can you see a theme in how the ONS are handling freedom of information requests?

The acronyms for the input reports can be found in my previous post about local authority finances.

It would be my ideal if, in the future, the ONS list and publish the raw data they aggregate in their reports. I’ve written before, but it bears repeating, that if the publication of COINS taught us anything, it should be that within a week of publishing data the government can expect an impressive range of free, and in many cases open source, products to display the data. Surely this is good for public understanding and the public purse?

UPDATE: this document(PDF) is quite useful for insights into ONS inputs.


Reports on local authority finances.

July 9, 2010

Each local authority has a database to store its finances. OK, sometimes it is more complex than that; sometimes authorities share a database and there is often sharing of data between databases. But for the purposes of this blog post, it is fair to say authorities store their finances in a database. I believe that this database has the most detailed data that exists about the authority’s accounts.

I’ve been plugging away to getting a local authority to publish their finances database, with reasonable redactions of personal details and individual payment details, of course. But so far the resistance to such a publication is high — as you would probably expect — mostly because of the work involved. But I’m asking for this information because authorities will be required to publish spending above £500 by January 2011, so if they are going to publish over £500 then why not go the whole way and do it all? Optimistic, hey?

As things aren’t moving so fast with that line of enquiry, I’ve been looking into all the different requirements for Local Authorities to report their finances. The conclusion is that there is a lot of reporting going on, and I haven’t found out about what CIPFA require or reached a point where I know all the requirements clearly. But here’s what I’ve found out so far:

The acronyms stand for:

As you can see, local authority finances are quite something, and there’s a fair bit of work to be done to figure out what it all means in a way your average man on the street will appreciate. But I hope it is a start for people skilled in communicating complex ideas in a meaningful way, to investigate more, and maybe to ask for the raw data from these reports before it is aggregated and published.


Nice visualisation of the Finance Bill

July 8, 2010

I’m really impressed at Parliament.uk as a website to aid public understanding. Take for example this page describing the Finance Bill 2010-11:

Aside from the horrible trick of links straight into PDFs, this website is really helpful for aiding my understanding. Very pleasing.


A big chunk of COINS was not published.

July 6, 2010

When I saw the COINS data that was published at the beginning of June, I suspected there was something missing. I had been reading about the Whole of Government Accounts (WGA) exercise, which I have documented in the previous post. So I was expecting local council assets and accruals data of the sort that is captured in the L-packs and central government spending as captured each year in the C-packs. But is wasn’t there.

I conducted some more investigation, speaking to the team at the Whole Of Government accounts. There team is really quite small — only two people in Communities and Local Government WGA team and five or six people in the Treasury — but they do an *amazing* job of documenting all public assets and accruals. What is more, they have been running it every year for 10 years, each year gathering a detailed picture of local authorities financial health.

Anyway, based on my existing knowledge and my conversations with the WGA team and others, I can now confidently confirm the WGA is completely absent from the COINS data that was released. This means there is no reporting of local authority’s spending in COINS. A report from the WGA is planned spring next year. But I believe this will be at a very high level of detail — the sum of the whole government’s assets and accurals, not the details of individual authorities and departments.

Anyway, I can now confidently confirm the WGA is completely absent from the COINS data released. I have requested the 2008/2009 WGA data, with the Department of Health and the Department of Defence data removed, as these two departments failed the Audit.

Now I’ll wait to see what happens.

UPDATE: See the reply to the request for the Whole of Government Accounts 2008/09.


The whole of government accounts: an exercise in elimination

June 25, 2010

When the large sample of COINS data was published on the 4th June it was accompanied by a guide to the data. The guide is very useful, but one thing it doesn’t explain in very much detail is where the COINS data comes from. The guide lists the inputs:

COINS – the Combined On-line Information System – is used by the Treasury to collect financial data from across the public sector to support fiscal management, the production of Parliamentary Supply Estimates and public expenditure statistics, the preparation of Whole of Government Accounts (WGA) and to meet data requirements of the Office for National Statistics (ONS).

But in what form are each of these different types of financial data  entered into COINS? To answer this question for the Whole Of Government Accounts (WGA) data, I’ve been looking more closely at the data gathered for this exercise.

After reading the WGA materials on the Treasury and the Communities and Local Government web pages, and chatting to some very helpful members of the WGA team, my understanding of the WGA exercise is that it identifies exchanges of funds between public bodies. These exchanges include the flow of funds from Central Government bodies down to Local Authorities and all the exchanges of funds between departments. When the transactions between public bodies are identified, the WGA exercise makes some adjustments to avoid double counting the money. So, if body A gives money to body B, then WGA would be responsible for subtracting the amount body B received from body A’s total.

As we know the COINS data is made up of spending or income records for each department. In these department records there is a Counter Party ID (CPID),  if that identifies another government department that means that some funds have been exchanged between the two departments.

There are scripts used on the COINS data to look for  eliminations using the CPID code.  which has the code in every department’s spending, if money of the public body money was exchanged with and the WGA team perform lots of checks on this. You can see this process happening in the adjustments table in COINS.

There are two guides to the WGA, one for  local authorities and the other for central government departments.

Central Government Accounts

The process of WGA for Central Government departments is simply that each central government department is required to fill in a C-Pack once a year, which is a spreadsheet constructed by the WGA team.

Point2.4. of the Guide for the C-Pack (PDF) says:

The key deliverable is the C-Pack, and the upload of Resource Accounts data and CPID data into COINS.

Local Authority Accounts

The WGA process for Local Authorities is a slightly different exercise. The Local Authority is asked by WGA to fill in an L-Pack once a year.  If you follow that link to the L-Pack excel spreadsheet that the Communities and Local Government branch of WGA prepare, they you will see that it is quite a complex looking creature. I’m going on a training course to understand it better, but I do know that the results of every local authority filling out this form amounts to quite a significant documentation of public spending and income.

In fact the WAG guide for local authorities states:

Local government controls over 50% of public sector fixed assets, accounts for about 25% of net public expenditure and represents 10% of UK GDP.

Now, here is the interesting part. The Local Authority spending and income that is recorded in the L-Pack is not in the COINS data that was published recently.

Now, I thought this missing detail in COINS might be because the WGA would be published separately.

There is a WGA report expected in spring 2011, but on further investigation it transpires that the level of detail will be the same as company accounts. We will get some extra details in this report, for example spending on PFIs will be included for the first time. But essentially we will miss out on all the lovely detail from the L-packs and C-packs.

Conclusion

The WAG is an exercise in eliminating excess data that clouds the picture of public spending and income.  The WAG team’s work seem to keep process of reporting spending and income more manageable. This is completely understandable. But on the other hand it would be great to have this detail of exchanges of funds so we can understand public spending as it really is.

UPDATE: I now understand the auditing going on it Whole Of Government Accounts better.

Auditing, I believe, means matching up buyers and providers:

A perfect match is:
Barnet Council purchases £5.5 m from Enfield Council.
Enfield Concil sales £5.5 m to Barnet Council.

The COINS scripts would eliminate this to zero as perfect match.

Another example:
Barnet Council purchases £5.0 m from Enfield Council.
Enfield Council sales £5.5 m to Barnet Council.
COINS would eliminate 5.0m and and put 0.5M into suspense.
The the suspense needs to be investigated more to see where the mistake is.
This investigation is the job of the Whole of Government’s Account team.
You can set a tolerance in COINS, which is the extent of the difference between two accounts it will put into suspense. The tolerance was set to 5.0m for 2008/09 accounts. I will be set to 1.0m for
2009/10 accounts.


COINS: the ins and outs

June 15, 2010

It’s been over a week, now, since the government spending recorded in the COINS database was free for everyone to see.

Before COINS was published I’d been looking into the different data gathering processes that ended with the data being stored in COINS. So you can imagine that having the actual COINS data is the final piece in the puzzle to understanding those processes. In short: it is all starting to make more sense.

In the next two blog posts here, I will describe two important processes that involve gathering data to be stored in COINS.

I’ve drawn a rough sketch of the processes in this diagram (click on the image to see more detail):

The two processes are:

  1. The Whole Of Government Accounts data gathering exercise which is performed by a branch of the Treasury. I’ve discovered this is an exercise in aggregating local spending, and avoiding double counting of spending in central government departments.
  2. The estimates and budgeting process which determines the resources departments are allocated.

I’m also writing a much more detailed guide to COINS, after more detailed talks with the very helpful people at HM Treasury, that is published on the Open Knowledge Foundation blog and will be published shortly on data.gov.uk.