COINS Frequently Asked Questions (FAQs)

28 November 2012

Accessing the data
Structure of the data
Data contents
Data updates

Accessing the data

How do I download the file? How long should it take?

Double click on the file to download it. With a fast broadband connection (8mbps), the compressed 70 MB 'fact table' should take about 10 minutes to download. The compressed ‘adjustment table' is approximately 40MB, and this will take approximately 5 minutes to download. Both these files have been compressed using ZIP archival and will decompress to sizes of about 5GB and 0.5GB respectively. The Whole of Government accounts 2010-11 data files are smaller, and have been compressed into a single file (2012COINS_extract_2010-11.zip). This file is approximately 22MB in size and should take less than 5 minutes to download with a fast broadband connect. When uncompressed, two files will be visible – the “fact table” and the “adjustment table”, with approximate sizes of 0.9GB and 0.1GB respectively.

Why can't I read the file?

The data have been provided in txt file format (but .csv format for 2011). The structure of the data is similar to a csv (comma separated variable) file, with a string of characters for each row of data from the database, and with each field separated by "|"in the 2011 release files, and “@” in the 2010 release files.

In order to read the data, they will need to be uploaded into appropriate database software (these files are not compatible with MS Access databases).

Why are there two files? Do I need to use these files together?

The 'fact table' contains the aggregate position for each department for each of the data streams. The 'adjustment contains some, but not all, of the data underlying the aggregates in the 'fact table'. Specifically, the 'adjustment table' contains all of the adjustments underlying the Plans data in the 'fact table', but for Outturn data it only contains those adjustments that are made to the final full year position. Updates to in-year estimates are loaded direct into the 'fact table'. The 'adjustment table' provides an audit trail behind the adjustments reflected in the 'fact table'. The following file name approaches have been used for the differing data releases: The following file name approaches have been used for the differing data releases:

2010: fact table extract 200x xx.txt and fact table extract 200x xx.txt
2011: 2011 fact table extract 20xx xx.csv and 2011 adjustment table extract 20xx xx. csv, and for WGA data 2011_adjustment_table_extract_2009_10_WGA.csv
2012: 2012facttableresult_20xx-yy.txt; and 2012adjustmenttableresult_20xx-yy.txt and for WGA 2010-11 data files: 2012facttableresult_wga_2010-11.txt and 2012adjustmenttableresult_wga2010-11.txt

Structure of the data

How are the data structured?

COINS uses seven dimensions (version, organisation, time, variable, reporting, segment and counter party ID). A number of attributes are reported against each of these dimensions to reflect particular characteristics of the data in Plans, Outturn, Forecast Outturn and Audited Outturn data. However Audited Outturn (used for Whole of Government Accounts - WGA), does not make use of the segment dimension.
These attributes are used to construct the reports that the Treasury uses to publish aggregate government spending. "Understanding the COINS data" (available on the right-hand side of the page) provides a description of the attributes.
At the heart of the data structure are two key fields:

account codes (within the variable dimension) are used to record the economic nature of expenditure (e.g. pay, procurement etc);
Programme objects (within the segment dimension) are used to record the activity (a similar concept to a cost centre). Please note that segment data are NOT used for Audited Outturn (WGA).

The Treasury sets the account code structure. Departments set the programme object structure best to reflect their activities, which at aggregate level is agreed with Select Committees and Parliament. There are a number of attributes within the variable and segment dimensions that allow combinations of account codes and/or programme objects to be 'grouped' where they share common characteristics.

What are the different columns for?

The columns are used to identify particular attributes of spending. "Understanding the COINS data" provides a description of the attributes. These attributes are combined in various different ways to feed into aggregates and/or reports published by the Treasury and ONS.
These include spending shown in the Budget report, Parliamentary Supply Estimates, Public Expenditure Statistical Analyses (PESA), the monthly Public Sector Finances statistical bulletin and the National Accounts.

How can I find out total spending for an individual department?

There are different ways in which 'total departmental spending' can be measured - the simplest is to look at spending within resource and capital budgets, which are set by the Treasury. Resource budgets comprise day to day expenditure (pay, purchase of goods and services etc), while capital budgets comprise spending on fixed assets and investments, as well as capital grants. "Understanding the COINS data" provides examples of how some of these totals can be identified.

Why don't all departments record spending against all categories?

Each department has its own finance system for recording its spending transactions. These transactions are aggregated and mapped to a standard list of accounts provided by the Treasury.

Departments can largely choose the level at which they map their spending to this list of accounts, as long as at predefined levels the aggregates are correct. This means that some departments enter their data into COINS at lower levels of aggregation than others, leading to differences across departments.

The pre-defined levels are set by the Treasury and match the aggregates used in publications. The Treasury quality assures the data only at this level. Mappings used for Audited Outturn are subject to audit review.

Data contents

What data do the files contain?

There are two files for each year containing data from the COINS database, including the separate WGA releases.

The 'fact table' contains data that summarise departments' actual spend (Outturn), their in-year estimates (Forecast Outturn) and planned expenditure for the year (Plans). It also contains statutory account information (Audited Outturn – used for WGA). It does not contain individual transactions - these are recorded in systems maintained by each department for their own purposes.

The 'adjustment table' includes adjustments to Plans, Outturn and Audited Outturn that have been uploaded by departments or made by Treasury. The aggregate of the Plans adjustments will match the data in the 'fact table' but the same is not true for Outturn as some data are input directly into the 'fact table' rather than being uploaded through the 'adjustment table'.

The files do not include data relating to national security or military operations. The ‘fact table' for 2009-10 released in June 2010 also excludes the final in-year return provided by departments in May 2010 as this was subject to National Statistics protocols.

What is the data used for?

The data in COINS are used for:

fiscal management (e.g. Budget and Pre-Budget reports);
operational publications (e.g. Supply Estimates);
statistical publications (e.g. Public Expenditure Statistical Analyses, ONS/Treasury Public Sector Finances statistical bulletin, the National Accounts, etc); and
Whole of Government Accounts.

Who puts the data into COINS?

Data are input into COINS by each central government department. Where an entity does not have on-line access to the system, the data provided by these entities are loaded centrally by the Treasury.

In addition, the Treasury loads onto COINS the local government spending data provided by certain Whitehall departments and the Devolved Administrations. The Treasury also loads data relating to central funds (such as the Reserve) and other aggregates (e.g. central government debt interest) that are required for publications.

Who is responsible for the data on COINS?

Each entity is responsible for its own data and retains the ownership of the data in the system. The Treasury performs quality assurance at the level at which the data are included within publications.

Will I be able to see how much each department spends on expenses such as taxis?

You will not be able to see how much money departments spend on taxis, or other specific items of expenditure. Departments aggregate their transactions and map these to a list of accounts maintained by the Treasury. Departments can largely choose the level at which they aggregate their data, as long as at predefined levels the aggregates are correct.

Will I be able to see information on spending by Arms Length Bodies (ALBs)?

Spending by individual ALBs will be included within the data supplied by parent departments, but cannot necessarily be identified separately – this depends on the level of detail at which departments have chosen to report and whether or not they have prepared separate resource accounts, or if accounts have been sub-consolidated prior to submission to COINS.

How does this data relate to departments' published resource accounts?

Departments' resource accounts are produced from departmental systems that record their individual transactions. Departments are required to update the Outturn data held in COINS to reflect the final position in the audited resource accounts, and to provide full resource accounts information, including balance sheet, for Audited Outturn (for WGA). Please note that the Outturn data held in COINS may subsequently be amended to reflect classification changes (changes to the categorisation of the spend) or to reflect machinery of government changes (such as the merging or creation of new government departments or the changes to arms length bodies).

Will I be able to tie the data back to publications made by the Treasury and the ONS?

While a number of Treasury and ONS publications use COINS data, it will not always be possible to replicate published information. COINS has the facility to record a version of the data that reflects the position as at a particular date and/or for a particular publication (snapshot).
These versions contain key fields which reference the latest structure of the data held in COINS. Changes to the structure can occur continually (e.g. to reflect classification changes) with the potential for amendments to the attributes associated with these key fields. In addition, not all the data in all publications are sourced from COINS. Consequently, it may not be possible exactly to recreate published figures.
Furthermore, the eliminations process used in the Whole of Government Accounts (WGA) is highly complex and subject to Intellectual Property Rights (IPR) protection and thus limits the possibility of recreating published figures.

Will I be able to add up the expenditure against each line to get to total government spend?

There are a number of reasons why adding up all lines of expenditure will not produce total government spending. Three reasons are as follows:

to maintain control and provide audit trails in the system, adjustments to planned expenditure can go through a number of states (draft, submitted, approved, etc). As a consequence the same data can appear multiple times so simply adding up the spending against each line will duplicate some items;
not all items used to calculate total government spend are sourced from COINS;
the use of eliminations and other complex processes in the Whole of Government Accounts.

However, total departmental spending within budgets can be derived from COINS. Guidance on how to do this is provided in "Understanding the COINS data" where an example is included.

Why do some data entries appear to be duplicated?

To maintain control and provide audit trails in the system, adjustments to planned expenditure can go through a number of states (draft, submitted, approved, rejected etc). The adjustment is captured when it is recorded on COINS as well as at each change in status. As a consequence the same data can appear multiple times.

Has any data been excluded and if so why?

Some of the data in COINS are protected and so have been excluded. This includes data relating to national security and military operations. Some of the data may also be covered by official statistics protocols, which means that they cannot be provided until the statistics to which the data relates have been published.

Why do some rows of data have n/a against them? Why are some fields blank?

Data on COINS do not necessarily take values against all attributes. This is indicated either by a blank field or by 'n/a'.

Why are there negative numbers on COINS?

There are a number of reasons why there will be negative numbers on COINS.

These include, but are not limited to, income generated by bodies (e.g. sales of goods, services and assets, receipts of licence fees, levies, dividends, loan repayments etc), funding contributions received from other bodies, the transfer of responsibility from one department to another (including Machinery of Government changes), or adjustments to figures previously reported.

Data updates

How often are the data in COINS updated?

The data in COINS are updated continuously. Departments have on-line access to the system and can make adjustments to the open parts of the system as required.

When will the data in these files be updated?

The Outturn, Plans and Forecast Outturn raw data will normally be updated in September to reflect the initial data release for the financial year just ended and changes to the four years’ data prior to that, reflecting the fact that only the five most recent historic years’ data are maintained on COINS (e.g.in September 2011, raw data for 2010-11 was released, along with updates to 2006-07 to 2009-10). Raw data relating to the Whole of Government Accounts (WGA) (the Audited Outturn data) will be released shortly after the Accounts for that year are published. For WGA 2009-10, published on 29 November 2011, the raw data were released in December 2011. We launched additional datasets in August 2010, which draw on the live data in COINS (i.e. they will exclude data that have been superseded). These datasets cover the main data streams, are available in Excel format and are interrogable. They are updated regularly, on a quarterly basis.

When will I be able to see data for earlier years?

Data going back more than five years from the current year are not maintained. We will be not be able to release data prior to 2005-06, as this was the earliest year for which the data were maintained in 2010. For data released in 2011, the earliest included year is 2006-07, and for the September 2012 release it is 2007-08.