FISIM in the National Accounts

Author: Andrew McCulloch

As it turned out, the National Accounts released last Thursday did not revise the figures for Gross Domestic Product (GDP) in the first quarter of 2012. The National Accounts estimate that GDP fell by 0.3 percent in the first quarter of the year placing the UK back in recession. Not good from an economic perspective but as a user of statistics it was encouraging to see that the previously published figures were not revised. One of the main things that users look for in statistics is consistency. If figures are frequently revised it undermines confidence in their reliability even when there may be good reasons for the changes that are made.

The next publication to look out for from the Office of National Statistics (ONS) is 2012's Blue Book on the 31st of July. The publication of the Blue Book, or the annual National Accounts, is much anticipated by all those interested in official statistics as it aims to give a comprehensive picture of what happened in the economy last year. The Blue Book also tends to be used to introduce significant changes in methodology. This year’s Blue Book is not going to contain any really big changes in methods, so today we’ll quickly look at changes made to the calculation of GDP in the 2008 Blue Book. The changes made to the calculation of GDP in 2008 concerned banking services and are interesting because they highlight the difficulty in measuring what might seem to be a fairly simple economic concept, particularly during periods of rapid economic change.

GDP is the main indicator used to measure the overall state of the economy and is defined as the value of all final goods and services produced in the economy over a certain time period. This includes the production of foreign companies located in the UK but excludes the production of UK companies located abroad. The value of GDP can be measured in several ways using the amount of spending, income or production in the economy. The notion here is that the flow of money through the economy is circular so one person's spending on goods and services is another person's income, so if we sum what has been spent it should equal income receipts which, in turn, should equal the value of what firms have produced.

The economy is constantly changing, however, and the nature of economic production today is quite different to what it was when the first set of National Accounts were produced in the 1950s. Of particular significance for the National Accounts has been the growth of the service sector and the decline in manufacturing production. In 1950, around 50 percent of employment was in industry and 50 percent in services but by 2000 the proportion of employment in industry had fallen to around 25 percent. The contribution of industry to GDP can be calculated relatively easily since most industrial output is sold in a market at a given price. The value of service sector production is however much more difficult to estimate and one reason for this is that households may use services which they do not pay for directly. One such component of services - banking services or Financial Intermediation Services Indirectly Measured (FISIM) - was included in the calculation of GDP for the first time in 2008 and also added to revised estimates of GDP in previous years.

We can look at the importance of the revisions by plotting the figures published in the Blue Books from 2005 to 2010 for money GDP in each year. In the figure below, each line is the GDP figures published in a given year. For example, the GDP series published in 2006, the pink line, gives GDP in 2000 as £953,227million but this has been revised upwards slightly in the 2007 figures, the turquoise line, to £958,931million. If there had been no revisions to the figures the lines would sit on top of one another and we would just have the one line. While the figure shows that, as noted previously, the size of the revisions to GDP have generally been quite small, with the lines lying quite close together, there is a clear jump in the figures published in 2008 (the dark blue line) which resulted from the inclusion of FISIM in GDP. The addition of FISIM to GDP resulted in an upwards revision to the previous figures of around 1.5 percent, or £15,000 to £20,000 million, not a small amount of money..

So how do we calculate the value of the production of FISIM?  Banks finance a large part of their activities by charging a higher interest rate on the loans they provide than on the deposits they take. Banks may also make money by trading but the traditional bank made money through the interest payments made by borrowers on loans. In order to place a value on banking services, the ONS use an estimate of the interest rate paid on loans and deposits above a risk free return, such as that on government bonds. As an example, if you had a loan of £1,000 for one year from your bank at an interest rate of 5 percent and the risk-free interest rate was 1 percent the ONS would add £40 (0.04 x 1,000) to GDP. This makes sense - the bank has been paid £40 for bearing the risk attached to not being repaid the loan and we include that as a measure of the bank's output. Excessive risk-taking by banks would also appear as a rise in GDP, however, even though, as we have seen recently, such behaviour is unlikely to lead to an increase in national output. Trying to value the contribution of banks to the economy is not without its problems.

The addition of the value of banking services to GDP in 2008 should mean, however, that the calculated value of GDP is now closer to the underlying real value of output in the economy. It is not that we found some loose change down the back of the sofa, nothing real has changed, but we now have a better estimate of the volume of production in the economy. GDP is better thought of as a concept rather than as a simply measurable fact and, as the economy changes, the way it is measured needs to be continually updated. As employees and consumers we might think of the value of economic production in terms of work done for pay or goods purchased in the shops but GDP aims to value the total output of the economy. It has even been argued that GDP should include the value of housework and I hope we are all doing our bit in that regard. The important point is not that published statistics are true but that we continue to find the GDP figures useful in understanding the economy and shaping policy.