Values for many of the datasets used in TIFF do not become available until later in the year. This means that the data published here for 2017 are often only a first estimate, with revised data being published next year and possibly a further update the following year. For this reason this publication contains updated data for 2015 and 2016.
A wide-ranging review of TIFF methodology was carried out during 2012 and 2013. Thereafter more minor methodology improvements have been identified, due to improved data sources, estimations or modelling. Where such changes have been made, back series are calculated where necessary. A document explaining the latest revisions to previously published TIFF estimates is available online at the following link
The main methodological changes in this year's publication relate to the inclusion of stockfeed, roughage, and camelids in the outputs, the inclusion of stockfeed and protein crops in the feed bill. Data changes affecting earlier years occurred in revised cattle data and in the ONS index used to separate finance costs into interest and FISIM.
Effect of changes
For the most recent years, compared to the results published in January 2017, TIFF has been revised downwards £14 million (two per cent) for 2015, and down £77 million (11 per cent) for 2016. This 11 per cent change was due to a two per cent revision in the income estimate and a 0.1 per cent revision in the costs estimate.
This year's revised trend in TIFF between 2015 and 2016 shows an increase of £33 million (five per cent) before inflation, a smaller increase than that published in January 2017 (which initially estimated a 15 per cent increase before inflation).
The largest revisions (in numerical rather than percentage terms) between 2016 initial estimates published last year and 2016 second estimates published here, which result in the £77 million change in estimate, are in
i. income from non-agricultural activities, revised down £89 million. This is due to actual expenditure reported in the Farm Business Survey being quite different from initial estimates based on a range of price indices and other data. The value of income from non-agricultural activities, which only includes those activities that are non-separable from the farm activity, is difficult to measure consistently, and therefore difficult to estimate in advance.
ii. income from other crops, revised upwards £39 million. This was due to the value of roughages and protein crops being included for the first time in the calculation.
iii. subsidies, revised downwards £24 million as based on actual claims rather than initial budgets.