Guidance

Local authority planned expenditure benchmarking tables 2021 to 2022: additional information

Updated 6 July 2022

Applies to England

1. Introduction

This document provides local authorities with additional information that can be used to understand how to use the planned expenditure benchmarking tables. These tables allow local authorities to compare their planned expenditure from the Department for Education financial data collection budget statements (also known as section 251 budget statements) for the 2021 to 2022 financial year against the national averages and their statistical neighbours.

2. How to view information on local authorities with similar characteristics

For all benchmarking tables, you can choose to view:

• all local authorities
• local authorities in a particular region or type of authority, such as metropolitan, unitary, London or upper tier (counties)
• your own local authority along with its statistical neighbours
• your own local authority with a manual choice of up to 10 others

For the purposes of these benchmarking tables, City of London and Isles of Scilly have not been included due to their small size, while North Northamptonshire and West Northamptonshire have not been included in every variable, due to data limitations associated with the local authority boundary changes.

3. Statistical neighbours

Statistical neighbours are calculated to enable comparison across ‘similar’ local authorities. Further information on these statistical neighbours is available.

When you view your statistical neighbours within a benchmarking table, they are ordered according to their ‘closeness’ (such as degree of similarity), with the closest at the top of the list.

4. Why there may be differences in funding across statistical neighbours

Statistical neighbours provide a basis for comparison between local authorities with similarities over a broad range of educationally relevant characteristics; they are calculated according to a number of criteria relating to the effectiveness of local authorities and educational outcomes. These criteria are not necessarily all relevant for evaluating how similar local authorities are to each other in terms of characteristics that are relevant to funding.

5. Variation across local authorities in the expenditure per capita figures

Differences in the structure of education services between individual local authorities will result in variations in certain budget lines. For example, some local authorities maintain no sixth forms and this will be reflected in the related budget lines. Similarly, there are differences in the structure of special educational needs (SEN) provision and the relative use of maintained special schools, other authorities’ provision, non-maintained and independent schools.

All figures are rounded to the nearest pound so components may not sum exactly to totals. For categories where there is a very small amount of planned expenditure, the per capita figure may be less than £0.50 and hence rounded to zero. A zero per capita figure does not necessarily imply that nothing has been spent by the local authority in that category.

6. Why you might see a large percentage change year-on-year

There are a number of reasons why there might be a large percentage change in year-on-year spending. The s251 Budget data collection was suspended in 2020 due to coronavirus, so year-on-year comparisons compare 2021 to 2022 data and 2019 to 2020 data.

If the underlying cash amounts are very small, a relatively small cash increase or decrease between years could result in a large percentage change. The averages, minima and maxima noted at the top of each column, together with the values on the per pupil table, will give a general indication of the relative size of expenditure on these budget items in 2021 to 2022. The raw budget data for 2021 to 2022 will be published in September. Large year-on-year changes could also be a result of accounting changes, or could also be a result of a local authority delegating increased levels of expenditure to its schools.

7. Why there can be a significant difference between the mean and the median and what this means

Throughout the benchmarking tables, both the mean and median are given for each line. The median is less sensitive to extreme values than the mean, and is therefore often used for benchmarking.

To show this, we consider an example of 8 local authorities with the following expenditure on three different lines:

Local Authority name	Line 1	Line 2	Line 3
Example 1	10	0	30
Example 2	20	0	30
Example 3	20	0	30
Example 4	30	0	30
Example 5	30	0	30
Example 6	40	70	30
Example 7	40	80	30
Example 8	50	90	270
Mean	30	30	60
Median	30	0	30

Table 1: Example calculation of mean and median

The mean is calculated by adding together the entries in each line and then dividing by the number of entries there are (in this case 8).

The median is calculated by putting the entries in order and then finding the ‘middle’ entry. In this case, since there are eight local authorities, the ‘middle’ is halfway between example 4 and example 5, so the median is halfway between the value of the fourth largest and fifth largest entries.

The mean is the same in line 1 and line 2, even though the data looks very different. The median is the same in line 1 and line 3 even though again, these two sets of data are very different. The mean is equal to the median in line 1, but they are quite different in lines 2 and 3. Giving both the mean and median tells us more about the distribution of the data than just giving one of these on its own.

As in line 2 of the example above, for some columns the median will be zero despite a large number of local authorities actually having significant expenditure. This is mathematically correct; if more than half the local authorities have no expenditure in the category, then the middle value of expenditure when values are arranged in ascending order will be zero.