Police costs methodology
Published 24 November 2025
Applies to England and Wales
Executive summary
This technical note sets out the methodology for producing updated police costs using data from the Police Activity Survey (PAS). The costs reported are in the form of crime response workforce[footnote 1] unit costs, and are presented by crime type below:
Table 1: Crime response workforce unit costs
| Crime incident type | Crime response workforce unit cost (2022 to 2023, on police recorded crime (PRC) basis) |
|---|---|
| Arson | £1,130 |
| Bicycle theft | £150 |
| Criminal damage | £270 |
| Death or serious injury - unlawful driving | £60,620 |
| Domestic burglary | £1,250 |
| Homicide | £577,530 |
| Miscellaneous crimes | £1,550 |
| Non-domestic burglary | £870 |
| Other sexual offences | £3,100 |
| Other theft offences | £330 |
| Possession of drugs | £1,290 |
| Possession of weapons offences | £2,320 |
| Public order offences | £280 |
| Rape | £5,110 |
| Robbery | £2,100 |
| Shoplifting | £330 |
| Stalking and harassment | £290 |
| Theft from the person | £880 |
| Trafficking of drugs | £6,850 |
| Vehicle offences | £640 |
| Violence with injury | £1,260 |
| Violence without injury | £470 |
1. Introduction
1.1 Background to PAS
The PAS was conducted during the first week of February 2023 with 35 out of the 43 territorial police forces in England and Wales taking part. The survey asked frontline police officers and staff to record their activities each day of that week. All activity recorded within PAS falls into one of the following 3 high level categories:
- Specific crime incident activity
- Specific public safety and welfare (PSW) incident activity
- General crime/PSW activity not related to a specific crime/PSW incident[footnote 2]
Within each of these 3 high-level categories are different activities that the police officers could select. See the Annex of the PAS publication for the full list of PAS activities under each of the 3 high-level categories as well as a full description of the survey results and approach.
1.2 Generated cost estimates and their associated definitions/interpretations
To calculate these costs, total police budget is split into 4 separate categories:
- Crime response workforce costs: These reflect the average cost of police responding to a crime based on the time officers spend handling the incident.
- Crime anticipation costs: These refer to the wider costs involved in maintaining the infrastructure that enable the police to effectively respond to criminal activity including police stations, vehicles, technology, and training programmes.
- PSW response workforce costs: These reflect the average cost of police responding to a PSW incident based on the time officers spend handling the incident.
- PSW anticipation costs: These refer to the wider costs involved in maintaining the infrastructure that enable the police to effectively respond to PSW incidents.
This publication focuses on the methodology to estimate crime response workforce costs (1). Crime anticipation costs (2) and PSW costs (3 and 4) are outside the scope of this publication, so cost estimates for these are not provided.
1.3 Methodology summary
The methodology to estimate crime response workforce costs is summarised below. Each of these steps is then explained in turn in this note.
- Obtain total workforce costs for the PAS activities.
- Apportion out crime response costs from the general crime/PSW activity costs.
- Allocate the general crime/PSW activity crime response costs to each crime incident type.
- Calculate the total crime response workforce cost for each crime type by adding its total workforce cost to the portion of general crime/PSW activity crime response costs allocated to it.
- Divide total crime response workforce costs by PRC volumes for each crime type.
2. Obtaining total workforce costs for the PAS activities
The first step is to obtain total workforce costs using PAS data. The PAS gives total hours spent by police officers and staff on different activities during the survey week for the 35 forces [footnote 3]. As part of PAS, forces also provided supplementary wage data, broken down by rank and department.
PAS activity data can be combined with this wage data to obtain the total workforce cost for the different activities across the 35 forces. These costs, which cover the survey week, are then scaled to a year.
Table 2 below shows the annual workforce costs for specific crime incident activities.
Table 2: Specific crime incident activities – total workforce costs
| Crime incident type | Annual workforce cost - 35 Forces |
|---|---|
| Arson | £21.4 million |
| Bicycle theft | £5.8 million |
| Criminal damage | £86.7 million |
| Death or serious injury cause by unlawful driving | £45.5 million |
| Domestic burglary | £141.0 million |
| Fraud offences to 2012 to 2013 | £101.8 million |
| Homicide | £193.2 million |
| Miscellaneous crimes | £119.7 million |
| Non-domestic burglary | £42.7 million |
| Other sexual offences | £42.7 million |
| Other theft offences | £88.0 million |
| Possession of drugs | £93.6 million |
| Possession of weapons offences | £85.8 million |
| Public order offences | £109.1 million |
| Rape | £218.9 million |
| Robbery | £68.7 million |
| Shoplifting | £69.5 million |
| Stalking and harassment | £139.9 million |
| Theft from the person | £28.1 million |
| Trafficking of drugs | £174.8 million |
| Vehicle offences | £139.5 million |
| Violence with injury | £452.2 million |
| Violence without injury | £249.4 million |
| Total 35-force annual workforce cost of specific crime incident activities | £2917.5 million |
3. Apportioning out crime response costs from the general crime/PSW activities
Table 2 presents the direct workforce costs associated with each crime type; however, it does not consider that, in the process of the police responding to a specific crime, they may also have to undertake activity that is defined as a general crime/PSW activity in PAS.
The next step therefore is to determine the proportion of the general crime/PSW activity costs that should be attributed to crime response. Incorporating these costs is necessary to obtain fully-formed crime response costs, since general crime/PSW activities such as ‘custody duties’ form part of the police’s response to crime. The apportioning approach involves evaluating each general crime/PSW activity according to 2 criteria:
- Whether the cost is relevant only to crime, or relevant to both crime and PSW. If it is relevant to only crime (for example, custody duties), then the cost is allocated fully to crime. If it is relevant to both crime and PSW (for example, training), then the cost is allocated based on the relative workforce costs in PAS for crime compared to PSW (as shown in Figure 1 below). This gives a weighting of 74% for crime. This is the proportion allocated to crime when costs are determined to be relevant to both crime and PSW.
- Whether it is a response cost or an anticipation cost. Response costs are those that are incurred when a crime happens, and the police must handle an incident. Anticipation costs are the wider costs involved in maintaining the infrastructure that enable the police to effectively respond to criminal activity including police stations, vehicles, technology, and training programmes. This distinction determines whether the cost is allocated to crime response or crime anticipation.
Figure 1: Calculating the weighting for allocation to crime
The apportioning approach can be explained through an example
- ‘assisting ambulance’ is a general crime/PSW activity with a total workforce cost of £12.1 million
- this is a cost relevant to both crime and PSW because an ambulance could be necessary for both crime and PSW incidents; the cost is therefore weighted, with 74% going to crime (£8.9 million)
- it is also deemed to be a response cost, as it is an activity and cost that is only incurred when an incident happens; hence this 74% (£8.9 million) gets allocated to a crime response ‘pot’
Following this approach for all the general crime/PSW activities creates a ‘pot’ of additional crime response workforce costs of £544.7 million.
4. Method for allocating the general crime/PSW activity crime response costs to individual crime incident types
The next step is to allocate this additional crime response workforce cost of £544.7 million to the specific crime incident types. This is done based on the share that each crime type makes up of the total crime workforce cost.
For example, robbery makes up 2.36% of the total crime workforce costs, hence it gets allocated 2.36% of the £544.7 million, which equals £12.8 million.
Table 3: Total crime response workforce costs
| Crime incident type | Annual workforce cost - 35 Forces | Proportion of total crime workforce cost | Allocation of wider activity cost | Total crime response workforce costs |
|---|---|---|---|---|
| Arson | £21.4 million | 0.73% | £4.0 million | £25.4 million |
| Bicycle theft | £5.8 million | 0.20% | £1.1 million | £6.9 million |
| Criminal damage | £86.7 million | 2.97% | £16.2 million | £102.9 million |
| Death or serious injury cause by unlawful driving | £45.5 million | 1.56% | £8.5 million | £54.0 million |
| Domestic burglary | £141.0 million | 4.83% | £26.3 million | £167.4 million |
| Fraud offences to 2012/13 | £101.8 million | 3.49% | £19.0 million | £120.8 million |
| Homicide | £193.2 million | 6.62% | £36.1 million | £229.3 million |
| Miscellaneous crimes | £119.7 million | 4.10% | £22.4 million | £142.1 million |
| Non-domestic burglary | £42.7 million | 1.46% | £8.0 million | £50.7 million |
| Other sexual offences | £42.7 million | 8.30% | £45.2 million | £287.3 million |
| Other theft offences | £88.0 million | 3.02% | £16.4 million | £104.4 million |
| Possession of drugs | £93.6 million | 3.21% | £17.5 million | £111.0 million |
| Possession of weapons offences | £85.8 million | 2.94% | £16.0 million | £101.8 million |
| Public order offences | £109.1 million | 3.74% | £20.4 million | £129.5 million |
| Rape | £218.9 million | 7.50% | £40.9 million | £259.8 million |
| Robbery | £68.7 million | 2.36% | £12.8 million | £81.6 million |
| Shoplifting | £69.5 million | 2.38% | £13.0 million | £82.5 million |
| Stalking and harassment | £139.9 million | 4.80% | £26.1 million | £166.1 million |
| Theft from the person | £28.1 million | 0.96% | £5.2 million | £33.3 million |
| Trafficking of drugs | £174.8 million | 5.99% | £32.6 million | £207.5 million |
| Vehicle offences | £139.5 million | 4.78% | £26.0 million | £165.6 million |
| Violence with injury | £452.2 million | 15.50% | £84.4 million | £536.6 million |
| Violence without injury | £249.4 million | 8.55% | £46.6 million | £295.9 million |
| Total | £2917.5 million | 100.00% | £544.7 million | £3462.3 million |
5. Obtaining the crime response workforce unit costs
The final step of the methodology is to take the ‘Total crime response workforce costs’ across all 35 forces for each of the different crime incident types and divide them by their volumes across all 35 forces for 2022 to 2023. This produces unit costs.
For the crime volumes, Crime outcomes in England and Wales 2022 to 2023 is used.
Table 4: Crime response workforce unit costs
| Crime incident type | Total crime response workforce costs (2022 to 2023) | 35-force police recorded crime volumes (2022 to 2023) | Crime response workforce unit cost (2022 to 2023, on PRC basis) |
|---|---|---|---|
| Arson | £25.4 million | 22,400 | £1,130 |
| Bicycle theft | £6.9 million | 47,472 | £150 |
| Criminal damage | £102.9 million | 381,244 | £270 |
| Death or serious injury cause by unlawful driving | £54.0 million | 891 | £60,620 |
| Domestic burglary | £167.4 million | 133,438 | £1,250 |
| Fraud offences to 2012/13[footnote 4] | £120.8 million | N/A | N/A |
| Homicide | £229.3 million | 397 | £577,530 |
| Miscellaneous crimes | £142.1 million | 91,890 | £1,550 |
| Non-domestic burglary | £50.7 million | 57,997 | £870 |
| Other sexual offences | £287.3 million | 92,791 | £3,100 |
| Other theft offences | £104.4 million | 320,515 | £330 |
| Possession of drugs | £111.0 million | 85,855 | £1,290 |
| Possession of weapons offences | £101.8 million | 43,935 | £2,320 |
| Public order offences | £129.5 million | 461,918 | £280 |
| Rape | £259.8 million | 50,832 | £5,110 |
| Robbery | £81.6 million | 38,934 | £2,100 |
| Shoplifting | £82.5 million | 249,895 | £330 |
| Stalking and harassment | £166.1 million | 564,955 | £290 |
| Theft from the person | £33.3 million | 37,917 | £880 |
| Trafficking of drugs | £207.5 million | 30,291 | £6,850 |
| Vehicle offences | £165.6 million | 260,719 | £640 |
| Violence with injury | £536.6 million | 425,287 | £1,260 |
| Violence without injury | £295.9 million | 625,731 | £470 |
6. Quartiles
To reflect the uncertainty in these estimates, lower and upper quartiles for the unit costs of the 35 forces have been included. Differences in unit costs between forces reflects both genuine regional variation (operational differences between forces) as well as random variation because of the survey approach. Before calculating the quartiles, any outliers (driven by low volumes) were excluded.
Given the survey was conducted over a week and response rates varied across forces, there is an inherent amount of uncertainty within all estimates produced from the Police Activity Survey. This is more pronounced for some of the low volume crimes such as homicide, where the force may not have had any homicides committed in the survey period or have low volumes across the year. Therefore, these estimates should be treated with caution and sensitivity analysis should be conducted wherever they are used to ensure decisions are not heavily influenced by uncertain estimates.
Quartiles are provided which express some of the uncertainty within the estimates. These are illustrative and do not represent confidence intervals or the only bounds within which it would take the police to respond to a crime. Instead, they are intended to give an idea of the level of uncertainty involved with the estimates, a broad range of potential costs in response, and the level of variation between crime types, with low volume crime types inherently having higher uncertainty.
It is not possible to produce statistical confidence intervals because the survey does not track individual crimes. Instead, unit costs for each force are calculated based on the annual number of crimes for that force and assuming the survey week is representative of time spent across the year. This is an oversimplification, and an assumption which has known limitations; however, it represents a practical option given data constraints.
Table 5: Lower and upper quartiles[footnote 5]
| Crime incident type | Crime response workforce unit cost (2022 to 2023, on PRC basis) | Lower quartile | Upper quartile |
|---|---|---|---|
| Homicide | £577,530 | £330,770 | £912,440 |
| Death or serious injury - unlawful driving | £60,620 | £37,900 | £117,480 |
| Trafficking of drugs | £6,850 | £5,160 | £12,530 |
| Rape | £5,110 | £3,660 | £7,440 |
| Other sexual offences | £3,100 | £2,580 | £4,120 |
| Possession of weapons offences | £2,320 | £1,430 | £2,930 |
| Robbery | £2,100 | £1,720 | £3,720 |
| Miscellaneous crimes | £1,550 | £1,150 | £2,050 |
| Possession of drugs | £1,290 | £1,080 | £1,860 |
| Violence with injury | £1,260 | £1,100 | £1,580 |
| Domestic burglary | £1,250 | £870 | £1,800 |
| Arson | £1,130 | £520 | £1,870 |
| Theft from the person | £880 | £770 | £1,720 |
| Non-domestic burglary | £870 | £750 | £1,230 |
| Vehicle offences | £640 | £530 | £1,040 |
| Violence without injury | £470 | £420 | £610 |
| Other theft offences | £330 | £230 | £420 |
| Shoplifting | £330 | £250 | £420 |
| Stalking and harassment | £290 | £230 | £400 |
| Public order offences | £280 | £250 | £340 |
| Criminal damage | £270 | £220 | £330 |
| Bicycle theft | £150 | £80 | £180 |
7. Use in appraisal
The crime response workforce unit costs reflect the opportunity costs saved by reducing crime volumes (the officer’s time, and associated wages, that could be spent on other tasks or duties). As such, these costs can be used in policy appraisal to assess the potential benefits of crime reduction efforts.
The response workforce unit costs are only monetised using police wages, and do not consider anticipation costs. When policies have long-term effects on crime volumes over an extended horizon, it may be important to consider these anticipation costs as well as the response unit cost in appraisal. Anticipation costs might involve investments in infrastructure, vehicles, technology or other expenses that could need to be increased or decreased if crime levels changed. These costs should be considered in some scenarios, where there are large changes in crime levels, to provide a more comprehensive evaluation of the policy’s impact. Please contact PEPFAteam@homeoffice.gov.uk in such contexts, and we can advise on how to approach this.
Additionally, since these costs are survey-based, without all 43 forces partaking, there is inherently a degree of uncertainty with these costs. Costs may vary between forces and over time, so it is recommended that the estimates are used with this in mind and that practitioners consult the quartiles listed in Table 5 above for broader context as to the degree of the uncertainty (for use in sensitivity analysis, for example).
8. Limitations
A clear limitation is that only 35 of the 43 police forces in England and Wales completed the survey. Notably, the Metropolitan Police Service (MPS) - which recorded around 16% of all crimes reported by the 43 forces in England and Wales for 2022 to 2023 – did not partake. This omission means a significant portion of crime volumes is not reflected in the PAS. The MPS is uniquely large and operates in a highly urbanised environment, likely leading to different unit costs compared to smaller forces. Additionally, its London location results in higher wages, further increasing unit costs – another factor not captured in PAS.
To estimate unit costs for different police activities, total costs recorded during the survey week are divided by average weekly volumes, derived from police recorded crime data. This approach assumes that the survey week is representative of typical weekly activity across the year. While this assumption is generally reliable for high-volume crimes, it presents challenges for low-volume crimes – most notably homicide and death or serious injury caused by unlawful driving. Because the PAS methodology is activity-based rather than incident-based, we do not know how many individual cases the recorded costs relate to. For low-frequency crimes, the number of cases being actively worked on during the survey week may differ significantly from the average, leading to greater uncertainty in these cost estimates.
PAS itself was only conducted over a one-week period in February 2023, therefore it won’t capture the end-to-end process of all incidents. It is therefore assumed that across the entries this is balanced out. For example, activities linked to incidents which took place before the survey began (capturing activities at the end of the process) balance out those which took place during the survey period but do not reach an end point.
Workforce cost estimates do not capture the role of Regional and Organised Crime Units (ROCUs) in investigating some crime incidents.
-
Within PAS, ‘workforce’ is used to refer to officers and staff (excluding corporate functions). ↩
-
In the survey, this was referred to as ‘Activities not related to a specific crime or non-crime incident’. For ease, it will be referred to henceforth in the rest of the note as ‘General crime/PSW activities’. ↩
-
The results are weighted to represent the entire force. See the published report (link) for further information. ↩
-
No unit cost is calculated for fraud as it was not possible to obtain fraud volume estimates for the 35 forces. ↩
-
All costs are rounded to the nearest £10. ↩