Research and analysis

Economic and social costs of crime 2019 to 2020

Published 27 May 2026

Applies to England and Wales

Amendments to 2019 to 2020 cost estimates

This report presents legacy estimates of the economic and social costs of crime for the financial year 2019 to 2020 and should therefore be interpreted as a technical update of previous iterations of the Economic and social cost of crime series (most recently in 2018 which was based on 2015 to 2016 data). The analysis reflects the most robust data sources and methodologies available at the time of development.

Users should note, however, that some components may have since been superseded by subsequent analytical advances. For example, at the time of publication in May 2026, the Home Office’s Economic and social cost of fraud (2026) provides more comprehensive and up to date estimates for fraud related harms, and the Police Activity Survey (PAS) (2025) offers a more current basis for estimating police response costs than the older Activity Based Costing (ABC) data used here.

While this publication remains a valuable historical baseline for understanding 2019 to 2020 costs, users should consult more recent reports where they are available and relevant. The Economic and social cost of crime: Amendments to unit costs provides more up to date unit cost estimates.

Acknowledgements

This report and analysis were authored and conducted by James Haslam, Becky Acres, Ellie England, Will Phillips, Amy Robinson and Will Vaughan.

The authors would like to thank all the individuals who supported with the development, analytical quality assurance of this work, and the individuals who peer reviewed the final report.

For any questions relating to this report, please email EconomicandSocialCostsofCrime@homeoffice.gov.uk.

Executive summary

The economic and social costs of crime estimates help to develop an understanding of the wider costs and benefits associated with changes in crime trends. This is key to informing future policy decisions, providing the basis of value for money assessments. The framework applied in previous versions of the Economic and social costs of crime report has since been reworked to generate more detailed and targeted analysis regarding the impact to society of specific crime types. These include Home Office analysis on the Economic and Social Costs of Modern Slavery (Reed et al., 2018), Understanding Organised Crime (Fell et al., 2019), the Economic and Social Costs of Domestic Abuse (Oliver et al., 2019), the Economic and Social Costs of Contact Child Sexual Abuse (Radakin et al., 2021), and the Economic and Social Cost of Fire (Home Office, 2023).

This report uses existing crime and cost data to update previous analysis by the Home Office to estimate the economic and social costs of different offences in England and Wales. It concentrates on more serious victim-based offences; it does not estimate the economic and social costs of every type of crime. Costs have been estimated for crimes against individuals and for a limited number of business sectors. Crimes which are not committed against an individual victim – so-called crimes against society – are excluded from the analysis (for example, possession of drugs).

The report considers 3 main cost areas:

• costs in anticipation of crime, for example, the cost of burglar alarms
• costs as a consequence of crime, for example, the cost of property stolen and/or damaged
• costs in response to crime, for example, costs to the police and criminal justice system (CJS)

The total costs of crime in England and Wales in 2019 to 2020 are estimated to be approximately £61 billion for crimes against individuals and £17 billion for crimes against businesses (Table 1). Violent crimes (homicide, violence with injury, violence without injury, rape, other sexual offences and robbery) make up the largest proportion of the total costs of individual crime (just over 70% of the total cost) but only 35% of the total volume of these crimes. This is mainly due to the higher physical and emotional costs to the victims of violent offences. These costs are particularly high for crimes that are more likely to result in emotional injuries, such as rape and violence with injury. The offence with the highest estimated unit cost is homicide (£3.3 million). Rape has the highest estimated unit cost of non-fatal offences (£48,000).

Commercial theft makes up over 85% of all crimes against businesses, but accounts for just over 30% of the total estimated costs of crime against businesses (£5.5 billion), as each crime has a relatively lower impact on society. In contrast, commercial robbery, estimated to cost £6.8 billion in total, makes up around 40% of the costs of crime while only accounting for approximately 6% of all crimes against businesses.

For crimes against individuals, physical and emotional harms make up almost half (over 45%) of the total cost of crime. Therefore, most of the cost to society, from crimes against individuals, comes from the decrease in the quality of life that is experienced by the victims of crime as a result of their crime-induced physical and emotional injuries. The total cost of physical and emotional harms was highest for violent crimes, such as violence with injury and other sexual offences.

Value of property stolen or damaged is the cost category that contributes the most to the total cost of commercial crime. Almost 40% of the total cost of commercial crime is comprised of the value of property stolen or damaged in these crimes. This is largely driven by the crime of commercial theft: the total value of property lost or stolen in incidents of commercial theft is estimated to be £3.6 billion in 2019 to 2020.

The 2019 to 2020 estimates adopt a similar overall approach to earlier attempts to calculate the economic and social costs of crime (Heeks et al., 2018; Dubourg et al., 2005; Brand and Price, 2000) and include a number of improvements in data quality and cost estimations. These updated figures use improved data inputs for fraud and cybercrime, which previously used Crime Survey for England and Wales (CSEW) experimental statistics, and include more complete estimates of the costs of crimes against businesses (where 9 sectors are estimated rather than the previous 7, out of a total of 21 business sectors).

The total cost for each crime is driven by both the volume of the crime and the unit cost of the crime. Therefore, when analysing crime trends and the impact on society of crime reducing interventions, both the volume of crimes and the unit cost of the crime should be considered.

For example, the contribution of homicide to the total costs of crime is overwhelmingly driven by the crime’s unit cost: in 2019 to 2020, the 673 homicides accounted for less than 0.01 per cent of the volume of crimes against individuals, but this crime accounts for almost 4 per cent of total costs of crimes against individuals. This is because each homicide has an estimated unit cost of £3.3 million.

It is also the case that crimes with comparatively low unit costs can have a large impact on the total costs of crime to society if the volumes of these crimes are large. This is the case for commercial theft. Out of the 6 crimes against businesses that are costed in this report, commercial theft has the smallest unit cost, yet it has the second largest total cost. This high total cost is driven by the high volume of commercial theft incidents.

Table 1: Unit costs of crimes by cost area (2019 to 2020)

Note: Some components of the costs below have since been superseded by subsequent analytical advances. For use in analysis, the Economic and social cost of crime: Amendments to unit costs provides more up to date unit cost estimates, including police response and fraud costs.

Crime type	Anticipation unit cost	Consequence unit cost	Response unit cost	Total unit cost	Estimated total number of crimes (2019 to 2020)	Estimated total costs of crime (2019 to 2020 prices)
Individual						£60.8 billion
Homicide	£68,700	£2,637,190	£640,780	£3,346,680	673	£2.3 billion
Violence with injury	£280	£13,390	£2,050	£15,720	885,970	£13.9 billion
Violence without injury	£120	£5,480	£1,540	£7,150	720,710	£5.1 billion
Rape	£820	£38,100	£9,090	£48,000	164,070	£7.9 billion
Other sexual offences	£130	£5,610	£800	£6,540	1,881,180	£12.3 billion
Robbery	£480	£8,560	£1,450	£10,490	164,610	£1.7 billion
Domestic burglary	£1,030	£5,510	£1,180	£7,720	582,000	£4.5 billion
Theft of a vehicle	£2,750	£14,960	£2,620	£20,330	80,000	£1.6 billion
Theft from a vehicle	£170	£1,130	£130	£1,420	633,000	£900 million
Theft from the person	£290	£1,120	£180	£1,590	389,890	£600 million
Criminal damage	£300	£1,430	£240	£1,970	990,440	£1.9 billion
Fraud (1)	£260	£1,600	£130	£2,000	3,675,000	£7.3 billion
Cybercrime (2) (3)	£150	£600	£0	£750	876,000	£700 million
Commercial (9 sectors only)						£16.9 billion
Commercial robbery	£1,030	£8,680	£1,450	£11,150	608,630	£6.8 billion
Commercial burglary	£3,690	£5,460	£870	£10,020	343,130	£3.4 billion
Commercial theft	£110	£380	£90	£570	9,623,180	£5.5 billion
Theft of a commercial vehicle	£6,970	£23,000	£2,620	£32,590	16,790	£500 million
Theft from a commercial vehicle	£480	£1,700	£130	£2,300	95,080	£200 million
Commercial criminal damage	£160	£700	£240	£1,100	362,140	£400 million

Notes:

1. The total cost of fraud is £500 million greater than that published alongside the Fraud Strategy: stopping scams and protecting the public (2023), due to the adoption of a revised methodology for estimating lost output.

2. The definition of cybercrime relates to computer misuse, namely computer viruses and unauthorised access to personal information (including hacking).

3. Unit response costs for cybercrime are £0 as (a) unit police costs cannot be calculated for cybercrime as it was not recorded as a category in the ABC data in 2006 to 2007 and (b) unit CJS costs for cybercrime are very low per cybercrime offence as the total CJS costs are low and the volumes of crime are high.

1. Introduction

Home Office estimates of the economic and social costs of crime were first published in 2000 (Brand and Price, 2000). This was the first comprehensive attempt to estimate the costs of crime to society in England and Wales. An updated Home Office study (Dubourg et al., 2005) developed the methodology further and updated the estimates in 2005. A minor update was published in 2011 based on the latest crime data (Home Office, 2011). This was followed by a further update and published report in 2018 using updated crime and price data (Heeks et al., 2018). The Economic and social costs of crime (Heeks et al., 2018) report will from here forwards be referred to as ESCC 2018.

The analysis presented here further develops the methodology used in the earlier publications. It uses 2019 to 2020 crime and price data, where available, to give a more comprehensive and up to date picture of economic and social costs of crime. It is similar to the previous publications in terms of the cost categories it considers. These include costs incurred in anticipation of crime, as a consequence of crime and in response to crime. However, relative to ESCC 2018, this report calculates the costs to a greater number of business sectors.

1.1 What is included?

The current analysis breaks down the unit and total cost estimates by offence type. As in the previous analysis, only the costs of notifiable offences, for which police forces are required to provide details to the Home Office, are included; therefore, non-notifiable offences, such as summary motoring offences, are not estimated (see ONS, 2021a). The analysis estimates the costs of crimes against individuals and certain business sectors. Financial costs are not always borne by victims of crime, for example businesses sometimes refund defrauded consumers or individuals working for a business may experience harms from a robbery against the business. However, costs presented in this report represent a cost to society. Cybercrime and fraud incidents are not captured in crimes against businesses despite both being notifiable offences, as there were no available datasets capturing both reported and unreported volumes of fraud and cybercrime against businesses at the time of development, and various cost components had evidence gaps (for example, lost output from cybercrime incidents shutting down network systems).

The costs of crime are split into 3 main cost areas, each of which contains relevant cost categories:

 1) In anticipation of crime

a. Defensive expenditure
Defensive expenditure is defined as money individuals and businesses spend on crime detection and prevention. This encompasses expenditure such as burglar alarms, CCTV equipment and car alarms.

b. Insurance administration
The value of insurance administration costs resulting from crime is included but not the value of insurance pay-outs to victims of crime. In economics terms the latter does not represent a cost to society as it is a transfer of money between an individual and a business and vice versa. The costs to the victim or business are also already captured in the value of property stolen or damaged. The insurance cost that crime creates is the cost of employees of insurance firms dealing with crime-related insurance claims (for example, premises, salary, and equipment costs) when they could be engaged in other productive activities in society. It also includes the administration costs of the Criminal Injuries Compensation Authority, a public body which administers the payments to victims of violent crimes.

 2) As a consequence of crime

a. Property stolen or damaged
This is the value of the property stolen or damaged as a result of crime.

b. Physical and emotional harm to the victim
This is the reduction in the quality of life of the victim from the physical and emotional harm suffered as a result of the crime.

c. Lost output
Lost output estimates the lost productivity from time off work and reduced productivity whilst at work for victims of crime.

d. Health services   
There are health service costs from dealing with the physical and emotional harms of crime. These include ambulance costs, medical procedure costs associated with physical harm, and counselling costs associated with the emotional harms.

e. Victim services   
Victim services estimates the cost of support provided to victims of crime.

 3) In response to crime

a. Police costs  
A large part of the police’s resource is spent dealing with crimes. The cost estimated here is the opportunity cost of the police investigating these crimes. Opportunity cost is the value of other options that are forgone, so this cost captures the value of police time and resources that are taken up by investigating a certain crime, rather than engaging in other productive activity, such as responding to non-crime activities (for example, public order or disaster relief, among others).

b. Other CJS costs  
The CJS is a set of agencies and processes established by the Government to control crime and impose penalties on those who break the law. The costs include those for the Crown Prosecution Service (CPS), courts, defence, prison, and probation.

Despite the wide range of costs included in the estimates there are inevitably some costs of crime that cannot be estimated due to the lack of available evidence or data. This includes the impact of crime on indirect victims (for example, the cost of fear to family members of victims), and the cost of non-Ministry of Justice Victim Services, for multiple offence types; and police costs, for cybercrime specifically.

This report contains estimates of the economic and social costs of crime to society in the year 2019 to 2020. This year has been chosen to avoid capturing the COVID-19 pandemic impact on crime trends. Instead, by using data from 2019 to 2020, the estimates provide a baseline of costs upon which future costs can be developed.

The persistence of pandemic-related changes in the nature and volumes of different crime types in the medium and long-term is unknown. By calculating costs using data unaffected by the pandemic, costs can be used for decision-making as trends stabilise. This is under the knowledge that these costs are for the year 2019 to 2020 and therefore may not reflect crime trends beyond 2019 to 2020.

The unit costs of crime presented in this report are estimates which reflect the average monetary impact of a crime. They demonstrate the relative magnitudes of the economic and social costs of different crimes and should not be treated as precise estimates of the cost of each crime. The cost estimates in this report use the best available evidence and data at the time of analysis but they are inevitably sensitive to changes in crime trends, organisational developments, and changes in technology. These factors should be considered when deciding whether these estimates can be used, as the multipliers (see Section 3.3) and unit costs may no longer be appropriate.

1.2 Summary of the total costs of crime

Table 2 provides a breakdown of the total cost of crime for the year 2019 to 2020. This shows that costs as a consequence of crime are the largest contributor to the total costs of individual and commercial crimes.

For businesses, the largest impact of crime is the items of value that are damaged or lost in these crimes, whereas for individuals, the largest impact of crime comes from the physical and emotional harms that they suffer as a result of these crimes.

Table 2 also shows whether the costs in each category are ‘financial’ costs or ‘non-financial’ (monetised) costs. ‘Financial’ costs are the direct cash and budget costs associated with crime. ‘Non-financial’ costs refer to all other costs, including costs that use ‘notional’ (non-market) values which represent estimated harm in monetary terms. Some cost categories include both financial and non-financial costs.

Harms are notated in monetary terms as a common metric, and while this does not directly correspond to a cash value, it is used to allow both financial and non-financial costs to be put on the same scale. The report is aware that this approach may seem reductive to those that have experienced any of the crimes analysed and recognise the profound human costs of these crimes to victims, and the entirety of these cannot be captured in this report. However, in assessing the costs of crime in this way, this report is better able to articulate the costs of crime to victims, and the wider economy, demonstrating the impact in a clear and quantified manner.

1.3 How to use the cost of crime estimates

The unit costs of crime estimates are designed to help policymakers and practitioners weigh up the crime reduction benefits of policies and help assess the cost effectiveness of particular interventions. Users should therefore understand how to use the costs of crime to ensure the net benefit from crime reduction policies is calculated correctly. The unit costs of crime capture all crimes and not just crimes recorded by the police. The total costs of crime for each offence are therefore divided by all crime (both recorded crime and crime not reported to the police) to calculate the unit costs. This is an important distinction to understand when estimating the benefits of any crime reduction policy. The unit costs can therefore only be used directly with reductions in all crimes (such as CSEW crimes). The unit costs should not be used in conjunction with Police recorded crime (PRC) data without applying the multipliers presented in Section 3. Box 1 outlines a simple example of how the unit costs of crime and multipliers should be used with PRC.

When using the unit costs in appraisal to estimate the monetised benefits of interventions leading to reductions in crime, users should assess whether to include anticipation costs on a case-by-case basis, considering the magnitude of the reduction in crime volumes resulting from an intervention. Individuals are assessed as being unlikely to reduce their spending on anticipatory measures, for example, burglar alarms, CCTV in response to marginal reductions in crime volumes, but may do so in response to larger reductions.

Box 1: Example of how to use the costs of crime for PRC reductions

A pilot of a crime reduction policy shows evidence of a 5% reduction in police recorded crime for robbery.

The policymaker would like to know what the estimated benefit would be from rolling this policy out to the rest of the country assuming the 5% reduction.

Say there are 50,000 police recorded robberies in England and Wales a year. To use the unit costs of crime, the number of police recorded robberies need to be converted into an estimate for all robberies (including those not recorded by the police). This can be done using the multiplier estimates in Section 3 of this report.

If the multiplier for robbery was 4 (implying that only 1 in 4 robberies are reported to the police), the estimate for the actual number of robberies would be 200,000.

A 5% reduction in recorded robberies is therefore estimated to result in a 10,000 reduction in all robberies.

The unit cost of crime for robbery, excluding the anticipation costs, can then be multiplied by the total reduction in all robberies (10,000) to calculate the benefit of this crime reduction policy.

Table 2: Breakdown of total costs of crime in England and Wales in 2019 to 2020

Note: Some components of the costs below have since been superseded by subsequent analytical advances. For use in analysis, the Economic and social cost of crime: Amendments to unit costs provides more up to date unit cost estimates, including police response and fraud costs.

Cost Area	Cost Category	Contribution to overall cost of individual crimes (£ millions)	Contribution to overall cost of commercial crimes (£ millions)	Method (1)	Proportion of total cost of individual crimes	Proportion of total cost of commercial crimes	Financial or non-financial (2)	Time to realise (3)
Costs in anticipation	Defensive expenditure	£2,540	£2,700	Top-down	4%	16%	Financial	One year
Costs in anticipation	Insurance administration	£740	£470	Top-down	1%	3%	Non-financial	One year
Costs in anticipation	Subtotal	£3,280	£3,180
Costs as a consequence	Property stolen or damaged	£7,360	£6,510	Bottom-up	12%	38%	Non-financial	One year
Costs as a consequence	Physical and emotional harm to the victim	£28,660	£2,840	Bottom-up	47%	17%	Non-financial	Up to 5 years
Costs as a consequence	Lost output	£9,780	£1,760	Bottom-up	16%	10%	Non-financial	Up to 5 years
Costs as a consequence	Health services	£3,320	£490	Bottom-up	5%	3%	Financial	One year
Costs as a consequence	Victim services	£70	£10	Top-down	0%	0%	Financial	One year
Costs as a consequence	Subtotal	£49,190	£11,600
Costs in response	Police costs	£5,100	£1,260	Top-down	8%	7%	Non-financial	One year
Costs in response	Other Criminal Justice System costs	£3,260	£890	Top-down and Bottom-up	5%	5%	Financial and non-financial	Up to 20 years
Costs in response	Subtotal	£8,360	£2,150

Notes:

1. ‘Bottom-up’ unit costs are a direct cost per victim measure. ‘Top-down’ costs refer to unit costs estimated by calculating a total cost and dividing by the prevalence estimate.

2. ‘Financial’ costs are the direct cash and budget costs associated with crime. ‘non-financial’ costs refer to all other costs, including costs that use ‘notional’ (non-market) values which represent estimated harm in monetary terms.

3. This is the time period over which these costs are realised. Physical and emotional harms to the victim take up to 5 years to realise because the 3 longest lasting harms (the rape-related harms of drug abuse, alcohol abuse and obesity/eating disorder) are assumed to last for 5 years. Other CJS costs include the prison costs of those convicted for the crime of homicide. Given that the average sentence length for someone convicted of homicide - rounded to the nearest 5 years - is 20 years, the time to realise the CJS costs is up to 20 years.

1.4 Methodological changes since ESCC 2018

Scope, data input, and methodological changes in this report compared to previous editions (for example, ESCC 2018) mean that the costs cannot be directly compared to previous reports. Examples of changes include, but are not limited to:

• removal of arson as a crime type

  Arson as a crime against individuals and businesses has been removed from the total and unit cost estimates as the methodology used in this report excludes significant elements of the cost of arson, for example, the cost of Fire and Rescue Services responses. Instead, the cost of arson should be considered in the context of other fire harms, such as deliberate fires, where more specific cost categories should be considered that would not be applicable to the crime types considered in this report. This is included in the Economic and Social Cost of Fire (ESCF, 2023).

• increased number of sectors included in the crimes against business estimates

  Due to increased data availability, the analysis now captures the costs of crime against 9 business sectors, compared to the previous 7 sectors.

• improved CSEW data quality for fraud and cybercrime

  Data used in ESCC 2018 for fraud and cybercrime was experimental CSEW data. Newer CSEW data on fraud and cybercrime has been upgraded to Official Statistics.

• prison costs ‘bottom-up’ methodology

  Within CJS costs, a new methodology is applied to understand the impact of crimes committed in the 2019 to 2020 financial year. A ‘bottom-up’ methodology is applied to estimate the likely prison cost of crimes committed in the year 2019 to 2020, rather than apportioning the 2019 to 2020 prison budget to different crime types as is the methodology in ESCC 2018.

• removal of volunteer time from Victim Services

  In contrast to the ESCC 2018, in this analysis a cost has not been estimated for volunteer time for Victim Services due to significant data gaps. Previous methodology used data from one charity, Victim Support. Changes to the operational model of this organisation, specifically the shift from volunteer to employee facilitation of services, while other organisations have remained mostly volunteer-led (including the National Witness Service), means that it is not appropriate or representative to produce estimates using data from this organisation only. Furthermore, the limited publicly available data in this area means both verifying the representativeness of Victim Support’s decline in volunteer hours across the whole victim services sector and estimating a robust cost from an alternative source difficult. While the analysis is missing costs for volunteer hours, the opportunity cost of paid staff is captured in the funding data presented in Section 5.5.

In some areas of the report no new data has been available. In absence of new data, costs have been adjusted where possible to reflect the crime landscape in 2019 to 2020.

2. Main results

The unit costs in Table 3 show the average estimated cost to society of one crime. These are the estimated total costs of crime for each crime type divided by all crimes committed (both reported and unreported) in England and Wales in the year 2019 to 2020 (Table 4).

Table 3: Unit costs of crimes by category (2019 to 2020 prices)

Note: Some components of the costs below have since been superseded by subsequent analytical advances. For use in analysis, the Economic and social cost of crime: Amendments to unit costs provides more up to date unit cost estimates, including police response and fraud costs.

Crime type	Defensive Expenditure	Insurance Admin	Value of Property	Physical/Emotional Harms	Lost Output	Health Services	Victim Services	Police	CJS	Total Unit Cost
Individual
Homicide	£68,700	£0	-	£2,247,260	£383,630	£1,230	£5,070	£11,710	£629,070	£3,346,680
Violence with injury	£280	£10	-	£9,060	£3,240	£1,060	£20	£1,100	£940	£15,720
Violence without injury	£120	£10	-	£3,850	£1,300	£320	£10	£790	£750	£7,150
Rape	£810	£10	-	£26,340	£10,730	£990	£40	£7,480	£1,610	£48,000
Other sexual offences	£130	£0	-	£4,210	£990	£400	£10	£670	£140	£6,540
Robbery	£190	£290	£1,600	£4,760	£1,740	£450	£10	£990	£450	£10,490
Domestic burglary	£720	£310	£3,080	£1,420	£580	£430	£0	£520	£660	£7,720
Theft of a vehicle	£2,120	£630	£13,830	£590	£290	£250	£0	£1,990	£630	£20,330
Theft from a vehicle	£140	£20	£740	£230	£100	£60	£0	£80	£40	£1,420
Theft from the person	£20	£260	£340	£440	£180	£150	£0	£40	£140	£1,590
Criminal damage	£30	£270	£440	£560	£210	£220	£0	£150	£90	£1,970
Fraud	£240	£20	£840	£360	£260	£140	£0	£60	£70	£2,000
Cybercrime	£150	£0	£60	£270	£180	£90	£0	-	£0	£750
Commercial (9 sectors only)
Commercial robbery	£990	£40	£1,080	£4,310	£2,590	£690	£10	£990	£450	£11,150
Commercial burglary	£3,640	£50	£4,340	£530	£440	£150	£0	£520	£350	£10,020
Commercial theft	£70	£40	£380	£0	£0	£0	-	£40	£50	£570
Theft of a commercial vehicle	£6,270	£700	£22,390	£370	£220	£10	£0	£1,990	£630	£32,590
Theft from a commercial vehicle	£440	£40	£1,510	£100	£90	£0	£0	£80	£40	£2,300
Commercial criminal damage	£110	£40	£560	£60	£40	£40	£0	£150	£90	£1,100

Notes:

1. There may be discrepancies in the total figures due to the effect of rounding.

2. Unit police costs cannot be calculated for cybercrime as it was not recorded as a category in the ABC data in 2006 to 2007. Unit CJS costs for cybercrime are very low per cybercrime offence as the total CJS costs are low and the volumes of crime are high.

Table 4: Total costs of crime in England and Wales (2019 to 2020)

Crime type	Unit costs	Number of crimes (1)	Total costs
Individual
Homicide	£3,346,680	673	£2.3 billion
Violence with injury	£15,720	885,970	£13.9 billion
Violence without injury	£7,150	720,710	£5.1 billion
Rape	£48,000	164,070	£7.9 billion
Other sexual offences	£6,540	1,881,180	£12.3 billion
Robbery	£10,490	164,610	£1.7 billion
Domestic burglary	£7,720	582,000	£4.5 billion
Theft of a vehicle	£20,330	80,000	£1.6 billion
Theft from a vehicle	£1,420	633,000	£900 million
Theft from the person	£1,590	389,890	£600 million
Criminal damage	£1,970	990,440	£1.9 billion
Fraud (2)	£2,000	3,675,000	£7.3 billion
Cybercrime	£750	876,000	£700 million
Total cost of crimes against individuals			£60.8 billion
Commercial (9 sectors only)
Commercial robbery	£11,150	608,630	£6.8 billion
Commercial burglary	£10,020	343,130	£3.4 billion
Commercial theft	£570	9,623,180	£5.5 billion
Theft of a commercial vehicle	£32,590	16,790	£500 million
Theft from a commercial vehicle	£2,300	95,080	£200 million
Commercial criminal damage	£1,100	362,140	£400 million
Total cost of commercial crime			£16.9 billion

Notes:

1. Estimated using the CSEW (individual) and CVS (commercial – 9 sectors only) with the exception of homicide which used the Homicide Index. The CSEW figures are based on the main survey, aged 10 to 15 survey and estimates for the number of crimes for individuals aged under 10. See Section 3 of this report for more detail.

2. The total cost of fraud is £500 million greater than that published alongside the Fraud Strategy: stopping scams and protecting the public (2023), due to the adoption of a revised methodology for estimating lost output.

3. Estimating the number of crimes and calculating the multipliers

This chapter outlines how the numbers of crimes committed in the year 2019 to 2020 have been estimated, and how these estimates feed into the multipliers. The estimated total number of crimes is used to calculate the unit costs for each of the cost and crime categories. The multipliers are produced to enable users of the costs of crime unit cost estimates to scale up PRC volumes to all crimes committed (reported and unreported) when required.

3.1 Crimes against individuals

To estimate the total number of crimes committed there are 2 key data sources: PRC and CSEW.

PRC statistics contain the annual volumes for the total number of crimes recorded by the police (Home Office, 2022). Although PRC data provides a good measure of crime-related demand faced by the police, it does not capture all crimes committed against individuals each year. Therefore, using PRC volumes alone would underestimate the number of crimes committed in a year and their subsequent impact on society. PRC also records incidents in the year in which they are reported and so historic crimes could be included in data from later years should it take time to be reported - homicides are assumed to occur in the year they are reported.

The CSEW is often used as the basis to estimate the total number of crimes in England and Wales, as in ‘The economic and social costs of crime’ first edition (Brand and Price, 2000), second edition (Heeks et al., 2018) and Dubourg et al. (2005). Only notifiable offences are included, therefore the costs associated with non-notifiable offences, such as summary motoring offences, are not estimated (see ONS, 2021a).

The CSEW in 2019 to 2020 surveyed 33,735 adults and 2,398 children aged 10 to 15 about their experiences of crime over the previous 12 months. It is a survey of households but does not cover people resident in institutions. Usually, the survey interviews 35,000 adults and 3,000 children, however, some interviews towards the end of the sample window were cancelled due to the COVID-19 pandemic. Despite the impact on interviews, the survey results were largely unaffected (ONS, 2020a).

The CSEW estimates the total number of crimes which occur in England and Wales against victims either aged 16 to 74 or aged 10 and over depending on the crime type. CSEW volume estimates use the main face to face survey for all areas except for sexual offences where a self-completion module is used.

To estimate the number of crimes against individuals aged 9 and under (for sexual offences, this is instead for those aged either 15 and under or aged 75 and over), it is assumed that the age distribution of PRC is equivalent to the age distribution for all victims of crimes in England and Wales. Following the approach taken by ESCC 2018, PRC data on the proportion of victims by age is split into 4 bands per crime type (ages zero to 9; 10 to 15; 16 to 74; and aged 75 and over), and these are used to estimate the proportion of the total victim population captured by the CSEW. As the analysis uses PRC age data, it may be affected by reporting bias; certain age groups may report more (or less) proportionally, resulting in a larger (or smaller) number of crimes being allocated to this age band than occur. The extent of reporting bias is uncertain but is more likely to impact analysis where a large proportion of victims are aged below or above the CSEW age range, such as other sexual offences.

These CSEW volumes are then divided by this proportion to estimate the total volume of crimes in society for the year 2019 to 2020, across all ages. To estimate the total number of incidents of violence with injury in the year 2019 to 2020, the combined 10 to 15 and aged 16 and over volumes are divided by the proportion of aged 9 and under victims from PRC by age data. This results in an estimated volume for victims of all ages.

All CSEW volume estimates are published alongside 95% confidence intervals to demonstrate uncertainty around the estimates: these uncertainties arise in all estimates produced using any survey. There is variation in the range of confidence intervals between crime types. To maintain comparison with the central estimate, confidence intervals have been manipulated in the same manner and extrapolated to provide upper and lower bounds around the estimates, for all ages. As a result of this manipulation, these upper and lower bounds should not be treated as confidence intervals. By extending the coverage in this way additional uncertainty is being incorporated, which this report is not able to quantify. Therefore, the upper and lower bounds around estimates for each crime type should be interpreted as providing a rough indication of uncertainty around the central volume estimate. These lower and upper estimates are used to inform total cost ranges within the report.

3.2 Crimes against businesses

As in ESCC 2018, the Commercial victimisation survey (CVS) is used to calculate the total number of crimes against businesses. The CVS surveys a few business sectors per year and data used in this report was collected annually between 2012 and 2018. Having been paused since 2018, the 2021 CVS was published in March 2021. The CVS previously ran between 1994 and 2002 but these results are not used due to the age of the data. From 2012 to 2018, the CVS only covered 3 or 4 industrial sectors each year, with 9 out of the 21 major standard industrial classification sectors (ONS, 2022a) being covered overall during this period. Therefore, data is used for each sector from the most recent CVS available for said sectors. The sectors and corresponding CVS year used are shown in Table 5.

Table 5: CVS data used

Sector	CVS year used
Construction	2015
Information and communications	2015
Administration and support	2016
Transportation and storage	2016
Arts, entertainment and recreation	2017
Manufacturing	2017
Accommodation and food	2018
Agriculture, forestry and fishing	2018
Wholesale and retail	2018

Notes:

The analysis does not include estimates for (i) mining and quarrying; (ii) electricity, gas, steam and air conditioning supply; (iii) water supply, sewerage, waste management and remediation activities; (iv) financial and insurance activities; (v) real estate, renting and business activities; (vi) professional, scientific and technical activities; (vii) public administration and defence; compulsory social security (viii) education; (ix) human health and social work activities; (x) other service activities; (xi) activities of households as employers; and (xii) activities of extraterritorial organisations and bodies.

CVS data gives the number of incidents for each commercial crime type, per 1,000 premises, for each industrial sector for each year the survey is conducted. In the absence of 2019 to 2020 data for all sectors surveyed, the change in the number of premises between the survey year and 2019 to 2020 is accounted for to better reflect the number of incidents in the reporting year of 2019 to 2020.

This is calculated by dividing the number of incidents per 1,000 premises for the latest survey year available for each crime type and sector by 1,000 and multiplying this by the number of premises for that sector in 2019 to 2020 (ONS, 2020b). The estimates across each crime type are then added together for each sector to calculate the total crime volumes per crime type in 2019 to 2020. For example, there were 211 burglaries per 1,000 premises in the manufacturing sector according to the central estimate from the 2017 CVS. The number of manufacturing premises in 2019 to 2020 was 133,610. Therefore, the number of incidents of burglary in the manufacturing sector for 2019 to 2020 is calculated by dividing 211 by 1,000 and then multiplying by 133,610. This gives a central estimate of 28,191 burglaries in the manufacturing sector in 2019 to 2020.

The commercial volumes estimates are shown in Table 6. As only 9 of the 21 business sectors are included in this analysis, the total volumes, and therefore total costs, of crimes against businesses are only partially captured. The Inter-departmental Business Register (IDBR) suggests that the sectors included in this report account for just over 60% of the number of VAT and/or PAYE enterprises (this does not account for enterprise size, only the number of enterprises) in England and Wales.

Low and high-volume estimates, with a confidence interval of 95%, are included in the analysis for each commercial crime type to account for the uncertainty of the analysis (Table 6). This uncertainty demonstrates some issues with data quality within the CVS, where certain sectors have a lower response rate than others, resulting in a varying picture of data robustness.

3.3 Multipliers

The estimated unit costs of crime shown in Table 6 are the average costs of each crime regardless of whether it was reported to the police or not. This is to ensure the unit cost reflects the cost of each crime committed, rather than the cost of each crime recorded by the police, as not all crimes are discovered by or reported to them.

In analysis PRC is often used as a measure of crime because it is available in both greater detail and frequency than the CSEW. In this situation, PRC volumes would need to be correctly converted to an estimate of all crimes if being used alongside the estimated unit costs in this report. Multipliers are calculated by dividing our estimated volumes through by the equivalent PRC volume for each crime type. The PRC volume for all crime types apart from fraud and cybercrime is sourced from PRC data for 2019 to 2020. The PRC volume for fraud and cybercrime is sourced from the National Fraud Intelligence Bureau (NFIB) statistics.

The multipliers in Table 6 allow changes in PRC to be converted into an estimate for changes in all crimes. For example, if a policy were to prevent 100 robberies recorded by the police, then by using the multiplier for robbery in Table 6, the estimate for the actual reduction in all robberies committed would be 202 (100 x 2.02). This is the number that would need to be multiplied by the unit cost in order to calculate the savings associated with preventing all crimes. Multipliers tend to vary across offence types and can also change over time as victim reporting dynamics and police recording practices change.

Multipliers cannot be calculated for all commercial crimes. This is because the total numbers of crimes committed do not apply to all sectors, therefore they cannot be compared with PRCs to calculate multipliers.

3.4 Summary of volumes and multipliers

The estimated total number of crimes, PRC volumes and resultant multipliers are summarised in Table 6. The table splits out the volumes between estimates for individuals aged 9 and under, CSEW estimates and CVS estimates. It then uses estimates for total crime volumes and PRC to estimate multipliers for converting PRC volumes to estimated total crime volumes.

Table 6: Total number of crimes committed, PRC volumes and resultant multipliers

Crime type	Aged 9 years and under (1), estimate	Aged 10 to 15 years (1), CSEW	Aged 16 years and over (2), CSEW	Low CVS commercial volume estimate	Central CVS commercial volume estimate	High CVS commercial volume estimate	Estimated total crime	PRC	Estimated multiplier
Homicide (2)	-	-	-	-	-	-	673	673	1.00
Violence with injury	29,994	264,980	591,000	-	-	-	885,974	540,808	1.64
Violence without injury (4)	39,778	32,932	648,000	-	-	-	720,710	733,136	0.98
Rape	11,729	23,530	128,811	-	-	-	164,069	58,959	2.78
Other sexual offences	253,681	652,876	974,623	-	-	-	1,881,181	104,082	18.07
Robbery	298	37,315	127,000	-	-	-	164,612	81,293	2.02
Domestic burglary	-	-	582,000	-	-	-	582,000	268,540	2.17
Theft of a vehicle (5)	-	-	80,000	-	-	-	80,000	113,162	0.71
Theft from a vehicle	-	-	633,000	-	-	-	633,000	285,311	2.22
Theft from the person	319	22,575	367,000	-	-	-	389,894	113,682	3.43
Criminal damage	1,235	4,207	985,000	-	-	-	990,442	534,108	1.85
Fraud (6)	-	-	3,675,000	-	-	-	3,675,000	312,040	11.78
Cybercrime (6)	-	-	876,000	-	-	-	876,000	26,215	33.42
Commercial robbery	-	-	-	300,818	608,634	923,918	608,634	-	-
Commercial burglary	-	-	-	201,755	343,132	484,510	343,132	-	-
Commercial theft	-	-	-	6,612,846	9,623,181	12,643,937	9,623,181	-	-
Theft of a commercial vehicle	-	-	-	4,370	16,788	29,376	16,788	-	-
Theft from a commercial vehicle	-	-	-	25,956	95,081	164,314	95,081	-	-
Commercial criminal damage	-	-	-	215,798	362,138	508,478	362,138	-	-

Notes:

1. Sexual offence volumes are only estimated in the CSEW for those aged 16 to 74 years. The figures in these columns include estimates for those aged 15 years and under. These are estimated using the same methodology as the estimates for aged 9 years and under for violence with injury, violence without injury, robbery, theft from the person and criminal damage.

2. Sexual offence volumes are only estimated in the CSEW for those aged 16 to 74 years. An estimate for sexual offences has been included for those aged 75 and over. These are estimated using the same methodology as the estimates for aged 9 years and under for violence with injury, violence without injury, robbery, theft from the person and criminal damage.

3. It is assumed that all homicides come to the attention of the police, therefore the total number of homicides is assumed to be the same as the number of recorded homicides.

4. The multiplier is less than one because police recorded crime covers all violence without injury whereas CSEW only includes crimes against people resident in households.

5. The multiplier is less than one as police recorded crime covers all vehicle thefts (including those against businesses) whereas CSEW only includes crimes against individuals. Given PRC’s larger remit, this leads to PRC volumes being higher than CSEW.

6. Estimates for 10 to 15 year olds cannot be produced for fraud and cybercrime as they are not included in the aged 10 to 15 years module. Estimates for aged 15 years and under also cannot be produced using the stated methodology with PRC by age data. The volumes of fraud and cybercrime are therefore likely to be an underestimate.

7. Fraud and cybercrime are not included. The CVS is a premises survey and therefore will significantly underestimate fraud and cybercrimes as they are likely to occur at head office level.

8. There may be discrepancies in the total figures due to the effect of rounding.

4. Costs in anticipation of crime

This chapter discusses the methodological approach and outlines the estimates of the costs in anticipation of crime. These include:

• defensive expenditure

• insurance administrations costs

4.1 Defensive expenditure

4.1.1 Approach

Defensive expenditure is the money spent by individuals and businesses on crime detection and prevention. This encompasses estimates for both security equipment (such as burglar alarms, CCTV equipment, car alarms) and private security (such as door supervision). The methodology followed in this analysis is similar to that used in ESCC 2018. The source of the estimates, details of how they have been split out by crime type and assumptions made are explained in the remainder of this chapter.

The cost of defensive expenditure is calculated as the total expenditure by individuals and businesses on the prevention of the specific crime(s) that relate to that type of expenditure. For example, total expenditure on car alarms is divided out across all ‘theft of’ and ‘theft from’ vehicle offences.

In total, 6 categories of security expenditure are estimated for crimes against individuals and businesses as shown in Table 7 below:

Table 7: Areas of defensive expenditure

Area	Individual	Commercial
General building security	🗸	🗸
General building security (specific to burglary prevention)	🗸	🗸
Vehicle security	🗸	🗸
Regulated security	🗸	🗸
Fraud	🗸	×
Cyber security	🗸	×

The defensive expenditure costs do not capture all the costs associated with precautionary behaviour to reduce the risk of being a victim of a crime. For example, someone may take a taxi home to avoid the risk of violence with injury. Additionally, the costs do not capture all potential products and methods used to defend against crime, such as in-house unregulated security guards. There is a non-exhaustive list of areas and products that are excluded due to evidential and feasibility constraints. Due to these difficulties, this report maintains consistency with the ESCC 2018 for the areas of defensive expenditure costed. These costs should therefore be interpreted as an underestimate.

4.1.2 Total costs

Total costs of market sizes are used to apply a top-down methodology for the defensive expenditure per total crimes in England and Wales. The following market sizes are used:

General building security

General building security expenditure is defined as a collection of common security measures, such as access control and intruder alarms, used to protect premises from crime Total building security expenditure for England and Wales is estimated in IHS Markit’s ‘Physical security equipment and services report 2015’ (Millar, 2015) for the UK and Ireland. This report considers expenditure on video surveillance, access control, intruder alarms, staffed security systems and wireless infrastructure among other things. The 2015 version of this report is used due to the lack of a readily available updated edition. It estimates the 2014 building security annual expenditure for the UK and Ireland to be $4.5 billion.

In this analysis, it is assumed that intruder alarms are only used for the prevention of burglary. This is assumed because an individual must be present in all other crime types (for example, homicide), or, in the case of vehicle security, there is already an alarm on the vehicle. The IHS report estimates that intruder alarms account for 55% of security expenditure (excluding vehicles) in the UK and Ireland. Intruder alarms are therefore estimated to cost approximately $2.5 billion of the $4.5 billion per annum (calculated by multiplying the $4.5 billion per annum by 55%).

The total UK expenditure is scaled to England and Wales, in line with the CSEW estimates, using the proportion of UK and Ireland’s Gross Domestic Product (GDP) that England and Wales make up (83%) (ONS, 2019; World Bank, 2021a; World Bank 2021b). Both estimates are then converted to GBP (£) and inflated to 2019 to 2020 prices to be comparable with the rest of the analysis.

Each estimate is then split out into individual and commercial expenditure based on the proportions listed in the IHS report. However, there is only overlap between the 2 reports in 3 sectors which are manufacturing; wholesale and retail; and transportation and storage. Therefore, general building security estimates are only calculated for these sectors and the remaining sectors are excluded from this analysis. Due to lack of data, it is therefore assumed that defensive expenditure costs and crime volumes for these sectors are representative of all other commercial sectors.

There have likely been changes in the total cost of general building security over time that is not reflected in the estimate from 2014. To account for this, the total cost is adjusted by the IHS report’s Compound Annual Growth Rate (CAGR) projection for the market (6% per annum), over the 5 years from 2014 to 2019. This results in final estimates for general security expenditure excluding burglary specific expenditure of £246.1 million for individuals and £364.5 million for commercial. For general security expenditure specific to burglary prevention, that is intruder alarms, it estimates £338.5million for individuals and £501.2 million for commercial.

Vehicle security

Data on the total security expenditure for vehicles were not available for England and Wales, therefore estimates have been made based on data from the USA in 2016, since it is assumed that the vehicle security market in the USA has similar characteristics to that of England and Wales. There are potential risks around the accuracy of this assumption, as the average household in the USA tends to own more cars than their UK equivalent and may therefore spend more per person on vehicle security components by purchasing them.

1. The USA share of the global vehicle security market in 2016 is calculated at £1.8 billion (38% of the global market size, £4.6 billion) (Marketsandmarkets.com, 2016). The total size of the USA market uses the average number of vehicle security systems that are installed on vehicles in each category (passenger car, commercial vehicle and off-highway vehicle) multiplied by vehicle production numbers. The US share of vehicle security system volumes is based on the proportion of US vehicles in each category (Marketsandmarkets.com, 2016).
2. To estimate total vehicle security expenditure in the UK, the ratio of the number of vehicles in the UK compared to the USA (14%) is applied to the estimated USA market. This is calculated based on 39.9 million vehicles in the UK (Department for Transport, 2020b) and 276.5 million vehicles in the USA (Statista, 2021).
3. The estimate is scaled using the ratio of vehicles in England and Wales compared to the UK (89%), resulting in a total vehicle security cost in England and Wales of £225 million (Department for Transport, 2020b).
4. The total cost estimate is split out into individual and commercial expenditure based on the proportion of company registered cars in the UK (Department for Transport, 2020b). For consistency with the data limitations from general building security expenditure, only manufacturing, wholesale and retail, and transportation and storage total vehicle security costs are included in the commercial costs.
5. The total vehicle security market estimate for commercial vehicles in England and Wales is multiplied by the combined market share from the IHS Markit report for the 3 sectors to reach a final estimate. The remaining commercial costs are not covered by sectors for which there is CVS volumes data, hence they are excluded from this analysis.
6. Finally, the estimates are inflated to 2019 to 2020 prices and adjusted for market growth over the period between the source report (2014 to 2015 prices) and the year of interest (2019 to 2020) with a compound annual growth rate (CAGR) of 4.1% (Future Market Insights, 2021).
7. This results in an estimate of the total vehicle security expenditure for individual and commercial vehicles in England and Wales of £205.6 million and £22.9 million, respectively. Due to a lack of data, it is assumed that security expenditure per vehicle is the same for commercial and non-commercial vehicles.

Regulated security expenditure

Regulated security expenditure relates to areas such as guarding, cash and valuables in transit, close protection, and door supervision. The total expenditure for these areas is estimated by Infologue (2019). It estimates the UK turnover of the top 33 companies in the sector and attempts to exclude business activities not related to regulated security. It extrapolates this to all UK businesses in the sector by adjusting from the combined market share of the 33 companies covered, reaching a total expenditure estimate for regulated security of around £4.4 billion.

For an England and Wales estimate, the figure is scaled by the proportion of total UK GDP which England and Wales are responsible for (ONS, 2019; World Bank, 2021a; World Bank, 2021b), resulting in a figure of £4.0 billion. The expenditure could relate to crimes against individuals or businesses; there is no information provided by Infologue to estimate this. This report therefore assumes the same percentage split as for general building security (17.8% individual and 26.3% commercial). Also, as before, commercial expenditure is only estimated for the 3 sectors included in the general building security source. This results in estimates of £700 million for individual expenditure and £1.1 billion for commercial expenditure.

Fraud

Fraud protection expenditure relates to activity undertaken by companies to protect their customers and operations from fraud, including staffing and technology costs.

To estimate a cost for individuals, the model uses UK Finance data on the cost of anti-fraud operations undertaken by UK Finance members, and the value of reimbursements (UK Finance, 2020). Data limitations prevent the isolation of individual-specific cost. However, following guidance from subject matter experts, it has been assumed this is a fixed cost unaffected by the volume or type of customer protected and individuals using services covered by this cost receive the benefits of protection. To estimate the total value of fraud protection expenditure, the value of reimbursements is removed from the total before dividing through by individual fraud volumes to reach a unit cost.

There are wider fraud protection industry costs that we are not accounting for due to lack of data. The model does not capture fraud prevention and detection analysts in sectors with a heavily online presence (such as retail and technology) or fraud prevention and detection businesses. Ravelin’s Online merchant fraud and payments survey 2022 showed that 86% of online merchants have fraud teams of 6 or more people, indicating the potential true scale of fraud prevention expenditure (not UK specific). Therefore, we are likely to be significantly underestimating the total cost of fraud defensive expenditure.

Cybercrime

Cyber security expenditure relates to expenditure on consumer security software estimated in Gartner’s ‘Forecast Analysis: Information Security and Risk Management, worldwide’ report (see Gartner, 2020). The source provides an estimate for worldwide spending on consumer security software of $6.25 billion. To reach an estimate for individuals, the global market is scaled by the UK’s share of global GDP (approximately 3%) (World Bank, 2022) and then by the proportion of UK GDP represented by England and Wales (approximately 90%) (ONS, 2019) resulting in a final market estimate of £134.9 million. This estimate is divided by individual cybercrime volumes to reach a unit cost.

Only an estimate for individuals is calculated. It is assumed that expenditure on consumer security software is solely in relation to protecting against instances of cybercrime against individuals.

4.1.3 Splitting out the total costs

The 3 main categories of security expenditure – general building, vehicle, and regulated security expenditure - need to be attributed across to the crimes they are intended to prevent. General building security and regulated security are split between all crime types apart from fraud and cybercrime. Vehicle security is split between the 2 vehicle-related crime types, theft of a vehicle and theft from a vehicle. The remaining categories – general building security specific to burglary prevention (intruder alarms), cybersecurity, and fraud prevention – are split out as follows: general building security specific to burglary prevention (intruder alarms) is assigned solely to burglary. Cyber security is assigned solely to cybercrime, which draws only from this category. Fraud protection is assigned solely to fraud, which only draws from this category. Assignments are presented in Table 8 below.

Table 8: Cost category assignments per crime type

Crime type	General building security	Intruder alarms	Vehicle security	Regulated security	Fraud protection	Cybersecurity
Homicide	🗸			🗸
Violence with injury	🗸			🗸
Violence without injury	🗸			🗸
Rape	🗸			🗸
Other sexual offences	🗸			🗸
Robbery	🗸			🗸
Domestic burglary	🗸	🗸		🗸
Theft of a vehicle	🗸		🗸	🗸
Theft from a vehicle	🗸		🗸	🗸
Criminal damage	🗸			🗸
Fraud					🗸
Cybercrime						🗸

Cost categories are not divided equally between the crimes which draw from them. It is assumed that individuals base their spending decisions on the expected loss incurred if they were to be a victim of the given crime. Consequently, the model uses weights that consider the expected cost incurred, as calculated and used in other models included in this report. The elements used for calculating the weightings are: value of property stolen or damaged; physical and emotional harm incurred; and volume of the given crime as a measure of its prevalence in relation to the other crime types. Overall, the weighting for a given crime is a function of the cost incurred by a victim, multiplied by the likelihood of becoming a victim.

A crime type with a high value of property stolen or damaged, physical and emotional harm incurred, volume, or a combination of several of these aspects, will have a larger weighting. Subsequently, this will result in this crime type attracting a higher proportion of cost from the expenditure categories assigned to it in relation to other crime types assigned to these categories with lower weightings. For example, homicide has the highest unit cost for defensive expenditure due to the relative size of its physical and emotional harms cost leading to it having a high weighting. Examples of defensive expenditure for homicide includes security systems, CCTV and physical security equipment.

It is not certain what drives individuals to invest in defensive expenditure, or to what extent individuals weight their spending decisions on expected harm over likelihood and vice versa. The assumption to split out spend based on harm and volumes may therefore result in defensive expenditure being disproportionately allocated to certain crime types. Additionally, this weighting method assumes that individuals have perfect information of the harms that could be incurred as a result of a crime, and volume of crimes committed. In reality, individuals may not know the true values of these, and therefore they cannot inform their spending decisions.

Total defensive expenditure per crime type is calculated as the sum of all costs assigned to a given crime type from each of the relevant defensive expenditure categories. Unit costs for each crime type are calculated by dividing the total defensive expenditure of a given crime type by the relevant crime volume. The unit costs for defensive expenditure are presented in Table 9.

4.1.4 Unit costs for defensive expenditure

Table 9: Unit costs of defensive expenditure per crime type

Crime type	Defensive expenditure (1)
Individual
Homicide	£68,700
Violence with injury	£280
Violence without injury	£120
Rape	£810
Other sexual offences	£130
Robbery	£190
Domestic burglary	£720
Theft of a vehicle	£2,120
Theft from a vehicle	£140
Theft from the person	£20
Criminal damage	£30
Fraud	£240
Cybercrime	£150
Commercial (9 sectors only)
Commercial robbery	£990
Commercial burglary	£3,640
Commercial theft	£70
Theft of a commercial vehicle	£6,270
Theft from a commercial vehicle	£440
Commercial criminal damage	£110

Notes:

1. In the estimation of the total costs, the unit costs have been applied to all 9 commercial sectors. The use of the unit costs to calculate the total costs implicitly assumes that the main and private security unit costs apply to the 4 other sectors not covered by the analysis (only 3 are covered by the analysis – wholesale and retail, transportation and storage, and manufacturing).

4.2 Insurance administration

4.2.1 Approach

The value of insurance premiums and pay-outs do not represent an overall cost to society. These are exchanges of money from one party to another and are therefore not included in the insurance administration analysis. These costs are instead captured in the value of property stolen or damaged (see Section 5.1 of this report).

Separately, an opportunity cost arises from the business activities related to issuing those insurance contracts that protect against crime and the processing of crime-related claims. This time and resource could have been otherwise been engaged in other productive activities and is therefore a cost to society.

It is not only insurers that protect and compensate victims and therefore generate this cost. Public bodies and financial institutions protecting individuals and businesses from crimes such as fraud will also experience this administrative burden and are estimated in this section.

There are 2 main activities carried out which result in crime-related costs. The first is the cost of issuing insurance policies that offer protection against losses from crime. This is an anticipatory cost as it is borne prior to a crime occurring and are referred to in this section as ‘anticipation administration costs’. The second is the cost associated with processing and investigating crime-related claims. This cost is only incurred if a crime occurs, and a claim is made against the insurance policy. This is a cost as a consequence of crime and is referred to as ‘consequence administration costs’ in this section. Despite these differences in the types of costs insurance activities can be categorised as, all costs are presented in the anticipation section of this report for consistency with ESCC 2018. It is also important to note that insurance administration estimates for some crime types are entirely costs as a consequence of crime, for example, fraud.

The approach to costing insurance administration varies between crime types due to different sources of compensation. This means that estimates cannot be derived from one data source.

4.2.2 Insurance administration costs for crimes involving stolen and damaged property

Background

The Association of British Insurers (ABI) is a membership organisation for the UK’s insurance industry. Similar to ESCC 2018, ABI data is the primary source for estimating the insurance administration costs for crime types involving stolen or damaged property. The ABI hold data related to claims and expenses for property and motor insurance products. Some data is only available at calendar year level. In this instance, the analysis uses 2019 data with the acknowledgement that it is not fully representative of 2019 to 2020.

Property Insurance

Anticipation administration costs

The analysis uses ABI data (unpublished, provided bespoke by ABI) on Acquisition expenses to estimate the anticipation administration costs. These are the costs incurred from issuing an insurance contract, for example, underwriting costs. In 2019 to 2020, approximately 11% of the volume of claims made against domestic property insurance policies were related to theft. This proportion is multiplied by the value for Acquisition expenses to estimate the total administration costs related to crime. This assumes that anticipation administration costs do not vary by claim cause or by value of items insured.

The resulting cost is divided by the proportion of the property insurance market captured by ABI members to estimate a cost for the full UK market. This value is then adjusted to England and Wales based on the proportion of PRC crimes in the UK attributed to England and Wales. This results in a total crime-related Acquisition expenses estimate of £214.2 million for England and Wales in 2019 to 2020.

Consequence administration costs

In 2019 to 2020, approximately 12% of the total value of domestic property insurance claims were related to theft. This proportion is multiplied by the value for Claims management expenses to estimate the total administration cost associated with processing theft claims.

Using the same method as the anticipation estimate, this value is adjusted to account for the full UK insurance market, and then for England and Wales only. This results in a total administration estimate of £28.2 million for England and Wales in 2019 to 2020. The same method is used for property crimes against businesses.

Overheads

The analysis also estimates crime-related overhead costs. The analysis assumes expenditure on overheads, such as IT costs, salaries, and premises costs, that enable the issuing of policies and processing of claims is proportionate to the value of expenses incurred through each of these activities. The analysis applies each proportion to the Other expenses total to split this between anticipation and consequence activities. A crime-related estimate is then estimated for each total using the relevant assumption, applying the volume of theft-related claims for anticipatory overheads, and the value of theft-related claims for consequence related overheads.

The analysis then adjusts estimates to the full insurance market, and to England and Wales using the same method outlined previously. These values are then added to the anticipation and consequence administration estimates, resulting in final administration estimates for domestic property insurance of £292.9 million for anticipation costs and £38.6 million for consequence costs.

Splitting totals by crime type

The approach to splitting the total administration costs by crime type differs between anticipation and consequence estimates. The analysis estimates an equal anticipation unit cost for the crime types protected against under ABI theft data for each type of insurance. For example, burglary, robbery and theft from the person are captured under theft-related domestic property insurance claims. This assumes that the administrative process for issuing an insurance policy is standardised and unlikely to vary significantly between the level of cover purchased or the value of items being insured. For the remaining crime types, the most similar offence is used as a proxy with the analysis assuming an equal unit cost.

To split the total consequence insurance administration cost by crime type, this total is multiplied by the proportion of the total value of property stolen or damaged attributed to each of the crime types captured in the ABI theft data (burglary, robbery, and theft). The remaining crime types not specified in the ABI claims data, criminal damage and commercial criminal damage, are estimated using the most similar offence as a proxy. Specifically, the relevant proxies are domestic burglary for criminal damage, and commercial burglary for commercial vandalism. The relationship between the proxied crime type’s insurance administration unit cost and value of property stolen or damaged unit cost is assumed to be the same for the crime type being costed.

For example, the insurance administration unit cost for domestic burglary (£50) is divided by its value of property stolen or damaged (£3,080). This value is then multiplied by the value of property stolen or damaged for criminal damage (£440). This results in a unit cost for criminal damage of £8. This assumes that administration costs are relative to the value of the claim, with greater resources and time employed in processing higher value claims.

Motor insurance

The insurance administration cost associated with vehicle theft is estimated using a similar approach to crimes protected by property insurance. In 2019 to 2020, 2% of domestic motor insurance claims resulted from theft. This proportion is multiplied by the value for Acquisition expenses to estimate the cost associated with issuing motor insurance contracts to protect against crime (the anticipation insurance administration cost). Meanwhile in 2019 to 2020, 4% of the total value of domestic motor insurance claims were related to theft. This proportion is multiplied by the value for Claims management expenses to estimate the administration costs of processing vehicle theft claims (the consequence insurance administration cost). Using the same method described under property insurance, crime-related overhead costs are split proportionally and added to each insurance administration cost area total.

Both the anticipation and consequence insurance administration total costs are split into theft of a vehicle and theft from a vehicle by the proportion of the total number (anticipation) or value (consequence) of theft claims attributed to each crime type according to the ABI data. The resulting values are divided by the proportion of the motor insurance market captured by ABI members to estimate the cost for the full market. This is then multiplied by the proportion of PRC crimes across the UK attributed to England and Wales. The resulting total anticipation insurance administration cost is estimated at £24.0 million for theft of a vehicle and £12.0 million for theft from the vehicle. For consequence insurance administration costs, final estimates are £26.3 million for theft of a vehicle, and £2.6 million for theft from a vehicle. The same approach is used to estimate the total administration cost for commercial motor claims.

4.2.3. Insurance administration costs for personal injury claims

Background

The approach to costing insurance administration for personal injury crimes relies on data from the Criminal Injuries Compensation Authority (CICA). ESCC 2018 also estimated costs associated with claims made through private personal protection insurance products such as illness, disability, and life insurance. It used data on the proportion of A&E attendances caused by assault and the proportion of deaths caused by homicide to estimate a value for crime-related claims. However, illness and disability insurance products protect individuals against specific conditions that tend to be serious, life-changing, and long-term or terminal. Many of these are naturally occurring and are therefore unattributable to crime, and there is a lack of evidence to robustly estimate the extent of victims who experienced the injuries protected by this insurance due to crime.

Therefore, it is difficult to robustly estimate a value for crimes that would meet an insurance policy’s eligibility criteria. The presented estimates are therefore likely to underestimate the administration costs associated with crimes resulting in injury to the victim.

Approach

The total administration cost for CICA in 2019 to 2020 was £15.4 million (CICA, 2020). Victims of crimes that occur in England, Wales and Scotland are eligible to claim through CICA. To estimate the administration costs associated with England and Wales only, the analysis multiplies the total CICA administration cost by the proportion of the total value of CICA pay-outs that were made to victims in England and Wales in 2019 to 2020. This results in a total administration cost of £14.3 million for England and Wales. To apportion this by crime type, a fixed cost for homicide is assumed by dividing the total administration cost by the number of CSEW crimes. This assumes that death claims demand less resource to administer than claims involving multiple or complex injuries as advised by industry experts.

For the remaining budget, the analysis calculates the proportion of the total physical and emotional harms costs attributed to each crime type using PRC volumes and applies the proportions to the CICA budget. PRC volumes are used to calculate the total physical and emotional harms cost to reflect the CICA requirement that victims reported the incident to the police, and therefore reflects the costs associated with the crimes that would meet the scheme’s core eligibility criteria.

The methodological update to estimate costs using the physical and emotional harms reflects discussions with experts in the insurance field, who advised that claims involving more severe, complex, and long-term injuries, are likely to involve ongoing administration activities and therefore incur increased administration costs. The analysis therefore assumes that the value of physical and emotional harms reflects the value of a CICA claim, with the value of the claim related to their administration costs.

4.2.4 Insurance administration costs for fraud and cybercrime

Background

The analysis uses an updated methodology to estimate the insurance administration cost of fraud due to improvements in the data landscape. It uses UK Finance data to estimate the total reimbursement paid to fraud victims where this was previously estimated using data from the CSEW. The cybercrime estimate follows the same approach to ESCC 2018, except for the input used for the average reimbursement per victim. This is due to the data being experimental CSEW data in 2015 to 2016. Instead, this analysis assumes the unit cost for the value of property stolen or damaged is representative of the average reimbursement, as this would ultimately inform the value of a claim.

Cybercrime

The CSEW shows the number of cybercrime incidents resulting in loss by whether losses were fully reimbursed or were not or only partially reimbursed. The data shows that for incidents resulting in loss, no incidents were fully reimbursed in 2019 to 2020 (ONS, 2022d). To estimate the value of partial reimbursements for cybercrime incidents, the analysis multiplies the number of cybercrime incidents resulting in loss by the proportion of fraud victims who received a partial refund. Data from the CSEW on fraud reimbursements is used due to lack of data on the proportion of partial reimbursements specific to cybercrime. This is then multiplied by the value of property stolen or damaged for cybercrime with an assumed 50% reduction due to these being partial reimbursements.

Due to lack of data specific to the administration costs associated with cybercrime reimbursement, the same ratio of consequence administration costs to the value of claims as motor and property insurance is applied to the value of reimbursements. Anticipation administration costs are not included in this estimate, as insurance products protecting against cybercrime are not currently available for individuals. Instead, it is assumed that the anticipation administration costs associated with protection against cybercrime are captured in defensive expenditure estimates. This results in a total administration cost of £7,630 for cybercrime.

Fraud

The analysis uses data from UK Finance (UK Finance, 2022) on unauthorised fraud losses and authorised fraud reimbursements to estimate the total reimbursement paid to fraud victims in 2019 to 2020. Due to only calendar year data being available, the analysis sums together 2019 and 2020 values for card fraud losses and remote banking losses to estimate the reimbursement paid to individual victims of unauthorised fraud, assuming that all incidents are fully reimbursed. It does not include cheque fraud due to the majority of these losses being attributed to fraud against businesses. The analysis then adjusts the UK figure by population to estimate the reimbursement cost attributable to England and Wales. This results in a total reimbursement figure of £830.5m.

To estimate the cost of administering reimbursement, the analysis multiplies this figure by the same ratio of consequence administration costs to the value of claims as motor and property insurance due to the lack of data specific to fraud. Similarly to cybercrime, the model only accounts for consequence administration costs in estimates. Legal protection against losses for victims of unauthorised fraud means that insurance products are not available for individuals to purchase. However, institutions such as banks invest in defensive expenditure to prevent reimbursement claims. The time and resources spent to minimise reimbursement claims (the anticipation administration costs) are therefore captured in defensive expenditure estimates. This results in a total administration cost of £71.4 million.

Table 10: Total unit costs of insurance administration

Crime type	Insurance administration
Individual
Homicide	<£5
Violence with injury	£10
Violence without injury	£10
Rape	£10
Other sexual offences	<£5
Robbery	£290
Domestic burglary	£310
Theft of a vehicle	£630
Theft from a vehicle	£20
Theft from the person	£260
Criminal damage	£270
Fraud	£20
Cybercrime	<£5
Commercial (9 sectors only)
Commercial robbery	£40
Commercial burglary	£50
Commercial theft	£40
Theft of a commercial vehicle	£700
Theft from a commercial vehicle	£40
Commercial criminal damage	£40

5. Costs as a consequence of crime

This chapter discusses the methodological approach and outlines the estimates of the costs as a consequence of crime. These include:

• property stolen and damaged

• physical and emotional harms to the victim

• lost output (both lost productivity and time off work)

• health services

• victim services

5.1 Property stolen or damaged

5.1.1 Approach

The value of property stolen or damaged represents a cost to the victim of a crime. In order to calculate the unit cost of this for different crime types, information is required on the average value of property that is stolen or damaged. For crimes against the individual, this information is provided using data from the CSEW. For crimes against the business, this information is provided using data from the CVS.

As these estimates are based on survey responses, it is possible that the values of property stolen and damaged are biased as respondents may tend to value the property either at the price they originally paid or at the cost of replacing the item. Given that most property loses value over time this is likely to be an overestimate of what the property is worth at the point at which it was stolen or damaged. Additionally, respondents of the CVS are primarily asked to discuss the value of property lost in the most recent crime against their business. Respondents may instead discuss the most memorable crime against their business. As these memorable crimes are likely to be those which have a higher value of property stolen or damaged, this may also lead to an overestimate of property lost.

These estimates reflect the value of property stolen, but not the value of property subsequently recovered. It is possible that respondents account for this themselves within their survey responses, but as it is not explicitly asked in the surveys, it is assumed that recovered costs are not accounted for.

For both crimes against the individual and businesses, the mean estimate of the value of property stolen or damaged provided by the survey results is used.

Value of property stolen or damaged: crimes against individuals

The CSEW data provides mean estimates for the value of property stolen or damaged for most crime types which involve some form of theft of, or damage, to a property. These estimates also account for those victims who do not have any items stolen and/or damaged, or who value their property at £0.

For most crime types which fit this description, the provided mean unit cost has been used to inform the unit cost for this analysis and to calculate the total cost. However, this differs for 3 crime types:

Robbery

The CSEW does not provide a mean loss for Robbery due to the low sample size of respondents. The analysis therefore uses an average of the mean loss for the value of stolen property for domestic burglary in a dwelling and Theft from the person as a proxy. These 2 crime types are used as a proxy due to the similarities in items stolen, location of crime, and level of threat used by the offender, to a robbery offence.

Fraud

The CSEW does not provide a mean value of financial loss suffered by victims of fraud. To estimate the total cost of fraud, the analysis adopts a different approach to most other crime types. The CSEW separates losses into bands based on the amount lost (for example, £50 to £99). By multiplying the proportion of incidents that fall into each band by the total volume of fraud, the number of incidents in each band is estimated. These are then multiplied by the midpoint of their respective bands (for example, £74.50 in the £50 to £99 band; the highest band of £40,000+ assumes a loss of £60,000, in line with the increase in loss per incident between bands), and summed, to estimate the total loss. This method ensures the broad, variable nature of fraud losses are accounted for using a robust estimate of how the volume of frauds differ across degrees of loss. This total cost is divided by the total estimated volume of fraud offences, to estimate the unit cost of fraud.

Cybercrime

To estimate the total cost of cybercrime, the analysis adopts a similar approach as for fraud. As the CSEW does not provided data on the financial loss from computer misuse incidents, the financial loss suffered by victims of computer viruses. The CSEW separates losses into bands based on the amount lost (for example, up to £50). By multiplying the proportion of incidents that fall into each band by the total volume of cybercrimes, the number of incidents in each band is estimated. These are then multiplied by the midpoint of their respective bands (for example, £25.50 in the up to £50 band), and summed, to estimate the total loss. This total cost is divided by the total estimated volume of cybercrime offences, to estimate the unit cost of cybercrime.

Table 11: Value of property stolen or damaged in crimes against individuals

Crime type	Value of property stolen	Value of property damaged	Total unit cost
Robbery	£1,600	-	£1,600
Domestic burglary	£2,350	£730	£3,080
Theft of a vehicle	£13,480	£360	£13,830
Theft from a vehicle	£380	£360	£740
Theft from the person	£340	-	£340
Criminal damage	-	£440	£440
Fraud	£840	-	£840
Cybercrime	£60	-	£60

Value of property stolen or damaged: crimes against businesses

The CVS provides mean estimates of the value of items stolen and damaged in the most recent incident of crime, both by crime type and by sector. As the CVS only surveys a few sectors each year, surveys carried out between 2012 and 2018 are used in order to source enough data to estimate a unit cost for every crime type and each sector, where available. As all survey cost estimates are from prior to 2019 to 2020, all costs have been inflated to account for price changes driven by inflation. As the data is from a wide range of years, there may be inconsistencies in the recording of survey responses over time and the value of property stolen or damaged today may differ greatly from the value in 2012, even when inflation is accounted for. Both of these caveats could mean that unit costs are over or underestimated when used to reflect 2019 to 2020 costs.

After sourcing the most up to date estimate for the value of property stolen or damaged for each crime type and sector, and accounting for inflation, the costs are then apportioned based on the total number of incidents of crime in each sector. This allows for a single unit cost per crime type to be estimated, which are presented in Table 12, rather than by sector, whilst weighting unit costs towards those sectors that are victims to the most crimes.

Table 12: Value of property stolen or damaged in crimes against businesses

Crime type	Unit cost
Commercial robbery	£1,080
Commercial burglary	£4,340
Commercial theft	£380
Theft of a commercial vehicle	£22,390
Theft from a commercial vehicle	£1,510
Commercial criminal damage	£560

5.2 Physical and emotional harms to the victims

5.2.1 Approach

Some victims of crime, especially violent crime, will suffer substantial physical and emotional harms. To quantify this cost to victims of crime, a Quality Adjusted Life Year (QALY) approach is applied. The QALY approach was adapted for the context of crime first by Dolan et al. (2005).

For an injury experienced, the QALY approach applies an assumed negative percentage impact on a person’s quality of life caused by experiencing different injuries. For example, if a person breaks their rib, they are estimated to suffer around a 10% reduction in their quality of life whilst they are recovering (QALY reduction as reported in Salomon et al., 2015).

Box 2 outlines how the cost of the injury as a result of being a victim of being a crime is calculated within the analysis.

Box 2:

LIKE – Likelihood of sustaining a specific physical and/or emotional injury

REDUCEQL – Percentage reduction in quality of life for that physical/emotional injury

DUR – Duration of the injury as a fraction of a total year

VOLY – Value of a year at full health

For each injury (i), for each crime type (c):

LIKE_i,c * REDUCEQL_i,c * DUR_i * VOLY = Average physical and emotional cost_ic

The QALY approach has been applied previously when calculating the economic and social costs of crime. For example, this approach was applied for violent crimes by Dubourg et al. (2005), who based their estimates on the research of Dolan et al. (2005). However, for non-violent crimes, Dubourg et al. (2005) continued with the methodology of Brand and Price (2000) where costs are calculated by asking victims to estimate the amount of money that they feel is necessary to compensate them for the inconvenience and physical and emotional harms they suffered as a result of the crime. This approach is potentially biased by individuals being unable to accurately estimate the value of the specific long-term emotional impacts they have suffered. Each of the components of the QALY approach will be discussed in more detail below to show how the estimates have been developed.

5.2.2 Prevalence of harms

The prevalence of harms indicates how likely a victim is to incur each type of physical and emotional harm as a result of being victim of an incident of crime. The CSEW survey for 2019 to 2020 produces estimates of the proportion of victims who experience various harms for each crime type against an individual, with the exception of rape and other sexual offences. Table 13 indicates the physical and emotional harms costed. Not all emotional harms that are stated in CSEW nature of crime tables are estimated in the analysis due to a lack of QALY reduction and duration estimates. The impact of this is unknown, however, duration, severity, and impact of the emotions such as ‘shock’ is expected to be low.

CSEW data on harms experienced by victims of rape and other sexual offences is collected via the self-completion module and does not contain data on all harms costed in this analysis. Therefore, it is assumed the emotional harms experienced as a result of rape and other sexual offences are equivalent to those in ESCC 2018, which are taken from the CSEW. For physical harms, 2 approaches are applied to estimate the physical harms experienced for rape and other sexual offences based on data available from the CSEW.

Physical and emotional injuries incurred by victims of commercial crimes are not covered in the CVS. This analysis uses data from ESCC 2018 (where, due to a smaller sample size, results are aggregated based on 7 years of the CSEW, years ending 31 March 2009 to 2016) from the CSEW on harms experienced by victims where the incident took place in the workplace or a business. This assumes that the prevalence of harms has not changed from 2015 to 2016, to 2019 to 2020.

The prevalence estimates may not fully capture less common injuries resulting from crimes as they are not recorded in the CSEW (such as burns) or are not captured in the sample surveyed. This may result in the analysis underestimating the prevalence of certain harms.

Table 26 and Table 27 in Appendix 1 show the prevalence of harms from the data sources outlined above for crimes against individuals and businesses respectively.

5.2.3 Quality adjusted life loss associated with harm

To calculate the negative percentage impact to a person’s quality of life (QALY loss) from the different injuries incurred by victims, disability weights are used to reflect the relative severity of the injury where zero represents full health and one represents death. The disability weights are taken from Salomon et al. (2015). Table 13 shows the assumed QALY losses associated with the various physical and emotional harms of crime. Table 28, in Appendix 1, explains in more detail the assumed harms for each CSEW physical and emotional harm category. Where the injuries in Salomon et al. (2015) do not exactly match the injuries suffered as a result of crime (from the CSEW), an explanation of the alternative source is provided.

5.2.4 Duration of harm

The QALY loss is assumed to be constant over the recovery period for a given injury. To estimate the duration of injury, various data sources are utilised. These durations largely come from Dolan et al. (2005), however when there is no duration estimated in the report, alternative sources are used or assumptions are made (Table 29 in Appendix 1 outlines the specific sources for each of the duration assumptions). For example, to estimate the remaining years of life for homicide victims, the average age of an adult victim of homicide is subtracted from the average life expectancy in the UK.

Academic literature suggests that depression and anxiety suffered following violent crime is likely to affect the victim for longer than depression and anxiety following non-violent crime (Wasserman and Ellis, 2007; Dolan et al, 2005), and therefore should be reflected in the analysis. To account for this, durations are estimated based on academic papers looking at the depression and anxiety consequences of violent and non-violent crimes respectively. Some crimes such as robbery can be either violent or non-violent. Therefore, the duration of the emotional harms incurred as a consequence of crimes that can be violent or non-violent are assumed to be the average of the duration of violent crime emotional harms and the duration of non-violent crime emotional harms.

Where the duration of harm is greater than one year, the cost of the harm is discounted in accordance with The Green Book, Central Government Guidance on Appraisal and Evaluation (2020).

5.2.5 Resulting costs of the specified harms

To calculate the costs of each of the harms we multiply the QALY losses by the durations and the estimated value of a statistical life year (derived from the social value of a change in the probability of losing or gaining a year of life expectancy). This gives an estimate of the cost to the victim associated with suffering each of the physical and emotional harms incurred. A QALY is equal to one year of life in perfect health and is used as a measure of injury or disease burden. In other words, one QALY is the value to an individual of being at full health for one year. The QALY value given by HMT (2022) is £70,000, and this is used throughout the analysis in line with DHSC best practice. Table 13 shows the results.

Table 13: Estimated unit costs of physical and emotional harms

Injury	QALY loss	Duration (years)	Cost of harm (1)
Physical
Minor bruising or black eye	2.6%	0.0288	£50
Severe bruising	5.2%	0.0575	£210
Scratches	0.2%	0.006	£0
Cuts	0.6%	0.024	£10
Puncture or stab wounds	10.3%	0.115	£830
Broken bones	10.3%	0.115	£830
Nosebleed	0.6%	0.0027	£0
Broken nose	6.7%	0.059	£280
Broken/lost teeth	3.4%	0.0192	£50
Chipped teeth	1.7%	0.0192	£20
Dislocation	6.2%	0.2307692	£1,000
Concussion or loss of consciousness	11.0%	0.0335	£260
Internal injuries	5.2%	0.0575	£210
Facial or head injuries (no bruising)	0.6%	0.024	£10
Eye/facial injuries due to acid/paint thrown	5.4%	0.0192	£70
Other	0.8%	0.0192	£10
Emotional – violent crime (2)
Fear	3.0%	1.25	£2,620
Depression	14.5%	1	£10,150
Anxiety/panic attacks	13.3%	3	£27,520
Emotional – non-violent crime (3)
Fear	3.0%	1.25	£2,620
Depression	14.5%	0.167	£1,700
Anxiety/panic attacks	13.3%	0.167	£1,550
Emotional – semi-violent crime (4)
Fear	3.0%	1.25	£2,620
Depression	14.5%	0.58	£5,920
Anxiety/panic attacks	13.3%	1.58	£14,660
Emotional – rape specific
Drug abuse	47.9%	5	£162,770
Alcohol abuse	37.3%	5	£126,750
Obesity/eating disorder	22.4%	5	£75,950
Sexual dysfunction	1.7%	0.17	£200
Death
Death	100%	42.2	£2,247,260

Notes:

1. Where the duration of harm is greater than one year, the cost of the harm is discounted in accordance with The Green Book, Central Government Guidance on Appraisal and Evaluation (2020).

2. Violent crimes are assumed to be homicide, violence with injury and rape.

3. Non-violent crimes are assumed to be burglary, theft, criminal damage, fraud, and cybercrime.

4. Semi-violent crimes are assumed to be other sexual offences, robbery, and violence without injury.

5.2.6 Resulting costs of the harms for each crime type

The final step is to combine the costs of the various harms (Table 13) with the likelihood of suffering those harms (Table 26 and Table 27). This results in unit costs for each physical and emotional harm for each crime type. The costs are accumulated for each crime type, to arrive at a final physical and emotional harm unit cost. Table 14 below shows the estimated costs.

Table 14: Estimated physical and emotional costs of each crime type

Crime type	Total
Individual
Homicide	£2,247,260
Violence with injury	£9,060
Violence without injury	£3,850
Rape	£26,340
Other sexual offences	£4,210
Robbery	£4,760
Domestic burglary	£1,420
Theft of a vehicle	£590
Theft from a vehicle	£230
Theft from the person	£440
Criminal damage	£560
Fraud	£360
Cybercrime	£270
Commercial (9 sectors only)
Commercial robbery	£4,310
Commercial burglary	£530
Commercial theft	£0
Theft of a commercial vehicle	£370
Theft from a commercial vehicle	£100
Commercial criminal damage	£60

5.3 Lost output

5.3.1 Approach

Lost output aims to estimate the cost of lost economic activity to society as a result of an individual being a victim of a crime, for both crimes against individuals and crimes against businesses (where the victim is employed by or self-employed). The analysis does not include the lost output of individuals who are sentenced to prison or probation.

Victims of crime may be less productive following an incident of crime for multiple reasons. Firstly, their output may be reduced because of the injuries incurred, for example, a broken arm will naturally impair an individual’s ability to type. Secondly, the victim may experience reduced output due to damage to or loss of equipment that enables them to work.

Due to a lack of data availability and evidence, the analysis focusses on the lost output incurred via injuries sustained, and therefore may be underestimating costs for crimes such as cybercrime where a large part of lost output is likely to come from damage to technology. Naturally, a loss of work equipment may result in feelings which may be captured as the emotional impact of a crime, and as such may be partially captured in the analysis.

The analysis focuses on 2 separate causes of lost productive hours as a result of crime victimisation:

1) Time taken off work as a result of the crime.

This is based on CSEW respondents who report the amount of time taken off work following victimisation.

2) Reduced productivity at work as a result of physical and emotional injuries.

Victims of crime who suffer physical and emotional injuries are assumed to be less productive at work. The QALY loss associated with the injury (discussed in Section 5.2) is used as a proxy for the extent of their reduced productivity. The physical and emotional harms costs are assumed to not capture lost productivity, which is supported by Shiroiwa et al. (2013) who conclude that any double counting between QALYs and productivity loss is negligible.

Box 3:

Outlines how the lost output cost to society is calculated in the analysis:

REDUCPROD – Percentage reduction in productivity

DUR – Duration of the injury as a fraction of a total year

AFFHRS – The number of hours affected by lower productivity

EMPRATE – The employment rate

LIKE – Likelihood of sustaining a specific physical and/or emotional injury

For each injury (i) and for each crime type (c) the formula is as follows:

REDUCEPROD_i,c * DUR_i * AFFHRS_c * HRLYLABCOST * EMPRATE_c * LIKE_i,c = Average hours of reduced productivity

HRSOFFWORK – The average number of hours taken off work

HRSREDPROD – The average number of hours of reduced productivity

For crime type (c) the formula is as follows:

HRSOFFWORK_c + HRSREDPROD_c = Average lost productive hours

LOSTHRS – The lost productive hours

HRLYLABCOST – The mean hourly labour cost

For each crime type (c) - the formula is as follows:

LOSTHRS_c * HRLYLABCOST = Average lost output cost

5.3.2 Time taken off work

Data on the time taken off work by victims following a crime is primarily taken directly from ESCC 2018, which uses CSEW data from 2009 to 2009. This question is no longer asked in the CSEW; therefore, the input data has not been changed. Although there has been minimal change in the average hours worked for all employees (ONS, 2021c), using this data will ignore the impact of changes to the nature of employment, for example, the shift towards more zero-hour contracts. The lack of new data in this area may impact the accuracy of some crime types more than others.

The CSEW data from 2009 to 2009 does not contain data on the time taken off work by victims of fraud or cybercrime. Data on the time taken off work by victims of fraud and cybercrime is taken from the Home Office report on the experiences of victims of fraud and cybercrime (2025). To estimate how much time victims of fraud and cybercrime take off work, the analysis uses a different method than for other crime types. The Home Office report on the experiences of victims of fraud and cybercrime groups the time taken off into ranges (for example, one to 2 days). The midpoint of each range (for example, 1.5 days for one to 2 days) is used alongside the weighted number of respondents to calculate an average number of days off.

This average is then adjusted based on the likelihood of a victim taking time off, according to CSEW data, and converted into hours to estimate the average time taken off work by victims of fraud and cybercrime. This approach ensures the wide range of experiences among victims are reflected.

The ‘lost hours’ outlined in Table 15 display variation across crime types in the average time taken off work following an offence. This may be driven by various factors, such as the severity of the crime, and the occupation the victim is employed in.

Table 15: Lost productive hours due to time taken off work by crime victims

Crime type	Lost hours (average time taken off work) following an offence
Individual
Homicide (1)	18,460.5
Violence with injury	4.9
Violence without injury	0.8
Rape	137.7
Other sexual offences	13.8
Robbery	2.7
Domestic burglary	4.0
Theft of a vehicle	3.8
Theft from a vehicle	0.8
Theft from the person	1.1
Criminal damage	0.5
Fraud	6.3
Cybercrime	4.0
Commercial (9 sectors only)
Commercial robbery	21.5
Commercial burglary	8.3
Commercial theft	-
Theft of a commercial vehicle	1.4
Theft from a commercial vehicle	1.7
Commercial criminal damage	0.4

Notes:

1. Lost hours from homicide are calculated as the annual number of hours worked by the average employed person, multiplied by the discounted average life expectancy of a homicide victim before retirement, multiplied by the average employment rate for victims of violent crime to give the total lost hours to society.

5.3.3 Reduced productivity

Beyond the time taken off work, a victim of crime may experience reduced productivity when they return to work whilst they are still recovering from the physical and/or emotional effects of the crime.

Reduced productivity is estimated by calculating the average number of hours affected by lower productivity and multiplying these by the assumed reduced productivity. The number of hours affected by lower productivity is combined with the likelihood of a victim suffering each harm to calculate the expected number of productive hours lost by victims of each crime type.

Hours of productive labour lost for each harm

To calculate the number of hours of productive labour lost as a result of victims suffering from physical and emotional harms, the following information is required:

• productivity loss associated with each physical and emotional harm

• the duration of the productivity loss

• the average number of hours worked by an individual in employment

• the employment rate for victims of crime

The first step is to estimate the specific productivity loss associated with each harm. There is no available data on the reduction in productivity resulting from crime. We therefore use the values presented in Section 5.2 on the QALY losses of physical and emotional harms as a proxy for the reduction in productivity resulting from the crime.

To estimate the number of hours that would be affected by productivity loss, it is assumed that the productivity loss would be incurred over the same duration as the harm, taken from estimates in Section 5.2. If the duration of the harm is shorter than the hours taken off work, then there is no reduced productivity beyond the hours taken off work. If the duration of the harm is longer than the hours taken off work, then the victim is assumed to work at a lower productivity for the remainder of their injury recovery period.

The mean weekly number of hours worked by an individual in some form of employment is estimated to be 32.7 hours (ONS, 2021c), which is the equivalent of approximately 1,705 hours annually. The employment rate for victims of each crime type is then applied to calculate the expected number of hours worked. This is estimated separately for each crime type, using the employment rate for crime victims sourced from the CSEW and accounting for the respective number of victims at working age. Implicitly this assumes the flow of victims into and out of employment over the duration of lost output is constant.

Likelihood of harm being suffered

The next step is to multiply the productivity losses by the likelihood of a person suffering each of the harms following victimisation of different crime types. These likelihoods (shown in Appendix 1) are used to create estimates of the number of hours of lost output because of reduced productivity from crimes and are displayed in Table 16. To avoid double counting, the reduced productivity is only applied to time when the victims had returned to work following the crime.

Table 16: Lost productive hours due to reduced productivity when returning to work after a crime

Crime type	Lost hours (reduced productivity) following an offence
Individual
Violence with injury	151.2
Violence without injury	62.0
Rape	378.7
Other sexual offences	33.9
Robbery	81.1
Domestic burglary	23.9
Theft of vehicle	10.0
Theft from a vehicle	4.0
Theft from the person	7.5
Criminal damage	9.5
Fraud	6.1
Cybercrime	4.5
Commercial (9 sectors only)
Commercial robbery	103.4
Commercial burglary	12.8
Commercial theft	-
Theft of a commercial vehicle	9.1
Theft from a commercial vehicle	2.5
Commercial criminal damage	1.5

5.3.4 Unit cost of lost output

The final step is to combine the lost hours from time off work with those from reduced productivity to understand the impact on working hours. The number of working hours lost is then multiplied by the average hourly labour cost to calculate the lost output costs of each crime type. It is assumed that the average wage of employed victims of crime is the same as the national mean wage for all employed individuals taken from the UK-wide dataset, ASHE (ONS, 2022e), this is then uplifted to the average labour cost by accounting for non-wage costs.

Table 17 shows the final estimates of the lost output costs of crimes against individuals and businesses.

Table 17: Average cost of lost output for crime victims

Crime type	Hours off work	Reduced productive hours post return to work	Total hours lost	Total lost output
Individual
Homicide			18,460.5	£383,630
Violence with injury	4.9	151.2	156.0	£3,240
Violence without injury	0.8	62.0	62.7	£1,300
Rape	137.7	378.7	516.3	£10,730
Other sexual offences	13.8	33.9	47.6	£990
Robbery	2.7	81.1	83.8	£1,740
Domestic burglary	4.0	23.9	27.9	£580
Theft of vehicle	3.8	10.0	13.8	£290
Theft from a vehicle	0.8	4.0	4.7	£100
Theft from the person	1.1	7.5	8.6	£180
Criminal damage	0.5	9.5	10.0	£210
Fraud	6.3	6.1	12.4	£260
Cybercrime	4.0	4.5	8.5	£180
Commercial (9 sectors only)
Commercial robbery	21.5	103.4	124.9	£2,590
Commercial burglary	8.3	12.8	21.1	£440
Commercial theft	-	-	-	-
Theft of a commercial vehicle	1.4	9.1	10.5	£220
Theft from a commercial vehicle	1.7	2.5	4.2	£90
Commercial criminal damage	0.4	1.5	1.9	£40

5.4 Health services

5.4.1 Approach

Health services costs aim to capture the costs to the NHS and other healthcare providers of responding to the physical and emotional harms of crime. Using NHS costs in this analysis may lead to an underestimate of the average costs if an individual was to use private healthcare. If fewer people were injured through crime, the resources used to treat them could be used in alternative activities. Therefore, costs to health services as a consequence of crime are an opportunity cost and therefore a cost to society of crime.

The estimates of health services costs are based on assumptions about the treatment that is likely to be required for certain injuries and the prevalence of these injuries (estimated using the CSEW). Physical harms are associated with ambulance and medical procedure costs, and the emotional harms from violent crimes are associated with counselling costs. The unit costs of healthcare activities used are from Curtis and Burns (2020) and NHS Reference Costs (NHS, 2021). This methodology is very similar to that used in the previous costs of crime publications (ESCC, 2018; Dubourg et al., 2005; and Brand and Price, 2000).

5.4.2 Medical requirements associated with injuries

The harms suffered as a result of crime are mapped to medical procedures that are assumed to be needed for such harms. For example, where a victim who suffers a broken nose requires medical attention, the type of medical attention needed is assumed to be a ‘Nose Procedure’ (see Table 18). The proportion of people who require medical attention following an injury is estimated using bespoke CSEW data for 2019 to 2020. In addition, the CSEW asks respondents whether they required medical attention from a paramedic after suffering a particular physical harm as a result of crime. This is used to estimate the proportion of people who required an ambulance following various harms. The average number of hours of physiotherapy and counselling required are based on assumptions made by Dubourg et al. (2005).

The proportion of victims who require the attendance of an ambulance, particular medical procedures, physiotherapy, and counselling following an injury are shown in Table 18.

Table 18: Proportion of victims requiring medical treatment following an injury and hours of physiotherapy and counselling required

Harm suffered (1) (2)	Ambulance	Bone fracture (3)	Other injury	Nose procedure	Strain, sprain, or minor open wound	Lowest cost head injury	Minor dental restoration procedure	Minor dental procedure	Physiotherapy (hours)	Counselling (hours)
Broken bones	23%	79%	-	-	-	-	-	-	10	-
Severe bruising	26%	-	55%	-	-	-	-	-	-	-
Puncture/stab wound	40%	-	86%	-	-	-	-	-	-	-
Internal injury	46%	-	100%	-	-	-	-	-	-	-
Broken nose	8%	-	-	54%	-	-	-	-	-	-
Cuts	10%	-	-	-	38%	-	-	-	-	-
Dislocation	-	-	-	-	47%	-	-	-	-	-
Concussion	48%	-	-	-	-	78%	-	-	-	-
Lost teeth	-	-	-	-	-	-	60%	-	-	-
Chipped teeth	-	-	-	-	-	-	-	54%	-	-
Scratches	5%	-	-	-	-	-	-	-	-	-
Minor bruising	3%	-	-	-	-	-	-	-	-	-
Facial injury	-	-	-	-	7%	-	-	-	-	-
Eye injury	-	-	-	-	-	-	-	-	-	-
Nosebleed	-	-	-	-	-	-	-	-	-	-
Other	-	-	-	-	62%	-	-	-	-	-
Fear	-	-	-	-	-	-	-	-	-	2
Depression	-	-	-	-	-	-	-	-	-	20
Anxiety/panic attacks	-	-	-	-	-	-	-	-	-	25
Drug abuse	-	-	-	-	-	-	-	-	-	50
Alcohol abuse	-	-	-	-	-	-	-	-	-	50
Sexual dysfunction	-	-	-	-	-	-	-	-	-	2

Notes:

1. ’-‘ indicates that there is assumed to be no medical requirement associated with the injury.

2. For some harms suffered, sample sizes are small which might mean the estimated prevalence of treatment for the related injuries may not capture the medical treatment required for the more uncommon and serious injuries. This may result in an underestimate of the medical requirements following injury.

3. Some broken bones (such as a broken toe) are not necessarily treatable, meaning that not all harms related to broken bones require treatment.

5.4.3 Costs of medical requirements

The next step is to estimate the costs associated with the various medical requirements. The hourly cost of counselling and the hourly cost of physiotherapy are based on Curtis and Burns (2020). These hourly costs are multiplied by the average number of hours of counselling and physiotherapy per injury (taken from Dubourg et al 2005) to give average health service costs associated with the harms. To estimate the health service costs associated with the other harms, such as a broken nose, the unit cost of the procedure to treat that harm is multiplied by the proportion of victims who require treatment after suffering that harm (from Table 18) to produce the expected cost of this procedure associated with this harm. The expected costs for all the procedures associated with a harm are summed to get the average cost of medical requirement associated with the harm. The results of this are shown in Table 19.

Table 19: Unit costs of health care associated with physical and emotional harms

Injury	Average cost of medical requirement
Broken bones	£2,470
Severe bruising	£740
Puncture/stab wound	£1,160
Internal injury	£1,350
Broken nose	£740
Cuts	£390
Dislocation	£450
Concussion	£740
Lost teeth	£180
Chipped teeth	£100
Scratches	£10
Minor bruising	£10
Facial injury	£70
Eye injury	-
Nosebleed	-
Other	£590
Fear	£100
Depression	£980
Anxiety/panic attacks	£1,230
Drug abuse	£2,450
Alcohol abuse	£2,450
Sexual dysfunction	£100

5.4.4 Costs associated with crimes

The unit health services costs for each harm are translated into costs associated with each crime type. For each crime type, the expected health service costs of the physical and emotional harms are multiplied by the proportion of victims of that crime type who suffer those harms. The proportions of victims who suffer harms as a consequence of each individual and commercial crime type are outlined in Section 5.2. For example, the unit health services cost of violence with injury is £1,060 (see Table 20), which is the sum of the expected health services costs of all injuries suffered as a result of the crime.

To arrive at the average health services costs of the injuries associated with each crime, the unit health services cost of each injury is calculated first. An example of how the unit health services cost of an injury is calculated is given in Box 4.

Box 4:

Example of how the unit health services cost for the injury of a broken nose is calculated:

i. The cost of an ambulance per episode is £260 and the proportion of crime victims who require an ambulance for a broken nose is 8% (Table 18). Therefore, £260 is multiplied by 8% to get the expected ambulance costs for a broken nose. The average cost of a nose procedure is £1,330 and the proportion of people who require a nose procedure for a broken nose is 54%. Therefore, £1,330 is multiplied by 54% to get the expected nose procedure costs for a broken nose.

ii. To estimate the average cost of a broken nose as a result of violence with injury, we must multiply the unit injury cost ((£260 x 8%) + (£1,330 x 54%)) by the proportion of violence with injury victims that suffer a broken nose, which is 2% (Table 26). This gives the expected health services cost of a broken nose as a result of violence with injury, which is approximately £15.

iii. The expected cost of a broken nose as a result of violence with injury is then added to the average costs for all other injuries suffered as a result of violence with injury, which are calculated in the same way as the broken nose cost.

This method should not be used to calculate the health services costs for homicide. The Department for Transport health service cost associated with fatal injuries (Department for Transport, 2020a) is therefore used instead to estimate the health services costs of homicide.

Table 20: Unit costs of health care associated with each crime type

Crime type	Unit cost
Individual
Homicide	£1,230
Violence with injury	£1,060
Violence without injury	£320
Rape	£990
Other sexual offences	£400
Robbery	£450
Domestic burglary	£430
Theft of a vehicle	£250
Theft from a vehicle	£60
Theft from the person	£150
Criminal damage	£220
Fraud	£140
Cybercrime	£90
Commercial (9 sectors only)
Commercial robbery	£690
Commercial burglary	£150
Commercial theft	£0
Theft of a commercial vehicle	£10
Theft from a commercial vehicle	£0
Commercial criminal damage	£40

5.5 Victim services

5.5.1 Approach

Victim services provide support to the direct victims of crime, their friends, and families. Without crime, resources and time spent on providing support to victims could be used for other public services and activities. Therefore, victim services costs represent a cost to society as a result of crime.

There are 2 types of costs associated with victim services; the opportunity cost of time spent supporting victims of crime and the expenditure on victim services in England and Wales. There is limited publicly available data on the number of hours volunteers spend supporting victims of crime, making it difficult to estimate a robust opportunity cost that is reasonably representative of their time. Additionally, some organisations have shifted from volunteer to employee facilitation of victim services. Therefore, some of the opportunity cost associated with paid employees are captured in the total victim services cost. Therefore, in contrast to ESCC 2018, the opportunity cost of volunteer time is not included in this analysis.

The analysis estimates the costs associated with public expenditure on victim services that can be used to support the direct victims of all relevant crime types in England and Wales in 2019 to 2020, based on publicly available data from the Ministry of Justice (MoJ) only. This comprises MoJ funding for PCCs to commission victim services in their local area and funding for Citizens Advice to deliver the National Witness Service.

Homicide is considered separately in the analysis as the victim services costs for homicide are wholly associated with indirect victims, for example, supporting the homicide victim’s friends and family. The victim services cost of homicide is therefore estimated using Victim Support data on MoJ funding for the National Homicide Service.

5.5.2 Expenditure on victim services

In 2019 to 2020 the MoJ provided around £63.2 million of funding to PCCs to commission support services for victims in their area (MoJ, 2019). The MoJ provided an additional £11.2 million to Citizens Advice to deliver the National Witness Service (MoJ, 2019). As the National Witness Service supports both direct and indirect victims, the analysis estimates the expenditure spent on supporting direct victims only. In 2019 to 2020, 65% of National Witness Service users were the direct victim of the crime they were giving evidence for (Citizens Advice, 2021). Therefore, the analysis multiplies the total MoJ funding for the National Witness Service, £11.2 million, by 65% to estimate a cost of £7.3 million for direct victims. It adds this to the MoJ funding for PCCs to commission support services, £63.2 million, resulting in a total expenditure of approximately £70.4 million.

Table 21: Breakdown of victim services funding

Funding Stream	Funding (£)
MoJ funding - PCCs	£63.2 million
MoJ funding – National Witness Service	£7.3 million
Total for direct victims (1)	£70.4 million
Victim Support - National Homicide Service	£3.4 million
Total for indirect victims	£3.4 million

Notes:

1. The numbers do not add up due to rounding.

There are multiple factors that influence service demand and the costs associated with each victim. These factors include the emotional harm experienced and the practitioner type providing support (for example, paid professional or a volunteer); however, there is a lack of evidence to link these factors to crime types.

In the absence of this data, the analysis uses the relative emotional harms (as in Section 5.2) to split the victim service budget by offence type. The emotional harm cost for each crime type is multiplied by the number of crimes committed. The proportion of the total emotional harm cost in 2019 to 2020 attributed to each crime type is then assumed to be the equivalent proportion of the expenditure spent on each crime type. This approach is guided by 2 core assumptions. The first is that crime types with a high prevalence are likely to have a higher volume of victims seeking support. The latter is that crime types with greater emotional harms are likely to require support for a longer period of time.

Finally, the analysis estimates the total expenditure on homicide using data from Victim Support. In 2019 to 2020, the MoJ provided £3.4 million to the charity Victim Support to deliver the National Homicide Service which supports people bereaved by murder and manslaughter in England and Wales (Victim Support, 2020). These aren’t direct victims but, given the nature of homicide and the large impacts on victims’ families and friends, is captured in this cost.

5.5.3 Unit costs for victim services

The unit costs of victim services are shown below in Table 22. These are estimated by dividing the total estimated cost for each crime type by the number of crimes committed (both reported and unreported).

Table 22: Unit costs for victim services

Crime type	Unit cost
Individual
Homicide	£5,070
Violence with injury	£20
Violence without injury	£10
Rape	£40
Other sexual offences	£10
Robbery	£10
Domestic burglary	<£5
Theft of a vehicle	<£5
Theft from a vehicle	<£5
Theft from the person	<£5
Criminal damage	<£5
Fraud	<£5
Cybercrime	<£5
Commercial (9 sectors only)
Commercial robbery	£10
Commercial burglary	<£5
Commercial theft	-
Theft of a commercial vehicle	<£5
Theft from a commercial vehicle	<£5
Commercial criminal damage	<£5

6. Costs in response to crime

This chapter discusses the methodological approach and outlines the estimates of the costs in response to crime. These include:

• police costs

• CJS costs

Not all crimes are reported or result in charges. As a result, the response unit costs captured are weighted by the probability that the response occurs. This methodology reflects the cost given the likely outcome of the crime, such that the unit cost can be applied to volume changes in all crimes as estimated by the CSEW (reported and unreported).

The unit cost per recorded crime can be calculated by dividing the total response costs by the volume of PRC.

6.1 Police costs

This section discusses the methodological approach and outlines the estimated costs of the police response to crime. This section also outlines the limitations of the approach and alternative information that can be used to inform the costs to the police of responding to crime.

6.1.1 Approach

The approach to calculating police costs maintains the approach taken in the ESCC 2018, which primarily uses ‘activity based costing’ (ABC) data from 2006 to 2007, explained within this section. A review of this approach was undertaken, including the possibility to explore potential drivers of ABC over time (for example, police recorded crime, charges, and days to outcome data) and use these to uplift and adjust the 2006 to 2007 data to reflect police activity in 2019 to 2020. However, it was deemed that maintaining the approach from the ESCC 2018 with minimal updates was still the most accurate method, despite being outdated. Information on how an approach using more up to date data sources could be established is detailed in Section 6.1.2.

ABC data

As with Dubourg et al. (2005), the costs associated with different crimes are based on ABC data. The data, provided by police forces, splits out their budget into the various different activities they performed (both crime and non-crime). These activities include a breakdown of the estimated cost of dealing with different crime types and associated overheads. All non-crime police spend is excluded from this analysis.

The last year for which reliable ABC data is available is 2006 to 2007 and as a result the costs require some adjustment to ensure that they are more reflective of the resources that the police spent dealing with and investigating crime in 2019 to 2020. The following section outlines how this was done in more detail.

Adjusting ABC data

The ABC costs data are full year estimates of cost type against activity type, based on the grossed-up figures from the 2-week survey data. The cost types are staff, operational support, and business support. The activities were organised into a series of domains (investigating crime, providing assistance, reducing crime, and promoting public safety). Within the ‘investigating crime’ and ‘providing assistance’ domains, spending on different types of crime is identified.

The following steps are undertaken to arrive at an uprated estimate of the police cost per crime in 2019 to 2020:

1. Crime categories from the original ABC classifications are identified.

2. The total direct annual cost of dealing with each crime is estimated using 2006 to 2007 data. The total cost figures are the sum of direct costs for staff, operational support, and business support.

3. In addition to the ‘direct’ crime costs, it is necessary to include some expenditure from the ‘investigating crime’ and ‘providing assistance’ domains which is not collected against individual crime and incident types. These cover costs that are ‘upstream’ of police attendance (call handling and control room) and those that are ‘downstream’ (for example, prisoner handling and custody duties, informants, ID parades and family liaison). The allocation of upstream/downstream costs across the two domains is relatively crude and represents an oversimplification of reality. For example, some ‘providing assistance’ costs (call handling and family liaison), though by definition about ‘providing assistance’, will partially relate to, and vary in accordance with individual crimes. Within each domain, the total upstream and downstream costs are allocated proportionately on the basis of the crime share of total cost. This gives an adjusted figure of total cost by specified crime type for 2006 to 2007.

4. Overhead costs should be included in the unit costs of crime incidents, so that cost estimates are not underestimated. It is assumed that overhead costs should be added onto the incident costs proportionately to the incident costs of each activity domain. That is, if ‘criminal damage’ accounts for 10% of the ‘investigating crime’ incident costs, it is assumed that 10% of the ‘investigating crime’ overhead costs should be added on top.

The costs per ABC category in 2006 to 2007, with the adjustments to account for overheads, can be seen in Table 23 below:

Table 23: Costs per ABC Category in 2006 to 2007

Investigating crime categories	Costs without overheads (£ million, 06/07)	Percentage	Costs with overheads (£ million, 06/07)	Percentage
Violence against the person – S20 and more serious	£554 million	13	£653 million	13
Violence against the person – S47 and less serious	£468 million	11	£551 million	11
Sexual offences	£294 million	7	£347 million	7
Burglary dwelling	£353 million	9	£416 million	9
Burglary – commercial and other	£181 million	4	£213 million	4
Robbery	£245 million	6	£289 million	6
Theft of or from a motor vehicle	£270 million	7	£319 million	7
Deception/fraud	£184 million	4	£216 million	4
Theft other	£356 million	9	£420 million	9
Drug offences	£357 million	9	£421 million	9
Criminal damage	£255 million	6	£301 million	6
Other crime	£602 million	15	£710 million	15
Investigating crime totals	£4.12 billion	(100)	£4.857 billion	(100)

5. To calculate the average cost per incident in 2006 to 2007, the adjusted total costs for each crime type are divided by the respective numbers of offences, using the crime categories in use at that time.

   However, between the years ending 31 March 2007 and 2020, there had been a number of changes to how crimes were recorded. The most significant change for this work was the groupings of offences codes within crime types changed. For example, some categories have changed name and scope – ‘more serious and ‘less serious violence in 2006 to 2007 have been recalibrated into ‘violence with injury’ and ‘violence without injury’ in 2019 to 2020. It was therefore decided that cost estimates should try and account for these changes.

   Home Office median days to outcome data (2014 to 2015) are used to estimate more granular costs for 2006 to 2007. The data is used to quantify the relative burden on forces for different crime types (to the extent that burden on the police per crime is relative to elapsed time between crimes being recorded and an outcome being assigned to them).

6. Due to changes in the classification of some offences, it is necessary to recalibrate 2006 to 2007 offence categories against their 2019 to 2020 counterparts. For example, blackmail was then categorised as a ‘miscellaneous other’ crime, but in 2019 to 2020 was categorised as a theft offence.

   ABC categories in 2006 to 2007 are split up by estimating the proportion of the total cost that would be spent on the subcategories within it. The estimates are made by weighting the volumes of crimes in each subcategory by their relative median days to outcome. The volume of these crime groups are weighted by the ratio of these durations, to give an estimate for the amount of the cost in the ABC category that could be expected to relate to each subcategory. These subcategory costs can then be recalibrated to match the 2019 to 2020 crime types.

7. Once cost estimates for categories as defined in 2019 to 2020 have been calculated in this way, it is then necessary to calculate a unit cost for each crime type. This involves dividing the cost estimates for each crime type by the corresponding volumes (from PRC) to give a cost per incident.

8. Then, the unit costs must be adjusted to reflect changes since 2006 to 2007. To do this, 2 methods were considered for how to adjust the cost figures.

   i. Adjust by inflation

   A simple way to account for changes over time would be to adjust by inflation. However, doing this assumes that the unit cost is independent of demand on the police over time, and of the police budget. That is, assuming that unit costs remain essentially constant might lead to, in cases where volumes have changed dramatically, notional national police spend greater than the known police budget. It is therefore useful to reference the police budget in some way. As such, an alternative method is used.

   ii. Uprate to the police budget

   An alternative method is used to account for the shift in volumes over time, labelled here as ‘ uprating’, to calculate 2019 to 2020 unit costs. Normally to calculate a unit cost, a cost and a volume are needed. In this instance, the costs in the 2015 to 2016 are not known. The unit costs in 2006 to 2007 are known, and the ratio of these unit costs is assumed to be the same in that year and 2020. Using the ratios of the unit costs, it is possible to express the total volume of crimes in terms of one single crime type. This ‘volume’, as it represents the volume of crimes that would cost the entire police budget, can then be the denominator used to calculate a unit cost, with the numerator being the total police budget.

   Through this method, the relative costs of each crime cost in 2006 to 2007 are held constant but adjusted in a way that accounts for changes in volume and police budget.

9. The final step is to account for the validation of the 2015 to 2016 figures with police forces for the ESCC 2018, which were asked to share any pertinent data they held that might inform that work.

   A data validation workshop was held with a variety of police force representatives, with 2 main aims:

i. Validation of the interpretation of data shared by police forces.

ii. Seeking professional expertise on the triangulation of uprated ABC data with police force data.

There was general agreement that the unit costs calculated through the uprating methodology for crimes looked reasonable except in the case of sex offences. It was suggested that those should be changed to match costs measured by police force exercises. In order to identify any force-led surveys which might generate useful data, all forces were invited to supply data and findings from contemporary (within 3 years of 2015 to 2016) activity-based type exercises undertaken within forces. In total, 13 forces submitted data or reports in response to this request. One study was particularly useful in generating data which was broadly comparable to ABC crime data. This was used to estimate the cost to the police of sex offences. The adjusted unit costs produced for ESCC 2018 were uprated to 2019 to 2020 using the police budget, alongside the other cost figures.

6.1.2 Future Developments

At the time that this update was conducted, 2 pilot waves of the Police Activity Survey including a total of 14 forces had taken place in 2021 and 2022 respectively. Unlike Activity-Based Costings, police staff and officers fill in a survey at the end of their day recording all of their time broken down by activities like crimes and incidents. Each respondent fills in the survey every shift over 7 days. This information, in combination with supplementary data around officer salary, can be used to create unit costs for different crime and incident types, which will serve to improve Home Office’s understanding of the police costs of crime. Plans for phase 3 were in development, but it had not yet taken place. As phase 2 was based on a small sample of 9 forces it was decided that Phase 3 should be considered in future updates.

Phase 3 of the Police Activity Survey has now taken place, but the finalised unit costs were not yet available to be used. In Phase 3 the Home Office invited 43 forces to participate of which 35 forces signed up. The survey ran for one week in February 2023 and includes all police staff and officers excluding those in corporate functions. The results from Phase 3 will be considered to update the police costs in future publications of this report.

6.1.3 Police unit costs

The full list of unit costs of dealing with and investigating crime including overheads by crime type are presented in Table 24. This is calculated by dividing the total costs to the police by crime type, divided by the estimated total number of crimes (including crimes recorded by the police and those not reported to the police). It is therefore not a unit cost to police of each crime recorded as it uses all crimes (rather than Police recorded crime) to estimate the unit cost for each crime and maintain consistency with the CSE.

Table 24: Unit costs for police costs (2019 to 2020)

Crime type (1)	Unit Costs
Homicide	£11,710
Violence with injury	£1,100
Violence without Injury	£790
Rape	£7,480
Other sexual offences	£670
Robbery	£990
Domestic burglary	£520
Theft of a vehicle	£1,990
Theft from a vehicle	£80
Theft from the person	£40
Criminal damage	£150
Fraud	£60
Cybercrime (2)	-

Notes:

1. Unit costs are presented as the average cost for all crimes in society to account for the likelihood of a crime being reported to the police. For police recorded crime unit costs, total cost should be divided by PRC rather than CSEW volumes.
2. Commercial crimes are assumed to cost the same per crime as individual crimes. For example, a domestic burglary is assumed to cost the same as a commercial burglary.
3. Costs cannot be calculated for cybercrime as it was not recorded as a category in the ABC data in 2006 to 2007.

6.1.4 Drivers of police costs

Despite not being able to update the costs and methodology to reflect 2019 to 2020 police costs due to limited direct evidence on what has happened to patterns in police investigative costs since 2006 to 2007, it is reasonable to assume that there have been developments in that period which will have had considerable influence on average police investigative costs. For example, the growth in the importance of digital evidence and relative changes in charge ratios. This is an important caveat to the analysis, especially when using the police unit costs in isolation within government appraisal.

6.2 Criminal justice system

This chapter presents the direct costs to the CJS (excluding the police) in response to crimes occurring.

6.2.1 Approach

The CJS costs capture the following 9 areas:

1. Crown Prosecution Service
2. Magistrates courts’
3. Crown Court
4. Jury service
5. Legal aid defence
6. Non-legal aid defence
7. Prison service
8. Probation service
9. Youth Offending Teams

These CJS categories are costed in the ESCC 2018 report, however this updated analysis has excluded costs associated with the National Offender Management Service Headquarters (NOMS HQ) - since renamed His Majesty’s Prisons and Probation Service (HMPPS)) - and has replaced Youth Justice Board (YJB) costs with the cost of Youth Offending Teams.

Since the production of the ESCC 2018 report, there have been changes to the operating and funding of the YJB, which now focuses on funding Youth Offending Teams. However, Youth Offending Teams also receive partnership funding, with around two-thirds of their funding being through partnership streams. Taking this into account, this report captures Youth Offending Team costs beyond those provided by the YJB to demonstrate the wider cost of youth offending to the CJS. Costs associated with the National Offender Management Service Headquarters (NOMS HQ) are predominantly fixed administration costs that do not vary significantly with the number of crimes. Therefore, whilst budgets associated with other areas of the CJS may respond to a change in crime volumes, NOMS HQ would not be expected to change in value with a change in the volume of crime. Therefore, following consultation with MoJ, this area has been removed from this analysis.

Costs associated with the Crown Court, Magistrates’ courts, legal aid, non-legal aid and Youth Offending Teams use data obtained from the MoJ. Data on Prosecution costs were provided by the CPS. The costs for the CPS, non-legal aid defence and Youth Offending Teams are calculated using a top-down approach. The costs for the probation service are calculated using a top-down approach using ONS data and cost estimates from ESCC 2018. The remaining 5 areas are calculated using a bottom-up approach.

6.2.2 Crown Prosecution Service

The total CPS cost in the year 2019 to 2020 is provided by the CPS. This cost is adjusted to reflect the proportion of all crime types that are captured in the ESCC by weighting the proportion of severity of all offences that are captured in the ESCC, using the Crime Severity Score. To attribute a portion of the total cost to each crime type, a case complexity weighting is applied which takes the CPS caseload for each crime type, alongside court hearing times for each crime type, as a proportion of the total hearing time for all crimes.

The unit costs presented are not the same as the cost per CPS case. The unit cost divides the CPS total cost by the number of crimes in society, whereas the cost per CPS case would divide the total CPS cost by the CPS caseload. Given that the number of crimes in society is different to the number of cases that progress to the CPS, the cost per CPS caseload is not equivalent to the unit costs produced in this analysis.

6.2.3 Magistrates’ courts

The Magistrates’ courts cost per crime type is calculated by multiplying the cost of a sitting day in a Magistrates’ court by the total annual sitting days in Magistrates’ court for each crime type. This bespoke data was provided by the MoJ for the year 2019 to 2020.

6.2.4 Crown Court

The Crown Court cost per crime type is calculated by multiplying the cost to the Crown Court per sitting day by the average number of sitting days of the Crown Court per crime type to give the cost per hearing for each crime type. The cost per hearing is then multiplied by the average number of hearings for each crime type to give the total annual cost to the Crown Court.

6.2.5 Jury service

To estimate the total jury costs, the number of jurors for each case (12) and the expected number of working days missed a year (ONS, 2021c) is multiplied by the average daily wage of a member of the jury.

The mean daily wage is calculated as the mean number of hours worked per day by someone in employment multiplied by the mean hourly earnings and non-wage costs (£20.78). Juries are only required in Crown Court. Therefore, the total number of Crown Court sitting days for jury members that are occupied by employed individuals is calculated by multiplying the total number of Crown Court sitting days by the employment rate for those aged 18 and over (which was 63% in 2019 to 2020 (ONS, 2022b)). It is assumed that the employment rate for those aged 18 and over is the same as the employment rate for those of eligible jury age. This employment-adjusted number of Crown Court sitting days is then multiplied by the mean daily wage to give the opportunity cost of jury service for employed jury members. A similar approach is taken to estimate the opportunity cost of jurors that are not employed by using the value of a non-working hour for an ‘other’ trip given by the Department for Transport (2021).

6.2.6 Legal aid

The MoJ provided data on the average costs of each category of legal aid for each of the crime types. These categories of legal aid were: payments to solicitor firms in Magistrates Court cases; payment of barristers or higher solicitor advocates in Crown Court cases; and payment of solicitor firms in Crown Court cases. For each crime type, the average cost of each type of legal aid is multiplied by the volume of cases to get the total cost of each type of legal aid associated with each crime. For each crime type, these legal aid costs are summed to give a total legal aid cost for each crime.

6.2.7 Non-legal aid defence

Non-legal aid defence represents the cost to defendants of private legal assistance. This is calculated based on the legal aid costs which are scaled up based on an estimate of the ratio of private defence costs to legal aid whilst adjusting for self-representation. Based on the Legal Aid Means Test Review expertise, it is assumed that private defence costs are 3.25 times greater than legal aid costs. Total non-legal aid costs in the Crown Courts are calculated by adjusting total legal aid costs by the ratio of 3.25, accounting for the proportion of Crown Court cases that were legally aided and the proportion of unrepresented litigants in the Crown Court. This process is also conducted using Magistrates’ court data to calculate the total non-legal aid costs in the Magistrates’ court. The non-legal aid costs for the Crown Court and Magistrates Court are combined to give the total non-legal aid costs.

Weightings are then assigned to each crime type based on their total combined Crown Court and Magistrates’ court costs. These weightings are used to divide the total non-legal aid costs so that the non-legal aid cost associated with each crime is produced.

6.2.8 Prison service

A bottom-up methodology is used to estimate the costs of prisons for crimes committed in the year 2019 to 2020, replicating the method used in Radakin et al. (2021). This report follows a bottom-up method to account for prison costs spanning over more than a year. This bottom-up methodology is used in order to avoid double counting of costs from crimes committed in previous years.

To estimate the total prison cost for each crime type, the number of people sentenced to immediate custody for that crime type in 2019 to 2020 is multiplied by the average length of a custodial sentence served for that crime type and the average cost per prisoner per year. The average cost of a prisoner per year is taken from MoJ statistics on prison performance (Ministry of Justice, 2020) and discounted for each year spent in prison. For example:

i. In 2019 to 2020, 571 people were sentenced for homicide. The average length of sentence for those who were sentenced for homicide was 18.1 years, and it is assumed that offenders serve on average two-thirds of their sentence, bringing an average sentence served of 12.1 years. This is multiplied by the average discounted cost per prisoner in that year, which discounts the average cost per prisoner per year of £42,670 in 2019 to 2020 over the 12-year period. The total prison cost for homicide is calculated by multiplying the number of people sentenced, with the average sentenced served, and the discounted average cost per prisoner per year.

6.2.9 Probation service

The costs associated with the probation service are calculated by inflating the total probation costs from ESCC 2018 to 2019 to 2020 prices. This is due to a lack of more up-to-date data preventing the ESCC 2018 approach from being replicated for 2019 to 2020 or a new approach from being developed. Total probation costs per crime type are provided by the MoJ in the ESCC 2018 report for the year 2013 to 2014. These are inflated to the price year 2019 to 2020 for this analysis. Where total probation costs are given at an aggregated level, the unit cost is assumed to be constant across the crimes included. For example, the unit cost for probation is the same for rape and for other sexual offences.

6.2.10 Youth Offending Teams

The cost of Youth Offending Teams is calculated using a top-down methodology which assigns the 2019 to 2020 budget for Youth Justice Services across crime types. This considers the total budget for Youth Offending Teams, taking the Youth Justice Board funding alongside funding from local authorities, health services, the Welsh Government and other funding. This excludes funding from the police, funding from police and crime commissioners and funding provided by the Probation Service, as these costs are captured elsewhere in the economic and social costs of crime.

The total budget is adjusted to reflect the costs spent in response to crime based on a survey of Youth Offending Teams. This remaining budget is then adjusted to capture ESCC crime types. This is calculated by taking the volume of children found guilty for each offence type alongside the severity of each offence type as a proportion of the total severity of all crimes. This estimate for the total cost of Youth Offending Teams within the ESCC is attributed to each crime type by weighting the severity of each offence against the total severity of all offences in the ESCC.

6.2.11 CJS unit costs

Estimates of the unit costs of each of the 9 areas of the CJS for each crime type are displayed in Table 25.

Table 25: Breakdown of average costs of crimes to the CJS

Unit costs are presented as the average cost of all crimes in society (CSEW) to account for the likelihood of a crime being reported to the police and being passed through to the CJS. For CJS case unit costs, total costs should be divided by the relevant caseload rather than the CSEW volumes.

Crime type (1)	CPS	Magistrates’ courts	Crown Court	Jury service	Legal aid	Non-legal aid defence	Prison Service	Probation Service	Youth Offending Teams	Total CJS unit cost (3)
Homicide	£47,280	£180	£21,190	£7,860	£96,160	£49,280	£363,850	£41,630	£1,650	£629,070
Violence with injury	£110	£10	£30	£10	£80	£100	£430	£100	£70	£940
Violence without injury	£80	£10	£10	<£5	£50	£50	£170	£370	£10	£750
Rape	£150	<£5	£100	£40	£110	£240	£940	£20	£10	£1,610
Other sexual offences	£10	<£5	£10	<£5	£10	£20	£70	£20	<£5	£140
Robbery	£10	<£5	£10	<£5	£20	£40	£300	£60	£10	£450
Domestic burglary	£20	£10	£10	<£5	£30	£40	£440	£110	£10	£660
Theft of a vehicle	£80	£50	<£5	<£5	£30	£130	£40	£290	£10	£630
Theft from a vehicle	<£5	<£5	<£5	<£5	<£5	£10	£10	£20	<£5	£40
Theft from the person	£10	£10	<£5	<£5	£10	£20	£30	£70	<£5	£140
Criminal damage	£10	<£5	<£5	<£5	<£5	<£5	£10	£60	<£5	£90
Fraud (2)	£10	<£5	<£5	<£5	£10	£10	£20	£20	<£5	£70
Cybercrime	<£5	<£5	<£5	<£5	<£5	<£5	<£5	<£5	<£5	<£5
Commercial theft	£10	<£5	<£5	<£5	<£5	£10	<£5	£30	<£5	£50
Commercial burglary	£10	£20	<£5	<£5	<£5	£40	£150	£130	£10	£350

Notes:

1. It is assumed that for all commercial crimes except commercial burglary and commercial theft, these are dealt with in the same way as their individual equivalents in the CJS. Therefore, it assumed that the unit costs for commercial crimes are the same as the equivalent individual crimes.

2. The unit CJS costs for cybercrime are low per cybercrime as the total CJS costs are low and the volumes of crime are high. This suggests a very low p of cybercrimes are prosecuted.

3. There may be discrepancies in the total figures due to the effect of rounding.

Annex 1: Total cost of fraud

The £500 million greater total cost of fraud in this report than that published alongside the Fraud Strategy: stopping scams and protecting the public (2023) - GOV.UK is due to the adoption of a revised methodology for estimating lost output. The 3 revisions to this methodology are (1) using the midpoint approach to estimate the average hours taken off work instead of the median values, (2) using the mean hourly wage instead of the median hourly wage and (3) using the mean weekly number of hours worked instead of the median weekly numbers of hours worked. Points (2) and (3) are consistent across all crime types in this report.

Appendix 1: Physical and emotional costs tables

Table 26: Prevalence of harms against individuals for each crime

Harm (1)	Violence with injury	Violence without injury	Rape	Other sexual offences	Robbery	Domestic burglary	Theft of a vehicle	Theft from a vehicle	Theft from the person	Criminal damage	Fraud	Cybercrime
Emotional harms
Fear	26%	28%	60%	23%	32%	33%	10%	6%	9%	10%	7%	6%
Depression	19%	6%	22%	8%	6%	10%	7%	1%	2%	9%	4%	1%
Anxiety/panic attacks	23%	19%	42%	18%	24%	24%	14%	3%	10%	9%	8%	6%
Physical harms
Minor bruising/black eye	57%	-	23%	2%	14%	-	-	-	-	-	-	-
Severe bruising	33%	-	8%	4%	2%	-	-	-	-	-	-	-
Cuts	27%	-	5%	0%	4%	-	-	-	-	-	-	-
Scratches	21%	-	15%	3%	11%	-	-	-	-	-	-	-
Puncture or stab wounds	4%	-	-	-	0%	-	-	-	-	-	-	-
Broken bones	4%	-	5%	2%	1%	-	-	-	-	-	-	-
Concussion or loss of consciousness	4%	-	1%	0%	0%	-	-	-	-	-	-	-
Nose bleed	3%	-	-	-	0%	-	-	-	-	-	-	-
Facial or head injuries (no bruising)	2%	-	1%	-	0%	-	-	-	-	-	-	-
Broken nose	2%	-	-	-	0%	-	-	-	-	-	-	-
Internal injuries	2%	-	-	-	0%	-	-	-	-	-	-	-
Chipped teeth	1%	-	-	-	0%	-	-	-	-	-	-	-
Dislocation	1%	-	-	-	0%	-	-	-	-	-	-	-
Broken or lost teeth	1%	-	-	-	0%	-	-	-	-	-	-	-
Eye/face injuries due to for example, acid/paint thrown into face	0%	-	-	-	0%	-	-	-	-	-	-	-
Other	1%	-	-	-	0%	-	-	-	-	-	-	-

Notes:

1. For crimes where respondents in the CSEW were not asked whether they experienced physical harms, no entry (shown by a ‘-‘) is given in the table. Where a question on experience of harms has been asked in the CSEW, but none of the respondents reported experiencing that harm, 0% is given in the table.

Table 27: Prevalence of harms against businesses for each crime

Harm (1)	Commercial robbery	Commercial burglary	Commercial theft	Theft of a commercial vehicle	Theft from a commercial vehicle	Commercial criminal damage
Emotional harms
Fear	63%	13%	0%	14%	4%	0%
Depression	13%	6%	0%	0%	0%	4%
Anxiety/panic attacks	13%	6%	0%	0%	0%	0%
Physical harms
Minor bruising/black eye	13%	-	-	-	-	-
Severe bruising	25%	-	-	-	-	-
Cuts	13%	-	-	-	-	-
Scratches	25%	-	-	-	-	-
Puncture or stab wounds	0%	-	-	-	-	-
Broken bones	0%	-	-	-	-	-
Concussion or loss of consciousness	13%	-	-	-	-	-
Nose bleed	13%	-	-	-	-	-
Facial or head injuries (no bruising)	-	-	-	-	-	-
Broken nose	0%	-	-	-	-	-
Internal injuries	-	-	-	-	-	-
Chipped teeth	0%	-	-	-	-	-
Dislocation	0%	-	-	-	-	-
Broken or lost teeth	13%	-	-	-	-	-
Eye/face injuries due to for example, acid/paint	-	-	-	-	-	-
Other	-	-	-	-	-	-

Notes:

1. For crimes where respondents in the CSEW were not asked whether they experienced physical harms no entry (shown by a ‘-‘) is given in the table. Where a question on experience of harms has been asked in the CSEW but none of the respondents reported experiencing that harm 0% is given in the table.

Table 28: QALY losses associated with physical and emotional harm

Injury	Corresponding Global Burden of Disease (GBD) (1)	QALY loss
Physical harms
Minor bruise/black eye	0.25 of broken bones (Dolan et al., 2005)	2.6%
Severe Bruising	0.5 of broken bones (Dolan et al., 2005)	5.2%
Scratches	0.25 of cuts (Dolan et al., 2005)	0.2%
Cuts	Open wound: short term, with or without treatment	0.6%
Puncture or stab wounds	Broken bones used as a proxy (Dubourg et al., 2005)	10.3%
Broken bones	Fracture of sternum or one or 2 ribs: short term, with or without treatment	10.3%
Nose bleed	Open wound: short term, with or without treatment	0.6%
Broken nose	Fracture of face bone: short or long term, with or without treatment	6.7%
Broken or lost teeth	0.5 of fracture of face bone: short or long term, with or without treatment (Dolan et al., 2005)	3.4%
Chipped teeth	0.5 of broken/lost teeth (Dolan et al., 2005)	1.7%
Dislocation	Dislocation of shoulder: long term, with or without treatment	6.2%
Concussion or loss of consciousness	0.5 of disability weight for intracranial injury (short term) (Dolan et al., 2005)	11.0%
Internal injuries	No associated injury in GBD. Severe bruising has been taken as an appropriate proxy	5.2%
Facial or head injuries (no bruising)	No associated injury in GBD. Cuts have been taken as an appropriate proxy	0.6%
Eye/face injuries due to for example, acid/paint thrown into face	Injury to eyes: short term	5.4%
Other	Other injuries of muscle and tendon (includes sprains, strains, and dislocations other than shoulder, knee, or hip)	0.8%
Emotional harms
Fear	Anxiety disorders: mild (Ohman, 2008)	3%
Depression	Major depressive disorder: mild episode	14.5%
Anxiety/panic attacks	Moderate anxiety	13.3%
Drug abuse	Moderate cocaine dependence	47.9%
Alcohol abuse	Moderate alcohol use disorder	37.3%
Obesity/eating disorder	Anorexia nervosa/bulimia nervosa	22.4%
Sexual dysfunction	Impotence	1.7%
Death
Death	Death	100%

Notes:

1. Salomon et al. (2015).

Table 29: Duration of negative consequences associated with physical and emotional harms

Injury	Duration (years) (1)	Source
Physical
Minor bruise/black eye	0.0288	Dolan et al. (2005)
Severe bruising	0.0575	Dolan et al. (2005)
Scratches	0.0060	Dolan et al. (2005)
Cuts	0.0240	Dolan et al. (2005)
Puncture or stab wounds	0.1150	Advance Tissue (2014)
Broken bones	0.1150	Dolan et al. (2005)
Nose bleed	0.0027	No source available, assumed one day
Broken nose	0.0590	Dolan et al. (2005)
Broken or lost teeth	0.0192	Dolan et al. (2005)
Chipped teeth	0.0192	Dolan et al. (2005)
Dislocation	0.2308	NHS (2020)
Concussion or loss of consciousness	0.0335	Dolan et al. (2005)
Internal injuries	0.0575	No source available, assumed same as severe bruising
Facial or head injuries (no bruising)	0.0240	No source available, assumed same as cuts
Eye/face injuries due to for example, acid/paint	0.0192	No source available, assumed one week
Other	0.0192	Dolan et al. (2005)
Emotional – violent crime (2)
Fear	1.2500	Norris and Kaniasty (1994)
Depression	1.0000	Dolan et al. (2005)
Anxiety/panic attacks	3.0000	Dolan et al. (2005)
Emotional – non-violent crime (3)
Fear	1.2500	Norris and Kaniasty (1994)
Depression	0.1670	Wasserman and Ellis (2007)
Anxiety/panic attacks	0.1670	Wasserman and Ellis (2007)
Emotional – semi-violent crime (4)
Fear	1.2500	Average of violent and non-violent crime
Depression	0.5800	Average of violent and non-violent crime
Anxiety/panic attacks	1.5800	Average of violent and non-violent crime
Emotional – rape specific
Drug abuse	5.0000	Dolan et al. (2005)
Alcohol abuse	5.0000	Dolan et al. (2005)
Obesity/eating disorder	5.0000	Dolan et al. (2005)
Sexual dysfunction	0.1670	Dolan et al. (2005)
Death
Death	42.1649	ONS (2021b)

Notes:

1. Results are presented in years so that they map more easily to the total QALY value.

2. Violent crimes are assumed to be homicide, violence with injury and rape.

3. Non-violent crimes are assumed to be burglary, theft, criminal damage, fraud and cybercrime.

4. Semi-violent crimes are assumed to be other sexual offences, robbery and violence without injury.

Appendix 2: Changing the price year

As explained in the report, all cost estimates in this publication are presented in 2019 to 2020 prices. Analysts and policymakers using the costs in this report may want to change the price year to reflect changes in price level over time. For example, if a policy report about the harms caused by violent crime is written in the year ending March 2024, it is more meaningful to the reader to be presented with costs in 2023 to 2024 prices as these will reflect the context of that year’s economy.

This appendix provides instructions for how to adjust the price year of costs presented in this report using the GDP deflator series.

A.1: GDP deflators series

The GDP deflator can be viewed as a measure of general inflation in the domestic economy. Using the series allows users to account for changes in the general price level.

Inflating the costs of crime using the GDP deflator is a simplification; in theory, the different cost components would move in line with different indices. For example:

• lost output should be inflated in line with wage growth
• physical and emotional harms, based on the Quality Adjusted Life Year (QALY) value of £70,000, should be held constant when adjusting price years, and discounted at the health rate of 1.5% instead of the standard 3.5% (HMT, 2022 and DHSC guidance)
• police costs should be inflated in line with growth in the nominal value of the police budget

However, given that source costs used in this report have come from a variety of years, and the documented limitations in the methodology, using the GDP deflators to change the price year of costs and discounting at the standard rate of 3.5% is a proportionate approach. Users should assess the proportionality on a case-by-case basis, and if of disproportionate impact should hold physical and emotional harms constant and discount at the health rate of 1.5%.

A.2: Practical examples – inflating figures to 2023 to 2024 prices

Table 30 is an extract from a GDP deflator series (HM Treasury, 28 March 2025) and provides the basis for 2 worked examples of how the series can be used.

Table 30: GDP deflator series

Financial year	GDP deflator at market prices	Percentage change on previous year
2018 to 2019	82.278	2.11%
2019 to 2020	84.227	2.37%
2020 to 2021	88.746	5.37%
2021 to 2022	88.228	-0.58%
2022 to 2023	94.452	7.05%
2023 to 2024	100.000	5.87%
2024 to 2025	103.801	3.80%

Notes:

1. GDP deflator data for 2024 to 2025 have been added in for the purpose of this example.

Q1.

If a ‘violence with injury’ offence has a unit cost of £15,720 in 2019 to 2020 prices, how much would this be worth in 2023 to 2024 prices?

A1.

Use the GDP deflator series which shows that in 2023 to 2024 prices are higher than in 2019 to 2020 by a ratio of 1.19 (100.000 ÷ 84.227).

That is £15,720 × (100.000/84.227) = £18,664

Therefore, £15,720 in 2019 to 2020 prices is equivalent to £18,664 in 2023 to 2024 prices.

Q2.

If the total economic and social cost of ‘violence with injury’ offences is estimated at £13.9 billion (£13,900 million) in 2019 to 2020 prices, how much would this be worth in 2023 to 2024 prices?

A2.

Use the GDP deflator series which shows that in 2023 to 2024 prices are higher than in 2019 to 2020 by a ratio of 1.19 (100.000 ÷ 84.227).

That is £13,929 million × (100.000/84.227) = £16,537 million

Therefore, £13.9 billion in 2019 to 2020 prices is equivalent to £16.5 billion in 2023 to 2024 prices.

To put the 2019 to 2020 values into a different price year, simply swap in the relevant GDP deflator.

A.3: Useful resources

• the HM Treasury publish new GDP deflators each quarter. Make sure to use the most recent series for analysis
• to accompany the GDP deflators, HM Treasury has also published a user guide
• there are also some practical examples of how to use the GDP deflator series (following the same approach used in this appendix)

References

Action Fraud (no date) ‘ActionFraud: National Fraud & Cyber Crime Reporting Centre’. Available from: https://www.actionfraud.police.uk/ [Accessed 21 June 2022].

Advanced Tissue (2014) ‘The signs and stages of wound healing’. Available from: https://advancedtissue.com/signs-wound-healing-stages [Accessed 27 January 2018].

Brand, S. and Price, R. (2000) ‘The economic and social costs of crime’, Home Office Research Study 217. London: Home Office Research, Development and Statistics Directorate.

Citizens Advice (2021) ‘The impact of the Witness Service during coronavirus 2020 to 2021’. Available from: https://www.citizensadvice.org.uk/about-us/information/impact-of-the-witness-service/ [Accessed 22 June 2022].

Criminal Injuries Compensation Authority (2020) ‘Annual Report & Accounts, 2019-20’. Available from: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/933212/6.6482_CICA_Annual_Report_and_Accounts_2019-20_FINAL_WEB.pdf [Accessed 21 June 2022].

Curtis, L. and Burns, A. (2020) ‘Unit Costs of Health & Social Care 2020’, Personal Social Services Research Unit, University of Kent, Canterbury.

Department for Transport (2020a) ‘Guidance: TAG data book’. Available from: https://www.gov.uk/government/publications/tag-data-book [Accessed 14 June 2021].

Department for Transport (2020b) ‘Vehicle Licensing Statistics: Annual 2019’. Available from: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/882196/vehicle-licensing-statistics-2019.pdf [Accessed 9 June 2022].

Department for Transport (2021) ‘Guidance: TAG data book’, Available from: https://webarchive.nationalarchives.gov.uk/ukgwa/20220506212708/https://www.gov.uk/government/publications/tag-data-book [Accessed 9 June 2022].

Dolan, P., Loomes, G., Peasgood, T. and Tsuchiya, A. (2005) ‘Estimating the intangible victim costs of violent crime’, British Journal of Criminology, vol. 45(6), pp 958 to 976. Oxford University Press, Oxford.

Dubourg, R., Hamed, J. and Thorns, J. (2005) ‘The economic and social costs of crime against individuals and households 2003 to 2004’, Home Office Online Report, 30/05.

Heeks, M., Reed, S., Tafsiri, M. and Prince, S. (2018) ‘The economic and social costs of crime, Second edition’, Home Office Research Report, 99.

Fell, E., James, O., Dienes, H., Shah, N. and Grimshaw, J. (2019) ‘Understanding organised crime 2015 to 2016: Estimating the scale and the social and economic costs, Second edition’, Home Office Research Report, 103.

Action Fraud, 2019-20 (2021) ‘Fraud Crime Trends 2019-20’. Accessed at https://data.actionfraud.police.uk/cms/wp-content/uploads/2020/07/Fraud-crime-trends.pdf [accessed 8 July 2023].

Future Market Insights (2021) ‘Car Security System Market’. Available from: https://www.futuremarketinsights.com/reports/car-security-system-market [Accessed 9 June 2022].

Gartner (2020) ‘Gartner Forecasts Worldwide Security and Risk Management Spending Growth to slow but Remain Positive in 2020’. Available from: https://www.gartner.com/en/newsroom/press-releases/2020-06-17-gartner-forecasts-worldwide-security-and-risk-managem [Accessed 20 June, 2022].

HM Treasury, HMT (2022), ‘The Green Book: Central government guidance on appraisal and evaluation’.

HM Treasury (2025), ‘GDP deflators at market prices, and money GDP March 2025’. Available from: https://www.gov.uk/government/statistics/gdp-deflators-at-market-prices-and-money-gdp-march-2025-spring-statement-quarterly-national-accounts.

Home Office (2011) ‘Revisions made to the multipliers and unit costs of crime used in the Integrated Offender Management Value for Money Toolkit’. Available from: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/97813/IOM-phase2-costs-multipliers.pdf [Accessed 8 June 2022].

Home Office (2022) ‘Police recorded crime and outcomes open data tables’. Available from: https://www.gov.uk/government/statistics/police-recorded-crime-open-data-tables [Accessed 20 June 2022].

Home Office (2025) ‘Experiences of victims of fraud and cyber crime’. Available from: https://www.gov.uk/government/publications/experiences-of-victims-of-fraud-and-cyber-crime.

Infologue (2019) ‘Infologue 2019 Top 30 UK Security Companies’. Available from: https://www.infologue.com/exclusive/infologue-2019-top-30-uk-security-companies/ [Accessed 8 June 2022].

Marketsandmarkets (2016) ‘Vehicle Security System Market by Type, Technology, Vehicle Type (Passenger Car, Commercial Vehicles and Off-Highway Vehicles), and by Region (Asia-Pacific, Europe, North America and Rest of the World) – Industry Trends and Forecast to 2021’. Available from: https://www.marketsandmarkets.com/Market-Reports/automotive-security-market-184165229.html [Accessed 28 January 2018].

Millar, N. (2015) ‘Physical security equipment and services report 2015’. IHS. Available from: https://www.emc.com/collateral/analyst-reports/physical-security-equipment-services-report.pdf.

Ministry of Justice (2019) ‘Guidance: Victim and witness funding awards’. Available from: https://www.gov.uk/guidance/victim-and-witness-funding-awards#national-homicide-service-for-england-and-wales [Accessed 8 June 2022].

Ministry of Justice (2020) ‘Prison Performance statistics 2019 to 2020’. Available from: https://www.gov.uk/government/statistics/prison-performance-statistics-2019-to-2020 [Accessed 22 June 2022].

National Statistics Scotland (2021) ‘Recorded Crime in Scotland, 2020-2021’. Available from: https://www.gov.scot/publications/recorded-crime-scotland-2020-2021/documents/ [Accessed 21 June 2022].

NHS (2020) ‘Dislocated shoulder’. Available from: https://www.nhs.uk/conditions/dislocated-shoulder/ [Accessed 9 June 2022].

NHS (2021) ‘National Cost Collection for the NHS’. Available from: https://www.england.nhs.uk/costing-in-the-nhs/national-cost-collection/ [Accessed 5 November 2021].

Norris, F. H. and Kaniasty, K. (1994) ‘Psychological distress following criminal victimization in the general population: Cross-sectional, longitudinal, and prospective analyses’. Journal of Consulting and Clinical Psychology, vol. 62(1), pp 111 to 123.

Ohman, A. (2008) ‘Fear and Anxiety’ in Lewis, M., Haviland-Jones, J. and Barrett, L. (Eds.), The handbook of emotions (pp. 709 to 729). New York, NY: The Guildford Press.

Oliver, R., Alexander, B., Roe, S. and Wlasny, M. (2019) ‘The economic and social costs of domestic abuse’, Home Office Research Report, 107.

ONS (2019) ‘Regional gross domestic product: all ITL regions’. Available from: https://www.ons.gov.uk/economy/grossdomesticproductgdp/datasets/regionalgrossdomesticproductallnutslevelregions [Accessed 8 June 2022].

ONS (2020a) ‘Crime in England and Wales: year ending March 2020’. Available from: https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/bulletins/crimeinenglandandwales/yearendingmarch2020#coronavirus-and-crime-statistics-in-the-year-ending-march-2020 [Accessed 20 June 2022].

ONS (2020b) ‘UK business: activity, size and location’. Available from: https://www.ons.gov.uk/businessindustryandtrade/business/activitysizeandlocation/datasets/ukbusinessactivitysizeandlocation.

ONS (2021a) ‘Crime in England and Wales QMI’. Available from: https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/methodologies/crimeinenglandandwalesqmi [Accessed 17 June 2022].

ONS (2021b) ‘National life tables: UK’. Available from: https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/lifeexpectancies/datasets/nationallifetablesunitedkingdomreferencetables [Accessed 9 June 2022].

ONS (2021c) ‘Earnings and hours worked, all employees: ASHE Table 1’. Available from: https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/earningsandworkinghours/datasets/allemployeesashetable1 [Accessed 21 June 2022].

ONS (2022a) ‘UK SIC 2007’. Available from: https://www.ons.gov.uk/methodology/classificationsandstandards/ukstandardindustrialclassificationofeconomicactivities/uksic2007 [Accessed 20 June 2022].

ONS (2022b) ‘Employment rate (aged 16 to 64, seasonally adjusted): %’. Available from: https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/employmentandemployeetypes/timeseries/lf24/lms/previous [Accessed 9 June 2022].

ONS (2022c) ‘Crime in England and Wales: Appendix tables’. Available from: https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/datasets/crimeinenglandandwalesappendixtables [Accessed 21 June 2022].

ONS (2022d) ‘Crime in England and Wales: Other related tables’. Available from: https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/datasets/crimeinenglandandwalesotherrelatedtables [Accessed 21 June 2022].

ONS (2022e) ‘Guide to interpreting Annual Survey of Hours and Earnings (ASHE) estimates’. Available from: Guide to interpreting Annual Survey of Hours and Earnings (ASHE) estimates - Office for National Statistics (ons.gov.uk) [Accessed 18 July 2022].

Police Service of Northern Ireland (2020) ‘Police recorded crime statistics Archive’. Available from: https://www.psni.police.uk/about-us/our-publications-and-reports/official-statistics/police-recorded-crime-statistics/police [Accessed 21 June 2022].

Radakin, F., Scholes, A., Soloman, K., Thomas-Lacroix, C. and Davies, A. (2021) ‘The economic and social cost of contact child sexual abuse’. Available from: https://www.gov.uk/government/publications/the-economic-and-social-cost-of-contact-child-sexual-abuse/the-economic-and-social-cost-of-contact-child-sexual-abuse [accessed 8 June 2022].

Reed, S., Roe, S., Grimshaw, J. and Oliver, R. (2018) ‘The economic and social costs of modern slavery’, Home Office Research Report, 100.

Salomon, J. A., Haagsma, J. A., Davis, A., de Noordhout, C. M., Polinder, S., Havelaar, A. H., Cassini, A., Devleesschauwer, B., Kretzschmar, M., Speybroeck, N., Murray, C. J. L. and Vos, T. (2015) ‘Disability weights for the Global Burden of Disease 2013 study’ in The Lancet Global Health, vol. 3(11), pp 712 to 723.

Shiroiwa, T., Fukuda, T., Ikeda, S. and Shimozuma, K. (2013) ‘QALY and Productivity Loss: Empirical Evidence for “Double Counting”’ in Value in Health, vol. 16(4), pp 581 to 587.

The Chartered Institute of Purchasing & Supply (no date) ‘Introduction to Purchasing Security’. Available from: https://www.perpetuityresearch.com/images/Reports/2007-02%20Introduction%20to%20Purchasing%20Security.pdf [Accessed 11 March 2022].

Victim Support (2020) ‘2019-20 Annual Report & Accounts’. Available from: https://www.victimsupport.org.uk/wp-content/uploads/2021/04/VS_Annual_Report_2019-20.pdf [Accessed 8 June 2022].

Statista (2021) ‘Number of motor vehicles registered in the United States from 1990 to 2020’. Available from: https://www.statista.com/statistics/183505/number-of-vehicles-in-the-united-states-since-1990/ [Accessed 9 June 2022].

UK Finance (2020) ‘Written evidence submitted by UK Finance’. Available from: https://committees.parliament.uk/writtenevidence/18646/pdf/ [Accessed 20 June 2022].

Wasserman, E. and Ellis, C. (2007) ‘Impact of crime on victims’. Chapter 6 in National Victim Assistance Academy Track 1: Foundation-Level Training. Available from: https://ce4less.com/Tests/Materials/E075Materials.pdf [Accessed 8 June 2022].

World Bank (2021a) ‘United Kingdom’. Available from: https://data.worldbank.org/country/GB [Accessed 22 October 2021].

World Bank (2021b) ‘Ireland’. Available from: https://data.worldbank.org/country/Ireland [Accessed 22 October 2021].

World Bank (2022) ‘GDP (current US$). Available from https://data.worldbank.org/indicator/NY.GDP.MKTP.CD?end=2020&start=1960&view=chart [Accessed 20 June 2022].