Last Updated on September 26, 2026 by Karl Thompson
Crime costs money — to victims, to businesses, to the NHS and to the criminal justice system. The Home Office publishes estimates of that cost, and the most recent figures are:
- £60.8 billion a year for crimes against individuals
- £16.9 billion a year for crimes against businesses, covering nine business sectors out of twenty-one.
Home Office, Economic and Social Costs of Crime 2019 to 2020, published 27 May 2026.
Two things about those numbers are more interesting than the numbers themselves, and both are examinable.
Most of the cost is not financial. Nearly half the cost of crimes against individuals — £28.7 billion of £60.8 billion — is the Home Office’s attempt to put a monetary value on physical and emotional harm to victims. Not stolen property, not police time. Harm.
The estimates deliberately leave things out. Crimes with no individual victim are excluded entirely. So are fear of crime, quality-of-life effects, and the harm to victims’ families. The Home Office says so explicitly, and treats its own figures as an underestimate.
This post covers material relevant to the crime and deviance module, AQA Paper 3 (7192/3). It is most useful for evaluating victimology and crime prevention policy.
What the estimates include
The Home Office splits costs into three areas.
In anticipation of crime — what we spend because crime might happen. Burglar alarms, CCTV, car alarms, private security, insurance administration. £3.3 billion for individuals.
As a consequence of crime — what happens because it did. Property stolen or damaged, physical and emotional harm, lost output from time off work, health services, victim services. £49.2 billion for individuals, by far the largest area.
In response to crime — police time, the Crown Prosecution Service, courts, defence, prisons and probation. £8.4 billion for individuals.
| Cost category (crimes against individuals) | Cost | Share |
|---|---|---|
| Physical and emotional harm to victims | £28.7bn | 47% |
| Lost output | £9.8bn | 16% |
| Property stolen or damaged | £7.4bn | 12% |
| Police costs | £5.1bn | 8% |
| Health services | £3.3bn | 5% |
| Other criminal justice system costs | £3.3bn | 5% |
| Defensive expenditure | £2.5bn | 4% |
| Insurance administration | £0.7bn | 1% |
| Victim services | £0.07bn | 0% |
The last row is worth pausing on. Victim services account for 0.1% of the total cost of crime — £70 million against £28.7 billion of harm to the people those services exist to support.
Unit costs: what one crime costs
The Home Office also estimates the average cost to society of a single offence.
| Offence | Estimated cost per offence |
|---|---|
| Homicide | £3,346,680 |
| Rape | £48,000 |
| Theft of a vehicle | £20,330 |
| Violence with injury | £15,720 |
| Robbery | £10,490 |
| Domestic burglary | £7,720 |
| Violence without injury | £7,150 |
| Other sexual offences | £6,540 |
| Fraud | £2,000 |
| Criminal damage | £1,970 |
| Theft from the person | £1,590 |
| Theft from a vehicle | £1,420 |
| Cybercrime | £750 |
The pattern that matters: violent crimes make up just over 70% of the total cost of crimes against individuals, but only 35% of the volume. Homicide is the extreme case — 673 homicides in 2019/20 were under 0.01% of all crimes against individuals and almost 4% of the total cost.
Of homicide’s £3.35 million unit cost, £2.25 million is physical and emotional harm and £384,000 is lost output. The criminal justice response accounts for £629,000.
Fraud is the volume problem
Fraud has a low unit cost (£2,000) and an enormous volume — an estimated 3.7 million offences, against 312,040 recorded by the police. A multiplier of nearly twelve.
The Home Office now publishes fraud separately, because the picture has moved so fast. Its most recent estimate puts the cost of fraud against individuals and businesses in England and Wales at £14.4 billion in the year ending March 2024.
Compare the multipliers. Homicide is 1.00 — every homicide is known to the police. Fraud is 11.78 and cybercrime is 33.42. The crimes that cost most per offence are the ones we know about; the crimes with the largest hidden volume are the ones we do not.
Why sociologists treat these figures with caution
This is where the marks are. The estimates are useful and they are also a construction, and saying how earns AO3.
They price harm using willingness-to-pay methods. The emotional cost of rape is estimated at £26,340. The Home Office acknowledges directly that this may seem reductive to people who have experienced these crimes, and that the human cost cannot be fully captured. Putting a price on suffering makes it comparable, which is the point — and makes it look more precise than it is.
Crimes against society are excluded. Drug possession, public order offences and anything without an individual victim are outside the estimates entirely.
Corporate and state crime are almost entirely absent. The figures cover notifiable offences against individuals and nine business sectors. Marxists would argue this measures the crimes of the powerless with precision and the crimes of the powerful not at all. Tombs and Whyte (2007) make exactly this point about workplace deaths recorded as accidents — no offender, no victim, no cost.
Some of the data is old. The 2019/20 year was chosen deliberately to avoid the pandemic, so the current published estimates describe a pre-Covid world. Some inputs are older still: the vehicle security estimate is scaled from a 2016 US market study, and building security from a 2015 industry report.
The costs are not evenly borne. A £7,720 burglary costs a wealthy household and a poor household the same on paper. It does not cost them the same in practice. Left realists make this point sharply: crime falls hardest on people least able to absorb it.
Where this is useful in the exam
Evaluating victimology. The figure that does the work is that 47% of the cost of crime is harm to victims. Positivist victimology takes that seriously; critical victimology asks why harms outside this framework are not counted at all.
Evaluating crime prevention. These unit costs exist so that policymakers can weigh interventions. If a scheme prevents a hundred burglaries, the Home Office method says it saved society £772,000. That is situational crime prevention logic, and you can evaluate it on the same terms.
Evaluating official statistics. The multipliers — fraud at 11.78, cybercrime at 33.42, homicide at 1.00 — are a precise illustration of the dark figure of crime, published by the Home Office itself.
For crime’s effects at the level of society rather than the individual, including Durkheim’s argument that crime performs positive functions, see functionalist explanations of deviance.
Signposting and Related Posts
This post has been written for students studying crime and deviance as part of the AQA A-Level Sociology course, Paper 3 (7192/3).
For what crime does to the people it happens to, see what are the impacts of crime on victims?
For the perspectives, see victimology: positivist and critical approaches.
For the functionalist argument that crime benefits society, see functionalist explanations of deviance.
For one category in detail, see what is the cost of cybercrime?
Please click here to return to the homepage – ReviseSociology.com
Sources
Home Office (2026) Economic and Social Costs of Crime 2019 to 2020, published 27 May 2026.
Home Office (2026) Economic and Social Cost of Fraud 2023 to 2024 (second edition).
Heeks, M. et al. (2018) The Economic and Social Costs of Crime, second edition, Home Office Research Report 99.
Tombs, S. and Whyte, D. (2007).