Measuring drug-related homicide methodology: February 2024

• The Home Office currently measures drug-related homicides as any case involving a drug user or drug dealer, or that is related to drugs in any way using record-level data from the Home Office Homicide Index.

• Further analysis of these data explored applying the Goldstein tripartite framework to further categorise "drug-related" homicides as systemic, economic compulsive, and/or psychopharmacological; drug-related homicides that did not meet the criteria for the three groups were defined as having an ambiguous link to drugs.

• Data from the Home Office Homicide Index shows that between April 2018 and March 2023, there were 3,148 homicides recorded by the police, 1,627 of which were drug-related (52%) according to the current definition of drug-related homicide.

• Out of these 1,627 drug-related homicides, 838 homicides could be categorised as systemic, economic compulsive, and/or psychopharmacological (27% of all homicides), with a further 789 homicides having an ambiguous link to drugs (25% of all homicides).

• In the year ending March 2023, 144 (24% of all homicides) met the criteria for one or more of these categories; 86 homicides were systemic (15% of all homicides), 24 were economic compulsive (4% of all homicides), 66 were psychopharmacological (11% of all homicides), with a further 166 homicides (28%) having an ambiguous link to drugs.

The Office for National Statistics's (ONS's) annual Homicide in England and Wales report publishes various measures on drug-related homicides, including the number and proportion of homicides that involve drug users or dealers or are related to drugs in any way. These figures are based on analyses of information held on the Home Office Homicide Index, which contains detailed record-level information about each homicide recorded by the police in England and Wales. The ONS publication is designated as National Statistics.

The Home Office definition of a drug-related homicide is broad. It includes any case where the police believed the victim or suspect was an illegal drug user or dealer; had taken a drug; had motive to obtain drugs or steal drug proceeds; or where the case was otherwise marked as drug-related.

Between April 2018 and March 2023, there were 3,148 homicides in England and Wales recorded in the Home Office Homicide Index, 1,627 (52%) of which fell under the Home Office definition of a drug-related homicide.

In the year ending March 2023, 310 homicides, or 53% of homicides, were drug related according to this definition.

In order to better understand the nature of drug-related homicide and trends over time, this analysis applied a widely used framework for understanding how illicit drug use drives homicide developed by Goldstein (1985) to Home Office Homicide Index data. The Home Office's trends and drivers of homicide research report reviewed the literature on drugs as a driver of homicide trends, with most of the 45 studies identified in the review referencing Goldstein (1985). This framework was therefore identified as the most suitable method to develop our understanding of drug-related homicide.

Goldstein outlines three main types of drug-related homicide.

Systemic homicides

Systemic homicides are those arising from the fact that prohibition of drugs creates an illegal market, in which grievances cannot be reconciled through normal judicial channels, so they may be settled through violence. For example, homicides resulting from disputes over drug-selling territory, hierarchy enforcement and punishment for failing to pay debts would all fall into this category.

Economic compulsive homicides

Economic compulsive homicides arise if illicit drug users have to steal to fund their drug use. As a result, it is possible that a homicide may occur in the act of robbery or burglary.

Psychopharmacological homicides

Psychopharmacological homicides occur as a result of the psychological effects that drugs have on those who take them, such as increased aggression or disinhibition. Goldstein argues this may make people more likely to be victims or perpetrators of homicide.

This analysis applied Goldstein's framework to "drug-related" homicides recorded between April 2018 and March 2023. We developed inclusion criteria for the framework based on existing variables in the Home Office Homicide Index. As the Home Office Homicide Index is an administrative dataset, rather than a bespoke dataset, we identified the best fit variables for each category within the Goldstein tripartite framework.

Following thorough interrogation of the Home Office Homicide Index, we identified the variables listed below as the most relevant to Goldstein's framework and used them to create the criteria for the systemic, economic compulsive, and psychopharmacological categories. The analysis only applies the criteria to homicides already falling under the Home Office's definition of drug related. Drug-related homicides can meet the criteria for more than one category within Goldstein's framework.

Drug-related homicides meet the criteria for the systemic category if:

• the suspect and victim are both known illegal drug dealers

• the suspect was a dealer with motive to obtain drugs or steal drug proceeds

• the suspect was a dealer and the main method of the homicide was shooting

• the victim's relationship to the principal suspect was criminal associate

• the victim and/or suspect were part of a gang

• it is gang related

• it is organised crime group (OCG) related

• it is county lines related

Drug-related homicides meet the criteria for the economic compulsive category if the suspect was a drug user and:

• the main circumstance of the offence was in furtherance of theft or gain

• the motive was to obtain drugs or steal drugs proceeds

Drug-related homicides meet the criteria for the psychopharmacological category if the suspect had taken drugs at the time of the homicides (with or without alcohol).

The original definition of psychopharmacological cases developed by Goldstein includes homicides where victims were under the influence of drugs at the time (regardless of whether the suspect was also under the influence of drugs). This implies wrongly (whether intentionally or unintentionally) that the victim may have been partially or wholly responsible for what happened to them. In this analysis, only those cases where the suspect was under the influence of drugs at the time of the homicide are included in the psychopharmacological group.

Ambiguous drug-related homicides are those that fall under the Home Office definition of "drug-related" that do not meet any of the criteria used to apply Goldstein's framework to the data. For example, cases involving a victim marked as a drug user (which could refer to historical use rather than use at the time of the homicide) in the absence of other evidence (the filters used to create the categories outlined above) to suggest that drugs were relevant to the homicide. Some cases with an ambiguous link to drugs may be recategorised as systemic, economic compulsive, and/or psychopharmacological if new information becomes available or new suspects are identified.

Between April 2018 and March 2023, 838 of 3,148 cases in the Home Office Homicide Index met the criteria for a systemic, economic compulsive, and/or psychopharmacological link to drugs. This represents 27% of all homicides.

A further 789 (25% of all homicides) had an ambiguous link to drugs and 1,521 (48%) had no known drugs link. Cases with an ambiguous drugs link are those that fall under the Home Office definition of a drug-related homicide but do not meet the criteria for any of the three categories within the Goldstein tripartite framework. They therefore have a less certain link to drugs.

As shown in Table 1, of the total number of homicides that met the criteria for the Goldstein framework between April 2018 and March 2023, 464 (15%) were systemic, 159 (5%) were economic compulsive, and 399 (13%) were psychopharmacological, with some overlap between them. Of the total number of homicides, those that met the criteria for a systemic case only were the most common of the Goldstein categories (337, 11%), followed by psychopharmacological only (307, 10%).

Applying the new framework to Homicide Index data over time

There were 590 homicide victims in the year ending March 2023. This was 94 fewer than the year ending March 2022 (684 homicides, a 14% decrease).

In the year ending March 2023, 144 cases (24% of homicides) met the criteria for one or more of the three Goldstein categories.

Figure 1 shows the proportion of homicides meeting the criteria for at least one of the Goldstein categories. This increased gradually from 26% (143 homicides) in the year ending March 2021 to 30% (203 homicides) in the year ending March 2022, before falling back to 24% in the year ending March 2023 (144 homicides). The year ending March 2021 covers the majority of the coronavirus (COVID-19) pandemic when restrictions were thought to have contributed to a fall in public space homicides. As COVID-19 restrictions eased, the numbers of homicides increased again, but have since fallen.

Trends in systemic, economic compulsive and psychopharmacological homicides

Figure 2 shows the trend in systemic, economic compulsive and psychopharmacological cases. The different categories of homicide are not mutually exclusive, so the figures below should not be summed.

In the year ending March 2023, there were 86 systemic homicides (15% of homicides), a decrease compared with the year to March 2022 (when there were 109 systemic homicides, 16% of all homicides) but an increase compared with the year to March 2021 (70 systemic homicides, 13% of all homicides). Since 1 April 2021, forces have been able to indicate in the Home Office Homicide Index whether homicides were related to organised crime groups (OCGs) or county lines. However, this has had a minimal impact on the volume of cases marked as systemic (see Section 4: Data source and framework section for more information).

In the year ending March 2023, there were 24 economic compulsive homicides (4% of homicides). This was a decrease compared with the year ending March 2022 (31 economic compulsive homicides, 5% of all homicides) and similar to the year ending March 2021 (23, 4% of homicides).

In the year ending March 2023, there were 66 psychopharmacological homicides (11% of all homicides). This was a decrease compared with the year ending March 2022, when there were 100 psychopharmacological cases (15% of all homicides) and the year ending March 2021 (74, 13% of homicides).

Home Office Homicide Index

Data presented have been extracted from the Home Office Homicide Index, which contains detailed record-level information about each homicide recorded by police in England and Wales. The Home Office Homicide Index is an administrative dataset and, therefore, the accuracy of the categories depends on the availability of relevant information. The Home Office Homicide Index is updated as new information becomes available, so historical data is subject to change in later editions of the publication. Further information on the data source is available.

The option to mark homicides as organised crime group (OCG) or county lines related was introduced on 1 April 2021. Homicides that have been marked by forces as related to OCG or county lines are included in the systemic homicide category. This has had a minimal impact on the volume of cases in the systemic category. Since 1 April 2021, two additional cases have been labelled systemic because of the county lines variable and because of the OCG variable. Other cases marked as OCG and/or county lines related in the Home Office Homicide Index also met other systemic criteria so would already have been labelled as such.

It is important to note that like the Home Office's definition of drug-related homicide, this analysis relies on variables in the Home Office Homicide Index rather than reviews of case files. In 691 homicides (22% of all cases) recorded between April 2018 and March 2023, a suspect had not been charged at the time of analysis. Many of the variables used to develop the criteria for Goldstein's framework require suspects, so once more suspects are charged, additional cases may meet the criteria for Goldstein's framework.

Table 2 shows the breakdown of suspects in different types of homicides. All cases labelled economic compulsive and/or psychopharmacological have a suspect because of the variables used to develop their inclusion criteria. Some of the variables used to create the systemic category also require a suspect. These are: suspect and victim are both known illegal drug dealers; suspect was a dealer with motive to obtain drugs or steal drug proceeds; suspect was a dealer and the main method of the homicide was shooting; victim's relationship to the suspect was criminal associate; and suspect was part of a gang. Once more suspects are charged, it is likely that additional homicides will be marked as systemic, economic compulsive, and/or psychopharmacological.

Goldstein’s tripartite framework

Goldstein's tripartite framework is based on three separate empirical investigations undertaken in New York City between 1976 and 1984. These involved interviews with sixty women about the link between sex work and drugs; an ethnographic study of the economic behaviour of 201 street opiate users; and a study of the relationship between drugs and violence.

Goldstein (1985) provides a tripartite framework for more focused understanding of drug-related homicides, which is commonly used in research into the topic. This includes the European Homicide Monitor, who uses it to categorise drug-related homicides across Europe. Moreover, the Home Office's Trends and drivers of homicide research report surveyed the literature on drugs as a driver of homicide trends. Most of the 45 studies identified in the systematic search referenced Goldstein (1985). We therefore identified it as the most suitable method to develop our understanding of drug-related homicide.

The approach has been criticised for being outdated, lacking relevance outside the specific New York City context in which Goldstein's original research took place, and in need of refinement to consider the range of factors that influence the connection between drug use and crime. However, it remains the prevailing theory used to understand drug-related homicide and can provide useful insights about homicides that fall under the Home Office's existing definition.

Drug-related homicide

The Home Office definition of a drug-related homicide includes any offence where any of the following variables are positive: victim illegal drug user; victim illegal drug dealer; suspect illegal drug user; suspect illegal drug dealer; victim has taken a drug; suspect has taken a drug; suspect had motive to obtain drugs; suspect had motive to steal drug proceeds; drug-related.

Homicide suspects

For the purposes of the Homicide Index, a suspect in a homicide case is defined as either a person who has been charged with a homicide offence, including those who were subsequently convicted and those awaiting trial, or a person who is suspected by the police of having committed the offence, but is known to have died or committed suicide. This analysis is based on the principal suspect only.