The ongoing debate about the death penalty in Trinidad and Tobago has been closely related to the rising rates of violent crime. In response to escalating concern about crime levels, in 2011 the government declared a temporary state of emergency in response to ‘specific and emerging threats’.

The table below shows the increase in murders over the past two decades.

Year Number of Murders Recorded by Police
1994 143
1995 122
1996 106
1997 101
1998 98
1999 93
2000 118
2001 151
2002 171
2003 229
2004 260
2005 386
2006 368
2007 395
2008 550
2009 509
2010 485
2011 354
2012 383
2013 408
2014 403
2015 410

Figures from Trinidad and Tobago Crime Statistics (there are a number of sources online, some of which offer slight variation on the figure for particular years).

Murder is often calculated as a rate per 100,000 of the population; this provides a useful way to compare murder rates across different countries.

In 1998, there were 7.6 murders per 100,000 in Trinidad and Tobago. In 2002 there were 13.6 per 100,000. In 2005, the rate was 30.7 per 100,000. In 2013, the rate was 31 per 100,000, and in 2014 the rate was 35.3 per 100,000.

For comparison, in 2005 the murder rate in Jamaica was the highest in the world at 58 per 100,000 (see UNODC/World Bank, 2007); in 2014 this had fallen to 36 per 100,000 (OSAC, 2015).

An escalating homicide rate has fuelled much of the debate on retaining the mandatory death penalty for murder. Consideration of the death penalty should therefore take into account the nature of homicide in Trinidad and Tobago as well as how the criminal justice system deals with homicide cases.

Research on Homicide in Trinidad and Tobago

Research conducted by Roger Hood and Florence Seemungal (2006) on homicide in Trinidad and Tobago showed that the arrest and conviction rates were very low.

Hood and Seemungal concluded that ‘Infliction of the death sentence was therefore both rare and arbitrary.’

Hood and Seemungal collected information on 633 reports of murder and 279 committals on indictment for murder to the High Court, between January 1998 and December 2002. They tracked these cases until the end of 2005 to monitor how the legal system dealt with cases.

Killings were grouped into five broad categories:

  • Killings arising from a gang dispute or related to the trade in drugs including a sub-category where the killing was carried out like an assassination;
  • Killings arising during the commission of another crime, such as robbery or burglary and killings arising from sexual assault;
  • Killings arising from a domestic dispute including not only all killings in which the perpetrator and the victim were related by marriage or other family bonds but also those which arose from common-law relationships or former common-law relationships as well as child abuse and infanticide;
  • Killings as a result of an attack or fight arising from other inter-personal altercations or conflicts, usually between persons known to each other, including killings by police and security personnel in the exercise of their duty, and also arising from inter-personal conflicts where innocent bystanders were killed;
  • Killings where the motive or relationship between victim and killer remained impossible to determine, the body having been found either ‘dumped’ or in other circumstances.

Hood and Seemungal found that murders were not distributed evenly across Trinidad and Tobago, with most occurring in the north, north eastern and Port of Spain areas of Trinidad, rather than the south of the island of Tobago.

The killings had taken place in the home of the victim in 28.8% of cases. Altogether, 36.3% of killings occurred in a general domestic setting; 28.6% occurred on the public street and 9% in commercial buildings or public institutions. In 12.8% of cases, the site of the death was unknown as the body had been dumped following death. The proportion of this last category increased between 1998 and 2002.

Hood and Seemungal found that killings by gunshot also increased between 1998 and 2002, as did the proportion of male victims compared to female victims. These changes reflect the increase in killings related to gang disputes or the drug trade over the years in question. The proportion of killings arising from the commission of another crime also increased.

However, although these types of killing increased, it was less likely that suspects would be apprehended in such cases. In fact, suspects were most often identified and prosecuted in domestic disputes and killings resulting from inter-personal altercations.

The research highlighted the problems with delay and low conviction rates in the criminal justice system. By December 2005, only 33 of 633 murders had resulted in convictions for murder (this was only 5.2%, although including manslaughter convictions the number was 17%). Conviction for murder was lowest for gang-related or murders where the body had been dumped, only two persons had been convicted of murder, and two for manslaughter by end of 2005.

Less than a quarter of defendants had been tried and judgment passed within two years.

The proportion of murders in which no suspects were identified was increasing while the proportion of murders resulting in a conviction was decreasing.

No category of killing had a ‘very high probability of conviction for murder and mandatory death sentence,’ likewise there was also no high probability of conviction for manslaughter in any category.

Conviction rates by category of killing:

  • Domestic conflict (35.7%);
  • Commission of another offence (28.9%);
  • Gang or drug-related (21.7%);
  • Inter-personal altercation (6.4%).

Acquittal rates were higher in gang-related murders and murders related to the drug trade, and killings from inter-personal altercations.

Murder arising from domestic conflict accounted for a fifth of all prosecutions but just over a third of all murder convictions. Murder during the commission of a non-sexual offence accounted for approximately 30% of prosecutions but 43.1% of convictions.

In two-thirds of all cases which led to a prosecution for murder, the victim and offender were known to each other. Within this group, a quarter had been accused of killing an intimate partner or former partner. This group were more likely to be convicted of murder than those who had no prior relationship. They accounted for 41.4% of all those convicted for murder.

279 persons were tried for murder in the five year period 1998-2002. The numbers of persons committed for trial fell over the five year period.

Only 58, or 20.8%, of persons were ultimately convicted of murder, 57 were sentenced to death (one had been under 18 when the killing was committed). All 57 appealed conviction and sentence:

  • In 25 cases conviction and sentence were reaffirmed;
  • In 6 cases the conviction was affirmed but the death sentence was substituted for a life sentence;
  • 4 convictions were quashed and defendant discharged;
  • 7 convictions were quashed and retrials ordered;
  • 8 convictions were quashed and manslaughter convictions substituted;
  • 2 appeals were refused;
  • 5 appeals had yet to be heard by the end of the research period in December 2005.

Additionally, eight of the defendants had their appeals heard by the Privy Council, none of these death sentences survived.

Ultimately, of 57 death sentences, only 23 persons remained under sentence of death and this was likely not the final count as the process of appealing was continuing for many defendants, either domestic appeals or to the Privy Council.

Hood and Seemungal concluded that the majority of persons ‘got away with murder.’ Even for those murders in which a suspect was found and indicted for murder, the rate of conviction was very low and falling. There was a high rate of failure in prosecutions for murder, especially in cases related to gangs or the drug trade.


Hood and Seemungal concluded that the arbitrariness of prosecution, and the very low rates of conviction, reduced the deterrent effect of conviction and sentence. As the risk of death was so low in gang-related killings, it could not be an effective penalty.

‘The evidence adduced by this study adds weight to the arguments of those who maintain that the retention  of the mandatory death penalty for murder in Trinidad and Tobago serves no useful purpose, is arbitrary and unfair in its enforcement, and may well be counterproductive.’

Ironically, ‘crimes of passion’, those arising from domestic altercations, were the most likely to end in murder convictions, and are the least amenable to deterrence. The killings which are most likely to end in conviction and death sentence are also not those which are typically deemed the ‘worst of the worst.’

Hood and Seemungal argued that law enforcement must increase the certainty of conviction.

‘The implications of the findings of this study are inescapable – the chances of a person who committed a murder in Trinidad and Tobago suffering sentence to death were very rare.’

The authors also concluded that there were numerous difficulties in bringing a successful prosecution, including: witness intimidation, reluctant witnesses, reliance on oral testimony, lack of scientific evidence, and long delays.

The authors argued that the mandatory death penalty may also have dissuaded some jurors from voting to convict of murder.

‘It may well be that if the death penalty were abolished, or failing that the mandatory death penalty, witnesses, prosecutors and juries would be more likely to make sure that those accused of murder who are guilty of murder are convicted of murder and not of a lesser offence.’