What is ‘Other’? Inside TTPS’ homicide classifications Pt1

Every year, Trinidad and To­ba­go mea­sures its bat­tle against vi­o­lent crime us­ing a sin­gle num­ber: mur­ders.

But the coun­try’s most im­por­tant crime sta­tis­tic de­pends on a clas­si­fi­ca­tion de­ci­sion the pub­lic nev­er sees.

This Guardian Me­dia two-part in­ves­ti­ga­tion is not about count­ing the dead. It is about un­der­stand­ing which draw­er each death is placed in—and who or what mech­a­nism de­cides which death makes it to the fi­nal year­ly mur­der toll.

Eliz­a­beth Gon­za­les

Se­nior Re­porter

eliz­a­beth.gon­za­[email protected]

Five hun­dred deaths record­ed by the Trinidad and To­ba­go Po­lice Ser­vice (TTPS) over more than a decade have been placed in a cat­e­go­ry sim­ply la­belled “Oth­er”—but the Po­lice Ser­vice has not pro­vid­ed a writ­ten de­f­i­n­i­tion of what that cat­e­go­ry means.

A lot of at­ten­tion is paid to the coun­try’s mur­der fig­ure month­ly and an­nu­al­ly. But there is one cat­e­go­ry that stands out be­cause it con­tains hun­dreds of deaths: “Oth­er”.

Be­tween Jan­u­ary 1, 2016 and June 3, 2026, the TTPS record­ed 500 deaths and placed them in that cat­e­go­ry, ac­cord­ing to a Crime and Prob­lem Analy­sis Branch (CA­PA) doc­u­ment ob­tained un­der the Free­dom of In­for­ma­tion Act.

CA­PA is the Po­lice Ser­vice unit that com­piles and analy­ses crime da­ta.

Those 500 deaths were part of 946 record­ed out­side the mur­der cat­e­go­ry in the Po­lice Ser­vice’s wider homi­cide ta­ble.

The ta­ble record­ed 6,157 To­tal Homi­cides: 5,211 Mur­ders, 321 Un­clas­si­fied Deaths, 68 In­quests, 57 Manslaugh­ter cas­es and 500 Oth­er.

Oth­er was the largest of the four cat­e­gories out­side of Mur­der.

Yet the ta­ble does not ex­plain what Oth­er means. It con­tains no de­f­i­n­i­tion, no break­down of the deaths it in­cludes and no in­di­ca­tion of why those cas­es were placed there in­stead of else­where in the po­lice clas­si­fi­ca­tion sys­tem.

The ex­is­tence of the cat­e­go­ry does not mean those 500 deaths were mur­ders that went un­count­ed. The records do not sup­port that con­clu­sion.

The ta­ble lists Oth­er sep­a­rate­ly from Un­clas­si­fied Deaths. TTPS has not pro­vid­ed Guardian Me­dia with a writ­ten de­f­i­n­i­tion show­ing what be­longs in Oth­er. If Un­clas­si­fied Death al­ready ex­ists for deaths po­lice have not yet re­solved, what ex­act­ly is Oth­er?

The CA­PA ta­ble treats Oth­er and Un­clas­si­fied Death as sep­a­rate cat­e­gories. On pa­per, they are not in­ter­change­able.

Deputy Com­mis­sion­er of Po­lice Ju­nior Ben­jamin al­so said they were not the same.

“I think that ‘Oth­er’ is clas­si­fied, but it is clas­si­fied out­side of a mur­der and a manslaugh­ter,” he said. “But un­clas­si­fied, it means that we have not had enough ev­i­dence, ei­ther for foren­sic or oth­er­wise, to de­ter­mine ex­act­ly what cat­e­go­ry it should be.”

But his next ex­pla­na­tion blurred that dis­tinc­tion. He de­scribed a de­te­ri­o­rat­ed body still await­ing foren­sic find­ings—a sit­u­a­tion he had just as­so­ci­at­ed with Un­clas­si­fied Death.

“We find a body, right? We do not know—the body is de­te­ri­o­rat­ed,” Ben­jamin said. “We would then need to find out from foren­sic whether or not they saw pro­jec­tiles in the body, or whether there was some stran­gu­la­tion. Dif­fer­ent things will tell you whether it was an un­nat­ur­al death, whether it was by nat­ur­al caus­es. So you have to wait un­til the foren­sic re­sults to de­ter­mine what di­rec­tion it will go and whether or not you will prob­a­bly place it in the Oth­er cat­e­go­ry or in the mur­der cat­e­go­ry.”

Ben­jamin said his un­der­stand­ing was that Oth­er in­clud­ed sui­cides (even though Oth­er made up part of the to­tal homi­cide fig­ure) and ve­hic­u­lar manslaugh­ter or ve­hic­u­lar homi­cides. His an­swer was not a writ­ten de­f­i­n­i­tion, and the bound­ary be­tween Oth­er and Un­clas­si­fied Death re­mained un­clear.

Guardian Me­dia at­tempt­ed to ob­tain a writ­ten de­f­i­n­i­tion. When Ben­jamin was asked about the method­ol­o­gy TTPS us­es to clas­si­fy deaths, he re­ferred the query to Se­nior Su­per­in­ten­dent Sean Dhilpaul, head of the Homi­cide Bu­reau of In­ves­ti­ga­tions.

When con­tact­ed, Dhilpaul said TTPS us­es a glos­sary to de­ter­mine the mo­tive as­signed to homi­cide cas­es.

“We have a glos­sary. TTPS has a glos­sary. And that is how we de­ter­mine. Based on the glos­sary, that is where the mo­tive will be giv­en from,” he said. He added that the glos­sary “itemis­es the cer­tain cat­e­gories” used to as­sign mo­tives.

Dhilpaul said while the glos­sary was used to as­sign, it did not de­ter­mine whether a death moved from Un­clas­si­fied Death to Mur­der.

“That would be based on the in­ves­ti­ga­tion,” he said. “Nor­mal in­ves­tiga­tive pro­ce­dures and ev­i­dence would de­ter­mine how a death moves from un­clas­si­fied to whether be re­clas­si­fied or re­main un­clas­si­fied.”

When Guardian Me­dia asked what types of deaths are placed un­der the cat­e­go­ry “Oth­er”, Dhilpaul did not pro­vide an an­swer.

“Let me ver­i­fy one thing with re­gards to Oth­er,” he said. No fur­ther ex­pla­na­tion was pro­vid­ed af­ter.

While the se­nior of­fi­cers could ex­plain parts of the sys­tem, the de­f­i­n­i­tion of Oth­er re­mained un­re­solved.

Guardian Me­dia turned to the Free­dom of In­for­ma­tion Act for the writ­ten rules and records be­hind the clas­si­fi­ca­tions.

On Jan­u­ary 26, 2026, Guardian Me­dia asked TTPS for “the to­tal num­ber of cas­es ini­tial­ly clas­si­fied as: Homi­cide; Mur­der; [and] Sus­pect­ed Homi­cide.”

It asked for cas­es “sub­se­quent­ly re­clas­si­fied from Homi­cide/Mur­der to an­oth­er cat­e­go­ry—in­clud­ing but not lim­it­ed to ac­ci­den­tal death, sui­cide, sud­den death or ‘no clas­si­fi­ca­tion’—be­tween 2016 (sic) present.”

It al­so re­quest­ed cas­es “marked as ‘not clas­si­fied’ or oth­er­wise re­moved from Homi­cide Sta­tis­tics.”

Then it asked for the ma­chin­ery be­hind the num­bers: “the cri­te­ria or pol­i­cy used to clas­si­fy and re­clas­si­fy Homi­cide cas­es”; “any in­ter­nal guide­lines, stand­ing or­ders or man­u­als gov­ern­ing case re­clas­si­fi­ca­tion”; who had au­thor­i­ty to ap­prove a change; whether su­per­vi­so­ry ap­proval was re­quired; the num­ber changed each year; and the gen­er­al rea­sons for those changes.

The first re­sponse showed an­oth­er gap.

On April 13, the TTPS ap­proved the first sta­tis­ti­cal item and sup­plied a ta­ble head­ed “Homi­cides and Mur­ders for the pe­ri­od 2016-2025 as at 18.02.2026.”

But the rules were still miss­ing.

The Le­gal Unit wrote: “We have made sev­er­al re­quests for the in­for­ma­tion; how­ev­er, due to the de­cen­tralised na­ture of the TTPS, we are await­ing the pro­vi­sion of same from the rel­e­vant di­vi­sions.”

TTPS sought an­oth­er 30 days “to ren­der ac­cess de­ci­sions on the out­stand­ing items” by May 13.

The ta­ble was use­ful, but it was not the fi­nal dataset used in this sto­ry. It pre­sent­ed “Homi­cide” dif­fer­ent­ly from the lat­er ta­ble and some fig­ures changed. It list­ed 369 mur­ders for 2025.

The CA­PA ta­ble list­ed 370.

One mur­der case had en­tered the mur­der col­umn af­ter the ear­li­er fig­ure was pro­duced. The records did not iden­ti­fy the case, the date of the change or whether the re­vised toll was for­mal­ly an­nounced.

The records es­tab­lished that fig­ures could change. The next ques­tion was what gov­erned those changes.

The next day, a sec­ond Le­gal Unit re­sponse an­swered part of the re­quest. This time, TTPS spoke di­rect­ly about re­clas­si­fi­ca­tion.

The doc­u­ment said: “We are in­struct­ed by the Se­nior Su­per­in­ten­dent of Po­lice, Homi­cide Bu­reau of In­ves­ti­ga­tions that the re-clas­si­fi­ca­tion of any un­nat­ur­al death is de­pen­dent on three (3) com­po­nents.”

The first was: “The pro­nounce­ment of the post mortem ex­am­i­na­tion con­duct­ed by a Pathol­o­gist of the Foren­sic Sci­ences Cen­tre, in­clu­sive of fur­ther foren­sic ex­am­i­na­tion of Bi­ol­o­gy, Chem­istry, Bal­lis­tics and DNA analy­sis.”

The sec­ond was: “The ad­vice re­ceived from the Di­rec­tor of Pub­lic Pros­e­cu­tions hav­ing pe­rused the con­tents of the in­ves­tiga­tive file sur­round­ing the cir­cum­stance where the death tran­spired.”

The third was: “The dis­cre­tion of the As­sis­tant Com­mis­sion­er of Po­lice, Crim­i­nal Di­vi­sion or Se­nior Su­per­in­ten­dent Homi­cide Bu­reau of In­ves­ti­ga­tions. This as­sess­ment re­lies on the ev­i­dence gath­ered ini­tial­ly on the scene and dur­ing the course of the in­ves­ti­ga­tion.”

It did not pro­vide the re­quest­ed man­u­al, stand­ing or­der or pol­i­cy. It did not de­fine Mur­der, Un­clas­si­fied Death, Oth­er, In­quest or Manslaugh­ter. It did not ex­plain who made the first clas­si­fi­ca­tion, who checked it or what writ­ten rea­son was pre­served when it lat­er changed.

And it still did not ex­plain Oth­er.

Guardian Me­dia re­turned to TTPS on May 7 with an­oth­er re­quest.

It asked, by year, for “the to­tal num­ber of deaths clas­si­fied as ‘Un­clas­si­fied Deaths’”; the num­ber lat­er re­clas­si­fied as Mur­der; the num­ber clas­si­fied as “Un­nat­ur­al Deaths”; the num­ber of those lat­er moved to Mur­der; and the to­tal fig­ures for Manslaugh­ter and In­quest.

It al­so sought “any ag­gre­gate records, sta­tis­ti­cal ta­bles or an­nu­al sum­maries show­ing move­ment or re­clas­si­fi­ca­tion be­tween the cat­e­gories Un­clas­si­fied Death, Un­nat­ur­al Death, Mur­der, Manslaugh­ter and In­quest.”

On June 15, the TTPS replied: “the num­ber of Un­clas­si­fied Deaths is 321”; “the num­ber of Un­clas­si­fied Death cas­es re­clas­si­fied as Mur­der is 60”; “the num­ber of deaths clas­si­fied as Un­nat­ur­al Deaths is 1,504”; “the num­ber of Un­nat­ur­al Death cas­es sub­se­quent­ly re­clas­si­fied as Mur­der is 60”; “the num­ber of cas­es clas­si­fied as Manslaugh­ter is 57”; and “the num­ber of cas­es clas­si­fied as an In­quest is 68.”

The at­tached ta­ble was where Oth­er ap­peared in full.

For 2025, it record­ed 492 To­tal Homi­cides. Of those, 370 were Mur­der. The re­main­ing 122 were di­vid­ed among 41 Un­clas­si­fied Deaths, eight In­quests, eight Manslaugh­ter cas­es and 65 Oth­er.

Across the full ten-year pe­ri­od, Oth­er reached 500. (See ta­ble 2)

TTPS pro­vi­sion­al fig­ures

Year - Mur­der - Un­clas­si­fied Death - Re­clas­si­fied from Homi­cide/Mur­der to In­quest - Re­clas­si­fied from Homi­cide/Mur­der to Manslaugh­ter - Oth­er - To­tal Homi­cides

2016 464 28 8 7 23 530

2017 496 11 4 9 5 525

2018 518 41 8 4 4 615

2019 539 44 14 4 4 605

2020 403 43 7 4 6 463

2021 452 33 7 4 15 511

2022 606 19 11 5 6 647

2023 580 16 6 4 6 612

2024 626 32 7 6 4 675

2025 370 41 8 8 65 492

2026 157 13 1 2 18 191

To­tal 5,211 321 68 57 500 6,157

Source: TTPS pro­vi­sion­al fig­ures cov­er­ing Jan­u­ary 1 to June 3, 2026.

The ta­ble showed the size of Oth­er. It still did not show what was in­side it.

Over the past three months, Guardian Me­dia re­viewed an archive of 494 in­ter­nal TTPS morn­ing-re­port files (dai­ly re­ports), cov­er­ing 187 re­port dates from 2023 to 2025.

The re­ports cov­ered mur­ders, shoot­ings, rob­beries, thefts, bod­ies found and oth­er in­ci­dents. Deaths ap­peared un­der head­ings in­clud­ing Homi­cide, Un­clas­si­fied Death and Sus­pect­ed Homi­cide.

The doc­u­ments showed that clas­si­fi­ca­tion was not on­ly a year-end ex­er­cise. An ini­tial la­bel en­tered the dai­ly re­port­ing process ear­ly enough to ap­pear in the morn­ing re­port.

The pub­lic, how­ev­er, nor­mal­ly re­ceived some­thing nar­row­er.

“What we would nor­mal­ly send out on a dai­ly ba­sis will not have all that,” Ben­jamin said. “I’m watch­ing what we send out now, and all it has on it, it will re­al­ly just show the to­tals in terms of mur­ders. That’s what we nor­mal­ly would send out.”

He said the wider analy­sis con­tained the dif­fer­ent types of homi­cide and po­lice cat­e­gories. “But that is not some­thing that we send out on a dai­ly ba­sis,” he said.

A death clas­si­fied as mur­der could en­ter the pub­lic toll sent to the me­dia dai­ly for pub­li­ca­tion. A death placed else­where re­mained out­side that num­ber while the in­ves­ti­ga­tion con­tin­ued.

Ex­am­ple of TTPS dai­ly toll no­tice sent to the me­dia:

Dear col­leagues. Please see the mur­der toll for to­day, Wednes­day 5 Au­gust, 2026

COM­PAR­A­TIVE YEAR BY YEAR TO­TAL

2025: 233

2026: 223

COM­PAR­A­TIVE MONTH­LY TO­TAL

Au­gust 2025: 4

Au­gust 2026: 0

The fol­low­ing da­ta con­cerns Un­clas­si­fied Death and Sud­den Death. It does not ex­plain the 500 deaths record­ed un­der Oth­er.

Ben­jamin said an Un­clas­si­fied Death was not placed un­der the mur­der cat­e­go­ry un­less po­lice lat­er clas­si­fied it as mur­der.

“Un­clas­si­fied mur­ders will not be count­ed at all in our mur­der fig­ures at the end of the year, be­cause it is un­clas­si­fied,” he said. “We don’t know whether it is a homi­cide, a mur­der or not. So it will not be on our mur­der fig­ures. On­ly when it is clas­si­fied, and it is clas­si­fied as a mur­der, it will be part of the mur­der fig­ures.”

He said that move­ment could hap­pen af­ter the year had end­ed.

“A lot of times, when it is re­clas­si­fied, some­times af­ter the year has passed, all the nec­es­sary foren­sic ev­i­dence [is need­ed] to make a de­ter­mi­na­tion as to whether or not it should be a mur­der or not,” Ben­jamin said. “Some­times it is nor­mal­ly af­ter the pe­ri­od when we nor­mal­ly check it, which is the year. So you might see one or two in­creas­es in the num­bers for that year, prob­a­bly a year or two af­ter, as we get the in­for­ma­tion. Be­cause some­times, we have to wait a while for foren­sics to give us the in­for­ma­tion.”

The at­tached CA­PA ta­ble showed that change.

Be­tween 2016 and June 3, 2026, 60 Un­clas­si­fied Death cas­es were moved to Mur­der. The largest group was 29 in 2020. The re­main­ing changes were spread across 2018, 2019 and every year from 2021 to 2026.

Six Sud­den Death cas­es were al­so re­clas­si­fied as mur­der. Those changes ap­peared in 2023, 2024, 2025 and 2026.

Year Mur­der toll Un­clas­si­fied Deaths % of Mur­der Toll

2016 464 28 6.0%

2017 496 11 2.2%

2018 518 41 7.9%

2019 539 44 8.2%

2020 403 43 10.7%

2021 452 33 7.3%

2022 606 19 3.1%

2023 580 16 2.8%

2024 626 32 5.1%

2025 370 41 11.1%

2026 (up to June) 157 13 8.3%

To­tal 5,211 321 6.2%

The ta­ble sup­port­ed Ben­jamin’s ex­pla­na­tion that Un­clas­si­fied and Sud­den Death clas­si­fi­ca­tions can change. But it failed to ex­plain any move­ment in­to or out of Oth­er. TTPS pro­vid­ed re­clas­si­fi­ca­tion fig­ures for Un­clas­si­fied and Sud­den Death, but no equiv­a­lent move­ment trail for the 500 deaths record­ed un­der Oth­er.

The doc­u­ments then cre­at­ed an­oth­er con­flict.

The June cov­er let­ter said 60 “Un­nat­ur­al Death” cas­es had been re­clas­si­fied as Mur­der. The at­tached ta­ble said 60 “Un­clas­si­fied Death” cas­es.

The records did not ex­plain whether that was a copy er­ror, over­lap­ping cat­e­gories or two dif­fer­ent mea­sures with the same to­tal.

That in­con­sis­ten­cy did not in­volve Oth­er, but it showed why the Po­lice Ser­vice’s cat­e­gories can­not be treat­ed ca­su­al­ly or in­ter­change­ably.

A se­nior of­fi­cer has claimed that some deaths which ap­peared to be mur­ders were be­ing kept away from the mur­der count un­der Un­clas­si­fied Death.

“Usu­al­ly, most of the time, those are mur­ders,” the source al­leged.

“But what hap­pens is they keep it in an un­clas­si­fied fold­er. They can just put it in un­clas­si­fied. That means they can keep the stats down,” the se­nior TTPS source claimed.

Asked when those cas­es were re­clas­si­fied, the source replied: “They will just leave it there. They’ll leave it in un­clas­si­fied.”

The source al­so claimed the pub­lic was not nor­mal­ly told about that cat­e­go­ry.

“They don’t tell you about the un­clas­si­fied,” the source said. “They don’t tell you that so the per­son won’t ask, ‘How much un­clas­si­fied do we have?’”

The al­le­ga­tion echoed wider sus­pi­cion about whether the falling mur­der count ac­cu­rate­ly de­pict­ed the coun­try’s crime sit­u­a­tion, in­clud­ing ques­tions raised on so­cial me­dia.

Guardian Me­dia test­ed the al­le­ga­tion against the FOIA records, the in­ter­nal morn­ing re­ports and in­ter­views with se­nior po­lice of­fi­cials. Those records con­firmed a broad­er clas­si­fi­ca­tion sys­tem. They did not prove the source’s claim.

Deputy Com­mis­sion­er of Po­lice Ju­nior Ben­jamin re­ject­ed the al­le­ga­tion.

“Again, I beg to dis­agree,” Ben­jamin said. “There must be cer­tain in­gre­di­ents that bring about a mur­der. If we can­not de­ter­mine the means by which a per­son died, even though a per­son might say, ‘I see some­body kill that per­son,’ you can’t just take a per­son at their word. The ev­i­dence must sup­port that. We do all due dili­gence.”

He ques­tioned why TTPS would ma­nip­u­late the toll on­ly af­ter mur­ders be­gan falling.

“When we got 600 and some­thing mur­ders, no­body charged us for math fix­ing,” Ben­jamin said. “If we gave the truth, the whole truth and noth­ing but the truth, why, when we bring some­thing good now, peo­ple think it’s math fix­ing? We showed that there was an in­crease in mur­ders. The ques­tion is what cre­at­ed the change.”

He main­tained: “I be­lieve that every sin­gle mur­der is ac­count­ed for. On­ly those that we can­not and have not re­ceived the rel­e­vant ev­i­dence to de­ter­mine the clas­si­fi­ca­tion; those are the ones that are not clas­si­fied. Oth­er than that, all homi­cides are clas­si­fied ac­cord­ing­ly.”