Honduras solves its crime problem

Cartoon by Dario Banegas, La Prensa, Honduras

Honduras has solved its crime problem in the only way the Honduras government knows how — manipulate the statistics!

Honduras' Observatorio de la Violencia (OV) has announced that they will no longer provide crime reports and statistics because the Honduras Minister of Security has refused to provide data to the OV.

Note: See update Observatorio will continue to provide crime reports using other sources.

The Observatorio has been the only objective source of Honduran crime information since 2005. It is supported by the United Nations and the Swiss Cooperation Agency. It operates out of and under the supervision of UNAH, the national university system. In 2011, after the UN proclaimed Honduras the "Murder Capital of the World" and the Rector of the UNAH system's son was killed by police, worldwide attention was devoted to Honduras' crime statistics.



Then President Lobo and his Minister of Security began putting on hundreds of "shows" — road blocks where police check auto registration and driver's licenses — to show that the police were hard at work on crime. Numerous government announcements began coming out that "Crime has been reduced in Tegucigalpa by 70%", "No murders have occurred in San Pedro for six days", and President Lobo once even had the audacity to proclaim that "crime has been reduced to practically zero". After a decade of rising murder rates, all this supposedly happened practically overnight. But the people weren't buying the propaganda. They knew better. So did the OV. The 2011 homicide report showed a significant increase over the already high levels.

Numbers within the columns indicate the homicide rate per 100,000

Shortly after Minister of Security Arturo Corrales was appointed by President Juan Orlando Hernández, the battle of the crime statistics began. Police would announce the number of murders or reduction in murders with nothing but their word to back it up, but the OV would announce a higher number. In addition to police and morgue reports, the OV compares all of the murders against the national newspaper reports. They were often finding that newspapers reported murders (complete with details and photos) that weren't included in the police statistics. In the months preceding Corrales' denial of data, the following chart shows the types of discrepancies that the OV was finding.

Click to view larger image

I've discussed in the past the many reasons why the police statistics are understated but just briefly, a) murders aren't reported, b) murders are reported but are deliberately not recorded by local police, c) bodies are just buried by perpetrators in clandestine mass graveyards or they just disappear and are never processed by the morgues, and d) there are no reporters in most of the towns in Honduras. Now I would add to that list deliberate government manipulation of murder statistics. Similarly, the OV statistics are understated because there are no reporters in much of the country's small remote towns. I believe that even though the OV statistics are more accurate, they are probably understated by a minimum of 10%, possibly more.

About mid-year 2013, Minister of Security Arturo Corrales began prohibiting police from talking to reporters and refusing to provide data to the OV. He then announcing that the OV statistics couldn't be accurate only the police could determine whether a death was a murder or not — that's a bold statement considering that the majority of violent deaths go uninvestigated. He is apparently saying that without an autopsy, a death cannot be considered a homicide. He realizes, of course, that many murder victims are not autopsied. For the state of Olancho alone, 232 bodies of a total of 278 violent deaths were not sent to Tegucigalpa for autopsy during the first eight months of 2010, resulting in an 83% understatement for that period. It doesn't take an expert to determine that someone shot five times in the back of the head is a murder victim, autopsy or not.

The OV struggled through 2013, gathering and analyzing what data they could obtain from the Criminal Investigative Unit, the police stations, the morgues, and newspapers, and completed the 2013 reports, which of course showed a reduction in murders from 2012, but still not as large a reduction as Corrales claimed. Despite outraged denuncias of lack of transparency, both within Honduras and internationally, no action was taken to support the OV. Corrales' decisions apparently have the support of President Hernández, since he certainly has the power to make Corrales provide the data.

Minister Corrales has additionally discussed setting up "Observatorio" committees in each of the approximately 300 municipalities to oversee the police data. One way to make sure that nothing gets done in Honduras is to appoint a committee to handle the project, even better 298 municipal committees. Committees usually consist of some number of government officials, one church member, one person representing civil society, one person representing labor unions, etc. Citizen appointees often have no particular expertise in the area or special knowledge or ability to analyze the issue, and may not even have an interest in the subject. It's not unusual that some committees never even meet, or if they do, they don't take their responsibilities seriously. In any case, they are always outnumbered by government representatives who control the show and make sure that even if the committee comes up with concrete results or suggestions, those results are never acted upon but instead languish in a Minister or President's desk for years.

The police circumvented the transparency issue by developing their own crime data website. While there are some good things about this website, let me count the ways it is inadequate and inaccurate:

1. Data from one page to another on the site isn't consistent. For example, a data report on 2013 murders in the state of Atlántida results in 464 murders. The interactive map reports 309 for the same state and period. Also shown on this map is an unknown figure of 56 which, if it is supposed to be included, results in a total of 365 murders.

2. The interactive map gives numbers, but no indication of the size of the areas covered by those numbers. The map is a major road map, with the state lines only lightly marked. Those state lines disappear at larger and smaller zoom levels. Numbers are sometimes reflected within one state's borders when they actually relate to another. Additionally, there is no population data. The worldwide standard manner of reporting homicides per 100,000 population is not used at all on this website. The various OV maps and charts were much more informative.

3. The interactive map is inaccurate. It gives different results for the same searches. As discussed in 1. above, the first couple of times I searched for homicides, all sexes, departamento (state) of Atlántida, all municipalities, from 1/1/13 through 12/31/13, the results showed 309 vs. the total of 464 reported in the data section. In another map search of all states, Atlántida was shown to have 457 murders. When I went back to double check, it showed 327 murders plus some miscellaneous numbers, some of numbers are shown in locations that aren't even in Atlántida. Here is a screen shot:

Click for larger view

If you individually select each of the eight municipalities in Atlantida, you get these results: La Ceiba* 280 murders, El Porvenir 27, Esparta 16, Jutiapa 21, La Masica 14, San Francisco 14, Tela 74, and Arizona 8, which totals 454, yet another number that doesn't agree with either the basic data or the statewide data. I have no idea what the 56 to the west of La Ceiba represents. No combination of municipalities in that area gives a result of 56. The blue 2 appears to be over Sonaguera which is not in Atlántida and the 78 appears to be in Yoro. However, if you add those four numbers, you get 463, which is close to the Atlántida total given in the data section.

*In the data section, the report for Registro de Fallecidos (register of deaths) shows that La Ceiba reported 288 deaths in 2013.

If all of that isn't bad enough, I rechecked Atlántida's eight municipality figures and the second time, San Francisco became 13 instead of 14, and Arizona became 9 instead of 8. Apparently, it is possible to make the same selections and receive different results each time.

3. The interactive map is unreliable. The numbers reported change as you zoom in or out. For example, if you zoom in just enough to see a number for Sonaguera (southeast of La Ceiba), the map shows 195 plus 139 in Tocoa and 2 in an unnamed town for a total of 336 murders in the state of Colón. Zoom in one more click and you see Sonaguera 53, Trujillo 62, Tocoa 103, Bonito Oriental 16, plus 5, 2, and 2 in which the towns aren't shown for a total in Colón of 243 murders. (The 22 in Jutiapa is in the state of Atlántida.) Zoom in one more click and you see Sonaguera 38, Trujillo 61, Tocoa 100, Saba 14, Taujica 2, and 2, 5, 2, 2 for which no towns are shown plus 2 blue pins with no numbers, for a total of 228 Colón murders. The data section of the website indicates there were 244 murders in Colón.

4. Data is not "linkable" or downloadable for researchers.

5. Data is missing or duplicated. On the Registro de Fallecidos for Atlantida, I happened to notice a 10-year-old victim named Bairon Noe Muñoz on page one, victim #145. Then I saw the same exact murder also reported on page two, victim #274. The total number on that report is 464 so either his murder was counted twice or at least one other murder victim is omitted.

Click for larger view

I then searched for Héctor Ramos, a well-known La Ceiba businessman who was kidnapped and then murdered after the family had agreed to pay the ransom in December 2013. He was not included on the list on the statewide report. However, in searching for Hector, I found three other Hectors who were also duplicated, Hector Cruz, victim 149 and 278; Hector Sevilla, victim 183 and 312; Hector Coca, victim 191 and 320. In searching the Registro de Fallecidos for La Ceiba, I again found that Hector Ramos' name was missing and the three other Hectors were duplicated, but a fourth Hector, Hector Rivas, was listed who was not included on the Registro de Fallecidos for Atlántida.

I searched for another high-profile murder victim, Nedenia Post Dye of the Post cereal family whose December 2013 murder on Roatán made international news. She was not listed among the 21 victims on the Islas de la Bahia Registro de Fallecidos. Then I searched for Juan Ramón Lopez, who was killed February 7, 2013 near Tocoa, Colón. He also was not listed.

6. Listings of murder victims are not arranged in any order within municipalities and states. To see if a specific murder victim is listed requires scanning through hundreds of names. To search for all the duplicates like I found in 3. above would be impossible.

7. 2014 data is not available in either the detailed data reports or the map. If you wanted to verify that a murder victim was listed or compare last week's newspaper reports with the police data, I guess you would have to wait until next year.

8. Data is laughable in some cases. For example:

a) The "Detinidos" (detained) section of the June 15, 2014 weekly report compares the number of detainees for faltas varias (minor issues) for January 1 through June 15 as 29 in 2013 and 23,978 in 2014. First of all, 'detained', like 'suspended', is a word with no defined meaning which is used by the police department to confuse the media and population by implying one thing when what actually happened is something different. Detained could mean that a criminal was apprehended and put in jail. It could also mean that police stopped a vehicle for 5 minutes or that a teenager was thrown into jail overnight for absolutely no reason and then released the next day with no charges. If I had to guess, these 23,978 would largely be made up of the number of innocent citizens stopped at road blocks to allow the police to check their driver's license and auto registration.

b) Similarly, the "Patrullajes" (patrols) section of that same report include numbers which can only be a result of creative accounting. I'd like to believe that the police are doing more patrols -- I want to believe that! -- but if that is happening, it is primarily in San Pedro and Tegucigalpa and I simply can't believe these numbers: patrols on foot, 2013: 4,837, 2014: 38,007; patrols in vehicles, 2013: 2,524, 2014: 63,855; patrols in motorcycles, 2013: 13, 2014: 19,755. And what is the definition of a patrol? Is it an 8-hour shift or a 10-minute drive around a couple of blocks?

c) For the week ending June 15, this report shows a 24% reduction in homicides. Sorry. I do not believe that for one minute! Overall, the police are showing an almost 18% year to date reduction in homicides. I don't believe that either. I imagine that there are file cabinets full of cases awaiting an investigation or determination that will never happen. I've seen one of those file cabinets in La Ceiba!

d) In the data section for La Ceiba, a municipality of around 250,000 population and the third largest city in Honduras, the police report one robbery of persons and two robberies of homes! The actual numbers would be closer to 1,000 times those numbers. This can't entirely be blamed on the police because the vast majority of crimes are never reported as people know that it is a complete waste of time or that the police might even be involved.

There are some good things about the website, if it was accurate. Listing the victims by name could result in more credibility for the police statistics. Unfortunately, it did the opposite for me since I couldn't find any of the victims I looked for. Additionally, being able to zoom in on the map to the aldea (town) and barrio level to show where the murders occurred is also helpful.

Overall, using the website is tedious and inefficient. To check several categories or departments for any period, you have reenter each of the selections all over again, and click back through 12 calendars to select dates each and every time. Compiling usable data for comparison purposes or even searching for a specific murder is near impossible. The website definitely was not designed with users needs in mind and was not adequately tested to ensure accurate results.

I'm convinced that this was the point of the Sepol website: To be able to say that the police are transparent, while at the same time not providing usable or accurate data and analyses. Having the notoriously corrupt police department provide the only source of information is like having soccer players decide whether or not they have committed a penalty. There is no reason that the Observatorio de la Violencia not continue as the only serious and objective source of Honduras' crime data. The Sepol website is not going to fool anyone, least of all the UN, who ignores government statistics that they find unreliable.

In order to provide real transparency, President Juan Orlando Hernández should order the Minister of Security to provide any and all data requested by the OV. Unfortunately, the truth is not helpful in attracting business and tourists to Honduras which is the President's focus. But the truth is what it is and the Honduran government should be doing more to protect its population and less in trying to obfuscate the facts.


See the update, June 23: The Observatorio will continue publishing crime reports