Despite representing just 8% of the world’s population, the region accounts for nearly one-third of global homicides.
Breaking Latin America’s Cycle of Low Growth and Violence
North and Sub-Saharran Africa are poorer than Latin America but have much less murder going on. What’s going on in Latin America?


Those are relative explanations as in relative to the region. Compare it to Africa where the crime rate and murder rate is quite different despite there being active wars on the continent. Violence is of course present but not at the same level as in South America.
Relative?
There is a long list of correlations.
A lot of the “relative” is within the same country and just a varied time frame so your different reguon interpretation is flawed.
You seem to need simplistic answers and the quote above pretty clearly points to recessions, gang activity (presumably related to drugs), inequality and governance as having some correlation (but I’m lazy and didn’t properly read again). You could look up rates and compare if you wanted to have an idea of where some things might be different. That would be hugely speculative though
Are you conflating violence (a broad term) with homicide?
This is entirely speculation but the fact that there are active wars in the continent might affect how the data is classified. I don’t know how the article you posted defines homicide. There are some rules here https://www.unodc.org/unodc/en/data-and-analysis/statistics/iccs.html and it seems that deaths during war conflicts might not be counted as intentional homicides. Latin America hasn’t had many wars but had/has many conflicts involving guerrillas, cartels and political groups. Is it possible that many of the resulting deaths are counted as homicides whereas similar violence in Africa is counted as, for example, civil war deaths?
I’m inclined to agree:
https://ourworldindata.org/grapher/homicide-rate-unodc
This source has made a clear definition of homicide that agrees with what you’re saying. But in this data, a lot of Africa is missing and I think that includes well know areas with wars. So depending on how that gap in data is filled in could introduce errors. I didn’t bother checking the OPs source
if things are static in africa, then the base level of violence may be maintained, wheras increases in factors can lead to increases in violence.