(Mexico) State highway improvements in Chihuahua

[mgk920]

And what is really sad is that ALL of the subject matter of this thread drift is directly related to the Drug War™.



Mike

[realjd]

Great job! And to think I've managed to avoid being randomly killed by roving patrols of criminals on both urban and rural highways!

The point was that travel warnings are often overblown. When I went to London in April, the US State Department had a terrorism warning for the UK. I saw huge labour party protests and mild anarchist rioting, but no terrorism.

[nexus73]

From today's news and you can count on more news like this coming soon:

.Mexico's northern state of Nuevo Leon, home to the industrial city of Monterrey, saw 33 killings Wednesday, in the most violent day in recent history, officials said Thursday.

The prosperous region, home to many foreign companies, was until recently considered a safe haven as drug violence increased in parts of Mexico.

But a bloody turf war between the Gulf cartel and its former hitmen the Zetas has spilled into the state in less than two years, producing daylight shootouts, grenade attacks and a climate of fear.

A spokesman for Nuevo Leon government, who requested anonymity, said 33 violent killings were recorded in and around state capital Monterrey, an area of some four million, on Wednesday.

Before then, the highest number recorded were 18, on May 20, including 14 in a deliberate fire in a jail in Apodaca town.

Many of Wednesday's killings bore the signs of drug gang vendettas, including the discovery of the mutilated corpses of two bodyguards of State Governor Rodrigo Medina, along with a message allegedly from the Gulf gang.

Parts of Mexico, particularly near the US border, have seen a rise in suspected drug violence, with more than 37,000 deaths blamed on drug gangs since the launch of a military crackdown on organized crime at the end of 2006.

Rick

[kphoger]

Off-topic, along with the rest of y'all:
The U.S. citizen death toll in 2010 (the most violent year) was 111.  The number of U.S. visitors to México that year was about 7 million, and the number of U.S. residents in México that year was about 1 million.  You do the math.  I did.  If I'm to assume that the average resident stays in México an average of ten months, and the average visitor stays 12 hours (both of these numbers are completely made up, of course), then the average length of stay altogether is about 0.73 years.  111 deaths, therefore, adjusted for lenght of stay, comes to a homicide rate of 1.9 per 100,000 population.  This is lower than both the Mexican population at large and all but nine U.S. states (2009 figures).  Chihuahua is the #1 state for drug-related homicides, most notably because of Cd Juárez, but also for Chihuahua City et al.  I haven't looked at the municipal-level monthly figures for the state myself, but I have a copy of them (12/2006 through 12/2010) on my computer at home.  Typically the numbers show most of a given state being peaceful with contrastingly high rates in drug trafficking cities; I suspect that towns like Creel and the others mentioned to show a spike for one month and then drop back down to normal.  We read the news report, and assume the violence continues on afterwards indefinitely, when in reality it's usually contained fairly well (well-known cities excluded, of course).  For comparison, 80% of all drug-related murders in the state of Coahuila in 2010 occurred in the municipal district of Torreón; all other districts had lower rates than the homicide rate of Tulsa, Oklahoma.  I could go on, but I might bore you with more statistics.  All that is to say that people usually have a mistaken view of the safety situation in México

ANYWAY.  Back on topic:
I have had limited exposure to Chihuahua's state highway system.  I was in the state over New Year's 2001/2002, and traversed the state from Juárez to Batopilas by bus (via Chihuahua City and Creel).  I remember the state highway from Creel to the Batopilas junction being in very good condition, with near-new blacktop and bright stripes.  The road the rest of the way to Batopilas is a state road, and is famous for being challenging — gravel switchbacks almost nonstop down into the canyon.  Considering the topography, it too was very well maintained.  Looking at Google Maps and its satellite and street views, it also appears they've done some extensive toll-road construction west of Cuauhtémoc since I was there, so it doesn't surprise they're also improving access to the falls.  I note also the existence of state 2-lane toll roads between Chihuahua City and Nuevo Casa Grandes.  If only they could figure out a stoplight-free bypass around the capital city!

[agentsteel53]

The U.S. citizen death toll in 2010 (the most violent year) was 111.  The number of U.S. visitors to México that year was about 7 million, and the number of U.S. residents in México that year was about 1 million.  You do the math.  I did.  If I'm to assume that the average resident stays in México an average of ten months, and the average visitor stays 12 hours (both of these numbers are completely made up, of course), then the average length of stay altogether is about 0.73 years.  111 deaths, therefore, adjusted for lenght of stay, comes to a homicide rate of 1.9 per 100,000 population.

this math is not correct.  1 million residents staying 10 months each is 10 million man-months.  7 million visitors staying 12 hours each is 0.12 million man-months.  a total of 10.12 million man-months divided by 8 million visitors is 1.265 months average stay, which is about 0.1 years, not 0.73

the homicide rate therefore is about 13.87 per 100,000 population, which is much higher than the US rate of 5... but is still significantly less than that of New Orleans, which is 50.

and, in general, your overall point stands.  stay out of the trouble zones, don't be an idiot, and enjoy your visit!

[kphoger]

OK, my head is swimming with math right now.  I'm redoing the math myself and coming up with a different answer than I did originally.  I'll just accept it for now that you're right. 

[J N Winkler]

Using your assumptions (10 months/year per US resident in Mexico, 12 hours/year per US visitor to Mexico), we have (10/12 * 1,000,000) + (1/730 * 7,000,000) = 833,333 + 9,589 = 842922 person-years of exposure distributed among a population of 8 million.  This means that the homicide rate has to be multiplied by (8,000,000/842,922) = 9.49 to obtain an estimate that is comparable to American homicide rates, which generally assume one person-year of exposure per person.  So the adjusted homicide rate is (9.49*111)/8,000,000 = 0.00013167375 = 13.1 deaths per 100,000.

Compared to the homicide rate (murder + nonnegligent homicide) for major US cities, this actually isn't too bad--it is roughly comparable to Memphis, for example, and about a third of the rate for New Orleans:

http://en.wikipedia.org/wiki/United_States_cities_by_crime_rate

It is, however, three times the murder rate for Wichita.  It is also slightly higher than the overall rate for Mexico in 2009:  9,600 homicides for a population of 100,000,000, or about 9.6 homicides per 100,000.  (For 2010 I have found data on drug-related homicides only--corresponding to 15.3 deaths per 100,000.)

The possibility of being killed is only one aspect of the ongoing concern about violence in Mexico, however.  There have been reports of the drug cartels running roadblocks on the highways; of kidnappings for ransom; and of mass graves, many of which are thought to have been filled with people dragged off intercity motorcoaches and then executed.  On foot in an urban setting I don't generally worry too much because I figure I can blend in without too much trouble, but as a roadtripper I find it particularly unsettling not to be able to count on the freedom of the road.

http://en.wikipedia.org/wiki/Jaime_Zapata_%28U.S._agent%29

http://en.wikipedia.org/wiki/2011_Tamaulipas_massacre

The mass graves also have an effect on the crime statistics since the people in them typically can't be counted as homicides until their remains are found.  Several mass graves in Durango have yielded 308 bodies, while another was found in Nuevo Leon with 70 bodies, a third in Coahuila with 38, and a fourth in Tamaulipas with 193 (or 500 if you believe reports which the Tamaulipas state government is said to have suppressed).  Mexico has a population of about 100 million, so it takes just 1,000 people found in mass graves to move the homicide rate one point when it is expressed as deaths per hundred thousand.

I appreciate that the violence tends to be localized but even when the obvious suspects like Juárez are left out, it is hard to know which cities and highway corridors to avoid.

Edit:  Jake posted to make the same point about the need to adjust for exposure as I was composing this post.  I second his point about avoiding known trouble-spots (the border is a good starting point), but at this point Baja looks like the only reliably trouble-free part of Mexico.  It used to be that you just had to avoid Sinaloa (Mex. 15 corridor), but traditionally reliable places like Chihuahua (outside Juárez), Nuevo Leon and Tamaulipas are no longer so.

[kphoger]

All right, I finally got the math straightened out and....what do you know....came up with the same answer as you did.

The biggest hole in the data, as you pointed out, is the lack of representation of missing persons which cannot legally be counted as fatalities.  However, it is reasonable to assume that these gaps in the data would be fairly uniform geographically and, if they are not uniform, would be more sizeable in those areas already shown to be violent.  Therefore the comparison of place to place remains valid.

As far as total homicide rate goes, I haven't found that figure either; however, one Harvard study cited that 45% of the murders since Decmeber of 2006 were drug-related; since the drug crime rate has increased substantially, I find it safe to assume that more than half are now drug-related; this, then, means that in areas not prone to drug violence, the remaining homicide rates are very low.

There have been reports of people being snatched off the road by the cartels.  These are few and far between, most occur at night, most are motorists travelling alone on deserted highways, etc.  Right back to common sense.  I actually find myself being passed by much more expensive cars than my own on Mexican highways, so I don't find blending in to be too much of a problem.

And, as far as avoiding minor cities or not avoiding them, I could investigate these on a case-by-case basis for you if you'd like.  I have a spreadsheet that breaks down the deaths by month and by municipal district.  Pretty easy to do a cross-section.

[realjd]

Edit:  Jake posted to make the same point about the need to adjust for exposure as I was composing this post.  I second his point about avoiding known trouble-spots (the border is a good starting point), but at this point Baja looks like the only reliably trouble-free part of Mexico.  It used to be that you just had to avoid Sinaloa (Mex. 15 corridor), but traditionally reliable places like Chihuahua (outside Juárez), Nuevo Leon and Tamaulipas are no longer so.

There are safe areas other than Baja. The Yucutan is still pretty much trouble free for tourists (especially Cozumel), and Mexico City is supposed to be a great place to visit. Just avoid the drug trade and you should be OK.

Of the 111 American citizens killed, I doubt very many (if any) were tourists.

[kphoger]

Of the 111 American citizens killed, I doubt very many (if any) were tourists.

In fact, most of the victims were themselves involved in illicit operations, such as drug or human trafficking.  A third of them were in Juárez and Tijuana alone.

[kphoger]

Here is the municipal-level (read: county-level) breakdown for the state of Chihuahua.  Note how far and away Juárez leads the state, and how much of the state is relatively violence-free.

[J N Winkler]

Out of interest, what is your source for the crime data?  I presume it can be combined with population data (INEGI?) to derive per-capita values.

[kphoger]

The Mexican federal government released its data back in January of this year.  I was able to download the Excel file directly from their website.  I believe it's a conglomeration of multiple government sources.  There had been a call for the government to release the data for a few years, but they were very slow-moving, not because they didn't want to but because they wanted to make sure it was accurate.  Too liberal, and they'd be called fear-mongering; too conservative, and they'd be accused to sweeping it under the rug.

And, yes, I've done the per-capita thing for other regions.  The problem is that the per-capita rate can be very misleading in rural areas.  Have one shootout in a years' time, and the per-capita rate goes through the roof, whereas in a more urban district the number doesn't fluctuate as much.  That's where month-by-month evaluation is useful.