5 reasons travel security data is so bad

At Anjin Secure Car we use algorithms to determine the most appropriate supplier and security configuration for a given trip.  The algorithm takes human expertise and instantly determines whether a trip requires, for example, an armored vehicle or if a soft skin vehicle will suffice.  One of the main variables that our algorithm uses to determine the risk of a trip is the incidence of crime at the destination, pickup location, and along the route.  For the algorithm to be effective, we need to have a very clear idea of the level of crime down to the neighborhood level. 

Ideally, our algorithm would be fed in the most objective manner possible: carefully cross-checked and verified crime data.  As this data is very hard to come by, we are forced to combine more subjective methods to inform our risk ratings. 

Anyone that has tried to track crime data outside of the United States and Western Europe knows that this isn’t easy.  Ironically, we have found that the safer and more affluent the neighborhood, the easier it is to get reliable data.  This can lead one to the incorrect conclusion that the safest neighborhood in the city is the most dangerous.

Those who try to get crime data for their travel security program have probably faced similar frustration.  The U.S. State Department’s travel warnings will broadly cover entire regions and states within a country which is akin to saying the entire state of Illinois is dangerous because of one neighborhood in Chicago.   It is easy to lose credibility with employees when you paint an entire city as high-risk just because the data isn’t specific enough to account for massive variations in neighborhoods. 

There are several factors that make it extremely difficult to get good crime data.  We deal with this every day and have listed out the factors that we have discovered in our hunt for the most accurate perspective.

  • 1. Good crime data doesn’t exist.  Simply put, the data doesn’t exist because no one is competently collecting it.  There are huge problems with crime data in the United States, as The Atlantic pointed out in this article about crime and Uber, but we are world’s ahead of places like Latin America or Africa.  Even many small police departments in the United States have crime analysts on staff that index and evaluate crime data looking for both actionable information for their officers and trends to determine longer term policy.  Not so much overseas.
  • Tracking data requires headcount which is easier to come by in the post 9/11 United States.  It also requires communication across various law enforcement agencies which is even more difficult in some countries where the military also partakes in fighting crime. 
  • The other major practical impediment to good crime data is that some law enforcement agencies are fighting wars, not crime.  The police in northern border cities of Mexico or the cops in Managua are not focusing on car-jackings and sexual assaults. These are serious crimes but require investigations and time – something that they don’t have. 
  • 2. Powerful people don’t want you to know what is really going on.  Why would the local government in Rio de Janeiro or Cabo San Lucas want transparency in crime? 
  • We have seen the lengths that governments have gone to eradicate “undesirables” from the view of tourists with the movement of favelas in Brazil during the Olympics.  Local governments undertook massive efforts that required spending hundreds of millions of dollars and drawing the ire of human rights watchdogs.  This public relations calculus is depressing: it is better to be accused of violating human rights than accused of having a crime problem that impacts tourists.
  • 3. The press has been killed or has been forced to leave.  You must be incredibly brave to report crime in many countries and face countless death threats and watch your colleagues get brutally killed.  Even in major business centers like Monterrey, Mexico, reporters who report the “wrong” things can end up dead.  And while most of can understanding while reporting from the front lines of Syria is dangerous, it might shock many that some of these cities are home to hundreds, if not thousands, of factories and offices run by U.S. companies. 
  • To get perspective on what is happening in many dangerous locations, one must scour blogs or forums.  Sites like BorderlandBeat.com can be very graphic and require constant reading to keep up with the macabre soap opera that is cartel politics.
  • 4. People don’t report crimes.  Ironically, if you plot reported crimes on a map for man of the World’s cities you will conclude that this “objectively” points to some of the safer neighborhoods being dangerous.  Why?  Affluent people, businesses, and hotels are more likely to report crimes in many cities than the middle and lower classes.  For this reason, the supposedly objective measures of data don’t work. 
  • If your car was broken into would you call the police?  For people in the United States the answer is “yes” but not because we expect that the police are going to launch a full-scale investigation but rather because you need the police report for your insurance claim.  Without the motivation for financial reimbursement why bother giving up what could be a couple of hours of your time to bother reporting the break-in?
  • Even for more serious crimes, the likelihood of an investigation coming to a meaningful conclusion is very low.  In many countries, the clearance rate for murder (clearance rate is the % reported crime that leads to an arrest) is less than 10%.  I have no data to support this take, but considering the cultural baggage that comes with defining sexual assault, it wouldn’t surprise me if the clearance rate is lower for rape and sexual assault. 
  • Take away the financial compensation, the expectation of an investigation, and now add in another factor: in many cases in high-crime cities and countries the police may be involved in committing the crime.  With all these factors there is no compelling reason for people, particularly poor people, to report crimes.  This phenomena becomes obvious when you look at crime plotted on maps and notice entire swathes of a city with zero incidents. 
  • 5. Crime mapping companies are numerous in the United States.  Aggregating data is not easy.  Police departments, journalists, federal agencies, and watchdog groups don’t all subscribe to the same methodology and software to collect and index crime data.  Fortunately, we have companies in the United States that take all this data and assemble it into a coherent and readable format.  They do this because they can sell this information. 
  • Policy makers, municipalities, and real estate developers among others keep this industry alive.  Not to say that this industry won’t spread to other countries, but it is a chicken and egg situation. While there are companies that are starting to aggregate data overseas (primarily risk management and security companies) it takes a tremendous amount of effort and manpower to collect this data. 
  • Until there is a massive market to buy this data and an abundance of data to collect, you won’t see the same offerings overseas.  As one member of a prominent think tank that studies violence told me, “collecting threat data takes a lot of effort and we can only do it effectively if we focus on one country or region at a time.”

With all these factors in place, it looks like it will be awhile before we have access to more accurate and affordable crime data in the developing world.  There are some bright spots out there.  A few startups are looking at crowd-sourcing this data with the idea that the masses can become a giant army of collectors much like Waze is for traffic.  Other startups have started scraping 9-1-1 calls to get a real-time perspective on crime. Whether this can work overseas remains to be seen given that the real crime fighters in many countries are the military and they often work with encrypted communications. 

One bright spot we have found is our suppliers.  Local security providers have incredible intelligence networks.  Most of these networks are informal and data is shared with a view to inform current operations. For these reasons, much of the data is not archived.  We are looking at methods to archive and collect this data, but it won’t be a panacea, rather just one more source to inform our perspective.