This post is adapted from the blog of data.world, a Priceonomics Data Studio customer. Does your company have interesting data? Become a Priceonomics customer.
***
Each year, the State Department issues dozens of advisories with the intent of keeping Americans safe as they travel abroad. What countries are targeted by these advisories, and what risks do Americans face by visiting them? Are State Department advisories effective in keeping American travelers safe?
We decided to investigate what are the most dangerous countries for American to visit as measured by State Department warnings and also by actual deaths. We used data from Priceonomics customer data.world, a platform that ties many different data sets together so it’s easy to analyze them (you can download their dataset here).
We found that Mexico, Mali, and Israel have been targeted by the most travel advisories in recent years, but that Americans are more likely to face life-threatening danger in Thailand, Pakistan, and Honduras. Indeed, warnings and deadly violence are correlated on the whole. And fortunately, some travelers – at least those headed to the Philippines or Egypt – seem to heed these advisories, as those countries see dropoffs in tourism following warnings.
***
We began by identifying the countries that are most often targeted by U.S. State Department travel advisories. The State Department has multiple mechanisms for advising American travelers, but we focused just on Travel Warnings, which are issued when lasting turmoil in a country poses such a danger that the State Department discourages any travel there at all.
We filtered out warnings that had been issued for natural disasters, then ranked countries based on the number of Travel Warnings issued against them in an 8-year period between 2009 and 2017. We display the top 25 below.
Source: data.world
Mexico tops the list with 28 warnings in an 8-year period. It’s worth noting that these warnings are regionally specific, targeting sites where crime syndicates are particularly active. Popular tourist destinations like Mexico City and the Yucatán peninsula (including Cancún) are generally regarded as safe.
Most other countries on this ranking are participants in ongoing international conflicts (e.g., Israel, Pakistan, Afghanistan), or are sites in which extremist groups regularly carry out terrorist attacks (e.g., Mali, Nigeria, Syria).
North Korea is an interesting exception, as the government itself presents a danger to American travelers. According to the State Department, foreigners are liable to be jailed for unspecified reasons, or for seemingly innocuous infractions like interacting with the locals or taking unauthorized photos.
How do State Department warnings square with the actual likelihood of crime abroad? Reliable, global data on crime is difficult to come by, but the State Department tracks the incidence and causes of American deaths abroad. We used that dataset to identify countries where Americans are most likely to experience life-threatening danger while traveling.
In the table below, we rank the foreign countries in which the most Americans were killed between 2009 and 2016. Before ranking, we filtered the data to include only homicides, executions, deaths in terrorist attacks, and drug-related deaths.
Source: data.world
In general, a violent death abroad is extremely unlikely. Between 2009 and 2013, 1,151 Americans – out of a population of 316 million – were killed abroad. For comparison, 15,809 homicides occurred in the U.S. in 2014 alone.
Of the 1,356 killings that occurred abroad, 1,193 (88%) happened in the 25 countries listed above. And just one country, Mexico, accounted for 50% of those deaths.
Of course, more Americans die in Mexico because vastly more Americans travel to Mexico than any other country. This holds true to a lesser degree for some of the other countries near the top of this ranking, including the Dominican Republic and Jamaica.
With that in mind, we adjusted our ranking to account for the volume of tourism from the U.S. We calculated the number of Americans murdered in a country per 100,000 American tourists, using travel numbers from a dataset gathered by the Bureau of Transportation Statistics. We excluded any country that received fewer than 100,000 American visitors between 2009 and 2016.
Source: data.world
Adjusting for travel volume shuffles our ranking, if not drastically changing it; 16 of the 25 countries listed here also appear on our ranking of countries by absolute number of American deaths.
Pakistan and Thailand jumped to the top of the list, each with a handful of deaths in a relatively small pool of tourists.
Surprisingly, there is only modest overlap between this ranking and the set of countries receiving the most travel warnings. Of the top 10 countries ranked here, only 4 — Pakistan, the Philippines, Nigeria, and Mexico — were among the top 25 countries targeted by travel warnings.
This led us to wonder about the connection between State Department warnings and American death abroad. Does the State Department issue more warnings for a country if Americans are more likely to be killed there? To find out, we correlated the number of Travel Warnings issued for each country with number of Americans (per 100,000 travelers) killed there.
Source: data.world
On the whole, there is a significant relationship between the number of American deaths abroad per capita and the number of travel warnings a country receives (r = 0.56, p = <.001).
But within this chart, we identified some interesting patterns. In some countries, the number of Travel Warnings a country receives does scale with the number of deaths. In others, no warnings are issued even while the risk of death is relatively high. In still others, many warnings are issued even though most Americans pass through the country intact.
In which countries are warnings well correlated with the risk of death? In which are they not? In the rankings below, we identify 5 countries that exemplify each pattern.
Source: data.world
A relatively high number of American travelers die in the countries in the left-most ranking above. Accordingly, these nations are often targeted by State Department warnings.
The center ranking features countries where warnings are “under-issued;” the risk of death is relatively high for Americans, but no warnings were issued in the 8-year period we examined. This ranking consists mainly of Central and South American countries where roughly 1 in every 100,000 U.S. travelers will be killed.
Finally, the countries in the right-most ranking are often targeted by warnings, but Americans have a low risk of facing life-threatening danger while visiting them. This may have to do with a regional pattern of unrest; in Turkey, for instance, tourists visiting the southeast may be subject to terrorist attacks spilling over from Syria, while travelers to Istanbul are comparatively safe.
Alternatively, few Americans may die in these countries because they are heeding the high number of Travel Warnings these nations receive. This raises the question of how travel advisories impact tourist behavior. Does traffic to a country drop after the State Department targets it with a Travel Warning?
To find out, we again put the Bureau of Transportation Statistics dataset to work, this time comparing travel numbers in the 6-month periods immediately preceding and immediately following the issuance of a Travel Warning. For this analysis, we only considered countries that had received at least 3 warnings.
Source: data.world
Egypt sees the largest drop-off in travel after a warning is issued with a 34% decrease in travel. Thailand travel appears to follow a similar trend; when warnings are issued, American travel to Thailand drops by 15%.
Travel declines modestly in Israel and Venezuela after the issuance of a warning, even though neither country lands a spot on our top 25 nations by American deaths per capita.
And strangely, travel to Ukraine, Nigeria, and Saudi Arabia rises by more than 10% after a warning is issued.
Overall, American don’t appear to be especially sensitive to State Department warnings. Of the 16 countries we featured here, double-digit travel change was seen for seen for only 5, and those changes could be driven by many other factors.
***
For simplicity, we restricted our analysis to just one advisory mechanism – the U.S. State Department Travel Warning – and one outcome measure – American deaths abroad. It’s possible different trends would have emerged if we had considered other data sources. But given these constraints, what did we see?
In absolute terms, more American tourists are killed in Mexico than in any other foreign country. This is partly owing to the strong flow of tourism between the U.S. and Mexico; when figures are adjusted to account for the volume of tourism, Pakistan rises to the top of the heap, with roughly 4 deaths per 100,000 travelers.
Despite this, Thailand does not rank among the top 25 countries for travel warnings. In general, warnings are not strongly correlated with American deaths abroad: countries like Israel, Turkey, and Saudi Arabia are relatively safe despite being subject to numerous warnings, and the converse is true in Belize, Guatemala, and Guyana.
Even warnings that are well correlated with violence are only valuable if travelers heed them, and tourism appears largely insensitive to travel advisories. In approximately half the countries we considered, tourism shifted by no more than 2% after issuance of a warning.
***
Note: If you’re a company that wants to work with Priceonomics to turn your data into great stories, learn more about the Priceonomics Data Studio.