How Do People Get Here?

One of the most interesting pieces of studying migration to me is the actual movement data; cases where we’ve got tracking of how people actually move from place to place. Domestically, we’ve got pretty limited data on that. But internationally, thanks to data generated by the visa-controlled nature of global travel, we actually have quite a lot of data. For example, the United States produces monthly statistics of entries for many different categories of visa-holders, and produces good summary-statistics for all border-crossers at the annual level. Today, I want to briefly look at that data.

We can get extensive data on land border crossing from the Department of Transportation, and can separately get international air traffic. In order to have harmonized statistics, I will track entry and exit; so a person making a round-trip shows up twice. If they arrive by car and exit by plane, then they add 1 car border crossing and 1 plane border crossing. I don’t care about direction for now.

Aquatic borders are much trickier, and unfortunately I have little data on them. The Department of Transportation does not seem to have much to say on the question of maritime arrivals. These are likely to be small overall, but when you consider international cruises, private vessels headed to Caribbean destinations, and shorter-haul ferries connecting to Mexico or Canada, Great Lakes or St. Lawrence waterway connections, and the inclusion of Hawaii, Alaska, and Puerto Rico, well, suddenly you realize maritime arrivals could actually be appreciable.

I do have a few spotty pieces of data. For example, I have the number of cruises by departure port and destination area, with the number of passengers and nights, from 2004 to 2012. I can sort it to exclude cases of cruises that we probably don’t want to include: <5 day cruises not identifying any foreign landfall. We can also exclude cruises going from west-coast ports to Hawaii or Alaska, though I’m inclined to think we should include cruises from Puerto Rico to the east coast, as other Caribbean stops seem plausible. We can then use this limited dataset and a few stylized figures from the cruise line industry about capacity in non-covered years to guess-timate the cruise-line component of our data.

There is also a national database of ferry ports, but alas, it does not include an estimate of passengers arriving at each ferry. It is possible that car/pedestrian border statistics count car ferries and pedestrian ferries… but I do not know for sure. And I do not think anyone traveling by passenger vessel across open waters, like say a Cuban fleeing to Florida, would be included under any land border estimates. But estimating these general maritime arrivals is even harder. For simplicity, I’ll just guess there’s 2 million in/out via other maritime sources in every year.

So when we do all of this, what does the data show?

The graph below is total border crossings estimated for the US from 1995 to 2016, broken out by transport category.

As you can see, the prevalence of entry via airplane has been steadily rising, even as personal vehicle passengers fell sharply from 2000 to 2010 or 2011. Other categories, like cruise lines or pedestrians, have been somewhat more stable. Here’s each category, indexed to its 2000 value:

Substantial growth can be seen for airlines and cruise lines, as well as some recent growth in train passenger volumes. Since 2011, personal vehicle traffic has risen slightly or is about steady, but remains well below 2000 peaks.

The main forces driving the total trend, of course, are changes in cars and airplaines. So here’s the share of border crossings made up by each mode:

As you can see, the airlines have eaten the car companies’ lunch.

This trend is being driven by many factors. Much of the rise in aviation is driven by non-U.S.-border countries. The Department of Transportation provides some breakouts for specific countries, which I’ve summarized below:

The fastest growth has been in “other” countries without regard to region, meaning that, broadly speaking, the geographic network of our airline connections has expanded. My suspicion is that “other” here is particularly heavily composed of the Gulf States, former Soviet states, Singapore, and Malaysia.

But we’ve also seen particularly rapid growth in flights from Mexico and Canada. As the north American market for flights has become more competitive from 1990 to the present, more and more former drivers opted for flights within the region.

Additionally, rising wealth and population around the world has boosted demand for travel to the United States in many destinations that were not our big traditional partners. So, for example, China has risen from 700,000 crossings in 1990 to 8.4 million in 2016. That’s a big increase.

But why have auto crossings declined? That’s a good question. We can break out these crossings for Canada and Mexico separately.

Both have seen declines. Now, note, data from several large customs regions is not available in some early years; complete coverage does not begin until 1998, and the problem is particularly severe in southern-border states like California and New Mexico, so treat the pre-2000 southern border figures with some skepticism.

But still. While both show a secular decline, Canada shows an uptick during some recession years, while Mexico a decline. What’s driving this?

Well, the uptick for Canada is driven by increased border crossings around Seattle and North Dakota: boom regions! Canada’s long-run decline is heavest in the customs regions covering depopulating parts of Maine, Michigan, and Minnesota, with a particularly severe decline around Detroit and northern Maine. No surprise given those regions’ last 20 years of economic history.

For Mexico, the decline is most severe in Texas, while we’ve actually seen higher border crossings in New Mexico, the only state where that’s true since 2000. I’ll admit that I don’t really know what’s driving these trends.

But here’s a thought. In the graph below, I compare apprehensions of illegal attempted border crossings from Mexico to legal crossings by car, as well as to all legal entrance from Mexico by any means.

You notice anything? Cuz I notice something. What I notice is that these trends seem pretty similar.

Maybe that’s a coincidence. But maybe what we’re seeing is that volumes of illegal migration (proxied for by apprehensions) closely mirror the volume of legal border crossings. That is, when there are more reasons to cross the border legally, there are also more reasons to cross the border illegally. And since there have not been too many big changes to the legal rights of Mexicans to cross the U.S.-Mexico border across this time, this is about what we might expect.

I’m sure others will come up with clever explanations for this relationship. But for me, it’s enough for today simply to note its existence and provide the data.

Check out my Podcast about the history of American migration.

If you like this post and want to see more research like it, I’d love for you to share it on Twitter or Facebook. Or, just as valuable for me, you can click the recommend button at the bottom of the page. Thanks!

Follow me on Twitter to keep up with what I’m writing and reading. Follow my Medium Collection at In a State of Migration if you want updates when I write new posts. And if you’re writing about migration too, feel free to submit a post to the collection!

I’m a native of Wilmore, Kentucky, a graduate of Transylvania University, and also the George Washington University’s Elliott School. My real job is as an economist at USDA’s Foreign Agricultural Service, where I analyze and forecast cotton market conditions. I’m married to a kickass Kentucky woman named Ruth.

My posts are not endorsed by and do not in any way represent the opinions of the United States government or any branch, department, agency, or division of it. My writing represents exclusively my own opinions. I did not receive any financial support or remuneration from any party for this research.