Winds know no borders.

However tightly controlled a state may be on the ground, however tall its walls and vigilant its patrols, the open sky is remains free. It is this tension that makes balloons and North Korea so fascinating: how absolute can a state be when news of the outside world drifts on the air? It’s a question tackled, in a roundabout fashion, in a new report from the RAND Corporation, released earlier this year. In “Low-Cost Options for Airborne Delivery of Contraband into North Korea,” author Richard Mason looks at the aeronautics of cheap flying devices, wind patterns and cost analysis, and attempts to find the most effective way to get a message across the DMZ.

Balloon propaganda between the Koreas dates back to the Korean War, as parties to the conflict used balloons to deliver propaganda leaflets across the front lines in an attempt to sway hearts and minds. For decades after, leaflet delivery by balloon was primarily a function of government, with the Republic of Korea officially launched balloons across the DMZ into the early 2000s, before a formal halt in 2004. Following the halt, part of a mutual ceasefire in propaganda delivery that lasted from 2004-2010, non-governmental organizations instead used balloons to carry messages across the way.

How effective were those messages? The RAND report doesn’t delve much into the actual content of leaflets attached to balloons and carried by the wind, nor does it spend a lot of time addresses the effect of those messages in the hands of the people who received them. (Besides leaflets, payloads include everything from socks and marshmallow peeps to bibles and radios). Instead, Mason looks at the first part of the task: are the winds even carrying the balloons where they’re supposed to go?

If the wind cooperates, RAND simulations show that the balloons can make it as far as North Korea’s border with China, at least during the spring and summer months. Flight time is the biggest determining factor here, but it’s not the only factor. There’s also the matter of North Korean machine guns.

”Could North Korea Shoot Down the Balloons, If It Made a Concerted Effort to Do So? “ the report asks, and answers:

Our analysis suggests the answer is marginally yes. The Korean NGO balloons typically fly at an altitude of around 3,000 m, where they can be engaged by North Korean antiaircraft guns. The economic exchange is very roughly even; i.e., shooting down the balloons is neither drastically cheaper nor drastically more expensive than launching the balloons, so a “war of attrition” would be about equally expensive for both sides.

The study also explores alternative delivery methods, but all of these are rightfully seen in light of the cost constraints. Balloons are cheap, helium is cheap, and even with an expectation that most balloons will not reach their destination, Mason puts the cost per delivered balloon payload at around $1,500.

But what about drones? A hexacopter drone runs slightly more than that, and is in theory reusable, but comes with far greater risks. Drone flight times and ranges while carrying a full payload are minuscule, likely enough for a single sprint from one side of the DMZ to somewhere not too far on the other end, and the drone is an obvious piece of contraband that would need to be recharged before it could be returned. Mason notes that hexacopters are possibly less risky than human couriers on the ground, but they cannot go as far as balloons, and can be jammed out of the sky, so they should not be seen as a viable alternative for distributing material indiscriminately. Other options discussed include fixed-wing drones (more expensive, would need multiple trips to be worth it), balloon-released single-use gliders (cheaper than single-trip fixed wing, more expensive if multiple trips possible) and uncrewed dirigible drones, which are delightful but also more expensive still.

Ultimately, Mason suggests that pressure release valves on balloons are likely the most cost-effective way to ensure balloons stay airborne until they can travel far and deliver a payload, and the method most likely to avoid any North Korean anti-air defenses.

As for what happens when the payloads reach people inside North Korea? Mason’s study cites testimony multiple North Korea defectors gave in the 1990s. Some of the people quoted encouraged the continuation of balloon drops, and multiple others noted that while propaganda had to be handed over to government agents, in between retrieving it and turning it in, people could read the leaflets. What people do with the propaganda is not the core focus of the report, though it’s still a tad underwhelming to see so much depth on how to get the leaflets into North Korea and so little on the impact those materials have.

Information on what happens within North Korea is hard to obtain in the best of circumstances, so as proxy the report looks at official state reactions to the balloon campaigns. These include threats to shell balloon launch sites, as well as anti-aircraft fire against the balloons themselves. North Korea also sends leaflet-carrying balloons and drones of its own into the South, a sort of airborne propaganda war carried out entirely by the wind and robots.

Should balloon launchers switch to higher-altitude flights, beyond the reach of ground-based anti-aircraft weapons, the only thing left for Pyongyang to do to stop the balloons? Scramble super low-tech jet fighters.