Last week, FiveThirtyEight’s Brittany Lyte wrote that casualties caused by land mines fell 25 percent in 2013, to 3,308, the lowest level on record, supposedly because of increasing compliance with the Anti-Personnel Mine Ban Convention. But the decline may not exist – and if it does, that treaty, which President Obama is trying to get the U.S. to comply with, can’t have much to do with it.

Lyte’s data comes from the International Campaign to Ban Landmines’ annual Landmine Monitor, released that same day. So do all of her quotes. That’s not a good sign: Journalism involves more than some cool new numbers from an NGO and reformatting a chart. Reading sources, securing contrary opinions, and knowing the context matters, too.

The data the ICBL produces is the best there is. But it’s still not good, as the organization itself admits: “Data collection in various countries such as Afghanistan, Chad, the Democratic Republic of the Congo, India, Iraq, Myanmar, Pakistan, Somalia, Syria, and Yemen was believed to be incomplete.” Of the 3,308 casualties, an easy majority — 1,802 — occurred in these nations. This, as the ICBL acknowledges, “makes trends difficult to discern.”

So the data isn’t good enough to allow us to say much with certainty. Much worse, though, is that Lyte doesn’t recognize that many of those 3,308 casualties weren’t caused by anti-personnel land mines (APLs).

The ICBL’s figures include both casualties caused by APLs and victim-activated IEDs (which are covered by the land-mine convention) and casualties caused by anti-vehicle mines, cluster munitions, and “explosive remnants of war” (not covered by the convention).

In fact, according to the ICBL’s figures, only about half the casualties it reports were caused by APLs and victim-activated IEDs. By far the deadliest kind of explosive device were the “explosive remnants of war,” which are bombs and shells that did not detonate when used: in 2013, they caused 1,006 casualties. These casualties were not caused by APLs, and the ICBL doesn’t claim they were.

Yet all of these numbers were related by FiveThirtyEight as land-mine deaths.

It’s even harder to deal with trends in the data. The ICBL adjusted the casualty figures it reported before 2012 in light of new information, which makes historical comparisons difficult. But even in its most recent reports, trends are hard to spot. In 2012, the ICBL reported 468 victim-activated IED casualties. But in 2013, after adjusting its data again, it reported that 2012 had actually witnessed 1,169 such casualties, while 2013 supposedly saw only 671. APL casualties, by contrast, declined only slightly, from 937 in 2012 to 812 in 2013.

In short, the fall in casualties from 2012 to 2013 was driven almost entirely by the supposed decline in victim-activated IED casualties. There’s no comparable fall in APL casualties. And if history is any guide, the IED casualty figures for 2013 will be revised up next year. I don’t blame the ICBL for improving its data as better information becomes available. But its data doesn’t support Lyte’s reporting.

So, even leaving aside the poor quality of the data, we don’t know if casualties are falling. Since four of the eight nations with the most casualties in 2013 aren’t party to the convention, it doesn’t deserve credit for anything that happened in those nations. And if casualties are falling, what’s declined in the past year, mostly, is IED use. If you really believe that al-Qaeda is using fewer victim-activated IEDs because an inter-state treaty banned them, well, good luck to you.

But this story isn’t just about numbers. It’s also about context. The amount of unexploded ordnance in the world – and the number of IEDs used and APLs laid – is a lagged function of the number and viciousness of the world’s wars. The late 1980s and 1990s saw a lot of these wars, in Afghanistan, Colombia, Ethiopia/Eritrea, and the former Yugoslavia, among other places.

It’s not surprising that, as some of these wars cooled and unexploded was cleared, the number of casualties recorded by ICBL for its Landmine Report has declined. As war has come to Syria and Ukraine, and returned to Afghanistan, the next decade is likely to see more casualties. The anti-land-mine treaty is largely irrelevant to these trends.

That treaty, the Anti-Personnel Mine Ban Convention, isn’t even the only treaty relevant to the problem. Land mines are also covered by Amended Protocol II of the Convention on Certain Conventional Weapons, which the U.S. ratified in 1999, and unexploded ordinance is covered by Protocol V (the U.S. consented in 2009). To the extent that casualties have fallen, it’s likely that these protocols deserve at least some of the credit.

But the ICBL doesn’t like the conventional-weapons treaty. In fact, the reason the newer anti-land-mine treaty exists is because its promoters hated the conventional-weapons treaty’s process, which sought to regulate – not ban – APLs. The NGO-driven process that created the anti-land-mine treaty has more or less substituted moral fervor for serious diplomacy. By jumping outside the conventional-weapons process, the ICBL actually made it harder to negotiate new, consensus-based agreements inside the CCW. It is entirely possible that this fit of moral fervor will end more lives than it saves.

For all the attention paid to them, casualties caused by APLs are rare: according to the ICBL, only 812 injuries and deaths in the world in 2013. That doesn’t mean that we shouldn’t try to reduce them further, but land mines have been demonized out of proportion to the casualties they cause: The ICBL’s own data shows that unexploded causes more harm than APLs. Moreover, APLs, if properly employed – as U.S. land mines are — can actually save the lives of civilians and U.S. forces alike by deterring attacks.

The defining aspect of the 2000s isn’t declining land-mine casualties, and the Obama administration’s push to comply with the APL convention is as irrelevant to humanitarian aims as it is dangerous militarily. But the real irony is that while the ICBL pats itself on the back for its progress in defeating the land-mine scourge, and the Obama administration applauds in the background, the most significant new weapon of war in the past decade is the insurgent’s IED. In other words, Lyte’s wrong: The land mine, so far as international law sees it, is not going away.

 - Ted R. Bromund is a senior research fellow at the Heritage Foundation’s Thatcher Center for Freedom.