Steve Hanke says crude oil tanker crossings through Bab el-Mandeb averaged under three a day in August and early September 2026, down from nearly seven in mid-July, arguing the Houthis control the strait

On 6 September 2026, Steve H. Hanke, Professor of Applied Economics at Johns Hopkins University, posted a chart on X with a compact claim about the Bab el-Mandeb Strait. In his words: “Tanker trackers show the weekly average of crude oil tanker crossings through the Bab el-Mandeb Strait has remained below 3 per day throughout August and now September, down from nearly 7 per day in mid-July.”

Then came the second half of the post, in capitals: “WHEN IT COMES TO THE BAB EL-MANDEB STRAIT, THE HOUTHIS RUN THE SHOW.”

Those are two different kinds of statement sitting in the same box. One is a measured series with a stated unit, a stated vessel class, and a stated comparison point. The other is a conclusion about power. Hanke is entitled to both, and readers are entitled to keep them apart.

What the numbers in the tweet actually say

Start with the metric as Hanke defines it, because the definition does most of the work.

The unit is a weekly average of daily crossings, not a raw daily count. That matters. A weekly average smooths out the days when nothing moves and the days when a cluster of vessels transits together. It is a stabilizing choice, and it is the right one if the goal is to show a trend rather than a spike. It also means no single day in the series should be read off the headline number.

The vessel class is crude oil tankers. Not product tankers, not LNG carriers, not container ships, not bulk carriers, not naval traffic. Bab el-Mandeb, the strait linking the Red Sea to the Gulf of Aden between Yemen and the Horn of Africa, carries a mixed flow of commercial shipping. Hanke’s chart is about one slice of it. A reader who converts “crude oil tanker crossings” into “shipping through the Red Sea” has widened the claim well past what was published.

The comparison is August through early September 2026, sitting below 3 per day, against mid-July at nearly 7 per day. Hanke does not name a peak date beyond “mid-July,” and this piece will not invent one. The shape of the claim is a level shift that has held for several weeks, not a one-week dip.

The sourcing is the part most likely to get dropped in retelling. Hanke writes “tanker trackers show.” He is citing a tracking source, and presenting figures drawn from it. That is a normal and honest way to publish a chart on social media. It is not the same as an official transit count from a maritime authority, and it should not be reported as one. Different tanker-tracking methodologies define a “crossing” differently, treat partial transits and course reversals differently, and classify vessel types differently. None of that makes Hanke’s series wrong. It means the series belongs to the methodology it came from, and readers should carry that label with the number rather than stripping it off.

Separating the count from the control claim

The capitalized line is where the post changes register.

“THE HOUTHIS RUN THE SHOW” is Hanke’s reading of what the traffic decline implies. It is an argument, offered by an economist who is well known for making blunt arguments in short form. It is not a verified finding about command and control over a waterway, and nothing in a chart of crude tanker crossings can establish one.

Here is why the gap is real. A count of transits records an outcome. It records how many crude tankers went through. It does not record why any individual operator chose not to. Routing decisions in a contested corridor are made by shipowners, charterers, insurers, flag states, and corporate risk committees, and those decisions respond to a stack of inputs that includes threat perception, insurance terms, charterer preferences, crew policy, contract structure, freight economics on the alternative route around southern Africa, and what competitors are doing. A single number at the end of that stack cannot tell you which input dominated.

So the honest phrasing is that Hanke has published a decline and offered an explanation for it. The decline is the data. The explanation is the take. Reporting that fuses them into one sentence is reporting his conclusion as though the chart proved it, and the chart is not built to prove it.

This is not a claim that Hanke is wrong. It is a claim about what kind of evidence a transit count is. Attribution is the whole discipline here: his figures, from his cited trackers, and his interpretation, in his own capitals.

What a crossings decline can and cannot tell you

A sustained drop in crude tanker transits through a chokepoint is genuinely informative. Read carefully, it tells you something specific.

What it can support: that fewer crude tankers, as counted by the trackers Hanke cites, chose this route during August and early September than during mid-July. That the change persisted across weeks rather than appearing as a single anomalous reading. That whatever is shaping routing behavior in this corridor was still shaping it at the time of the post. For anyone tracking freight patterns, that is a real signal, and it is the reason charts like this circulate.

What it cannot support, on its own: a specific attribution of cause. It cannot quantify who is deterring whom. It cannot establish an insurance story, because the post reports no premium data. It cannot establish an attack tempo, because the post reports no incident counts. It cannot describe naval posture or escort arrangements, because the post says nothing about them. And it cannot settle the question of who “controls” a strait, which is a legal, military, and political question that a transit series is not equipped to answer.

There is also a quieter limitation. Vessels that reroute do not disappear from the world. Crude that avoids Bab el-Mandeb still moves, on longer voyages, at different costs, on different schedules. A falling transit count at one chokepoint is a statement about that chokepoint, not about global crude flow. Nothing in Hanke’s post claims otherwise, and readers should not add the claim for him.

The Direct Message is straightforward. Hanke is publishing a sharp decline in his cited weekly average of crude tanker crossings through Bab el-Mandeb, from nearly 7 per day in mid-July to below 3 per day through August and into September, and he is attaching a political conclusion about who runs the strait. The first is a figure with a stated source. The second is his opinion. Keep them in separate boxes, and the post is useful. Merge them, and you have reported an argument as a measurement.