Last year around this time, the tropics provided some unintentional basis for what would become hysteria. Hurricane Harvey by soaking a major metropolitan area with a biblical amount of rainfall delighted Keynesians everywhere. So much destruction, so much economic growth potential on the rebuild. Then Irma flirted with the Gulf Coast spine of Florida for good measure.

For several months, if you didn’t watch the Weather Channel you might have thought the economy was accelerating. Combined with other factors, the narrative of globally synchronized growth chief among them, inflation hysteria reached full volume.

Economists are hoping for something similar out of Florence. Like the middle of 2017, the middle of 2018 isn’t quite shaping up as expected. Sure, Q2 GDP was revised up to 4.1% but even the most hardcore of inflationists know that the trick is not to occasionally touch 4% growth but do so quarter after quarter after quarter. Those two near 5% in the middle of 2014 proved, in the end, nothing other than a tortuous tease.

One of the more consistent economic indications has been Markit’s Composite PMI. Part services, part manufacturing, the index is supposed to capture near realtime conditions for a broad enough cross-section. Like most PMI’s, it’s not quite so simple. What matters most, and what this version has been best at, is when it starts moving.

That’s why some renewed appreciation for storm-broken windows. If IHS Markit’s PMI is right, the US economy is slowing. The flash estimate for September, released today, was 53.4 down from 54.7 (final) in August. That’s a 17-month low.

In an attempted response, to begin the fourth paragraph of the press release Florence pops up. “Anecdotal evidence that some of the slowdown in overall output growth reflected company shutdowns on the east coast ahead of hurricane Florence.” Maybe, but early indications are that Florence was no Harvey, nor Irma for that matter, and the composite PMI barely noticed the downside when the latter struck in August 2017.

Besides, this slowdown began months ago. The peak for the index was May 2018, consistent with a lot of other data. Over the four months since, the composite PMI has declined in each one. Collectively, it’s the greatest setback since early 2016.

That’s very likely why skipping past Florence Markit had to note the following:

Future expectations meanwhile fell to the lowest so far in 2018, and the second-lowest in over two years, as optimism deteriorated in both the manufacturing and service sectors.

Over in Europe, the performance isn’t any better just different timing. These eurodollar-based downturns never strike all at once, nor with the same intensity in every place. Even 2008 was a staggered blow only eventually hammering the whole world. Some places were hit long before others; some far worse than others.

The current slowdown, which this increasingly appears to be, registered in Europe almost immediately. It wasn’t trade wars nor imminent nuclear war in Korea (or whatever else has been offered as an excuse along the way). Europe’s setback began in 2017 with its banks’ unwise reentry into the eurodollar business. While Harvey was ravaging Texas, European financial institutions were about to be reminded why they stopped being the center of the eurodollar world in the first place.

The initial decline for the IHS Markit European Composite PMI was months ahead of its American cousin. Beginning in February, the index dropped quickly and hasn’t yet rebounded. The longer it doesn’t, the more likely it won’t. The flash estimate for September was 54.2, indicating that there is still more of a downside tendency than not.

Macro trends tend to be snowballs, full of momentum either way. If it starts to slow and doesn’t quickly recover, the economy adjusts by doing more negative things that make it slow even more. And so on.

None of these estimates suggest the global economy is in imminent danger of falling off a cliff. Rather, they indicate more of the same shift we’ve seen three times already over the last eleven years. Or, as Reuters put it in June:

A rare synchronized expansion among major world economies that was cheered on by markets and policymakers may be already unraveling, with only the near-term prospects for the United States looking significantly better.

That was June. Now in September, there may be less decoupling and more synchronization than people in the US might like. This isn’t unexpected, as I wrote elsewhere today:

From 2003 to 2009, it went: globally synchronized growth, decoupling, globally synchronized downturn. From 2010 to 2012, it went: globally synchronized growth, decoupling, globally synchronized downturn. From 2013 to 2016, it went: strong global growth (not synchronized), decoupling, synchronized downturn.

Last year to this year, it has gone: globally synchronized growth, decoupling. What comes next?

Is this time really different? Jay Powell sure thinks it is. Janet Yellen thought so, too, in 2014 and 2015. Markit’s PMI would show her otherwise.