A few days ago, I asked someone in the United Arab Emirates why the Covid vaccines seemed to have failed so badly there. He had first emailed me in the winter, at a time the UAE and Israel were leading the world in first-dose vaccinations – and seeing a sharp rise in cases and deaths.

What was so striking about the increase was the fact that the Emirates and Israel were using very different vaccines. Israel offered its citizens only the Pfizer/BioNTech mRNA vaccine, which depended on novel technology and had produced apparently strong clinical trial results. The UAE had gone with a cheaper Chinese vaccine that used an older technology and hadn’t done quite as well in trials.

Yet both countries had had cases and deaths soar in the month after they began mass vaccination campaigns.

That pattern would repeat in almost every country. No matter which mix of vaccines were used, no matter the season or the size of the country, infections and deaths rose shortly after health authorities offered the first dose to the most vulnerable people.

That increase may happen because the vaccines temporarily grab the attention our immune systems and make us more vulnerable to Covid and other infections. Or it might have to do with the very process of vaccination, as people congregate in vaccination centers and nursing homes open their doors for their first time. Or it might be some other, unknown factor. We don’t know yet.

But the fact of the spike cannot be denied.

Then the UAE and Israel diverged. The Israeli spike topped out and reversed. By June Israel was reporting almost no deaths or new infections, compared to more than 8,000 infections a day in January. But in Emirates cases remained stubbornly high. After topping out at 4,000 a day in late January, they were stuck above 2,000 in June.

The results appeared to be a clear win for Israel and the mRNA vaccines. Sinopharm and other Chinese vaccines just weren’t powerful enough to provide widespread immunity.

I wondered if my source had another explanation. So I emailed him: “What on earth is happening in the UAE? The steady vaccinations seem to be doing no good at all. Is it just the Chinese vaccines?”

His response (which I have edited for length):

You're right that it's hard to see a major effect of vaccines in the UAE - both in the first and second waves there was a similar-shaped decline to a moderate baseline:

However, with the exception of the US, we actually see a very similar pattern in the other countries with the highest vaccination rates:

The UK:

France:

Germany:

So the UAE doesn't look too different from those other high-vaccine-rate countries.  All of these countries had just as quick a decline last year, when there was no vaccine, as they do now, and in some countries the decline was actually quicker last year.  The outlier is actually the United States, because your northern seasonal decline last year coincided with the start of your southern wave:

Because of this data artifact, I wouldn't be surprised if many Americans were overestimating the visible impact of the vaccine on declining death rates, since the southern first wave obviously can't happen again this year and southern spread since last year has been more uniform.

So the UAE is not too different from other high-vaccination countries, but it DOES seem to be stabilizing at a higher baseline than last year.  Why?  I think the reason mostly has to do with the UAE's sky-high testing rate, which has actually doubled since last fall:

To sum up, I don't think there's evidence just from the case and death rates that the Sinopharm vaccine is performing noticeably worse than the mRNAs -- mostly because the effect of the mRNAs themselves are so hard to disentangle from seasonality and herd immunity.

His analysis is fascinating.

When I released my Unreported Truths booklet about the vaccines in March, I spent considerable time on the question of whether the mRNA vaccines from Pfizer/BioNTech and Moderna would work in people over 80. Those people are at far higher risk from Covid than anyone else and were essentially excluded from vaccine clinical trials.

By June, though, the very sharp declines in cases and deaths from Israel, the UK, and US nursing homes seemed to have proved me wrong.

This analysis has made me reconsider.

To be clear: one can certainly argue with the positioning of the blue “vaccine” boxes — the rollouts have gone all spring throughout Europe, and given the time lag before vaccines kick in, we may only now seeing their peak effect.

But to look at these graphs is to understand that we still cannot disentangle the effects of seasonality, natural immunity, and the way the post-first-dose spike pulls cases forward from the vaccine effect. The evidence from 2020 makes clear that seasonality is by far the most powerful force governing the spread of Covid. Look at those late spring and summer declines.

Not lockdowns or masks or vaccines. Just simple weather patterns.

Further, in the United States and most European countries, many, many people have already been infected and recovered from Covid and thus have immunity that appears at least as good as vaccine-generated protection. The CDC estimates that about 35 percent of Americans have had Covid, more than 100 million people, and most were infected before vaccinations began in earnest.

Finally, we do not know how long vaccine-generated immunity lasts. We have some evidence that it declines more quickly than natural immunity, although no one has offered a definitive answer yet.

In fact, looking at the world the biggest outlier so far has been not been the UAE but Israel – where both cases and deaths have been close to zero this month.

But in the last several days, new positive tests in Israel have suddenly spiked, rising from 20 or so last week to more than 100 in the last couple of days.

So it turns out that six months in to what has become the largest medical experiment in human history, we still don’t know exactly how well - or for how long - vaccines prevent Covid.

It would be nice if we did.