The Trump administration’s new executive order on immigration and refugee issues — the so-called travel ban — was supposed become effective on Thursday, but it was blocked by a federal judge. Like the judges who blocked the first version of the order, federal judge Derrick K. Watson of the District of Hawaii said opponents of the order could likely prove it discriminated on the basis of religion, in violation of the First Amendment.
So the legal fight and policy arguments continue. Unfortunately, there’s been some crucial data missing from the debate about the effects of a refugee ban — namely, information about what happened during the short period that the first one was in place. Until now, that is: This article will fill in some of the gaps.
Where does one get such information? Refugee admission data is produced on a daily basis by the Refugee Processing Center, operated by the State Department. It lets us see how many refugees were admitted into the country each day, as well as their religion, age, sex, nationality, and other demographic markers.
The first executive order was implemented with little warning or guidance to officials on the ground (let alone to refugees and other visa holders) — and with no exemption for people who were in transit. Nor was it implemented as written: Under pressure, the White House relented and allowed previously chartered refugee flights to continue for several days. So there are caveats involved in extrapolating from what happened during the first order to what might happen under the second.
Undoubtedly, any new order will face resistance and on-the-ground logistical challenges, too. Nonetheless, the figures here offer a useful window into refugee admissions in general, and into the effect of the first ban — with all its attendant confusion.
The new executive order differs from the old one in several ways. The number of countries singled out for special scrutiny has dropped by one, with Iraq left off the new list. The first also order included an explicit preference for “religious minorities,” which has been abandoned. This, however, is a bit misleading: Preference for religious minorities and persecuted religions was US law before Trump’s executive order, during it, and remains so today. Refugee status can be granted for many reasons, and religious persecution is one of the most common, along with racial or ethnic persecution.
However, the religious preference was seen on both sides as signaling — by progressives, that the administration would discriminate against Muslims in refugee admission; by conservatives, that the administration had heard their concerns about the increasingly unsafe environment for Christians in the Middle East. Dropping that preference ends both signals. In any case, we’ll see below what impact the religious-minority preference may have had during the first order’s brief implementation.
Second, the old executive order indefinitely barred refugees from Syria (with the possibility of exemption for religious minorities, namely non-Muslims). This indefinite ban is missing from the new order.
Finally, the old order had no exemptions for refugees already in transit, little guidance on how to make exceptions, and little help for agencies looking for smooth administration. The new order remedies those omissions to some extent.
Besides these three changes, the two executive orders have similar approaches: All new refugee admissions are paused for 120 days subject to certain exceptions, and the fiscal-year cap is lowered from 110,000 to 50,000.
We can cut the data many ways. Here’s monthly refugee admissions through February for fiscal year 2017 — as well as the maximum, median, and minimum monthly admissions for each month going back to 2007:
Fiscal year 2017 began with the highest rate of admissions seen in a decade. But after the new administration took over, refugee admissions fell to historically more normal levels.
We can be a bit more precise. Let’s look at the average daily rate of refugee inflows for the periods January 1 to 26, January 27 to February 3, and February 4 to March 6. Why those dates? Well, I queried this data on March 6, the old executive order was active from January 27 to February 3 — and it seems reasonable to compare that period to the January pre-ban-period, and to the period after the ban was lifted. For context, I also show information for previous years, going back to 2015, using those same slices of each year.
(For years before 2017, I also include information for the balance of the year, beyond March 6.) As you can see, the period of the refugee ban had far lower rates of arrivals than in late 2016 or early 2017, but inflow rates were not far out of line with many recent periods. This is not what might have been expected, to say the least, from a blanket 120-day pause on admissions. Aside from the admission of the chartered flights mentioned earlier, it also appears that case-by-case waivers were granted extensively. The “pause” was at best a slowdown, where refugee vetting had an additional step in which waivers had to be formally granted.
Since the executive order was halted on February 3, refugee admissions have continued, in theory with no legal impediment. Still, they remain below late-2016 levels (although inflows remain comparable to other prior periods).
We can also break the data down further, looking at the share of refugees who were Muslims:
And from this data, we can see that the Muslim share of all refugees fell sharply during the ban, to 17.3 percent, roughly the share seen in 2011 or during the mid 2000s. This 17.3 percent represents a dramatic departure from the policies of President Barack Obama’s second term, but, again, not a huge departure from much of recent US history.
We can also look at the share of refugees who were Syrians:
The Syrian ban appears to have been enforced with effectively zero waivers. I could not identify any Syrian refugees admitted during the ban period. No Muslim Syrians, no Christian Syrians: no Syrians of any kind. The administration seems to have rigorously enforced this particular policy.
But wait: If the huge reduction in Syrians is also a huge reduction in Muslims, maybe there was little to no effect on admission of non-Syrian Muslims? We can explore whether that hypothesis holds true:
It does not. As it turns out, non-Syrian Muslims got hit too, with their share of admissions reaching the lowest level in a decade. It seems that the executive order, possibly due to its religious preference clauses, did result in lower relative admissions of all Muslims. Here’s the daily admissions of all Muslims, showing the steep fall-off in admissions:
Finally, complicating the picture somewhat, here’s daily admissions of all non-Muslims:
As you can see, the ban period does coincide with reductions in the share of Muslim and Syrian refugees, but also a cut, if less severe, for non-Muslim refugee admissions. Both Muslim and non-Muslim refugee admissions were slowed to some of their lowest rates in recent memory, while the total ban on Syrian refugees was implemented exactly as the administration intended — and as critics feared.
However, while Muslim and Syrian refugees have seen their admissions rise again since February 3, non-Muslim admission reductions have continued apace even after the ban was overturned. In reviewing the old order, I predicted that it would not achieve many conservatives’ hopes of “prioritizing” Christian refugees, but instead would just abandon suffering people of all faiths. That’s what ended up happening, in the chaotic period of the ban and even subsequently, with Syrians of all faiths hardest hit by far, followed by Muslims around the world.
What might happen next?
Using data from the Refugee Processing Center as the baseline, we can also try to see what would happen in the next few months if the new ban were implemented. (At this point, that would require a Trump victory in the courts.) If we take the administration at its word, there will be a 120-day pause, with just a few exceptions for transit or special cases.
I assume here that the pause will be implemented more thoroughly the second time around. I also assume the 50,000 refugee cap will be enforced. Here’s a reasonable forecast of what monthly admissions might look like:
This would, of course, be an incredibly low amount of refugee admissions, near the slowest pace since the current refugee program began in 1980. (Restrictions immediately after 9/11 did result in even lower rates.) If numerous waivers are granted during the 120-day period, as seems likely if refugee advocates resist the new order as much as the previous one, then admissions at the tail end of the year must fall from my forecast, in order to stay under the cap. Although the new order no longer includes an indefinite ban on Syrians, it is possible de facto targeting could continue, just as Muslim admissions generally were de facto reduced under the old order. Time will tell how this all shakes out.
In the meantime, however, severe declines in refugee admissions could begin as soon as the 120-day suspension begins and in-transit refugees finish arriving. Would-be refugees will continue to languish in places that range from uncomfortable to truly dangerous, without any regard to whether or not they pose any risk of terrorism. Vetting procedures may be improved, or they may not; no public word has been spoken yet about how vetting will be changed. And within a few weeks, we will begin to have data showing us how many “exceptions” to the pause are going to be made. I, for one, will be watching very closely.
Lyman Stone is a regional population economics researcher who blogs at In a State of Migration. He is also an agricultural economist at USDA. Twitter: @lymanstoneky.
The Big Idea is Vox’s home for smart, often scholarly excursions into the most important issues and ideas in politics, science, and culture — typically written by outside contributors. If you have an idea for a piece, pitch us at email@example.com.