During his short appearance in the White House press briefing room on Thursday, President Trump dropped a number that Homeland Security Secretary Kirstjen Nielsen had mentioned during a Cabinet meeting the day before.

Introducing Brandon Judd, president of the National Border Patrol Council, the president said: “This group [Border Patrol] has apprehended, last year, 17,000 criminals trying to get across the border — 17,000. And that’s one category, there are plenty of others.”

This number is important because of how Trump was deploying it: as evidence of the need for a wall on the border with Mexico. The almost two-week-old government shutdown, affecting a number of departments including Homeland Security, was initiated when Trump refused any funding deal that didn’t include money for a wall. To bring Democrats to his point of view, he has repeatedly argued for the necessity of the wall to stem an alleged influx of drugs and criminals.

On Friday morning, Trump sent a letter to Congress including a presentation from Nielsen that sought to make that argument using a more fleshed-out set of data. I reached out to DHS on Thursday evening to try to evaluate the boundaries of that 17,000 number and ended up getting a better sense of how effective the numbers in Nielsen’s presentation were at making Trump’s case.

Let’s look at three particular pages of that presentation.

Criminals

Here are the figures, including that 17,000 number, meant to reinforce how dangerous it is not to have a border wall.

Or, shown as a bar chart:

Let’s start by looking at the number for gang members. This has been a frequent discussion point for Trump, who has warned about the horrors of the spread of MS-13 across the country, implying frequently that members of that gang enter the United States by crossing the southern border illegally.

You’ll note, though, that the graphic above isn’t specific to the border. It mentions gang members (“including MS-13”) apprehended at the border — and removed from the country by Immigration and Customs Enforcement. According to a Homeland Security spokesman, about 5,000 of those 6,000 gang member arrests/apprehensions happened somewhere besides the southern border. Only 1,019 were apprehended at the southern border, about a fifth of them at designated ports of entry.

In other words, the figure for gang members attempting to enter the United States by crossing the border with Mexico illegally isn’t 6,000, it’s about 800.

The other figures are similarly inflated. It’s probable that most of those 17,000 criminals (a figure that hasn’t been finalized), who were apprehended at the border with Mexico, were also caught when entering at designated border entry points, I was told, though a specific breakdown wasn’t available.

And as for that terrorists number? We’ve been over this before: It includes far more ports of entry than those at the southern border such as airports. It also includes people prevented from flying to the United States because of suspicion of ties to terrorist organizations. Exact details of these numbers are classified.

As we’ve noted, the State Department last year offered an estimate of how many terrorists had entered the United States by crossing the border with Mexico: zero.

Update: On Monday, NBC News reported a more concrete number. Six non-Americans on the terror watch list were stopped on the border with Mexico in the first six months of fiscal year 2018. By contrast, 41 non-Americans were stopped on the border with Canada.

Surge in migrants

To argue that a surge in arrivals at the border made the construction of a wall more pressing, the DHS presentation includes this bit of data.

We can see that surge in family units in particular, in monthly data on apprehensions provided by Customs and Border Patrol.

But notice that the surge follows a sharp drop-off in early 2017. In fact, the number of apprehensions at the end of 2018 was about where it was in mid-2016. Normally, apprehensions surge in the spring, not at the end of the year, but the raw counts aren’t themselves unusual. What was unusual was that plunge in 2017.

That plunge, of course, is why there was such a big percentage-point increase in fiscal year 2018 vs. 2017. (Federal fiscal years run from October to September.) Our figures, from data on the DHS website, are slightly different, but the point remains: It was 2017 that was the unusual year more than 2018.

That distinction is important for another reason, too.

Over the course of 2017, as those apprehension numbers remained low, Trump celebrated that decline as a demonstration of how effective his administration had been. Recently, the administration’s tune has been different, framing the apprehension number as an indicator of how effective the administration has been at catching illegal border-crossers.

Drugs

In that light, consider this page from the DHS presentation.

We’ve noted that most drug apprehensions occur at existing ports of entry, a fact attested to in testimony in 2017 by none other than Nielsen’s predecessor and Trump’s former chief of staff, John F. Kelly.

That the number of drugs being seized could mean that there’s a new push to get drugs across the border — or it could mean that the administration has been more effective at catching attempted smugglers. This is another one of those metrics that can be presented either as evidence for or against the effectiveness of the existing system.

A review of fiscal year 2017 produced by the department indicated that 2.14 million pounds of narcotics were seized that year. The presentation released on Friday has the 2018 figure at 1.7 million pounds.

What’s not included in that presentation is another figure that the department has touted and which has been used by Trump’s allies to make the case for the wall.

Those 900 children “saved from human trafficking” comes from a number released to a conservative news outlet last year.

We evaluated it at the time. The 900-odd people referred to above includes anyone who came to the border claiming to be part of a family with children but who weren’t. That figure, though, includes disqualifiers such as the children being at least 18 and therefore not children.

There were only 170 people who were determined to be part of family units that were separated because they weren’t related. Those people, though, weren’t necessarily human traffickers; they could also have been family friends accompanying younger people to the border. Not family — but not necessarily anything nefarious.

To the department’s credit, that figure is not included in the day’s presentation. Perhaps because it, too, had declined over the course of the year.