U.S. Customs and Border Protection reported a sharp decline in apprehensions along the southwest border last month. In March, 16,600 people were apprehended or determined “inadmissible,” compared to 23,570 people in February and 42,477 people in January. These are the lowest monthly numbers in the past seventeen years. The second lowest number dating back to 2000 was in December 2011 with 18,983 people.

Southwest Border Migration, via U.S. Customs and Border Protection

CBP credits the marked decline to the “implementation of executive orders to enforce immigration laws,” and the Department of Homeland Security Secretary John Kelly, called the dip, “no accident” in a written testimony. But according to Faye Hipsman, a policy analyst at the nonpartisan think tank the Migration Policy Institute, it’s not so much the implementation of new policy affecting immigration as it is President Trump’s rhetoric. It’s what some experts are referring to as the “Trump effect.” 

“There is a tremendous amount of fear of interior immigration enforcement in immigrant communities right now because of some of the policies laid out in the Trump executive orders and things he’s said,” Hipsman says. “A lot of the fear is a bit misguided because he hasn’t really ramped up operations significantly, which is conducting a huge increase in raids or checkpoints, but a lot of fear has been drummed up.”

Certainly, Trump has backed his promises in terms of immigration with presidential action. But Hipsman says that implementation of some parts of Trump’s executive orders, such as opening more detention spaces, will take time to produce real results. She also adds that there currently isn’t enough information on ICE arrests under Trump to compare them to those under previous administrations, but it’s the nature of the arrests that stand out.

Those arrests are what ICE refers to as “collateral apprehensions.” In February, more than half of the people arrested by ICE in some of the highly reported raids in Austin were non-criminals. Other arrests made during routine check-ins with ICE and even of “Dreamers” have made undocumented immigrants unsure of their security in the U.S., and that seems to be influencing how they interact with local law enforcement. News of an ICE arrest of an undocumented trans woman outside of family court in El Paso just after she obtained a protective order spread across the nation. The incident prompted concerns from both immigration and sexual violence advocates that her arrest would result in a “chilling effect” in which fewer undocumented victims of sexual or domestic violence would file police reports.

And last week, the Houston Police Department announced that the number of Hispanics reporting rapes from January to March of 2016 had dropped 42.8 percent compared to last year. “When you see this type of data, and what looks like the beginnings of people not reporting crime, we should all be concerned,” Houston Police Chief Art Acevedo said at a news conference. According to the Houston Chronicle, Acevedo didn’t directly attribute the decrease in reports to the actions or rhetoric of the Trump administration, but he did note that similar trends are happening in other parts of the country, such as Los Angeles.

The “Trump effect” could also be accountable for surges in immigration that took place before his inauguration. In October, the number of apprehensions and inadmissible people rose to 66,710—second highest to 68,804 in May 2014—before beginning to drop. According to the Waco Tribune-Herald, some experts think that increase took place because people moved up their immigration plans in order to make it to the U.S. before the election.

Hipsman says it’s hard to tell whether the number of apprehensions in March are accurate indicator of what’s to come in the future. Numbers of apprehensions tend to swell in early spring, and whether or not they do this year might be a more accurate indicator of the “Trump effect” on immigrants.

“We’ve seen a trend of three months, but by no means does it mean that this is a lasting trend or that it will continue once people’s fears and also the policy changes. Once all that gets ironed out and the dust settles, the numbers may increase again” Hipsman says. “It just depends on when the dust settles on these policies and what changes actually take place. People aren’t coming right now because they think that they may be detained for longer periods or that they may be prosecuted, but if the policies don’t actually change people will realize that and could start coming again.”

Kate Christensen Mills, a former deputy director for congressional relations for ICE, is also unsure about how implementation of Trump’s executive orders could shape up. She does, however, have ideas about what the decrease in border apprehensions could mean for ICE arrests. “If you’re going to have a decrease of people coming across the border, you’re going to have an increase in interior enforcement,” Mills said at the Border Security Expo. According to the Huffington Post, Mills predicts that the increase of interior arrests will be of people who have been in the U.S. for longer and who don’t have criminal records, because there are “only so many criminal aliens.”

ICE removals began dropping after 2012, going from 409,849 that year down to 240,255 in 2016. Over 90 percent of last year’s interior removals were of convicted criminals, something ICE concludes came from their continued focus on targeting “individuals who threaten public safety, national security, and border security.” That percentage has steadily increased from 2011, when it was just 67 percent. Whether those numbers will continue on their current trend is hard to say just months into Trump’s first term, but Hipsman also believes the percentage of criminals removed may have decreased.

“It’s not likely that the numbers of removals from the country has increased substantially since Trump took office,” Hipsman says. “What is likely changing is the composition of people who are removed. Under the Obama administration there were enforcement priorities which meant that people who were removed tended to be serious criminals, people with recent removal orders, recent arrivals and under this administration, that prioritization system is no longer in place. So what we’re likely to see is a higher share of removals that are of non-criminals, people that would not have been priorities under the previous administration.”

As for border crossing, Hipsman adds that there’s always the chance that even when Trump’s policies are solidly in place, they might not affect some immigrants at all. “A lot of people are fleeing very violent and dangerous situations,” Hipsman says. “And for some of those people no level of deterrents or detention is going to stop them.”