Misleading News on Deportation “Record”

On Oct. 22, the Associated Press reported on a Migration Policy Institute estimate of deportations under the second Trump administration.

“The numbers fall short of previous highs in the early 2000s.”

The comparison ran everywhere the story did. The administration’s 660,000 deportations in fiscal 2026 sat “well below the 1.86 million deportations recorded in 2000.” Newsweek and NewsNation used the same figure. One of them called it removals.

It was not removals. The 1.86 million is a real number from Table 39 of DHS’s Yearbook of Immigration Statistics. But the table has two columns. In fiscal 2000:

• Formal “removals” came to 188,467.
• “Returns” came to 1,675,876.

Nine of every ten “deportations” that year were “returns,” not “removals.” If you understand the difference, you understand how misleading the recent coverage was.

The Difference

A formal removal carries a bar on reentry of five years to life, and coming back after one is a federal felony. A return carries neither.

A return in 2000 usually meant a Mexican national caught by the Border Patrol who signed a form and walked back across the line. DHS’s report on 2005 enforcement says more than 99 percent of voluntary departures that year “involved aliens who were apprehended by the Border Patrol and returned quickly.” The Congressional Research Service says agents “permitted most Mexicans apprehended at the border to voluntarily return to Mexico without any penalty.”

The Government Accountability Office later ranked voluntary return the “least effective and efficient” consequence the Border Patrol had. Ninety percent of the “record” deportations in 2000 were of this “least effective” variety.

But the misleading reporting misses another key point:

The “return” counts also measured events, not people. DHS said so in its 2008 enforcement report: the returns figure “includes recidivists and thus is a measure of events rather than unique individuals.” A man turned back on Monday and caught again on Thursday counted twice.

The count followed the border

Returns rose and fell with Border Patrol arrests, because they were the same cases. CRS called the two “closely tied.” The Border Patrol made 1,676,438 apprehensions in fiscal 2000. DHS recorded 1,675,876 returns. Ninety-eight percent of those arrested were Mexican nationals, and most were returned within hours.

So the 2000 record counted how many times agents caught someone at a border they could not hold.

Then the government changed what it did with the people it caught. Starting around 2005, CRS reports, CBP “limited voluntary returns in favor of formal removals.” Crossers were processed for removal, with the reentry bar and the felony exposure attached. Returns fell from more than a million a year to 178,972 by 2013. MPI’s own report credits the same shift for the drop. Fewer returns meant more removals, by design.

What the numbers say now

MPI’s estimate splits the same way DHS’s table does. Of about 980,000 deportations from January 2025 through July 2026, 580,000 were formal removals. That is 59 percent, against 10 percent in 2000. MPI puts the current pace at about 380,000 removals a year. The 2000 figure was 188,467.

Part of that doubling is the processing change. The man who got a return in 2000 gets expedited removal today. That change was deliberate. The government now applies the consequence Congress wrote into law instead of the one that cost nothing.

The other part is the border itself. The Border Patrol recorded 98,708 apprehensions on the Southwest border in fiscal 2026, the lowest in the agency’s history. In 2000 it recorded 1,643,679. In other words, there were far fewer returns in 2026 because the border was secure. A total built on returns cannot approach the 2000 figure when the border produces one-seventeenth of the cases. September was the 17th straight month with zero Border Patrol releases.

MPI’s report did make a fairer comparison. It set today’s 380,000 removals a year against the Obama years. Removals then averaged 392,000 in the first term and 373,000 in the second, so the current pace sits in between. Most Obama-era removals were border cases too. In 2013, ICE’s record year, 235,093 of its 368,644 removals were people caught at or near the border. The outlets took the 2000 line instead, and that line was mostly returns.

The data gap

DHS has not published its monthly enforcement tables since the November 2024 data. The only current estimate with a published method is MPI’s, built from records obtained through the Freedom of Information Act. Publishing the removal and return columns for fiscal 2026 would settle the question.

What the comparison gets backwards

A border that releases no one and produces few crossers will never post a 2000-style total. The figure that matters is removals, and they are running at twice the 2000 rate. Those removals and the secure border behind them rest on executive policy. The asylum rules, the detention policy, and the end of parole releases came by executive action, and a future administration could reverse them with the stroke of a pen. The border would look like 2021 again within months, and the next “record” would be another tally of a border that had failed. H.R. 9773, reported by the House Judiciary Committee, and S. 5371 in the Senate would write those rules into law. That would keep the border Americans have now, whoever sits in the White House next.