By Jeremy Schwartz, Austin American-Statesman

AUSTIN, Texas — Jaime Zaldana was driving to work on Interstate 35 through the San Antonio suburb of Schertz on a winter morning in 2010 when his red pickup passed a U.S. Border Patrol unit on the side of the highway. He was 150 miles from the nearest border crossing, in Eagle Pass.

The agents would later write that Zaldana and two co-workers “appeared to be startled” and looked “straight ahead,” never acknowledging them — reason enough, they claimed, to pull the truck over. Zaldana was eventually deported as an undocumented immigrant, but not before the U.S. government paid $25,000 to settle his claim the agents stopped him only because of his ethnicity and race, according to public court documents.

Most Texans probably think of the Border Patrol as doing what the agency’s name suggests: interrupting illegal activity along the line separating the U.S. from Mexico. Yet over the last decade, agents have regularly made arrests deep inside Texas, according to an Austin American-Statesman investigation into the little-known realm of the Border Patrol’s interior enforcement operations.

Between 2005 and 2013, agents apprehended more than 40,000 subjects at the nine most inland Border Patrol stations representing locations as far as 350 miles from Mexico — an average of more than 10 per day, according to numbers obtained under the Freedom of Information Act.

In numerous cases, the arrests occurred more than 100 miles from the border, a blurry line of demarcation drawn a half-century ago. Some resulted from the same controversial roving highway patrols that nabbed Zaldana, in locations not typically associated with border enforcement: San Antonio, Odessa and San Angelo.

While the agency has insisted the 100-mile line doesn’t limit its activities, courts have ruled that agents need stronger justification to stop motorists the farther they are from the border. And critics say roving interior patrols inevitably lead to agents pulling over motorists based only on their racial appearance, resulting in unconstitutional stops of American citizens and legal residents — the Border Patrol says it does not keep track of the numbers of those stops.

Judges have voided Texas stops because of agents’ flimsy pretexts, though the Executive Office for Immigration Review, which adjudicates immigration cases, doesn’t track that number, either. Advocates say the number of decisions understates the reality because immigrants rarely challenge the stops.

“You ask yourself: What else do (Border Patrol) agents have to go on besides race?” said San Antonio immigration attorney David Armendariz, one of the few attorneys nationally to regularly pursue such cases. “What does an immigrant without papers look like? Because your average immigrant without papers looks like your average Hispanic.”

The Border Patrol will reveal few details about its work far from the border. Yet evidence suggests the government might have moved to limit its scope in Texas in recent years.

As recently as 2012, roving patrols made up the principal activity of agents in such places as San Angelo, 130 miles from border, according to court documents reviewed by the Statesman. That same year, however, the agency proposed shuttering six interior stations in Texas, including San Angelo’s, and shifting the agents to the Mexican border.

Members of the Texas congressional delegation have so far resisted such deactivations, which lawmakers have said would weaken law enforcement efforts in rural areas, and they have funded the stations through the budget cycle ending Dec. 11. In the meantime, Shawn Moran, spokesman for the national union representing Border Patrol employees, said the agency has ordered agents to “do as little interior enforcement as they possibly can.”
___

In recent years, the Border Patrol has aggressively resisted revealing information about its more controversial operations. The agency has deflected media requests for information regarding fatal shootings and its use of force policy, leading to questions about its handling of those incidents.

In response to the Statesman‘s requests seeking details of the agency’s enforcement actions far from the border, it provided only general numbers on arrests — and those only after the newspaper formally appealed the agency’s initial refusal to respond to a Freedom of Information request.

Those numbers show apprehensions have fallen steeply in the Texas interior to 1,459 last year, down from 4,448 in 2010 and 9,234 in 2005. During the same time, staffing levels at those interior stations have remained relatively stable — ranging from 47 to 66 agents over the last decade. The agency wouldn’t reveal staffing levels or apprehension numbers at individual stations.

But in important ways, the information fails to provide a full picture of the agency’s activities deep within the state.

Officials wouldn’t provide the location of the interior arrests, for example. So, while all of the substations are close to, or beyond the 100-mile zone, it is impossible to determine which apprehensions occurred beyond that line. Nor would the agency say how many of the inland arrests were the result of roving patrols, rather than checkpoint stops or joint operations with other agencies.

Others seeking information about the agency’s work away from the border have also encountered obstacles. American Civil Liberties Union chapters in Arizona and Washington state sued the Border Patrol last year in an attempt to obtain roving patrol information in those states. A congressional effort to require annual reports of the Border Patrol’s interior activities failed earlier this year.

The Border Patrol also says it doesn’t track how many U.S. citizens or legal residents it pulls over during its roving patrols far from the border. The agency denied a Statesman request to visit an interior station and speak with agents.

But the 2012 deposition of one veteran Border Patrol agent in San Angelo, obtained by the Statesman, provides perhaps the most detailed account yet of how the agency’s officers operate far from the border.

The agent testified that he spent much of his time patrolling roads, pulling over work crews of illegal immigrants traveling to and from such places as Austin and Houston. But only about half of his roving patrol stops resulted in arrests, he said, suggesting that large numbers of American citizens and legal residents might have been stopped and released by Border Patrol agents over the last decade.

Border Patrol spokesman David Vera told the Statesman that agents “enforce the nation’s laws while preserving the civil rights and civil liberties of all people. … (Customs and Border Protection) does not tolerate racial profiling or agent misconduct and appropriately investigates allegations of wrongdoing.”

In a court filing, Armendariz argued that the Border Patrol’s use of roving patrols resulted in “willful and abusive traffic stop-type seizures of Latino drivers in and around central Texas.”

___

The Border Patrol’s authority to stop and search is almost unlimited at the border, where agents can question subjects without providing a reason or obtaining a warrant. They also enjoy wide latitude at checkpoints located close to the border.

The law allows Border Patrol agents to make warrantless stops at “a reasonable distance” from the border, defined as 100 air miles from the border by the Justice Department in 1953. But the actual limit of Border Patrol’s authority is blurry, its policies peppered with vague directives.

Federal rules technically allow agents to operate beyond the 100-mile line by asking permission from top agency brass in the event of “unusual circumstances.”

“Chief Patrol Agents can submit operations orders … when they believe certain vulnerabilities exist that require the reasonable distance from the border to be greater than 100 miles for a certain amount of time and/or in a certain geographic area,” Vera said.

It is unclear, however, whether such special permission is given case by case, or as a blanket approval for a specific period of time or geographical area. The agency didn’t answer questions about the policy. Yet some inland operations appear to be permanent and ongoing: In Texas, the agency has half a dozen stations well beyond the 100-mile mark.

Agents working far from the border also must have both a “reasonable certainty” that a suspect recently crossed a border as well as a “reasonable suspicion” that he or she is engaged in criminal activity, which can include being in the country illegally.

Courts have identified eight separate factors that agents can cite to make stops, including proximity to the border, recent trafficking activity and number of passengers. Traffic violations alone don’t provide justification for a Border Patrol stop, nor does race or ethnicity. Several courts also have ruled that the farther away from the border, where motorists have a greater expectation of privacy, the stronger the Border Patrol’s justification for the stop must be.

Yet generally, judges have interpreted those standards broadly, permitting agents to stop suspects, for example, because of a tip that they’d recently crossed the border. Especially heavily loaded trucks have also been adjudged reason enough for a stop, including one case in which agents spotted a spare tire in the back seat of a truck to make room in the covered bed, which sagged noticeably under the weight of what turned out to be human cargo.

___

Border Patrol agents have stopped motorists deep inside Texas on much more questionable pretexts.

In 2012, the 5th U.S. Circuit Court of Appeals declared unlawful the apprehension of an undocumented immigrant driving on Interstate 20 near Odessa, 170 miles from the border, after dismissing agents’ claims the stop was warranted because the truck’s passengers avoided eye contact.

Nor did the court buy the agency’s claim it had reasonable suspicion to stop the vehicle because an agent had observed one of the passengers pointing to an open field. “This passenger could have been pointing to anything: an animal, a tree, etc.,” the judges wrote.

In 2007, a Border Patrol agent on roving patrol in San Antonio — about 130 miles from the Mexico border — stopped a Guatemalan immigrant based on his apparent nervousness and the appearance of his “cheekbones, jaws, ears, and forehead.” The agent said the facial features gave the man the look of an “OTM,” or Other Than Mexican immigrant.

In declaring the stop illegal, immigration Judge Glenn McPhaul wrote that the arresting agent couldn’t answer “how one Hispanic person might stand out from another as an illegal alien when the Hispanic population is so high in San Antonio.”

Border Patrol agents operating in the state’s interior have cited other vague suspicions to stop drivers. In 2010, agents stopped four men driving on FM 187 in the Hill Country town of Vanderpool, 90 air miles from the border crossing in Del Rio.

“The typical drivers on these roads are very friendly and courteous, especially when they see our marked Border Patrol vehicles,” one of the agents said in a sworn statement. “Instead of waiving (sic) to me like the typical drivers in the area, the driver’s facial expression changed immediately upon making eye contact with me.”

One of the passengers was Alejandro Garcia de la Paz, a 24-year-old graduate of Harlandale High School in San Antonio whose parents had brought him to the U.S. from Mexico when he was 1 year old. With few relatives remaining in Mexico, he worried about returning to a country he didn’t remember.

“I guess they just want to get the most people out of here that they can,” Garcia de la Paz said. “We’re already here, already working. It’s not like we came from Mexico yesterday.”

Garcia de la Paz is currently protected from deportation by the Deferred Action for Childhood Arrivals program. His federal lawsuit claiming an illegal stop is pending.
___

In San Angelo, 130 miles from the border, roving patrols constituted the principal activity of agents, according to Border Patrol agent John Finney, whose 2012 deposition in a deportation court case provides a rare description of the agency’s otherwise hidden operations.

Finney said he mainly looked for groups of men in work trucks with “tools … maybe bags, luggage … ice chests.” The vast majority of targets were Hispanic. “Maybe not 99 percent, but in the high nineties,” he said.

Finney also described how he determined if people were acting suspiciously: If motorists did a “classic double take” when they saw him on the side of the road, Finney said he would often follow them, observing them from behind or from a parallel lane, sometimes slowing down for the vehicle to pass before speeding up again to gauge facial expressions and movements.

In court filings, Armendariz argued Finney’s tactics might cause a driver to act suspicious: “This bizarre behavior no doubt bewilders most drivers.”

The strategy yielded mixed results. Finney estimated that “a little better than 50 percent” of stops based on such reasonable suspicion were of undocumented immigrants and so resulted in arrests — a low record of success, Armendariz wrote in response: “By his own admission, Mr. Finney gets it wrong and stops U.S. citizens or persons in lawful immigration status almost half the time.”

At least two immigration judges have questioned the agency’s use of inland roving patrols. In 2010, immigration Judge Bertha Zuniga declared unconstitutional the stop of an immigrant pulled over in San Angelo because he appeared Latino, writing: “To allow roving-patrol stops of all vehicles in San Angelo, Texas, carrying Hispanic-looking persons without further evidence of suspicious activity would subject residents to ‘unlimited interference with their use of the highways.’ ” That same year, immigration Judge John D. Carte made the same ruling in a San Antonio case.

Roving patrol stops of U.S. citizens also sparked a 2012 lawsuit by the Washington state ACLU, which accused the agency of failing to establish reasonable suspicion before stopping motorists in the Olympic Peninsula. According to the lawsuit, the Border Patrol’s actions often seemed “based on nothing other than the ethnic and/or racial appearance of a vehicle’s occupants.”

Last year, the Border Patrol settled the lawsuit without admitting wrongdoing and agreed to provide its agents in the area with additional training in constitutional privacy rights. The agency also agreed to publicly disclose its traffic stop information for 18 months. According to that data, in the year following the settlement, Border Patrol made only seven roving patrol stops in the area.

MCT Photo/Michael Robinson Chavez/Los Angeles Times

Interested in more national and political news? Sign up for our daily email newsletter.