In 2015, the University of Arizona interviewed LGBT students across three states asking them about their experiences with over-discipline and school pushout. In response, one student said, “The teachers … they thought we were selling weed in school, they thought that me and her were both selling weed ‘cause like, the way we were dressing, ‘cause we were the only girls at that middle school that dressed like boys. So it was like ‘now we’re bad.’”

New research out of Princeton University confirms years of such anecdotal and empirical evidence: Girls, especially those of color, who experience same-sex attraction face disproportionate levels of discipline at school. And that over-discipline is a direct part of the school-to-prison pipeline.

Joel Mittleman, a doctoral candidate at Princeton University, analyzed data earlier this year from the Fragile Families and Childhood Wellbeing Study (FFCWS), a longitudinal study that has tracked the births, lives, and outcomes of nearly 5,000 U.S. children born between 1998 and 2000. The FFCWS is the only longitudinal study of this kind to also look at respondents’ sexual orientation. This particular survey determines respondents’ same-sex and different-sex attraction by asking the youth—now age 15—if they have “ever liked a girl as more than just a friend?” and “ever liked a boy as more than just a friend?” Two percent only said yes to ever liking someone of the same sex as more than a friend, and 8 percent said yes to both questions. (The majority of those answering yes to being attracted to both were girls.)

Mittleman’s analysis revealed that girls who said yes to experiencing any same-sex attraction faced far higher rates of discipline than the girls who only reported opposite-sex attraction. There was no difference among the rates of discipline for the boys.

A girl with same-sex attraction is 95 percent more likely to face discipline at school. Crucially, even though caregivers of girls with same-sex attraction were somewhat more likely to report “aggressive behavior,” that was not responsible for the disparity.

And while Mittleman controlled for race (meaning he made sure that this specific measure of over-discipline could only be attributed to sexual orientation), it is important to remember that race is still the greatest predictor of over-discipline in schools.

In fact, a much larger survey of nearly 100,000 public schools’ disciplinary actions, the results of which were released in late April, found that Black and Native students in particular face far greater rates of suspension, expulsion, and arrest than white students. And the disparity, compared to previous years, is getting worse, not better. According to a Washington Post analysis of the Civil Rights Data Collection, Black students made up 15 percent of all students in the 2015-16 school year, but 31 percent of arrests.

For LGBTQ students of color, then, the situation can be especially devastating. A series of reports released by the Movement Advancement Project (MAP) look at the impact of the broken criminal justice system on the safety of LGBTQ people. In particular, MAP’s 2016 report on LGBTQ youth showed that LGBTQ girls of color, and gender nonconforming girls of any race, face a higher risk of harsh school disciplinary policies than white girls and straight cisgender girls. GLSEN’s biennial survey of LGBTQ youth found that LGBTQ students of color, especially Black students, Latino students, and multiracial students, face far higher levels of detention, suspension, or expulsion than white students.

In examining why LGBTQ and gender nonconforming girls of color face over-discipline, MAP points in part to the work of Monique Morris and Kimberlé Williams Crenshaw, who found that Black girls who deviate from biased, so-called typical gender norms face harsher discipline.

“When Black girls do engage in acts that are deemed … a deviation from the social norms that define female behavior according to a narrow, White middle-class definition of femininity, they are deemed nonconformative and thereby subject to criminalizing responses,” Morris writes.

Girls who appear to fall short of this imaginary benchmark, they theorize, are therefore punished.

The impact of over-discipline of LGBTQ girls of color at school is not a surprise. Suspension, expulsion, and school arrest make these girls much more likely to interact with the criminal legal system. Nearly 40 percent of girls in juvenile justice facilities identify as lesbian or bisexual. Of the LGBTQ youth in juvenile incarceration facilities, 85 percent are youth of color. And while in detention, LGBTQ girls of color face rampant mistreatment and abuse from staff and other youth. According to MAP, more than 20 percent of lesbian, gay, and bisexual youth in juvenile detention report sexual victimization by staff, twice the rate of heterosexual youth (a rate that is already shameful in itself).

And when queer femme youth face harassment, targeting, and push-out, they find it even harder to access necessities later in life. Nicole “Co’Bella” Bennet, member communications coordinator for BreakOUT!, a youth-led organization working to end criminalization of LGBTQ young people in New Orleans, told Rewire.News via email, “When the school-to-prison pipeline pushes young queer girls of color out of school, it makes it twice as hard for these girls to later gain housing, a job, and further educational opportunities.”

The author of the Princeton article analyzing the Fragile Families and Childhood Wellbeing Study hopes that by “incorporating gender into the study of sexual minority students,” researchers and advocates can move their focus “away from individual acts of homophobic treatment and toward broader ideologies that affect all students.”

Holding all students to the same prejudiced norms of behavior, gender, and appearance—and refusing to confront authority figures’ own biases—is pushing vulnerable students into interactions with school officials, interactions with law enforcement, and ultimately, into the criminal legal system.