A century after this chapel was built to house open inquiry, a Middlebury honors thesis set out to measure what students will actually say out loud. Its finding is not that anyone was forbidden to speak. It is that silence, at an institution that formally protects speech, is still produced.
Ignacio Nicolás Gamero · High Honors, Department of Political Science · Middlebury College · May 2026. Thesis chair: Matthew Dickinson, Ph.D.
A survey of 184–216 students, interviews with students and faculty, and ordered logit regression testing how a student's politics predicts what they are willing to say in class. Published here with the author's permission.
Liberal to conservative students in the survey sample
Of conservatives report self-censoring - against 34.6% of liberals
Of conservatives call class “open to debate” - against 75.7% of liberals
Of conservatives say politics is framed in moral terms - against 36.8% of liberals
Increase in the odds of self-censoring for each step rightward (p<0.001)
Of all respondents describe the campus as liberal or very liberal
Student survey, N=184–216 depending on item completion. Ordered logit models, thesis pp. 37–51.
A single campus survey invites an easy dismissal: that the students who feel silenced are simply the ones who bothered to answer. So each headline figure was set against large datasets the author had no hand in collecting - FIRE, Heterodox Academy, HERI, and Middlebury's own institutional Zeitgeist surveys. They agree.
| The College’s own surveys | Middlebury surveys its own students. Those surveys found the campus is 79-82% liberal and only 8-10% conservative - an even wider gap than this thesis reports. The free-speech group FIRE measured the college separately and came out almost exactly the same. |
|---|---|
| The 52-point openness gap | Zeitgeist 7.0 (N=1,028) found 53% of liberal students describing the campus as open, against 21% of conservatives. |
| The rate of self-censorship | FIRE’s College Free Speech Rankings put 43% of Middlebury students at self-censoring monthly or more, against a national baseline of 45-48%. |
| The conservative-liberal asymmetry | Heterodox Academy’s national Campus Expression Survey finds 32% of conservative students “very reluctant” to discuss politics against 8% of liberals - the same roughly 4:1 shape. |
| Where the imbalance comes from | HERI’s longitudinal data shows students do not arrive this way. Incoming cohorts lean only slightly left, 33.5% liberal to 21.6% conservative. By senior year it is 77% to 18%. |
External validation, thesis pp. 52–55.
“Self-censorship in elite higher education is not primarily the result of formal restriction, but of an environment in which certain viewpoints are consistently perceived as more legitimate than others. In such contexts, silence is not incidental, it is produced.”
— Gamero, conclusion, p. 120
One campus, two erasures - a name taken off a wall, and views taken out of a room. Both start with the same instinct to settle a question by removing it.
This case has always turned on whether decisions get argued openly or arrived at quietly. So does a classroom.
This is not a grievance. It is 149 pages of surveys, interviews, and regression analysis, submitted for honors and defended.
The thesis does not stop at diagnosis. Its final chapter sets out what Middlebury could do, pp. 102–112.
The first pages are previewed below. Open or download the full 149-page document to read the methodology, interviews, and appendices.
A 12-page preview loads in your browser. The full document is 4.2MB.