Ami Horowitz builds the video around a simple and potentially revealing question: do students who identify with the Democratic Party actually hold positions that correspond with that label? At a Los Angeles community college, he asks Black students how they identify politically and then presents a series of questions without explicitly labeling the answers Republican or Democratic. The eventual reveal creates the central dramatic payoff, as many participants who initially identify as Democrats are told that their responses place them predominantly on the Republican side. Several appear genuinely surprised, and the strongest moments come when that surprise leads them to reconsider the relationship between party identity and individual beliefs.

The questions cover abortion, illegal immigration, parental versus government influence over education, taxes, policing, medical transition for children, transgender participation in women’s sports, social spending and debt, voter identification, and climate policy. That breadth gives the exercise more substance than a single-issue political gotcha, and the students do not respond uniformly. Some support abortion rights, increased social spending, or climate action, while others favor lower taxes, stronger immigration enforcement, more policing, voter identification, or restrictions related to gender policy. The footage therefore does show participants whose views cross conventional partisan boundaries rather than fitting neatly into one ideological package.

The difficulty is that the test itself is never explained well enough to justify the precision of its conclusions. Horowitz reports results such as 60%, 70%, or 80% Republican, but viewers are not shown a scoring methodology, weighting system, source for the ideological classifications, or explanation of why each binary answer should map cleanly onto one party. Several questions also bundle multiple propositions together. The climate question, for example, contrasts more regulation and higher taxes to address climate change with less regulation and lower taxes to spur the economy, while the social-program question links higher spending with higher debt and reduced spending with debt reduction. A respondent may agree with one part of an option but not another, yet the format forces a single partisan classification.

Some questions are framed in ways that further complicate the experiment. Asking whether parents or government should have “more control” over education reduces a complicated division of responsibility to a binary choice. The question about medical transition for children does not define age, intervention, circumstances, or what “medically transition” includes. The sports question similarly uses a broad formulation without distinguishing different policies or levels of competition. These subjects can legitimately reveal political differences, but wording matters if the resulting answers are going to be treated as measurements of partisan alignment. Without consistent definitions or neutral alternatives, the exercise is better understood as an informal street interview than a rigorous political test.

The student reactions are nevertheless the video’s most valuable material. One participant says the result encourages looking more closely at issues and elections rather than relying on party letters, while another discusses the social association between Republican identity and whiteness and the criticism Black Republicans can face. Those comments point toward a more interesting question than whether someone is secretly Republican: political identity can involve community, history, priorities, candidate choices, and differing levels of importance assigned to individual issues. A voter can agree with one party on six questions and still rationally support the other because the remaining issues matter more to that person. The percentage score cannot capture that hierarchy.

Horowitz’s closing claim goes further than the footage can comfortably support. He says that not a single Black person tested on the campus aligned more with Democrats than Republicans, but the video does not establish how participants were recruited, how many people completed the test, whether every interview was shown, or whether the sample was representative of the campus, Los Angeles, Black voters, or any broader population. The introductory statement that nearly 90% of Black people have voted Democratic over the years is also offered without sourcing or defining the elections and period being referenced. Even assuming the displayed results accurately represent everyone interviewed, they demonstrate what happened within this particular exercise, not a general mismatch between Black voting behavior and political beliefs.

As a piece of political entertainment, the video has an effective structure: establish identity, conceal the partisan scoring, ask recognizable policy questions, and reveal an unexpected result. The students are given enough room to explain some answers, and their willingness to engage produces moments of genuine reflection rather than only confrontation. As evidence for the larger implication suggested by the premise, however, the methodology is too opaque and the questions too reductive. The experiment succeeds at illustrating that party identification does not automatically predict every policy preference, but it does not establish that these students are “actually” Republicans, much less support broader conclusions about Black voters without a transparent test, representative sampling, and more nuanced measurement of political priorities.

Pros

• Uses a clear reveal structure that effectively contrasts participants’ stated political identities with their answers to individual policy questions.
• Covers enough different subjects to demonstrate that some students hold a mixture of positions rather than following a uniform partisan line.
• Several participants respond thoughtfully to unexpected results and explicitly consider judging issues independently of party labels.
• A student’s discussion of the social associations surrounding Black Republican identity introduces a deeper dimension that the numerical scores alone cannot capture.
• Allows participants to explain some of their reasoning instead of reducing the entire presentation to reaction shots.

Cons

• Provides no transparent scoring methodology, weighting system, or source explaining how answers become precise Republican or Democratic percentages.
• Several questions bundle multiple policies together or use broad binary framing that can push nuanced positions into simplistic partisan categories.
• Does not disclose the sample size, participant-selection process, omitted interviews, or other information needed to evaluate the closing claim about everyone tested.
• Treats agreement with a majority of selected positions as evidence of partisan identity without considering how voters prioritize issues differently.
• The introductory voting statistic and several political claims are presented without sourcing or sufficient context.
• The final implication reaches beyond what an informal campus exercise can establish about either the participants’ actual political identities or Black voters more broadly.

The experiment produces engaging reactions and a worthwhile reminder that political beliefs often cross party lines, especially when people consider policies individually rather than beginning with a partisan label. Its larger conclusion is much less persuasive because opaque scoring, loaded binary choices, unknown sampling, and the absence of issue weighting make the supposed Republican percentages look more definitive than the exercise warrants; the students’ own reflections on political identity ultimately provide more insight than the test designed to classify them.