Democracy depends on a feedback loop: citizens express preferences, representatives respond, citizens evaluate the response. When that loop breaks, institutions lose the signal they need to function, and concentrated interests fill the void. The evidence suggests we are deep into this failure mode.
The measurement tools are broken
Professional polling — the primary instrument of democratic feedback for nearly a century — is in structural decline. Response rates for telephone surveys, once above 35%, have fallen below 6% in recent years. Gallup ended 88 years of presidential approval polling. Online surveys — the replacement most researchers have turned to — face a new threat: AI chatbots that infiltrate them with a 99.8% success rate at evading detection. Researchers publishing in the Proceedings of the National Academy of Sciences called this a potential existential threat to survey-based social science.
The bot problem is not a marginal concern. If nearly all online surveys are vulnerable to synthetic respondents generating plausible but fabricated answers at scale, the data feeding into policy decisions, academic research, and media narratives may not represent anyone's actual views at all.
The silence is documented — and self-reinforcing
Even where measurement tools exist, they may not capture what people actually think. A 20-year longitudinal study tracking one million people found that between 40–67% of Americans self-censor their political views. The Cato Institute independently confirmed that 62% of Americans say they have political views they're afraid to share.
This rate of self-censorship exceeds the McCarthy era. In the early 1950s, when congressional investigations targeted political dissidents, approximately one in eight Americans reported feeling afraid to speak freely. Today, the proportion is between two in five and two in three.
The sociologist Elisabeth Noelle-Neumann described the mechanism: when people perceive their views as minority positions, they remain silent, which makes those views appear even more marginal, which induces further silence. This "spiral of silence" concentrates visible discourse among those least representative of the broader population — the loudest, the most ideological, and increasingly, the automated.
The consequences are economic, not just civic
The standard framing treats democratic erosion as a civic problem. The evidence suggests it is equally an economic one.
Economists Daron Acemoglu, Suresh Naidu, Pascual Restrepo, and James Robinson published a rigorous causal analysis in the Journal of Political Economy finding that democratization increases GDP per capita by approximately 20% over 25 years. Their methodology — using waves of regional democratization as instrumental variables — addresses the endogeneity problems that plagued earlier research. The finding is not that wealthy countries become democratic; it is that democratic institutions directly produce economic growth through investment, education, public goods provision, and reduced social conflict.
The political scientists Martin Gilens and Benjamin Page analyzed 1,779 policy issues and found that average citizens' preferences had a statistically near-zero, non-significant impact on policy outcomes. Economic elites and organized interest groups, by contrast, had substantial and statistically significant influence. When the feedback loop fails, policy optimizes for concentrated interests rather than broad-based prosperity.
This is the mechanism through which democratic erosion becomes economic extraction. Strip away democratic voice, and you strip away the institutional foundation that enables broad-based growth. Pew Research reports that 85% of Americans believe their elected officials don't care what they think — up 30 percentage points since 2000. When that proportion of the population has functionally exited the democratic feedback loop, the institutions theoretically accountable to them have lost the signal they need.
The gap Open Caucus is designed to fill
The problem is not that people don't have political views. It is that no existing infrastructure can surface those views with three simultaneous properties: resistance to bot manipulation, protection from surveillance, and verifiable geographic granularity.
Traditional polling fails on bot resistance and is declining in reach. Social media fails on all three. Voting — the ultimate expression of political will — happens too infrequently to serve as a continuous feedback mechanism and is itself under increasing pressure.
Open Caucus is designed to occupy this specific gap: a continuous, privacy-preserving, bot-resistant instrument for measuring what communities actually care about — built on infrastructure that cannot be compelled to disclose individual responses.
Whether it succeeds depends on execution, adoption, and trust earned over time. But the gap itself is documented, widening, and consequential. Someone will build infrastructure to fill it. The question is whether that infrastructure serves democratic accountability or surveillance capitalism.
← Back to Open Caucus