The Origins of Modern Polling and Electoral Mathematics American political polling emerged systematically in the 1930s through figures like George Gallup, transforming elections from purely intuition-based campaigns into data-driven operations. By the 2016 election, major polling aggregators like FiveThirtyEight were tracking 2,000+ individual surveys with methodological sophistication that would have seemed impossible two decades prior. However, polling accuracy has declined measurably since 2000—the average error margin has widened from ±2-3% to ±4-5%—driven by rising mobile-only households, demographic sorting, and lower response rates (declining from 36% in 1997 to under 6% today). This shift matters because it's fundamentally altered how campaigns allocate resources, how media frames races as competitive or foregone, and how citizens develop expectations about electoral outcomes—meaning the infrastructure that shaped modern democratic participation is itself becoming less reliable.