Forecasting in a Polarized World: The Role of Political Disagreement in Analyst Forecasts and Information Production

This paper provides novel causal evidence that political polarization affects the production of financial information by analysts. Analysts politically misaligned with firm CEOs issue systematically more pessimistic and less accurate earnings forecasts. The identification strategy compares forecasts by two analysts covering the same firm at the same time but differing in political alignment with the CEO, or the same analyst’s forecasts across firms led by CEOs from different parties. The distortions are not priced by markets: investors react similarly to aligned and misaligned forecasts. However, partisan misalignment weakens the sensitivity of firm investment to Tobin’s Q, indicating reduced investment efficiency. To investigate the mechanism, we use a large language model to analyze earnings conference calls. We find that analysts interpret CEOs’ statements through the lens of their own political views, placing different weight on firm-level versus macroeconomic information without needing to observe the CEO’s political affiliation. Finally, we show that analysts systematically sort into brokerage firms whose analyst workforce shares their political affiliation, indicating that political identity also shapes the organizational structure of information production. Overall, the findings reveal that ideological bias among financial intermediaries transmits political polarization into asset prices, the allocation of analyst talent, and ultimately firms’ real economic decisions.