This document discusses autocorrelation and its consequences. Autocorrelation occurs when error terms in a time series regression model are correlated over time. This violates the classical linear regression assumption that error terms are independent. If autocorrelation is present, it can bias standard error estimates and invalidate statistical tests. The document outlines various causes of autocorrelation like inertia in time series data, omitted variables, incorrect functional form, lags, and data manipulation. It also discusses the consequences of autocorrelation like biased standard errors and underestimated variance estimates. Methods to detect autocorrelation graphically and through statistical tests like runs tests are presented.