This article discusses the modeling of time dependence in binary data, focusing on the limitations of time dummies and splines, and proposing a simpler alternative using cubic polynomials. The authors argue that time dummies can lead to estimation problems due to separation, while splines, though more flexible, are often misused by researchers who do not fully understand them. The cubic polynomial approach is presented as a straightforward alternative that avoids these issues and performs well in simulations. The authors reanalyze data from Crowley and Skocpol (2001) using nonproportional hazards and find new empirical support for the historical-institutionalist perspective. The article also highlights the practical consequences of separation in binary data and the importance of properly modeling time dependence in political science research. The cubic polynomial method is shown to be effective in recovering the true hazard across various hazard shapes, outperforming time dummies and splines in most cases. The authors conclude that the cubic polynomial approach is a simple and effective alternative for modeling time dependence in binary data.This article discusses the modeling of time dependence in binary data, focusing on the limitations of time dummies and splines, and proposing a simpler alternative using cubic polynomials. The authors argue that time dummies can lead to estimation problems due to separation, while splines, though more flexible, are often misused by researchers who do not fully understand them. The cubic polynomial approach is presented as a straightforward alternative that avoids these issues and performs well in simulations. The authors reanalyze data from Crowley and Skocpol (2001) using nonproportional hazards and find new empirical support for the historical-institutionalist perspective. The article also highlights the practical consequences of separation in binary data and the importance of properly modeling time dependence in political science research. The cubic polynomial method is shown to be effective in recovering the true hazard across various hazard shapes, outperforming time dummies and splines in most cases. The authors conclude that the cubic polynomial approach is a simple and effective alternative for modeling time dependence in binary data.