Scott Alexander explores the Median Voter Theorem, its limitations in real politics, and why recent US elections have been surprisingly close despite these complications. Longer summary

This post explores the Median Voter Theorem and its application in real-world politics. Scott first explains the theorem, which states that rational candidates will converge on the position of the median voter. He then discusses three reasons why this doesn't perfectly describe reality: primary elections, voter turnout, and party goals. Despite these complications, Scott notes that recent US elections have been surprisingly close to 50-50, suggesting some validity to the theorem. He explores how parties might target different median voters for different elections (House, Senate, Presidency) and considers how parties might adjust their positions in response to structural advantages. The post ends with questions about how quickly and effectively parties can shift their positions in response to changing political landscapes. Shorter summary

Scott proposes adapting DW-Nominate, a political spectrum analysis tool, to study morality through real-life decisions, exploring its potential implications and limitations. Longer summary

Scott proposes using a method similar to DW-Nominate, a tool used to calculate politicians' positions on the political spectrum, to investigate morality. He suggests applying this method to real-life moral decisions people make, from major choices to everyday actions. The post explores how this could potentially group people into 'good' and 'bad' categories, identify the most moral actions, and possibly provide insights into contentious moral issues. Scott also discusses how this method might work in morally inverted societies like Nazi Germany, presenting three hypothetical outcomes and their implications for understanding morality objectively. Shorter summary

The post explores the contradiction between studies suggesting elite influence on policy and those indicating money's limited impact on politics, offering several hypotheses to reconcile these findings. Longer summary

This post discusses two seemingly contradictory political science findings: one suggesting that elite opinion strongly influences US policy, and another indicating that money has little impact on politics. The author presents several hypotheses to reconcile these findings, including the possibility that legislators are themselves elites, that elites control cultural institutions, or that money influences politics through lobbying rather than campaign contributions. The post critically examines each hypothesis, considering their strengths and weaknesses in explaining the apparent contradiction. Shorter summary

Scott analyzes DW-NOMINATE data which unexpectedly shows no leftward drift in US politics over 120 years, despite historical evidence suggesting otherwise. Longer summary

Scott Alexander examines the claim that society is drifting to the Left by looking at DW-NOMINATE, a statistical system that analyzes legislators' voting patterns over time. The system shows that while Congress has become more polarized and Republicans have shifted right since 1980, there has been no significant liberal drift in either party over the past 120 years. This result contradicts the expected leftward shift given historical developments like the expansion of the welfare state and civil rights movement. Scott finds this counterintuitive and struggles to explain it, noting that the system is well-regarded and its developers defend its use for intertemporal comparisons. Shorter summary