The nine Supreme Court justices base their decisions on their interpretation of both legal doctrine and the precedential application of laws in the past.
In modern discourse, the justices of the Court are often categorized as having conservative, moderate, or liberal philosophies of law and of judicial interpretation.
A growing body of academic research has confirmed this understanding, as scholars have found that the justices largely vote in consonance with their perceived values.
The simplest way to approximate the ideological leanings of Supreme Court justices is by the political party of the president who appointed them.
In a 2000 paper, Segal, Timpone, and Howard found that, in their study area (civil liberties and economics cases from 1937 to 1994), presidents appear to be reasonably successful in extending their policy preferences by appointing like-minded justices to the court, though they found that justices appear to deviate over time away from the presidents who appointed them.
[6] In all the non-unanimous decisions made by the Supreme Court in the terms from 1937 to 2023 in which there was a specifiable ideological direction, justices appointed by Republican presidents (red bars) generally cast liberal votes much less frequently than justices appointed by Democratic presidents (blue bars).
As the graph clearly shows, in every term since 1970, the Court majority (consisting of at least 5 of the justices) has been appointed by Republican presidents.
As a result, "... between 1790 and early 2010 there were only two decisions that the Guide to the U.S. Supreme Court designated as important and that had at least two dissenting votes in which the Justices divided along party lines, about one-half of one percent.
[13]: 561 [14] Epstein, Walker, and Dixon found they could explain and predict rulings in criminal justice cases (in the study period 1946–1986) using a simple model with four inputs: the political party affiliation of the majority of justices, the political party affiliation of the current president (representing the current political climate), the Supreme Court rulings in criminal justice cases in the previous year, and the percent of criminal cases the Court decides to hear in the current year (how much interest they take in the issue).
For example, Grofman and Brazill performed multidimensional scaling (MDS) using SYSTAT 5.0 of the entire range of cases considered by the Supreme Court, 1953–1991.
Analyzing terms with an unchanging membership ("natural courts") and a complete bench of nine members (3,363 cases), they found that a one-dimensional scale provided a satisfactory explanation of votes and that the degree of unidimensionality generally rose over the years, writing: "On average, over the 15 courts, the mean r2 values are 0.86 for a one dimensional metric MDS solution, and 0.97 for a two dimensional metric MDS solution.
[18] Using increasingly sophisticated statistical analysis, researchers have found that the policy preferences of many justices shift over time.
[19][20][21] The ideological leanings of justices and the drift over time can be seen clearly in the research results of two sets of scholars using somewhat different models: Andrew D. Martin and Kevin M. Quinn have employed Markov chain Monte Carlo methods to fit a Bayesian statistic measurement model[22] of ideal points (policy preferences on a one-dimensional scale) for all the justices based on the votes in every contested Supreme Court case since 1937.
[27] Note that the scale and zero point are strongly dependent on the particular set of cases chosen to be reviewed by the Court each year; the relative distance between the lines for each justice is most informative.
This additional information gave him a richer dataset and also enabled him to deduce preference values that are more consistent with the DW-Nominate Common Space scores used to evaluate the ideological leanings of members of Congress and presidents.
[32] However, he only used votes and cases related to the major topics addressed by the courts in the postwar era: crime, civil rights, free speech, religion, abortion, privacy, and labor unions.
In the early 1930s, which is earlier than the data on the Martin–Quinn graph, the "Four Horsemen" (justices James McReynolds, Pierce Butler, George Sutherland, and Willis Van Devanter) mostly opposed the New Deal agenda proposed by President Franklin D. Roosevelt.
The liberal "Three Musketeers" (justices Harlan Stone, Benjamin Cardozo, and Louis Brandeis) generally supported the New Deal.
Dissenting in many key cases are justices Sonia Sotomayor and Elena Kagan (appointed by President Barack Obama).
[46] The most volatile seat appears to be Seat 10 (light blue lines), which was held by conservative Pierce Butler until 1939, then liberal Frank Murphy until 1949, then moderate-conservative Tom Clark until 1967, then strong liberal Thurgood Marshall until 1991, and then strong conservative Clarence Thomas.