The U.S. Supreme Court is entrusted with resolving disputes about how the United States Constitution and other federal laws should be applied to cases that have been appealed from lower courts. The justices base their decisions on their interpretation of both legal doctrine and the precedential application of laws in the past. In most cases, interpreting the law is relatively clear-cut and the justices decide unanimously. However, in more complicated or controversial cases, the Court is often divided.

It has long been commonly assumed that the votes of Supreme Court justices reflect their jurisprudential philosophies as well as their ideological leanings, personal attitudes, values, political philosophies, or policy preferences. A growing body of academic research has confirmed this understanding: scholars have found that the justices largely vote in consonance with their perceived values.[1][2] Analysts have used a variety of methods to deduce the specific perspective of each justice over time.

Ideological leanings over time

Researchers have carefully analyzed the judicial rulings of the Supreme Court—the votes and written opinions of the justices—as well as their upbringing, their political party affiliation, their speeches, editorials written about them at the time of their Senate confirmation, and the political climate in which they are appointed, confirmed, and work.[3] From this data, scholars have inferred the ideological leanings of each justice and how the justices are likely to vote on upcoming cases.[4]

Using statistical analysis of Supreme Court votes, scholars found that an inferred value representing a Justice's ideological preference on a simple conservative–liberal scale is sufficient to predict a large number of that justice's votes.[5] Subsequently, using increasingly sophisticated statistical analysis, researchers have found that the policy preferences of many justices shift over time.[6][7][8] 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 measurement model 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.[9][10][11][12] The graph below shows the results of their analysis: the ideological leaning of each justice from the term beginning in October 1937 to the term that began in October 2016.[13][14] Note that the scale and zero point are arbitrary—only the relative distance of the lines is important. Each unique color represents a particular Supreme Court seat, which makes the transitions from retiring justices to newly appointed justices easier to follow. The black lines represent the leanings of the Chief Justices. The yellow line represents the estimated location of the median justice—who, as Duncan Black’s median voter theorem posits, is often the swing vote in closely divided decisions.[15]

Michael A. Bailey used a slightly different Markov chain Monte Carlo Bayesian method to determine ideological leanings and made significantly different scaling assumptions.[16][17][18] He analyzed cases by calendar year and supplemented the data regarding votes in each Court case with additional information from the majority, concurring, and dissenting opinions in which justices commented on previous cases, as well as with votes made by members of Congress on similar legislation, amicus filings by Solicitors General and members of Congress, and presidential and Congressional positions on Court cases. 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.[19] However, he only used votes and cases related to the major topics addressed by the courts in the postwar area: crime, civil rights, free speech, religion, abortion, and privacy. He did not include federalism or economic issues.[16][20]

The graph below shows the ideological leaning of each justice by calendar year from 1950 to 2011.[21] The scale and zero point roughly correspond to DW-Nominate Common Space scores, but otherwise are arbitrary. As in the graph above, each unique color represents a particular Supreme Court seat. The black lines represent the leanings of the Chief Justices. The yellow line represents the median justice.

These two graphs differ because of the choices of data sources, data coverage, coding of complicated cases, smoothing parameters, and statistical methods. Each of the lines in these graphs also has a wide band of uncertainty. Because these analyses are based on statistics and probability, it is important not to over-interpret the results.[22][23] Also, the nature of the cases the Supreme Court chooses to hear may lead the justices to appear more liberal or conservative than they would if they were hearing a different set of cases. And all cases are valued equally even though, clearly, some cases are much more important than others.[24][25] Still, they offer an indication of the overall ideological orientation of the justices and provide a visualization of changes in the Court's orientation over time.

Ideological shifts since 1937

In the early 1930s (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. Two justices (Chief Justice Charles Evans Hughes and Justice Owen Roberts) normally cast the swing votes.

As the Martin–Quinn graph shows, by the 1939 term, Roosevelt had moved the Court to a more liberal position by appointing four new justices including strong liberals Hugo Black, William O. Douglas, and Frank Murphy. However, led by the increasingly conservative Chief Justices Harlan Stone and Fred Vinson, the Court shifted in a more conservative direction through the early 1950s.

President Dwight Eisenhower appointed Earl Warren to be Chief Justice in 1953, and both graphs indicate that the Court then turned in a more liberal direction as Warren grew substantially more liberal and especially when he was joined by strong liberal justices William Brennan, Arthur Goldberg, Abe Fortas, and Thurgood Marshall (though Justices Black and Felix Frankfurter became more conservative over time).

In the 1970s, the Court shifted in a more conservative direction when President Richard Nixon appointed Chief Justice Warren Burger and strong conservative Justices Lewis Powell, William Rehnquist, and Harry Blackmun, and more so when President Ronald Reagan elevated Rehnquist to Chief Justice (though Blackmun became more liberal over time). The Court shifted to an even more conservative orientation when it was joined by strong conservative Justices Antonin Scalia (appointed by President Ronald Reagan), Clarence Thomas (appointed by President George H.W. Bush), and Samuel Alito and Chief Justice John Roberts (both whom were appointed by President George W. Bush). During this time, Justice David Souter became more liberal over time.[26]

Both graphs indicate that the current Roberts Court remains conservative, with four conservative justices (including Chief Justice Roberts, though he has become more liberal) and the median position held by Justice Anthony Kennedy (appointed by President Ronald Reagan), who has also become more liberal. So far, Justice Gorsuch (appointed by President Trump) is following the conservative path taken by Justice Scalia. Dissenting in many key cases are Justices Sotomayor and Kagan (appointed by President Obama) along with Justice Ginsburg, who has continued to become more liberal, and Justice Breyer (both appointed by President Clinton).[14][27]

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. The path of Justice Harry Blackmun illustrates the ideological drift shown by many justices.[8] Blackmun (purple line) had a conservative score (Quinn–Martin = 1.767; Bailey = 0.43) in the 1970 term, his first on the bench, but had shifted to a liberal score (Quinn–Martin = −1.943; Bailey = −0.81) by the 1993 term, his last. The median justice was Byron White for most of the time from 1970 to 1993, Sandra Day O’Connor from 1994 to 2005, and Anthony Kennedy since 2006.

Career "liberal" voting percentage by issue area from 1946–2017

The following sortable table[a] lists the lifetime percentage "liberal" scores of Supreme Court justices as compiled in the Supreme Court Database.[28] The table shows data for justices whose service began at or after the 1946 term; the data ends with the 2016–2017 term.

The term "liberal" in the Supreme Court Database represents the voting direction of the justices across the various issue areas. It is most appropriate in the areas of criminal procedure, civil rights, and First Amendment cases, where it signifies pro-defendant votes in criminal procedure cases, pro-women or -minorities in civil rights cases, and pro-individual against the government in First Amendment cases. In takings clause cases, however, a pro-government/anti-owner vote is considered liberal. The use of the term is probably less appropriate in union cases, where it represents pro-union votes against both individuals and the government, and in economic cases, where it represents pro-government votes against challenges to federal regulatory authority and pro-competition, anti-business, pro-liability, pro-injured person, and pro-bankruptcy decisions. In federalism and federal taxation cases, the term indicates pro-national government positions.

A highlighted row indicates that the Justice is currently serving on the Court.

# Justice Year Confirmed Position Criminal Procedure Civil Rights First Amendment Union Economic Federalism Federal Taxes
85 Vinson, Fred M.Fred M. Vinson 1946 Chief Justice 27.0% 45.3% 26.2% 34.1% 59.9% 51.5% 74.5%
86 Clark, Tom C.Tom C. Clark 1949 Associate 32.1% 53.5% 37.1% 61.9% 72.5% 51.3% 74.8%
87 Minton, ShermanSherman Minton 1949 Associate 23.1% 36.2% 27.0% 55.9% 62.0% 53.3% 75.6%
88 Warren, EarlEarl Warren 1953 Chief Justice 73.3% 81.0% 83.5% 70.2% 81.3% 72.1% 80.0%
89 Harlan II, John MarshallJohn Marshall Harlan II 1955 Associate 37.8% 42.0% 44.2% 53.8% 36.7% 50.0% 70.6%
90 Brennan, Jr., William J.William J. Brennan, Jr. 1956 Associate 76.2% 82.5% 84.5% 65.8% 69.8% 66.3% 70.3%
91 Whittaker, Charles EvansCharles Evans Whittaker 1957 Associate 39.8% 44.2% 36.5% 40.5% 33.6% 52.4% 63.2%
92 Stewart, PotterPotter Stewart 1958 Associate 44.9% 48.4% 63.0% 57.3% 45.0% 56.5% 66.4%
93 White, ByronByron White 1962 Associate 32.9% 55.2% 39.2% 62.6% 59.3% 66.9% 84.7%
94 Goldberg, ArthurArthur Goldberg 1962 Associate 75.0% 98.0% 96.2% 61.1% 63.2% 53.3% 77.3%
95 Fortas, AbeAbe Fortas 1965 Associate 81.5% 82.1% 79.5% 60.0% 72.2% 64.3% 46.7%
96 Marshall, ThurgoodThurgood Marshall 1967 Associate 80.1% 84.5% 83.4% 68.9% 63.0% 68.6% 74.2%
97 Burger, Warren E.Warren E. Burger 1969 Chief Justice 19.0% 37.4% 30.9% 43.3% 45.0% 66.7% 72.1%
98 Blackmun, HarryHarry Blackmun 1970 Associate 42.3% 62.1% 56.2% 61.8% 55.0% 67.4% 74.4%
99 Powell, Jr., Lewis FranklinLewis Franklin Powell, Jr. 1971 Associate 28.4% 40.3% 46.1% 51.1% 45.6% 59.0% 56.1%
100 Rehnquist, WilliamWilliam Rehnquist[b] 1971 Associate 14.1% 24.1% 18.1% 48.9% 43.5% 29.1% 62.5%
101 Stevens, John PaulJohn Paul Stevens 1975 Associate 66.9% 64.3% 66.9% 63.9% 56.2% 56.9% 59.4%
102 O'Connor, Sandra DaySandra Day O'Connor 1981 Associate 26.4% 45.4% 41.8% 42.7% 46.5% 48.3% 58.9%
100 Rehnquist, WilliamWilliam Rehnquist[c] 1986 Chief Justice 19.6% 30.6% 26.1% 33.3% 49.4% 46.1% 77.8%
103 Scalia, AntoninAntonin Scalia 1986 Associate 27.4% 30.6% 30.5% 33.8% 46.2% 52.0% 69.7%
104 Kennedy, AnthonyAnthony Kennedy 1988 Associate 33.3% 44.0% 46.0% 39.7% 44.6% 56.3% 81.0%
105 Souter, DavidDavid Souter 1990 Associate 57.2% 69.3% 70.8% 58.8% 54.2% 64.8% 70.7%
106 Thomas, ClarenceClarence Thomas 1991 Associate 22.4% 25.0% 32.2% 30.8% 46.2% 47.4% 58.7%
107 Ginsburg, Ruth BaderRuth Bader Ginsburg 1993 Associate 62.3% 70.0% 69.4% 72.2% 56.6% 60.6% 80.6%
108 Breyer, StephenStephen Breyer 1994 Associate 55.6% 69.7% 55.8% 77.1% 50.0% 67.7% 79.4%
109 Roberts, JohnJohn Roberts 2005 Chief Justice 31.1% 42.4% 50.0% 41.2% 44.5% 73.2% 90.9%
110 Alito, SamuelSamuel Alito 2006 Associate 18.6% 38.8% 39.0% 31.3% 42.4% 68.4% 81.8%
111 Sotomayor, SoniaSonia Sotomayor 2009 Associate 66.9% 70.7% 67.7% 61.5% 50.0% 63.0% 100.0%
112 Kagan, ElenaElena Kagan 2010 Associate 62.5% 71.9% 68.0% 66.7% 50.0% 75.0% 80.0%
113 Gorsuch, NeilNeil Gorsuch[d] 2017 Associate 0.0% 50.0% 100.0% No cases 33.3% No cases No cases
  1. ^ To sort, click on the arrow in the header. To sort by multiple columns, click on the first column's sort arrow, then shift-click on subsequent columns' sort arrows.
  2. ^ William Rehnquist's terms as Associate Justice and Chief Justice are listed separately.
  3. ^ William Rehnquist's terms as Associate Justice and Chief Justice are listed separately.
  4. ^ Percentages for Neil Gorsuch may be misleading since he has participated in only a few cases so far.

See also


  1. ^ Segal, Jeffrey A.; Spaeth, Harold J. (2002). The Supreme Court and the Attitudinal Model Revisited. Cambridge: Cambridge University Press. ISBN 978-0521783514. 
  2. ^ Epstein, Lee; Martin, Andrew D. (2012). "Is the Roberts Court Especially Activist? A Study of Invalidating (and Upholding) Federal, State, and Local Laws" (PDF). Emory Law Journal. 61: 737–758. …for Justices appointed since 1952, Epstein & Landes’s findings parallel ours: the vast majority were opportunistic restraintists (activists), willing to uphold laws that were consistent with their policy preferences and strike those that were not. 
  3. ^ Zorn and Caldeira provide a good overview of these methods and their limitations: Zorn, Christopher; Caldeira, Gregory A. (5 February 2008). "Measuring Supreme Court Ideology" (PDF). Retrieved 30 November 2012. 
  4. ^ For example:
    • Segal, Jeffrey A.; Cover, Albert D. (June 1989). "Ideological Values and the Votes of U.S. Supreme Court Justices" (PDF). American Political Science Review. 83 (2): 557–565. doi:10.2307/1962405.  Segal and Cover found (p. 561) a strong correlation (0.80) between justices' perceived ideological perspectives on civil liberties and civil rights issues as attributed to them in elite newspaper editorials written just before their confirmation (their Segal–Cover score) and their later votes in the study period 1953–1988.
    • Epstein, Lee; Walker, Thomas G.; Dixon, William J. (November 1989). "The Supreme Court and Criminal Justice Disputes: A Neo-Institutional Perspective" (PDF). American Journal of Political Science. 33 (4): 825–841. doi:10.2307/2111111.  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). In this analysis, the political party affiliation of the majority of justices provided about one-fourth of the predictive power.
    • Pinello, Daniel R. (1999). "Linking Party to Judicial Ideology in American Courts: A Meta-Analysis". The Justice System Journal. 20 (3): 219–254.  Pinello conducted a meta-analysis of 84 studies of American courts covering 222,789 cases and found that political party affiliation was a dependable indicator of rulings: Democratic judges voted in favor of liberal solutions more often than Republican judges did, especially in federal courts (the U.S. Supreme Court, U.S. Courts of Appeal, and U.S. District Courts).
  5. ^ Some examples:
    • Grofman, Bernard; Brazill, Timothy J. (2002). "Identifying the median justice on the Supreme Court through multidimensional scaling: Analysis of 'natural courts' 1953–1991" (PDF). Public Choice. 112: 55–79. doi:10.1023/A:1015601614637.  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. "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."
    • Poole, Keith (10 July 2003). "The Unidimensional Supreme Court". Retrieved 27 November 2012. The bottom line is that the current Court is basically unidimensional.  Poole used various statistical measures to show that a unidimensional scale provides a good measure of the Rehnquist Court during the 8-year period 1995–2002.
  6. ^ Martin, Andrew D.; Quinn, Kevin M. (2007). "Assessing Preference Change on the US Supreme Court" (PDF). The Journal of Law, Economics, & Organization. 23 (2): 365–385. doi:10.1093/jleo/ewm028. 
  7. ^ Epstein, Lee; Martin, Andrew D.; Quinn, Kevin M.; Segal, Jeffrey A. (2007). "Ideological Drift among Supreme Court Justices: Who, When, and How Important?" (PDF). Northwestern University Law Review. 101 (4): 1483–1503. 
  8. ^ a b Ruger, Theodore W. (2005). "Justice Harry Blackmun and the Phenomenon of Judicial Preference Change" (PDF). Missouri Law Review. 70: 1209. 
  9. ^ Martin, Andrew D.; Quinn, Kevin M. (2002). "Dynamic Ideal Point Estimation via Markov Chain Monte Carlo for the U.S. Supreme Court, 1953–1999" (PDF). Political Analysis. 10 (2): 134–153. doi:10.1093/pan/10.2.134. 
  10. ^ Martin, Andrew D.; Quinn, Kevin M. (2 May 2001). "The Dimensions of Supreme Court Decision Making: Again Revisiting The Judicial Mind". 
  11. ^ Martin, Andrew D.; Quinn, Kevin M.; Epstein, Lee (2005). "The Median Justice on the United States Supreme Court". North Carolina Law Review. 83: 1275–1322. 
  12. ^ Jackman provides a simple description of this kind of statistical analysis: Jackman, Simon (2011). "Statistical Inference, Classical and Bayesian". International Encyclopedia of Political Science. doi:10.4135/9781412959636.n585. ISBN 9781412959636. 
  13. ^ "Martin–Quinn dataset". 
  14. ^ a b Silver, Nate (29 March 2012). "Supreme Court May Be Most Conservative in Modern History". New York Times.  In his FiveThirtyEight blog, Silver graphed the Martin–Quinn data to show the movement of the median, most conservative, and most liberal justices over time.
  15. ^ Black, Duncan (February 1948). "On the rationale of group decision-making". Journal of Political Economy. 56 (1): 23–34. doi:10.1086/256633. JSTOR 1825026. 
  16. ^ a b Bailey, Michael A. (August 2012). "Measuring Court Preferences, 1950–2011: Agendas, Polarity and Heterogeneity" (PDF). Retrieved 20 November 2012. 
  17. ^ Bailey, Michael A.; Maltzman, Forrest (2011). The constrained court: law, politics, and the decisions justices make. Princeton: Princeton University Press. ISBN 978-0691151052. 
  18. ^ Clinton, Joshua; Jackman, Simon; Rivers, Douglas (May 2004). "The Statistical Analysis of Roll Call Data" (PDF). American Political Science Review. 98 (2): 355. doi:10.1017/S0003055404001194. 
  19. ^ Epstein, Lee; Martin, Andrew D.; Segal, Jeffrey A.; Westerland, Chad (May 2007). "The Judicial Common Space" (PDF). Journal of Law, Economics, and organization. 23 (2): 303–325. doi:10.1093/jleo/ewm024. The [NOMINATE] Common Spaces scores are bounded below by −1 and above by 1, whereas the Martin–Quinn scores are theoretically unbounded (currently, they range from about –6 [Justice Douglas] to 4 [Justice Thomas]). 
  20. ^ Bailey, Michael A. (July 2007). "Comparable Preference Estimates across Time and Institutions for the Court, Congress, and Presidency" (PDF). American Journal of Political Science. 51 (3): 433–448. doi:10.1111/j.1540-5907.2007.00260.x. 
  21. ^ "Bailey dataset". Retrieved 22 November 2012. 
  22. ^ Ho, Daniel E.; Quinn, Kevin M. (June 2010). "How Not to Lie with Judicial Votes: Misconceptions, Measurement, and Models" (PDF). California Law Review. 98 (3): 813–876. 
  23. ^ Carroll, Royce; Lewis, Jeffrey B.; Lo, James; Poole, Keith T.; Rosenthal, Howard (November 2009). "Comparing NOMINATE and IDEAL: Points of Difference and Monte Carlo Tests" (PDF). Legislative Studies Quarterly. 34 (4): 555–591. doi:10.3162/036298009789869727. 
  24. ^ Farnsworth, Ward (September 2007). "The Use and Limits of Martin–Quinn Scores to Assess Supreme Court Justices, with Special Attention to the Problem of Ideological Drift". Northwestern University Law Review. 101 (4): 1891–1903. SSRN 1000986Freely accessible. 
  25. ^ McGuire, Kevin T.; Vanberg, Georg; Smith, Jr., Charles E.; Caldeira, Gregory A. (October 2009). "Measuring Policy Content on the U.S. Supreme Court" (PDF). The Journal of Politics. 71 (4): 1305–1321. doi:10.1017/s0022381609990107. 
  26. ^ Liptak, Adam (24 July 2010). "Court Under Roberts Is Most Conservative in Decades". New York Times. 
  27. ^ Epstein, Lee; Landes, William M.; Posner, Richard A. (2013). "How Business Fares in the Supreme Court" (PDF). Minnesota Law Review. 97: 1431–1472. We find that five of the ten Justices who, over the span of our study (the 1946 through 2011 Terms), have been the most favorable to business are currently serving, with two of them ranking at the very top among the thirty-six Justices in our study. 
  28. ^ Epstein, Lee; Walker, Thomas G.; Staudt, Nancy; Hendrickson, Scott; Roberts, Jason (November 1, 2017). "The U.S. Supreme Court Justices Database". Retrieved November 2, 2017.