The US Supreme Court is at the centre of economic policymaking. Trade policy, immigration enforcement, climate regulation, labour law, and the scope of federal agencies routinely hinge on judicial review. As political conflict increasingly bypasses Congress, the Supreme Court has become a decisive arena for resolving distributional disputes. This has intensified long-standing concerns that the Court’s decisions systematically favour economic elites.

Narrative accounts have long suggested such a tilt. In his book Supreme Inequality, Adam Cohen (2021) gives many examples illustrating how, since the 1970s, the Court has repeatedly sided with powerful economic actors against workers, consumers, and beneficiaries of the social safety net. Yet, translating these claims into systematic evidence has proven difficult. Standard measures of judicial ideology that rely on abstract political labels (‘liberal’ versus ‘conservative’) or malleable doctrines of constitutional interpretation (‘living constitutionalism’ versus ‘originalism’) obscure the economic implications of the Court’s rulings (e.g. Segal and Cover 1989).

Our recent paper moves away from ideology to formalise Cohen’s intuition with hard evidence (Prat et al. 2026). Instead of classifying justices by their words, as the literature has done, we measure what their decisions do: whether they directly move economic resources toward parties more likely to be wealthy or toward those less likely to be wealthy. This outcome-based perspective allows us to construct a consistent metric for each justice with an intuitive unit of measurement. We measure and combine each justice’s tendency to ‘rule for the rich’ so we can assess the Court’s shift over time and as its composition changed.

We create a comprehensive dataset of Supreme Court cases from 1953 to 2022. We restrict attention to cases with economic content (omitting criminal cases and some others) and non-unanimous decisions. Disagreement is crucial: because justices vote differently on the same case, and the set of justices on the Court changes over time, our identification lets us compare them based on a metric with absolute units.

Many decisions have second-order downstream economic effects, but we deliberately do not try to analyse those effects. Instead, we ask a narrower question: does the ruling itself directly shift economic resources between parties? When it does, we code the outcome as ‘pro-rich’ or ‘pro-poor’ depending on which side is more likely, on average, to be wealthy. Corporations are owned disproportionately by the wealthiest Americans (approximately 80% of the stock market is held by the wealthiest 10% of households) and therefore a corporate win in a labour or regulation case will be pro-rich in expectation (Economic Policy Institute 2004).

For consistency, these determinations are made without regard to the specific characteristics of the corporation (whether it is for-profit or non-profit, healthcare, or transportation, etc.). In a competition case, in which a victory for the plaintiff delivers lower prices for consumers, a justice’s vote for the plaintiff is pro-poor. Likewise, cases between citizens and the government that strengthen the social safety net are pro-poor.

This protocol yields 1,782 cases with clearly identifiable direct economic incidence. These span a wide range of disputes: employees versus firms, customers versus firms, competition policy, taxation, and the social safety net. The coding is carried out by trained research assistants following a transparent set of rules, with extensive overlap and consistency checks. While human judgement is unavoidable, and the focus only on direct economic effects is constraining by design, the approach is replicable and disciplined.

A key advantage of this framework is that it avoids self-fulfilling ideological labels. We do not ask whether a justice is conservative or liberal, or whether they claim to be analysing the constitution based on its meaning 250 years ago. We ask only whether, in a given case, their vote directly reallocates resources up or down the wealth distribution.

A stark partisan divergence

The descriptive patterns are striking. In the 1950s, Democratic- and Republican-appointed justices voted in favour of the wealthier party at roughly similar rates – around 40 to 45% of coded cases. Over time, this symmetry disappears. By 2022, the average Republican-appointed justice votes pro-rich roughly 70% of the time, while the average Democratic-appointed justice does so about 35% of the time.

This divergence is not confined to a narrow set of issues. When cases are broken down by legal category – such as unions, taxation, or federalism – the polarisation appears almost everywhere. The pattern is therefore not driven by a single doctrinal shift or policy domain. Instead, it reflects a broad and sustained separation in how justices appointed by the two parties resolve conflicts with a distributional effect.

To summarise these trends more formally, we estimate a hierarchical Bayesian ‘ideal point’ model, adapted from the standard tools used to study judicial ideology (Martin and Quinn 2002, Bafumi et al. 2005). In our model, a justice’s latent position represents their propensity to vote pro-rich rather than pro-poor. Crucially, this latent scale has an explicit economic interpretation, anchored in observed redistribution rather than in an abstract ideology.

The results reinforce the raw data. In the early decades of the sample, the estimated positions of Democratic- and Republican-appointed justices are statistically indistinguishable. Over time, Republican appointees drift steadily in a more pro-rich direction, while Democratic appointees move only modestly – and not always significantly – in the opposite direction. By 2022, the implied gap between a ‘typical’ Republican- and Democratic-appointed justice reaches nearly 50 percentage points in the probability of voting pro-rich on a neutral case.

Why the median justice matters

Polarisation alone does not determine outcomes. Supreme Court decisions depend on majorities, and on a nine-justice Court, the pivotal actor is the median justice – the one whose vote determines which side commands a majority (Poole and Rosenthal 1991).

When we track the position of the median justice over time, the distributional implications become clearer. Since the late 1960s, shifts in the Court’s composition have repeatedly moved the median in a pro-rich direction. Three moments are particularly consequential: the early Nixon appointments around 1969–70, the replacement of two liberal justices in 1990–91, and the 2018 appointment that replaced a relatively moderate median with a significantly more pro-rich one.

Because many economically salient cases are decided by narrow margins, these changes in the median justice have real consequences. Small shifts in Court composition translate into large jumps in the median pro-rich justice, and therefore in how often the Court shifts resources from poor to rich in disputes.

Implications for policy debates

How does our research contribute to current debates over judicial reform and economic policy? First, it suggests that economic distribution may be a useful predictor of how the Court will decide a case. Any given mode of constitutional interpretation, even if a justice claims to consistently follow it, does not yield predictable results. Instead, an explicit economic metric can illuminate for voters the Court’s pro-wealthy trend and how future cases are likely to be decided.

Second, the findings underscore how appointment politics, in particular the strategy of Republicans to appoint pro-rich justices, shape economic policy over decades. There has been increasing alignment between the appointing party and distributional outcomes, meaning polarisation has steadily risen. Each Republican appointment moving the Court in a pro-rich direction has enduring implications for the distribution of resources, and perhaps for economic inequality in the US.

Finally, the analysis helps clarify why the Supreme Court should be a focal point for reform. Debates over the Supreme Court’s ethics and ability to preserve democratic government are no longer only critical constitutional questions; they are central to the political economy of redistribution. By providing a transparent, outcome-based measure of the Court’s pro-wealthy bias, we offer a framework for holding judicial power to account – one grounded not in ideology, but in who gains and who loses.

References

Bafumi, J, A Gelman, D K Park, and N Kaplan (2005), “Practical issues in implementing and understanding Bayesian ideal point estimation”, Political Analysis 13: 171–87.

Cohen, A (2021), Supreme inequality: The Supreme Court’s fifty-year battle for a more unjust America, New York: Penguin.

Economic Policy Institute (2004), “Stock ownership and wealth distribution”, State of Working America 2004/2005 fact sheet, Economic Policy Institute.

Martin, A D, and M K Quinn (2002), “Dynamic ideal point estimation via Markov chain Monte Carlo for the US Supreme Court, 1953–1999”, Political Analysis 10: 134–53.

Poole, K T, and H Rosenthal (1991), “Patterns of Congressional voting”, American Journal of Political Science 35: 228–78.

Prat, A, F Scott Morton, and J Spitz (2026), “Ruling for the rich: The Supreme Court over time”, NBER Working Paper 34643.

Segal, J A, and D A Cover (1989), “Ideological values and the votes of US Supreme Court justices”, American Political Science Review 83: 557–65.