On cert decision-making

The process by which the US Supreme Court selects cases for review reflects a dynamic interplay between judicial preferences, litigant strategies, and institutional norms.Footnote ¹ The literature of judicial behavior converges on three core frameworks for explaining cert decisions: cue theory, error correction, and jurisprudential development. These frameworks are not mutually exclusive. Rather, they offer complementary lenses on how justices, operating through a two-tiered process that starts with the clerks’ screening and ends with the Justices’ conference (Perry Reference Perry1991), sort through a large, litigant-driven petition pool (Mak, Sidman, and Sommer Reference Mak, Sidman and Sommer2013).

Cue theory treats certiorari as a high-volume screening problem. Justices and clerks rely on informational signals that help identify petitions likely to be “certworthy,” given limited attention and docket capacity (Black and Boyd Reference Black and Boyd2013; Caldeira and Wright Reference Caldeira and Wright1988; Caldeira, Wright, and Zorn Reference Caldeira, Wright and Christopher1999). These cues operate beyond purely attitudinal strategy and include both “low-information” and “high-information” indicators (Black and Boyd Reference Black and Boyd2013). Low-information cues include familiar markers of inconsistencies in the law or importance such as an inter-circuit conflict (“circuit split”), dissents below, reversals of the court of first instance, and institutional signals generated by the courts of appeals themselves—most notably en banc decisions, which are independently and positively associated with Supreme Court review even after accounting for classic cert variables (George and Solimine Reference George and Solimine2001).

Additional work also treats dissents as whistleblowing that draws hierarchical attention (Beim, Hirsch, and Kastellec Reference Beim, Hirsch and Kastellec2014). The ideological configuration of the lower court relative to the Supreme Court may itself operate as a cue, with review more likely when the panel’s ideological composition is in tension with the Court’s (Black and Owens Reference Black and Owens2012; Hall Reference Hall2009). Such cues economize on screening time by signaling importance or doctrinal inconsistencies without requiring full review on the merits.

High-information cues center on the Solicitor General (SG), whose recommendations and participation operate as unusually credible indicators of legal and policy consequence (Bailey, Kamoie, and Maltzman Reference Bailey, Kamoie and Maltzman2005; Black and Owens Reference Black and Owens2011)—a dynamic arguably amplified by the contemporary use of the shadow docket (Vladeck Reference Vladeck2019, Reference Vladeck2023). Amicus support on the petitioner’s side likewise communicates broad interest and downstream stakes, and is consistently associated with higher grant probabilities (e.g., Caldeira and Wright Reference Caldeira and Wright1988; Caldeira, Wright, and Zorn Reference Caldeira, Wright and Christopher1999; Collins, Corely, and Hamner Reference Collins, Corely and Hamner2014; Hansford and Johnson Reference Hansford and Johnson2014; Abi-Hassan, Box-Stefensmeier, and Christenson Reference Abi-Hassan, Box-Stefensmeier and Christenson2025). Additional signals, such as elite counsel and the presence of former Supreme Court clerks on briefs, reinforce perceptions of vehicle quality and certworthiness (Liu and Kastellec Reference Liu and Kastellec2023; Feldman and Kappner Reference Feldman and Kappner2017). Petitions that raise novel or broadly important constitutional questions are also more likely to make the discuss list and advance to conference, consistent with the Court’s preference for cases that can produce generally applicable guidance (Black and Boyd Reference Black and Boyd2013). Cue theory thus connects the form of the petition (e.g., clear questions presented) to the function of docket selection. Collectively, this suggests that cues do not replace legal judgment; instead they prioritize attention and certify quality. In practice, cues often interact with the other frameworks by helping discover where error correction is needed and locating strong vehicles for jurisprudential development.

Error-correction-based scholarship presents certiorari as means of hierarchical control. The Court deploys its limited plenary space to police legal error and resolve inconsistencies in the law, while moving doctrine toward the preferences of the median justice (Grant, Hendrickson, and Lynch Reference Grant, Hendrickson and Lynch2012; Cameron, Segal, and Songer Reference Cameron, Segal and Songer2000). Inter-circuit conflict and visible disagreement between lower courts (e.g., dissents, en banc activity) are robust predictors of grants because they signal the presence of conflict in national law, requiring the Court’s intervention to restore uniformity. The Court acts as a selective auditor, focusing on cases with significant doctrinal or ideological implications (Cameron, Segal, and Songer Reference Cameron, Segal and Songer2000), and the broader theoretical treatment emphasizes that discretionary jurisdiction “privileges error-correction” given capacity constraints (Clark and Kastellec Reference Clark and Kastellec2013). Consistent with strategic auditing, grants are more frequent where lower-courts are ideologically distant from the Supreme Court median (Lindquist, Haire, and Songer Reference Lindquist, Haire and Songer2007; Black and Owens Reference Black and Owens2012; Bryan and Owens Reference Bryan and Owens2017).

All else equal, greater ideological distance between a lower court’s disposition and the Supreme Court’s preferences increases the likelihood of review (Cameron, Segal, and Songer Reference Cameron, Segal and Songer2000). However, the authors point out that when that distance is extreme or the factual position of the lower court is a poor vehicle for resolving conflict, the informational value of review may decline. In such instances, the Court may bypass ideologically distant cases in favor of those offering clearer doctrinal leverage. Critically, litigant strategy shapes what the Court is able to audit—losing litigants are more likely to file cert petitions when the lower court case outcome and the Supreme Court are more closely aligned, resulting in a nonrandom pool of cases that reflects perceived ideological alignment (Yates, Cann, and Boyea Reference Yates, Cann and Boyea2013; Hinkle Reference Hinkle2024). This contributes to higher rates of ideological voting among appealed cases (Songer, Segal, and Cameron Reference Songer, Segal and Cameron1994; Songer, Cameron, and Segal Reference Songer, Cameron and Segal1995). Building directly on this monitoring logic, Bonica, Chilton, and Sen (Reference Bonica, Chilton and Sen2025) introduce the “odd party out” theory, which adds a party-configuration signal of potential error. The Court is more likely to grant cert when the petitioner is ideologically distant from both the respondent and the panel that ruled against them, a configuration that plausibly flags bias or misapplication by an aligned lower court. Empirically, the authors’ measure predicts grants even after accounting for standard indicators such as dissent, en banc, and SG involvement, functioning as a cue that enables error-correction by helping the Court target outlier or potentially skewed outcomes (Bonica, Chilton, and Sen Reference Bonica, Chilton and Sen2025).

Jurisprudential development perspectives foreground the justices’ interest in building, clarifying, and extending doctrine. Certworthiness depends on a petition’s capacity to produce clear, generally applicable law—clean records, preserved questions, and narrow, dispositive issues increase the expected payoff of doctrinal clarification (Grant, Hendrickson, and Lynch Reference Grant, Hendrickson and Lynch2012; Cross Reference Cross2003; Segal and Spaeth Reference Segal and Spaeth2002). In this register, Black and Owens (Reference Black and Owens2009) show that at the cert stage justices weigh expected policy payoffs against jurisprudential criteria such as legal importance and conflict, and that when those collide, jurisprudential considerations constrain—or even override—pure policy maximization, whereas alignment between policy and law “liberates” justices to pursue their preferred outcomes. Related scholarship casts the lower courts as sites of doctrinal learning and experimentation that the Supreme Court can leverage when selecting cases (Beim Reference Beim2017; Clark and Kastellec Reference Clark and Kastellec2013), a perspective consistent with classic accounts of agenda-setting (Perry Reference Perry1991).

The plenary docket often reflects an alignment between the justices’ jurisprudential priorities and litigants’ strategic case construction, with sophisticated parties crafting vehicles responsive to perceived Court signals (Baird Reference Baird2007). Status asymmetries shape access to this doctrinal agenda. Notably, high-status petitioners—such as the federal government represented by the SG—enjoy structural advantages in both securing a grant and shaping the Court’s agenda (Black and Boyd Reference Black and Boyd2012). Empirical work shows justices frequently follow SG recommendations even when those recommendations cut against ideological priors or case-specific legal factors, reflecting the SG’s institutional credibility and repeat-player status (Black and Owens Reference Black and Owens2011). Amicus participation can amplify or mitigate these advantages by signaling the breadth and configuration of interests implicated (Black and Boyd Reference Black and Boyd2012; Collins, Corely, and Hamner Reference Collins, Corely and Hamner2014; Hansford and Johnson Reference Hansford and Johnson2014; Abi-Hassan, Box-Stefensmeier, and Christenson Reference Abi-Hassan, Box-Stefensmeier and Christenson2025).

In practice, these three frameworks—cue theory, error correction, and jurisprudential considerations—operate jointly and through a petition pool that is neither random nor static. Petitioners deploy dissents below and amici to elevate perceived importance (Mak, Sidman, and Sommer Reference Mak, Sidman and Sommer2013), but structural advantages—resources, repeat-player status, elite counsel—systematically mediate success (Yates, Cann, and Boyea Reference Yates, Cann and Boyea2013; Black and Boyd Reference Black and Boyd2012). Recent scholarship suggests filing behavior itself shifts with Court ideology, with individuals and organizations representing marginalized groups filing relatively fewer petitions, and business and conservative interests increasing filings and amicus participation during more conservative periods (Widner and Gunderson Reference Widner and Gunderson2024). Ignoring selection into filing risks biased inferences about grant determinants; modeling the decision to file alongside the decision to grant is essential to recover the mechanisms driving agenda-setting.

In a parallel body of research in law and economics, Eisenberg (Reference Eisenberg1990) offers a foundational theoretical and empirical account of how the observable universe of litigated cases is systematically shaped by a “selection effect”—litigation outcomes are endogenously filtered through the incentives and constraints of litigants, creating selection bias in the cases courts ultimately decide. This logic is echoed in Blume and Eisenber’s (Reference Blume and Eisenberg1999) study of death penalty appeals, which shows that prosecutorial discretion, adjudicator behavior, and state-level institutional practices jointly determine which capital cases are reviewed, with profound implications for understanding variance in reversal rates and judicial responsiveness. Building on this foundation, Chang and Hubbard (Reference Chang and Hubbard2018) use trial-level data to demonstrate how the structure of legal incentives affects not just case outcomes, but the underlying decision to litigate or settle. Hylton (Reference Hylton2002) further advances this logic through a formal model of litigation under asymmetric information, demonstrating that when defendants possess superior knowledge of their own liability, selection effects emerge that not only skew win rates but shape the very pool of cases that proceed to trial.

Taken together, these works underscore the notion that the certiorari docket does not merely reflect the legal importance or ideological salience of petitions—it is the outcome of a multilayered filtering process structured by institutional incentives, regional discretion, asymmetric information, and actor-specific strategies at the point of filing. The filtering process distorts the inferences one can make from litigated outcomes alone because the causes that survive to higher levels of adjudication are not representative of the broader legal and factual universe. Integrating these insights enriches the agenda-setting framework by anchoring it in broader theories of case selection, informational dynamics, and legal stratification.

Justice Douglas’s (Reference Douglas1960, p.413–14) admonition that understanding certiorari demands “the economist’s understanding, the poet’s insight, the executive’s experience, the political scientist’s understanding, and the historian’s perspective” underscores the breadth of forces at work in the process of reviewing and granting cert. Historically, the Court’s campaign for greater discretion, culminating in the Judiciary Act of 1925, was justified in part by a promise of equitable geographic representation across jurisdictions (Cordray and Cordray Reference Cordray and Cordray2004; Perry Reference Perry1991). Law is not created in a purely conceptual vacuum but in legal space—the origin of disputes, the regional legal cultures that generate them, and the institutional pathways they travel all shape the menu of questions available to the Court (Economides, Blacksell, and Watkins Reference Economides, Blacksell and Watkins1986; Pue Reference Pue1990). Existing work links region and issue mix—such as the South’s disproportionate role in Eighth Amendment litigation, the Northeast’s prominence in search-and-seizure, and the West’s in environmental disputes (Brunn et al. Reference Brunn, Shelley, Webster and Ahmed2000)—and shows non-proportional geographic contributions on issues like abortion and redistricting (Stoler and Roman Reference Stoler and Roman2016). Recognizing this spatial heterogeneity and the petition supply that reflects it clarifies why some cues arise more frequently than others, why certain conflicts ripen sooner, and why some vehicles prove especially attractive for doctrinal development. Placed in this broader ecosystem, the Court’s discretionary power is best understood as the product of screening signals, hierarchical control, and jurisprudential ambition operating within, and conditioned by, the strategic behavior of litigants and the spatial-temporal distribution of legal questions.

Building the legal landscape

To better understand the dynamics of agenda-setting judicial behavior, we must examine the supply of petitions from which the Supreme Court chooses cases for review on the merits. Comprehensive data of this nature—spanning geography, time, and issue area—have largely eluded the discipline. We therefore begin by assembling a dataset encompassing all petitions filed with the US Supreme Court between 1946 and 2019. Using the Supreme Court Journals (SCOTUS 2020), we compiled identification data for 313,441 petitions. Employing an error detection protocol within the Optical Character Recognition (OCR) process,Footnote ² we meticulously extracted the petitioner and respondent names and status (federal government, state/local government, business/organization, individual), case origins, and the status of each petition—granted or denied—with a 95% confidence level.Footnote ³ The resulting dataset categorizes all petitions according to geographic origin, source court, and petitioner type, juxtaposed with the number of cases granted review. This also includes an identification of paid versus in forma pauperis (non-paid) cases.

For the plenary docket data, we extract from the US Supreme Court Database (SCDB) (Spaeth et al. Reference Spaeth, Epstein, Martin, Segal, Ruger and Benesh2022) case identification variables, including state of origin and case disposition. The SCDB contains case-related attributes such as dates, issue area, disposition of the case, winning party, and justice-related variables such as justices’ votes, opinion writing, and inter-agreements between justices. Thereafter, we merge our original cert dataset with the selected covariates from the US Supreme Court database (Spaeth et al. Reference Spaeth, Epstein, Martin, Segal, Ruger and Benesh2022). The final product is a dataset containing all petitions filed between 1946 and 2019, sorted by court of origin, contrasted with the number of cases on the docket, and the percentage granted.

Spatial-temporal patterns in the law

The evolution of the federal court system has been primarily driven by the aim of solving the wide variety of disputes arising from both federal and state legal systems to maintain a consistent interpretation of the law across the country. Although the justices exercise control over their docket, this control is institutionally limited to cases that are appealed. In other words, the Supreme Court depends on losing litigants to bring cases before it (Mak, Sidman, and Sommer Reference Mak, Sidman and Sommer2013). As a result, the decision-making process for granting certiorari fundamentally relies on the decision to appeal, making the pool of cert petitions neither random nor static. This invites inquiry into how the supply of petitions has changed over time and where the most salient volumes of cert petitions originate from? More importantly, how does the influx of petitions compare to the grant rate across different sources of origin over time?

Over the past 75 years, the Court’s docket has expanded substantially, driven largely by the growing scope of the federal government and the establishment of federally enforceable individual rights. This expansion is evident in the increase in certiorari petitions from approximately 1,300 in 1950 to over 5,000 in the 2000s. In contrast, the number of cases granted review has fluctuated from roughly 100 in the 1950–60s, to 150 in the 1970–80s, and down to about 80 cases per year following the 1988 jurisdictional reforms—amounting to roughly 1% of the Court’s docket. This divergence underscores the Court’s highly selective decision-making process, where grants are the result of a filtering process that begins with clerks privileging cases that score high on objective certworthiness, and credible proximate cues (Benesh, Armstrong, and Wallander Reference Benesh, Armstrong and Wallander2020). This is further compounded by the justice’s own strategic and legal considerations (Black and Boyd Reference Black and Boyd2012). Given the statistical rarity of formal review, it is essential to focus not only on the cases that receive full consideration, but also on the broader pool of petitions available to the Court.

The stark disparity between the total number of cert petitions filed and the cases granted review over time is evident in Figure 1. Starting in 1946, the number of petitions begins to increase at a constant rate until the mid-1980s, with a sharper increase up until 2005, when petitions begin to drop. In contrast, the number of cases granted review remains relatively steady until 1988 and then drops to about 80–100 per year. This drop coincides with the removal of the Court’s mandatory appeals jurisdiction in the Judiciary Act of 1988, which converted the rulings in state court on the validity of federal statutes or treaties and appeals from circuit courts to discretionary review by certiorari, reducing the number of cases the Court was required to hear to practically zero.

Figure 1.

Total Cases & Petitions, 1946–2019.

Note: Shading categorizes the year variable into courts by Chief Justice: Fred M. Vinson (1946-1953), Earl Warren (1953-1969), Warren E. Burger (1969-1986), William Rehnquist (1986-2005), John Roberts (2005-incumbent).

Figure 1. Long description

The x-axis represents years from 1946 to 2020 in five-year increments. The y-axis represents the Total Number from 0 to 8000 in increments of 500. Background shading denotes five eras: Fred M. Vinson (1946 to 1953), Earl Warren (1953 to 1969), Warren E. Burger (1969 to 1986), William Rehnquist (1986 to 2005), and John Roberts (2005 to 2019).

* Cases on the Docket: Represented by a solid orange line. This data remains nearly flat and stable near the bottom of the graph, fluctuating slightly between 0 and 250 cases throughout the entire period.

* Writ of Cert Petitions: Represented by a dashed blue line. This data shows a significant upward trend. It begins around 1200 in 1946, rises steadily through the Warren and Burger eras to reach approximately 4000 by 1986. During the Rehnquist era, it climbs sharply to a peak of nearly 7700 around 2005. In the Roberts era, the trend reverses, declining to approximately 5500 by 2019.

In addition to the expansion of federal authority and the creation of new individual rights—which we explore further in the paper—the dramatic growth in the Supreme Court’s docket has also been linked to the rise of the Supreme Court bar. Beginning in the mid-1980s, a modern, elite bar emerged, composed of former Solicitors General, deputies, and clerks who joined major law firms and brought sophisticated expertise in packaging and presenting petitions. As Lazarus (Reference Lazarus2008) documents, private firms—most notably Sidley’s recruitment of former SG Rex Lee in 1985—deliberately built elite SCOTUS practices, even as the Court’s merits docket contracted. This development was reinforced by broader transformations in corporate legal strategy during the same period, including the growth of in-house legal departments and increased demand for specialized Supreme Court advocacy. Together, these changes sustained a high volume of petitions and contributed to a supply-side professionalization that made sophisticated parties more likely to file cert petitions—and to frame them in ways that enhance their certworthiness (Nelson and Epstein Reference Nelson and Epstein2022).

The expansion in the Supreme Court bar’s membership coincides with the turnover of the Court’s leadership to Chief Justice Rehnquist along with the appointment of Antonin Scalia as associate justice in the same year. Both favored a limited role for the Court in resolving societal disputes and strongly believed that not all lower court conflicts presented questions salient enough to merit its attention (Merrill Reference Merrill2003). This makes the 1980s an institutional and ideological inflection point in the Court’s operation, which changed both the docket size and litigants’ expectations about payoffs to filing (Lazarus Reference Lazarus2008).

Another notable change in the relation between cert filings and the Court’s plenary docket is the drop in the number of petitions beginning in the 2006 term. This coincides with the beginning of the Roberts’ Court, which was expected to accentuate the conservative shift in the Court’s ideological composition that had started during Chief Justice Rehnquist’s tenure. Coupled with an emphasis on hearing more business cases, and their exceptional success starting in the 2006 term,Footnote ⁴ this drop in petitions may be the outcome of liberal advocates being discouraged from pursuing litigation in areas of individual liberties, including litigants filing in forma pauperis, as evidenced by the empirical findings of Widner and Gunderson (Reference Widner and Gunderson2024). The downward trend in petitions starting in 2005 is also consistent with Baird’s (Reference Baird2007) argument that politically salient Supreme Court decisions operate as signals to litigants and interest groups about the Court’s policy priorities; strategic actors respond by generating and framing additional, higher-quality litigation in those specific policy areas. Policy-specific waves of filings are predictable supply-side responses to politically salient Supreme Court decisions. Such decisions signal policy priorities to litigants and interest groups, who then generate and frame additional cases in those domains.Footnote ⁵ We would expect to encounter the opposite effect if the Court’s salient cases signal to litigants that some policy areas are less favored by the Court. Changes in the Court’s docket can also be linked to the emergence of information technology, which facilitated the communication and monitoring between circuit courts to reduce the possibility of legal conflicts (Narechania Reference Narechania2022).

As the primary contributors to the Supreme Court’s plenary docket, fluctuations in the volume and characteristics of rulings in lower federal and state jurisdictions impact the Court’s case composition, even though it does not directly impact the rate at which the Court grants review. As shown in Panel (a) of Figure 2, approximately two-thirds of cert petitions originate from the federal courts of appeal, with the remainder distributed between state courts and other sources.Footnote ⁶ Panel (b) of Figure 2 illustrates that, in accordance with observed trends, the bulk of the cases on the Court’s plenary docket are sourced from federal courts, followed by state courts and other sources. However, discernible temporal variations are evident. From 1970 to 1985 federal petitions were granted at a disproportionate rate in comparison to cases originating from state courts. Starting in the 1980s, the Court markedly reduced its plenary docket regardless of fluctuations in the volume of petitions received, for all three jurisdictions alike.

Figure 2.

Petitions & Cases Categorized by Source, 1946–2019.

Note: Shading categorizes the year variable into courts by Chief Justice: Fred M. Vinson (1946–1953), Earl Warren (1953–1969), Warren E. Burger (1969–1986), William H. Rehnquist (1986–2005), John Roberts (2005–incumbent).

Figure 2. Long description

A two-panel line graph. Both panels share a horizontal x-axis representing years from 1946 to 2019, with vertical shaded regions indicating the tenures of Chief Justices Vinson, Warren, Burger, Rehnquist, and Roberts.

Panel a, titled Writ of Cert Petitions, has a y-axis for Total Number ranging from 0 to 7000.

- Federal (solid red line) shows a significant upward trend, starting near 500 in 1946, rising sharply during the Rehnquist era to a peak of approximately 6000 around 2005, then declining toward 4000 by 2019.

- State (dashed blue line) shows a gradual increase from near 500 to a plateau of approximately 1500 starting in the 1990s.

- Other (dotted green line) remains consistently low, near zero, with a slight bump to 300 around 2010.

Panel b, titled Granted Cases, has a y-axis for Total Number ranging from 0 to 150.

- Federal (solid red line) fluctuates between 70 and 100 until the mid-1970s, peaks at 130 during the Burger era, and then shows a steady decline through the Rehnquist and Roberts eras to approximately 55 by 2019.

- State (dashed blue line) starts at 55, drops to 30, peaks again at 50 in the late 1960s, and then steadily declines to approximately 10 by 2019.

- Other (dotted green line) remains flat near zero throughout the entire period.

A legend at the bottom indicates: solid red line for Federal, dashed blue line for State, and dotted green line for Other.

Table 1 summarizes the total number of petitions and cases per court of origin at the federal and state levels. Between 1946 and 2019, the average grant rate for all cases was 2.85%; however, there are significant differences between jurisdictions.Footnote ⁷ There is variation in the number of cert petitions filed by the different jurisdictions—going from 1,427 petitions appealed from US District Courts to 35,651 and 38,727 petitions coming from the Ninth Circuit Court of Appeals, and State Appellate courts, respectively. This is contrasted by grant rates that range between 0.75% for state trial courts and 55.47% for cases coming from U.S. District Courts.Footnote ⁸ As the Court grants so few cases in comparison to the number petitioned, even a fraction of a percentage difference can represent dozens of cases. To explore this level of variation at a more granular level, in the following sections we analyze the over-time variation for petitions filed and cases granted among different jurisdictions, specifically: federal lower courts and state courts.

Table 1.

Descriptive Statistics, 1946–2019

Table 1. Long description

The table consists of four columns: Source Type, Petitions, Cases, and percentage Granted.

Total Federal: 228,035 petitions, 6,674 cases, 2.93% granted.

- U S Court of Appeals, First Circuit: 6,166 petitions, 180 cases, 2.92% granted.

- U S Court of Appeals, Second Circuit: 17,780 petitions, 637 cases, 3.58% granted.

- U S Court of Appeals, Third Circuit: 16,060 petitions, 407 cases, 2.53% granted.

- U S Court of Appeals, Fourth Circuit: 21,992 petitions, 377 cases, 1.71% granted.

- U S Court of Appeals, Fifth Circuit: 33,383 petitions, 684 cases, 2.05% granted.

- U S Court of Appeals, Sixth Circuit: 21,164 petitions, 496 cases, 2.34% granted.

- U S Court of Appeals, Seventh Circuit: 14,916 petitions, 476 cases, 3.19% granted.

- U S Court of Appeals, Eighth Circuit: 14,041 petitions, 346 cases, 2.46% granted.

- U S Court of Appeals, Ninth Circuit: 35,651 petitions, 1,068 cases, 3.00% granted.

- U S Court of Appeals, Tenth Circuit: 11,694 petitions, 277 cases, 2.37% granted.

- U S Court of Appeals, Eleventh Circuit: 21,085 petitions, 224 cases, 1.06% granted.

- U S Court of Appeals, D C Circuit: 7,175 petitions, 491 cases, 6.84% granted.

- U S Court of Appeals, Federal Circuit: 3,502 petitions, 110 cases, 3.14% granted.

- U S Special Courts: 2,000 petitions, 110 cases, 5.50% granted.

- U S District Courts: 1,426 petitions, 791 cases, 55.47% granted.

Total State: 79,203 petitions, 2,055 cases, 2.59% granted.

- State Supreme Courts: 36,101 petitions, 1,693 cases, 4.69% granted.

- State Appellate Courts: 38,727 petitions, 329 cases, 0.85% granted.

- State Trial Courts: 4,375 petitions, 33 cases, 0.75% granted.

Other: 6,203 petitions, 201 cases, 3.24% granted.

Grand Total: 313,441 petitions, 8,930 cases, 2.85% granted.

Note: There are 100 cases from the 1940s to 1950s originating in US District Courts that also fall under a State court jurisdiction; in the maps these cases have been included in the State Courts count. The “Other” category contains petitions that were not clearly labeled in the US Supreme Court journals, as well as Original Jurisdiction cases.

Federal lower courts

Trends in the volume of cert petitions are primarily driven by cases coming from the lower federal courts. In general, shifts correspond with the expansion of the US courts of appeal dockets, going from 30,000 to 60,000 annual cases between 1985 and 2005, before stabilizing at approximately 45,000 cases per year thereafter (Lazarus Reference Lazarus2008). Providing a more detailed level of analysis, Figure 3 shows the breakdown of petitions and cases granted for the US circuit courts of appeal. Although most circuit courts show an increase in the number of cert petitions, it is most significant, in order, for the Ninth, Fifth, Eleventh, Fourth, and Sixth Circuits (see Panel (a) in Figure 3). More importantly, there is great variation between the number of petitions coming from each circuit and the rates at which petitions are granted over time. Below we synthesize explanations for why the rate of certiorari petitions from different US Courts of Appeal varies over time considering the overall supply of “cert-worthy” questions a circuit produces, institutional shocks that differently affect circuits, and the demand-side that changes litigants’ incentives to petition.

Figure 3.

US Courts of Appeals, 1946–2019

Note: Shading categorizes the year variable into courts by Chief justice: Fred M. Vinson (1946–1953), Earl Warren (1953–1969), Warren E. Burger (1969–1986), William H. Rehnquist (1986-2005), John Roberts (2005–incumbent).

Figure 3. Long description

Two side-by-side line graphs with a shared legend at the bottom. The X-axis for both represents years from 1946 to 2020 in five-year increments. Vertical shaded bands indicate the tenures of Chief Justices: Fred M. Vinson, Earl Warren, Warren E. Burger, William H. Rehnquist, and John Roberts.

Panel A, titled Writ of Cert Petitions:

* The Y-axis measures Total Number from 0 to 1100.

* Most circuits show a gradual increase from 1946 until the mid-1980s, followed by a sharp rise during the Rehnquist era (1986-2005).

* The Ninth Circuit (teal dashed line) and Fifth Circuit (olive dashed line) show the highest volume, peaking near 1000 petitions around 2005 before declining.

* The D C Circuit (solid red line) remains relatively flat and low compared to others.

Panel B, titled Granted Cases:

* The Y-axis measures Total Number from 0 to 25.

* Trends are more volatile with lower values than Panel A.

* Most circuits peaked between 1970 and 1990, followed by a general downward trend through the Roberts era.

* The Ninth Circuit consistently maintains one of the highest counts of granted cases, though it drops from a peak of nearly 20 in the early 2000s to approximately 11 by 2019.

The legend identifies twelve circuits: First (green dotted), Second (light blue dashed), Third (pink dashed), Fourth (teal dash-dot), Fifth (olive dashed), Sixth (purple solid), Seventh (blue dotted), Eighth (brown dashed), Ninth (teal long-dash), Tenth (magenta solid), Eleventh (gold dashed), and D C (red solid).

Circuits differ, persistently, in both volume and composition of appeals. Statistics for the federal judiciary document large and shifting circuit caseloads where, unsurprisingly, larger pipelines (e.g., the Ninth and Fifth) tend to produce more potential petitions, while specialized pipelines (e.g., the D.C. Circuit’s administrative docket; the Federal Circuit’s patent docket starting in 1982) generate issues the Court regularly polices. Hence, we would expect higher per-decision pressure when a circuit’s docket generally produces cases in nationally consequential areas of the law (e.g., admin law in D.C.; securities in the Second; patents in the Federal Circuit).Footnote ⁹ For instance, the Ninth Circuit Court of Appeals is known to handle a large proportion of cases dealing with immigration, environmental, and First Amendment questions (Feldman Reference Feldman2025). In the early 2000s, the DOJ’s Board of Immigration Reform was associated with a surge of immigration cases in the Second and Ninth Circuits, sharply altering those circuits’ dockets, which is consistent with the trend observed in Panel (a) of Figure 3. In addition, the Fifth Circuit Court of Appeals Reorganization Act of 1980 redistributed appeals between the Fifth and the new Eleventh Circuit Court of Appeals. In our series, this appears as a step-down in the Fifth Circuit’s petition share and a corresponding step-up in the number of petitions in the Eleventh Circuit beginning in the 1981–82 term. Moreover, as the Court’s cert policy emphasizes resolving splits, circuits that more frequently generate or attract conflicts will produce more petitions. Hellman (Reference Hellman1978) emphasizes how conflict production varies by circuit—for instance, the Ninth Circuit’s sheer size and unique limited en banc procedure can foster inconsistency and sustain splits longer, which is known to increase incentives to petition and potentially the rate at which they are granted as evidenced by Panel (b) in Figure 3.

Agenda-setting accounts show that cert is more likely when the Court perceives lower-court decisions as ideologically adverse or doctrinally problematic. When a circuit’s median is far from the Supreme Court, losing parties (and the SG) have stronger reasons to petition, expecting a receptive Court (e.g. Black and Owens Reference Black and Owens2011; Caldeira and Wright Reference Caldeira and Wright1988; Bustos and Jacobi Reference Bustos and Jacobi2019; Cameron, Segal, and Songer Reference Cameron, Segal and Songer2000). In periods when, for example, the Ninth Circuit was relatively liberal vis-à-vis the Burger/Rehnquist Courts, you often see more petitions driven by policy disagreement and conflict correction. This is also consistent with the observed surge in cases granted review for the Ninth Circuit beginning in mid-1970s, as shown in Panel (b) of Figure 3.Footnote ¹⁰ In addition, the modern Supreme Court strategically is more likely to select cases with high-status petitioner and advocates, as well as those supported by amicus briefs (Collins Reference Collins2008; Feldman and Kappner Reference Feldman and Kappner2017). A professionalized pipeline of cases increases both the quality and salience of petitions from certain circuits, raising grant probability and upstream filing incentives, especially for those cases that are attractive to the Supreme Court bar (e.g., D.C. administrative law cases, Federal Circuit patents, 2nd Circuit post-2008).

Beyond the imposition that jurisdictions have on where cases can be appealed, litigants from across the political spectrum are also known to engage in venue shopping when filing suits in district and appellate courts in search of the most advantageous pool of judges (Atkinson, Marco, and Turner Reference Atkinson, Marco and Turner2009). For instance, liberal litigants have generally preferred the Ninth Circuit Court of Appeals, while conservatives are more likely to file suits in the Fifth Circuit Court of Appeals. This variation in petitions unsurprisingly contributes to the shift in the stream of petitions, as well as the Court’s grants mediated by the ideological distance between the Supreme Court and lower courts (Cameron, Segal, and Songer Reference Cameron, Segal and Songer2000).

State courts

In the post-War period, the Court maintained a comparatively small merits docket and received a limited number of state criminal judgments each year, setting the baseline against which later expansions are judged. Beginning in the 1950s, the rights revolution drives up the number of cert petitions originating from state courts (Epp Reference Epp1998). The Warren Court’s (1953–1969) expansion of criminal procedure protections (e.g., Mapp, Gideon, Miranda) produced a surge of direct review from state courts, particularly for criminal defendants. This is evident in the surge observed through the 1950s to 1960s—for both appellate and state supreme courts—in Panel (a) of Figure 4. Contemporary and retrospective overviews emphasize that the Warren Court dramatically expanded the federally protected constitutional rights of state criminal defendants, which necessarily channeled a large share of the Court’s attention to state-court decisions (Pye Reference Pye1968).

Figure 4.

State Courts, 1946–2019.

Note: Shading categorizes the year variable into courts by Chief Justice: Fred M. Vinson (1946–1953), Earl Warren (1953-1969), Warren E. Burger (1969-1986), William H. Rehnquist (1986-2005), John Roberts (2005–incumbent).

Figure 4. Long description

A two-panel line graph. A shared legend at the bottom indicates a green dotted line for Supreme Court, a red solid line for Appellate, and a blue dashed line for Trial. Vertical shaded regions represent the tenures of Chief Justices Vinson, Warren, Burger, Rehnquist, and Roberts.

Panel a, Writ of Cert Petitions. The X axis shows years from 1946 to 2020. The Y axis shows Total Number from 0 to 1100. The Appellate line (red) shows a steady increase, peaking near 1000 around 2005 before a slight decline. The Supreme Court line (green) fluctuates, peaking near 550 in the late 1960s and 650 in the late 1990s. The Trial line (blue) remains relatively flat and low, staying below 100.

Panel b, Granted Cases. The X axis shows years from 1946 to 2020. The Y axis shows Total Number from 0 to 40. The Supreme Court line (green) starts high at 35, fluctuates between 20 and 33 until the late 1980s, then drops sharply to approximately 10 by 1995, remaining stable thereafter. The Appellate line (red) shows two distinct peaks near 10 in the late 1960s and early 1980s, then drops to near zero after 1990. The Trial line (blue) remains consistently near zero throughout the period.

In the early 1970s, the conscious expansion of the Court’s docket, undertaken in part to preserve room for civil rights cases while also addressing other federal law areas, helped sustain substantial review of state decisions in the early years of the Burger Court (Hellman Reference Hellman1978). Despite a slight reduction in state filing mid-1970s, which could be attributed to Justice Brennan’s Reference Brennan1977 call for “New Judicial Federalism,”Footnote ¹¹ the Court adopted the rule, where absent a clear indication that a state decision rests on adequate and independent state grounds, the US Supreme Court presumes a federal basis and may review. This doctrinal move re-opened review channels for some state-court decisions that were ambiguous, producing countervailing pressure toward more grants from state courts in the mid-1980s (Hellman Reference Hellman1978).

As Figure 4 shows, the 1980s witnessed a sharp increase in petitions from state appellate courts, even though the overwhelming majority of state-origin cases granted review continued to come from state supreme courts (Panel (b)). This divergence became especially pronounced as the Supreme Court’s plenary docket began to contract after 1986 while cert petitions from both state appellate and state supreme courts kept rising. A central reason was the Judiciary Act of 1988, which effectively eliminated most of the Court’s remaining mandatory appellate jurisdiction and converted many state-level federal-question appeals into discretionary petitions for certiorari. This jurisdictional shift increased the number of state-origin petitions entering the cert pool while simultaneously contributing to the sharp decline in state supreme court cases on the plenary docket beginning in the early 1990s, as shown in Panel (b) of Figure 4.

Beginning in the late 1990 to early 2000s, court observers have noted—as shown in Figure 4—that petitions from state appellate and supreme courts begin to experience a modest cool down and eventually a drop. Associated with this decrease is likely a marginal reduction in the average number of state supreme court decisions producing full judicial opinions between 1995 and 2010 (Hall and Windett Reference Hall and Windett2013), a factor that significantly influences the probability of certiorari petitions being filed and granted (Black and Boyd Reference Black and Boyd2013).

Similar to the grant rates among the US circuit courts of appeal, there is some variation in the grant rate per individual state, as shown in Figure 5. The rate of cert petitions granted ranges between 1% and 7%, excluding US territories. In the upper bound, we find states including South Dakota, Vermont, Montana, Utah, New Hampshire, and Idaho, which have a case-to-petition-ratio twice as much as most other states. This suggests that in cases coming from state courts, beyond these few states with high rates, the Court’s plenary docket is better at proportionally representing the various disputes across the country (see correlation between petitions filed and granted by state origin in Appendix C). However, the slight divergence we observe may also be an instance of the type of cases most likely to emerge in each state in different eras.

Figure 5.

Percentage of Cert Petition to Case by State, 1946–2019.

Figure 5. Long description

A choropleth map of the U S A with a color scale legend on the right ranging from 2 (light blue) to 7 (dark blue). Each state is labeled with its two-letter initialism and a numerical value.

* Legend: Values range from 2 to 7. S D (South Dakota) has the highest value at 7.1, shown in the darkest blue.

* Northwest: W A 3.5, O R 3.6, I D 4.6, M T 4.7, W Y 2.3.

* Southwest: C A 2.1, N V 1.5, U T 5.1, A Z 3.4, N M 3.7.

* Midwest and Plains: N D 3.1, S D 7.1, N E 3.2, K S 3.8, O K 2.8, T X 1.9, M N 2.3, I A 3.7, M O 3.2, A R 3.7, L A 4.2, W I 3.9, I L 1.9, M I 2.6, I N 2.2, O H 2.9, K Y 3.

* Southeast: T N 2.2, M S 3.3, A L 4.5, G A 3.3, F L 2.2, S C 4.7, N C 3.1, V A 2.2, W V 3.3.

* Northeast: P A 1.8, M D 1.3, D E 3, D C 1.8, N J 2.2, N Y 3.9, C T 2.6, R I 5.4, M A 2.8, V T 5.5, N H 5, M E 2.6.

* Insets: A K (Alaska) 3.6 in the bottom left; H I (Hawaii) 2 in the bottom center.

Another noticeable trend is the drop or leveling off in the volume of cert petitions during transition periods between chief justices. While further inquiry is needed, it may be that legal activists go through a cooling period until they can ascertain the type of Court each chief justice will preside over before moving forward with petitions.Footnote ¹² To explore this variation, we expand our review to a dynamic overtime analysis of cert petitions originating from state courts. Figure 6 shows that the trends observed in the aggregate view hold, with some caveats. The states that have a high case-to-petition ratio vary over time. South Dakota, Montana, and the District of Columbia have grant rates up to 80% during the Vinson Court, and about 25% for South Dakota, North Dakota, and the block of Southern states during the Warren Court, with similar rates for South Dakota in addition to Utah and Vermont during the Burger Court. Although we see a drop in the overall grant rate beginning in the 1980s, we observe grant rates of 5% for Arkansas, West Virginia, Maine, and Montana during the Rehnquist Court, and Vermont, Kansas, and Montana during the Roberts Court.

Figure 6.

States Ratio of Cases to Petitions by Court.

Figure 6. Long description

The figure contains five panels, each displaying a map of the U S A with states shaded in blue to represent the ratio of cases to petitions. Darker blue indicates a higher ratio.

* Top-left: Vinson Court, 1946 to 1953. The scale ranges from 0 to 80. S D shows the highest ratio, appearing as a dark blue outlier compared to the rest of the country which is mostly light blue.

* Top-right: Warren Court, 1953 to 1969. The scale ranges from 0 to 25. Higher ratios are concentrated in the Southeast, specifically G A, A L, and L A, as well as S D in the North.

* Middle-left: Burger Court, 1969 to 1986. The scale ranges from 0 to 20. S D remains a high-ratio state. Darker shading is also visible in V T and U T.

* Middle-right: Rehnquist Court, 1986 to 2005. The scale ranges from 0 to 5. Higher ratios are more dispersed, with notable concentrations in the Northwest (I D, M T), the South (A R, M S), and the Mid-Atlantic (W V).

* Bottom-center: Roberts Court, 2005 to 2020. The scale ranges from 0 to 6. The highest ratios are seen in M T and K S, with moderate ratios in I D and V T.

Across all eras, the maximum ratio value on the legend decreases significantly over time, from 80 during the Vinson Court to 6 during the Roberts Court.

While it is difficult to narrow the precise mechanism driving these patterns, given case- and justice-specific factors shaping cert grants, we believe that at a macro-level, a combination of geography-based issue areas and time is at play. For instance, some of the cases coming from Montana and South Dakota in the earlier years of our data involve federalism and tribal questions, including Baldwin v. Fish & Game Commission of Montana, 436 US 371 (1978). During the Warren Court, a concentration of cases from southern states is consistent with a focus on solving questions on criminal procedure, and the high grant rate for in forma pauperis (IFP) petitioners. It is evident that the supply of petitions is not constant and potentially speaks to strategic behavior by legal activists or the justices themselves during the writ of cert process.

Bringing the federal and state origin stories together, the throughline is that institutional source shapes the menu of cert-worthy questions and the Court’s willingness to spend scarce plenary slots, while those same forces condition the incentives of would-be petitioners. The next section shifts our focus from where cases originate to who is doing the filing—how organizational capacity and repeat-player advantages (OSG, state SGs, the elite Supreme Court bar), alongside resource constraints and access regimes (IFP), translate these federal-versus-state institutional landscapes into observable filing volumes and grant probabilities.

The petitioning party

Petitioning parties play an important role in shaping the US Supreme Court’s docket. In order to reflect the general status of those filing cert petitions, the litigants are sorted into four categories, ordered by their status, institutional capacity, and access to resources, following the classification by Songer, Sheehan, and Haire (Reference Songer, Sheehan and Haire1999). From strongest to weakest, we start with the federal government at the top, followed by state/local governments, businesses/organizations, and individuals last. As shown in Figure 7, there is wide variation in the level of participation between the different types of petitioners seeking review.

Figure 7.

Percentage of Petitioner Type from Total Cert Petitions Filed, 1946-2019.

Note: Shading categorizes the year variable into courts by Chief Justice: Fred M. Vinson (1946-1953), Earl Warren (1953-1969), Warren E. Burger (1969-1986), William H. Rehnquist (1986-2005), John Roberts (2005-incumbent).

Figure 7. Long description

The x-axis represents years from 1946 to 2019 in five-year increments. The y-axis represents the Percentage of Petitions from 0 percent to 75 percent and above. Vertical shaded regions delineate the tenures of Chief Justices Fred M. Vinson, Earl Warren, Warren E. Burger, William H. Rehnquist, and John Roberts.

* Individual: Represented by a pink dash-dot line. It starts at approximately 50 percent in 1946 and shows a steady upward trend, peaking near 90 percent around 2011 before a slight dip toward 2019.

* Company or Organization: Represented by a blue dotted line. It begins at approximately 33 percent in 1946 and shows a consistent long-term decline, ending below 10 percent by 2019.

* Government: Represented by a solid green line. It starts at approximately 12 percent in 1946 and declines steadily, flattening out near 0 percent from the mid-1980s onward.

* State: Represented by an orange dashed line. It remains the lowest and most stable category, fluctuating slightly between 0 percent and 5 percent throughout the entire period.

The vast majority of cert petitions are filed by individuals, which exhibit a constant upward trend through the studied period. The first substantial increase takes place between the mid-1940s and mid-1960s, with a drop during the Burger Court (1969-1986), followed by another significant increase between 1986 and 2006, with another drop beginning around 2010. Notably, these shifts perfectly align with the leadership of chief justices, suggesting that the rate of cert petitions filed by individuals could be shaped by perceptions of how the Court is likely to treat different issues. As shown in Figure 8, petitions by individuals maintain a high grant rate of approximately 2.5%, with slight declines towards the mid-1960s, and remain low during the Burger Court and the first half of Rehnquist’s Court, with a sharp increase in the second half, up until the beginning of the Roberts Court, where the grant rate begins to drop again. This association between court and the behavior of litigants may be most true during the Warren Court, given its known receptiveness towards individual petitioners, specifically those filed in forma pauperis (IFP), which make up the majority of individual petitions (Widner and Gunderson Reference Widner and Gunderson2024, p. 35).

Figure 8.

Percentage of Cert Grant per Petitioner Type, 1946–2019.

Note: Shading categorizes the year variable into courts by Chief Justice: Fred M. Vinson (1946-1953), Earl Warren (1953-1969), Warren E. Burger (1969-1986), William H. Rehnquist (1986-2005), John Roberts (2005-incumbent).

Figure 8. Long description

The x-axis represents years from 1946 to 2019 in five-year increments. The y-axis represents the Percentage of Grants from 0 percent to 25 percent. Vertical shaded regions delineate the tenures of Chief Justices Vinson, Warren, Burger, Rehnquist, and Roberts.

Four petitioner types are tracked:

* Government (solid green line): Starts near 3 percent, peaks at 8 percent in the early 1950s, drops to 2 percent in the mid-1960s, then rises sharply to a global peak of 20 percent around 1986. It gradually declines to approximately 11 percent by 2019.

* State (dashed orange line): Follows a similar but lower trajectory to the Government line, peaking at nearly 14 percent in the early 1980s before declining to roughly 4 percent by 2019.

* Company or Organization (dotted blue line): Remains relatively stable and low, fluctuating between 1 percent and 5 percent throughout the entire period.

* Individual (dash-dot pink line): The lowest category, consistently staying between 0 percent and 3 percent, with a slight dip in the late 1990s.

As a baseline, the Court’s IFP docket is large and distinct. For example, the Court’s own year-end report shows that in the 2005 term the Court received 8,521 filings, 6,846 of which were IFPs (approximately 80% of all filings). This is consistent with the long-run pattern of IFP shares, as shown in Figure 9. Empirical work and Court commentary consistently show that the IFP docket is heavily composed of criminal, habeas, and prisoner civil-rights matters and that the Court treats IFP matters as a recognizably separate stream (Watson Reference Watson2006). Fluctuations in the rate of these petitions can also be associated with expanding statutory access for indigent litigants at the turn of the 20th century, followed by several landmark Supreme Court decisions that lowered the barriers for filing IFP and pro se petitions.Footnote ¹³

Figure 9.

Share of Paid vs. in forma pauperis Cert Petitions, 1946–2019.

Figure 9. Long description

A stacked area chart where the x-axis represents the Year from 1946 to 2019 and the y-axis represents the Share of total petitions from 0% to 100%. The chart is divided into two colored regions: a blue area at the top representing Paid petitions and a grey area at the bottom representing In forma pauperis petitions.

* From 1946 to approximately 1970, In forma pauperis petitions dominate the share, consistently staying above 85%.

* Between 1971 and 1981, there is a sharp decline in the share of In forma pauperis petitions, dropping to a low of approximately 55% as the blue Paid petition area expands downward.

* From 1981 to 1995, the share fluctuates significantly, with a notable spike in In forma pauperis petitions around 1995 reaching nearly 90% before dropping back down.

* From 1996 to 2016, the In forma pauperis share stabilizes between 75% and 85%.

* In the final years leading to 2019, there is a sharp final decline in the In forma pauperis share, ending at approximately 60%.

Following the Warren Court expansion, doctrinal cutback slowed the rate of habeas-driven cert petitions by individuals, consistent with the dip in Figure 7 between 1971 and 1981. During this period, through various decisions, the Court also restricted federal habeas review by excluding many Fourth Amendment claims, tightening procedural default (Stone v. Powell 1976), and further limiting relief on collateral review (Wainwright v. Sykes 1977). This doctrinal retrenchment towards habeas petitions is likely associated with the decrease in petitions by individuals, as well as grant rates, during the above-mentioned period.

Other factors may also contribute to the change in the rate of individual petitions, including broader access to legal resources for low-income individuals, increased public awareness of the ability to file cert petitions, and potentially, shifts in the types of cases stemming from lower courts. Despite the establishment of structural or policy limits on indigent and pro se petitioners, the incarceration boom of the late 1970s into the 1990s could explain the second surge in petitions by individuals. A 700% rise in the prison population between 1972 and 2009—with a modest decline thereafter (Ghandnoosh Reference Ghandnoosh2020), coincides with large run-ups in prisoner petitions at earlier adjudicatory stages (district and circuit courts). This constitutes a much bigger base of potential litigants, holding other costs constant, that would raise the volume of IFP petitions.

By volume of petitions, those filed by businesses/organizations follow the individual category (see Figure 7). In contrast to individuals, petitions by businesses are marked by a sharp decline during the Warren Court, followed by a more steady decline over the studied period. In the 1940s, companies and organizations make up approximately 20% of the petitions filed, then declining to about 12% by the mid-1960s. The volume of petitions by these entities remains steady throughout the 1970s and 1980s, followed by another decline to a current rate of about 5%.

This downward trajectory reflects the Warren Court’s agenda displacement. Empirical work on the Court’s plenary docket shows that across the 1960s, civil-rights/civil-liberties questions steadily displaced other categories of federal law on the merits docket, which size remained unchanged for most of the period. As those issues crowded out “general federal law” (which includes much business and regulatory litigation), expected grant rates for business petitions plausibly fell, rationally depressing corporate filings at the cert stage (Hellman Reference Hellman1978), consistent with grant rates for businesses/organizations in Figure 8. In part, driven by this crowd-out effect, the Burger Court expanded the plenary docket in 1971 to almost double that of the merits cases, which increased the grant rates for these entities from 1% to almost 5%. This would create space for non-civil-rights cases, which should raise expected grants for business litigants and, in turn, increase corporate cert filings.

The creation of the US Court of Appeals for the Federal Circuit in 1982 centralized patent appeals, leading to fewer inter-circuit conflicts—one of the most reliable cert cues (Black and Owens Reference Black and Owens2012). This, in addition to the removal of the Court’s mandatory appeals jurisdiction in 1988, partially explains the decline in business petitions beginning in the 1980s. Narechania’s (Reference Narechania2022) “patent puzzle” shows how the lack of field splits reshaped cert dynamics in patents. In addition, driven in part by this change in the Court’s jurisdiction, the justices decided far fewer cases per term beginning in the second half of the Rehnquist Court (Owens and Simon Reference Owens and Simon2012). A smaller merits capacity filters upstream to lower expected success and therefore fewer business petitions at the margin. Lastly, from 2008 onward, securities class actions surged around the financial crisis, increasing the pipeline of appellate securities disputes from which cert petitions could arise. Moreover, Lazarus (Reference Lazarus2008) documents how a small cadre of expert advocates, with the expansion of the Supreme Court bar—disproportionately representing corporate clients—achieved higher grant rates and actively shaped the petition pipeline, especially from the mid-2000s forward. This kind of institutional change increases the quality and volume of business cert petitions.Footnote ¹⁴

The volume of petitions by state governments and the US government follows a similar trend, with varying proportions overtime. The same Warren-era docket displacement (toward civil rights/liberties) reduced the expected payoff for many regulatory or business petitions that the states or US might otherwise have filed. At the same time, the OSG’s traditional role includes self-restrained petitioning—the office files only in a small fraction of government losses, prioritizing institutional interests and credibility—so when expected grants fall, the OSG reduces filings rather than “spray and pray” (Hellman Reference Hellman1978). Moreover, until 1988, several federal government matters reached the Court by appeal of right, reducing the need to petition in key categories.

When the Court increases plenary throughput, states and the OSG can justify more petitions (and more recommendations to grant in private cases) because grant chances improve at the margin. This is consistent with our observed slight increase in petition volume for states and the US government in Figure 7, and the sharp increase in grants, going from approximately 3% for both in 1969 to an all-time high of 13% for states in 1982, and 20% for the US in 1986. As the Court began to reduce its docket, the OSG reduced its own cert requests by more than 70%, filing in the order of about 15 petitions per term by the 2000s (Cordray and Cordray Reference Cordray and Cordray2010). A similar pattern is observed for state filings and grants—classic supply-side self-selection in response to falling expected returns.

The justices have also become more reliant on the Call for the Views of the Solicitor General (CVSG). Empirical work shows that a cert petition is much more likely to be granted after a CVSG—the Court agrees with the SG in more than half of the cases (Thompson and Wachtell Reference Thompson and Wachtell2009). This channel allows the United States to steer the agenda without always filing as petitioner, reducing the raw number of OSG petitions even as its influence rises. Post-2000s, the Court has continued to follow the OSG’s cert-stage recommendations at striking rates (overall and especially in patent cases), reinforcing the incentive for the United States to engage more assertively where it sees national uniformity stakes (Gugliuzza and Koivula Reference Gugliuzza and Koivula2023). This is consistent with the stability in cert filings by state and US governments observed in our data.

The decline in the US government’s role as a petitioner in more recent decades corresponds to an increased selectivity in pursuing certiorari and a shift in the Court’s preferences, such as favoring cases with broader doctrinal significance over those narrowly impacting government policy. This interpretation aligns with broader themes in judicial behavior and litigation strategy, suggesting that the changing landscape of Supreme Court litigation affects institutional actors like states and the federal government, as much as private litigants.

Taken together, the patterns in Figures 7 and 8 make clear that who petitions the Court is a first-order determinant of the petition stream the justices confront. Individuals—disproportionately filing in forma pauperis—supply the vast majority of petitions and display pronounced era-linked surges and retrenchments, while organizational and governmental filers contribute more selectively in ways that track shifting institutional incentives and perceived payoffs. These petitioner-specific regularities do not merely diversify the docket; they structure the selection environment the Court can audit, with volume, capacity, and status advantages jointly shaping both baseline access and observed grant propensities. Conceptually, then, any account of agenda-setting must model the decision to file alongside the decision to grant, treating shocks to the filing environment (doctrine, jurisdiction, and professionalization) as causal, not incidental, features of cert dynamics.

The supply-side framework

This research advances our understanding of the US Supreme Court’s agenda-setting process by emphasizing the critical, yet often overlooked, role of the supply of cert petitions in shaping the Court’s docket. Our findings show that the Supreme Court’s agenda is not simply the product of the justices’ preferences applied to a fixed set of disputes. Rather, it emerges from a dynamic petition environment that varies across geography, time, institutional source, and petitioner identity.

By assembling the complete universe of certiorari petitions since 1946 and linking them to merits-stage attributes, we uncover a system in which the Court’s choices are continually conditioned by where disputes arise, who brings them, and when structural shocks alter the cost-benefit calculus of filing. Three empirical regularities emerge. First, the origin of the petitions matters. The uneven distribution of petitions between federal courts of appeal and state courts is mirrored—though not mechanically—in the merits docket, with notable shifts across different eras. Second, petitioner status matters. While individuals account for the majority of filings, they receive the lowest grant rates. In contrast, institutional actors such as the federal and state governments enjoy significantly higher success rates, reflecting structural advantages in resources, expertise, and credibility—though they move with era-specific shocks. Third, capacity and professionalization matter. The sharp increase in output from lower courts coupled with the rise of a specialized Supreme Court bar has intensified selection pressures, even as the Court’s merits docket has contracted.

These patterns yield clear theoretical implications. Models that focus exclusively on demand-side cues, including legal importance, conflict, or ideology, risk conflating the Court’s preferences with the availability of vehicles that exhibit those cues. The petition pool is neither random nor static; it is shaped by heterogeneous regional dockets (e.g., immigration, environmental, or federalism clusters), evolving jurisdictional and administrative regimes, and the professionalization of Supreme Court advocacy. Recognizing that heterogeneity—across federal and state pipelines, as well as petitioner types—clarifies how the structure of the petition pool conditions which issues reach the Court, when conflicts mature, and how doctrinal questions surface for review.

Overall, this research provides a foundational framework for understanding the supply-side dynamics of the Supreme Court’s certiorari process. Substantively, the supply-side lens reframes several familiar debates. Apparent “under-selection” of state-origin petitions, for example, reflects not only justices’ preferences but also doctrines and drafting practices that shape reviewability and steer many state disputes to summary correction rather than argument. Similarly, fluctuations in federal-origin grants track changes in lower-court pipelines and case types (not just shifts in cert policy). In short, the Court’s discretionary power is exercised within an uneven petition environment, not above it.

In sum, our findings suggest that to understand what the Supreme Court decides, we must first understand what the Supreme Court chooses from. Treating shocks to the filing environment—across regions, institutions, and petitioner types—as causal inputs rather than background noise yields a more complete and accurate account of agenda-setting in the United States Supreme Court. Methodologically, our contribution is an infrastructure for studying agenda-setting as a two-stage process: the decision to file and the decision to grant. This dataset creates a foundation for designs that: (i) treat filing as endogenous to doctrine, geography, and organizational capacity, and (ii) separate composition effects in the pool from selection effects at conference. We expect future research to build on this work by exploring the interactions between supply-side patterns and specific issue areas, the evolving role of legal advocacy, and the potential implications for access to justice and legal equity in the highest court of the United States.