We’ve all seen it on television: the well-dressed lawyer rocketing from her chair with an adroitly timed “Objection!” that turns the tide of the entire case. The jury gasps, the judge sustains, and the client wins. The attorney winks at the camera as the credits roll. But this is not the realty for many Americans. Combined, the state and federal court systems hear approximately 67 million cases per year. Nearly 50% of litigants will face these cases without legal representation.
Despite the critical role legal information plays in protecting rights, access to legal assistance remains severely limited. A study conducted by the Legal Services Corporation’s Justice Gap found that 92% of the civil legal problems experienced by low-income Americans receive little to no legal help. Law libraries increasingly serve as one of the few public-facing resources designed to help individuals better engage with the legal system. This does not apply purely to unrepresented litigants. Anyone seeking help in the face of an active lawsuit can, and often will, turn to a law library.
This begs the question: how can law libraries better curate their collections to meet actual patron need? We discovered that case filing data offers an answer. It proposes a preemptive approach to help law libraries equip patrons to resolve the most trying circumstances of their lives by foregrounding collections that statistically meet patron demands before they walk in the door. So, with the preliminaries out of the way, what’s all this about data?
We encountered a significant difficulty early on in our analysis. After much research, we were unable to locate sufficiently granular state court data suited for our purpose. We ultimately relied on public information regarding federal court filings. The difference between the two court systems and their disparate case types is acknowledged. However, we found that federal court data allowed actionable recommendations not readily emulated by the lack of state data. That said, total reliance on our recommendations for state matters is neither appropriate nor accurate given the data’s federal nature.
We also acknowledge that our data’s generalized nature does not perfectly capture every patron need. Some seek assistance for issues beyond the most commonly filed matters in their district. Law librarians should be mindful of the potential harms that may arise by backgrounding materials outside the averages. The recommendations that follow should only be used to enhance collections, not diminish them.
One may be inclined to think, “Surely the Government keeps a steady thumb on the pulse of yearly case filings?” One would be correct, but in spirit only. The most exhaustive studies to this end have been conducted by the United States Court System. These reports, however, are limited in scope. They present (1) general conclusions about case percentages across identified metrics and (2) charts identifying the number and type of cases filed. Though useful for general percentages, the reports do not record the exact number and type of case in a manner useful for making recommendations to individual law libraries. Other public and cost-based databases, such as Court Listener and the Statistical Abstract of the United States, also tender no conclusions about patron needs. In short, while there is no scarcity of case filing data, there is far less data granularity allowing specific recommendations.
We then discovered TRAC. Nearly a quarter of a century ago, the Transactional Records Access Clearinghouse (“TRAC”) was launched at Syracuse University. Its mission was to “provide the American people and the institutions of oversight with the authoritative information they require to independently judge the actual performance of the federal government.” Over two decades, TRAC employed Freedom of Information Act requests supported by constant litigation to gather internal Government data files concerning federal cases. Through an involved process, TRAC transformed the specific files into usable research cataloging federal civil pending and closed cases and federal convictions across the ninety-four federal districts in the United States. It conducted extensive reliability testing and developed the data into databases updated annually until June 2023.
The civil dataset may be found at: https://tracreports.org/tracfed/judges/interp/judgelist.html?tracdecor=1. It consists of 938 observations organized among four columns (Judge, District, Cases Pending, Cases Closed). The criminal dataset may be found at: https://tracreports.org/judges/aboutCivil.html. It consists of 845 observations organized among three columns (Judge, District, Convictions). Both datasets are webpage tables that may be filtered with a Case Type tool.
In order to convert both tables into a master dataset in manipulable R format, some legwork was required. Because we did not have access to the raw data, it was necessary to create separate datasets for each of the seventeen fields of civil law and twelve of criminal law. We copied the entire dataset for each field of law into Excel, which we then imported into R.
Once imported into R, we created a new column labeled “Case Type” in each of the twenty-nine data frames to capture the field of law for each case. We then aggregated all civil and criminal data frames into a master dataset and added a “Law Type” column specifying which data entries were civil verses criminal. The final dataset consisted of 16,985 observations across seven variables indicating every type of case per judge per federal district. Through it, we wrangled the data to identify case trends across states and federal districts.
The final dataset that we used can be found here:
https://raw.githubusercontent.com/Graysonscott116/cases_dataset/refs/heads/main/all_cases.csv
As with any legal enterprise, ethics is vital. The reader should bear in mind several ethical considerations and limitations with our data. It is important that TRAC’s work relied nearly exclusively on support from foundation grants, gifts, and donations of legal service and time. Apart from institutional support provided by Syracuse University, TRAC preserved its independence by rejecting Government funding. In preserving its neutrality, TRAC made two significant choices regarding the collection and categorization of the data. First, by making the affirmative choice to seek data directly from primary sources via FOIA requests, internal judicial researchers, and litigation, TRAC did not rely on secondary sources lacking authority. Second, it made interpretative choices in resolving data discrepancies by using an internal usability procedure.
While we had to rely on TRAC’s evaluations given our inability to verify TRAC’s work, its dedication to these datasets ameliorates verification concerns. Finally, we found it significant that TRAC’s databases were born from the Government’s reticence to supply similar data. Federal reticence implies a desire to keep such data private, possibly due to a concern of making public judicial inefficiencies or because the story it tells about charging decisions. Nonetheless, TRAC’s dedication in gathering the data gave us confidence that its data collection procedures were unbiased.
Regarding limitations: First, the data is constrained by its twelve-month time period, which some may find challenges the accuracy of modern-day conclusions. Case dockets are not prone to drastic change over a three-year span, as can be seen in the United States reports. Second, while it is not uncommon for judges to rotate dockets or retire over a three-year period, judges inherit their predecessors’ dockets, including their condition. It is acknowledged that some of these dockets may have improved since June 2023, but such improvements given the U.S. Reports are de minimis.
Our data is legal in nature and designed to be used by law libraries. Because of this, anyone engaging with it will need some familiarity with the law. Federal districts are abbreviated for ease of reading, and our recommendations assume a level of legal knowledge. We chose TRAC’s datasets with this firmly in mind. We are both lawyers, and our recommendations are ultimately directed at law libraries. Our data should serve as a guidebook, not a gospel. In this blog post, we walk through how to synthesize it for specific states, districts, and regions.
To investigate cases on a per state basis, we totaled the sum of all civil and criminal cases and organized each district under its respective state. Plotting these totals surfaced important conclusions for whether law libraries should prioritize civil or criminal federal materials. Civil totals were calculated as TotalClosed cases, while criminal convictions were calculated as Total Convictions. Civil and criminal cases are represented by two graphs for universal readability. States may be easily compared via the interactive legend.
Figure 1. Distribution of Cases by State
As Figure 1 illustrates, criminal cases far outweigh civil matters in several large jurisdictions: Texas, California, Arizona, and New Mexico. In Texas, for example, civil matters (18,192) are far fewer than criminal matters (79,030). From this, it can be concluded that Texas is prosecuting criminal offenses far more than closing civil cases, implying a more active criminal justice system. This analysis reveals that law libraries in Texas should foreground resources addressing federal crimes rather than civil law.
Figure 1 conversely offers advisements for other high-filed, high-population states. In Florida during the 2022-2023 period, 81,555 civil cases were closed verses 12,934 criminal convictions. The data suggests that Florida law libraries should prioritize federal civil legal reference materials. While we will not examine every state here, Figure 1 provides general state overviews allowing analogous conclusions for any state.
Legal data is often toothless without specificity. Our computational analysis therefore included totaling the sum of the types of cases per state to investigate the prevailing field of law among identified state metrics. The heatmap below, when accounted for severe outliers based on state population, allows a measure of prophylactic analysis.
Figure 2. Heatmap of Cases by State
Figure 2 confirms our conclusions drawn from Figure 1. The area of law with the most cases in Texas, indicated by the highlighter yellow color, is Immigration with 55,787 convictions. Immigration represents 70.6% of the total criminal convictions from 2022-2023. Narcotics & Drugs is the second highest field of law with 12,027 convictions. A clear recommendation emerges that law libraries in Texas should ensure their immigration and narcotics legal references materials are robust and updated.
Modern context supports this recommendation. According to the United States Government, filings in the Southern and Western District of Texas for 2025 rose 21% and 19%, respectively, with notable increases in incarceration rates in defendants charged with regulatory, drug, and immigration offenses. These increases are shared for the remaining boarder states. The modern political milieu likely explains some of this disparity and demonstrates that our data holds for 2025.
Indeed, our data confirms the importance of immigration reference materials, which have become more important with the expansion of Immigration and Customs Enforcement (ICE). According to one recent account, ICE agents detained Irvin González, a transgender woman and undocumented immigrant, while she in court seeking a protective order against an abusive ex-boyfriend. Several media outlets reported that her ex-boyfriend called immigration authorities on her. These stories have only increased in the last few years, providing anecdotal impetus to our data’s recommendations.
A similar analysis can be conducted for civil cases in Texas. Surprisingly, Contracts arises as the most common civil case. Unsurprising, Civil Rights cases closely follow. This is to be expected in a state with such a high number of criminal convictions. Texas displays the highest number of Civil Rights cases other than California, which shares a criminal conviction overbalance when compared to civil cases. Law libraries should recognize that Civil Rights claims are a companion to criminal convictions and, as demonstrated by Figures 1 and 2, proportionally rise with criminal prosecution. It is thus critical that law libraries in states with higher criminal convictions like Texas and California do not neglect legal resources covering companion topics.
We next converted Figure 2 into a percentage heatmap in order to homogenize de minimis case types and quickly surface actionable information.
Figure 3. Heatmap of Percentages of Cases by State
Figure 3 places conclusions regarding Figure 2 into relief. In Texas, for example, Immigration is 57.4% of all cases and in New Mexico 72.4%. Of note, it further emerges that the high civil-criminal disparity in Florida is due to 68.2% of all cases being Personal Injury – Products Liability. The specificity here is critical. Products liability is a unique form of negligence that implicates different legal rules than traditional personal injury cases. Based on this difference and their predominance, Florida law libraries should confirm they are adequately surfacing products liability materials for their patrons. Anecdotal reports concerning the number of personal injury lawyers in Florida confirm these recommendations. According to one law firm, there are 10,000 personal injury attorneys in Florida.
Figures 1-3’s conclusions can be reframed in the context of their broader regions. To investigate whether granular per-state recommendations are captured by regional trends, we organized the data into four regions: Northeast, South, North Central, and West. We then calculated the total case type per region as a proportionally stacked bar chart for easy computational comparison.
Figure 4. Distribution of Cases by Region
Regional data reveals distinctions in the legal makeup per region. Immigration as predicted arises as a leading case type in the South and West regions. Conclusions where immigrants may reside who need immigration resources can be made. According to Figure 4, product liability claims are disproportionately filed in the South. However, reading Figure 4 in concert with Figure 3, we can attribute this to Florida. When Florida is filtered out, Immigration stands as the top case type.
The unlabeled column represents American territories such as Puerto Rico and the Virgin Islands. Its balanced percentages show a relatively even delineation and support recommendations of a wide legal resource collection for U.S. territories.
Civil Rights cases are a higher percentage of the overall region in the Northeast, while Narcotics & Drugs and Weapons charges lead in the North Central. Based on Figure 4, state trends are confirmed by regional differences when accounting for the larger states. On the whole, law libraries located within these regions should be aware of the general trends and investigate whether their particular state shares them.
This project has at its center federal data. Its utility would be greatly diminished should we neglect to calculate the most common case types per federal district. This allows law libraries to match the data with their specific district in curating their collections. Using a similar heatmap as Figure 3, we plotted each federal district’s most commonly filed case types.
Figure 5. Distribution of Cases by District
It confirms our previous conclusions and recommendations, but with even more granularity. Rather than merely “Texas,” we now see that Immigration cases are highest in the Southern and Western Districts of Texas, the Northern District of Mexico, and the Southern District of California. Using California as an example, Immigration cases in its Southern District constitute 39.5% of cases while in its Northern District they plummet to 0.8%. A focus on immigration cases in law libraries in the Northern District would not be as efficacious as the same focus in the Southern. Should law libraries desire to strike a fine balance in their collection based on internal data of where patrons reside, Figure 4 serves a more accurate representation of federal cases on a per district basis.
It further allows some universal conclusions across districts. Personal Injury-Products Liability emerges as one of only three common case types for most districts. They do not include Immigration. Products liability cases are filed as higher percentages of their home district’s total than most other case types. Given that federal law applies identically across all federal districts, law libraries, even those in non-products liability dominated districts, would be well-served by housing operative products liability reference materials. The universality of certain trial cases across districts implicates the civil-criminal split for appellate cases.
The districts of the federal court system make up circuits, where federal appellate courts lie. An examination of civil-criminal case breakdown provides clues to the accuracy of our analysis. It further shows the drastic difference between circuits, of which law libraries should be aware. To calculate this, we organized each district into its respective circuit and plotted them alongside one another for easy comparison.
Figure 5. Distribution of Cases by Circuit
We can make several fascinating conclusions. First, civil cases overshadow criminal cases in ten out of the twelve circuits. One measure of a law library’s success is preparing patrons for appeal as much as trial. Because the case numbers in 83% of circuits show majority civil law representation, so too will their appeals. Particularly in the 11th Circuit courts, where civil cases outpace criminal cases by a factor of sixteen, law libraries should be aware that appellate courts will examine civil matters far more than their criminal counterparts. Thus, collections prioritizing appellate law should favor civil law in all but the 5th and 10th Circuits.
Second, because no one circuit is dominated by criminal law at the expense of civil law, we cannot make broader recommendations for which law libraries in which circuits should increase criminal law representation in their collections. Instead, law libraries in circuits where the numbers are close should rely on state and district data. As an example, we will conduct this analysis for a state of our choosing, Washington State.
We filtered and visualized the case type breakdown for Washington State to show how our data may be used to make more specific conclusions for law libraries.
Figure 6. Washington State
Figure 6 reveals that Washington is a rather balanced state. Narcotics & Drugs is the highest case type, with Immigration, Contracts, Civil Rights, and Prisoner Petition close behind. Its more uniform nature when compared to other states answers the question of how Washington law libraries should structure their federal collections. They should be balanced, with a slight favoritism toward Immigration and its related civil causes of action. Washington’s balance, compared to states like Texas and Florida, shows that law library collections should not be uniform across the country. Our computational analysis uncovers the importance of accounting for more granular differences in patron need.
Qualitative, report-based studies have ascertained the same. How law is accessed has radically grown over time. Much modern-day argument centers “rightsizing,” or making thoughtful decisions about aligning law library collections with institution mission and community need (see Watson, 2023).
Our analysis of TRAC’s data leads to one conclusion: law libraries should rely on granular data in curating their collections. The problems facing pro se litigants vary. The types of problems they face depending on the state or district even more so. Reliance on general data, like those reported by the United States Government, have limited practical utility. The plots shown here demonstrate that general conclusions about case filings cannot be made in the face of state and district individuality. Law libraries should adjust their collections to meet the federal needs shown by the cases actually filed in their neighborhoods. Guidance on how to do so should not rely on “feeling out” patron needs. Instead, it should be found in an often-overlooked part of law: data.
. . .
Free.Law, Court Listener (2026), https://www.courtlistener.com/opinion/; United States of America, Statistical Abstract of the United States, 131st ed. (1878-2012), https://heinonline-org.offcampus.lib.washington.edu/HOL/Index?index=usfed/statabs&collection=usfed.
Judge Information Center, Criminal Cases in District Court (2023), TRACREPORTS, https://tracreports.org/tracfed/judges/interp/judgelist.html?tracdecor=1.
Judge Information Center, Criminal Cases in District Court (2023), TRACREPORTS, https://tracreports.org/tracfed/judges/interp/judgelist.html?tracdecor=1.
Legal Services Corporation. (2022). The Justice Gap: The unmet civil legal needs of low-income Americans. https://justicegap.lsc.gov.
Perkins Law Offices, Navigating the Complexities of Personal Injury Claims in Florida: A Guide for You (last accessed June 5, 2026), https://www.perkinslawoffices.com/how-many-personal-injury-lawyers-are-there-in-florida/.
Stephen Rushin, Federal Enforcement of Police Reform, 82 Fordham L. Rev. 3189 (2014), https://ir.lawnet.fordham.edu/cgi/viewcontent.cgi?article=5011&context=flr.
Michael Morris & Lauren Renee Sepulveda, TDCAA, A New ICE Age
(2017), https://www.tdcaa.com/journal/a-new-ice-age/.
The Pew Charitable Trusts, How Many Cases—and What Kind—Do State and
Local Courts Handle? (Mar. 6, 2025), https://www.pew.org/en/research-and-analysis/data-visualizations/2025/03/how-many-cases-and-what-kind-do-state-and-local-courts-handle.
United States Courts, Federal Judicial Caseload Statistics (March 31, 2025), https://www.uscourts.gov/data-news/reports/statistical-reports/federal-judicial-caseload-statistics/federal-judicial-caseload-statistics-2025.
United States District Court for the District of Kansas, What Is the Difference Between U.S. (Federal) District Courts and State Courts?, https://ksd.uscourts.gov/content/what-difference-between-us-federal-district-courts-and-state-courts (last visited June 5, 2026).
Watson, Amanda, Beyond Titles: Curating Relevant Law Library Collections (March 1, 2023). Law Library Journal, Forthcoming, Yale Law School Legal Information Symposium – The Legal Treatise: Past, Present, and Future, U of Houston Law Center No. 2023-A-15, Available at SSRN: https://ssrn.com/abstract=4376450.