US Capital Punishment (1999): A Curated Dataset of Judicial Executions for Criminology and Data Science

Description

This curated dataset provides a comprehensive and high-fidelity record of the judicial executions carried out in the United States during the year 1999, which represents the historical peak of capital punishment activity in the modern era.

Unlike generic data lists, this resource is the result of a rigorous process of digital curation and technical validation. It captures the complete universe of 98 records, documenting key variables such as demographic profiles (age, gender, ethnicity), legal jurisdiction (state), and the specific methodology of the execution, along with a detailed classification of the underlying criminal offenses.

A core value of this collection is its technical and structural integrity. The dataset has been engineered to meet the highest standards of data science: it features a verified ASCII encoding (1.0 confidence) and follows the RFC 4180 standard for handling complex text fields (such as multiple criminal charges). This ensures seamless interoperability across different analytical environments, preventing common parsing errors in professional workflows.

The purpose of this resource is to serve as a foundational "Gold Standard" table for criminologists, legal researchers, and data scientists. It provides a clean, verified, and multi-platform foundation (fully compatible with R, Python, and Excel) for any longitudinal or statistical project involving the American capital punishment system and its most active historical period.

Language and Encoding

To ensure maximum technical compatibility and international reach, this dataset follows these rigorous standards:

1. Metadata in English: All variable names (column headers) are in English (e.g., full_name, execution_date, offense), facilitating seamless integration with global data science libraries and professional workflows in R and Python.

2. Data Integrity and Standardization: The records maintain high-fidelity accuracy for legal and geographical data. Criminal offenses are documented using official judicial terminology, ensuring the dataset meets the requirements for formal criminological research.

3. Encoding (ASCII 1.0): The files were validated using readr::guess_encoding() in R, which detected pure ASCII encoding (confidence = 1.0). As ASCII is a strict subset of UTF-8, this dataset is natively compatible with all modern computing environments (Windows, macOS, and Linux) without any risk of character corruption or "mojibake" errors.

4. Cross-Platform Compatibility (R, Python, and Excel): Both formats were successfully stress-tested in the industry's leading analytical environments:

   • In R: The .csv and .xlsx files were imported using readr::read_csv() and readxl::read_excel(), correctly parsing the 98 rows and 9 columns.
   • In Python: Both files were imported via pandas.read_csv() and pandas.read_excel(), yielding identical DataFrames. The execution_date field is pre-configured for high-precision temporal analysis, being automatically recognized as a datetime64[ns] object.
   • Data Engineering: Complex text fields containing multiple offenses (e.g., "Capital Murder, Robbery") are correctly escaped following the RFC 4180 standard, ensuring that commas within fields do not break the dataset structure.

The dataset is ready for immediate deployment in standard data science workflows, providing a robust and "zero-configuration" experience for researchers.

Data Structure

The dataset is provided in .csv and .xlsx formats, containing the following variables:

Variable	Type	Description
id	String	Unique identifier for each record (Format: EXE-1999-XXX).
full_name	String	Legal name of the executed individual.
gender	String	Gender of the individual.
ethnicity	String	Ethnicity/Race of the individual.
state	String	US State where the execution was carried out.
age_execution	Integer	Age of the individual at the time of execution.
execution_date	Date	Date of execution (Format: YYYY-MM-DD). Validated as datetime64[ns] in Python.
execution_method	String	Method used for the judicial execution (e.g., Lethal Injection, Electrocution).
offense	String	Criminal charges/offenses. Multiple values are escaped using RFC 4180 standards.

Methodology and Quality

1. Data Sources: The records in this dataset have been meticulously cross-referenced and validated against official reports and investigative journalism to ensure historical accuracy:

   • Death Penalty Information Center (DPIC): Execution List 1999.
   • The Washington Post: Historical analysis of execution trends in the United States.
   • Wikipedia - List of people executed in the United States in 1999: Comprehensive index used for cross-referencing individual case details and judicial outcomes.
   • Criminal Records - True Crime: "The Year with the Most EXECUTIONS in US History: 1999" – Detailed historical review of cases and final statements.

2. Curation Process:

   • Triangulation: Every entry (98 records) was verified across the three sources mentioned above to confirm the precision of the execution_date, age_execution, and state.
   • Normalization: Column names follow the snake_case convention (full_name, execution_method, execution_date), ensuring standard compatibility with Python, R, and SQL workflows.
   • Integrity Validation: The dataset uses a unique primary key (id) with the format EXE-1999-XXX to prevent duplication and facilitate relational database integration.

3. Technical Interoperability:

   • Standard Compliance: The .csv file was engineered under the RFC 4180 standard to handle complex text fields in the offense column without structural breakage.
   • Multi-Platform Testing: Both .csv and .xlsx formats were stress-tested in Visual Studio Code, RStudio, and Jupyter Notebooks, yielding 100% error-free imports.

License

This dataset is distributed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. This allows for the redistribution and adaptation of the data for any purpose, even commercially, provided that appropriate credit is given to the author and the original source is cited.

Author

Renzo Cáceres Rossi

• ORCID: 0009-0005-0744-854X
• GitHub: us-capital-punishment-1999