STCA Dataset

Welcome to the US-CAN Safe Third Country Agreement Dataset repository.

We collect public datasets about the entry of asylum seekers into Canada from multiple departments of the Government of Canada, in the hopes of providing researchers with better data in order to study the effects of the US-Canada Safe Third Country Agreement. These data will be released under the HAVEN: Asylum Lab project for public usage.

Using the Data

Information on what datasets we have, what they include, and what they measure, are all included in our master metadata document.

• To access the complete metadata, check out data/master_doc_compressed.pdf.

We have the cleaned data in .csv and .rds file format under the data/clean/ directory. Accompanying metadata are available for each table and sub-dataset.

Reproduce the Cleaned Data

Data Source

The raw data files are located under the data/raw/ directory. Manifests are available for each raw file which includes the source URL.

Processing the Source

Requirements:

• A working installation of R (at least version R-4.3.2, tested on Windows 11, MacOS Monterey, and MacOS Sonoma)
  • If you don't have it in your environment, install the tidyverse package by running install.packages("tidyverse") in the R console
• Any .xlsx and .csv editor

Scripts under the R/ directory will parse the source webpage files in data/raw/web_ircc/ and data/raw/web_rad/ and output the cleaned tables to data/clean/ircc/ and data/clean/rad/ directories respectively. IRB Irregular crossers claims data (data/clean/quarterly_irregular_claims/) are edited manually using a spreadsheet editor, for instructions check the "Modules" section in the complete metadata.

Generating the Metadata Document

Requirements:

• A working installation of Typst (tested on typst-0.11.1 on Windows 11)
• A working installation of Ghostscript (optional, tested on ghostscript-10.03.1 on Windows 11)
• A working shell on this repo's root directory. Make sure to run the commands below using the right path.

The individual .csv metadata for each table/sub-dataset are manually generated. To generate the complete metadata document (data/master_doc.pdf), run the following command:

typst compile data/master_doc.typ

This will read all the individual metadata and compile it into a single human-readable document.

OPTIONAL: Compress data/master_doc.pdf for distribution by running the following command:

ghostscript -sDEVICE=pdfwrite -dNOPAUSE -dBATCH -dSAFER -dPDFSettings=/screen -sOutputFile="data/master_doc_compressed.pdf" "data/master_doc.pdf"