Datasets Available through CPRC

Columbia Population Research Center has various datasets made available through the Columbia Data Platform (CDP). CDP is a cloud-based platform where researchers can store, discover, analyze, and share data, and where datasets can be archived. It is powered by Redivis and runs on Google Cloud Platform. It is approved for sensitive data, including HIPAA, PII, PHI, and RHI, and Columbia users can get a free account.

L2 Data

L2 Voter Data supports research into voter behavior, political participation, representation, demographics, and electoral geography in the United States. The database brings together voter-registration records from across the United States in a standardized format and supplements them with demographic, geographic, consumer, and modeled information. The data include millions of individual-level observations over time.

L2 data may indicate whether someone participated in an election and, in some cases, the method used to vote.The data does not reveal which candidate or ballot option the person selected.

About the Data:

Available fields vary by state and year. L2 data may include:

  • Voter-registration status 
  • Political-party affiliation where recorded by the state
  • Election-participation history
  • Voting method, such as absentee or in-person voting
  • Precinct, legislative district, and other geographic information
  • Basic demographic and household characteristics
  • Contact information
  • Consumer-preference information
  • Modeled or inferred demographic and political characteristics

Researchers should consult the current L2 data dictionary before beginning an analysis. In particular, it is important to distinguish information obtained from official records from variables that are appended, modeled, or inferred.

IPUMS Ancestry Full Count Census Data

The IPUMS Ancestry Full Count Census Data contains complete historical census enumerations from 1850 to 1950, cleaned, coded, and harmonized by IPUMS. The public IPUMS full count files are free, but this restricted version adds variables that are not publicly available. Those include names, sub-county geography such as addresses, identification of the original enumeration form, and string versions of the public variables. These variables make it possible to link individuals across census years, study neighborhoods and households at a fine geographic scale, and work with name-based information.

About the Data: 

Available fields vary by year but may include: 

  • Cleaned, coded, and harmonized census records for every enumerated person
  • Restricted variables: names, sub-county geography (e.g., addresses), original enumeration form identifiers, and string versions of public variables