Linking individuals across historical data sources opens up exciting new research possibilities.

The Census Linking Project allows researchers to create longitudinal datasets from historical US Census records (1850-1940) in order to study processes that unfold over time, such as intergenerational mobility and immigrant assimilation. But linking requires large investments of time and resources, limiting their accessibility to scholars with limited research budgets. With the Census Linking Project, we aim to reduce these barriers by providing customizable linked historical datasets to the broad research community.

The project uses several different automated methods to link individuals across Census pairs using first and last name, year of birth and state/country of birth. We provide a series of fully anonymous crosswalks between each pair of Censuses. We advise researchers to test the robustness of their results using multiple linking algorithms and use their judgement to determine whether a method is right for their context.

The website will be updated periodically as links using additional matching methods become available. Links made using ABE-JW and EM methods are in progress. We also invite researchers to get in touch with us if they want to add their own Census links to our crosswalks using alternative methods or linking specific populations that might be difficult to link using automated methods.

We acknowledge generous funding from the Industrial Relations Section at Princeton University and support from the Minnesota Population Center and the National Bureau of Economic Research. We are grateful to Ancestry.com for providing the underlying data.

Citation

Publications using data from the site should cite the Census Linking Project appropriately as follows:

Ran Abramitzky, Leah Boustan, Katherine Eriksson, Santiago PĂ©rez and Myera Rashid. Census Linking Project: Version 2.0 [dataset]. 2020. https://censuslinkingproject.org

Please follow the IPUMS citation guide and terms of use when utilizing publicly available Census data and adhere to the security agreement and appropriate citation in your license when using restricted data.

If using the Census Tree links, please also include the following citation for those data:

Joseph Price, Kasey Buckles, and Mark Clement. Census Tree Data: Version 1.0 [dataset]. 2021. https://censuslinkingproject.org

Using our work?

Tell us about your published work using data from the Census Linking Project here.