Doing Archival Research in Political Science By Scott A. Frisch, ...

and were astonished to learn that there was anything of worth in those boxes.

Are there gaps in the archival record? Absolutely, there are. Upon close examination, all political science data are somehow incomplete. The question is whether the gaps in a researcher’s data—in this case, archival data—are systematic. If those gaps are systematic, it is a potential problem; if the gaps are random, they are a nuisance (even a big nuisance), but they do not constitute a fundamental problem. And, we argue, there are means by which a researcher can fill these gaps.

Why are there gaps in the archival record? It may be not so much for nefarious reasons, but for perfectly explainable ones, as Mark Greene explained:

Gaps in the collections occur primarily because of inconsistent record-keeping (including the destruction of inactive records taking valuable storage space in the attics of Congressional office buildings; destroyed simply because the office no longer needed them and no archival repository had yet contacted the member to explain the value of some of those records) and deliberate appraisal by the repository (either carried out prior to transfer by the office, at the advice of the repository, or carried out at the repository after transfer).¹⁰

There are three likely sources of systematic bias in the archival records. The first are the decisions that are made at archival repositories (appraisal). Libraries face difficult choices about which collections and what material to maintain in the context of limited resources; one of the main limiting factors is the space to house political collections. According to Leigh McWhite, a political papers archivist at the University of Mississippi,

Repositories are…making tough decisions about whether they have the space and resources to store entire congressional collections…a number of archives are choosing to reject or to destroy selected series within a collection, like case files or academy