data-datasets

Harvard Dataverse

Harvard's instance of the Dataverse platform, a large repository of research data. The search API is keyless and our search queries it directly; the website offers the same search for browsing. A standard stop when looking for academic datasets.

Datasets API

Search this archive

Why it’s useful & how it works

Harvard Dataverse is one of the most widely used academic data repositories, operated by Harvard's Institute for Quantitative Social Science and open to researchers from any institution and discipline. Datasets are organized into subject-specific dataverse collections, and each deposit receives a persistent DOI. You search by keyword, author, or subject on dataverse.harvard.edu, and Echo queries the same search API to surface relevant datasets for you.

What’s inside

Tens of thousands of datasets and hundreds of thousands of data files from researchers across the social sciences, natural sciences, and humanities. The repository has been running since 2007 and is especially strong in social-science and political-science data.

API access

https://dataverse.harvard.edu/api/search?q= (keyless read)

What we measured

Our own probes, not the archive’s own claims. Re-run periodically; every reading below is dated.

Reachability

Direct request: Responded HTTP 202 563 ms
Through a datacenter proxy: Responded HTTP 202 562 ms
API, direct: Responded HTTP 200 355 ms
API, through a proxy: Responded HTTP 200 902 ms

Did a real search return anything?

Returned results · 21404 results for the sample query · median 1.1 s · stable across runs

Sample query: US 2020 General Official Election Results in Tabular Format

Benchmark coverage

Found 10 of 10 test items (100%) in the Datasets cluster. Median response 556 ms.

By how well known the item is: 3/3 well-known, 6/6 mid, 1/1 obscure .

One item it found: Replication Data for: Learning Concept Bottleneck Models from Mechanistic Explanations

Reachability measured 2026-08-22. Search checked 2026-08-22. Benchmark run 2026-08-22.