Forum Thread Context, November 2004

A discussion thread was posted in November 2004.

Context note

In November 2004, a forum member posted several hundred links to outside commentary and reports about that month's election, introduced by a quotation.

The thread’s title, quotation and linked claims are not verified or endorsed. It was written under a private individual's forum name, consisted of links to and extracts from third-party publications, and falls outside this neutral historical collection. The general policy is set out on the forum overview. No claim is made about the 2004 election or any other election.

Reliable information about American election administration should come from current official sources.

Why the 2004 answer was a link list

It is worth understanding why a compilation of blog links seemed like a reasonable resource in 2004, because the answer is not simply that its author was undiscriminating. There was very little else. The federal election administration agency was months old and published almost nothing. The large-scale surveys of how jurisdictions run elections did not exist. Academic study of election administration was a small field without the datasets it now takes for granted. Jurisdiction-level records of what equipment was used where were maintained by volunteers, if at all.

In that environment, aggregation by hand was a rational if unreliable response to scarcity. The problem is that a link list carries no filter: a careful county-level report and an unfounded rumour appear as adjacent entries, and the compilation confers a false equivalence on both. The scarcity that produced it is described further on the resources page.

Where the data is now

Federal survey and research data

The U.S. Election Assistance Commission publishes a programme of studies and reports, the centrepiece of which is a biennial national survey of election administration collecting data from local jurisdictions across the country on registration, turnout, mail and absentee voting, provisional ballots, polling places, poll workers and equipment. It is the largest systematic dataset on how American elections are actually administered, and nothing comparable existed in 2004.

Academic data and research

The MIT Election Data and Science Lab publishes datasets, documentation and peer-reviewed research on election administration, voting technology and election performance, including compiled precinct- and county-level returns and methodological guidance on how to use them. For a researcher, this is the natural starting point, and its published methods make its work checkable in a way a link compilation never can be.

Comparative state law

The National Conference of State Legislatures maintains nonpartisan comparative summaries of state election law, including current state audit law summaries and parallel summaries on standards, testing and certification. Because so much of American election practice is set at state level, comparative summaries of this kind are often the fastest route to an accurate answer.

Equipment by jurisdiction

For the specific question of which voting equipment is used in which county — the question the 2004 volunteers spent most effort trying to answer by hand — Verified Voting maintains a public database recording equipment by jurisdiction, including the type of system and whether it produces a paper record. It is cited here solely as a current reference dataset. That organisation is independent, its present-day positions are its own, and no relationship of any kind is implied.

How to evaluate a source on this subject

A general observation for research into a contested subject: Sources on election administration divide fairly cleanly into those that publish their method and their underlying data, and those that publish conclusions. The first kind can be checked, corrected and built on; the second cannot. Federal survey data, academic datasets and published certification test reports are all of the first kind — which is what makes them useful even to a reader who is sceptical of the institutions producing them, since the material can be examined directly rather than taken on trust.

The certification records that answer most equipment questions are described on the certification background page, and the sequence of events across the years this site was active is set out on the timeline page.