Reports : November 13, 2001
Uniform Poll Closing and Reporting

Summary of Conclusions

1. The effect of most early projections on voter turnout is small. In most cases, the early projections simply confirm what voters expected to happen. In some circumstances, however, the effect of projections may be large enough to influence outcomes further down the ticket. The effect is concentrated in the western region of the country.
2. Uniform poll hours, such as exist in Canada, would probably meet resistance from western states, where polls would have to close earlier than they now do, and from eastern states, where polls would have to open later than they now do. Resistance would arise partly from concerns about convenience for voters, and partly from potential difficulties in staffing polling places and conducting the count, which could range far into the night in the East.
3. A uniform closing would tend to restrict turnout opportunities for workers in blue collar occupations, who tend to vote Democratic. A uniform opening would restrict opportunities for white collar workers, who tend to vote Republican. Most voters so affected would find ways to vote at other times, but uniform poll times would probably have a greater effect on lower status and less educated citizens, who typically are less motivated to vote and therefore more sensitive to convenience.
4. Restrictions on the official reports of election outcomes could probably not prevent early projections absolutely, but they could raise the cost of a projection of a close election considerably. In closer races, exit pollsters could defeat restrictions on official reports by increasing the size of exit poll samples, by polling in more precincts, and as they already do by observing counts in precincts, which activity by law is open to public scrutiny in most states.