Historically, signature petitions have been used to determine whether a proposed initiative is of sufficient interest to the electorate to place it on the ballot. Currently all state initiative procedures use signature petitions for this purpose. However, it has become increasingly obvious that signature petitions are a flawed way to assess the interest of the electorate and cannot be made to work, especially for large jurisdictions (see Stern, 2002). The Ballot Initiative Strategy Center has put together a web site to combat ballot fraud. They recommend a variety of remedies. A number of these ideas have been put into practice in state initiative law (see Waters, 2003, pp. 20-24). However, none of these reforms address the fundamental problem that signature petitions are a flawed way to assess the public interest. Polls are more accurate and less expensive in all but the smallest jurisdictions (see Act, 3-F-2). The use of polls to qualify initiatives is a new idea and has not been tested in state initiative law. All states currently use signature petitions.

In order to show that there is reasonable interest among citizens for the possible enactment of an initiative, a petition may be circulated to secure the signatures of an Electoral Trust-specified number or percentage of voters registered within the relevant jurisdiction. This practice is followed in states that have initiative laws. The practice is continued, but made considerably easier and less expensive, by the provisions of the Democracy Act, which permits signatures to be captured electronically as well as manually. The Electoral Trust will specify the forms and format to be used to collect manual signatures and the mechanisms through which sponsors may create on-line petitions. To ensure the integrity of the process, data captured on-line will be stored and managed on information systems under the direct control of the Electoral Trust. The number of signatures required on any petition will be related to the number of voters who voted in the last presidential election, and will be determined by the Electoral Trust for each government jurisdiction.

Stern & Holman, 2002, pp. 4-5 wrote:
The Act requires that 2% of the number of voters in the last presidential election sign petitions for statutory measures and 5% of the number of voters in the last presidential election sign petitions for constitutional or charter changes. The percentages seem reasonable. Some persons in smaller states, however, will be concerned that a few states, such as California and New York, may be able to qualify a measure on their own. Some states currently require a geographical representation to qualify initiatives within their own states.

This suggestion was adopted by allowing the Electoral Trust to determine petition thresholds.