Changing a position is allowed and expected in the My Vote Forward framework. A person can revise their priorities and opinions at any time; the revised answer replaces the earlier one going forward, and an opinion is never silently carried over when the wording of a proposal changes. For elected representatives, the framework treats a changed position as something to explain, not something to score. One limit applies to everyone: an opinion that was published can be withdrawn from future releases, but copies already distributed cannot be recalled.

The objection in its strongest form

The objection comes in two versions, and they point in opposite directions.

The first says a system that records positions over time will punish people for changing their minds. Records invite comparison. If a person’s earlier opinion is visible, anyone can accuse them of inconsistency; if a representative’s earlier stance is on file, opponents will use it as proof of a flip-flop. People will learn to avoid recording positions they might later revise, and the system will collect only safe, rehearsed views. A framework meant to encourage honest participation would end up encouraging caution.

The second says the opposite: a system that lets positions change freely makes them meaningless. If a participant can revise a priority every day, aggregates become unstable and easy to manipulate by last-minute campaigns. If representatives can change positions without consequence, voters lose the ability to hold them to what they said. On this view, accountability requires that positions have some weight and some memory, and a framework that treats every change as legitimate removes that weight.

Why the objection matters

The objection matters because My Vote Forward is built around ongoing participation rather than a single moment. “Participation over time” (see participation over time) only makes sense if positions can change. But how changes are recorded decides whether the record helps people think or helps others attack them. For representatives, the stakes are higher: a framework that makes position changes look like failures would push officials toward rigidity, and one that ignores them would weaken accountability.

Defined terms

  • Revision: a person’s new answer to a priority or opinion question, which replaces the earlier one going forward.
  • Stale opinion: an opinion recorded against wording that has since changed. The framework marks it stale and invites review instead of applying it to the new wording.
  • Snapshot: a published opinion fixed to a particular question version and date. See consent.
  • Cutoff rule: the stated date or rule that decides which answers a published result uses.
  • Trustee and delegate: two classic views of representation. A delegate follows constituents’ expressed wishes; a trustee exercises independent judgment on their behalf.

Evidence

Public opinion changes, sometimes a great deal. Pew Research Center’s long-running measure of attitudes on same-sex marriage illustrates scale. In 2004, “Americans opposed same-sex marriage by a margin of 60% to 31%”; by 2019, “a majority of Americans (61%) support same-sex marriage, while 31% oppose it”.⁠1 These are aggregate figures, so they show that measured public opinion shifted, not how many individuals changed their minds. They are cited here only to show that large changes in measured opinion are part of normal democratic life.

Deliberation changes views. Stanford’s Deliberative Democracy Lab describes Deliberative Polling as a process in which “A random, representative sample is first polled on the targeted issues,” then deliberates, and “After the deliberations, the sample is again asked the original questions”.⁠2 The lab reports that “Each time, there were dramatic, statistically significant changes in views,” and argues that the results represent conclusions the public would reach if more informed and engaged.⁠2 That characterization is the lab’s own interpretation of its method.

The trustee view of representatives. Edmund Burke argued to the electors of Bristol in 1774 that “Your representative owes you, not his industry only, but his judgment,” and that a representative who sacrifices that judgment to constituents’ opinion betrays rather than serves them.⁠3 On this view, a representative’s changed position after new evidence is not a breach of duty but an exercise of it.

Survey practice on wording. Pew reports that “Even small wording differences can substantially affect the answers people provide”.⁠4 That finding is why the framework treats an opinion as tied to exact wording: an answer to one version of a question is not evidence of what the person would say to another.

How comparable systems have handled it

Comparable systems handle change in different ways. Estonia’s internet voting lets a voter cast a new i-vote, with “Only the last i-vote cast is taken into account, and earlier votes are annulled”.⁠5 That rule exists to protect free choice, but it also shows a system designed so that the latest choice governs.

Deliberative Polling measures views before and after deliberation and reports change as a finding, not a failure.⁠2 Long-running survey series, like Pew’s same-sex marriage measure, report the public’s position at each point in time and present the shift as a trend.⁠1

Competing interpretations

One interpretation holds that a civic record should be forgiving: people should be able to revise without their earlier answers being visible, because the point is to capture what people think now. This favors private histories and public aggregates only.

A second interpretation holds that public accountability needs memory, at least for officials. Voters are entitled to know what a representative said before an election and what they did after. This favors public, dated records of official positions, with explanations attached.

A third interpretation, which the framework adopts for representatives, distinguishes changing a position from concealing the change. A changed position with a public explanation respects both the trustee’s judgment⁠3 and the voter’s right to know. Whether explanations satisfy voters is for voters to judge, not for a framework to score.

What the framework answers

My Vote Forward’s rules are:

  • Revisions are allowed and dated. A person who changes a ranking creates a new dated record, and published results state which snapshot or cutoff they use (see measurement and opinions).
  • An opinion is stored with its exact item wording and version. When official wording changes, the earlier opinion is marked stale and the person is invited to review it; it is never silently applied to new wording.
  • Opinions are private by default, so a person’s revision history is not public unless they published specific snapshots (see consent).
  • Representatives may publish explanations, including of changed positions, without being rated; the framework does not endorse, match, or score officials (see representation).
  • A published opinion can be withdrawn from future releases. The interface warns before sharing that copies may persist.

Hypothetical example. A participant ranks housing first in March, then moves it to third in June after a local rent decision. A release with a cutoff of April 30 counts the March ranking; a release with a cutoff of July 31 counts the June ranking. Neither release shows the individual’s history, and neither treats the change as inconsistency.

Where the objection still wins

Both versions keep some force. Against the first: anything a person chose to publish can be copied and compared with later views, and the framework cannot prevent that. Against the second: allowing free revision makes aggregates sensitive to timing, and a coordinated push just before a release cutoff could move results. The framework addresses this only by publishing cutoffs and timing, not by preventing it. For representatives, the framework’s refusal to score officials means it offers no accountability mechanism of its own; it relies on voters and the press to judge explanations.

What My Vote Forward can and cannot do here

Can:

  • Let people revise priorities and opinions at any time, with dated records.
  • Tie each opinion to exact wording and mark it stale when wording changes.
  • Keep revision histories private and publish only aggregates and chosen snapshots.
  • Give representatives a place to explain changed positions without rating them.

Cannot:

  • Recall copies of a published opinion.
  • Stop others from comparing someone’s earlier and later public statements.
  • Prevent last-minute coordinated revisions from affecting a release, though it can disclose timing.
  • Decide whether a representative’s explanation is adequate.

Unresolved questions

  • What cutoff rules best balance current views against manipulation near a release date?
  • Should a release report how many participants revised their answers, and would that reveal anything about individuals?
  • Should a representative’s explanations be versioned like questions, so that edits are visible?
  • How long should private revision histories be kept, and should people be able to delete them?

Sources

  1. Pew Research Center, “Attitudes on Same-Sex Marriage,” fact sheet. pewresearch.org/…
  2. Deliberative Democracy Lab, Stanford University, “What is Deliberative Polling?” deliberation.stanford.edu/what-deliberative-pollingr
  3. Edmund Burke, Speech to the Electors of Bristol, 3 November 1774, in The Founders’ Constitution, University of Chicago Press. press-pubs.uchicago.edu/founders/documents/v1ch13s7.html
  4. Pew Research Center, “Writing Survey Questions,” Methods. pewresearch.org/writing-survey-questions
  5. State Electoral Office of Estonia, “Introduction to i-voting.” valimised.ee/…
Author
Publisher
My Vote Forward
Version
1.0
Last reviewed
Evidence
5 numbered sources on this page, and the full source list.
Corrections
Report an error on this page. Revisions are recorded in the changelog.