South Shore News | Ballot Question Series, Week 4 of 9
This is the fourth in a nine part series looking at each of the ballot questions coming to Massachusetts voters in November.
Please take the poll in the middle of the article.
On Saturday, Aug. 22, Massachusetts residents could register to vote and cast a ballot in the same visit.
It was an accident of the calendar. Early in-person voting for the Sept. 1 primary opened that morning. The in-person deadline to register closed at 5 p.m. the same afternoon. And because registration is available at early voting locations, a resident could walk into a town hall, sign up, and vote before leaving.
Here is the thing worth understanding about that day: it is not what Question 4 would do.
The measure on the November ballot creates registration on Election Day itself. It says nothing about the early voting period. What happened on Aug. 22 — registering at a clerk’s counter, staffed by clerks, during a quiet week with no line — is the version that is not on the ballot. The version that is would put every late registrant into a polling place on the single busiest day of the cycle.
It isn’t clear is anyone actually registered and voted on August 22, but it was possible.
What Question 4 would do
A voter would register at the polling place by showing either a Massachusetts driver’s license or state ID, or a document with their name and current address — a utility bill, bank statement, lease, paycheck, phone bill, government check or correspondence, or a student fee statement. In every case they would also sign an oath swearing they are a citizen, at least 18, not under guardianship, not disqualified for corrupt election practices, and not voting elsewhere in that election. A false statement is a felony carrying up to five years in prison, a $10,000 fine, or both.
A voter who cannot produce documentation votes provisionally and then has a window to bring the missing document to the clerk: until the polls close in a municipal election, two business days in a state primary, six days in a state election.
The measure also lets an already-registered voter update their address at the polls, which is probably the more common use. Registrations made this way carry forward to future elections. It does not let anyone change party enrollment at a primary — a provision worth watching, since Question 3 on the same ballot would abolish party primaries entirely. It applies to municipal elections; the only carve-out is town meeting.
One provision has gone almost unmentioned. Massachusetts currently lets a voter who moves within the state continue voting in their former city or town for up to six months. Question 4 would end that, on the logic that anyone who moves can now simply register at their new polling place on Election Day.
It takes effect Jan. 1, 2028.
Who is behind it
Secretary of the Commonwealth William Galvin, who runs elections in Massachusetts, personally signed and filed the petition — the first ballot measure he has ever signed. His office later assigned it its number on the ballot.
That is a turn, because eight years ago Galvin was on the other side of this argument in court. In a 2018 case brought by the ACLU, he was the defendant who successfully defended the state’s then-20-day registration cutoff. Writing for the Supreme Judicial Court, Justice Kimberly Budd held that the deadline “does not disenfranchise any voter” — while noting in the same opinion that the blackout period “may need to be reconsidered.” Galvin said afterward the decision would help his push for same-day registration.
He went to the ballot after the Legislature repeatedly declined to act. Asked why: “It hasn’t happened.”
Backers gathered 87,408 valid signatures, certified in December 2025, then cleared the 12,429-signature second round required after lawmakers let the measure pass them by in May. Attorney General Andrea Campbell endorsed it in August. The Massachusetts Democratic Party, the ACLU of Massachusetts, Common Cause, MassVOTE, the League of Women Voters and the Massachusetts Voter Table all support it.
Supporters told a legislative hearing in March that the measure could add roughly 100,000 voters.
What the research actually shows
That 100,000 figure, and a turnout increase of “3 to 11 percent” that has circulated alongside it, deserve scrutiny.
The percentage range does not come from a study. It traces to a state legislator’s paraphrase at a hearing, and no research this newsroom could locate produces it — it appears to combine the floor of a general-population estimate with the ceiling of a projection about a specific subgroup.
The Tufts Center for State Policy Analysis reached the same conclusion this month. It estimates that roughly 85 percent of eligible Massachusetts voters already have active, current entries in the voter rolls, and notes that the widely cited 3-to-6-point turnout estimates "generally come from states that lacked the kind of automatic registration system that Massachusetts has developed." Its bottom line: the effect on turnout here "is likely to be small."
Minnesota and Wisconsin adopted it in the 1970s, have high turnout, and drive most of that correlation. When researchers instead compare a state to itself before and after adopting the policy, the measured effect shrinks sharply — a 2013 study in Political Analysis argued the cross-state designs rest on untestable assumptions and found little effect in Minnesota and Wisconsin. And the most recent multi-state analysis, from researchers at Tufts using 2020 voter file data, found no statistically significant turnout effect in the general population once state and county differences were controlled for.
What survives is a youth effect. A 2022 study in the Journal of Politics found same-day registration raised turnout among 18-to-24-year-olds by 3.1 to 7.3 percentage points. The Tufts researchers found the same. That is the group the ten-day deadline most often catches — students who move, first-time registrants, people who change addresses. Massachusetts often compounds this issue by scheduling the State Primary on or near September 1, by far the most common moving day in Greater Boston.
Massachusetts is also a hard case for a large effect. The state already registers voters automatically at the RMV, MassHealth and the Health Connector, and has online registration, no-excuse mail voting and early voting. The barrier this removes has largely been removed already, and no published study estimates what Election Day registration adds in a state that has all of that.
The closest real-world comparison is Washington State, which has had all-mail voting since 2011 and adopted automatic and same-day registration together in 2018. In 2024, Washington processed about 29,000 Election Day registration transactions — less than one percent of ballots cast, with brand-new registrations at about a third of one percent. Minnesota, which built its registration culture around the policy starting in 1974, saw 9.1 percent.
One study points the other way, and it belongs here. Researchers writing in the American Journal of Political Science in 2014 found that early voting by itself reduces turnout, by dispersing what had been a single focal day — but that adding same-day registration reverses the effect. Under that theory a state that already has early voting is precisely where this reform does more. Its early-voting finding has failed some replications, but the study is peer-reviewed and directly on point.
The measure polls at 69 percent — though the survey question described registering “on Election Day and during early voting periods,” which is broader than what Question 4 contains.
What Massachusetts can actually count
The strongest evidence for Question 4 is not a turnout projection. It is a number from the last election.
In November 2024, Massachusetts voters cast 5,106 provisional ballots. 3,309 were rejected — about two-thirds. Common Cause Massachusetts, which analyzed the state’s data and is a member of the coalition campaigning for Question 4, says nearly all of those rejections were administrative: address changes, voting in the wrong precinct after moving within a town, transcription errors. Its conclusion is that same-day registration would have made almost the entire provisional process unnecessary.
That conclusion is the campaign’s, and it is worth one degree of skepticism. A voter who turns up at the wrong precinct after moving still has to be registered where they are standing, by someone with the authority to do it, while they are willing to wait. Some share of those 3,309 would have left either way.
The scale also matters: 3,309 is roughly a tenth of one percent of the 3.5 million ballots cast in Massachusetts that year. It is a real harm to real people and a poor basis for claiming a large turnout increase.
The count understates the problem in one direction, though. Someone told at the polls that they missed the deadline, who leaves without casting anything, is never recorded anywhere.
On fraud
The standard objection is that same-day registration invites fraud. The evidence does not support it.
The most systematic study, published in 2007, reviewed nearly 4,000 news reports across six same-day registration states over seven years and found ten incidents with any apparent merit. In New Hampshire, about 240 Election Day registrants were investigated; six cases of false information emerged, and most of those turned out to involve legally eligible voters.
A more recent figure comes from a federal court in New Hampshire, which found in May that of 8.3 million ballots cast in that state between 1998 and 2024, 47 involved wrongful voting, eight of them by non-citizens. Two caveats belong with it. That finding came in litigation over a 2024 New Hampshire law that tightened its own Election Day registration by requiring documentary proof of citizenship — meaning the state Massachusetts holds up as a 30-year precedent decided its own system needed hardening. And New Hampshire appealed the ruling in June, so the finding is not final.
The serious version of the integrity concern is different, and it is not about fraud rates. Under advance registration, eligibility is checked before a ballot enters the count. Under Election Day registration, the ballot is cast and mixed in before verification happens. If the post-election check then fails — bad address, duplicate record — there is no way to retrieve the vote. The error rate is very small. The errors are uncorrectable rather than merely rare.
The capacity problem, and Boston
Whether Massachusetts can administer this is a separate question from whether it should, and there is a large piece of recent evidence.
In November 2024, at least 14 Boston precincts ran out of ballots. Galvin’s office investigated and found “serious problems,” the central one being “the inability of the Boston Election Department to directly communicate, in real time, with each voting precinct.” In February 2025 he ordered an overhaul and installed a receiver over the department. A state monitor is scheduled to oversee Boston’s elections through 2026, and the city’s new elections director starts this month.
That is the same official now asking voters to approve a system that requires checking the statewide voter file in real time at every precinct in the state, to prevent someone registering and voting twice.
The Tufts analysis puts the scale plainly: Election Day alone requires nearly 10,000 poll workers at some 1,200 polling places. And not every Massachusetts town uses electronic poll books — which means in some places a duplicate registration could not be checked in real time at all. The analysis also draws a distinction the campaign hasn't: a provisional ballot can be set aside and sorted out later, but a registration has to be completed on the spot, while the voter waits.
The architecture makes it harder. Massachusetts does not run one central voter file the way the phrase “statewide voter registration system” suggests. A federal court described the arrangement last year: the state lawfully delegates custody of voter records to its 351 individual cities and towns. Most precincts still check voters in from static lists.
Rep. David Vieira of Falmouth put a narrower version of the objection on the record: “until we expand broadband throughout the commonwealth, it’s not going to happen at every polling location in every community in the commonwealth.” Rep. Michael Moran, now House majority leader, framed the scale: “There are 2,000 precincts in Massachusetts. All of those would need to be looked at.”
The Tufts analysis, which takes no position on the question, notes that the Secretary of the Commonwealth is “in the awkward position of championing and implementing Question 4,” and suggests the Legislature authorize an independent assessment of any rollout.
Question 4 sets no implementation deadline and specifies no system.
What happens at the table
There is also a practical question the campaign has not addressed publicly: what registering someone actually does to a polling place.
Washington’s official fiscal analysis assumed ten minutes per same-day registration — checking documents, entering data, issuing a ballot. Ten minutes is six people an hour at one station. A precinct with fifty late registrants needs a dedicated worker for most of the day, or it grows a line.
The cautionary case is Michigan in 2022, where a registration surge at the University of Michigan produced a queue in which one voter waited roughly six hours and the last ballot was cast shortly after 2 a.m.
Two access questions follow, and neither has been raised in this campaign.
