Special Election Results Are Not A Reliable Way To Test Public Opinion

Special election results are not a reliable way to test public opinion, and for that reason, we believe long-term bonding should be placed on general election ballots whenever possible.

The Charlestown Citizens Alliance (CCA) did our best to get out the vote. Between April 12 and May 5, CCA published six posts at our website describing different aspects of the construction bond in detail and alerted voters to the availability of mail ballots, early voting, and where and when to vote. These posts were delivered to our readers in seven newsletters in the same three-and-a-half-week period. Finally, all posts were shared on the CCA Facebook and Twitter pages and on a large community Facebook group. Some were posted on Nextdoor.com.

A message contained in all these CCA communications was that the cost of operating aging buildings would be the same or more than the cost of new construction and that the May 7 school bond was the most fiscally responsible and educationally sound plan for the district.

None of the other political groups or parties in Charlestown provided any leadership, information, or even notice that there was a vote. Some members of the Charlestown Republican Town Committee posted as individuals on Facebook in opposition to the bond and other education issues, but from the parties themselves and the Town Council majority’s political PAC—Charlestown Residents United (CRU)—there was total silence.

Otherwise, the Westerly Sun did publish many letters but wrote only two news stories about the bond. Charlestown’s website had a tab that led to a description of the bond’s fiscal impact, but the town sent no written notice of the vote in any form of communication. Even the town’s Facebook page had no mention of the upcoming vote.

Perhaps the greatest awareness district-wide of the referendum came from the postcards sent out by BuildRI. The effect of these on the vote was mixed, but it did give anyone not on CCA’s newsletter list or active on Facebook notice of the vote.

CCA’s communications were likely responsible for a large portion of the 1,179 people who voted in Charlestown, and especially for the 847 who voted yes. After all, we had lots of positive communications and nearly every other group was silent. However, even with our efforts, the approximately 1,200 citizens who voted were a fraction of the about 5,200 and 4,200 who voted in the 2020 and 2022 general elections, respectively.

What these data show in stunning detail is that more citizens vote in general elections and the results are therefore different. In November 2018—a general election—71% of voters in Charlestown, 62% of voters in Richmond, and 64% of voters in Hopkinton voted to have the state allocate $250 million of their tax dollars on new school buildings. But, in 2024, a majority of CHARIHO voters refused to accept this money from the state, leaving their share of the money to go to other projects.

Many things may have changed since 2018, but the biggest difference in these two elections is the number of people voting. If there had been an option (there was not), putting the CHARIHO construction bond referendum on a general election ballot wouldn’t have guaranteed its passage, but it would have guaranteed a better representation of the electorate in all three towns.

Now that the special election is over, many interpret the results as representing the opinion of all voters in each town. It is true that those who didn’t vote lost an opportunity to receive once-in-a-lifetime funding from the state. However, because the number of voters is so small and the margin of error so large, the opinion of any majority shouldn’t be assumed based on special election results.


Cliff Vanover, the author of this post, is Treasurer of the Charlestown Citizens Alliance.