Calculation of Likely Voter Intent

Likely Voter Intent

A total of 6042 residents voted in this election, and each could cast up to 2 votes for Township Committee. Here’s how the election results tallies might have changed if all the election results were accounted for:

Polling machines were defaced.  Four polling machines in three different polling locations were defaced:  Julia Allen’s and Frank Gatti’s names were written-in with a marker on the machine itself on the write-in line under columns for some other office.  Until the defacement was detected and the machines taken out of service, an unknown number of subsequent voters likely thought they were casting votes for Allen and Gatti.  In fact they were casting a blank vote because the voting machine can only register write-in names that are typed in in the correct place.

Voters felt intimidated.  Many voters reported being intimidated and rushed by poll workers when attempting to cast their ballots, which may have contributed to voters failing to correctly cast their write-in ballot.  Although state law specifies a 2-minute limit per voter in the polling booth, Hunterdon County has not in recent memory enforced this limit.  Challengers for the opposition demanded that voters be held to a strict 2 minute time limit in the polling booth.   With two write-in candidates and two long referenda questions on the ballot, many felt that two minutes was not enough time in the booth.  Clearly, enforcement of the 2-minute limit was targeted at voters who were attempting to write-in their votes.

Write-in Voting Errors.   According to election data, 266 write-in votes of approximately 133 individuals casting a ballot at the polls or by mail were not counted toward the election results due to errors of various types. It is safe to assume that all of their votes were intended for Allen & Gatti.

Massive ‘Undervote’.  It appears that 663 individuals did not register votes for Township Committee at all.  In hotly contested elections, this ‘undervote’ is usually extremely low.   In this election, it was almost 11%.   It is very likely that these people thought they voted, but their vote didn’t count for any number of reasons:   by voting in a defaced voting machine, by failing to push the “enter” button next to the candidates name after laboriously typing in the names of Allen & Gatti, or by making other errors due to time pressures in the polling booth.   Analogous challenges were present in mail-in ballots. It is also very unlikely that these voters intended to vote for Broten or Tropello since these names were on the machine or mail-in ballots and voting for them was trivially easy.

Although Broten and Tropello won the election, they should not assume that they have a voter mandate to dismantle the policies and programs that have made Readington a beautiful place to live.  

Winning With Less Than A Majority Of The Votes

How could someone win office with less than a majority of the votes?

In any election, a voter may decide not to vote for candidates for some offices.  However, the Township Committee race was the only hotly contested race on the ballot and was clearly the reason that voter turn-out was so high in Readington (over 50% vs 30% in the rest of NJ).  It’s not likely that over 13% of the people who showed up to vote didn’t bother voting for the Township Committee race. A more likely explanation is that many voters tried but were not able to cast their write-in votes for Allen & Gatti correctly.  To cast a successful write-in vote on Readington’s voting machines, you must locate the correct column, type in the candidate’s name correctly on the keypad, and then press enter.  If you failed to press enter before beginning your second write-in vote, or before pressing the “Cast Vote” button, your vote did not register. Mail-in ballots presented similar challenges.

​Contributing Factors

Other factors also contributed to a win with less than a majority of the votes:

January 12, 2015

Only 47% of the 6,042 Readington residents who voted in November election voted for John Broten or Sam Tropello.

In fact, there is a credible argument that the majority of voters intended to vote for Julia Allen and Frank Gatti.  Approximately 796 voters who participated in the election were not successful in casting votes that counted toward the results for Township Committee, due to election irregularities and the difficulties of casting a write-in vote.  If their ‘unsuccessful write-in votes’ had counted, the outcome of the election would have been very different.

No Voter Mandate for Broten & Tropello

  1. Source of data:  publically available certified votes and the Hunterdon County Statement of Vote, HUN+20141104_E, page 39.
  2. Certified Total Vote = Recorded Votes - Disallowed Votes
  3. 1160 ‘undervotes’ were reported for the Township Committee race (1160 fewer votes than expected for the number of voters who went to the polls).  95 voters chose to vote for Broten but not Tropello.  It is likely that the remaining 1065 undervotes represented voters who went to the polls to cast a ballot for Allen and Gatti but were not successful.
  4. No vote for Township Committee was recorded for 79 mail-in ballots.  Since voting for Broten and Tropello was trivially easy, it is likely that these votes were not counted due to write-in errors.
  5. 50 undervotes were reported on provisional ballots.  17 of these were cast for Broten and Tropello.  It is likely the remaining missing votes were intended for Allen and Gatti.  
  6. Likely Voter Intent = Certified Election Totals + Disallowed Votes + Undervotes

These results were derived from the publically available certified votes and the Hunterdon County Statement of Vote, HUN+20141004_E, page 39.   Calculations and assumptions are shown in the following table.