The Illinois General Assembly is considering SJRCA0001, a constitutional amendment to remove the prohibition against multiple personal income tax rates, essentially permitting an Illinois state graduated income tax. The Senate has approved the amendment and if the House does as well it will go before the voters for ratification on the 2020 ballot.
Per Article XIV, Section 2(b) of the Illinois state constitution there are two possible ways the proposed amendment could be ratified by the voters:
A proposed amendment shall become effective as the amendment provides if approved by either three-fifths of those voting on the question or a majority of those voting in the election.
That means it will either need 60% of the Yes votes or it will need a simple majority of Yes votes among all ballots cast. There will be some voters who cast a ballot but for whatever reason don't vote on this item on their ballot, those voters are included in the simple majority calculation for the latter ratification method.
For those who are working to try to ratify the amendment which of the two methods is easier? That depends, specifically it depends on the dropoff rate, the percentage of voters who cast a ballot but don't vote on this measure. If the dropoff rate is 0.00%, if every voter who takes a ballot votes on the amendment, then the denominator of both calculations is the same and you only need a simple majority of voters to ratify the amendment. However as the dropoff rate increases so does the threshold for ratification. If the dropoff rate is less than 16.667% it's easier to get a simple majority of those voting in the election than three-fifths of those voting on the question, otherwise the 60% threshold is the only path.
Here's a table showing the ratification and dropoff history for amendments going back to the 70's, based on this data source (see notes 1 & 2 below for data concerns). In the 80's and 90's it was common to have dropoff rates that were very large but that has changed in recent years. Looking back on the last three amendments to go before the voters all three had a dropoff rate lower than 16.667%.
| Cycle | Issue | Total Votes | Ballots Cast | Dropoff |
|---|---|---|---|---|
| 2016 | Transportation Lockbox | 4,811,115 | 5,627,149 | 14.50% |
| 2014 | Victim's Bill of Rights | 3,382,466 | 3,680,417 | 8.10% |
| 2014 | Voter Discrimination | 3,310,295 | 3,680,417 | 10.06% |
| 2012 | Pension Benefits | 4,337,888 | 5,236,120 | 17.15% |
| 2010 | Governor Recall | 3,285,617 | 3,780,779 | 13.10% |
| 2008 | ConCon | 4,555,927 | 5,539,172 | 17.75% |
| 1998 | Courts Commission | 2,084,123 | 3,541,379 | 41.15% |
| 1994 | Rights After Indictment | 2,431,908 | 3,219,122 | 24.45% |
| 1994 | Effective Date of Laws | 2,144,200 | 3,219,122 | 33.39% |
| 1992 | Crime Victim's Rights | 3,680,194 | 5,164,357 | 28.74% |
| 1992 | Education | 3,300,089 | 5,164,357 | 36.10% |
| 1990 | Tax Sales | 1,390,318 | 3,420,720 | 59.36% |
| 1988 | Voting Qualifications | 3,249,002 | 4,697,192 | 30.83% |
| 1988 | Delinquent Tax Sales | 2,533,075 | 4,697,192 | 46.07% |
| 1988 | ConCon | 3,627,253 | 4,697,192 | 22.78% |
| 1986 | Bail and Habeas Corpus | 1,771,133 | 3,322,657 | 46.70% |
| 1986 | Veterans' Property Tax Exemption | 1,588,346 | 3,322,657 | 52.20% |
| 1984 | Veterans' Property Tax Exemption | 2,190,345 | 4,969,330 | 55.92% |
| 1982 | Bail | 1,629,176 | 3,856,875 | 57.76% |
| 1980 | Legislative Article (Cutback Amendment) | 3,074,549 | 4,868,623 | 36.85% |
| 1980 | Revenue Article (Delinquent Tax Sales) | 2,656,407 | 4,868,623 | 45.44% |
| 1978 | Personal Property Tax | 1,686,261 | 3,342,985 | 49.56% |
| 1978 | Veterans Organizations' Post Homes Exemption | 1,554,486 | 3,342,985 | 53.50% |
| 1974 | Governor's Amendatory Veto | 2,632,032 | 3,047,822 | 13.64% |
Oddly enough, of those three the transportation lockbox amendment probably had the most media attention, but it had a higher dropoff rate than either of those in 2014. That may be due to the fact that it was the only one of the three on the ballot in a presidential election year, where you have more low-information voters participating who may not vote all the way down the ballot.
Why have modern elections had a much lower dropoff rate than those from late last century? It's hard to say but a likely explanation has to do with changes in the way we vote. It varies county to county but many counties in the state now have touchscreen voting machines or other electronic methods and it's common for those voting systems to warn voters when they've undervoted any questions and we're seeing a significant reduction in dropoff votes now.
Here's a table of the Yes vote needed to pass the measure with a simple majority of all ballots cast at different dropoff rates:
| Dropoff | Yes/BC | Yes % |
|---|---|---|
| 0.00% | 50.01% | 50.01% |
| 1.00% | 50.01% | 50.52% |
| 2.00% | 50.01% | 51.03% |
| 3.00% | 50.01% | 51.56% |
| 4.00% | 50.01% | 52.09% |
| 5.00% | 50.01% | 52.64% |
| 6.00% | 50.01% | 53.20% |
| 7.00% | 50.01% | 53.77% |
| 8.00% | 50.01% | 54.36% |
| 9.00% | 50.01% | 54.96% |
| 10.00% | 50.01% | 55.57% |
| 11.00% | 50.01% | 56.19% |
| 12.00% | 50.01% | 56.83% |
| 13.00% | 50.01% | 57.48% |
| 14.00% | 50.01% | 58.15% |
| 15.00% | 50.01% | 58.84% |
| 16.00% | 50.01% | 59.54% |
| 16.67% | 50.01% | 60.01% |
If the dropoff rate for the upcoming vote is only 1% then it will simply need a Yes vote of 50.52% to pass, if the dropoff rate is 15% it will need a Yes vote of 58.84% to pass. It will automatically pass with a Yes vote of 60% or greater via the "three-fifths of those voting on the question" method.
Both sides are already spending on TV at a significant rate and that is before the measure has passed both houses of the General Assembly, it's likely that if the measure goes before the voters for ratification the media spending will grow even further. That kind of paid media attention will probably reduce the dropoff rate lowering the threshold needed for passage. That presents a difficult predicament for opponents, by bringing more attention to the measure, even negative attention, it will likely reduce the dropoff rate lowering the threshold needed for passage but they really don't have better options. If opponents encourage voters to skip voting on the measure then they run the risk of only Yes voters voting, which isn't an improvement.
What is clear though is that unless the Yes votes are well above 60% the outcome may not be known for certain until well after election day. It's entirely possible we have to wait until the vote totals are certified to be able to run the simple majority calculation on ballots cast.
Data Note 1: this was the data source for much of the historical data in the table above, however it has at least one error on it, probably a copy/paste mistake, the ballots cast number on the 2014 Voters Rights question is definitely wrong.
Data Note 2: for recent elections the ballots cast number is available via the State Board of Elections website in the Official Vote Total Book. However that report lists ballots cast numbers that are different in different locations. For example in the 2016 book it is listed as 5,666,118 on page ii for the purpose of calculating turnout but page vi lists it as 5,627,149. The 2014 book has the same issue, ballots cast is listed as 3,680,417 on page ii for the purpose of calculating turnout but on pages 4 and 8 it's listed as 3,667,206. Last week I emailed the State Board asking for a clarification, if I hear back from them I'll provide an update.
