I am often asked why the likely voter polling data, especially in New York State, has been consistently off target on the margin of victory over the last decade and a half. My response is always the same: The polling data itself is not usually the problem, it is the likely voter screens that are often suspect, sometimes projecting very inaccurate spreads and margins.
A better way to predict electoral outcomes is to project the current year’s turnout based upon recent trend lines, then re-weight the polling data to fit the truly “likely” electorate. When you do that, you will more often than not beat the accuracy of the public polls.
So to better understand the closing rush of the 2014 campaign in New York State, let’s project what turnout will be and where it will come from.
First, on macro turnout, a low turnout in New York’s gubernatorial race would be 4.3 million (2002) vs. a high turnout at 5.3 million (1994). Most New York gubernatorial races fall in the 4.5-4.8 million range (as opposed to 6.5-7.0 million in a presidential race). Given the very low voting in this year’s Democratic primary for governor and the low primary voting in races all around the country this year, I project turnout at 4.1 million on Tuesday. If, however, turnout dropped under 4 million in this gubernatorial election, for the first time in anyone’s memory, that would not surprise me.
In three of the last four gubernatorial elections, the regional shares have been 46 percent of the vote coming from Upstate, 30 percent from New York City and 24 percent from the suburbs. I believe that 46-30-24 percent regional split will probably hold within a percent or two in each region.
Women have consistently cast 53 percent of the general election vote statewide, both in the gubernatorial and presidential cycles. In New York, the Republicans have not faced so much a gender gap as a gender gulch; women voted for Andrew Cuomo by 67-33 percent in 2010 and for Kirsten Gillibrand by a 73-27 percent margin in 2012, according to exit polls. The GOP must find a way to cross over this political ravine among women voters to regain their competitive edge—especially since the minority vote has been growing every four years.
The aggregate minority vote (black, Hispanic, Asian and multi-racial) is now an enduring majority of New York City’s total turnout, and there is also growth in the minority vote coming from both cities Upstate and the downstate suburbs. In 2010, the minority vote rose to 29 percent of the statewide electorate and Democrats carried it by over 4 ½- to-1. If the recent growth trend continues, the minority share will hit 30 or 31 percent next Tuesday. The registration base among minority voters is there for the minority share to hit a full third of the statewide electorate, but 30 or 31 percent seems a more prudent projection for this year.
The white Catholic vote usually weighs in at 33-35 percent of the statewide general election vote and therefore stands astride the ideological center of the State as a veritable colossus: Fully half of November’s vote from the downstate Suburbs and over 40 percent Upstate is cast by white Catholics.
In terms of ideology, about 30 percent of New York voters are self-described liberals, versus a full third who are self-described conservatives, and just shy of the 40 percent who consider themselves moderates. Consequently, the melding of moderate and independent voters usually determines the victors of New York State’s general elections.
Overall, New York is a Democratic state, not a liberal one in general elections. The Democrats’ real edge comes from the party’s 3.1 million voter registration advantage over Republicans, since conservatives narrowly outnumber liberals in the state’s voting booths.
If you distill all these figures down the equilibrium point for New York State elections has been, will be next Tuesday, and will remain for the foreseeable future, moderate white Catholic women outside of New York City. As these women go, so goes the State of New York politically.
Who When the final public polls of likely voters come out on the eve of the election, check their samples by region, race, religion, gender and ideology, before betting a lunch or dinner on either the outcome or the ultimate margin of victory.
Bruce N. Gyory is a Political and Strategic Consultant at Manatt, Phelps & Phillips LLP and an Adjunct Professor of Political Science at SUNY Albany.
NEXT STORY: Following in the Statue of Liberty’s Footsteps