The New Wave That Will Leave The GOP Underwater In 2018

(Editor's note: This is an opinion piece posted by a Patch reader. If you'd like to post on Patch yourself, find out how here.)

Republicans who are honest with themselves, know that 2017 wasn't a good year for the GOP. 2018, in a few short months, has been considerably worse, and by the November mid-term elections the damage could intensify by several magnitudes.

The potential for this impact is real, the odds of it happening are great, and every piece of reliable data backs this up, as I will demonstrate once we progress further into the article. That's for the GOP to worry about.

On the Democratic side, however, there are concerns and assumptions that are not so easy to track. Namely, the concern about how to ride this electoral "Blue Wave" without dying at the shore. And the assumption, not shared by all, but by a number of individuals in the Democratic establishment locally and nationally, that the best way to win back House and Senate seats is to appeal to older, white, right-leaning voters, many of whom once identified as Democrats (and might still, on their registration), with a politically moderate platform.

Here in New York's 1st Congressional District (NY CD-1), there is a mix of excitement about greater Democratic turnout, and fear that incumbent Republican Lee Zeldin, coming off a large re-election victory in 2016, will win again unless Democrats figure out how to appeal to Trump voters.

I have read a few different local attempts at delving into data analysis on this topic, from guestimating Democratic turnout based on past party-line vote tallies at the ballot (this is not how turnout is estimated, because it's a terrible basis for turnout analysis, as Unaffiliated Voters have to vote for candidate on a party line, and many voters identify with a party that they are not registered with), to looking at 2010 Census data to determine which demographics might have the largest impact in the upcoming election (again, not how proper forecasting is done).

To date, there has not been a sound analysis of how a Democratic wave election will impact NY CD-1, in terms of who the "surge voters" will be that form the wave itself, and carry Democrats to victory.

So what is really happening in New York's 1st Congressional District?

Perhaps the better question is, what is happening in elections across the nation, what it the best available polling and forecasting telling us, and what does that forebode for NY CD-1? Keep in mind a few bedrock principles of vote and turnout forecasting:

• In any given voting district, year to year, no two voting populations are the same. The voting population that turned out in CD-1 in 2008 is not identical to the one that turned out in 2010, or 2012, 2014, etc. These populations are siblings, not clones: related, but with packed with distinct variations that can be, at times, significant.
• Mid-term elections, especially in unstable political environments, are often rebukes of the Presidential election results. The last three mid-term “wave elections” in 1994, 2006, and 2010, were the result of the Congressional minority party’s voters feeling threatened by the ambitious policy designs of the majority party’s President. When White House scandal and corruption is involved as well, as was the case in 1994 and 2006, opposition party turnout can be intensified. In all three Presidential elections prior to those mid-term waves, the winning party controlled both houses of Congress and was poised to control the levers of government for years to come, until the mid-term happened and suddenly, they weren’t.
• Pollsters and pundits have a tendency to over-correct for the results of the last major election. Nobody likes being wrong. A mediocre pollster will note that during a Presidential election, one particular voter demographic only had 30% turnout, and will control for exactly that turnout percentage when they conduct a poll for the next mid-term. A good pollster will watch the trend lines of state, local, and special elections leading up to the next election, recognize a higher turnout pattern, and adjust their poll accordingly. Because a good pollster understands that forecasting is not about understanding what happened in the last election, but predicting what will happen in the next one.
• Not all polls are created equal. The best polls use a combination of landline, mobile phone, and online surveys. Polling “Likely Voters” will give a different result than polling “Registered Voters” or “Adults”. This is why, when evaluating polls, it’s best to look at an average of all polls rather than cherry picking the one whose methodology fits your narrative.

“Wave elections” are loosely defined federal elections where one party gains a significant amount of seats in the House and/or Senate, enough to overturn a majority or create a supermajority. When a wave election happens, it often coincides with seat changeover at the state and local levels as well.

To gauge the potential for a wave election, forecasters look at a combination of Presidential approval ratings, the number of Senate seats the majority is defending, and the Generic Congressional Ballot poll, which polls voters nationally on who they generally prefer in Congress, Democrats or Republicans. Additionally, they will monitor state and local elections, and special elections, leading up to the federal election.

If a sitting President’s approval ratings are under 50%, and their party in Congress is at a Generic Ballot poll deficit of 5 points or more, that indicates the potential for a wave election.

As it stands, the 18-day polling average of President Trump’s approval rating sits at 42.1%, and has stayed firmly in the range of 38-44% over the past year. In the Generic Ballot poll, Democrats currently have a +5.8% point average lead, which has fluctuated in the range of 5-13% points for months. Just last week, their polling average lead was over 9% points.

What is even more telling, is the favorable voting trend towards Democrats in elections all over the country since 2017.

Daily Kos has compiled and maintained a spreadsheet of all state and local special elections in 2017 and 2018. The spreadsheet records the vote margin between the Democratic and Republican candidates, the margin between Donald Trump and Hillary Clinton in the same district, and the margin between Barack Obama and Mitt Romney. It then lists the margin differential for the Democrat in the local race, compared to how Clinton and Obama fared in that district.

For example, in Kentucky’s local HD-89 race last month, the Republican beat the Democrat 67%-33%, for a Democratic loss margin of -34%. In 2016, Donald Trump beat Hillary Clinton in that same district 79%-17%, or a Democratic loss margin of -62%. The margin differential between the 2018 local race and the 2016 Presidential race in this district, however, was a 28% improvement for the Democrat. Compared to President Obama’s performance against Romney in the same district in 2012, it was a 6% improvement.

Essentially, the margin differential % measures how Democrats have performed against Republicans all over the country in 2017 and 2018, compared to how the Republican and Democratic Presidential candidates performed in the 2012 and 2016 federal elections in those specific districts. This includes special elections such as the PA-18 House seat won by Conor Lamb, and the AL Senate special election won by Doug Jones.

The results have been staggering. In all 2017 special election races, the Democrat improved over Clinton’s 2016 margin by 10%, and improved over Obama 2012 by 7%. In 2018, those margins improved to 24% over 2016 and 12% over 2012, respectively.

For all of 2017-18, the average margin differential improvement has been 13% over Clinton, and 8% of Obama.

That 13% two-year average improvement over Clinton 2016 jives with what the Generic Congressional Ballot poll has been predicting all year, which is a Democratic advantage of between 5-13 points.

In terms of seat pickups, for all 2017-18 special election races, Democrats flipped 42 seats, and Republicans flipped 4, for a total Democratic pickup of +38 seats. Again, squarely in the prediction zone of what Democratic and Republican insiders are saying will be a 30-45 seat House pickup for Democrats.

Republicans in Congress know what’s coming - 40 Congressional Republicans have announced their retirement ahead of the mid-term elections, almost twice the number of Republican members who retired before the 2006 Democratic wave.

For a swing district like NY-1 represented by a firmly right-wing Trump Republican like Lee Zeldin, who himself succeeded a 12-year liberal Democratic incumbent in 2014, this climate bodes well for Democrats. Famed political analyst and election forecaster Larry Sabato has rated Zeldin’s seat as “Leans Republican”, an ominous distinction in a year when many “Safe Republican” seats are likely to flip Democratic.

If Zeldin’s Democratic challenger in NY-1 were to match the 2017-18 margin differential improvement over Clinton 2016, the Democrat would beat Zeldin by 5 points. If the 2018 Democrat in CD-1 improves by the margin of the Democrats’ Generic Ballot advantage (roughly 8 points) over what CD-1 Democrat Anna Throne-Holst did in 2016, they would finish with just under 50% of the vote.

For perspective, look at the two previous mid-term wave election results in CD-1. In 2010, a Republican wave year, Republican candidate Randy Altschuler improved by 8 points over the Republican vote in 2008 (ironically, that Republican was Zeldin), losing to incumbent Democrat Tim Bishop by less than 700 votes. The actual generic ballot advantage for Republicans nationwide that year was +6.8.

In 2006, a Democratic wave year, Bishop improved upon his own victory in 2004 by 6 points. The actual generic ballot advantage for Democrats in 2006 was +7.9.

Zeldin’s 17-point win margin in 2016 will not matter in 2018, just as Bishop’s 17-point win margin in 2008 didn’t matter in 2010, when he came within several hundred votes of losing.

If the generic ballot polling stays in the 5 to 9 point range in favor of Democrats, Zeldin will come within a point or two of winning or losing. If the generic ballot ticks up to the 10 point range or higher, the smart money would be on Zeldin being voted out of office. As it stands, a really, really good outcome for Zeldin would be to fall within 51-52% of the vote.

The more likely scenario is that this race comes down to a photo finish that is decided by a few thousand or even a few hundred votes, no matter who Zeldin’s challenger is.

But therein lies the difference between winning and losing. A Democratic challenger that does not maximize turnout and hopes for the generic ballot to bail them out, will be the one who loses by a percent.

Defeating Zeldin will require a challenger who can not only deliver left-leaning registered Democrats to the polls in great numbers, to counter and exceed Zeldin’s base turnout, but can also translate favorable national voting trends at the local level to put themselves over the top.

So what do I mean by “translate favorable national voting trends”?

When a district radically shifts its vote from an incumbent to a challenger two years later, the assumption people make is that the voters who represented the margin of victory the first time around, simply changed their minds the second time.

The reality is that voting populations are much more volatile than we assume. Understanding the trends and forces that bring one demographic to the polls and causes another to stay home, or switch their vote, causes inactive voters to become suddenly active or motivates first-time voters to participate, is key to election forecasting.

Because proper exit polling is not done in most Congressional districts, political scientists and election analysts rely on things like the generic ballot poll, state and local election trends since the prior federal election, and state-wide exit polling from those races (where it exists), along with opinion polls, to paint a broad picture of the likely voting population for the next mid-term or general election.
What does the available data tell us about Democrats’ dramatic improvement at the polls in 2017-18?

For the first time, Millennial voters (generally in the 18-34 age group) will outnumberBaby Boomers at the polls, starting this year.

A recent study conducted nationally by Cosmopolitan/SurveyMonkey that interviewed 4,706 adults ages 18-34, found that 68% plan on voting in the 2018 mid-term elections, and 60% intend on voting in their party’s primary. By contrast, only 25% reported voting in the 2014 mid-terms.

The information in this poll reflects the recent youth voter surge documented in the 2018 Alabama Senate special election, the 2017 Virginia Governor’s race, and the 2018 Texas state primaries. In the case of Virginia, youth voter turnout more than doubled in 2017 compared to previous years, specifically in suburban areas that are demographically similar to Suffolk County.

And whereas Hillary Clinton won both the 18-29 and 30-44 age demos over Trump with 55% and 51% respectively, the 18-34 demo has been going for Democrats in the past year by over upwards of 60%-70%. These are not just registered Democrats, but Unaffiliated voters who behave like Democrats at the polls.

The Cosmo/SurveyMonkey study also found that 53% of the young respondents said they were more likely to vote in the 2018 mid-term election as a result of Donald Trump’s victory in 2016. This age group reports strong disapproval levels of the President by 52%, far higher than the general electorate.

Now let’s put the microscope on NY CD-1.

Since November 9, 2016, 11,827 people have registered to vote in CD-01. Of those new registrants, 2,501 (22%) are Republicans, 3,744 (32%) are Democrats, and 4,815 (41%) are Independents or Unaffiliated. This is excellent news for Democrats, but the more striking number is that of these newly registered voters, 8,325 (70%) are between the ages of 18-36. [These numbers and all other CD-1 voter data citations were provided through the VoteBuilder campaign software, which uses data sets from the state Board of Elections]

As demonstrated, these voters will trend substantially to the Democratic side by upwards of 55% (at the low end) to 60% or higher.

In total, there are 103,569 voters in CD-1 ages 18-34. If that demographic reaches 60% turnout (62,141), we’re potentially looking at 25,000-30,000 newly activated and/or registered voters being injected into the 2018 voting population in our home district, 15,000-18,000 of those same voters supporting the Democratic candidate.

If that isn’t alarming enough for Lee Zeldin, consider this: Kristen Soltis Anderson, a Republican political analyst and researcher, recently penned an op-ed in the New York Times which made the case that Donald Trump is driving young voters out of or away from the Republican party. She cited a Harvard study that estimated President Trump has a 66% approval rating among Republicans between the ages of 18-29, far lower than his approval among Republicans of all ages. And of the remaining 34%, immigration and the environment are cited as their biggest points of disagreement with Trump.

There are just over 17,000 registered Republicans ages 18-29 in CD-1. If 34% of those Republicans do not approve of Trump, that means close to 5,700 young Republicans in CD-1, or 1.1% of the electorate, are potential crossover votes for a Democrat who can distinguish themselves as the opposite of Trump/Zeldin on the environment and immigration.

Turnout among black women and hispanic voters has also been up all over the country since 2017, two demographics that trend hard for Democrats. This was specifically the case in not only Alabama but Virginia as well.

Conversely, the much-lamented “Trump Democrats”, who only made up an estimated 9% of total Democrats who went for Trump in 2016 (compared with 7% of Republicans who voted for Hillary), i.e. the “white working class Trump vote” may not be nearly the factor that some in the national and local media have depicted.

Trump’s sub-50% approval ratings in the mid-west region of the country, where white male Baby Boomers strongly support Trump and are demographically similar to Suffolk County’s “Trump Democrat” population, combined with Trump not being on the ballot, suggests that Republicans will underperform with this voting population in the 2018 mid-terms, as they did for the past year and a half in state and local elections.

Simply stated: Democrats may not need Trump Democrats to cross back, but may benefit when they fail to show up at the polls entirely. We’re talking about a statistically small number of Democrats, compounded by the number of female and young Republicans who have been turned off to the GOP by the Trump Administration. The number of voters in both parties who have decided that the party “left them” may balance each other out.

In any case, if a Conservative Democrat (or Liberal Republican) has decided for years that the national party’s branding is a turn-off, there is almost nothing that a local candidate can do to change that opinion. And attempting to “win back” that portion of the electorate, at the risk of endangering turnout among newly active voters, is a bad gamble.

What is the proper turnout strategy, then?

The unifying turnout factor for 2018, without a doubt, will be “anti-Trump.” Any local canvasser who has knocked on doors for a candidate in the past few months will tell you that people’s eyes light up when you mention opposition to Donald Trump. Every piece of election data since 2017 began, will tell you the same thing.

There is no need to complicate what is simple: a significant portion of the electorate wants a chance to strike back at the GOP over the result of 2016, and there aren’t enough voters on the other side who care enough to stop it.

Anti-Trump sentiment will be enough to make the CD-1 race a close contest. Moving to the right on key issues like immigration, ICE administrative warrants, tax cuts, or healthcare, will not win back enough “Trump Democrats” to account for the depression it causes in turnout.

The trick for local Democrats in CD-1 will be to keep the iron hot on the Trump opposition vote, while reinforcing to newly activated voters that they are not “same old, same old” establishment Democrats. People want “new”, so give them “new”. The newly activated “surge voters” of 2018 will trend much further towards the Progressive left than in years past, while establishment Democrats will vote for anyone with a D next to their name.

Surge voters must be targeted by an aggressive, efficient canvassing and digital ad strategy that focuses on top-line messaging (Anti-Trump, “Here’s Why I’m Different”), and date reminders (not knowing when the Primary and/or General election is being held is a huge barrier for newly activated voters).

The progressive grassroots, which has grown locally by the thousands, must have a platform they can believe in, as these individuals will be the evangelists who form an army of canvassing teams. It is not simply a matter of getting them to vote, but to vote, walk, share, and donate.

Finally, the right-leaning independent voter who is not crazy about Trump, may or may not have voted for him, but can’t be swung by progressive ideology. How does a progressive candidate turn this voter? Simple: do not insult this voter’s intelligence by “moderating” to sound more like a digestible vita-gummy version of Lee Zeldin or Donald Trump.

They won’t believe or respect that tactic, so you tell them the truth. You, the Democratic candidate, are running on a platform that might not always conform to their personal belief system. But what the average, non-ideological voter will always agree on, is the need for infrastructure and local services.

Securing funding from Congress for sewers, road repair, education, or inpatient rehab facilities to help stem the opioid epidemic are not Republican or Democrat issues. They are “I’m the type of leader who can get things done” issues, and people will forgive ideological differences if you can prove yourself to be this individual.

A dynamic Progressive platform that appeals to a burgeoning generation of voters, rooted in the age-old principle of delivering the infrastructural goods for your district, is going to be the difference between a CD-1 Democratic challenger who winds up with 49% of the vote, and one that secures 50% plus one.

(Opinions stated in this article are my own and do not reflect that of my employers)