The vaunted Obama Ground Game has been slain … at least for one day in Washoe County, Nevada. Republicans outpaced Democrats in ballots cast on day 3 of early voting in the 2nd most important county in Nevada with Clark County clearly the top dog. Washoe made up 19% of the Nevada vote in 2008 and was carried by President Obama 55 to 43 — a total of 22, 791 votes. If Mitt Romney is going to win Nevada he must flip Washoe back into the GOP column which was won by George Bush in 2004, 51 to 47. If he can not do this, there is little reason to believe he will achieve success in the far more difficult terrain that is Clark County. Early voting was the secret weapon of the Obama 2008 campaign and in Washoe alone 11,987 more ballots were cast by Democrats than Republicans during this time which would have accounted for 52.6% of his final vote margin.
We will stipulate up front, when it comes to ballots being cast versus actual votes, there is a leap of faith involved. Neither candidate will secure 100% of their base but for analysis purposes we will hold this as a constant. We do know the candidates get over 90% of their party’s vote so for simplicity we may refer to a ballot cast as a vote despite the very real crossover vote as well as the Independent vote, both of whom broke decidedly for Obama in 2008.
That said, the early vote in Nevada has been a tale of two elections. The GOP in 2008 was nowhere to be found in Nevada and it showed in early voting as well as the election day ballot box. Not only did President Obama achieves an early vote advantage in Washoe County, on election day also outpaced John McCain in ballots cast that day. 2012 is a completely new ballgame and Washoe County is changing the narrative. In just the first three days of early voting, Washoe shaved 3052 ballots off the Democrats advantage versus 2008. That’s over 25% of Obama’s entire advantage in the county during the last election. And worse for Obama, the trend from 2008 was a slow Republican start that steadily gains steam while the Democrats start strong and fade into the close.
This chart shows the 2008 % makeup of the two parties ballots cast versus the total over the 14 days of early voting:
As you see, there is an aggressive convergence between the two parties heading into the final day of early voting. Democrats start strong but fade while Republicans gain steam. And this was in a campaign where an underfunded John McCain had largely ceded the state in these final weeks.
This time the GOP is under no such disadvantages and it already showing in only the first days of early voting. For the Monday October 22 early voting, Republicans cast more ballots than Democrats in Washoe County — an unheard of proposition considering the alleged indomitable strength of the Obama ground game as well as the disarray of the Nevada state GOP party. It looks like the shadow party, Team Nevada, has a few tricks up their sleeves…Here is how the same chart looks after 3 days in 2012:
Republicans begin the early voting season at a higher point than 2008 eating directly into the Democrat advantage as Independents were mostly constant election-over-election. Opening with a 10.3 percentage point (pp) gap, the Republicans quickly closed the deficit taking the lead on Day 3 of early voting neutralizing Obama’s great advantage that went a long way towards him carrying Nevada in 2008. This is a watershed moment for Nevada Republicans who have seen their fair share of let’s say “unique” candidates as well as office holders recently. If the GOP can keep anywhere near this pace of trend for the early voting season, President Obama may be in far deeper trouble in Nevada that his over-generous internal models predict.
The Washoe Model — predictive value low (at this point), illustrative value high
We previously created a crude model for early voting in Clark County and thanks to a helpful reader are able to construct a similar model using Washoe’s 2008 early voting numbers. Rather than reinvent the wheel, we simply mimicked the Clark model for Washoe based on the first day’s data. Using such limited data makes the models far less reliable for predictive purposes but still valuable to illustrate comparisons between the two elections. The forecasting value is inherently low based on only one election’s data (2008) and one day’s voting (2012). It is like extrapolating a heavily Democrat precinct’s data nation-wide in a Presidential election. It tends to lead to wrong conclusions, right President Kerry? But the model is very helpful for tracking purposes because deficits can signal signs of trouble while election-over-election advantages can indicate one party is making waves that could upset an election’s results. Over the next day or two we will tweak both models to make them more predictive than simply illustrative.
Washoe County had some similarities to Clark County in that after a slow start, Republicans gained steadily on Democrats throughout the early voting — only Washoe started from a higher base:
From this we see the average rate of change over the 14 days of early voting is a 3.85% average daily increase in the Republicans % of the Democrat vote almost reaching parity by election day. We use this as a basis for our growth assumption of Republican turnout relative to Democrats in our 2012 model. Although I use conservative estimates in my models to help offset the low predictive value we see quickly the Republican outperformance causes problems for the model on only day 3.
Democrat turnout estimate
Looking at the 2008 actual early vote #s, Democrat turnout after Day 1 in Washoe County averaged 86% of the day 1 result:
Plugging the 3.85% Republican growth assumption and the Democrat turnout at 86% of its first day total, our model is as follows:
If you look at the left-bottom row you will see that Day 2 of the model came close to forecasting the Republican % of the Dem vote. The model expected the GOP to improve to 82% of the vote and the actual results were 81% of the Democrat vote.This success wouldn’t last long.
From day 1 to day 2 we see a sharp drop-off in turnout. This was unsurprising and consistent with 2008 since it falls on a Sunday. The major difference was Day 3 when 2 important things happened — one predictable, the other not so much. The GOP kicked it into overdrive and increased their turnout on Monday over 1000 votes. This was consistent with 2008 when the GOP increased its Day 3 turnout nearly 1200 votes. What didn’t happen was a bounce-back by the Democrats . After a big splash on Day 1 and the expected Day 2 drop-off, Democrats only improved their turnout only 572 ballots day-over-day, dramatically lower than 2008 when their Day 3 turnout increased nearly 2200 ballots. This lead to Republicans outperforming the Democrats in early voting for the first time in either election — a preposterous proposition were you to predict turnout based solely on reputation.
Although aggregate numbers are up election-over-election, the poor day 3 turnout puts Democrats far below their 2008 pace while Republicans outstrip their 2008 pace by leaps and bounds which is where the model is of some help. For Democrats to proportionally meet their 2008 turnout, they need to average 3985 ballots cast every day until election day. Making matters worse for the Democrats, their turnout #s they were surprisingly steady with the noted exceptions of Sundays. Over the 7 subsequent regular days following the Sunday drop-off Democrat turnout only fluctuated between 3100 and 3600 each day before ramping up in the finals days of the early vote. So far they have woefully failed to meet that level in the two days since the early voting kickoff. If the Democrats are already performing below that band, they could be in for some trouble in the coming days if these new turnout results form new upper or lower bounds for them.
Republicans, on the other hand have nearly doubled their 2008 turnout over these three first days of early voting and clearly have momentum and history on their side. They have already cut 3052 ballots into the Democrats lead and if past in prologue, we can expect even more surprises in early voting returns from the Silver State. As I said up front, the above model is more illustrative than predictive at this juncture. As you can see it predicts net gains for Republicans of only 2727 ballots cast but that was based on the GOP not catching up to the Democrats for another 4 days. Unfortunately for my crude model, but good for the GOP, in 2012 the Nevada Republicans started strong and shocked most every election watcher surpassing President Obama’s vaunted ground game on only the third day. If this is a sign of things to come, Nevada’s 6 electoral votes may just end up in Team Romney’s column.