Tuesday, November 4, 2014

Some thoughts going into Election Evening

Here are some random thoughts on the US and NC elections prior to the polls closing at 7:30 (in the Old North State).

1. NC's US Senate Race could be a nail-bitter throughout the night.

The term 'constantly consistent' could be used to describe how the contest between Kay Hagan and Thom Tillis has been since the early summer.  The polls leading up to today have shown a race within the bounds of margins of error:


Based on the mid-term election fundamentals in North Carolina, this should be (by all accounts) a lean-GOP seat this year, but I think the level of polarization and competitiveness that North Carolina experiences in presidential years has bled over into mid-term years (more on both the lean-GOP and competitiveness in a moment).

The result is that most predictions are for a pure toss-up among those analysts (Charlie Cook, Stu Rothenberg) who use a variety of quantitative and qualitative measures, while the 'modeling' folks (538, Washington Post, HuffPost) are more likely to show a lean-Hagan tilt to the race.

It is still up in the air as to whether the U.S. Senate will need the N.C. race to determine whether the GOP claims control of the upper chamber of Congress or not. My thinking right now: there are many pathways (and a few obstacles) that the GOP can have to claim the 6 additional seats to capture majority control of the Senate. Both North Carolina and New Hampshire are probably more "canary in the coalmine" indicates of how the GOP may fare: if NC and NH remain in Democratic hands, then the Republicans will need some combination of red-states, such as Arkansas, Alaska, Louisiana (most likely to go to run-off) and then one competitive purple state, say Iowa or Colorado, to make their numbers work--combined with holding both Georgia and Kansas, two states they probably hadn't bet on contesting.

2. Do we continue to see a polarized electorate?

Based on research from the 2012's general election by the American National Election Study, the traditional electorate looks something like this in a presidential year:


Granted, this isn't a presidential election year, so mid-term electorates generally become more Republican, more white, and older in comparison.

But when it comes to the polarization effect, the classification that one identifies with has a strong connection to one's voting behavior.


Between strong partisans to leaning-independents, the likelihood is that they will vote for their party 85 to 99 percent of the time.  Only "pure independents"--usually less than 15 percent of the national electorate--are the swing voters.

And here in North Carolina, we see a similar, albeit more "center-right" when it comes to independent voters, in voting behavior:



So we have an idea of how these voters will behavior when they show up--that's the next big question, though.

3. Who Shows Up?

In North Carolina, the lean-GOP factor in mid-terms are quite noticeable. Granted, the rise of North Carolina's unaffiliated voters has been quite dramatic over the past few election cycles:


But the question remains: do unaffiliated voters show up at the same rate as partisan registered voters?


With the marked boxes of the past two mid-term elections (2006 and 2010), registered unaffiliated voters don't show at the same level of partisan registered voters.

What this tells us is that the electorate, in past mid-terms in North Carolina, should lean to the right, in comparison to the pool of eligible, registered voters overall.


So, what might we see this year in terms of an electorate? Well, since we already have some votes cast (1.1M+ in early ballots), we see something that may appear "out of the ordinary" when it comes to mid-terms in the Old North State:


Registered Democrats are 48 percent of the early in person ballots that came in with 7 reduced days, while registered Republicans were 32 percent and registered unaffiliated voters were 20 percent.  This is much more in line with a presidential year percentage basis of early in person ballots than a mid-term electorate, which is typically 45 D/37 R/18 U.

Among those 2014 voters who cast in person early ballots and how they voted in 2010:


The notable thing is that one third of registered unaffiliated voters and one quarter of registered Democrats did not participate in 2010, either by not being registered, not living in the state, or simply not voting.

4. But even if these folks do show up to cast ballots, will they have anything in North Carolina beyond a competitive U.S. Senate race?

Not likely.

Nationally, the percentage of competitive U.S. House of Representative races has declined substantially, to basically a hand-full of races that aren't in one camp or the other.


In North Carolina, we mirror that trend with 13 non-competitive congressional districts, and with only one flipping tonight due to the current incumbent (Democrat Mike McIntyre in the 7th) not running for re-election.


This is due partly to redistricting and gerrymandering, and due partly to voters becoming locked into voting for their party candidates both at the top and going down the ballot.


This impacts even the North Carolina General Assembly, with the connection between how a legislative district voted for Obama and how the district voted for the Democratic legislative candidate.


So, in the end, the predictions I would make are the following:

  • Republicans retain control of the U.S. House of Representatives.
  • Republicans gain majority control of the U.S. Senate, but the pathway there is unclear. I suspect Alaska, Arkansas, and Colorado flip to make the 51 needed, and there may be some other states that flip as well--most likely Iowa, Louisiana (in a December run-off). If Kansas goes independent, then the GOP needs Georgia, either tonight or in January, to build on the numbers.
  • North Carolina's Senate race may come down to a 2-3 percent margin of victory; right now, it seems like it could stay Democratic, but no bets as it is a coin-toss.
  • Republicans continue to have majority control of both chambers of the North Carolina General Assembly: biggest question is, can Democrats break the 3/5 supermajority to crack the veto-override numbers in one or the other? If Democrats can, slight favor to doing so in the House rather than the Senate.
  • With GOP control of Congress, expect more partisan gridlock between both ends of Pennsylvania Avenue. And it won't be just in Congress as well--one of the key questions out of the National American Election Studies from 2012 was a question, "is there anything you like about the Democratic or Republican parties?" Sorting by partisanship and identification, you get this:

Partisans love their party, don't like the opposition, and even independent leaners are mirror images of each other.  Pure independents--well, a 'pox on both parties' is best to describe them.

Oh, and one other prediction: we won't get through Wednesday before someone asks "so what about 2016's presidential election?"


Monday, November 3, 2014

The Day Before Election Day, A Few Readjustments to NC's Early Voting Numbers

We've seen a slight revision in the numbers of NC early votes coming in today right before the big day on Tuesday.

The early ballots--civilian mail-in, military mail-in, in person one-stop, and overseas mail-in--are 1,191,162, with the vast majority--1,097,563--come in via in person one-stop balloting.

Of these 1.19 million ballots, 1,155,666 have been accepted and recorded as votes for tomorrow's election, with 35,496 ballots remaining outstanding that could come in via mail or in-person, due to sick or disabled voters having until 5 PM Monday to return request (corrected) their ballots.

So we will probably see some further slight movement, especially with mail-in ballots coming in over the next few days. 

As of today's numbers regarding those who have cast early in person ballots, the composition of the electorate remains the same as we've seen over the past few days: 


Cumulatively, registered Democrats are 48.5 percent of the ballots cast via in person early voting, registered Republicans are 31.1 percent of the ballots cast, and registered unaffiliated/Libertarian voters are 20.4 percent of the early ballots cast in person.  

These numbers are all significantly above 2010's numbers, which may be due to the competitive nature of the U.S. Senate race, the anger and resulting mobilization of Democrats against the GOP's state government, and the influence of $100 million into the state's airwaves.

For registered Democrats, they saw an additional 106,000 voters cast early in person ballots, registered unaffiliated voters saw 192,000 more voters, and registered Republicans saw 16,000 more voters than in the last mid-term election.  

While these total numbers of in person early ballots doesn't match the 2012 presidential year numbers, the proportion of this electorate is more similar to a presidential year than a traditional mid-term year in North Carolina.

The key, though, is that this year's mid-term election isn't anything like 2006 or 2010 in North Carolina, so we may have to deal with a new ballgame when it comes to 2014.

White voters are 71 percent of the in person early ballots cast, while black voters are 26 percent of these same ballots. That is a closer ratio to a presidential year performance than the 2010 performance by racial groups.

Among the party registration and how these 2014 in person early voters participated in the 2010 mid-term elections:


The key here is the significant plurality of registered unaffiliated voters (33 percent) and registered Democrats (25 percent) who did not participate (either due to the fact that they weren't in the state, weren't registered, or simply didn't vote) in 2010.  

Let the actual Election Day activities begin! GO VOTE! 

Sunday, November 2, 2014

NC's In-Person Early Voting Comes to An End

We have the final totals for the vast majority of early votes coming in for North Carolina's general election; a small number of mail-in ballots will continue to trickle in, but the numbers point to a substantial turnout by North Carolina voters through the early voting period.

Of all ballots requested:


Overall, using both the mail-in requests (including those ballots that have been requested and those returned and accepted so far) and the in-person early voting method, North Carolina has the potential of seeing 1,192,190 votes cast before Tuesday's election.

Of all ballots accepted as votes:

Out of this 1.19 million votes cast, 1,155,131 ballots (both by mail-in and in-person) have been accepted as votes for Tuesday's election counting, with the following breakdowns by the different methods of balloting:


Among these accepted mail-in and in-person ballots:

  • registered Democrats are 47.6 percent
  • registered Republicans are 31.9 percent
  • registered Unaffiliated voters are 20.3 percent
  • Women are 54.2 percent
  • White voters are 71.6 percent
  • Black voters are 25.0 percent

Of all accepted ballots that were cast in-person:


Out of the accepted ballots that were cast in-person by North Carolinians, 1,097,560 have been recorded for votes on Tuesday.  This represents 121 percent of the 2010 numbers on the last day of early in-person voting in 2010.  The cumulative totals for these accepted in-person early votes are:



  • registered Democratic voters cast 48.5 percent of the in-person accepted ballots, for a total of 532,026 ballots, representing 125 percent of the final day Democratic totals in 2010
  • registered Republican voters cast 31.1 percent of the in-person accepted ballots, for a total of 341,523 ballots, representing 105 percent of the final day Republican totals in 2010
  • registered Unaffiliated & Libertarian voters cast 20.4 percent of the in-person accepted ballots, for a total of 224,011 ballots, representing 145 percent of the final day unaffiliated/Libertarian totals in 2010.
  • Female voters ended up casting 54.1 percent of the in-person accepted ballots
  • White voters are 70.8 percent of the total in-person accepted ballots
  • Black voters are 25.8 percent of the total in-person accepted ballots
Saturday's daily total of 103,128 for accepted in-person ballots was:




  • registered Democrats: 48.5 percent
  • registered Republicans: 30.4 percent
  • registered Unaffiliated/Libertarians: 21 percent
  • Women: 53.5 percent
  • White voters: 65.4 percent
  • Black voters: 29.9 percent
The trend line in comparing the daily cumulative totals of in-person accepted ballots against the numbers in 2010, as measured by the days out from the election, show the sizable performance of registered Democrats and registered Unaffiliated voters over their numbers from four years ago:


Finally, the voters who have participated in this year's in-person early voting and comparing their voting behavior in 2010's mid-term election shows a significant number of registered Unaffiliated and Democratic voters showing up who didn't cast ballots four years ago:


Additional Analysis (as of 2 PM):

In looking at the voters who were registered to vote in 2010 but didn't vote in the mid-term four years ago, but did show up to cast an early ballot this year, we see some interesting numbers that could give us a clue on the ground game and interest level among the different groups of voters.

Among these 75,616 voters who cast 2014 in-person early ballots and were registered to vote in 2010 but didn't vote in 2010:

  • 40,986 are registered Democrats, representing 54 percent of these voters
  • 17,892 are registered Republicans, representing 24 percent of these voters
  • 16,597 are registered Unaffiliated voters, representing 22 percent of these voters
Among each party registration in terms of race:


Among registered Democrats who cast 2014 in-person early ballots and were registered in 2010 but didn't vote in that year's mid-term election, 54 percent are black voters, with 42 percent white.  Among registered unaffiliated voters, 77 percent are white, while 14 percent are black voters.

Additional Analysis (as of 5 PM):

Among the 2014 NC in-person early voters who were registered in 2010 but didn't vote that year, fifty percent of these voters were in twelve counties (in order of the largest number of total voters): Mecklenburg, Wake, Guilford, Cumberland, Forsyth, Buncombe, Gaston, Durham, Union, Iredell, Pitt, and Catawba counties.

For registered Democrats, half of their 40K votes came from Mecklenburg, Wake, Guilford, Cumberland, Forsyth, Durham, Buncombe, Gaston, Pitt, Union, Wayne, and Orange counties.

For registered Republicans, half of their nearly 18K votes came from Mecklenburg, Wake, Guilford, Union, Gaston, Iredell, Randolph, Buncombe, Catawba, Forsyth, Cumberland, Davidson and Wayne counties.

For registered unaffiliated voters, half of their 16K votes came from Mecklenburg, Wake, Guilford, Buncombe, Forsyth, Cumberland, Union, Durham, Gaston, Orange and Iredell counties.

For all the NC Counties, here are their numbers of 2014 in-person early voters who were registered in 2010 but did not vote in 2010:


DEM              LIB REP   UNA   TOTAL
MECKLENBURG 5481 15 1558 1952 9006
WAKE     3591 22 1250 1912 6775
GUILFORD 2806 11 910 815 4542
CUMBERLAND 1679 2 434 485 2600
FORSYTH  1551 3 520 496 2570
BUNCOMBE 1109 6 530 704 2349
GASTON   1002 4 690 412 2108
DURHAM   1482 4 161 421 2068
UNION    704 0 702 463 1869
IREDELL  633 0 578 378 1589
PITT     868 1 335 265 1469
CATAWBA  485 3 521 329 1338
WAYNE    691 1 388 184 1264
NEW HANOVER 603 4 293 357 1257
ALAMANCE 608 3 321 244 1176
RANDOLPH 381 2 558 228 1169
ORANGE   643 7 117 384 1151
CRAVEN   562 3 254 218 1037
NASH     517 1 310 173 1001
CABARRUS 452 1 305 226 984
ONSLOW   398 0 326 240 964
ROCKINGHAM 543 0 245 167 955
JOHNSTON 400 3 295 248 946
DAVIDSON 325 1 390 185 901
BURKE    422 1 232 220 875
ROWAN    367 0 301 182 850
CLEVELAND 529 0 165 112 806
HENDERSON 237 2 242 261 742
BRUNSWICK 280 1 197 209 687
ROBESON  506 1 44 79 630
MOORE    262 2 209 145 618
CALDWELL 208 1 277 126 612
RUTHERFORD 301 0 198 103 602
FRANKLIN 362 1 122 114 599
LENOIR   391 1 106 96 594
CARTERET 226 0 203 164 593
SCOTLAND 399 0 64 128 591
LEE      353 0 140 97 590
CHATHAM  301 0 102 179 582
GRANVILL 392 0 78 108 578
HARNETT  336 1 148 91 576
EDGECOMB 485 0 43 41 569
SURRY    242 1 216 98 557
LINCOLN  217 1 196 113 527
WILSON   354 0 114 59 527
BEAUFORT 319 2 110 91 522
HALIFAX  376 1 32 79 488
VANCE    348 1 63 62 474
SAMPSON  283 1 103 67 454
STANLY   189 2 155 106 452
HOKE     292 2 44 79 417
PERSON   242 1 79 91 413
RICHMOND 275 0 62 73 410
PASQUOTANK 212 1 56 79 348
PENDER   175 3 83 81 342
HAYWOOD  182 1 56 85 324
TRANSYLVANIA 129 1 67 102 299
WILKES   88 0 144 65 297
COLUMBUS 214 0 36 42 292
ALEXANDER 112 1 114 64 291
DARE     117 1 71 80 269
STOKES   114 1 88 51 254
WATAUGA  90 3 49 97 239
MCDOWELL 77 1 93 60 231
DAVIE    66 0 97 60 223
WARREN   185 0 12 21 218
MACON    71 0 59 68 198
BLADEN   129 0 25 36 190
JACKSON  102 0 39 44 185
BERTIE   146 0 12 23 181
PAMLICO  106 0 37 36 179
ANSON    151 0 10 14 175
DUPLIN   118 0 36 18 172
MARTIN   119 0 21 27 167
MADISON  85 1 36 40 162
CASWELL  118 1 18 23 160
ASHE     47 0 52 41 140
POLK     37 0 44 58 139
SWAIN    72 1 30 34 137
GREENE   85 0 14 32 131
YANCEY   55 1 42 30 128
CHEROKEE 34 0 56 30 120
ALLEGHANY 63 1 29 25 118
MITCHELL 19 0 64 34 117
MONTGOMERY 68 0 19 25 112
WASHINGTON 92 0 9 11 112
CHOWAN   65 0 26 16 107
NORTHAMPTON 89 0 6 8 103
YADKIN   27 0 54 19 100
HERTFORD 64 0 7 18 89
PERQUIMAS 49 0 15 14 78
CAMDEN   31 0 20 24 75
CURRITUCK 24 0 19 27 70
CLAY     22 1 14 27 64
GRAHAM   19 0 21 17 57
AVERY    7 1 33 14 55
JONES    27 0 20 8 55
GATES    43 0 2 7 52
TYRRELL  20 0 1 3 24
HYDE     13 1 0 0 14