MikeBlog

Showing posts with label 2010 election. Show all posts

16 November 2010

A mostly-final review of Anne Arundel County Question A

I made my projections about Question A here:

...based upon my projected turnout numbers in Anne Arundel County, I think Question A will pass with 51.0-52.3% for, and 47.7-49.0% against. While this is not strictly a partisan issue, we've seen that Democrats tend to be for it (56 for, 37 against), Republicans and unaffiliated voters tend to be against it (44-51), and U (39-54). Since it is a relatively close vote, stronger than expected turnout by either party could tilt the vote either way. Even then, this is the prediction I am least confident about. Small changes in support among members of either party could swing the results either way.

The results were better than I expected:

My least confident pick was never in doubt. There are still a few votes left to be counted, but Question A leads with 55.7% to 44.3%. As I noted above, I think this is another race in which turnout may have had more an impact. Also remember that these projections were based upon two polls in which we didn't have the crosstabs and made some major assumptions based on shaky data. Still, to be within 3-4 points both ways (and of course, on the correct side of the projection), is a good feeling.

Official numbers show Question A passing by 55.52% to 44.48%.

In this post, I tried to project how Question A was doing among early voters; I came to the conclusion that it was split 49.3% for and 44.5% against (if you split the undecideds evenly, you'd come to 52.4% for and 47.6% against). The latest numbers show that in fact 54.64% of early voters favored Question A. So it turned out I was just a couple of points off, but looking back I could have been closer. In my overall projections, pushed 6 of 10 late deciders into the "no" column for (I suspect) no good reason other than to try to find a "safer" outcome rather than a closer outcome. If I had split the undecideds in proportion to the actual vote (hindsight is 20/20, of course), my early voter projection would have been less than 1% off.

Finally, I explained in this post how Arundel Mills voters cast their ballots in the 2008 statewide slots referendum:

Let's take a look at the final certified votes for Anne Arundel County voting precinct 1-22-32 in 2008 on the statewide ballot question:

1-22-32: 1393-904 (60.64%-39.36%)

Now let's also include the results of the three precincts directly east of this one:

1-06-32: 806-453 (64.02%-35.98%)

1-05-32: 570-403 (58.58%-41.42%)

1-20-32: 984-641 (60.55%-39.45%)

For the sake of this post, we'll call these four precincts "Arundel Mills voters." Now let's take a look at the aggregate totals:

Arundel Mills voters: 3753-2401 (61.0%-39.0%)

Anne Arundel County: 149,604-103,814 (59.0%-41.0%)

Maryland 1,482,295-1,043,129 (58.7%-41.3%)

In other words, Arundel Mills voters voted for slots in a greater proportion as Anne Arundel County as a whole, and even more so than compared to voters statewide.

I closed by saying that "...it will be interesting to see what the true level of support and opposition is in these precincts," so let's do just that (these numbers are unofficial, but close to final):

1-22-32 Jessup Elementary School 1774 of 3892 (45.58% turnout), 497 for (28.53%)
1-06-32 Anne Arundel Fire Station #21 947 of 1756 (53.93% turnout) 287 for (31.09%)
1-05-32 Hebron-Harman Elementary School 688 of 1404 (49.00% turnout) 180 for (26.55%)
1-20-32 Hebron-Harman Elementary School 1069 of 2123 (50.35% turnout) 344 for (33.05%)

All together, "Arundel Mills voters" (all voters in these four precincts) voted against slots by a 29-71 margin. Yet, turnout was only 44.48% in these precincts compared to 61.45% county-wide. This county-wide turnout trend would have netted more than 800 votes against slots in these four precincts alone.

Final conclusions: The end result was correct. Projections were off by a few percent, but looking back and eliminating unfounded assumptions (like 6 of 10 undecided voters saying "no") would have gotten us even closer. Still, as I said early on, these projections were based upon two polls in which we didn't have the crosstabs and made some major assumptions based on shaky data. So in the end I am very satisfied with the projections vs. actual results. I'll be sure to take some learning points away from this experience and apply it to the next project.

15 November 2010

A mostly-final review of turnout projections in Anne Arundel County

I made my original projections here:

I am estimating that Republicans will turn out in 1994/2002 levels at 69%. Democrats are certainly less enthused, but turnout in Anne Arundel County has always been higher than statewide. It is difficult to imagine Democrats turning out in better numbers then they did in 2006, but I'll give them the benefit of the doubt and put their number at 65%. With third party and unaffiliated voters typically at 50% or less, this brings our overall turnout to about 64%.

I reviewed the initial turnout numbers here with the caveat that absentee and provision ballots were not yet counted:

If we take 8-10% and presume those as reasonable limits, we find that perhaps somewhere in the range of 208,000 to 213,000 total people voted; this puts overall turnout in the range of 63-64%. These are educated guesses based upon previous elections, and we'll refine them as more comes in, but those numbers have been pretty good indications of what to expect.

Well, the final numbers are in. According to the Anne Arundel County Board of Elections website, 203,475 total votes were cast from a total of 331,098 registered voters, which represented 61.45% overall turnout. The difference is that the 11,682 absentee and provisional votes represented only 5.7% of the total votes cast, as compared to the projection of 8-10%. No partisan turnout numbers yet, but based upon what we have seen, we can expect they too will be a few percentage points short of the projections.

Generally speaking, I didn't nail turnout as close as I would have liked, but ~2.5% off seems pretty close given this was my first time doing any serious analysis of the numbers. I'll have a few more 2010 election-related posts including a final look at the slots voting (comparing final results to projections, and a look at voting by precinct) and a review of statewide turnout.

05 November 2010

Turnout is down in Maryland despite early voting...but that's not the whole story

FOX45 Baltimore is reporting that "[d]espite six days of early voting, turnout in Maryland's general election was five percent less than four years ago." They cite turnout as 52.5% in 2010 as compared to 57.5% in 2006. The story goes on to say this happened despite early voting, and they question whether or not early voting is worth it. However, there is a problem this number because it doesn't include most of the absentee and provisional ballots as you can see from this screenshot:

None of the state's most populous counties (Anne Arundel, Baltimore, Montgomery, Prince George's) nor Baltimore City has even reported their first canvass of absentee ballots, and no jurisdictions within the state have reported any provisional ballots or started their second canvass of absentee ballots.

If you look back at the 2006 numbers, you would find that absentee and provisional ballots represented over 11% of total votes cast statewide. With 3,468,287 registered voters as of the close of registration prior to this election, 52.5% would represent approximately 1,821,000 total votes cast; yet the possibility exists that upwards of 200,000 absentee and provisional ballots will eventually be counted. This would push total votes over 2,000,000 and at 58%+ which would match 2006 turnout (and possibly represent a small increase). But that wouldn't make a news story, would it? :-)

03 November 2010

A review of turnout projections in Anne Arundel County

Here were my projections regarding turnout in Anne Arundel County:

Turnout: During the last four Gubernatorial General Elections, turnout here in Anne Arundel County (as a percentage of eligible voters) about the parties has been the following:

1994 Democrats 69 Republicans 69 Overall 67

1998 66D 65R 63O

2002 66D 69R 62O

2006 64D 65R 61O

That averages to 66.25 Democrats, 67.00 Republicans, and 63.25 Overall. Based upon the overall "enthusiasm gap" in favor of Republicans (and respecting the fact that this gap is probably less evident in Maryland), I am estimating that Republicans will turn out in 1994/2002 levels at 69%. Democrats are certainly less enthused, but turnout in Anne Arundel County has always been higher than statewide. It is difficult to imagine Democrats turning out in better numbers then they did in 2006, but I'll give them the benefit of the doubt and put their number at 65%. With third party and unaffiliated voters typically at 50% or less, this brings our overall turnout to about 64%.

The official numbers aren't yet in, and won't be for a little while due to absentee and provisional ballots. Also, the county is showing 194/197 precincts (or 98.5%) reporting; I'm not sure if this is due to slow counting somewhere, or if the absentee and provisional ballots somehow make up their own precincts (or both). That being said, we can look at the numbers so far and compare them to the projections.

Together with early voting (but not absentee ballots) and Election Day itself, 191,793 of 331,098 (or 57.93%) of registered voters turned out. We won't have absentee numbers yet, but in 2006, 15,930 absentee ballots and 3,558 provisional ballots were cast (all of these numbers count toward overall turnout). Those 19,488 votes represented over 10% of the total 189,150 votes cast (in 2008, absentee ballots represented 8.2% of the total). If we take 8-10% and presume those as reasonable limits, we find that perhaps somewhere in the range of 208,000 to 213,000 total people voted; this puts overall turnout in the range of 63-64%. These are educated guesses based upon previous elections, and we'll refine them as more comes in, but those numbers have been pretty good indications of what to expect. So it looks like overall turnout projections may have been spot on.

I'll have partisan turnout numbers when a better breakdown of the overall numbers becomes available.

Quick and dirty election postmortem (Congress)

Projections vs. results for U.S. Congress:

...Translation: unless these numbers change dramatically toward the Democrats in the next week (and there isn't any indication to suggest they will), next Tuesday will be an electoral bloodbath for Democrats; on the order of 50-60+ seats in the House of Representatives (+39 would take control), and 8+ seats in the Senate. Republicans will regain control of the House and will probably end up with 49 or even 50 seats in the Senate (they'd need 51 to control the Senate since the Vice President could break a tie); although a majority in the Senate is not out of the question.

In regards to the House of Representatives, there are some numbers left to count, but it appears that the Republicans will exceed 60 seats and end up somewhere around 65. To put this in perspective, it will be the most seats won by either party since the 1940s. Bob Beckel and Doug Schoen (both Democrats) were on FOX last night trying to argue that the Republicans should have won 90 seats, and thus 65 would be somewhat disappointing all things considered. Seriously! I caveated this prediction with the possibility of even 70 seats, and the GOP came pretty close to that, outperforming expectations (but just slightly so). This is doubly interesting because most prognosticators (Cook, Sabato, Silver, et al) largely and accurately predicted the wave and its size. In other words, the news is more about results versus mostly accurate expectations, and lacking much difference, the newsworthiness doesn't seem to have as much impact.

In the Senate, the Republicans have picked up 6 seats with two yet to be undecided (CO and WA). The press is trying to spin this as disappointing results for the Republicans, but even if both of those seats end up in Democratic hands, the +6 will be nearly double the average gain of seats in a mid-term election. The Nevada Senate seat (Harry Reid) was really the only race where the GOP showed a lead and lost. CO would be the next closest. Manchin's win in WV is only remarkable because he ran so far away from Obama and literally fired a bullet through the cap-and-trade bill. Seats in CO and WA would bring us to eight, but it seems like getting both of those is unlikely. So Democrats outperformed expectations slightly here. NV goes down as incorrect with the possibility that CO could as well (I had CA, WA and WV already in the Dem column).

Quick and dirty election postmortem (Maryland and Anne Arundel County)

I'll get into more specific numbers later, but let's take a look at projections vs. results here in Maryland and Anne Arundel County:

U.S. Congress: MD-1 (Harris), MD-2 (Ruppersberger), MD-3 (Sarbanes), MD-4 (Edwards), MD-5 (Hoyer), MD-6 (Bartlett), MD-7 (Cummings), MD-8 (Van Hollen, Jr.). The only race that promises to be close is the MD-1 race between Andy Harris (R) and Frank Kratovil (inc. D). I've already chimed in on that race here, but generally speaking, the edge should go to Andy Harris by 2-3 points.

7 for 7 among the Maryland congressional delegation. While the final, official numbers aren't yet in, Andy Harris is currently shown leading (soon-to-be former Congressman) Frank Kratovil 54.6% to 41.5% so I missed the margin there considerably (I'll have some later comments on turnout in Maryland in general, and why this may have had a role in this particular election). This race, however, was not a "major upset" as I heard on the radio this morning. The race has been Harris' to win all along.

U.S. Senate: Like most of the House races, not a close contest. Mikulski (inc. D).

Easiest prediction of the night.

Maryland Governor: Maryland Governor Martin O'Malley has been somewhat immune to the nationwide political environment due to relatively low unemployment (based upon, for example, large numbers of government employees). Add to this the more than 2-1 advantage for Democrats in registration numbers and I think this is just too much for former Governor Bob Ehrlich to overcome. Ehrlich should win Anne Arundel County with at least 55% or so, but that won't be enough to overcome the large advantages that O'Malley receives in the D.C. suburbs. I'm thinking about 53% to 45% in favor of O'Malley over Ehrlich.

Second easiest prediction of the night. I was hoping Ehrlich could prove me wrong and made it a race, but it wasn't to be. Current numbers show O'Malley at 55.8% to Ehrlich's 42.2%. In Anne Arundel County, Ehrlich lead by 54.6% to 43.0%. Prediction-wise, a few points off in the state, very close in Anne Arundel County.

Question A: I've talked in length about this subject, so I won't spend too much time here other than to say that based upon my projected turnout numbers in Anne Arundel County, I think Question A will pass with 51.0-52.3% for, and 47.7-49.0% against. While this is not strictly a partisan issue, we've seen that Democrats tend to be for it (56 for, 37 against), Republicans and unaffiliated voters tend to be against it (44-51), and U (39-54). Since it is a relatively close vote, stronger than expected turnout by either party could tilt the vote either way. Even then, this is the prediction I am least confident about. Small changes in support among members of either party could swing the results either way.

My least confident pick was never in doubt. There are still a few votes left to be counted, but Question A leads with 55.7% to 44.3%. As I noted above, I think this is another race in which turnout may have had more an impact. Also remember that these projections were based upon two polls in which we didn't have the crosstabs and made some major assumptions based on shaky data. Still, to be within 3-4 points both ways (and of course, on the correct side of the projection), is a good feeling. I'll have a full post later dedicated to Question A.

Among the Maryland and Anne Arundel County projections, there were really no surprises, and everything went more or less as expected.

02 November 2010

Some very early turnout numbers (from one single precinct)

In the precinct I was working today (registration is 43.7R, 41.5D, 14.8O), 38.7% of Republicans and 37.6% of Democrats had turned out as of 4PM. It is impossible to know how many people in this precinct voted early because the state doesn't break down numbers that far. Projecting the rest of the day, this looks like about 62% for Republicans and 60% for Democrats. When you add in the percentages for early voting (~2.5% for each party), this comes pretty close, but slightly lower, than the numbers I projected county-wide.

I had heard some stories of lines at other polling places, but the precinct I was working was slow but steady. It was picking up as I left just after 4PM.

01 November 2010

Hedging my bets

I made my House and Senate predictions here, and my Maryland-related predictions here. I'm going to stick with those predictions, but I want to hedge my bets just a little bit and talk about some things that could happen tomorrow, in order of likelihood:

Even more massive House gains. I predicted a 50-60+ seat gain for the Republicans (even then, I was hedging the high side with the + sign), but we could even see 70+. As Gallup noted in their final generic congressional ballot poll commentary, "It should be noted, however, that this year's 15-point gap in favor of the Republican candidates among likely voters is unprecedented in Gallup polling and could result in the largest Republican margin in House voting in several generations. This means that seat projections have moved into uncharted territory, in which past relationships between the national two-party vote and the number of seats won may not be maintained." Generally speaking, I think a 70 seat gain is probably in the order of 40% or so.
GOP Senate takeover. While none of the major prognosticators have predicted this (and to be fair, neither have I), it would not surprise me in the least to see this happen. I have the GOP picking up at least eight seats: AR, CO, IN, IL, ND, NV (that's right, Harry Reid is going down), PA and WI; that brings the GOP to 49. Fifty seats would be, technically speaking, a tie, but since the Vice President can break all ties, the GOP effectively needs 51 seats. Of those additional two seats, they could come from CA, WA, or WV. The polls show small Democratic advantages in these three seats, but with a "wave election" any or all of these seats could switch.
GOP Governor surge. The RCP Average shows the Republicans picking up a net of six statehouses (gaining 12 and losing 6). Among those that are leaning Democratic but could go GOP in a huge wave would be CA, CO, MA, MN, and OR. Again, of these three predictions, this is perhaps the least likely to happen, since all of these races show small but solid leads for the Democratic candidate.

Reading the tea leaves

Here are a few last-minute tea leaves to read in regards to tomorrow's election:

Where are people campaigning? President Obama spent the weekend in Chicago, Cleveland, Philadelphia, and Bridgeport (CT). While there are competitive races in all of these states, if your chief campaigner is spending his time firing up the voter base in cities the candidates should already be solid in, it's not a good sign.
Final polls: The final Gallup generic congressional ballot numbers favor Republicans 55-40 among likely voters, and 48-44 among registered voters. This is consistent with Gallup's polling last week, and also matches big GOP leads on similar surveys from Pew, CNN, Fox News, among others. In other words, nothing has changed in the last week of the election to suspect that Democrats have closed the gap. If at all, Republicans have pushed their lead even higher. Also, don't be surprised to see a last day poll or two (or few or more) showing a very close race in a district that, even until now, was supposedly safe). Finally, 10 of 11 polls in RealClearPolitics' poll averages of Governor and Senate races showed GOP advantages widening or Democratic leads dwindling.
What are the Democrats saying? This article says it all:

There is nearly uniform consensus among Democratic campaign professionals that the House is gone — the only question, it seems, is how many seats they will lose.

While few will say so on the record for fear of alienating party officials or depressing turnout, every one of nearly a dozen Democratic House consultants and political strategists surveyed expect a GOP majority to be elected Tuesday — the consensus was that Democrats would lose somewhere between 50 and 60 seats.

A senior party consultant who was on the low end with his predictions said the party would lose between 40 and 50 seats. On the high end, one Democratic consultant said losses could number around 70 seats.

All spoke to the grimness of the mood.

“It sucks,” said Dave Beattie, a Florida-based Democratic pollster who is working on a slate of competitive House races and who acknowledges that the lower congressional chamber is lost. “I’m resigned to the fact that it sucks.”

While there was optimistic talk within party circles early this month that the electoral environment was improving for the party, the operatives said those conversations don’t take place anymore.

“If some Democratic consultant told you they are feeling better, they must have dropped some heavy drugs,” said a senior pollster who is working for candidates in competitive races. “It’s hard."

...

Still, among those in the Democratic consulting class, there’s a gloomy acknowledgment that many of the incumbents the DCCC has spent millions of dollars to protect won’t be coming back to Congress.

“Everybody that is tied will lose, and everyone that is ahead by a few points will lose because of the GOP wave,” said one party media consultant who is involved in a wide array of House races. “There are going to be some surprises.”

Some strategists have resigned themselves to an election night that will bring an early end to the promising careers of Democrats they shepherded to victories in 2006 and 2008.

“In a wave election, part of the problem is that you feel powerless. Everything I feel I know how to do, that I’m trained to do, I can’t do. And that feeling is pervasive,” said the pollster. “There’s a sense that there’s nothing you can do about it. When you know your friends are on the chopping block, it’s hard.”

“There’s nothing worse than talking to an incumbent member of Congress who’s been cut off by the DCCC and who has no money,” said another Democratic consultant who has worked on crafting some of the party’s TV ads this cycle. “It’s like talking to a dead man walking.”

30 October 2010

Maryland & Anne Arundel County election predictions

Anne Arundel County: Anne Arundel County has 331,101 eligible voters: 145,707 Democrats (44.0%), 120,970 Republicans (36.5%), and 64,424 third party and unaffiliated voters (19.5%). Despite this edge of Democratic voters, the county overall has been reliably Republican: in 2008, the county voted for McCain over Obama (50.0% to 48.2%) and voted for every Republican candidate for President since at least 1988 (data older than that is not difficult to find, but more so to read!). In 2006, the county voted for then-Governor Ehrlich by a margin of 56.9% to 42.0%.

Turnout: During the last four Gubernatorial General Elections, turnout here in Anne Arundel County (as a percentage of eligible voters) about the parties has been the following:

1994 Democrats 69 Republicans 69 Overall 67
1998 66D 65R 63O
2002 66D 69R 62O
2006 64D 65R 61O

That averages to 66.25 Democrats, 67.00 Republicans, and 63.25 Overall. Based upon the overall "enthusiasm gap" in favor of Republicans (and respecting the fact that this gap is probably less evident in Maryland), I am estimating that Republicans will turn out in 1994/2002 levels at 69%. Democrats are certainly less enthused, but turnout in Anne Arundel County has always been higher than statewide. It is difficult to imagine Democrats turning out in better numbers then they did in 2006, but I'll give them the benefit of the doubt and put their number at 65%. With third party and unaffiliated voters typically at 50% or less, this brings our overall turnout to about 64%.

Projections

Based upon the registration numbers and turnout numbers, along with some intelligent guesswork based upon past trends and future expectations, I have come up with the following predictions:

U.S. Congress: MD-1 (Harris), MD-2 (Ruppersberger), MD-3 (Sarbanes), MD-4 (Edwards), MD-5 (Hoyer), MD-6 (Bartlett), MD-7 (Cummings), MD-8 (Van Hollen, Jr.). The only race that promises to be close is the MD-1 race between Andy Harris (R) and Frank Kratovil (inc. D). I've already chimed in on that race here, but generally speaking, the edge should go to Andy Harris by 2-3 points.

U.S. Senate: Like most of the House races, not a close contest. Mikulski (inc. D).

Maryland Governor: Maryland Governor Martin O'Malley has been somewhat immune to the nationwide political environment due to relatively low unemployment (based upon, for example, large numbers of government employees). Add to this the more than 2-1 advantage for Democrats in registration numbers and I think this is just too much for former Governor Bob Ehrlich to overcome. Ehrlich should win Anne Arundel County with at least 55% or so, but that won't be enough to overcome the large advantages that O'Malley receives in the D.C. suburbs. I'm thinking about 53% to 45% in favor of O'Malley over Ehrlich.

Question A: I've talked in length about this subject, so I won't spend too much time here other than to say that based upon my projected turnout numbers in Anne Arundel County, I think Question A will pass with 51.0-52.3% for, and 47.7-49.0% against. While this is not strictly a partisan issue, we've seen that Democrats tend to be for it (56 for, 37 against), Republicans and unaffiliated voters tend to be against it (44-51), and U (39-54). Since it is a relatively close vote, stronger than expected turnout by either party could tilt the vote either way. Even then, this is the prediction I am least confident about. Small changes in support among members of either party could swing the results either way.

29 October 2010

Final numbers for early voting in Maryland

Day 1 (9/3/10): 14,049 (18.18% of all early voters)

Day 2 (9/4/10): 8,891 (11.50%)

Day 3 (9/6/10): 10,076 (13.04%)

Day 4 (9/7/10): 12,347 (15.98%)

Day 5 (9/8/10): 13,844 (17.91%)

Day 6 (9/9/10): 18,081 (23.39%)

Here are the numbers so far for the early voting period prior to the November 2010 general election (with percentage increase from the primary election):

Day 1 (10/22/10): 32,131 (+229% since Day 1 of primary election)

Day 2 (10/23/10): 28,282 (+318%)

Day 3 (10/25/10): 34,818 (+346%)

Day 4 (10/26/10): 35,970 (+291%)

Day 5 (10/27/10): 38,831 (+280%)

Day 6 (10/28/10): 49,578 (+274%)

Over the course of the six day early voting period, 219,610 people voted; this represents a 284% increase over the 77,288 people that voted during the primary election early voting period.

Wednesday turnout review: The primary election turnout model estimated ~51,900; the general election turnout model estimated ~51,700. The final result of 49,578 voters on day 6 shows that both of our models came pretty close to capturing the final number, if overestimating it a bit.

Overall turnout: The total turnout of 219,610 compared to our most recent prediction of 220,000 (!). This represents 6.33% of all eligible voters as compared to our prediction of 6-7%.

Partisan turnout: The percentages of Democrats and Republcians voting over the six day early voting period: D+7 and R+0 as compared to party registration; D-3 and R+0 as compared to the primary election early voting turnout.

Election day projections: I'm working on a post for the weekend that will use the turnout numbers from the early voting period combined with the primary election turnout numbers, to project overall turnout and turnout by party for Election Day. I'll do this for the state and hopefully for Anne Arundel County too, which will help to predict the outcome of Question A.

2010 Maryland General Election Early Voting Turnout (Final)

28 October 2010

Early voting in Maryland, projections after five days

Through the first five days, 170,032 people voted; this represents a 287% increase over the 59,207 people that voted through the first five days of the primary election early voting period.

Tuesday turnout review: The primary election turnout model estimated ~40,000; the general election turnout model estimated ~44,000. Clearly, the last two days is not matching the fantastic 300%+ rates we saw on Saturday and Monday. It appears as if the numbers are coming back down in line with the primary election turning model which was <1,200 voters off.

Thursday turnout preview: As with the primary election early voting period, we might expect that the last day of early voting will show the highest turnout as all those voters who intended to vote early but haven't yet will stop procrastinating and get to the polls! Under the primary election turnout model we estimate that ~51,900 people will vote on Thursday; the general election turnout model estimates that ~51,700 will turnout (the self-updating nature of this model is self-correcting itself toward the primary election voting model).

Overall turnout: Falloff in the rate of increase of voters the last two days has brought the overall turnout numbers closer to 220,000 and between 6-7%.

Partisan turnout: The percentages of Democrats and Republcians voting remained very close to the overall numbers D+7 and R+<1 as compared to party registration; D-3 and R+-0 as compared to the primary election early voting turnout.

How is Question A doing among early voters?

Update: Based on the fifth day of early voting, the newest projections are 49.3% for and 44.5% Against.

Two recent polls have shown a very tight race in regards to Question A, the referendum to approve the Anne Arundel County zoning bill authorizing slots adjacent to Arundel Mills Mall. The first poll was 47-45 for, while the second was 48-45 for. If we ignore the margins of error and presume the actual numbers are correct (a dangerous proposition, to be sure), and then use the breakdown of partisan voting in the second poll to allocate votes, can we find if Question A is winning or losing so far in early voting? There are a lot of assumptions there, and a lot of accumulated errors that we're ignoring, but it can at least help us make an educated guess. The State Board of Elections gives us early voting turnout numbers by county; and the Gonzales poll showed a party breakdown of D (56 for, 37 against), R (44-51), and U (39-54).

Through the first four days of early voting, 9,148 Democrats have voted, 7,015 Republicans, and 2,246 Unafiliated or other party voters. Taking the percentages from the Gonzales poll and applying those proportions to those that have already voted, we find that an estimated 9,085 people would have voted for Question A, and 9,324 have voted against it; or 49.4% for, 44.4% against, and 6.2% undecided. This appears to show an edge to the For Question A side, on the strength of higher Democrat turnout in early voting. Democrats represent 44% of registered voters in Anne Arundel County, as compared to 37% for Republicans and the remainder other parties or unaffiliated; Democrats have accounted for nearly 50% of early voters (nearly +6 as compared to registration), while Republicans have accounted for 38% of early voters (less than +2 as compared to registration). Expected stronger Republican turnout on Election Day will inevitably make this projected lead for Question A smaller.

Projected Question A votes (through the first four days of early voting):
49.4% For
44.4% Against

27 October 2010

Early voting in Maryland, projections after four days

Through the first four days, 131,201 people voted; this represents a 289% increase over the 45,363 people that voted through the first four days of the primary election early voting period.

Tuesday turnout review: Under my old model I projected ~35,600, but modified that formula to give more weight to previous days of general election turnout which bumped the projection to ~41,000. The actual turnour of 35,970 shows that maybe I was too quick to abandon the old formula! Either way, between the two models we're now starting to get closer each day.

Wednesday turnout preview: Under the primary election turnout model we estimate that ~40,000 people will vote on Wednesday; the general election turnout model estimates that ~44,000 will turnout.

Overall turnout: Remains the same at 230,000+ and projected to be somewhere between 6-7%.

Partisan turnout: The percentages of Democrats voting increased slightly on Tuesday, and likewise the share of Republicans declined slightly; but the overall numbers are the same. D+7 and R+1 as compared to party registration; D-3 and R+-0 as compared to the primary election early voting turnout.

"Democratic voters hit the polls early," but what does that mean?

The Baltimore Sun is reporting that Democratic voters hit the polls early, which is true:

In the first three days of early voting, registered Democrats, who make up 56 percent of the Maryland electorate, cast 63 percent of the ballots, according to elections board. Republicans, who make up 26 percent of the electorate, cast 27 percent of the ballots.

Right. This matches up with what I've been reporting. They even go so far as to point out that

[b]ecause this is the state's first general election with early voting, there are no historic data on which to base predictions.

Good! But they fail to do any comparison to early voting in September's primary election early voting period, as I did yesterday, where Democrats turned out at a +10 advantage to their registration numbers as compared to +7 now. So if the September early voting has any predictive value at all, it actually shows that Democrats are turning out -3 less than you would expect.

26 October 2010

Kratovil, Harris Race Evenly Split (but Harris will win)

From WBAL Baltimore:

According to a recent poll, it appears as if the race for Maryland's 1st Congressional District will be another close one.

In 2008, the race between Republican Andy Harris and current Democratic Rep. Frank Kratovil took nearly a week to call because it was so close.

According to a recent poll of 520 voters by The Baltimore Sun, the candidates are tied with a 40 percent split, with 14 percent of those polled not sure who they'll vote for, 5 percent not saying and 1 percent not voting at all.

Interesting poll, because of the report that 1 percent said they were not voting at all. This indicates that the poll was done of registered voters. This close to the election, polls are typically done of likely voters, and likely voter polls tend show better numbers for Republican candidates.

MD-1 was extremely close in in 2008: Kratovil 177,065 (49.1%), Harris 174,213 (48.3%), Davis 8,873 (2.5%), and that was with the very high turnout in a presidential election year. Kratovil might get some support from O'Malley at the top of the ballot, but he won't get the same kind of support he received from Obama being on the ballot in 2008. Incredibly, every single county that Kratovil won was also won by John McCain except for Kent (Obama beat McCain there by 48 votes). Lower overall turnout means smaller margins for Kratovil in the places he won in 2008. Together, those margins should evaporate his overall lead.

I think this race will be close, although not as close as 2008. Andy Harris should win by 2-3 points.

"Locking up child sexual predators before they strike!"

Baltimore County State's Attorney Scott Shellenberger is running a radio ad about sex offenders. He talks about proactive prosecution by saying that he will be "locking up child sexual predators before they strike." Yeah, go get 'em! Wait, what? How can you lock up someone before they do something? That's like saying your crime-fighting platform will be to "lock up murderers before they kill!"

I heard this ad yesterday and I honestly thought I misunderstood it, so I didn't really think much of it. Then, I heard it again today so I made an effort to check the language.

This YouTube video has some of the same language, but he doesn't go all Minority Report like the radio ad. Unfortunately, I haven't found any audio of the ad, yet.

One week away; congressional electoral predictions!

Check out this graph of Gallup's 1994 Generic Congressional Ballot (Registered Voters):

The Republicans' largest lead was in June, when they lead 50-44. By the time of the election in November, the lead had swapped several times, and in fact the last poll showed the parties tied at 46-46. Of course, 1994 was a banner year for Republicans, picking up 54 seats in the House of Representatives and 8 seats in the Senate, gaining control of both bodies.

As of today, Gallup's Generic Congressional Ballot shows Republicans leading 48-44 (+4) among Registered Voters; 55-41 (+13) among Likely Voters in a Low Turnout Model, and 52-43 (+9) among Likely Voters in a High Turnout Model.

Translation: unless these numbers change dramatically toward the Democrats in the next week (and there isn't any indication to suggest they will), next Tuesday will be an electoral bloodbath for Democrats; on the order of 50-60+ seats in the House of Representatives (+39 would take control), and 8+ seats in the Senate. Republicans will regain control of the House and will probably end up with 49 or even 50 seats in the Senate (they'd need 51 to control the Senate since the Vice President could break a tie); although a majority in the Senate is not out of the question.

Do you think I am way off, that you're still be seeing Speaker Pelosi next year? Ok then, who wants to bet me a beer that Democrats will retain control of the House? Anyone?

Early voting in Maryland, projections after three days

During the early voting period prior to the September 2010 primary election, 77,288 people voted early, which represented 2.44% of all eligible voters. Over the six day period, the numbers were as follows (with percentage of early voting total by day):

Day 1 (9/3/10): 14,049 (18.18% of all early voters)
Day 2 (9/4/10): 8,891 (11.50%)
Day 3 (9/6/10): 10,076 (13.04%)
Day 4 (9/7/10): 12,347 (15.98%)
Day 5 (9/8/10): 13,844 (17.91%)
Day 6 (9/9/10): 18,081 (23.39%)

What you see is strong initial enthusiasm, followed by an immediate drop and then a gradually increasing rise to the last day.

Here are the numbers so far for the early voting period prior to the November 2010 general election (with percentage increase from the primary election):

Day 1 (10/22/10): 32,131 (+229% since Day 1 of primary election)
Day 2 (10/23/10): 28,282 (+318%)
Day 3 (10/25/10): 34,818 (+346%)

What we see so far is that the turnout model during the primary election early voting period ("strong initial enthusiasm, followed by an immediate drop and then a gradually increasing rise to the last day") has not been predictive of turnout during the first three days of the general election early voting period, even when accounting for massive increases in turnout.

Through the first three days, 95,231 people voted; this represents a 288% increase over the 33,016 people that voted through the first three days of the primary election early voting period.

Monday turnout review: Based upon numbers we have seen thus far and comparing them to numbers from the primary election early voting period, we estimated ~26,500 people would vote statewide on Monday (although the estimate was a range anywhere from 23,000 to 32,000). In fact, turnout exceeded even the highest prediction, which again shows that the primary election turnout numbers are not proving to have much predicative value.

Tuesday turnout preview: Under our previous "primary election turnout" model, we could estimate that ~35,600 people would vote statewide on Tuesday. However, we have already seen flaws in the predicative nature of the primary election numbers, so I have modified the formula to give more weight to previous days of general election turnout. Likewise, this formula should self-modify itself a bit more each day. In this case, we find that we get an estimate of ~41,000. Obviously, we can tweak this again tomorrow based upon what we see.

Overall turnout: Our initial projection was for at least 200,000 people (and quite possibly upwards of 250,000. Given the strong turnout so far, I would be willing to move the floor to about 230,000 while still considerable upward potential to 250,000 and beyond.

September's early voting turnout was 2.44% of all eligible voters; general election early voting turnout was initially projected to be somewhere between 6-7%, and now looks closer to 6.5-7.0% with higher being possible.

Partisan turnout: Democrats are turning out so far at approximately 7 percentage points higher than their percentage of party registration; Republicans are turnout out less than one percentage point higher than their percentage of party registration. While this would seem to suggest a enthusiasm gap on the part of the Democrats (which would also counter the nationwide trend), there is no historical data from previous election cycles with which to compare this (i.e., if Democrats typically turn out +10, than this election would show less enthusiasm). This data is less useful now, but will be more useful in 2012.

However, we can look at the turnout during the primary election early voting period. Statewide voter turnout in early voting was 2.44%; Democrat turnout was 2.82% of their eligible voters; Republican turnout represented 2.37% of their eligible voters. Democrats represented nearly 71% of early voters, with Republicans at 28%. So by a slight margin, Democrats showed a slight turnout edge in early voting.

Compare this to Election Day: 21.83% of eligible Democrats turned out; 27.24% of eligible Republicans turned out. Democrats represented 62% of Election Day voters, with Republicans at 36%. In total, Democrat turnout was 24.65% while for Republicans it was 29.60%; Democrats represented 63% of total voters, with Republicans at 36%. So that significant edge in early voting only accounted for one additional percent on Election Day. This 5 percentage point overall turnout edge for Republicans was not predicted by the early voting turnout.

What we do know is that in September, registered Democrats represented 61.39% of all eligible voters and cast 70.87% of all early votes; a +10 advantage compared to registration numbers. Republicans represented 28.90% of all eligible voters and cast 28.02% of all early votes; a +1 advantage. Now, if we go back our original numbers, we see Democrats turning out in early voting at +7 and Republicans at +1. If the September early voting has any predictive value at all, it actually shows that Democrats are turning out -3 less than you would expect. Again, caveat: there is no historical data from previous election cycles with which to compare this.