Saturday, November 19, 2016

The Road Ahead - Facing Trumpism

Nearly 91% of all republican women and 90% of republican men voted for Trump.  Think about it for a moment.

Evangelical Christians, -who, for years, have passionately argued against the corroding influence of Hollywood on the society,  argued for Christian tenets of service, humility, and compassion,  - voted for a bombastic, boorish, libertine, Hollywood talk show host.

Foreign policy conservatives, -who for years, have supported our alliances such as NATO and harbor a deep skepticism about Russia's intentions against its neighbors, voted for someone who was sympathetic to strongmen such as Putin and Saddam. 

Trump server communicating with Russia? (read:

Economic conservatives who espouse free trade almost as a religion voted for a failed businessman arguing for tearing up free trade agreements and promising to start a trade war with our partners.

Trump has derided Trade agreements like NAFTA, TPP, etc.

Think about it for a minute.  How come, all these republicans voted for Trump?  Because, Hillary Clinton had a private email server? I am not sure that was the case.

Throughout the election, it was evident - even to many republicans, that Hillary Clinton had the experience, preparation, and policy chops that would serve her well, if elected.  For nearly 30 years, republicans had primed the American electorate to distrust her by launching one investigation after another, from Whitewater, to Benghazi, to Clinton Foundation, and the email servers.  Although these investigations, often led by partisan republicans, did not reveal any wrongdoing on the part of Hillary Clinton that would merit prosecution, they did serve the one goal that republicans had in their mind all along -to drive up Hillary's negatives, which they did. 

Her opponent, Donald Trump started his campaign by insulting Mexican immigrants, insulted his fellow republican primary opponents with petty name calling, insulted women, insulted a parent of a soldier who died in combat in service of his nation, proposed a blanket ban on Muslims entering the country, boasted about not paying taxes, and changed policy positions at the drop of a hat.  This is but a small sample of his hateful rhetoric over the course of his campaign.

The differences between the candidates could not have been stark.    I followed the arguments between the two candidates closely, checked on reputed pollsters such as Nate Silver and Sam Wang on almost a daily basis, and I convinced myself that Hillary Clinton will win.  My conviction was not just based on polls and pollsters, but also in believing that the immediacy of a Trump presidency would jolt Americans to their senses. It is clear now that it was not the case.

I am sure many a thesis will be written about how and why Trump won.  Hillary was a weak candidate; Trump was a master salesmen; It is difficult for a party to win three times in a row, etc.  There will be a kernel of truth in each, but the big question in front of progressives is:

To understand the mindset of the republican voter, who is willing to faithfully cast his ballot even for a candidate like Trump.

Once that is understood, we can craft a message to reach them.  Obviously, whatever that we have been doing is not working.  Perhaps, the first place to start would be to read the book by George Lakoff, who addresses this issue head on.  More on that in a later post. 

Monday, November 7, 2016

Can Trump Lose Texas? Final thoughts on Texas Early Voting

In 2016, nearly 4.5 million (or 46%) of the 9.8 million Texans living in the 15 largest counties had voted early which is a record  Although it is a full 4% less than the predictions that I made based on the first four days of early voting, it is sharply higher than the 42% and 39% early voting percentages (EV %)in 2008, and 2012.

Looking at 2016 data, a few things jump out that favors democrats:

1. More new RV live in Democratic Leaning counties.  Three fifths (61%) of the newly registered 1.1 million voters from the top 15 counties live in counties that were won by Obama in 2008, and 2012.  It is noteworthy that Harris county, the largest county in Texas with nearly 2.2 million voters, which Obama won by the slimmest of margins in 2012, accounts for nearly 20% all new RV in these counties.  The population of Harris county is diverse with 40% Latino, 30% Anglo, 20% Black, and 8% Asian, and EV analysis suggests that the Latino turnout is significantly higher than in 2012.  There is a good chance that Hillary Clinton may carry Harris county by a healthier margin (around 5%) than Obama did either in 2008 or 2012.

% of newly registered voters living in the 15 most populous counties in Texas.  The horizontal axis provides the average victory margin for Obama from 2008 and 2012 elections.   Nearly 250,000 more newly Registered voters live in Democratic leaning counties than Republican counties.  The size of the bubble indicates the size of the RV in the county.

2. How is the early vote (EV %) turnout in democratic leaning counties compared to 2012?

While the EV turnout compared to 2012 is higher both in republican and democratic leaning counties, it is clear that several large counties, which tend to be democratic leaning, have had significant increase in early voter turnout compared to 2012.

The increase in EV% in 2016 compared to 2012 shows a number of democratic counties increasing their turnout compared to 2012.  While republican counties also show an increase, the population of these counties is smaller than democratic counties.
 3) Changing Demographics: Counties such as Fort Bend, and Neuces which went republican in 2012 with fairly small margins but may flip to democratic column for two different reasons.  Fort Bend county, the richest Texas county, has an influx of highly educated professionals from around the world, which might make it difficult for Trump to carry that county.  Neuces county has a large Latino population, and EV% is about 6% higher in 2016, and an increased turnout from Latino population may be enough to turn this county blue.

4. Depressed Republicans: This may surprise some.  There are principled conservatives in Texas.  I would expect that at least a small fraction of principled conservatives and conservative women could not bring themselves to vote for Trump, depressing the republican vote slightly even in rural republican counties. 

In short, EV shows that the democratic enthusiasm in the state of Texas is more evenly matched with the republicans in this cycle, there are more newly registered voters in democratic leaning counties, and EV percentages in democratic leaning counties is higher than in 2012.  The changing demographics with increased Latino turnout in select counties, and a slightly depressed republican turnout in rural counties, all favor Democrats.

The combination of these might, just might, be enough to make it an interesting night for Texas.  All it may take is a few percentage swing toward the democrats.  So, pick up the phone, and call up your conservative relative or friend in Texas ( I know you have one! ), and test your powers of persuasion!

Thursday, November 3, 2016

Brazoria county: Vote for John T Floyd

Brazoria county is the prototypical republican dominated Texas county.  Democrats are a rare breed in these parts of Texas.  But, something is changing this time around.  Brazoria county is the 16th most populous county in the state of Texas, and as such does not get the honor of getting its early vote totals tallied by the office of the Texas Secretary of State

I sent an email to the office of the county clerk of Brazoria, and received a ton of information from Ms. Janice Evans, about the early voting. 

With one more day to go (last day of early voting typically gets around 4-5% of votes), Brazoria county early voting totals have already eclipsed the records set in 2008, and 2012.

Brazoria county is well on its way to cast nearly 50% of the 197,500 registered voters in the county.  This will set an all time record for Brazoria county. 

This time around Democrats have some good candidates on the ballot.  For example, we have an excellent candidate, John T Floyd running for the office of the state representative TX-29 who has waged an excellent campaign based on issues, such as:

  • Prioritizing public education to all.
  • Public School books must be historically accurate
  • Support law enforcement agencies, but monitor them and train them in community policing.
  • End corporate welfare and subsidies to big businesses.
He is eloquent, organized, and you find John Floyd signs everywhere. 

The success of his campaign hinges on running up the score in precincts with educated population such as Shadow Creek, Pearland West, Silverlake, etc. to offset some of the rural precincts that tend to lean republican.  If Democrats want to loosen the republican stronghold across all levels in Texas politics, they have to start with people like John Floyd competing for positions.

So, if you live in any of these areas, go and vote for John Floyd!  If you have friends living in those areas, ask them to vote as well. 

Wednesday, November 2, 2016

Texas Blue? Reading the tea leaves after 10 vdays of early voting [updated]

This is an update to the previous post about early voting in Texas.  A couple of things are worth noting.

First, the pace of early voting in the 15 most populous Texas counties continues to be higher than what was in 2008, and 2012. 

Compared to 2008, and 2012, percent of eligible voters who have voted during the first 9 days is substantially higher in 2016.  This is particularly impressive, given that the total number of registered voters increased by about a million voters between 2012 and 2016.

We still have two one days of early voting to go and if current trends hold, we might even get close to 50% of the votes cast (I think we might end up a bit shy of 50%) when the EV comes to a close this Friday.  If 2008/2012 patterns hold true, then the total number of votes cast in this election might be among the highest ever.

Second, it is clear that voting is brisk in both republican and democratic strongholds.  At the end of day 9 of early voting, compared to 2008 and 2012, EV in both republican and democratic counties are higher by significant margins.

In the vertical axis, we see the percent difference in EV at the end of 9 days across the 15 most populous Texas counties, between 2016 and 2012, and between 2016 and 2008.  The voting is higher in both republican (red shades) and democratic counties (blue shaded).
Both in 2008, and 2012, the number of votes cast in the 15 most populous counties was nearly equally shared between republicans and democrats.  The final point differential in favor of republicans in 2012/2008 elections is mostly due to republican dominance in smaller counties, and this amounts to about a million votes.

While it is possible, that the increased turnout in republican leaning counties is due to democrats who were spurred into voting due to the uniqueness of this election cycle, there is no concrete evidence to suggest this as of yet.  For example, strong republican counties such as Collins and Fort Bend, now have a significant influx of highly educated Asians with a higher median income.  In these counties, the margin of victory for republicans might shrink substantially.  It will be very interesting to see how this turns out.  Secondly, at least a couple of polls conducted in Harris county, the largest county, suggest that Hillary is leading by a fairly significant margin.  If that trend holds true for other large democratic leaning counties, then the election might be very close. 

So, if you have friends in Texas, ask them to vote!

Saturday, October 29, 2016

Texas Early Voting Results - Harbinger of a coming Tsunami?

Early voting in Texas has been very strong and this has attracted widespread attention from poll watchers, and recent polls have suggested that Trump's lead in the Lone star state is less than 5 points

Could Texas turn blue in this cycle?  

What can we reasonably infer from the Texas early voting results?

I will use the early voting results from the last two presidential elections (2008, 2012) as my baseline, as there is a significant drop in Texas mid-term voter participation and may not reflect what happens in 2016.

Texas has 12 days of early voting.  If we look at the day by day voting of the top 15 most populous counties in Texas (from Texas secretary of state) for 2008 and 2012:

Day by day cumulative percent of Texans voting during early voting (EV) from the 15 most populous counties in 2008, 2012 presidential elections.  The EV data for the first four days of the 2016 election cycle are plotted in circles, and the dotted line indicates a linear fit.

With respect to early voting, in 2008, 3.28% of registered voters cast their ballot per day, and in 2012, that number dropped ever so slightly to 3.2%/day. 

On election day, the percent of registered voters who cast their ballots in 2008, and 2012 were 17.47%, and 19.20% respectively, adding up to a total of 59.5% and 58.5% votes cast.  Although on the surface, 2008, and 2012 election voting (early and election day voting) looks roughly similar, the victory margins for Romney swelled by a full 5 points compared to McCain (11.8% in 2008 to 16.8% to 2012). 
For the 2016 cycle, we have data for the first five days of EV for the 15 most populous counties in Texas.  If the current voting trends hold - 4.73% per day, more than 51% of the registered voters would have cast their ballot before election day. If we model the election day voting as average of the 2008 and 2012 cycles, then we get an astounding 69.5% of registered voters casting their ballot in 2016, a full 10 points higher than the landmark election in 2008.

The voting is high, but does it necessarily portend good news for democrats?  If we calculate %EV/day for each of the fifteen counties for 2008, and 2012 cycle, we might be able to discern if 2016 is, in any way, different than 2008, or 2012.

Let us look at 2008 results.

2008 Early Voting Pattern for the 15 most populous counties in Texas.  The vertical axis shows the EV percent per day, and the horizontal axis, shows the final margin between Democratic and Republican votes.  The size of the bubble shows the number of voters in that county.  A strongly democratic county such as Hidalgo will appear on the far right, and has far fewer registered voters than the Harris county (the largest bubble right in the middle).

In 2008, although some strong democratic counties voted at rates higher than 3% per day during the EV period, some counties (lower right) voted at rates of around 2% per day.  In contrast, almost all republican leaning counties voted at rates around 3.5% day per higher.  Obama lost to McCain by 11.8 points. 

Let us look at 2012, results.

Compared to 2008 early voting results, very few strong democratic counties (to the right) clocked in a voting rate of 3.5 points per day, but many republican counties also (to the left) also clocked in 3.5 points per day. 
In 2012, a few  democratic counties clocked in early voting rate of 3.5% points, but the overall trend in EV remained the same - republicans outperformed the democrats.  One can easily imagine a negative trend line going down from top-left to bottom right.  Obama lost to Romney by 16.8%.

If we repeat the same analysis based on the EV data available to-date (six days), and using the democratic/republican margins of victory in 2012, one gets a very interesting picture.

In contrast to the negative trend lines in 2008/2012 cycle, this graph graph looks like a 'U'.  In strong democratic leaning counties (based on 2012 election results), the early voting rate per day exceeds 4%/day (counties on the right).  The same is true for strong republican counties (on the left), with Harris county (the big ball) right in the middle.
The 'U' shaped curve, at first glance, suggests that voter enthusiasm is high on both sides, with the most republican leaning county and the democratic leaning county raking in more than 5% of votes per day.  Does this suggest that Democrats will suffer yet another loss (like in 2008) despite the apparent voter enthusiasm?

There are a couple of encouraging signs for the democrats in the Lone Star state.  One of the key assumptions in the graph is that the D-R advantage is the same as in 2012.  There is some evidence to suggest that this is a very conservative estimate.  For example, the largest county in the state, Harris County,  democrats bested republicans by 1.63% in 2008, and by 0.08% in 2012.  In a poll conducted by the University of Houston - before the debates, before the Access Hollywood tapes, Clinton had a sizable lead over Trump.

A Harris county poll done in September shows a substantial advantage for Hillary Clinton among extremely ikely (4%), likely (9%), and registered (10%) voters.  This is surprising given that Obama won Harris county by only a sliver in 2008, and 2012.
If Harris county, where republicans and democrats have essentially tied in the past two presidential cycles, can swing strongly democratic (5-6% swing), then Democrats have a decent chance of performing even better in strong democratic leaning counties.  That may be enough to offset the republican advantage in rural areas.

Second, many republican counties such as Collin county, Fort Bend county, and Brazoria county, which are fast growing, have a greater influx of educated Asians working in high-tech, and medical fields, who tend to vote more democratic.  In fact,

Available evidence, though limited, strongly suggests Texas turning blue - possibly even in this cycle. 

If you have friends in Texas, call them and ask them to vote.  They may become part of a truly historic election in Texas!