bernie sanders

DEMOCRATIC PRIMARY PROJECTIONS: CALIFORNIA, MONTANA, NEW JERSEY, NEW MEXICO, NORTH DAKOTA, SOUTH DAKOTA

/ Tyler Pedigo / 50 Comments

If you don’t typically read my entire analysis, this is the time to read it all. There are many caveats to the below numbers.

Tuesday is effectively the last of the Democratic primaries for 2016 (with the exception of D.C.), and is indeed the last opportunity for the Sanders campaign to close the pledged delegate gap. As we all know, all eyes are on the state of California, which is set to allocate 475 delegates between the two candidates. Bernie Sanders will need a massive majority of these 475 delegates (and, realistically, several large wins elsewhere too) to take the lead before the Democratic convention. Unfortunately for Bernie, my models are not suggesting that a pledged delegate lead is currently possible, despite four projected wins on Tuesday. However, it does look like California will be a very close race, and there is a very real possibility that Sanders could win there after all. Here are my projections:

Screen Shot 2016-06-05 at 8.38.52 PMIf you would like to support my work and want me to be able to afford Top Ramen (or maybe even Mellow Mushroom pizza if you all are extraordinarily generous) while I’m working on these statistics, please click this link to donate to Tyler’s Food & Rent Fund!

CALIFORNIA

Bernie Sanders should do better in California than is currently expected. Though I do believe that the state is leaning towards a Hillary win, there are several reasons why it should be a close race:

  • Sanders has 74.0% of Democrat Facebook likes in California. This is similar to Kentucky (73.9%), Oklahoma (75%), and West Virginia (75%). Hillary lost OK and WV, and won KY by 0.4% of the vote.
  • California is a semi-closed primary, which Sanders has usually done quite well in. Other semi-closed primaries that Sanders won are New Hampshire, Oklahoma, Rhode Island, and West Virginia. Semi-closed primaries that Clinton has won are Massachusetts and North Carolina.
  • Demographically, California is fair for both candidates. It is indeed a diverse state, but not necessarily to Sanders’ detriment. California has an African American population less than the country average, though it has a significant Hispanic population. Hispanics do tend to prefer Hillary, though my finding is that this effect is not very substantial. More on this effect in New Mexico below.
  • Of the twenty-one states that Sanders had a greater proportion of the relative number of campaign contributions, he has lost only five (IA, MA, NC, AZ, KY). Furthermore, he has won four states in which he had a lower number of relative campaign contributions than he has in California (CO, OK, NE, RI).

However, California typically has a large amount of early voting, which has historically been very beneficial to Hillary Clinton. I am admittedly unsure of the exact percentage of Californian likely voters that have chosen to already vote by mail, but it is my personal belief that this number will be around 40-45% after all of the vote comes in. I expect Bernie to capture only about 40% the early vote while Hillary secures 60%, but at the same time I am expecting Bernie to win about 57% of the day-of vote while Hillary trails with about 43% on Tuesday.

With all of that being said, I am very uncertain of what will happen in California. My personal feeling is that Bernie Sanders has a decent chance of winning, but gut feelings aren’t typically based on numbers, so take that with a grain of salt.

MONTANA

Bernie Sanders should clean house in Montana. This is for the following reasons:

  • Bernie Sanders has 83.87% of Democrat Facebook likes in Montana. Only in five states does he do better, Vermont, Maine, Idaho, Alaska, and Oregon, and he won all of these states by large margins.
  • Only in three states does Bernie have a higher relative number of campaign contributions, in Vermont, Alaska, and Oregon.
  • Montana is only 0.4% African American. Sanders does exceedingly well under these circumstances.

We should see a blow out in Montana. Clinton has surprised us all before (Wyoming…), but I would be exceptionally surprised to see a margin of victory of less than 15% here.

NEW JERSEY

Hillary Clinton should win New Jersey with ease due to the following reasons:

  • New Jersey is 13.7% African American, which is above the country average. As many of you know, Hillary does very well with African American voters.
  • Bernie has only 70.0% of Democrat Facebook likes in New Jersey, and in only ten states is this measure more unfavorable for him.
  • Bernie’s relative number of Google searches over the last three days is only 0.68. I do feel like this measure is losing its relevance as the campaign has went on, but regardless, this is the worst of all fifty states. Alabama comes in second with 0.832.
  • The only redeeming factor in New Jersey for Bernie Sanders is the fact that it is a semi-closed primary.

I believe that the margin of victory in New Jersey will be anywhere from 5-20%.

NEW MEXICO

New Mexico stands to be one of the most interesting elections on Tuesday. Other outlets are projecting a Hillary blow out here (and I’m not so sure about my projection that I would say that they are definitely wrong), but I am projecting a Sanders win for the following reasons:

  • New Mexico is Sanders 12th best state when it comes to the relative number of campaign contributions for him (he has not lost any state where he had a higher number than in this state), and he has won twelve other states in which he had a lower relative number of campaign contributions than in New Mexico.
  • Sanders has 79.49% of Democrat Facebook likes in New Mexico. He has not lost any state where he has a higher number than 79.49%, and he has won seven other states in which he had a lower percentage of the Democrat’s Facebook likes (i.e., a number lower than 79.49%).

These are the two primary drivers of my New Mexico Sanders win, but there are two major factors that may produce the alternative outcome.

  • New Mexico holds a closed primary. As many of you know, Sanders seems to generally do very poorly in this contest format, as he typically has a substantial reliance on independent voters. He has only won one closed primary, in Oregon, and that state was uniquely predisposed to give Sanders a win regardless of contest format type. However, Sanders did just come within a half a percentage point of Hillary Clinton in Kentucky two weeks ago in Kentucky’s closed primary. Sanders has a far greater relative number of campaign contributions in New Mexico than in Kentucky (2.42 versus 1.79), has a far greater percentage of Democratic Facebook likes in New Mexico than in Kentucky (79.49% versus 73.91%), and has a much smaller African American population in New Mexico than in Kentucky (2.1% versus 7.8%). It could be the case that Sanders is increasingly being viewed more favorably among lifelong Democrats, or that Kentucky was a lone anomaly.
  • New Mexico has the largest Hispanic population of any state, about 47%. Hispanics do seem to prefer Clinton (hypothetical example, if you were to randomly select 100 Hispanics in Kentucky, instead of voting 46.8% for Hillary as was the state result, these 100 Hispanics would vote 61.0% for Hillary. This effect is based on my analysis of about two hundred randomly selected counties from all over the U.S.), but my finding is that this effect is not so substantial as to produce a Hillary victory in New Mexico as it is fighting against the above factors; abnormally high campaign contributions and Facebook presence.

If Hillary does win New Mexico, a very real possibility, I believe that it will be because of the closed primary format, and not the effect Hispanics have. Hispanics do not seem to vote as “monolithically” as African Americans do according to all of my county level analysis (an analysis performed almost entirely to answer the New Mexico Hispanic question). Perhaps I am totally in left field predicting a Sanders win here, but all of the indicators that I rely on are pointing solidly in that direction.

NORTH DAKOTA

In case anyone wasn’t able to already predict that Sanders would win North Dakota, I’ll make the case.

  • North Dakota is only 1.2% African American, similar to Maine (1.2%), Vermont (1%), and Utah (1.1%). As you all know, Hillary lost these states by margins between 30-67%.
  • Sanders has a very high number of relative campaign contributions in North Dakota, among his best states, and not too dissimilar from the aforementioned Montana.
  • North Dakota also holds an open caucus, which Sanders does extremely well in (MN: 61.6%, ID: 78.0%, UT: 79.3%, WA: 72.7%). It literally does not get any better for Bernie unless the state is named Vermont.

I expect the margin of victory in North Dakota to be anywhere from 30-50%

SOUTH DAKOTA

South Dakota is highly similar to North Dakota, and for that reason Sanders should probably win by around the same amount.

  • South Dakota is also only 1.2% African American, similar to Maine (1.2%), Vermont (1%), and Utah (1.1%).
  • Sanders also has a very high number of relative campaign contributions in South Dakota, albeit slightly less than North Dakota. It is still among his best though.
  • Bernie has 78.57% of all Democrat Facebook likes in South Dakota, slightly higher than in North Dakota (76.9%). Only in a handful of states does he do better.
  • South Dakota, unlike North Dakota, holds a semi-closed primary. As mentioned before, Sanders still does well in semi-closed primaries, but not nearly as well as in open caucuses. This reduces his expected vote share compared to North Dakota, but together with the difference in Facebook presence, the differences are mostly a wash within my model.

Like North Dakota, I expect the margin of victory to be quite large. Though it is of course possible that Hillary will keep South Dakota closer than I estimate, I would be surprised if the margin of victory was less than 15%.

 

In conclusion, major kudos to both candidates for such a hard fought race. As you all probably could’ve guessed, I am a Bernie supporter, but despite my perceived shortcomings of Hillary, I do believe that she would make a good president. Obviously I would prefer that Bernie win, but with the system we have now (super-delegates), Bernie probably came as close as any seemingly unknown outsider ever could.

Lastly, I want to say THANK YOU to all of you that have consistently tuned in and listened to what I have had to say over this election season. It really means the world to me that hundreds of thousands of people care about and appreciate my work. Also, a special thanks to all of you who have donated to help fund the work that I do here, and have ensured that I have a roof over my head! I am truly blown away at the support I have received, and I’m very excited for all of the upcoming elections later this year. What started as just an experiment to see if Facebook and Google data were useful in predicting elections has blossomed into something much bigger, and, in my opinion, we are witnessing the birth of an entirely new methodology of predicting elections; one that will be more powerful in the future than we could have ever imagined.

-Tyler

 

DEMOCRATIC PRIMARY PROJECTIONS: KENTUCKY AND OREGON

/ Tyler Pedigo / 36 Comments
As you all know, I’m just a graduate student with no income. If you would like to support my work and want me to be able to afford Top Ramen (or maybe even Mellow Mushroom pizza if you all are extraordinarily generous) while I’m working on these statistics, please click this link to donate to Tyler’s Food & Rent Fund!
Also, if you are an employer and have an open position, ideally in the D.C. area, I need a summer job! Please contact me if you think I would be a good fit for your organization. Paid positions only, please.


 I am estimating that Bernie Sanders will win both primaries tomorrow, in Kentucky and Oregon. Using my metrics, Oregon seems poised to be a blowout Sanders victory, while Kentucky stands to be a hard-fought battle between both candidates for a win. I have put together an entirely new framework over the past week to account for votes going to other candidates, which is where my West Virginia projection fell most short. It is a more comprehensive model, and should be more accurate. For anyone concerned, my old model is generating very similar estimates for tomorrow. Here they are:

Screen Shot 2016-05-16 at 6.47.10 PM

KENTUCKY

The demographics of Kentucky favor Bernie Sanders. It is a white state with only a 7.8% African American population, similar to that of Kansas (5.9%), Wisconsin (6.3%), and Indiana (9.4%), all states that he has previously won. 7.8% is approximately at one standard deviation from the mean Black population percentage (4.2%) of the states that Bernie has won, meaning that it is not too far out of the ballpark for a Sanders victory. Bernie has also done quite well with campaign contributions in Kentucky, with the logged value of the relative number of <$200 contributions being 0.337. This is slightly under the average of states that he has won, 0.366, but far higher than the average of the states that he has lost, -0.07. These reasons are the primary drivers of my estimated Sanders victory.

Bernie’s Facebook presence in the state is poor, 73.91%, which is lower than any state he has won at this point. Bernie’s relative search interest in Kentucky is poor as well, with the three-day average currently at 0.927. The mean three-day average for all the states that he has won is 2.167, though just last week he won West Virginia at a relative search interest value of 0.94.

Lastly, Kentucky has a closed primary electoral format, which Bernie has never won before. Regardless of who actually wins the Kentucky primary tomorrow, I believe it will be a very close race.

Screen Shot 2016-05-16 at 7.32.23 PM

OREGON

Oregon is Bernie’s best state with the exception of Vermont when it comes to Facebook data. He has 84.314% of Democrat Facebook Likes, similar to Idaho (84.0%), Maine (84.09%), and Alaska (83.87%) (Vermont was 95.00%). Demographically, Oregon is about as good as it gets for Bernie. The African American population is only 1.8%, similar to Hawaii (1.6%), Utah (1.1%), and Alaska (3.3%). Only in Vermont and Alaska did Bernie outpace Hillary to a greater extent than in Oregon in the relative number of <$200 campaign contributions. These are the primary drivers of the massive margin of victory that I am projecting. It is difficult to reconcile the one and only poll (that showed Hillary with a 15% lead… but also had 19% undecided… and was also conducted well after ballots had already been received and presumably had already been mailed off by many voters) conducted in Oregon with this projection, but I refuse to arbitrarily tack on extra points because I have a hunch about something.

Oregon is entirely vote by mail. Clinton has traditionally dominated early voting, but Oregon’s format is unique to all the states that have already voted, so it is difficult to predict how much of an effect this will have. Personally, I doubt it will be significant due to practically every metric being overwhelmingly in Bernie’s favor (imagine Vermont was only vote-by-mail, would that have really changed the result?). Also, the party registration deadline was recent, April 26th. Bernie’s current relative search interest is quite low, but Oregonians began receiving their ballots two-three weeks ago. If we go back in the Google Trends data to April 26th (around the day voters began receiving ballots), Bernie’s relative search interest for the next week and a half was around 1.45; not bad. The average for all the states he has won is 2.167, but the standard deviation is 0.61, so 1.45 is not indicative of anything particularly remarkable.

Lastly, Oregon is also a closed primary, which Bernie Sanders has never won before.

Screen Shot 2016-05-16 at 7.33.22 PM

If the above estimates are correct, this should give Hillary Clinton a ~24 delegate deficit tomorrow. Good luck to both candidates, and happy voting to all you Oregonians and Kentuckians!

-Tyler

DEMOCRATIC PRIMARY PROJECTIONS: KENTUCKY AND OREGON

/ Tyler Pedigo / 19 Comments
As you all know, I’m just a graduate student with no income. If you would like to support my work and want me to be able to afford Top Ramen (or maybe even Mellow Mushroom pizza if you all are extraordinarily generous) while I’m working on these statistics, please click this link to donate to Tyler’s Food & Rent Fund!


 I am estimating that Bernie Sanders will win both primaries tomorrow, in Kentucky and Oregon. Using my metrics, Oregon seems poised to be a blowout Sanders victory, while Kentucky stands to be a hard-fought battle between both candidates for a win. I have put together an entirely new framework over the past week to account for votes going to other candidates, which is where my West Virginia projection fell most short. It is a more comprehensive model, and should be more accurate. For anyone concerned, my old model is generating very similar estimates for tomorrow. Here they are:

Screen Shot 2016-05-16 at 6.47.10 PM

KENTUCKY

The demographics of Kentucky favor Bernie Sanders. It is a white state with only a 7.8% African American population, similar to that of Kansas (5.9%), Wisconsin (6.3%), and Indiana (9.4%), all states that he has previously won. 7.8% is approximately at one standard deviation from the mean Black population percentage (4.2%) of the states that Bernie has won, meaning that it is not too far out of the ballpark for a Sanders victory. Bernie has also done quite well with campaign contributions in Kentucky, with the logged value of the relative number of <$200 contributions being 0.337. This is slightly under the average of states that he has won, 0.366, but far higher than the average of the states that he has lost, -0.07. These reasons are the primary drivers of my estimated Sanders victory.

Bernie’s Facebook presence in the state is poor, 73.91%, which is lower than any state he has won at this point. Bernie’s relative search interest in Kentucky is poor as well, with the three-day average currently at 0.927. The mean three-day average for all the states that he has won is 2.167, though just last week he won West Virginia at a relative search interest value of 0.94.

Lastly, Kentucky has a closed primary electoral format, which Bernie has never won before. Regardless of who actually wins the Kentucky primary tomorrow, I believe it will be a very close race.

Screen Shot 2016-05-16 at 7.32.23 PM

OREGON

Oregon is Bernie’s best state with the exception of Vermont when it comes to Facebook data. He has 84.314% of Democrat Facebook Likes, similar to Idaho (84.0%), Maine (84.09%), and Alaska (83.87%) (Vermont was 95.00%). Demographically, Oregon is about as good as it gets for Bernie. The African American population is only 1.8%, similar to Hawaii (1.6%), Utah (1.1%), and Alaska (3.3%). Only in Vermont and Alaska did Bernie outpace Hillary to a greater extent than in Oregon in the relative number of <$200 campaign contributions. These are the primary drivers of the massive margin of victory that I am projecting. It is difficult to reconcile the one and only poll (that showed Hillary with a 15% lead… but also had 19% undecided… and was also conducted well after ballots had already been received and presumably had already been mailed off by many voters) conducted in Oregon with this projection, but I refuse to arbitrarily tack on extra points because I have a hunch about something.

Oregon is entirely vote by mail. Clinton has traditionally dominated early voting, but Oregon’s format is unique to all the states that have already voted, so it is difficult to predict how much of an effect this will have. Personally, I doubt it will be significant due to practically every metric being overwhelmingly in Bernie’s favor (imagine Vermont was only vote-by-mail, would that have really changed the result?). Also, the party registration deadline was recent, April 26th. Bernie’s current relative search interest is quite low, but Oregonians began receiving their ballots two-three weeks ago. If we go back in the Google Trends data to April 26th (around the day voters began receiving ballots), Bernie’s relative search interest for the next week and a half was around 1.45; not bad. The average for all the states he has won is 2.167, but the standard deviation is 0.61, so 1.45 is not indicative of anything particularly remarkable.

Lastly, Oregon is also a closed primary, which Bernie Sanders has never won before.

Screen Shot 2016-05-16 at 7.33.22 PM

If the above estimates are correct, this should give Hillary Clinton a ~24 delegate deficit tomorrow. Good luck to both candidates, and happy voting to all you Oregonians and Kentuckians!

-Tyler

DEMOCRATIC PRIMARY PROJECTION: WEST VIRGINIA

/ Tyler Pedigo / 39 Comments

Demographically, West Virginia seems like it should be a blowout in favor of Bernie Sanders. The polls, despite being all over the place, also seem to imply that West Virginia will be a blowout, with some giving Bernie a margin of victory as high as +28%(!). Recent polling has come back down to more believable margins, like +4% and +8%, which I think are much more in line with reality. Here is my estimate for what we will see tomorrow in the Mountain State:

Screen Shot 2016-05-09 at 6.50.51 PMWest Virginia is 3.4% African American, similar to states like Washington, Nebraska, and Colorado, all of which Hillary Clinton lost by margins that were quite large. But, there are other factors that point towards a <10% margin of victory. First, the relative search interest between the two candidates is leaning Hillary in West Virginia. The average relative search interest in all of the states that Bernie has won thus far is 2.23, whereas in West Virginia the relative search interest is 0.94; meaning that West Virginians are actually searching for Hillary Clinton more than they are Bernie Sanders. In fact, the lowest his three day relative search interest has ever been in a state that he ultimately won was 1.54 in Minnesota, and just last week in Indiana the relative search interest was 1.66. Secondly, Bernie’s share of Facebook Likes in West Virginia is only 75%, compared to an average of 81.2% for all of the states that he has won. I know that 6.2% doesn’t sound like much, but it really does make a big difference when you consider that the entire range of values for this measure is between 63.6%-95% (Mississippi-Vermont). West Virginia is also a relatively old state, with a median age of 41.9, compared to 37.78, which is the average median age of all the states that he has won.

Aside from this, West Virginia is a semi-closed primary, which usually helps Bernie slightly. In lieu of the Republican race being tied up at this point, it will be interesting to see if the semi-closed primary format attracts all of the independents to vote in the Democratic primary tomorrow. Also, Hillary’s comments on the coal industry have gotten quite a bit of media attention over the past week. This will probably help Bernie (even though Bernie pretty much has the same position on the issue). Also, the logged value of the relative number of campaign contributions is 0.36 in all of the states that Bernie has won so far, and in West Virginia this number is 0.405. This obviously looks positive for the Sanders campaign.

Lastly, on average, I underestimate Bernie’s vote share in semi-closed primaries by 3.27%, so I think we will end up with Hillary performing slightly worse in West Virginia than the numbers above indicate. Regardless, West Virginia is possibly the most unique state thus far with regard to the data that I look at. It’s really surprising how different pieces of data are really trying their best to pull my model very far in different directions in this state.

Thanks everyone, and happy voting to all of you West Virginians out there!

-Tyler

As you all know, I’m just a poor grad student. If you would like to support my work and want me to be able to afford Top Ramen (or maybe even Mellow Mushroom pizza if you all are extraordinarily generous) while I’m working on these statistics, please click this link to donate to Tyler’s Food & Rent Fund!

DEMOCRATIC PRIMARY PROJECTION: INDIANA

/ Tyler Pedigo / 74 Comments

I’m only aware of one other outlet that is projecting a Clinton loss tomorrow in Indiana. Though Bernie Sanders has scaled back spending in Indiana, Hillary has cut all spending from states that have yet to vote in the Democratic primaries, presumably to save funding for the general election campaign against Trump (edgy assumption, I know). This Clinton spending cut seems to be showing up in the Google Trends data for Indiana, as Bernie has seemed to have drastically increased his search interest relative to Hillary. Here are my estimates for what we will see tomorrow night:

Screen Shot 2016-05-02 at 8.01.28 PM

Something about this projection doesn’t feel very right to me, though I suppose this concern is rooted in the surprisingly consistent polling results showing Hillary with a win. Despite this, every different configuration of my model, six total, is showing a Bernie win in Indiana. I have devoted quite a bit of time over this past week trying to see if it was possible to generate a different result, but it just wasn’t possible within my framework. Perhaps Hillary will win, and perhaps Hillary will lose, but regardless, I do think it will be very close.

There doesn’t seem to be any particular factor within all of my data that is significantly driving this result, though if I had to choose one I suppose it would be his slightly higher-than-average share of Facebook likes within the state, 78.5%, similar to Michigan (80.0%), Kansas (78.5%), Illinois (76.1%), and Missouri (76.9%). The Republican open primary that is being held tomorrow in Indiana will help Hillary by stealing independents that would’ve voted for Bernie, likely by around 0.4% (already factored into the above projection).

Michigan is perhaps the best analogue for Indiana. The Facebook data is very similar, and they are both open primaries. I believe that both have also had a minimal amount of early voting. One somewhat stark difference between the two states are the respective portions of the populations that are African American, with Michigan at 14.2% and Indiana at 9.4%, so Bernie should gain an advantage over Michigan in this respect. However, in Michigan, Bernie had a higher relative search interest measure (from Google Trends), about 14% higher. Bernie also had a higher number of individual campaign contributions in Michigan compared to Indiana; where he had about 1.56 times more <$200 contributions than Hillary in Michigan, and only 1.31 times more <$200 contributions in Indiana.

Thanks for the interest everyone, and happy voting to all you Hoosiers!

-Tyler

As you all know, I’m just a poor graduate student. If you would like to support my work and want me to be able to afford Top Ramen (or maybe even pizza if you all are extraordinarily generous) while I’m working on these statistics, please click this link to donate to Tyler’s Food & Rent Fund!

(edit: Also, if you are an employer in the DC area and have an open position, I need a summer job! Please contact me if you think I would be a good fit for your organization. Paid positions only, please.)

FINAL PROJECTIONS: APRIL 26th DEMOCRATIC PRIMARIES

/ Tyler Pedigo / 33 Comments

At the beginning of April, when I posted the April outlook, I had initially believed that Bernie Sanders would be able to make enough progress to at least secure two to three victories on April 26th. For whatever reason, it seems as though Hillary has been able to prevent much of this progress from being made in a majority of the states that have elections tomorrow. Bernie has picked up a few tenths of a point in both Pennsylvania and Rhode Island, while losing support in Connecticut, Delaware, and Maryland. Here are my final projections:

Screen Shot 2016-04-25 at 9.37.55 PM

I am estimating a single Hillary loss in the state of Rhode Island, though with a margin of only 6.4%. This close margin leads me to believe that it could still be a toss up. The estimated margin of the Bernie loss in Connecticut is about the same, 6.2%, and the estimated margin of the Hillary victory in Pennsylvania is 9.4%. If these margins hold true, this will produce a sixty delegate surplus for Clinton. Furthermore, if Bernie loses four or five of these states, I suspect he may drop out of the race. An additional sixty delegate deficit tomorrow will only strengthen the perception that obtaining the nomination is essentially impossible for him at this point.

-Tyler

FINAL PROJECTION: NEW YORK DEMOCRATIC PRIMARY

/ Tyler Pedigo / 76 Comments

A few weeks ago, I posted an outlook for all of the April Democratic primaries. This outlook included my initial estimate for the New York primary, showing Bernie Sanders at 38.8% in the Empire State. We are all aware that both candidates have been campaigning relentlessly in New York, and for that reason I didn’t believe that the needle would really move much from the initial estimate. This assumption of mine is based on the concept of “dynamic equilibrium” that I learned about in my Political Science senior seminar during my undergrad, from a book called The Gamble that covered the 2012 general election. The idea is that, if both candidates are campaigning with approximately the same vigor and intensity in a state, they will both likely get about the same amount of media coverage there, capture the same number of votes in that time frame, etc. It’s a useful way to think about elections. Anyways, it does appear that Hillary Clinton has lost a very small amount of ground compared to my original estimate. Here are my final numbers for New York:

Screen Shot 2016-04-18 at 6.22.46 PM

I am expecting Bernie to do a couple of points better than the original outlook for a couple of reasons. First, his Facebook presence has become barely more favorable than it was, settling out at 70.00% of Democrat likes. This is similar to Virginia (70.37%), Florida (69.56%), and Iowa (71.87%). Secondly, his relative search interest on Google is decent; with the three day relative average coming in at about 2.05-2.1. This is in the ballpark of Illinois (2.05), Oklahoma (1.98), and Nebraska (2.02). The demographic makeup of the state, as well as the closed primary contest format still remain the greatest hurdle to any good Sanders performance in New York.

I am somewhat inclined to believe that New York’s unique primary rules will alter the results of tomorrow’s primary more in favor of Hillary. To vote in the Democratic primary, voters had to have been registered a Democrat by the end of 2015. My belief is that this will most certainly disproportionately affect Sanders supporters. A good friend of mine refers to the New York primary as the most closed primary of the season, and it will be interesting to see if that setup produces results that vary widely from the above projection. Thanks for reading.

-Tyler

WYOMING CAUCUS: FINAL PROJECTION

/ Tyler Pedigo / 59 Comments

The Wyoming Caucus is tomorrow, and though not very consequential in terms of delegates (with only fourteen up for grabs), Wyomingites are apparently poised to give Bernie Sanders another win. And yes, the official demonym of those who reside in Wyoming is “Wyomingite.” Here is my projection:

Screen Shot 2016-04-08 at 8.24.51 PM

Hillary Clinton will perform poorly for the following reasons:

  • Wyoming has a closed caucus, which Bernie has consistently done very well in. He has won all seven closed caucuses to date (counting the recent Nevada flip).
  • Wyoming has the second lowest population of Black voters in the nation, 0.8%, second only to Idaho.
  • Bernie Sanders has a greater social media presence in Wyoming than every other state that has voted so far, with the exception of Vermont.
  • Hillary Clinton has a very low amount of Google search interest in Wyoming; it’s her third worst state in this regard, behind only Idaho and Vermont.

I wouldn’t be surprised to see Hillary drop below 20% in Wyoming by the time the caucus results are in. As we witnessed in Alaska, at caucus locations where Hillary has very little support, she runs the risk of being deemed a non-viable candidate (this threshold is 15% at all caucuses as far as I’m aware) and being awarded no delegates at that location.

Thanks for the interest folks, and Wyomingites, happy caucusing!

-Tyler

POLITICAL FORECASTING: AN IMPRECISE SCIENCE

/ Tyler Pedigo / 11 Comments

Lately, there have been several occasions where people have compared the accuracy of my forecasts to FiveThirtyEight’s. As a big fan of Nate Silver and the rest of his team, I’m flattered that my work is even being talked about in the same conversation as theirs. I’m actually indebted to FiveThirtyEight anyways, because I use the Facebook data that they publish for free; and that variable is pretty much the cornerstone of my entire model. Granted, I’ve tried contacting the Facebook Data Science team on four different occasions to get the data personally, but a small fish like me can’t get a response (which is just fine, I know they have many other important things they are working on).

With all of that being said, I’d like to set some things straight:

  • I deeply respect the work of Nate and FiveThirtyEight, and I think they do a fantastic job.
  • It’s nice to be accurate, but if some other institution turns out to be more accurate than myself, it’s not as if I resent them for it. I’m happy for others when it turns out that they’ve done a good job through a solid analysis. That’s what this is all about.
  • This is a hobby of mine, and I do this because I think it is a fun mental exercise.

It’s tough to estimate the outcome of any election with perfect accuracy. There are hundreds (or maybe even thousands) of variables that can be used in a model, and the goal is to choose the fewest number of variables that possess the most predictive value. I receive criticism all the time because I’m not taking into account a certain thing, and I totally understand that many of those suggestions have predictive value. For many, however, I’ve already tested them and decided not to include them, most likely because the effect wasn’t statistically significant and/or I’m already capturing that effect through another variable.

Now, to address those who have suggested that I’m outperforming FiveThirtyEight. Simply put, I’m not. FiveThirtyEight has been closer than I have on average. Here is a breakdown of why that is the case:

  • FiveThirtyEight has, on average, been more accurate than me in the elections that we have both released projections for, which is 19 out of 32 states. I have published projections for 29 of the Democratic primaries/caucuses (I started after Nevada), and FiveThirtyEight has published projected results for 22.
  • FiveThirtyEight’s overall average error for the contests that they publish projections for is 3.2%, and their median error is 2.6%. My average error is 5.8% and my median error is 5.3%. I have only been closer than FiveThirtyEight in seven contests, whereas they have been closer than me twelve times. See the following graph to visualize this.

Screen Shot 2016-04-06 at 2.49.38 PM.png

  • As for calling wins and losses correctly: For Missouri and Illinois, FiveThirtyEight projected Hillary to win, and I was projecting her to lose. FiveThirtyEight was correct both of those times. On the flip side, in Michigan and Oklahoma I projected Bernie to win, and FiveThirtyEight was projecting him to lose. Bernie won both of those. In Minnesota and Arizona, FiveThirtyEight didn’t publish any projections that I’m aware of.
  • If you wish to recreate these numbers yourself, keep in mind that I publish projections for the two candidates excluding votes for other candidates. This means that my projections between Hillary and Bernie always add up to 100. FiveThirtyEight publishes projections based on polling, so that necessarily includes votes for other candidates. Thus, my projections must be measured against the results after adjusting them to the aforementioned scale (=100/(BernieVote + HillaryVote)*BernieVote), while FiveThirtyEight’s can be compared outright.
  • I have underestimated Bernie Sanders overall. The sum of all of my errors (AdjBernieResult – MyBernieProjection) is 1.4. The sum of all of the absolute values of my errors is 174.1. See the graph below (excludes the first three states because I started publishing results after Nevada). The x-axis denotes the contest number, i.e. Iowa would be 1, New Hampshire 2, and so on.

Screen Shot 2016-04-06 at 2.47.09 PM.png

So if I’m doing worse overall, why are my projections valuable?

I think my work is worth something. Maybe you don’t, and that’s perfectly fine. However, I know that I would want to get an idea of what would happen in some states when Nate hasn’t been able to publish any projections, and that has happened eleven times so far (mostly in caucus states). Let me be clear that it’s through no fault of FiveThirtyEight when they don’t publish projections, and they would likely publish projections for every state if there was always enough recent polling data, but that’s not always the case.

This is, in my opinion, the beauty of using the data sources I’m using. There is zero reliance on polls. Not that polls are a bad thing, because FiveThirtyEight demonstrates that enough polling data is an extremely powerful predictor all the time (See their Ohio, Vermont, Georgia, Virginia estimates all within one point of the result(!)). But, once again, polling isn’t always performed everywhere, and that’s where I think my work has the most value. Obviously my work isn’t particularly great all the time and I’ve had some major misses, but it’s still pretty cool to be in the infancy of this new methodology. I’m confident that this approach will become the new standard for political forecasting in the future and replace polling as the primary data source for predicting elections, and until then I will continue to refine my work to produce better results. Thank you everyone for the interest in what I do.

-Tyler

WISCONSIN PRIMARY: FINAL FORECAST

/ Tyler Pedigo / 32 Comments

Hello everyone,

Bernie Sanders seems to have maintained about the same projected lead as indicated in my previous post. Recent polling, the Benchmark Politics’ benchmark, and the FiveThirtyEight projection seem to corroborate this. Here is my final estimate:

Screen Shot 2016-04-04 at 6.21.48 PM

This is a particularly strong number, because Hillary Clinton has generally done much better in the open primary format, with the exception of Vermont and Michigan (though Michigan may as well have been a tie). Still, the demographics of Wisconsin favor Bernie more than Hillary, with an 83.3% non-Hispanic White, and 6.3% Black population. If the above numbers are accurate, this would produce a delegate allocation of 35 for Clinton, and 51 for Sanders.

However, it has been reported that there has been a record breaking number of early voting in the last two weeks in Wisconsin. This has strongly favored Hillary in previous contests, and it stands to reason that it will likely hold true in Wisconsin as well. For this reason I expect Hillary’s vote share to be slightly higher than the above number (10:11 PM edit: A recent Emerson poll shows Clinton trailing Sanders in early voting 38% to 52% which seems to indicate the opposite is true).

-Tyler