Bernie’s Narrow Path to Victory: a Statistical Analysis

shutterstock_335861132 (1)

Bernie Sanders is not yet realistically eliminated from the race against Hillary Clinton for the Democratic Presidential nomination but conceding Florida without a fight would make victory nearly impossible.

Wins by Sanders in Minnesota and Colorado along with his home state of Vermont were expected, as were tight races in Oklahoma and Massachusetts. While a narrow defeat in the quintessentially liberal Massachusetts would have ended Sanders’ chances if Nate Silver’s July 2015 White and Liberal narrative of the campaign were reality based, the size of Sanders’ victories in Minnesota and Colorado and especially Oklahoma breathed new life into Sanders’ campaign, as did news of a $45 million dollar fundraising haul from small donors in the month of February.

***

I have made projections in the map below, should Sanders chart a path to victory, by way of an analysis of states similar to Oklahoma, Minnesota, Colorado, Vermont, and New Hampshire as well as of the states most favorable to former Secretary of State Hillary Clinton. While there are a few wild card states, the greatest outstanding factor to date is how Senator Sanders can perform in delegate rich Rust Belt states Illinois, Ohio, Pennsylvania, and Michigan. Michigan votes next Tuesday, and Senator Sanders’ campaign is hitting the boom in campaign ads with opposition to massive trade deals supported by Clinton during her time in office but repudiated whenever she campaigns. With the exception of a single poll in Ohio showing Sanders with a narrow edge, most polls have showed him trailing by double digits in all four states. Beyond the Rust Belt and the Wild Cards, top ten, Democratic establishment states may, like Massachusetts, may prove tough ground for Sanders. I have assessed his reasonable chances in them through to California in the final section.

delegate-map-colored-numbered-final

A Path to Victory in Pledged Delegates for Bernie Sanders? http://wp.me/p7gBVo-3j


18 States Sanders Should Win, In Three Groupings

Comparing and contrasting the fundamentals of Minnesota and Colorado on the one hand and Oklahoma on the other yields enormous insight beyond the tired White and Liberal narrative set out by FiveThirtyEight. It is becoming clearer that the race is not about white voters versus all people of color. The campaign, as on the Republican side, should primarily be seen through the lens of establishment versus anti-establishment politics. Black voters are naturally proud of Barack Obama as the first President to snap the unbroken chain of white, male United States Presidents and are therefore heavily siding with Secretary Clinton who served as third in command in his administration and is vowing to continue his center-right policies. Where Bernie Sanders fails to compete, such as in Texas, he also risks losing other voters of color by substantial margins.

In Oklahoma, however, exit poll break down of the non-white vote makes it clear that Sanders won the Latino (4%), Native American (grouped under ‘other’ at 8%), and Asian (1%) vote by approximately 36%. While no exit or entrance polls were taken, Sanders won a substantial victory in Colorado, where Latina/os make up 15% of the state’s eligible voters. No side has clearly won the argument over Nevada’s hispanic vote, but it is clear that Sanders was extremely competitive for the Latina/o vote in that state as well.

Minnesota and Colorado are Democratic leaning purple or swing states while Oklahoma was named in 2015 by Gallup as one of the top ten most solidly red states in the United States. According to numbers from Business Insider for the fourth quarter of 2015, Colorado’s economy was the 3rd strongest in the nation and Minnesota’s was the 15th strongest; Oklahoma’s was 49th out of 51 (including Washington D.C.’s).

9 States Similar To Oklahoma

These nine states could vote similarly to Oklahoma based on sharing three out of four important factors: economics, region, ethnic composition (i.e. states with a black population below the nation average of 13%), and red state political leanings:

West Virginia – is a mostly reliable red state with an economy rated dead last out of fifty-one in the final quarter of 2015. One poll has already suggested that Democratic voting Mountaineers may go for Sanders by as much as 28% points.

Kansas – is a reliably red state with an economy ranked 41st. It is located just North of Oklahoma and 86% of its ethnic composition is of non-hispanic, white only people, according the United States census estimates for 2014.

Missouri – a few states North and East of Oklahoma, Missouri’s economy was ranked 42nd of 51. Thirteen percent of the state’s population is African American (putting it right on the line of states in this category – worth watching) and just 2% identify as Latina or Latino.

Wyoming – reliably red, 47th ranked economy, 84% white, 10% Latino.

North Dakota – it’s economy plummeted from the top ten to 43rd in the nation between the 3rd and 4th quarter based exclusively on the bottom dropping out of the price of oil, creating havoc for its fracking-heavy economy. 87% of its population is white, 5% Native American.

Alaska – the state that gives us the gift of Sarah Palin has the 40th ranked economy in the United States, with 62% of its population identifying as white only (compared to 74% for Oklahoma), 15% as Native Alaskan or American, and 7% Latino.

Kentucky – reliably red, 39th worst economy, 85% white alone, 8% black, 3% Latino.

Indiana – a traditionally red state with the 36th best economy, 80% white, 10% black, 6% Latino.

Finally, while reliably a blue state (though never in the top ten most blue states in the last seven years according to Gallup), I am including New Mexico in this grouping as it is geographically close to Oklahoma, has an economy ranked 50th, and its very large not-solely-white population is nearly 50% Latino and 10% Native American.

Together, these nine states will contribute 358 delegates to the Democratic race for president. I am going to project a number similar to Oklahoma’s 10.5% spread at 56.5% for Sanders. Approximately 202 delegates would go to Mr. Sanders, 155 to Ms. Clinton.

Screen Shot 2016-03-03 at 4.29.41 PM-1

7 States Similar to Minnesota and Colorado

In this category, I am expecting states to average a 22-25% spread for Bernie Sanders. These are Western or Midwestern states with healthy economies, black populations of 10% or less, and none are in the top ten most reliable blue states. In fact, based on polling in Utah suggesting a 22% lead for Sanders, I am including four reliably red states. Interestingly, all four red states show approximately an average liking for Sanders per FiveThirtyEight’s “Facebook Primary” (compared to his national average) while Secretary Clinton performs below or well-below her average in each of these four red states (in Oklahoma, Sanders was performing about 35% below his national average in the Facebook Primary).

The three blue leaning or solidly purple states are Oregon (13th economy), Washington (1st), and Wisconsin, which has only the 30th best economy but is included in this group rather than in the rust-belt grouping because of its close proximity to Minnesota geographically, ethnically, and culturally. Its black population, at 6.6%, is well below the next Rust Belt state’s (Pennsylvania’s at 11.6%).

The four red states are Utah, Nebraska, South Dakota, and Idaho ranked 4th, 5th, 8th, and 11th respectively in the last economic quarter of 2015.

Together, these states should give Sanders an approximately 85 pledged delegate advantage over Clinton.

Two States Like Vermont (24th economically) and New Hampshire (14th)

I expect Rhode Island (33rd) and Maine (34th) to go for Sanders by 30 percentage points if he is on track to victory. This would give Bernie just a 15 delegate advantage in this subcategory.

For these eighteen states as a whole, Bernie could approach a 150 delegate advantage, nearly enough to overcome Clinton’s current 193 pledged delegate head start (if it weren’t for the next section).

Six Remaining “Firewall” States Where Clinton Wins Handily

There are six remaining states (including Maryland and Washington D.C.) that have well above average African American populations and which therefore make up the last portions of Senator Clinton’s infamous “firewall.” We should expect Ms. Clinton to win again by around 50% in Mississippi, Louisiana, and Washington D.C. for an additional 53 delegate advantage. The Sanders campaign really ought to consider competing more strongly in North Carolina and Maryland then it did in Virginia. Absent that, we should expect two more 30% losses and another 60 delegate bump for the Clinton campaign. At this point, absolutely the only way I see for Sanders to compete through June is if the campaign goes all in and somehow holds Florida to a 10 point loss on March 15. A repeat of Texas, with a +30% loss in a state worth more than 200 pledged delegates, would be disastrous. If he can hold it to a 10 point defeat, Clinton would gain 22 additional delegates for a 135 delegate advantage on top of the 193 noted already on the bottom of the map.

The Rust Belt Four

Michigan (March 8th), Illinois and Ohio (March 15), and Pennsylvania (April 26) may well be the deciding factor in whether Sanders can survive the lead built by Clinton with the Southern Firewall. Each of the states has a black population roughly equivalent to the U.S.’s overall black population of 12-13% and Michigan and Illinois also have Latino populations which qualify it for The Latino Gauntlet (11 of the top 20 Latino states by population that vote within a single month during the primary and caucus cycle in 2016). I am projecting that Sanders needs to win these states by an average of 15% to have a chance at the nomination. This would give him a 92 delegate jump on Clinton. Based on where they line up in the primary calendar and their racialized compositions, I have suggested that this could be accomplished with a 6% victory on March 8th in Michigan, an even race in Illinois on the 15th with an important 18% point win in Ohio, which was hit hard by NAFTA, on the same day, and finally a large, 30 point victory in Pennsylvania (drawing from the lessons in previous Rust Belt states) on April 26th. The Sanders campaign has bet big on Illinois and Michigan showing a different face of the African American community. We simply have no idea if a 15% average win in these states is possible as no voting has taken place in similar states yet and polling is virtually non-existent in three of the four states. In Michigan, five polls over the last two weeks of February ranged from a Clinton lead of 10 to 33 points for a 19.2 point average deficit for Sanders. This would be quite a turnaround, indeed.

Wild Cards

Provided Sanders can hold serve in 18 states similar to previous victories, make Florida competitive, and win handily in the Rust Belt, to track with my victory in pledged delegates projection he will need to secure something like an eight delegate advantage in three wild card states with middling, diverse economies: Hawaii (21st best economy, reliably Blue), Arizona (25th, reliably Red, a large Latino population), and Montana (31st, reliably Red) for which we have very little information on which to make a projection. For now, I have suggested that a small, 5% victory in each state could reach the goal.

The Big Blues

If I am correct in assessing the anti-establishment versus establishment nature of the campaign-to-date, large and reliably Democratic states with healthy economies (in the top twenty or even ten) may prove to be too much like Massachusetts for a Sanders win. The five states in this category are New York, California, New Jersey, Delaware, and Connecticut. I am projecting tight races in Connecticut and Delaware, similar to Massachusetts, with hopes that Sanders may squeeze a couple of delegates out of the tension to take back what was lost in Massachusetts. I am going to give Clinton some home field advantage in New York and neighboring New Jersey for a 5% advantage in each. This would give Senator Clinton an additional 19 or 20 delegate advantage.

All of this, if my math and projections are correct, would leave Bernie Sanders with a need for a closing 20 point win in California on June 7th. I’d like to think my home state and the land of my mothers and grandmothers going back five generations is capable of delivering such a win, but to even have such a chance available, Senator Sanders will have win almost everywhere, all the time, in every region of the country except the South, and by substantial margins in many many places. At this point, the Rust Belt, Florida, and New York and New Jersey seem most daunting and therefore most likely to deliver Secretary Clinton the early elimination of Sanders she has basically already claimed.

To me, it would be quite shameful and dangerous for the Democratic party if she could do so without ever scoring a decisive victory in a reliably blue or very large blue leaning state.

Doug Johnson Hatlem writes on polling, elections data, and politics. For questions, comments, or to inquire about syndicating this weekly column for the 2020 cycle in your outlet, he can be contacted on Twitter @djjohnso (DMs open) or at djjohnso@yahoo.com (subject line #10at10 Election Column).