Hillary Clinton was expected, by all realistic analysis, to win the 2016 Presidential Election. It was in the polling, the demographics, and the electoral map. Moody’s, which has correctly predicted every Presidential Election since 1980, predicted Clinton to win.[1] FiveThirtyEight, one of the most accurate statistical analysis sites out there, gave Trump a 35% chance, at best, of winning in the week before the election.[2] This was actually more generous than some other sources.The New York Times gave Clinton an 85% chance of winning while The Huffington Post gave her a 98% chance.[3]
As we now know, everybody got it wrong. Donald Trump won the election, arguably the biggest upset since 1948 when Harry Truman defeated Thomas Dewey despite the polls, most newspapers and experts expecting the opposite. How the polling got it so wrong will have to be dissected in the future. I had a more immediate question: Would Bernie Sanders have won the election? I believe the answer is yes. I fully admit it’s impossible to ever make such a state with certainty, we can never know what may have happened. However I will make the argument using data, and hopefully some logical reasoning. Just for the record, this isn’t simple second guessing after the fact. In March 2016 I made a post explaining why I believed Bernie Sanders could win the election, if he was to won the nomination.
First, we must understand why Clinton lost the election. Many argue it’s because of the Electoral College. It’s true Clinton won the popular vote, currently by a margin of 2.2 million, yet lost the electoral vote. However, since we have the Electoral College, (and I’ll discuss this issue in the future) I feel it’s best to look within the system we have. With several weeks of data and analysis out, it’s safe to conclude that Hillary Clinton lost the election because of 4 reasons:
1: White working class voters, especially in the rust belt
2: Lower youth turnout than in 2012
3: Independents
4: General unpopularity
It’s been well documented that white working class voters were Clinton’s weakest area of support, and Trump’s strongest. These white working class (WWC) voters constitute a large bloc in the rust belt, and their strong support for Trump is why he won the reliably blue states of Pennsylvania, Wisconsin and Michigan, and thus the election. They also drove states like New Hampshire, Maine and Minnesota to surprisingly close results. Even strongly blue New York state gave the Democratic nominee their weakest support since 1992, again driven by WWC voters who dominate the upstate region, also part of the rust belt.
Macbomb County, MI a suburb of Detroit, is the epitome of the WWC. Macomb backed Obama both times with over 50%, and backed the Democrats 4 out of the last 5 elections. Macomb backed Trump by over 48,000 votes, Clinton lost Michigan by under 11,000 voters.
The Lehigh Valley is a Democratic stronghold in Pennsylvania. Luzerne County, which has backed the Democrat every time since 1992, (usually by over 50%) favored Trump by 26,211 votes. Northampton County has also favored the Democrat every time since 1992, and favored Trump, by 5,448 votes. These two, usually reliable blue, counties are responsible for nearly half of Trump’s victory in Pennsylvania. Erie County is another unionized, working class hub that has voted Democrat every election since 1988, (both times for Obama with over 57%) that broke for Trump.
By comparing results from 2008, 2012 and 2016 you can see Democratic support in Wisconsin, a stronghold of union/labor politics, was wiped out. (In the maps used below, Democrats are red, Republicans are blue).
In Florida, 4 counties flipped from blue to red this year from 2012: Pinellas, Jefferson, St Lucie and Monroe. All voted for Obama twice and have generally supported Democrats since the 90s. In addition, 3 other counties: Pasco, Hernando and Volusia came out for Trump far stronger than they did for Romney, or McCain or Bush. Trump’s victory in these 7 counties was over 125,000 votes. Trump won Florida with about 113,000. These 7 counties account for Trump’s unexpected victory in Florida, and there is a common thread among these counties: They are largely white, suburban, working class areas. [4][5][6] Clinton did exceed Obama’s 2012 numbers in some Democratic areas of Florida, and scored stronger than usual results in the Republican counties of Duval and Seminole, yet all this was wiped out by Trump’s strong support from the working class suburbs of Tampa-St Petersburg and Daytona Beach.
Trump’s strong support from this bloc is largely due to his criticism of NAFTA, TPP and trade with China, part of his overall anger about the decline of US manufacturing and job loss. This is an area Clinton was weak in. Even though she campaigned against TPP, her previous calling it the “gold standard” of trade deals, then abruptly turning against it, as well as her previous support of NAFTA and insider status made it difficult for her opposition to be taken too seriously. Trump also pledged to not touch Social Security, (as well as Medicare and Medicaid) a total break with Republican orthodoxy since Reagan. This is an essential program for the working/middle class, and Trump’s promise, (we have to see what reality brings) to not touch it likely helped comfort working class voters about him.
While many point to race as a factor, and this is an issue that certainly can’t be ignored in this race, the states of WI, MI and PA voted for Obama twice. Same for Ohio and Iowa. All predominantly white. In fact they all backed Obama stronger than they did for Kerry, Gore or Bill Clinton. Obama won white, working class counties in these states, (as well as New England) that Clinton lost. Polls show Trump’s support from black, hispanic and Asian voters actually improved from Romney’s.[7] The gender issue is another crucial one, but again polls show women voted in 2016 along normal lines. Women support for Clinton actually fell 1 point from 2012, while support for Trump fell just 2 points.[7] A 538 analysis shows white women, especially those without college degrees, backed Trump and Republican women voted for Trump in largely normal numbers.[8]
While racial and gender issues are crucial and can’t be ignored, it’s also inaccurate to boil the results down to just these. Especially since a 538 analysis shows Trump did best in counties where the economy was weaker (lower job growth, avg wage) and with higher economic anxiety (where jobs are easier to automate/be sent overseas). This held true in counties that swung from blue to red.[9]
Youth voters (below age 30) turned out for Clinton with 55% support. This is down from 60% in 2012.[7] Given their support for Clinton, and Democrats in general, this drop in turnout was a big loss for Clinton especially with several states being so tight. Particularly large drops in youth turnout from 2012, 19 and 20% were seen in the states of Wisconsin and Pennsylvania.[10]
Independents, which make up the largest bloc of voters in the US, backed Trump 48% to 42% a gap that rose 1% in favor of the Republicans from 2012. This is the lowest Independent support the Democrats have received since 1996.[7]
Finally, Clinton was simply not a popular candidate. Neither was Trump, who was consistently more unpopular. However, Clinton did have trust and image issues. The emails, the DNC scandal, Wall Street ties and refusal to release her paid speeches, questions about the Clinton Foundation all hurt her trustability. James Comey’s questionable last second reopening of the email issue was damaging, most of her lead in the polls got obliterated. She has the image of an insider, and elitist, part of the problem of government that has failed to help the people. Trump, despite his disturbing personal issues, comments and actions, was an outsider, and spoke to the people of the rust belt, became a vehicle for their economic frustration and worry. An article sheds light on some Trump voters in Erie County, PA (including former Democratic voters) and it seems many were voting not out of like for Trump, but dislike of Clinton, and a yearning for change.
It mentions many of the good paying manufacturing jobs were being sent to Texas, not overseas, but this may not have mattered. Trump gave these people a voice, one many feel has been abandoned by the Democratic Party (at least through candidate Clinton).[11] This, with the hope maybe Trump can achieve some positive change, was enough for people across the rust belt to hold their noses. Clinton’s opposition to TPP was simply not believed by many people. Her dedication to Social Security, jobs, wages, the middle and working class which were all likely sincere, simply were difficult to believe for some voters.
So, to Bernie Sanders. Would he have fared better? Remember the 4 factors that lead to Clinton’s loss: WWC voters, lower youth turnout, loss of independents and image/trust issues. Sanders had very strong support from the white working class. This was one of his strongest blocs of support in the primary. He has a superb record on the issues: voted against every trade bill, came out early and strong against TPP, pro union, higher wages, and ran an undisputed pro “main street” anti elitist message. His other strong bloc of support came from the youth. In the Democratic Primaries, Sanders consistently won the under 30 vote, even in many southern states he lost badly, often winning with over 70 or 80%[12] Sanders won more youth voters in the primaries than Clinton and Trump combined, and his support didn’t waver even as it became clear Sanders would not win the nomination.[13] His youth support was massive.
Independents also favored Sanders fairly strongly and consistently through the primaries.[12][14] Unlike Clinton and Trump, Sanders had a positive favorability rating, the only candidate of both parties to achieve this.[15] As Sanders’ name recognition grew, so did his popularity. As he became more known his image did not become more mixed, the opposite in fact, the more people got to know Sanders, the more popular he became finishing with a +16.6 favorable rating. Contrast with Clinton who’s numbers got worse over the campaign finishing with a -15.2 favorable rating. [16][17] Note Clinton’s downward trend begins Feb 2013, and takes another downward trend at the start of 2015, both before Bernie Sanders hit the scene.
So, what would’ve happened if Sanders won the nomination and squared off against Trump? We can’t ever know definitively, who knows how the campaigns would’ve been run, the tactics used. I will say though, the areas that cost Clinton, were Sanders strong suits. We can’t know how white working class voters in the rust belt would’ve broken, but given the fact they are usually blue states, and Clinton lost all 3 by a total of just 107,000 votes, it’s not unrealistic to imagine Sanders winning enough votes, in these normally Democratic states, to win all 3 and thus the election. Especially when you take his popularity vs Trump’s into account.
While it’s also impossible know how youth turnout would’ve been, given Sanders’ extreme popularity with the youth bloc, I imagine there would’ve been some amount of improvement over Clinton’s result. With MI, PA and WI being so close it would not have taken much to tip the results Democratic. Same with Independents. We can’t know how they would’ve broke, but given his popularity with Independents one would imagine he would’ve fared better than Clinton.
To conclude, I believe Bernie Sanders’ strength with white working class voters, the youth, independents and his general popularity would’ve been enough to win the usually Democratic states of Michigan, Pennsylvania and Wisconsin and thus the election. Especially against the immensely unpopular Donald Trump, who didn’t manage to win even 49% in any of the 3 states. There would’ve been no email story to constantly hammer home, nor could Comey have dropped a last second bomb, no DNC leaked email scandal, no Wall Street or Clinton Foundation $ ties to raise suspicion. The 4 major factors behind Clinton’s loss would’ve been strengths in a Sanders campaign, and the issues that dogged Clinton would no longer have been issues.
Sources/References:
1: http://money.cnn.com/2016/11/01/news/economy/hillary-clinton-win-forecast-moodys-analytics/
2: http://projects.fivethirtyeight.com/2016-election-forecast/
3: http://www.huffingtonpost.com/entry/nate-silver-election-forecast_us_581e1c33e4b0d9ce6fbc6f7f
4: http://www.tampabay.com/news/politics/stateroundup/donald-trumps-strength-in-florida-shows-in-working-class-pasco-county/2293878
5: http://www.orlandoweekly.com/Blogs/archives/2016/11/10/blue-collar-counties-key-for-trump-in-florida#
6: http://www.miamiherald.com/news/politics-government/election/article113786328.html
7: http://www.nytimes.com/interactive/2016/11/08/us/politics/election-exit-polls.html
8: http://fivethirtyeight.com/features/clinton-couldnt-win-over-white-women/
9: http://fivethirtyeight.com/features/trump-was-stronger-where-the-economy-is-weaker/
10: http://www.npr.org/2016/11/14/501727488/millennials-just-didnt-love-hillary-clinton-the-way-they-loved-barack-obama
11: http://www.nytimes.com/2016/11/13/us/politics/pennsylvania-trump-votes.html?_r=0
12: http://www.cnn.com/election/primaries/polls
13: https://www.washingtonpost.com/news/the-fix/wp/2016/06/20/more-young-people-voted-for-bernie-sanders-than-trump-and-clinton-combined-by-a-lot/?utm_term=.4af10160baa5
14: http://fivethirtyeight.com/features/the-hidden-importance-of-the-bernie-sanders-voter/
15: http://fivethirtyeight.com/features/donald-trump-is-really-unpopular-with-general-election-voters/
16: https://elections.huffingtonpost.com/pollster/bernie-sanders-favorable-rating/edit#!minpct=0&maxpct=70&mindate=2008-01-01&maxdate=2016-11-09&smoothing=moderate&showpoints=yes&showsplines=yes&hiddenpollsters=&hiddensubpops=&partisanship=S,P,N&parties=D,R,I,N&selected=favorable,unfavorable&fudge=1
17: http://elections.huffingtonpost.com/pollster/hillary-clinton-favorable-rating
The World at Large 