In the sixth of a series entitled U.S. Election Notebook, Athabasca University professor Paul Kellogg looks behind the numbers, to see just who likes and dislikes the two main U.S. presidential candidates – Hillary Clinton and Donald Trump. In the fifth of the series, professor Paul Kellogg sees the shadow of the anti-civil rights figure George Wallace in the current U.S. presidential election.
It has become commonplace this election cycle to call Donald Trump and Hillary Clinton the two most disliked major party standard-bearers in history. An August 25 poll showed that 53% of voters had a strongly unfavourable view of Trump, compared to just 21% with a strongly favourable view. Clinton’s figures were only slightly better, 46% having a strongly unfavourable view of the Democratic candidate, just 25% with a strongly favourable view.
But these statistics actually conceal more than they reveal. When we restrict the survey universe to just white voters, Donald Trump’s approval ratings are much higher than his overall figures, while Hillary Clinton’s are much lower. By contrast, restrict the universe to African-Americans and Latina/Latinos, Trump’s support is unbelievably low, while Clinton’s support is consistently high. In this election cycle, emotion has a colour.
This can be quantified. The polls conducted by Quinnipiac University are quite reputable (given an “A-“ by Nate Silver in his pollster ratings) and have the advantage of breaking down their polling sample into different groups by gender, ethnicity and age. In their June 29, 2016 poll, Quinnipiac showed a virtual dead-heat – Democrat Hillary Clinton being the choice of 42% of those polled, Republican Donald Trump the choice of 40%.
However, the profile of the two candidates’ support was completely different. Figure 1 shows overall support for Donald Trump, followed by key categories identified by the pollsters, in descending order of by level of support. Figure 2 repeats this for Hillary Clinton. The results are fascinating.
One common view of the Trump phenomenon, is that its bedrock is among one of the poorest sections of U.S. society, those without a college education. On winning the Nevada Republican caucus, he famously proclaimed: “We won with poorly educated. I love the poorly educated!”.
He certainly does have the support of large sections of poor white voters, and not just in Nevada in the west. Devastated by job losses in coal mining, the Appalachian region in the east is overwhelmingly populated by poor, working class, white voters – and the region overwhelmingly backed Donald Trump in the Republican primaries. “Trump won a complete landslide in the region and carried 295 county units” out of the 311 which comprise Appalachia.
However, a careful analysis of the statistics presented here, troubles this common view. In the category of voters, “no college degree” while Trump does have more support than Clinton, the difference is not dramatic – 43% to 37%. And interestingly, while Clinton’s 47% support of those with a college degree is comfortably ahead of Trump’s, her lead is not insurmountable. Fully 37% of U.S. voters, with a college degree, are planning to vote for Trump. His base cannot be reduced to the uneducated, manual working class. Thousands of people with college and university degrees will walk to the polls and put an “X” beside his name. To find the key divide between Clinton and Trump, we are going to have to look somewhere other than a divide based on education.
Trump’s strongest support
Trump’s strongest support is among white men (56%), followed by seniors (51%), the overall category of white (47%) and the overall category of men (47%). There is nothing terribly surprising in these figures. They exactly match the identity of the candidate – a 70 year-old white man. Trump does quite poorly among women (33%), slightly better with the narrower category of white women (39%), is struggling with the Latina/Latino vote (33%), and has very little support among young adults (23%). What jumps off the page is his level of support among African-American voters. The astonishingly low figure of just 1% is the kind of figure rarely found in samples of support for candidates of major parties. The margin of error in the poll conducted was plus or minus 2.4 percentage points. That means statistically, the 1% figure should really be presented as the mid-point of a range between plus 3.4% and minus 1.4%. Of course it is impossible to have negative support for a candidate. A more reasonable way for a statistician to approach this, would be to take Trump’s support among African-Americans, and round it to zero.
Clinton’s strongest support
By contrast, the profile of Clinton’s support only partially syncs with her own identity, a 68-year old white woman. She does have strong support among women (50%), but has much weaker support in the narrower category of white women, only 42% of whom support her. In fact Donald Trump only trails her by 3% in that category. Her support among seniors is just 35%, a lower figure than the 48% of young adults who support her. She is doing very well with the Latina/Latino vote, with 50% support. Finally, two areas jump off the page. One is her extraordinary level of support among African-American voters, an astonishing 91%. Second is her extremely low level of support among white men, in this poll having the support of just one in four (25%).
In 1903, the great African-American scholar W.E.B. Du Bois picked up a term used earlier by Frederick Douglass. “The problem of the twentieth century,” he wrote in one of his masterpieces, The Souls of Black Folk, would be “the problem of the color line.” The picture painted here, of the United States in the 21st century, is of a country divided in large measure by that colour line, one camp being mostly white, mostly male and mostly older; the other populated disproportionately by people of colour (in particular African-American), young people and women.
In the next post, entitled Glass ceilings and backlash, Paul Kellogg examines the shocking sexism frequently directed against presidential candidate Hillary Clinton.
U.S. Election Notebook #1: Grief of the Parents
U.S. Election Notebook #2: Memories of Goldwater
U.S. Election Notebook #3: Calculus of ‘misspeaking’
U.S. Election Notebook #4: Birther origins
U.S. Election Notebook #5: Memories of George Wallace
U.S. Election Notebook #6: The Colour Line
U.S. Election Notebook #7: Glass ceilings & backlash
U.S. Election Notebook #8: Slaves in the White House
U.S. Election Notebook #9: Locking up the vote
U.S. Election Notebook #10: The New Jim Crow
U.S. Election Notebook #11: Not a dime’s worth of difference?
U.S. Election Notebook #12: ‘A sort of public and psychological wage’
U.S. Election Notebook #13: Political futures
Paul Kellogg (Ph.D. Queen’s, M.A. York) is an Associate Professor, Centre for Interdisciplinary Studies, Master of Arts – Integrated Studies Program, Athabasca University. Among his courses at Athabasca is POLI 345, “American Government and Politics”. He is the author of Escape from the Staple Trap: Canadian Political Economy After Left-Nationalism (University of Toronto Press, 2015). His teaching and research interests are political economy, social movements and global governance.