During the build up to the US 2016 General Election an analysis conducted by a senior economist at Gallup found that the widely assumed link between economic hardship and support for Donald Trump and his policies was overstated.
The analysis (which was based on a sample of over 80 000 respondents) found that:
- those with a favourable view of Trump were no more likely to be unemployed than those who see him unfavourably.
- those who viewed Trump favourably earn household incomes 6% higher (unadjusted) than those who see him unfavourably.
- having a blue collar occupation predicted a more favourable view of Trump, but so did higher household income.
The Gallup analysis also found that because older, less educated men (the prototypical Trump voter) tended to live in industrialised metropolitan areas, demographics may also be a key driver of the correlation between support for Trump and the manufacturing orientation of a local area.
Another interesting finding was that workers who appear to have been just as exposed as others, if not more so, to competition from immigrant and foreign workers were no more likely to hold favourable views of Trump, a finding consistent with contact theory.
The findings of the Gallup analysis question widely held assumptions about both the type of people Trump appeals to, and why he appeals to them. The key swinging states in this election featured large swathes of industrialized metropolitan areas. It may be that the demographics of these locations ensured Trumps victory rather than his assumed resonance with working class Americans.
In the aftermath of Trump’s stunning upset exit polling by Edison Research validated many of the findings discussed above.
Of those surveyed by Edison (just over 24 000 respondents):
- Those who earned under US$30 000 a year overwhelmingly voted for Hillary Clinton over Donald Trump – 53% for Clinton versus 41% for Trump.
- For the next income tier – US$30 000-$49 999 per year – the results were similar – 51% for Clinton versus 42% for Trump.
Conversely, when we reach the US$50 000 + income tiers the trend is reversed:
- Those polled who earned between $50 000-$99 999 per year, 50% voted for Trump while 46% voted for Clinton.
- Those earning US$100 000-$199 999 per year also backed Trump over Clinton – 48% for Trump versus 47% for Clinton.
And when we reach the $250 000 + per year income tier Trump (48%) begins to pull away from Clinton (46%).
These findings should not come as a surprise. It is well established that Republican Party voters are wealthier relative to their Democratic Party counterparts, and this is borne out in the data.
Edison’s polling shows that Donald Trump’s win was not a victory for the ‘little guy‘, unless the ‘little guy’ is a strange way of saying ‘individuals who earn over US$50 000 per year’ (for context, the annual earnings for an American working full-time on the Federal minimum-wage of US$7.25 an hour is US$15 080).
But despite the evidence there remains an almost universal narrative which portrays Trumps win as a ‘working class revolt against X’. But this interpretation raises more questions than answers.
Hillary Clinton’s failure to connect with economically vulnerable working class people is cited by both conservative and progressive pundits as being instrumental in her failure to secure the Presidency. But Clinton not only received more votes from lower income working class voters than Trump, she did overwhelmingly so.
If what we’ve just seen was really a ‘working class revolt against X’, why did working class people earning less than US$30 000 per year still vote for Clinton by a margin of over 10%?
One explanation for this margin is strong Democratic Party support among black households (the median income for black households is US$35 000 as opposed to US$67 000 for white households). But if black and white households ultimately share common class interests, why are the voting patterns among these racial groups so distinct?
Black communities have experienced a slightly worse decline than their white counterparts in terms of total manufacturing jobs so the white community isn’t being particularly impacted by the overall decline in US manufacturing.
White communities also fared much better throughout the Global Financial Crisis (GFC); between 2007-2010 median household income for blacks dropped by 10.1 % compared to 5.4% for whites. At the height of the GFC unemployment in the black community peaked at just under 17%. For the white community it peaked at just over 9%. And the post-GFC recovery has all but failed to materialize in black communities where unemployment remains double that of the white community.
So despite bearing the brunt of America’s manufacturing decline and the GFC, 88% of black voters still backed Clinton. If Trumps win really represents a ‘working class revolt against X’, wouldn’t this have been reflected in the black vote? Or is class consciousness only the preserve of white voters?
The obvious answer to this is that the ‘working class revolt against X’ narrative itself is simply too reductive to offer a meaningful explanation of non-white voting patterns in this election. But more worryingly, the narrative is beset by a white/Euro-centric view of what constitutes ‘working class’ in America today. It also implies that non-white Americans effectively just voted against their class interests.
The ‘working class revolt against X’ narrative is at best implausible, and at worst paternalistic and racist. Its time to drop it and pay attention to the evidence. And the evidence indicates that the distinct racial dimension to this election warrants deeper consideration by both conservatives and progressives.