by Lance Selfa
Putting this in historical context, Trump’s shift of union household voters is actually less dramatic than the swing from 1976 to 1980 for Reagan and even less so than the fourteen-point desertion of union household voters from Carter in 1980, half of which went to independent John Anderson rather than Reagan in 1980, when the union-householders composed 26 percent of all voters.9 In other words, Trump attracted both a smaller proportion and number of these voters than Reagan or Anderson. These same voters have swung between Democrats, Republicans, and high-profile third party candidates such as Anderson; Perot, who got 21 percent of union household voters in 1992; and Nader, who got 3 percent in 2000, for some time.10 The meaning of the 2016 shift was more sinister to be sure, but it was also long in the making as the Democrats moved to the right.
This is not to say that the swing of union household or white, working-class voters away from the Democrats doesn’t reflect the conservative social views, racism, and in the case of Clinton, sexism of many white, working- and middle-class people as well as their anger at their deteriorating situation. Nor is Trump the same as Reagan. Clearly Trump won almost 10 million union household votes, compared to almost 12 million for Clinton. We might assume that the broader nonunion, white, working-class electorate is even more conservative-leaning than union members, but since the level of education turns out to be a dubious measure of working-class status, we have no way of measuring just how many white, working-class people actually shifted from the Democrats, either in 2016 or at some point in the past. We only know that the numbers are significant, but that many are not as new to voting Republican as is often thought. This, of course, is not something to take comfort in, but it is an indication of the results of the Democratic Party’s choice to emphasize higher-income people who are more likely to vote that began under Bill Clinton and the Democratic Leadership Council.11
Nevertheless, it seems clear that a significant number of white, working-class people who had voted for Obama in 2008 or 2012 voted for Trump—even if more just didn’t vote. To get a closer look at how this might have worked across the Rust Belt, where the Democrats lost both the vote and the Electoral College, we will look more closely at Ohio and at heavily white and blue-collar counties that moved from Obama in 2012 to Trump in 2016.
In this Rust Belt state, the union household vote was a dramatic 54 percent for Trump. This was a huge change from 2012, when only 37 percent went for Romney and 60 percent for Obama. The Republicans lost Ohio in 2012, but won it in 2016 by 446,841 votes. This would seem to justify the media story. The total two-party vote, however, declined by 253,859 votes between 2012 and 2016, and the union household vote by 74,366. Trump’s gain of 151,054 union household votes over Romney, and 166,867 more than Clinton, is obviously significant, but less so than the huge drop in the Democratic vote from 2012 to 2016, and not enough to explain Trump’s majority. The Ohio Latino vote, alone, slumped from 10 percent of the total vote to 3 percent in 2016 according to the exit polls. In the actual results, Clinton got 433,547 fewer votes than Obama, while Trump got 179,598 more than Romney, not enough to absorb all the missing Democrats.12 The biggest story here was the drop in Democratic voters.
Thus, the biggest shift was not to Trump, despite the large percentage of union household votes for him, but away from Clinton and the Democrats. The third-party vote increased by two-and-a-half times to 261,318, but not in a way that would have harmed Clinton, as the Libertarian took the lion’s share. The Greens at about forty-five thousand were not a “spoiler” and couldn’t have a made a difference. Clinton’s overall loss of votes was almost two-and-a-half times larger than Trump’s gain.13 Trump won Ohio because the total Democratic vote had declined more than the drop in the total two-party vote, and significantly more than the Republican increase. Furthermore, the 2016 Ohio electorate saw an increase in the proportion of higher-income voters. The Ohio electorate saw those earning $50,000 or more go from 59 percent in 2012 to 63 percent in 2016, while those earning $100,000 more increased from 28 percent to 30 percent. Trump got 57 percent of voters in the $100,000 plus income level, compared to 39 percent for Clinton.14 This certainly smells like a petty-bourgeois/middle-class movement.
To examine the Ohio vote a little more deeply, we will look at four of Ohio’s Rust Belt counties along Lake Erie, stretching from Cleveland to Toledo, both of which, like most cities, went Democratic in both 2012 and 2016. These four counties, Lorain, Erie, Sandusky, and Ottawa, all went for Obama in 2012 by a total of 124,330 to 98,564. In 2016, this vote was somewhat reversed, with Trump getting 113,081 votes to Clinton’s 98,789.15 Again, the two-party vote dropped by 11,024, while the loss of 25,541 from the Democratic column was larger than the shift of 14,517 to Trump and larger than his margin of victory in these four counties. Presumably, the story is similar in other blue-collar counties, such as those in northeast Ohio, home to Youngstown and the legendary GM Lordstown plants. What seems clear about these deindustrialized blue-collar counties is that disillusioned Democrats and demoralized labor leaders forced to “sell” Clinton as an establishment neoliberal could not prevent either the drop in Democratic voters or the shift to Trump, even though the latter was relatively small. A look at Lorain County, the largest of the four, will tell us more.
Lorain County’s population of just over 300,000 is 80 percent white. It includes the small, upper-income Cleveland suburb of Amherst, the small, liberal-college community of Oberlin, a number of small farming communities and townships, as well as larger working-class Lorain City and Elyria City. The county’s workforce is still 42 percent blue-collar. While it still has many employed industrial workers, the local Ford plant has closed, and employment at its biggest steel mill went from 15,000 jobs to fewer than 4,000 in 2014. This is, of course, the county where labor elected three independent candidates on the “Lorain Independent Labor Party” slate in 2014, due to disgust with the local Democrats.16 In 2012 it chose Obama 78,112 to Romney’s 58,092, a majority of over 20,000 votes. In the 2016 primaries, Bernie Sanders got almost as many votes in the Democratic primary (16,587) as Trump did in the Republican primary (16,776).17
In 2016 the county went narrowly for Trump—65,346 votes to Clinton’s 64,958, a margin of just 388 votes. Voter turnout was down by 5,900 voters, but the Democratic vote dropped by over 13,000 votes from that in 2012. The fact that Lorain County is 80 percent white means a lot of white people voted for Clinton, but a significant number just stopped voting. Trump won not only the upper-income Cleveland suburb of Amherst, but all the more rural small towns and townships in the county, with the sole exception of the tiny college community of Oberlin. Trump’s slight majority did not come from the two largest and most “proletarian” towns in the county: Lorain City and Elyria City, the two places where the Independent Labor candidates won in 2014. Lorain City is just over two-thirds white, while Elyria City is closer to 78 percent white. Thus, both have substantial white majorities, but also a significant African American and/or Latino population.18
Table II
Number of Votes per Candidate in the 2012 and 2016 General Elections: Lorain County, Lorain City, Elyria City
Sources: Lorain County Board of Elections, 2016, 2102, “General Election”; New York Times, 2012, “President 100% reporting,” http://elections.nytimes.com/2012/results/states/ohio.
Both towns went for Obama by large margins: three-to-one in Lorain City and two-to-one in Elyria City in 2012, and for Clinton by smaller two-to-one and five-to-four margins respectively in 2016. As the figures in table II show, the turnout in 2016 was down by over 6,000 votes in these two towns, while the increase in the Republican vote for Trump of 2,094 votes was almost four times smaller than the drop in the Democratic vote of 8,209. In other words, the drop in the Democratic votes surpassed that of the decline in voter turnout and the increase in the Republican vote in these two heavily blue-collar towns. This is consistent with the larger statewide drop in the Democratic vote of 433,547 compared to the drop in the total two-party t
urnout in Ohio of over 250,000.
One can hardly avoid the conclusion that the Democrats lost the 2016 presidential election, and those for state office as well, in these and other similar Rust Belt counties due to the fact that former Democratic voters in Rust Belt states with relatively large white, blue-collar populations failed to turnout for Clinton and other Democrats, but did not necessarily vote for Trump. More failed to vote at all. While there was a swing among white, blue-collar and union household voters to Trump, it was significantly smaller than the overall drop in Democratic voters. Below we will see why.
While recent voter-suppression laws demanding state-issued photo IDs in some seventeen states, along with the racial cleansing of voter rolls in many states, have undoubtedly limited voting for Blacks, Latinos, and low-income whites, most nonvoters don’t vote because they don’t see anything compelling to vote for. At the same time, working-class voter participation has remained low in part because the political parties have reduced the direct door-to-door human contact with lower-income voters in favor of purchased forms of campaigning from TV ads to the new digitalized methods of targeting likely voters.19 Vast amounts of personal data are accumulated by firms specializing in this, turned into voter-targeting algorithms, and sold. Among the firms specializing in voter targeting are Artistole, Xaxis, Voter Contact Services, and DSPolitical, the latter of which specifically serves Democratic campaigns. According to John Aristotle Phillips, the CEO of Aristotle, they can provide customers with “up to 500 different data points on each individual.”20
The parties or campaigns that purchase this service, in turn, use it to spread targeted messages to specific groups or even individual voters mostly via the internet through various platforms, including Facebook, which apparently made a bundle off the 2016 election. Digitalized politics cost more and more money. As one recent academic study of voter mobilization said of these new techniques, “campaign consultants have a business interest in deploying these kinds of tactics and the fact that no one knows for sure whether such tactics generate votes allows sub-optimal campaign tactics to persist.” Spending on digital political ads rose from $22 million in 2008 to $158 million in 2012 and is expected to hit $1 billion for the 2016 election and over $3 billion by the 2020 elections. No doubt these amounts will continue to soar as digital political ads are increasingly available for elections way down the ballot to the local level, according to the Democratic digital outfit DSPolitical.21
Aside from the soaring costs this invasive digital targeting adds to US elections and the further erosion of our privacy, it further removes political campaigning from any direct human contact. As reporters for the Guardian put it, “campaigns of the future will depend as much on being able to track people across screens and apps as knocking on doors or sending out flyers.”22 It’s not that no doors are knocked on or phone calls made, but that it is the algorithm that decides the limited number of actual voters to be visited or called to turn out the vote. In practice, this has meant identifying those most likely to vote; that is, the better-off part of the population. The Get-Out-The-Vote campaign has become the Get-Out-The-Well-To-Do-Vote canvass. Most of the persuasion side of campaigning by these outfits is done online through ads on Facebook, YouTube, your smart phone, etc. The knowledge about local voters once in the head of the precinct captain is no longer needed, as far more data are fed through an algorithm that has your number. More importantly, the shaping of the political process, already an auction, is being even further outsourced to the profitmaking “expert” firms that provide this service. The result is that the proportion of relatively higher-income people in the electorate has grown, as the votes for both Clinton and Trump show.
In other words, despite all the vast amounts of money raised and deployed, all the digital and “expert” sophistication available to this “party of the people,” and Clinton’s allegedly massive “ground war” force in the “battleground” states, the Democratic Party as a whole cannot or no longer tries to mobilize enough of those among its traditional core constituencies—Blacks and Latinos, as well as white workers and union members—to win national and even state offices in these key states and possibly elsewhere, despite the demographic trends in its favor. While race was certainly a factor in Trump’s appeal, the fact is, in much of the country, the Democrats at every level could not get the turnout of African Americans or Latinos they needed to balance out their losses among white voters of all classes. To be sure, Clinton won a majority of the popular vote nationally, perhaps, as John Nichols gloated in the Nation, an “unprecedented” majority that might run as high as two million by the time the official state votes are counted. [Editor’s note: The final tally, according to certified results recorded for the Electoral College at the National Archives, showed a difference of 2.839 million votes in Clinton’s favor.] The problem is that 1.5 million of that majority can be accounted for from Clinton’s vote in New York City alone; that is, by just her majority over Trump, not her total vote.23 The majorities in the coastal states of California and New York by themselves account for more than her net final majority count of 2.8 million. The rest of the country continues to see its Democratic vote stagnate or decline. Aside even from the digital demobilization, the reason is not hard to find.
The Democrats are and have been for decades the party of the (neoliberal) status quo, when millions of all races have seen their living standards shrink and future prospects disappear and, as a result, have come to despise the status quo. And, as not only the Clintons’ social position, but the many millionaire Democrats in Congress (average wealth of a Democratic Representative is $5.7 million), and their business buddies demonstrate for all to see, they are part of the nation’s elite. The decline in manufacturing jobs, the shrinking of union representation, the creation of more and more lousy jobs, the withdrawal of aid to the cities, etc. have created, not just “angry white men” who voted for Trump, but angry white, Black, Latino, and Asian men and women who, for good and sound reasons, no longer see the Democrats as their defenders. Many in the ranks of this legion have voted with their feet, and it wasn’t to the polls. According to one estimate, over 100 million eligible voters didn’t cast a ballot in the 2016 presidential election. In 2014, the last off-year congressional election, nonvoters numbered almost 128 million adult citizens—a majority of eligible voters.24 And, the vast majority of these were middle-to-lower-income, working-class people.
Strong evidence that the Democrats can no longer motivate or mobilize the majority in much of the country is found in the fact that the millions of nonvoters are on average and in their majority politically to the left of those who do vote on key economic issues. As one study put it, “Nonvoters tend to support increasing government services and spending, guaranteeing jobs, and reducing inequality” more than voters and do so by about 17 percentage points. This includes whites as well as Black and Latino nonvoters. Furthermore, these nonvoters are as likely to support legal abortion and gay marriage as those who vote. Specifically, they are to the left of voters on economic issues in that a consistently higher percentage of them think the government should “make union organizing easier,” increase funds to the poor and for schools, “guarantee jobs,” and “provide health insurance.”25 In short, the Democrats cannot mobilize the forces needed to defeat the right, including Blacks and Latinos, in part because they cannot implement any policies capable of addressing the plight of the majority that might attract these left-leaning nonvoters.
In fact, nationally, the Democrats have been losing elections at just about every level since 2009. In that year, during the 111th Congress, the Democrats had 257 members in the House of Representative. By 2015, in the 114th Congress, that was down to 188 Democrats, the lowest number since the 80th Congress in 1947–49, over which time voter participation rates fell from 48 percent to 42 percent in off-year congressional elections. In 2016 the Democrats won back just six seats in the House.26 Between 2009 and 2015, the Democrats lost 203 seats in State Senates a
nd 716 in State Houses or Assemblies. An indication of what was to come in Ohio and Pennsylvania in 2016 could be seen in the loss of 21 Democratic seats in the state legislatures of each of these states between 2009 and 2015.27
This time, however, the falling Democratic vote meant the victory not of a run-of-the-mill conservative or even a Tea-Bagger, but of a racist demagogue bent on doing serious damage. And he already is. There will be resistance. Rather, there is increased resistance. And this will offer new possibilities for organizing, even in a more hostile atmosphere. At the same time, many, including not a few on the socialist left, will run for cover in the Democratic Party’s “Big Tent,” arguing now is not the time to take on the Democrats, that the great task is to elect a Democratic Congress, any Democratic Congress, in 2018 to rein in Trump just as the Republicans blocked Obama after 2010, and so on. It will not even be an argument for reforming the party, just stopping the Trump rampage. It will be a tempting and effective argument, particularly if Trump has, indeed, pulled out all the stops or hasn’t already appointed someone to the Supreme Court. But such a political direction will only reinforce the Democrats’ neoliberalism, digital-dependency, and failed strategies.
In addition to all the structures and money behind the Democrats at all levels and the reality of party discipline in most legislative bodies, there stands the mass of elected Democratic Party office holders. In what was surely a real-life test of their politics, Bernie Sanders’s campaign for the presidency, they stuck with the neoliberal mainstream in their vast majority. While Sanders got nearly 13 million votes from rank and file Democrats, his support among Democrats in office was marginal, even among self-styled progressives: 10 out of 232 Dems in both houses of Congress endorsed Sanders; five of the 75 of his fellow members of the Congressional Progressive Caucus; 91 out of 3,170 Democratic state legislators; three of the 48 Democrats in New York’s city council; only two of that council’s 19 Progressive Caucus members; and so on.28