Counting all the voters

October 15, 2018

In his column for Inside Higher Ed, Scott McLemee reviews two books about elections — one about why people vote the way they do, and the other about those who don’t vote.

ON THE way to my caffeine fix most mornings, I go past a building with a large digital clock out front counting down the number of days, hours, minutes and seconds remaining until the midterm elections. It has been there for a couple of months now and is too big to miss, though it’s not exactly necessary considering the location is downtown Washington. People working nearby are no more likely to forget the time left until an election than the residents of a coastal town are to overlook when a hurricane is expected to make landfall.

Making more detailed predictions about the impact seems like a bad idea in either case. I’ve just finished reading a couple of new scholarly books on the American electorate and its discontents. Neither is focused on the 2016 presidential race and its implications in particular, though both make their long-term survey of data with recent developments in mind. You would be hard-pressed to extract a short-term forecast from either volume — a good thing, on the whole. Their interest should survive any incoming wave.

Voters wait in line to cast their ballots in Florida
Voters wait in line to cast their ballots in Florida

Provocative just in its title alone, Sharon E. Jarvis and Soo-Hye Han’s Votes That Count and Voters Who Don’t: How Journalists Sideline Electoral Participation (Without Even Knowing It) — from Penn State University Press — is more insightful than the usual complaints about American politics devolving into a horse race in an echo chamber. That it has, of course, is not in question. The authors (both associate professors of communication studies, at the University of Texas at Austin and Kansas State University, respectively) accept an analysis of the 1968 presidential race as a turning point: campaign strategists “began limiting the press’s direct contact to their candidates” and so “journalists began to report more heavily on the artifice of campaigning.” I remember thinking that this tendency had reached the point of unsustainable absurdity in 1988 — a campaign season in which much of the coverage was polling data, including polls about whether too much attention was going to polling data — though things have gotten worse somehow.

The long-term trend, the authors say, has been for “the reportorial gaze” to shift “away from the relationship between candidates, voters and policy positions over to the power of strategists and political operatives in American life.” The reference to “gaze” may seem to imply a driving role for television coverage, but the authors focus most of their efforts on content analysis of the newspaper reporting of presidential campaigns for the past 70 years.

In particular, they find a pronounced tendency for “vote” and “voter” — nouns specific to the electorate and its activity — to shift in weight and connotation in presidential campaign coverage from 1948 to 2016. A graph showing the frequency with which each word appeared (whether in singular or plural form) relative to the total number of words in campaign news coverage. Between 1948 and 1968, the term “vote” appeared up to six times more often than “voter,” and in context it typically “signaled something that was pursued, that mattered, and that added welcome citizen input to the news narrative.” After 1968, the long-term trend was for “vote” to appear less frequently and “voter” to become the predominant term. If my eyes and calculator are to be trusted, it was used twice as often as “vote” in 2016.

And substitution did not entail equivalence. Campaign reporting on “the voters” tended to be about their mood or their level of engagement. How could a campaign mobilize voters who were “apathetic,” “cynical,” “troubled”? Was their participation itself in question — were they “intimidated,” “misinformed,” “puzzled”?

Not all coverage painted the electorate in such somber hues, of course. But in general, Jarvis and Han find that references to “the vote” around midcentury tended to imply something “honored and unsullied,” while more recent consideration of “the voters” has skewed toward “adverse situations and complex emotions.”

The semantic drift here reflects the long-term reorientation of electoral coverage. With “the reportorial gaze” focused on the candidates’ output of spin and spectacle, there’s almost certainly a tendency to absorb some of the campaigns’ perspective. “The vote” as civic ritual is irrelevant, while “the voters” are raw material: an only somewhat reliable means to the end of getting a sufficient number of ballots marked a certain way.


AFTERWARD, SOMEONE has to puzzle through the enigmas of who voted — and why or why not. To extrapolate anything from the findings is pretty much the last thing you can do with Bernard L. Fraga’s investigation of The Turnout Gap: Race, Ethnicity, and Political Inequality in a Diversifying America (Cambridge University Press). The author, an assistant professor of political science at Indiana University, combs the demographic and voting data — as well as “decades of survey data along with new voter file-based analyses” — to try to account for the continuing disparity between the voter turnout of minority groups and that of what the census form calls “non-Hispanic whites.” The latter, Fraga notes, “were 73.6 percent of the eligible to vote population” in 2010 “but 77.5 percent of the population that actually voted.”

Now, chances are you can come up with a number of plausible-seeming hypotheses for why other groups might vote at lower rates than whites do — including differences in socioeconomic background or rates of voter eligibility, or deliberate legal maneuvers to suppress minority voting. That these factors are not mutually exclusive possibilities is obvious. But Fraga’s crunching of the numbers suggests that none of them is sufficient to account for the overall disparity in turnout, or differences in relative turnout over time or among the groups making up the “non-Hispanic white” electorate. Relatively high minority turnout “can and does occur even in the face of tremendous institutional barriers,” such as legal efforts to make voting more difficult.

The reader can get lost among the bales of statistical data and all the confirmed noncausalities. The largest point here, I take it, is that any assumption that a long-term increase in demographic diversity will more or less automatically generate significant diversity among who goes to the polls — or even that short-term change will involve noticeable regularities of the kind politicians and their electoral machines can bet on — would be misguided. The one factor he identifies as playing a clear role in mobilizing members of a group to vote is their sense of “having the potential to drive election outcomes.” That is putting it positively. The obverse formulation would be that “the root cause of low minority turnout is the fact that, in most elections and in most places, African Americans, Latinos and Asian Americans are perceived to be a less electorally relevant force.”

But again, this is a way of framing issues likely to be most congenial to campaign professionals. Breaking down the electorate into census-based racial/ethnic categories makes a kind of sense when creating a spreadsheet for calculating the most cost-effective way to move clusters of voters around. There has to be more to a democratic society than that.

First published at Inside Higher Ed.

Further Reading

From the archives