Voter suppression's last stand: North Carolina's new Jim Crow counties
A high-level look at the wide range of changes North Carolina's 100 counties made in the average number of early voting days, per site, they will offer in 2016, ranging from massive cuts to dramatic increases. But at this altitude you can't see the faces involved - which demographic groups are the beneficiaries of those changes, and which are their victims? This Insightus investigation provides the answers. And the news is not good.
Our review of the state's early voting plan for this year finds that many boards did just the opposite. Still, a defiant band of renegades – the state’s New Jim Crow counties – did answer that call with cuts disproportionately falling on minority voters and promising election day chaos. But voting rights advocates are fighting back.
Related Content (blog): NC’s Early Voting 2016 Dashboard
The Supreme Court's 2013 Shelby County v. Holder decision overturned a key protection of the Voting Rights Act of 1965 (VRA), instantly transforming North Carolina into the epicenter of the nationwide battle over minority voting rights.
Within weeks of that decision (which freed the state from VRA's requirement for federal oversight of changes to its election practices) North Carolina's Republican-controlled General Assembly passed, and Gov. Pat McCrory signed into law, the state's "monster voter suppression bill," HB589. The law slashed early voting days, imposed a cumbersome voter ID requirement, and ended voter registration during the early voting period, among many other restrictions.
The good news for North Carolina voting rights in 2016 was a July decision by the U.S. 4th Circuit Court of Appeals overturning HB589, finding that the law unconstitutionally "targeted African Americans with almost surgical precision."
But the bad news is that Shelby still grants North Carolina's political ne'er-do-wells a free hand to experiment with ever more creative means to suppress minority voters, leaving early voting – and the citizens who rely on it – in their crosshairs.
In North Carolina, early voting plans are redrawn every two years on a county-by-county basis, by county boards of elections that are, by law, dominated 2–to–1 by the current governor's party (today, the GOP). Our review of all one hundred counties' recently finalized early voting plans for 2016, detailed here, reveals that this year the war on minority ballot access has retreated to the county level, where a band of rebel counties have pushed back against what would otherwise be a broadly improved statewide voting climate.
What We Looked For In NC's Early Voting Plans
State law requires each North Carolina county to offer at least one day of early voting in at least one location, up to a maximum of 17 days across numerous sites. Within those limits, county boards of elections have broad discretion to craft their own unique plans. Key factors under the boards' control, which combine to make early voting easier or more difficult, include:
Number of polling sites per county: More sites mean voters travel shorter distances to vote, putting younger and poorer voters, whose transportation options are often restricted, on a more nearly equal footing with older and more affluent voters with unlimited access to cars. Shorter distances in turn mean less time spent travelling to and from the polls - a key concern for the working poor, who must squeeze voting into their typically inflexible work schedules.
Days per site: The more days from which voters may choose to cast their ballots, the more likely they are to do so. That benefits election turnout generally, and particularly among the working poor.
Total site-hours: This is the sum of all open hours across all of a county's early voting sites. More hours yield shorter waiting lines, plus increased scheduling flexibility for wage-earners.
Off-hours: The fraction of each site's hours that fall outside of the conventional work week (8 AM to 5 PM weekdays). Off-hours are particularly important to the working poor.
Sunday hours: African American churches have a strong tradition of busing their congregations to the polls following Sunday services, making Sunday voting an important factor in minority ballot access.
These factors are unique to early voting, which differs from election day voting in that the latter is confined to a single site per precinct, on a single Thursday, during hours that are fixed by law both across the state and from each election to the next.
Nonetheless, a voter-friendly early voting plan benefits the county's election day voters as well, by keeping lines short on November 8th (thus encouraging turnout). Conversely, a bad early voting plan threatens election day pandemonium with long lines and jammed parking lots at precinct polling places, the potential for administrative snafus like ballot shortages, and election results that may be delayed or – worse still – even contested.
Our Findings: Devils in the Details
(See Supplemental Material 1 for data sources used in this analysis)
At first glance, the outlook for 2016 early voting in North Carolina is improved over 2012:
- The number of voting sites statewide is up 22%, to 445 this year from just 365 in 2012
- Total site-hours across the state are up 17%, to more than 42,000 from only about 36,000 in 2012
- That increase in total site-hours more than adequately tracked the state's population growth, which was up by 4.1% from 2011 to 2015
- Early voting hours outside of the conventional work day are up 27%, to nearly 11,000 hours statewide from 8,600 in 2012
- Sunday hours are up 14%, to 781 hours from 683 for the last presidential election
- Average open days per site declined by 10%, to 11.6 days from 12.9 in 2012
Yet despite these improved aggregate statistics, our analysis of the state's early voting plan at the level of its 100 individual counties – where the rubber meets the road – uncovered a small band of renegade counties that imposed multiple voter-suppressive cuts to key early voting features (Figure 1), which had the net effect of making early voting more difficult, particularly for African American voters.
Figure 1: Numbers of NC's 100 counties making changes to key early voting features for 2016. Red: Cuts to early voting for 2016. Yellow: No change from 2012. Green: Improved for 2016.
Notes: (1) Off-hours are voting hours outside of the conventional work week (8 AM to 5 PM weekdays), expressed as a fraction of total site-hours. (2) Per 25,000 population. (3) See text for discussion.
(click to open in new tab)
Supplemental Material 1:
Data Sources Used In This Analysis
Seventy-eight North Carolina counties made changes to their open days for 2016, 17 of which had negative impacts on voters. Those changes were far from colorblind.
Changes to a county's open days can be tricky to judge. Each of a county's early voting sites can be open for as many as seventeen days, or as few as one, and total open days can differ between those sites. On top of that, changes to a county's open-days-per-site average also hinge on changes in the number of sites in a county. To be as even–handed as possible in the face of these complexities, we considered all possible combinations of those factors (see Supplemental Material 2) to classify each county's net change in open days into one of four possible categories: "No Change," "Winners," "Tradeoffs," and "Losers" (Figure 2).
Figure 2: Comparison of the racial balance of NC counties making any of four possible changes to open voting days for 2016. For a complete description of the four classifications used here, see the text and Supplemental Material 2. Counties classified as Winners increased total open days without decreasing site numbers. Tradeoff counties either increased total open days but decreased site numbers, or vice versa. Loser counties decreased total open days while either decreasing site numbers or leaving them unchanged from 2012. Racial balance is expressed as each group's white/black (W/B) ratio (the sum of the group’s non-Hispanic white registered voters as of May 2016 divided by the sum of its black registered voters). For North Carolina as a whole the W/B ratio is 2.7 (dashed red line).
The majority of the county boards either made no changes to their site numbers and open days for 2016 (No Change in Figure 2), or changed these parameters in ways that had a net favorable impact on early voting (Winners). Both groups were whiter than the state as a whole, with 3.2 or 2.8 white voters for each black voter (W/B ratio), respectively.
Another 10 counties imposed tradeoffs of debatable merit on their early voters, either decreasing total open days while increasing the number of sites across the county, or vice versa. As a group, these Tradeoff counties were slightly blacker than the state as a whole (W/B ratio = 2.6).
But by far the largest deviation from the state's racial balance revealed by our analysis was that of the 7 Loser counties - those decreasing their total open days while either holding sites constant or, in the case of Columbus County, cutting both sites and open days. Loser counties were substantially blacker than the state overall, with just 2.1 whites for each black voter (Figure 2). Among them, the biggest loser of all (Columbus) is blacker still, with just 1.8 whites per black voter.
In short, the sacrifices in open days and total sites imposed by a small band of 17 counties (Losers and Tradeoffs in Fig. 2) fell disproportionately on black voters.
The impact of those sacrifices on North Carolina’s African American vote will prove to be substantial this year. In raw numbers, Loser counties are home to 17% of the state's black voters, but only 13% of its white voters. In a battleground state like North Carolina, where the margin between winning and losing candidates' totals is often razor thin (and frequently tipped one way or the other by African American turnout), this disparity is a real concern for friends of democracy.
And among those worrisome losers, Mecklenburg County is the most troubling of all. Leading its fellow losers in both population (1.0 million) and the sheer size of its cut to total open days (-22%) Mecklenburg is the state's largest county. An urban and highly diverse community (49% non-Hispanic white, 32% black, 13% Latino), its elections often hang by a thread. Bush edged out Gore there by just a few hundred votes in 2000, while Kerry topped Bush by only a few thousand in 2004. But Mecklenburg's black voters, when they turn out in strength, are the deciders: in 2008 and 2012 Obama took Mecklenburg County by landslides of 61% and 62%.
If Mecklenburg County decides the state (and perhaps the nation?) in Republicans' favor this year, the party's presidential candidate may well have Mecklenburg's majority-white Board of Elections to thank for its leading role in race-based election re-engineering.
(click to open in new tab)
Supplemental Material 2:
Days/Sites Classification Matrix
Twenty-four counties made voting more difficult for wage-earners generally and the working poor in particular by cutting the proportion of early voting hours that span what we call 'off-hours' (voting hours outside of the conventional work week of 8 AM to 5 PM weekdays). Here too, these changes disproportionately impacted black voters. Counties cutting off-hours (expressed as a fraction of total hours) are blacker than the state as a whole, while those that improved their off-hour performance are whiter (2.2 and 3.0 whites per black voter, respectively; see Figure 3).
Figure 3: Racial balance of the groups of NC counties that increased or decreased off-hours for 2016. Off-hours are defined as the fraction of a county's total hours that fall outside the conventional work week of 8 AM to 5 PM weekdays.
Here too, the raw numbers of African American voters impacted are substantial: 44% of the state's black voters saw their ability to vote outside of work hours restricted, while just 36% of non-Hispanic whites did.
Giant Mecklenburg County was joined as an off-hours cutter by the state’s second and third largest counties: Wake (pop. 1 million) and Guilford (pop. 520,000). Between just these three counties (with their very low combined W/B ratio of 1.9) their off-hours cuts will impact about a third of the state’s black voters (32%) but less than a quarter of its whites (22%).
African American voters won increased Sunday voting hours for 2016. But, paradoxically, fewer blacks will be permitted to exercise the right to vote on Sunday.
Only a relative handful of counties offer Sunday voting opportunities, which have become increasingly important in the life of the black church across the South. In 2012, 21 North Carolina counties offered Sunday voting, increasing to 22 in 2016. Yet that apparent gain came at an important and hidden cost. Three counties abolished Sunday voting for 2016 while 4 others offered Sunday hours for the first time, thus accounting for the net increase in Sunday-voting counties. But the abolishing counties comprise 6% of the state’s black voters, while the initiating counties comprise only 2%. On balance then, a net 4% of the state’s African American voters lost the right to vote on Sunday in 2016.
Twenty-two counties attacked early voting in 2016 by cutting total site-hours, thus insuring longer waiting lines this year and risking decreased turnout. But the added burden fell about equally on black and white voters alike (36% and 35% of their statewide electorates, respectively).
The state's larger counties proved to be particularly enthusiastic hour-cutters, led once again by Mecklenburg (pop. 1 million), and joined by Buncombe (253,000), Guilford (518,000), and New Hanover (220,000), which together are home to about one quarter (24%) of the state’s black voters but just 16% of whites.
While cuts to total hours across all counties were not racially disparate this year, they are still a cause for concern. As Figure 4 illustrates, the state’s two largest counties, Mecklenburg and Wake, have fallen far behind the curve of what may be considered typical site-hour allocations in North Carolina. And while Wake showed improvement this year (increasing its total hours over 2012), Mecklenburg’s inexplicable cut to total hours seems almost intended to flirt with election day chaos.
Figure 4: Total site-hours versus total registered voters for each of North Carolina’s 100 counties in 2016. The black dashed line is the linear regression line for all counties except Wake and Mecklenburg (which are indicated in red). We take this line to indicate the norm for the state’s counties.
Number of Sites
Among all of the metrics by which we judged county early voting plans, sites per county saw the most lopsided positive changes for 2016. And unlike the other metrics reported here, their large benefits and limited costs fell about equally on the state's black and white voters.
Losses were small in number, with only 2 counties’ voters receiving bad news (cuts of 1 site each for Cleveland and Columbus counties). Those cuts affected tiny - and roughly equivalent - fractions of the state’s black and white voters (1.7% and 1.5%, respectively). Still, they were hard losses for these two counties’ low-income (and thus lower-mobility) voters, whose per capita incomes are just 80% and 75% of the state’s, respectively. Cleveland’s cut was particularly painful, leaving the county (which spans 25 by 25 miles) served by only 3 early voting sites.
On the positive side, 50 counties gained additional sites in 2016 - potentially good news for low income and lower-mobility voters. Combined, those counties account for 63% of the state’s white voters and a nearly equivalent 64% of blacks.
But for the moment we must defer final judgment regarding these apparent improvements, while Insightus analysts finish the computer-melting job of calculating how voters' distances from their homes to their nearest early voting locations have been affected. Because in poll-counting, as in real estate, location is everything. Early voting sites move from year to year - sometimes closer to white enclaves, sometimes closer to more diverse neighborhoods. As we previously documented in our analysis of location changes for the last election, in 2014 polling place shifts across North Carolina moved the ballot box a combined 350,000 miles further away from black voters, but less than a tenth that far (just 21,000 miles) from whites. In the upcoming second installment of this series we’ll report our distance-to-poll findings for this year’s early voting map.
North Carolina’s New Jim Crow Counties (and the leaders who urged them on)
Voters of every color who find themselves facing rogue counties’ long waiting lines this election will have one man in particular to thank: the state Republican Party’s executive director, Dallas Woodhouse.
In August, Woodhouse was embarrassed by the revelation of a private email in which he called upon the Republicans controlling the state's county boards of elections to "make party-line changes to early voting" for 2016. Calling for the abolition of Sunday voting (which, as we have seen, three counties did), Woodhouse wrote "many of our folks are angry and are opposed to Sunday voting." Encouraging the strategic placement of polling places away from neighborhoods that are home to demographics less sympathetic to Republican candidates, he asserted that "no group of people are entitled to their own early voting site."
And in a broad call to hamstring early voting in every other way possible, Woodhouse observed that "same–day registration is only available during early voting," made the unsupportable claim that "same–day registration is ripe [sic] with voter fraud," and completed the non sequiter by concluding that "we [Republicans] are under no obligation to offer more opportunities for voter fraud."
A minority of county boards wholeheartedly embraced Woodhouse's call for election tampering, even as the majority of NC's counties rejected it.
Table 1, below, lists the worst of the worst in that rogues' gallery: the 17 counties that each made multiple voter-suppressive changes to the metrics we analyzed for this report, in what many observers take to be concerted efforts to suppress minority voting. Between them these counties account for 28% of the state’s population, including nearly a third (32%) of all black voters but only about a quarter (26%) of non-Hispanic whites. As a group they are distinctly blacker than the state as a whole, with a combined 2.2 white voters per black voter (compared to the state's W/B ratio of 2.7).
Table 1: North Carolina counties imposing two or more different voter-suppressive changes on early voting for 2016.
By any measure, the leader among these rogue counties is the state's largest, Mecklenburg, whose Republican board of elections chair, Charlotte resident Mary Potter Summa, channeled Woodhouse by declaring "I’m not a big fan of early voting. The more sites we have, the more opportunities exist for violations." Mecklenburg's early voting plan is notable for substantially cutting open days, off-hours, and total site-hours (falling well behind the county’s population growth) - guaranteeing voter frustration and advancing the cause of voter suppression.
Postscript: How Voting Rights Advocates Are Fighting Back
Since early August, when the outlines of county boards' early voting plans first began to emerge, the North Carolina chapter of the NAACP (lead plaintiff in NC-NAACP v McCrory) and its members across the state began advocating directly to the State Board of Elections to take corrective actions in keeping with the 4th Circuit's ruling in that case. But those arguments went largely unheeded by the State Board. As a consequence, last Saturday a group known as the "Duke Intervenor plaintiffs" filed an emergency motion in U.S. District Court, asking it to order the State Board to "immediately modify the early vote plans for Mecklenburg, New Hanover, Nash, Guilford, and Forsyth Counties" to comply with NC-NAACP v. McCrory.
Plaintiffs in that emergency motion declared "The challenged plans are blatant attempts to make an end run around [...] this Court's injunction. They seek, at least in part, to accomplish on a county-by-county basis what the Fourth Circuit barred the General Assembly from doing: suppressing African American voting strength by limiting access to early voting. To 'fully correct and eliminate root and branch' the State's racially discriminatory effort to suppress the vote of African Americans, the Court should order these plans be modified."
The emergency motion seeks quick action by the court, in time for the start of early voting in North Carolina on October 20th.
Insightus is proud to have provided pro bono data analytics support to attorneys advising NC-NAACP.
Related Content (blog): NC’s 2016 Voter Turnout (Part 1)