Thomas B. Hofeller was known in political circles as a master of drawing political districts and using demographic data to further the aims of Republicans. But before his death a year ago, the public record of his work was limited to gerrymandered political maps in North Carolina and a handful of other states, and a sheaf of expert-witness depositions in lawsuits.
That only skims the surface of his political work over the last decade.
Tens of thousands of maps and documents, kept on computer backups that were uncovered after Mr. Hofeller’s death in August 2018, show his work on a broad range of projects nationwide, from drawing political maps that maximized Republican representation during the 2011 redistricting cycle to a host of smaller projects to advance the party’s fortunes.
He kept at it to the end, promising to “bedevil the Democrats” with Republican-friendly voting districts shortly after receiving a likely fatal medical diagnosis in 2016.
The few documents made public from his work have loomed large in two major lawsuits. A legal analysis found in his files appeared verbatim in a draft proposal to add a citizenship question to the 2020 census, pointing to his personal involvement in a plan that most experts have concluded was intended to skew the census process toward Republicans. And a North Carolina state court cited data from the backups last week in striking down partisan gerrymanders in State House and Senate maps that Mr. Hofeller had drawn.
But files from the Hofeller backups recently made available to The New York Times offer a much broader view of Mr. Hofeller’s partisan work. The New Yorker reported on the contents of the Hofeller files in an article published on Friday.
Among the highlights from the files obtained by The Times:
Mr. Hofeller played a crucial role in drawing partisan maps nationwide in 2011.
After the Republican Party’s sweep of 2010 elections, documents from his computer files show, Mr. Hofeller and a business partner, Dale L. Oldham, helped draft political maps in an array of states, from G.O.P. strongholds like Texas and Alabama to states trending Republican like West Virginia to swing states like Florida.
The documents indicate Mr. Hofeller advised Republicans on the boundaries of new congressional maps in Florida in 2011, even though voters had mandated nonpartisan maps for all political offices in an amendment to the State Constitution the previous year. State courts later nullified those congressional maps as a violation of the constitutional requirement.
Mr. Hofeller’s files are virtually devoid of any paper trail documenting his agenda, but his 2011 efforts to draft congressional maps in Texas offered a rare exception.
It was a Saturday in June, and very late on a Saturday at that. But Mr. Hofeller was still at his computer, mining mounds of Texas demographic data.
“I can give you about .8 percent increase in SSVR within Austin only,” Mr. Hofeller wrote in an email, using an abbreviation that denotes residents with Spanish surnames. With a few keystrokes, Mr. Hofeller apparently was shuttling 30,000 mostly Hispanic residents from a Republican district west of Austin into a Democratic one.
“Every .8 helps at this point,” his fellow strategist, Mr. Oldham, replied.
The map the two men produced at about 7 a.m. the next morning was ungainly, Mr. Oldham wrote. But it did the job, sending fingers from three neighboring Republican districts deep into Austin and giving the party a lock on all but one of the House seats in heavily Democratic Travis County.
Mr. Hofeller was the architect of Republican political dominance in the swing state of North Carolina.
More than any other state, North Carolina was shaped politically by Mr. Hofeller’s talents. A trove of files shows his involvement in virtually every aspect of the State Legislature’s effort to promote and defend Republican control.
In 2014, Mr. Hofeller compiled a database of more than 318,000 college students who had registered to vote in the state, broken down by a host of factors including age, sex, race and Hispanic ethnicity, party and address. Mr. Hofeller also matched lists of registered college students with state driver’s license data; a license was one of the acceptable proofs of eligibility under North Carolina’s voter ID law.
Another project singled out registered voters who lacked a driver’s license or state ID card, and then tagged their addresses to determine how far they lived from an office that issues driver’s licenses.
Both databases later figured in a defense of the state’s voter ID law against a lawsuit charging that the law discriminated against certain classes of people who predominantly voted Democratic. But the information also had great political value. For example, Mr. Hofeller proposed analyzing the list of voters without IDs to determine whether any were noncitizens or lived outside the state.
The database of college students, who largely support Democratic candidates, could have proven valuable in 2016 and 2017, when parts of the state’s political maps were found to be racial gerrymanders and had to be redrawn.
The remaps were notable for how several college campuses were singled out, and sometimes split in half, for apparent political gain.
Race was another constant of Mr. Hofeller’s work in the state, even when his Republican employers insisted it was not. A 2015 chart broke down the state’s Superior Court districts by race and political leanings; the court’s 95 judges are elected in partisan races.
The three-judge panel that struck down North Carolina’s State House and State Senate maps this month cited Mr. Hofeller’s files in concluding that he had used racial statistics to shape his maps despite public claims to the contrary. This map of metropolitan Greensboro taken from the files shows the racial composition of precincts and State House districts.
The same court cited the Hofeller files in another deception: Metadata on maps of state legislative districts showed they were almost completely drawn months before Republican legislative leaders publicly adopted the standards for drawing them.
Mr. Hofeller pursued other efforts to shore up Republican competitiveness.
In 2011, he took on a request by an ostensibly nonpartisan organization, Fair Districts Mass, to draw a redistricting proposal that would have subsumed most of Boston into a single congressional district, weakening Democratic representation in the region. The proposal failed to gain traction.
Virtually all political mapmaking relies on census counts of the entire population to set the size of districts. But as this summer’s trial over the 2020 census showed, Mr. Hofeller had extensively studied how using a different count — only American citizens of voting age, shortened to CVAP — would boost Republican fortunes in Texas.
The files reveal that Mr. Hofeller collected CVAP data for other states, led by Arizona and North Carolina. The data is needed for some mapmaking under the Voting Rights Act, but Mr. Hofeller may have had a political interest as well. “We never look at CVAP data in North Carolina,” said Allison Riggs, the senior voting rights attorney at the Southern Coalition for Social Justice, but the Hofeller files contain detailed county-by-county data for the state, collected in both 2011 and 2015.
The data can provide demographic insights for mapmakers, and some Republican states are widely expected to attempt to base their next political maps on CVAP data in 2021.
“A savvy map drawer (and he was) is looking at changing demographics and voting patterns throughout the decade in any state he or she might work on in the next cycle,” Ms. Riggs wrote in an email, “because we’re often playing around with maps based on what we think the 2020 census data will show.”
Mr. Hofeller was no fan of independent, nonpartisan redistricting commissions, which he viewed as tools of Democrats. But he made an exception in 2009, mulling over ideas to lend Republican support to a ballot initiative to set up such a commission in overwhelmingly Democratic California.
“The Democrats will find themselves in the same position as the Republicans in Florida,” he wrote, referring to a constitutional amendment there that banned partisan maps altogether. In an email, he endorsed a proposal to lend the backing of the Republican National Committee chairman at the time, Michael Steele, to a $ 10 million fund-raising campaign to support the ballot initiative. It is not clear whether those plans came to pass, but the ballot initiative was eventually approved.
Mr. Hofeller was a partisan to the end but worried about his party’s future.
“I am a little shell-shocked from receiving this news,” Mr. Hofeller wrote a friend in May 2016, a day after receiving a grim cancer diagnosis. But “I still have time to bedevil the Democrats with more redistricting plans before I exit. Look my name up on the internet and you can follow the damage.”
Exactly three months later, after Donald J. Trump had been crowned the Republican presidential nominee, Mr. Hofeller had turned deeply pessimistic about the party’s fortunes in November.
Still, he figured artfully drawn political districts could outweigh political tides.
“Do not worry about us in North Carolina in terms of redistricting,” he wrote. “Even in the coming political bloodbath we should still maintain majority control of the General Assembly.”
Mr. Hofeller was right. North Carolina Republicans kept their state legislative supermajorities in the 2016 general election. And even in the 2018 midterms — when Democrats took 51 percent of the statewide vote — Republicans won 29 of the State Senate’s 50 seats, and 65 of the 120 seats in the State House.