A Communist Cell in South Carolina?

Front page headline from the January 17, 1958 issue of The Columbia Record

"Carolinas Have 147 Known Reds" screamed the headline of the January 17, 1958 issue of the Columbia Record newspaper. Only two days earlier the governor of South Carolina, in his annual address, had made a shocking accusation: there was a communist cell in South Carolina. He even disclosed that the cell was located less than two miles away from the statehouse, at Allen University. 

The alleged cell was not made up of the 147 "known Reds" from the Record headline. The subheading clarified that only "17 Reside Within S. C." Even this number had only weak support. Deep into the article,  the reporter clarified that the membership numbers came from a public report published seven years earlier. They "do not necessarily reflect a perfectly accurate count as of today," an unnamed FBI agent is quoted as saying.

The tone of the article invites to reader to imagine that the membership numbers had grown dramatically since 1951, but the reality was that the Communist Party USA had been in free fall for much of the 1950s. Support for the Communist Party had peaked in the immediate aftermath of World War II, a time when the Soviet Union was celebrated as an important ally. This situation changed quickly. Once the Cold War broke out, Party members found themselves on the frontline of the conflict. Mainstream political groups distanced themselves from communists, and the federal government, empowered by new anti-subversion laws, pursued criminal charges against much of the CP leadership. By 1958, the Party was in disarray, with almost one-hundred members criminally convicted. 

The unnamed FBI agent the journalist spoke to was tightlipped about recent developments, but he had come to the conclusion that there were no serious concerns. The FBI had just completed a thorough investigation of the Allen professors. One professor, Edwin Hoffman, had been considered for inclusion in the FBI's index of people who were a threat to national security. Their conclusion? He "should not be included in the Security Index at this time."

The FBI did find that there were communists in South Carolina. Internal data included that estimate that five communist "registrants" were living in the state in 1956. This was actually a peak: there were no members in 1955 or 1954.

While the number of communists was tiny, the FBI was concerned about a recent reorganization within the CP. Communist leadership had abandoned the south in 1952 as part of a shift to "full-scale underground activity." By the summer of 1957, the Party had begun reemerge, and it reorganized its activities in the south. A newly created Southern Region Committee was "controlling unit" for most of the southern states including South Carolina.

FBI agents were eager to learn more about the SRC. They had determined that two CP members, James Edward Jackson and Nemmy Sparks, had represented South Carolina at a recent meeting. However, Jackson and Sparks lived out-of-state, and the regional organizer, Sylvia Hall, was based in New York City. The FBI had been unable to identify any of the Communists believed to live in the state.

The FBI did have its suspicions as to who the communists were. They believed they were the accused professors at Allen along with professors at Allen's neighbor, Benedict College. The conjecture was not unreasonable. The available information (age and race) matched.

The FBI agents were able to tie two of the professors to the Communist Party albeit with only a thin thread. In May 1957, an unnamed source provided the FBI with documents obtained from Sylvia Hall's apartment. The source had evidently searched the apartment and found papers referencing two of the South Carolina professors. The papers included an envelope with "The old man" written in a corner. Inside was a note that read "Anyone coming to see people will go to Ed Hoffman's house (phone book) or at Allen University" as well as the address "JSK, 1408 Senate St., Allen University." An address found on a torn envelop clarified the initials. They were those of "Rev. J. Spencer Kennard, Jr., 1408 Senate St., Columbia S. C." on a torn envelope placed in box containing the operational records of the SRC. 

A third note referenced one "Morris" which agents believed was Shirley Juanita Morris, a student at Benedict College. Intriguingly, she appears nowhere in the public record even though the FBI had identified her as a CP member. Her alleged communist created to problems for her. While the professors were forced out of the state by the end of the year, Morris remained at Benedict and graduated magna cum laude in 1960. 

The FBI used the tools it had to investigate the professors. They examined the professor's mail for a month and consulted with their informants (likely state law enforcement and sympathetic faculty at the schools), but ultimately obtained nothing of substance. Edwin Hoffman did a good job of summarizing the event in a letter to a friend. After explaining the situation, he wrote, "It can be said our state suffers from a cultural lag and McCarthyism is first limping into the picture at this time."

The Security State Spots Edwin Hoffman

After waiting over five years, I finally got the FBI files on Edwin D. Hoffman! Hoffman was one of the three professors at Allen University who were accused of being "communist workers" by South Carolina Governor George B. Timmerman  in the late 1950s.

The governor's accusations drew national attention and ultimately resulted in the dismissal of Hoffman and the other professors. However, the core of his claims is, at best, unclear under close inspection. The governor claimed that he'd seen records showing that Hoffman had been a member of various "Communist Party affiliates," but the groups he cites were obscure ones like the Queens County Financial Station. The name "Financial Station" certainly does not suggest troubling communist political activity. Moreover,  in most cases, Hoffman's involvement dated to a decade or more in the past. The FBI records shed light on the governor's claims and show how Hoffman first caught the attention of the surveillance state. 

The FBI files show Hoffman enthusiastically participated in left-wing political activities, essentially all of which the governor would have condemned. However, there is no evidence that he was a member of the Communist Parties, and his documented activities were well within constitutionally protected activities. 

Hoffman was first investigated by army counter-intelligence during the Second World War, and this occurred almost by accident. Shortly after Hoffman joined the army (on August 8, 1942), an army investigator was examining records at the City College of New York as part of an investigation unrelated to Hoffman. In the course of doing this, the investigator came across evidence that Hoffman had engaged in "communistic activities" while a student in the late 1930s.

The investigator's claim wasn't baseless, although it is striking that City College retained records on Hoffman several years after he graduated. When Hoffman attended the college, the student population drew heavily from the city's working-class immigrant families. Many had seen their families devastated by the Great Depression and were open to political and economic change. This manifested itself by a large wave of campus activism.

Hoffman was a leader among student activists. He helped form and then lead a chapter of the American Student Union. The governor of South Carolina described the ASU as a Communist Party affiliate, and he wasn't wrong. The ASU had been formed from various socialist and communist groups. However, the City College chapter focused on issues like opposing war and fascism. They successfully advocated for removing mandatory ROTC training. Hoffman's proudest achievements were those related to African-American rights. A student protest he participated in succeeded in getting the college to hire its first African-American instructor, Max Yergan. 

Hoffman's old student records were enough to trigger an army investigation. As Hoffman was enrolled in the army during war time, the military had extensive powers to investigate his activities. Investigators interviewed a number of the soldiers who worked with him as well as former employers, reviewed all the records they could get their hands on, and inspected his mail.

Investigators found a few additional causes for concern. Hoffman's political activities hadn't ended with college. When he registered as a voter, he had registered as a member of the American Labor Party. This was a legitimate, if minor, political party, but it promoted a socialist democratic platform and was viewed by some as communist-aligned. 

A more serious cause for concern were his Selective Service records. They showed that he told the draft board that he was a conscientious objector. Moreover, he had actively protested against the war. On May 29, 1941, Hoffman was stopped by a police officer for distributing "Communistic" anti-war pamphlets on a street corner. While the officer did not arrest him, he did make a report to Hoffman's draft board.

May 1941 was before both the invasion of the Soviet Union and the bombing of Pearl Harbor, and Hoffman's feelings about the war had evidently changed by the time he was inducted into the army. Every soldier interviewed spoke highly of Hoffman, although a few acknowledged that he held leftwing political beliefs. One officer said he was "pinkish" but loyal to the US. Two other officers said that Hoffman had discussed being an active communist during college and acknowledged that he had maintained friendships with active communists.

None of the soldiers expressed serious concerns about Hoffman, but investigators felt differently. They noted that he was friendly with another soldier (Richard J. Crohn) who had already been flagged as "proven to be a dangerous Communist." 

On January 22, 1944, army investigators completed their work and concluded that Hoffman was at least a communist sympathizer. He was deemed potentially subversive and a few months later (on March 17, 1944) he was restricted from sensitive duties.

Hoffman, it appears, was unaware of the investigators' findings or even the existence of an investigation. All evidence shows that he was a loyal soldier who dedicated himself to the war effort. He was commissioned as an officer and volunteered for combat duty at the time of the Normandy invasion. He served as an artillery forward observer in western Europe and earned three Battle Stars. 

Hoffman's military service did nothing to erase the army's concerns about his communist sympathies. A few years after the war (on March 26, 1948), the army provided the FBI with a copy of their records on Hoffman. Using those records as a basis, the FBI opened an investigation. The reasons for sending records to the FBI are unclear. In Hoffman's file, an FBI cover letter attached to Hoffman's military record is followed by four pages that are fully redacted under a national security exception. These pages may contain a description of why the files were transferred.

However they made their way to the FBI's files, the records were hugely consequential. The governor's accusations against Hoffman were a regurgitation of the army's findings together with some activities that the FBI had investigated in the early 1950s. 

Hoffman's Anti-War Pamphlet.

The text below is a transcription of the anti-war pamphlet that Hoffman was distributing in May 1941. The pamphlet is a response to one of President Roosevelt's radio addresses in which he encouraged Americans to support the war. At the time, America was nominally neutral, but it was sending supplies to the Allies via Naval convoys, something the pamphlet openly opposes. The pamphlet also opposed organizing an A.E.F., which was an acronym for an American Expeditionary Force (the term for the US military forces that fought in World War One). 

The pamphlet encourages readers to listen to an answer to the president's speech that Congressman Vito Marcantonio would deliver. The reference to Marcantonio is highly significant. He represented East Harlem and was one of the most left-wing members of Congress. At the time, he was a member of the American Labor Party. During the early years of the war, he was among the staunchest opponents to American military involvement. His position appears to have reflected his sympathy for communism. Up until June 1941 (a month after the pamphlet was distributed), the Soviet Union and Nazi Germany had formed a non-aggression pact, and  the Soviets encouraged American communists and their allies to advocate for American neutrality. 

 

A War Hero turned Communist Worker?: Lewis Smith's military service

Aleck Lewis Smith was one of the professors that the governor of South Carolina publicly accused of being a "communist worker." When I first started researching the governor's accusations, all I knew was that Smith was employed by Benedict College, and that he was accused of the following: 

The governor's claims about Smith

Taken at face value, Smith's dishonorable discharge from the Navy is one of the most notable accusations. The others were considerably thinner: Forrest O. Wiggins was accused of having a subscription to the National Guardian newspaper, Rideout of being a member of the Progressive Party during the 1948 election. Today I got a copy of Smith's notice of separation from the Navy, so I can now seriously evaluate the governor's accusations.

The most important fact? The governor was wrong! Smith was discharged on March 1, 1955 because of lack of progress. There was nothing suspicious about this. In 1955, Smith was a middle-aged English professor who had been in the Naval reserves for a decade. He had little to offer the Navy, and he certainly had better things to do. There was nothing unusual about his discharge for "lack of progress."

Notice that Smith was discharged from the Navy

The governor also left a lot out. Whatever Smith's politics were, he had served in the U.S. Navy with distinction. He enlisted on November 11, 1940. Smith left no record of his reasons for joining, but he had just graduated from college, and the Navy offered a promising career path for men like him. During the winter of 1940, the U.S. Navy was rapidly expanding as everyone anticipated that America would soon enter the Second World War. 

The war, of course, began for America in December 1941, following the Attack on Pearl Harbor. Smith spent most of the war stationed on the USS North Carolina, a fast battleship. The North Carolina was in the Pacific for almost the entire duration of the war, and it was used to screen aircraft carriers and cover troop carriers. The ship saw combat and was even torpedoed twice during the Guadalcanal campaign. 

In February 1944, Smith was transferred to the USS Pasadena. The Pasadena was a light cruiser, a smaller and more modern ship than the North Carolina. However, it performed a similar service: screening aircraft carriers. 

On both ships, Smith served as a damage control specialist. This was an important and stressful role on a battleship that saw combat. Uncontrolled fires could be catastrophic as the North Carolina was loaded with fuel and ammunition. Combat was an especially intense time. When the ship was torpedoed, fire broke out, and the ship took on tons of water, and damage control specialists had minutes to limit the damage and save the ship, and its crew, from destruction.

At the end of the war, Smith was awarded a bronze star for participating in the liberation of the Philippines and three additional bronze stars and two silver stars for service in the Pacific theater. These awards honor his participation in fourteen separate campaigns, an impressive record of military service by any measure. 

Smith left active service shortly after the surrender of the Empire of Japan, on October 28, 1945. Although he remained in the Naval reserves for a decade, he appears to have lost interest in the military. He moved to the midwest and went into higher education. He does not appear to have recorded his thoughts about his naval service, but his activities suggest mixed feelings. He married a Japanese woman and, in 1970, moved to Hiroshima, Japan, where he lived for two decades.

Smith's trajectory is suggestive of one seen with other men his age. Shocked by the destruction he witnessed during the Second World War and deeply concerned about the prospect of nuclear war, he rejected mainstream American politics, deeming it inadequate to the moment. He explored ideas like communism as they offered an alternative, but in Jim Crow South Carolina, even such modest actions made one a target for state politicians. Whatever his political beliefs, Smith certainly represented some of the best of mid-century America. He not only served the country with distinction during war, but during peace, he endured a dishonest political attack made for an abominable goal and he did so with dignity.

The Melton Family After Reconstruction

In Chesterfield County, South Carolina, Reconstruction ended early in the morning of April 16, 1871. On that day, after months of harassment, Robert Melton and his wife were murdered in their home by a group of Ku Kluxers. 

Robert was a small farmer who lived in the northwestern part of the county, near the modern location of Pageland. After the Civil War, he was one of the few white men who actively supported the Republican Party. The party was especially active in the region. The newly elected state senator worked with northern investors to develop the township. They planned to purchase plantations, subdivide them, and sell small plots to farmers. Alongside those efforts, they were going to connect the region to larger towns via a railroad. The outcome, they hoped, would improve the economy and secure the county as a long-term base of Republican support.

Robert was well-positioned to support their efforts. He was one of the few South Carolinians who had supported the Union during the Civil War. When Gen. Sherman's army passed through the area, Robert and his family eagerly provided them with all the provisions they had.

At least in the beginning, Reconstruction uplifted Robert and his family in ways that would have been impossible during antebellum. Because of the development efforts, Robert was able to purchase his own farm, and he was appointed to positions in local government, serving as an election manager and a tax assessor. 

Robert's murder by Ku Kluxers brought an abrupt end to all of this. Not only was Robert dead, but the political changes that had empowered him were reversed. Investors abandoned their development plan, and fearing for their lives, Republican politicians fled the county. Chesterfield's wealthy lawyer / planter class regained political power and would govern the county for generations.

The Ku Kluxers killed both Robert and his wife Harriet, but they left behind their children, seven in total. All were young adults, and four of them were still living on the family farm. They witnessed the murder of their parents, and one daughter, Sarah, suffered a serious gunshot wound.

The Melton murder made the national news. The press mentioned that Sarah was hurt so badly that her leg would likely be amputated at the thigh, and she might succumb to the injuries. However, by the next month, the press cycle had moved on, and the Melton children passed out of the public spotlight.

The Melton children did not, however, escape the fallout of their parents' murder. Locally prominent Republican leaders were able to simply abandon Chesterfield County, secure employment elsewhere, and move on with their lives, but the Melton family had far fewer options.

Robert and Harriet's seven children were: (1) Mary Jane, aged thirty-two,  (2) Alexander L., aged thirty, (3) William J., aged twenty-eight, (4) Harriet Ann, aged twenty-six, (5) Sarah S., aged twenty-four, (6) Emaline R., aged nineteen, and (7) Robert C., aged seventeen. 

At the time of the Ku Klux attack, the oldest son, Alexander, was married and running his own family farm near Chesterfield Courthouse. He does not appear to have been very involved with his parents and siblings during this period. The others, however, were in the thick of things. Mary Jane, Sarah, Emaline, and Robert C. were all living at home, and Mary Jane, Sarah, and Robert C. had even been present during the assault. (Emaline is not mentioned in any accounts. She might have been absent, or she might have been shielded from public scrutiny because of her youth.)

Harriet Ann had married the son of a neighbor, Robert A. Allen, and they were running their own farm. However, they remained near Harriet Ann's parents, and Robert A. was one of the first people at  the scene after the attack. Harriet Ann's brother, William J., was also living in the area, although the records of his activities are thin.

One reporter wrote Robert Melton's daughters had "imbibed much of his courage," and both his sons and daughters demonstrated that in the years after their parents’ murders. Although the murderers went unpunished and likely remained in the area, the family was fully unrepentant. The year after the murders, Mary Jane submitted a financial claim to the federal government for compensation for the provisions they had given the Union army. As part of her claim, she had to submit a list of witnesses who would support her claim. She offered a full-throated response: "All of Old Store, Chesterfield Co."

Mary Jane's claim was barred. The reason for the decision was not recorded, but it may have been  rejected for technical reasons. She hadn't owned the provisions, they belonged to her deceased parents owned them. In any case, Mary Jane was in dire financial straits, and she had sacrificed much for her loyalty to the Union, so the rejection must have been hard to accept.

Hardship continued to be the lot of the Melton family for the duration of the 1870s. Emaline, Mary Jane, and Sarah all married, but Mary Jane and Sarah each lost their husband within a few years. Mary Jane's husband had a heartbreaking end. He suffered from severe mental health problems, and in March 1879, he committed suicide by jumping down a well. 

It is hard to imagine how Mary Jane and Sarah survived. They were middle-aged widows living in the community that harbored their parents’ murderers, and they had to provide for themselves and their children despite having no clear means to earn a living. They lived near Emaline, so they may have been supported by her family and her family friends, although in rural South Carolina, nobody would have had much money to spare.

Intriguingly, Mary Jane appears to have remained close to the few Republicans who remained in the area. In 1900, she was boarding with John McCulla. McCulla had served as county treasurer during the beginning of Reconstruction. By the end of Reconstruction, he had become a hated figure and was criminally prosecuted for misuse of his office and for election fraud. 

Despite the criminal charges, McCulla had fared better than the Melton family. He owned a great deal of farmland which he rented to Black sharecroppers. His boarding of Mary Jane was likely an act of generosity towards an old friend who had long ago fallen into hard times.

Mary Jane later moved in with the daughter of her sister Sarah. It’s unclear what happened to Sarah herself. She disappears from records after 1880.

The rest of the Melton family followed the path that many in rural South Carolina followed. They and their children eked out a living by farming until the growth of cotton mills and manufacturing work drew them to townlife. Likely in search of better work opportunities, Harriet Ann and her brother Robert C. moved to North Carolina. Harriet Ann's husband found work as a factory hand in Charlotte, while Robert C.'s son worked at a cotton mill in the town of Stanley.

Emaline remained in South Carolina, but her family moved to Lancaster County where her husband found work as a cotton mill laborer. Mill work was unhealthy and physically demanding, yet Emaline's husband was still working in the industry when he was in his sixties. 

All in all, the later history of the children of Robert and Harriet is one of quiet persistence in the face of financial hardship and bitter political conflict. Robert C. appears to have been the longest lived of the siblings. He died on November 20, 1932 and was buried in a Baptist cemetery in Mount Croghan. His death marked the end of a remarkable family. Although they were poor farmers, the Meltons had stood up against the Confederacy and offered nearly everything they had to the Union army. After the war, the father and mother gave their lives trying to build a new South Carolina, and Robert C. had personally fought the Ku Klux Klan. Despite all this, Robert C.'s death does not seem to have received any public notice. His gravesite, located at a simple county church, remains the sole physical marker of his presence. His tombstone is inscribed with the phrase "Gone but not forgotten," and the story of his family's lives remains to be remembered by those who explore South Carolina's history.

 

Sources

1) The National Archives in Washington D.C.; Record Group: Records of the Bureau of the Census; Record Group Number: 29; Series Number: M653; Residence Date: 1860; Home in 1860: Chesterfield, South Carolina; Roll: M653_1217; Page: 145; Family History Library Film: 805217  

TESTIMONY AT INQUEST FOR MARY JANE MELTON'S HUSBAND 

Miss Elizabeth S. Hill

Witness is sister to deceased. Witness has noticed for some time that deceased was not right in his mind. Heard him say on yesterday morning that he wished he was dead. Don't think that any person threw him in the well.

Robert Brewer

Witness was to work for Mr. Gathing today. Was going from the barn to where he was to work about eight o'clock this morning and saw deceased's coat and hat hanging on the windlass of the well.

Witness and others let down a pair steelyards, the hooks caught in the clothes of deceased and brought him up.

Witness thinks deceased went into the well voluntarily. There was plenty of water in the well to drown a man.

Robt. Brewer

John Gathings

Witness recognises the body here lying dead to be the corpse of A. M. Hill. Has known deceased intimately for fifteen months.

Witness has heard deceased use expressions which led him to believe that deceased would rather be dead than alive. After seeing deceased's hat and coat on the windlass of the well, witness with others proceed to search for him and found his body in the well.

Witness thinks that deceased went into the well voluntarily. Witness has noticed that deceased has been despondent and in low spirits for some time.

John Gathings

Mrs. Mary J. Hill Sworn

Was married to deceased in April 1878. Have noticed several times that his mind was not right.

He complained of his head hurting him not long ago and said that he wished he was dead. Felt uneasy all the time that deceased might commit suicide. Deceased got up at daylight this morning and walked off.

He seemed quite cheerful last night.

Witness does not think that deceased committed suicide by throwing himself into the well.

Mary J. [her mark] Hill

Pickle's story of leaving South Carolina

The account below was written by Frances Emmaline Allen ("Ema Pickle"), and it follows her account "Emaline's Tale" of Sherman's Union army in Chesterfield County, South Carolina. "Emaline's Tale" ended with the departure of Sherman's troops. 

The Allen family lived on a small farm near the town of Mount Croghan. They owned one-hundred thirty-five acres and farmed twenty-five acres on which they raised food provisions and a small amount of cotton. Emmaline mentions workers who were enslaved on the farm, and evidently, the workers made up at least a few families since she mentions multiple cabins on the farm. Unfortunately, its unclear exactly how many people were enslaved there as their presence went unrecorded in the 1870 census.

In the story below, Emmaline begins with the departure of Sherman's troops from the area around Mount Croghan, which was early March 1865. As was the case throughout the region, the Allen family was left destitute.

Emmaline describes Confederate soldiers, including members of her family, returning home after the Sherman's army left. She mentioned by name her father Eli, brother Robert Alfred, and an "Uncle Jas." I have not been able to verify the military service of all these family members. Eli does not appear in any records I examined, but he was in his fifties during the war (very old for military service), so he have served the Confederate government in a non-military capacity. Robert does appear but the extent record does not agree with Emmaline's recollection. He enlisted at the start of the war (August 25, 1861) in the Chesterfield Light Artillery but was discharged for health reasons. within a year, long before Sherman came to Chesterfield County.  "Uncle Jas" I have not been able to identify. The discrepancy could be attributed to missing record to Emmaline's mistaken memory. 

The narrative ends with Emmaline's family leaving for North Carolina and starting a new life for themselves. By 1870, the family had moved to Cabarrus County, North Carolina. Emmaline says that the family abandoned cotton growing and worked on the railroads after after the move, but the 1870 census records show the family continue to farm. However, they are recorded as farm hands, rather than farmers, indicated that they likely didn't own their own farm, showing a significant drop in economic status. 

Emmaline married a man by the name of James T. Burris in the 1870s (probably around 1872). Burris was a farmer in Stanly County, and Emmaline focused on home life, raising three kinds and helping care for her father-in-law William.  They remained on their farm until Emmaline's death in 1930.

Emmaline's discusses her sister, Nancy, who became a pariah after becoming romantically involved with a Union soldier. Independent records indicate that Nancy indeed was scarred by her experiences during the war. In 1880, she was living with Emmaline's family. Despite being in her late thirties, the census taker recorded that she was single, unemployed, and disabled. 

Homecoming – Leaving Home

After the army left, and the others that were to follow, had come and gone we had virtually nothing left. Our spirit was broken along with furniture, and all belongings. Mama managed to repair the secretary and safe with tin panels.

A lot of collard greens were left unharmed so we cooked messes of the stuff. Once Mama had stored a pot of them in the pie safe for overnight, and it came up a bad storm. Mama was scared to death of thunder and lightning. She woke us all up in the middle of the night and made us sit in the hall, lined up along a wall. Sleepy headed as we were, some would start snoring and she would pound our head and say that we should be prepared in case we have to run. After awhile the rain slacked off from making so much racket on the roof and it had been some time since the last crack of thunder. Mama told Eli to get up and look out the door and see how bad the storm was. He had been the one snoring and was half asleep. He accidentally opened the safe door instead and stuck his head in. There was a long period of silence and Mama hollered at him. 

"It's black as pitch and smells like collards!" He reported.

We faced another enemy: poverty and starvation. It was soon late spring and we needed to get seeds into the ground. Luckily, some seed was saved, but we didn't have a mule, plow or even a hoe to plant with. We had no saw to cut stove wood. One of my brothers tried to make a plow from piece of the old one, it turned out crooked and useless. Te saying "make do" had to be born here. There was no source of assistance available. The handful of help which were left behind were very aggressive towards us. The rest  either went on their own or were forced to join Sherman's army.

Somehow we did get another vegetable garden started and found cornmeal that was hidden. There were no salt to put in it or anything to make it rise, so it was flat and tasteless, but it was hot and filling. We did have a lot of fish fries in those days.

For months, I would awaken in the still of night and be very distressed. Squads of ragged looking refugees and our beloved soldiers in grey had been passing by daily.

One cold Saturday afternoon in December, we saw a column of weary men coming around the bend of a road, a dismal train of returning soldiers. A lanky, giant of a man with rounded shoulders and sunken chest broke apart from the rest. There was a tangle of black hair sprouting from holes in his hat like lots of ruffled grouse feathers and he looked quite a spectacle to us with his scraggly beard as he came closer. The  walk was very familiar but his stooped posture didn't fit my remembrance of Papa.

Us children, tow headed and thin as young pines standing there in rags must have looked like a spectacle to him as well.

"It's Papa! Eli shouted. The words "Papa" ran from every mouth except mine. There was no way that is Papa. You could not make me believe it. I remembered him being tall, but not this thin. The weather was bitter cold, and the wind whistled through the leafless trees. As he drew near us, the dozens of bearded men stopped to see what would happen. 

He definitely was coming to our place. There was a familiarity to his dusty, haggard face, a face which wore a weary expression that was painful to look at. He had no soles on his shoes, just tops. He walked from Pig Point, Virginia all the way to Chesterfield County.

But it was Papa. We surrounded him joyously. Hugged him and ground our faces into his rough worn coat, making it soaking wet from tears of joy.

The old wash pot that served many purposes and plugged with tallow was full of hominy Mama had prepared for our meal. She invited the soldiers to have some and never in my life have I witnessed such ravenous hunger. They filled their dirty hats with food to use for bowls. Some even used pieces of wood found on the ground. We drank some poor tea made from sassafras leaves and bark.

We had managed to save some of the silver by hiding it in the woods under the sod. The rest we could not find, but for one of the boys had placed it under a fence post. After the soldiers took up the posts to burn, its whereabouts could not be located. One of the neighbors had saved some household furniture, gave us a couple of chairs, two stools and an old pine table. We located several pieces of crockery and a wash basin, but we had to drink from gourds.

Papa had not seen home in four years. Yet here he was and he intended to try to regain the life he left. It was a hopeless cause. Try as he might, he did not succeed. We were not even left a Bible to read. The lawless men had took all the best horses and cows, then shot the rest and left them lying to rot. There was no way for us to get rid of them. The stench was unbearable. Everything was gone, not a building fit to use left standing, the house itself beyond repair.

The soldiers had even pushed Mama down snatched the shoes right off her feet and threw them in the fire when the woodpile was burning. As she fell, her snuff box tumbled from her picket, and of course, was promptly taken. I always looked forward to going in the yard and breaking off twigs to chew up the ends for Mama's toothbrushes. I would chew and chew to make it soft for her to dip with, now there was no reason to do so. We had only one needle to sew with; the one Margaret had hidden in her bonnet. 

Franklin and Eli were thin, sickly and often plagued with horrid nightmares. Eli never fully recovered from his shock at the creek. He used to be a fleshy child, but was now very peaked. The unhiding in the bushes the night of his torture was our Uncle Jas. He had returned from Virginia, traveling by night to avoid capture. He didn't survive, for he was caught when the found his haversack. He was tortured severely, and he died as a prisoner of war with the pneumonia. They were kind enough to let us retrieve his body which we laid him to rest under the magnolia tree. Such a lonely grave there by itself.

My brother, Robert, had walked most of the way home from Pig Point, Virginia. He had to hide out to avoid being taken and endured most horrid circumstances eventually being taken prisoner a short time. 

We were children that had lost childhood. Our family had led a life of hunger and terror. Nancy was in a bad way since she returned. Thin and claimed to have headaches. She would hardly eat and didn't do any work. Not that she was so inclined to do alot of anything before this, but now it was either an excuse or the fact that apparently she was still in love with our enemy. Somehow Papa found out, he always finds out everything.

He had not expected all that. When Papa returned to find her in this state. I think that's where his world ended. Nancy was his pick. He always believed every word she said, but her actions and attitude broke his heart. He felt that Chesterfield County was no longer our home.

I knew a change was coming before Papa decided to move us. I watched as daily, Mama stood, face streaming with tears when she thought no one could see. She had been caught wringing her hands for she could not bear to see him this dreadful shape. It sickened her heart. She grew even leaner and her face had a pinched look to it. At first, I thought it was because she missed her snuff. In the past, John would play pranks on her and all of us for that matter. He would hide her spittoon when she was in the house cooking or cleaning and didn't want to go out to spit. She would look around trying to find it and her face had this same pinched look. Lack of a place to split was not the reason for this.

 There was not a fence post anywhere. The peach and apple trees were gone. We did find the silver that was buried on the creek bank and a barrel of salt that we had disguised by burying it and covering  it over with leached ashes. But we couldn't even restuff our bedticks, there wasn't a chicken or duck on the place and the soldiers had taken their swords and stirred up the molasses in the feathers they took out of the pillows.

Such a sight of desolation and the prospect was gloomy at trying to resume our former life. Papa had tried but the coming of the carpetbaggers and those "bummer" outlaws was the last straw. We were attempting to resew our bed clothing when an outlaw came in and used his bayonet to reduce the clothing to small pieces. Finally, he gave up. He hitched a scrawny horse that he bought off someone to our ragged buggy and loaded what we had salvaged, of our property which wasn't much, and we left Chesterfield forever. 

Papa said we were going to a different world. A world with no more sand spurs and swamps. We would be in North Carolina. To us children, that place sounded grand compared to what we had been accustomed to. Also, he gave me a new nickname, 'Ema Pickle'. Maybe he is taking a cotton to me now, being as I am older and help mama the most around the house. He seemed to pay more attention to me than before.

He decided to take us to his brother in Mecklenburg, North Carolina. We lived for awhile in Morningstar Township which is which is Matthews before moving to down town Charlotte. My grandpa, uncles and aunts went along. too. The whole family left that barren, God forsaken country and it's bad memories. Grandpa and all the rest ended up in Cabarrus County, North Carolina. Up at Rocky River Township. Even help start a church in the area. There, we tried to forget the past by starting a new life. I know things are going to be better now. Mama's hands don't shake like they used to. I watched her part her hair in the middle and pull it smooth over her temples, then put it up in a top knot with a new jeweled comb that Papa brought her. Our cotton life was over, railroads were our new way to make a living. 

A neighbor's girl named Mary went with us when we left. A lot more wanted to go, but there wasn't enough room. It seemed she was with child and rumor was that it belonged to a Yankee. The family rejected her and she had to leave home. Papa felt sorry for a young girl in a situation like this especially since she was a friend to Nancy.

Mary's baby was born by the light of a tallow candle in our new place in Matthews, North Carolina. It was a beautiful boy. No matter, though, for Papa said she couldn't stay with us any longer. She would have to go to Anson County, to an all female college along with Nancy for they needed book learning. Mary took the baby with her. Later, the poor dear was discovered dead. It was told that she poisoned him and put the body in her suitcase. It was found by the lady that ran the boarding house where she stayed. For some reason, nothing every came of it. We didn't breathe a word about what she'd done even though we felt Nancy had a hand in it.

Nancy never married, but she went to church every Sunday. Why, she even went so early as to sit on the porch steps out front until it was time for the people to arrive. She was there every time the doors opened for the rest of her life. Papa told us tales in the evenings after supper about his long walk home. There was no money for to pay his way.

Coming through a steep, rugged pass in the mountains of North Carolina, he spent the night in an abandoned school house. He knew it was abandoned because it was thick with cobwebs and mice had knawed all the books into  nest materials.

In the middle of the night there was a slam and a bang. The benches were shaking and started turning over row by row. Before it got to the row on which he laid down, he fled out the door and ran as fast as he could without looking back. 

Then he had hidden in a briar patch and watched two outliers kill a man and take his horse. Both of them rode it off together and it was too much for  the horse bare. Horses were scarce for so many were killed in the war.

In one town he crouched in the weeds beside the train tracks until it was near time for the train to leave the station. When the whistle started blowing, he ran and pulled himself into an open boxcar. The workers walked up and down looking inside each one but somehow they didn't see him. The rattling of the wheels sang him to sleep and he almost didn't get off in time before the next stop.

Uncle Willie lives in the edge of the mountains, and it wasn't out of his way to stop by for a visit with the kin folks. His brother had not yet returned to help the family with the spring planting, so he stayed for a few days. He was in a hurry to get the last field plowed so that he could continue on his way and each evening he worked until there was no light to see by.

The last  few rows were down near a thick patch of woods. Suddenly, the mule didn't want to go in that direction. It rared up with eyes wild looking like an animal in a trap and snorted until it's nose bled. Then after a lot of tossing around, the harness came a loose and that mule ran off faster than any one Papa had seen go. There was deadly silence, not even a bird was heard. Papa froze, for standing in front of him was a huge bear. Someone up above must be looking after him, because he made it back to the house. Walking backwards the whole way. He said that he didn't want to turn his back on it.

Papa's advice is for all us southerners to work hard and wait. Wait for their time to come. One of these days, it will. 

Stopping the Steal in the 1876 South Carolina election

In this post, I want to take a look at the 1876 election in South Carolina through the lens of statistics. The post will be riffing on Ronald F. King's article "Counting the Votes: South Carolina's Stolen Election of 1876."

The election was among the most eventful in American history. It resulted in white Conservatives sweeping the pro-Black Republican Party out of office. The state's electoral votes played a key role in Rutherford B. Hayes's presidential victory, and for generations, South Carolina would be governed by white Conservatives who almost wholly excluded Black people from political life. 

The Conservative victory was all the more remarkable because demographics were against them. Politics were sharply divided along racial lines, and white voters made up only one-third of the electorate. How did their achieve victory? By "shoot[ing] negroes to get relief from the galling tyranny to which we had been subjected," in the words of the influential politician Benjamin Tillman.

Political violence played two significant roles. First, it allowed the radical faction of the conservative party to dominate the party. In the summer before the election, the conservatives were divided between moderates who felt it best to accept Republican rule and extremists who wanted to try and throw out all Republicans. Going into the summer, it appeared that the moderate wing would prevail, but starting in July, radical Conservatives provoked several horrific violent political riots. This polarized politics so much that all moderation was abandoned. 

The second role of political violence involved events after Election Day. The electoral outcome was close, and each candidate claimed to be the victor. The result was chaos. For almost a year, two different bodies claimed to be the state legislature, and a great deal of ink was split trying to resolve the issue through political and judicial procedures. Ultimately, the matter was decided not by pen and ink but rather by rifle and saber. Over the years, conservatives politician had organized a state-wide paramilitary group that was more than capable of overwhelming the state government's meager law enforcement and military forces. They were temporarily held back by the presence of the US army (the one organization capable of mustering greater armed manpower). However, in March, the president ordered the army to disengage from the political conflict. Continuing to press his claim would have put the Republican claimant's life in danger, so he finally conceded defeat.

The goal of this post is to closely examine the election returns. Conservatives leaders were only able to challenge the electoral outcome because the vote was so close. When the votes were counted, the Conservative candidate for governor was ahead by 1,324 votes. However, the returns from three counties, Barnwell, Edgefield, and Laurens, were challenged by Republicans who claimed that Conservatives had committed illegal ballot fraud and voter suppression. The ballot total with these three counties dropped was a victory for the Republican candidate.

The political process for resolving the dispute was complicated and, ultimately, irrelevant. Under the state constitution, the newly elected state legislature was to declare the vote for the governor, but two different organized bodies claimed to be the legislature and each declared a different candidate the governor. While each body offered legal justifications for their decisions, the only argument that the Republican-dominated body ultimately found persuasive was that an armed paramilitary group would murder them if they continued to deny legitimacy to the Conservative-dominated body.

Here I want take a close look at the the actual votes using statistics. The election is particularly amenable to statistical analysis because the political and legal processes generated a great deal of data. Moreover, no statistics were done at the time, so we can hope to get new insight into what happened with this historic event. 

The Statistics Analysis

There are three basic questions that I think are important to answer:

1) Were the ballot counts from the counties of Edgefield and Laurens legitimate?

2) What about the counts from other counties? Although Republicans ultimately accepted the counts for Barnwell County, they claimed that those elections results were improper as well. Conservatives charged that Republicans themselves had committed fraud, especially in the counties of Beaufort and Charleston. Do these results hold up?

3) What would the election outcome have been if there had been no election fraud?

Statistics alone can't answer these questions. Regarding the first question, a statistically unlikely voter count for Edgefield could represent voter fraud or it could represent unusual local political conditions.

In his paper, Ronald F. King uses a very simple but also very useful statistical model of the election. He assumes that race is the only determiner of voting behavior, and the each candidate gets fixed proportion of the white vote and the Black vote. These proportion can be estimated using statistical techniques.

A bit more formally, King's model is that

        Y = p X_1 + q X_2,

where

    Y = the proportion of the vote that the Republican candidate gets

    X_1 = the percentage of voters that are Black

    X_2 = the percentage of voters that are white

    p, q = are the fixed percentages of votes that the Republican candidate gets.

The quantities Y, X_1, X_2 are treated as random variables, and we treat each county as a sample of the random variables. Since X_1 + X_2 = 1, the previous equation can be rewritten as

    Y = q + (p-q) X_1,

and King estimates p and q by a least squares linear regression on the county data. The Democratic candidate is handled analogously. 

The model clearly has a number flaws. Voter behavior was certainly impacted by more characteristics than just race. For example, birth state played a huge role in determining the behavior of white voters. White men who were veterans of the Union army and had moved to South Carolina after the war were among the most loyal Republicans. The counties were also far from homogeneous. Almost everyone in South Carolina was a farmer, and in most part parts of the state, they grew mostly cotton and corn. However, in a few counties along the coast, rice was the main agricultural product. Despite these flaws, we will see that the model works quite well. 

There are five sets of data that we will use to make our statistical estimates:

1) the county-level election returns

2) a county-level record of election day sign-ins,

3) the 1870 federal census record of the populations of the counties

4) the 1880 federal census record of the populations of the counties

5) a 1875 state census of the population of the counties

Each of the data sets 2), 3), 4), and 5) can be used to estimate the percent of white residents / black residents. 

The plot below shows the data. Each data point is a county, the x-axis measures the percentage of the population that was recorded as "Black" in the 1870 census, and the y-axis measures the share of votes that went to the Republican candidate. Visually, we see that most points approximately lie on a straight line, but there are a few outliers. 

The equation of the red line is Y = -.033 + .891 X. In other words, the least squares estimate is that the Republican candidate got -3.3% percent of the whites vote and 85.8% of the Black vote. Certainly the negative percent does not make logical sense and is an artifact of our statistical methods, but it indicates that the share of the white vote was negligible. The 85.8% figure is also reasonable. Historical accounts report widespread dissatisfaction with the Republican party, and some Black voters turned to the Democratic party either out of opportunism or after being physically threatened.

The adjusted R-squared value is 0.857, confirming that a linear model is a good fit for the data. The standard errors for the constant term -.033 and the linear coefficient .891 are .0348 and .0663 respectively. The 95% confidence interval for the constant term is (-0.111, 0.046); for the coefficient, it is (0.755, 1.026).

Our estimates differ from those reported by King in his paper because we are using different data. Instead of using the 1870 census, King used two different models, one based a 1875 census and one on a record of voter sign-ins. A second difference is that I am assuming all eligible voters cast ballots, but King's model allows that some voters do not.

Despite the differences, the models are broadly consistent. For example, using the 1875 census (and dropping Charleston County) King gets the linear model Y = -.033 X + .946 X. The constant term of this equation and the constant term I got are the same up to rounding error. The linear coefficients differ but by a quantity that is approximately .83 standard errors. 

Predicting a Fair 1876 Election

Now that we have the statistical model, what can we say about the ballot counts in Laurens and Edgefield? The plot below is a plot of residuals, that is the difference between the actual value of Y and the value predicted by the model. If the linear model is a good fit to our data, we should expect the data points to be clustered around the x-axis but otherwise not display any discernible pattern. This isn't quite what we see. Three counties are flagged for having unusual values. These are the data points labeled "3", "4", and "12." They correspond to the counties of Anderson, Barnwell, and Edgefield. Just as Republicans claimed, the votes for Republican candidate in Barnwell and Edgefield are suspiciously low. The returns from Anderson County were not challenged, but it isn't surprising to see the county appear as it shares many characteristics with Edgefield (both are in the northwestern part of the state and lie on the Georgia border), so it wouldn't be surprising if there were voter suppression efforts and ballot fraud there.

The statistics don't flag the counties of Beaufort and Charleston although some Conservatives claimed that Republicans living there engaged in voter fraud. Laurens also doesn't appear, but the voter returns look suspicious for other reasons. The reported number of Republican votes precisely matches the recorded number of Black voters and similarly with the Conservative votes. Presumably, the election managers didn't actually record the number of Black voters and instead just assumed that they all voter Republican. 

What about the votes for the Conservative candidate? In our model, everyone votes, so the vote share that goes to Conservative candidate is 1-Y. In particular, we do not get more information by looking at the Conservative votes.

What should the election outcome have been? We can answer this question by using the linear model to estimate what the votes in Edgefield, Barnwell, and Anderson should have been. The table below shows what we get:

County	Actual Republican Vote	Predicted Republican Vote	95% Confidence interval, lower bound	95% Confidence interval, upper bound
Anderson	21.29%	32.25%	19.75%	44.76%
Barnwell	41.25%	51.94%	39.59%	64.29%
Edgefield	33.14%	50.04%	37.71%	62.38%

 

We can only reject the Edgefield result at the 95% confidence level, but Anderson and Barnwell are close. The table below shows what the votes predicted by the model are:

	Actual	Predicted
Anderson County, Republican Vote	1124	1702.48
Barnwell County, Republican Vote	2778	3497.64
Edgefield County, Republican Vote	3107	4690.75

The differences in these three counties are significant enough to have changed the election. The numbers are displayed in the table below. If voter fraud in Edgefield had been prevented, Republicans would have won the election. Our numbers differ from those in King's article, but the general conclusion is the same: in a fair election, the Republican wins by a few thousand votes.

	Republican	Conservative	Republican – Conservative
Reported Votes	91,127 (49.69%)	92,261 (50.31%)	-1,134
Votes with Edgefield replaced	92,711 (50.55%)	90,677 (49.45%)	2,034
Votes with Edgefield, Barnwell, Anderson replaced	94,009 (51.26%)	89,379 (48.73%)	4,630

While a statistical analysis shows that the Republican party should have won the election, it also shows that the Republican Party was in serious trouble. The Republican Party had won the previous election by a comfortable 11,585 votes.

The Other Elections 

The unusual nature of the 1876 election shows up even more clearly if analyze the earlier elections. Plots of Black population versus Republican voter share for the elections in 1874, 1872, and 1870 are shown below. For the years 1874 and 1872, there is a clear correlation between voting behavior and the Republican voter share, but it is weaker than it was in 1876. Another notable point is that the linear models predict that Republicans pick up a significant share of the white vote, more than one-third in 1874. These features can be explained by the nature of the elections. There was not really a Conservative gubernatorial candidate. Instead, the two main candidates represented opposing factions within the Republican Party. The candidate labeled as "Republican" was the regular Republican who ran against a Reform Republican. Conservatives endorsed the Reform Republican in 1874, but they ignored the gubernatorial race and focused on the presidential election. 

The 1874 election

 

The 1872 election

 

The predicted large white vote for the Republican candidate might be an artifact of the model we are using. The model does not account for voters not casting a ballot, and I speculate that many white voters stayed at home on the election day in 1874 and 1872.

A glance at the 1874 election showed the presence of three outliers. Three counties had large Black majorities but very few people voted for the Republican candidate. The counties are Charleston, Sumter, and Clarendon. I was surprised to see them showing up because I haven't encountered any discussion of these highly unusual outcomes. I checked the outcomes, and there do not appear to have been any accusations of fraud or misconduct, and certainly, Conservative newspapers would have crowed about any reports their received. One possible explanation of the outliers is local politics. The outgoing governor, Franklin J. Moses, was from Sumter County. Moses is regarded as a disaster as governor, and he was one of the most hated in the state. The Republican candidate was an outspoken enemy of Moses, and low votes in Charleston, Sumter, and Clarendon may indicate that Moses had retained a loyal base in those counties. Moses was from Sumter, and the county shared a border with Clarendon.

Of the three elections, the 1870 election shows the strongest correlation between race and voter. The adjusted R-squared value is 0.8625, approximately the same as in the 1876 election. This is consistent with the historical record. The year 1870 was a year in which Conservative played an active role in state elections. They formed a "fusion" ticket with moderate Republicans and ran under the name of the "Union Reform" party. The Union Reform candidate for governor was a moderate Republican, but the candidate for lieutenant governor was M. C. Butler, a former Confederate general from Edgefield who played a central role in the 1876 election. 

The 1870 election

 

 

An inspection of the residuals suggests that, unlike the 1876 election, a linear model is not the best one to use for the 1870 election. The residual plot is displayed below. You'll see that the residuals do not appear to be uniformly distributed near the x-axis. Instead, their plot appears to approximate a concave up function. This typically suggests that a non-linear model is a better fix for the data, and a pattern to the residuals disappears if we instead us instead use the quadratic model 

    Y = q + (p-q) X_1 + r X_1^2.