Who Fought for the Confederacy?

Harrison County, Texas, as a Test Case


The question “Who fought for the Confederacy?” seems simple enough at first glance. It probably brings to mind images of thousands of young men such as Josiah Perry Alford, who was photographed in Harrison County, Texas, a little before he went to war in 1861. However, as is true of many historical questions, a broad inquiry leads immediately to more specific queries. And the answers that can be found in the evidence are usually anything but simple. Asking “Who fought for the Confederacy?” leads to at least three more specific questions, each with significant implications for the impact of the Civil War on Texas.

First, what percentage of the military-age men in Texas entered Confederate service in any capacity between April 1861 and April 1865? Was it one-third, one-half, two-thirds, three-fourths? The answer matters because the proportion of the male population who left for the army affected virtually every aspect of life on the home front during the war. To give just one example, the proportion of men who left determined how many farms and plantations had to be worked or managed by women.

Second, what was the economic status of Texans who served in the Confederate military? Were men from wealthy, slaveholding families more or less likely to serve in the military than those from poorer, nonslaveholding families? This is another way of asking: Was the Civil War a “rich man’s war and a poor man’s fight?” Granted this question has become well worn over the years, but the answer still matters in understanding the war.

Third, what happened to men who served in the military in terms of wounds, illnesses, and imprisonment? How many died in battle or as prisoners or war or from disease? The importance of this question really needs no argument. A large number of deaths among the military-age male population of any state or community can be a devastating blow requiring years to overcome.

Now, having asked three of the most important questions that arise from the more general inquiry “who fought for the Confederacy?” I want to offer answers, using Harrison County, Texas, as a test case. (If you are thinking why only one county?—why not the whole state?—why not the whole Confederacy? Please bear with me until you see what is involved in an analysis of just one county.)

Harrison County, which is located on the Louisiana border in northeast Texas, was among the most “southern” of the state’s counties during the antebellum years. Natives of the South headed more than 90 percent of its households in 1860, and 61 percent of those households owned at least one slave. The county had 8,784 slaves, the largest population of bondsmen living in any Texas county at that time. The 1859 cotton crop was 21,440 bales, the second largest grown in any Texas county. Slavery and cotton made Harrison County very southern. When the secession crisis began in November 1860, Harrison County was one of the first to call for a convention to consider leaving the Union, and on February 23, 1861, its voters overwhelmingly endorsed disunion—866 to 44.1

Secession, of course, soon led to war. When word of the firing on Fort Sumter reached Marshall, the county seat of Harrison, on April 17, 1861, the town resounded with cannon fire and patriotic speeches. And the county’s men prepared to fight.2 During 1861–1862, Harrison County provided most of the men in thirteen companies—a company having approximately 100 officers and men—that entered Confederate service. These companies saw varying amounts of action, and I do not have time to give even a brief account of each one. However, let me tell you a little about two of them as examples of the war experiences of Harrison County men from 1861 to 1865. Perhaps the most notable company raised in the county was one that called itself the “Marshall Guards.” Formed soon after fighting began, the “Guards” left for Richmond, Virginia, in May 1861. They became Company E of the 1st Texas Infantry, which later joined three other regiments to form Hood’s Texas Brigade, one of the most famous units in the Confederate Army. As part of Hood?s Brigade, the Guards fought in thirty-eight battles and skirmishes including Antietam, Gettysburg, Chickamauga, and the Wilderness and took massive casualties along the way. At Antietam alone, for example, the Guards suffered nineteen casualties.3

Another Harrison County company, one formed by men from the eastern half of the county soon after the war began, adopted the name, ?Texas Hunters? and became Company A of the 3rd Texas Cavalry. The Hunters first saw action in the Arkansas-Missouri border area, most importantly in the Battle of Pea Ridge in the spring of 1862. Later in 1862 the Hunters crossed the Mississippi and fought in the effort to hold northern Mississippi after the Battle of Shiloh. In 1863 they became part of Ross?s Texas Brigade (commanded by Lawrence Sullivan “Sul” Ross) and served in both the Atlanta and Tennessee campaigns in 1864.4

The “Marshall Guards” and “Texas Hunters,” along with the the eleven other companies raised in 1861–1862, included most, but not all, of the men from Harrison County who served the Confederacy in a military capacity. In 1863, some of the county?s men who had remained at home during the first two years of the war enlisted in a battalion of state troops. These men were generally older, only served six months, and did not engage in battle; still, they should be counted. Also, a significant number of Harrison County men served in units identified with other counties. For example, residents of the northern part of Harrison often joined the 18th or 19th Texas Infantries, both of which were raised primarily in neighboring Marion County. These units remained in the Trans-Mississippi as part of Walker’s Texas Division and fought at the battles of Mansfield and Pleasant Hill in 1864.5

This brief sketch of military activity clearly establishes heavy participation in the Civil War by the white male population of Harrison County, but it does not provide the information necessary to answer key questions about who fought for the Confederacy. To go further requires quantitative evidence. So, let’s turn now to look at a database that I compiled on the military-age population of Harrison County at the time of the Civil War, a population defined as all white males aged 13 to 46 in 1860. These age limits were used because by 1864 Confederate conscription laws applied to all men aged 17 to 50, meaning that a 13-year old in 1860 would have reached draft age in 1864 and a 46-year old in 1860 would have remained eligible four years later.

The 1860 United States Census for Harrison County reported 1,728 white men and boys between the ages of 13 and 46 inclusive, that is 1,728 males who were of military age sometime between 1861 and 1865. (See the first line of Table 1) Once this military aged population of white males was identified, I collected data on the age, state of birth, marital status, and occupation of each individual and on the wealthholding and slaveholding status of their families. Once I had this basic information from the census, the next step was to search the compiled service records of Confederate veterans by going case-by-case through the companies raised primarily in Harrison County, to determine which of these individuals served and what happened to them during their time in the military.6 Confederate service records, however, are notoriously incomplete, especially for the later years of the war, and many names required confirmation beyond location in the service records. So, other sources also were searched for evidence on military service. For example, by an act of the Texas legislature, muster rolls containing the names of all men aged 18 to 50 in each county in the state were created in March 1862. These rolls indicated the men who were already in the military at that time as well as those who were eligible to serve.7 Also by act of the legislature, county judges in 1864 and 1865 made lists of servicemen whose families, widows, or dependents were eligible for relief payments from the state. These lists of Confederate Indigent Families were a good source to determine who actually served.8 Confederate pension rolls, created after Texas began in 1899 to pay benefits to disabled and indigent veterans or their widows, contained records on the service of many Harrison County residents. 9 Finally, vertical files of family records in the Harrison County Historical Museum often provided direct evidence on service by family members.10

Even all this research did not complete the database because, as I already mentioned, sizable numbers of men who lived in Harrison in 1860 served in companies that were not recruited primarily in that county. Therefore, the service records of those units also were searched carefully and with considerable success. In total, 215 of the 1,728 men in the database were identified as members of companies raised primarily outside Harrison County.

Once the collection of data from the census and six sources on military service was completed, basic questions, beginning with “what percentage of Harrison County’s white men aged 13 to 46 in 1860 entered the Confederate Army or Texas State Troop units?” could be answered. Table I shows that almost half (49 percent) of those 1,728 men entered military service at some time between 1861 and 1865.11 This proportion—one of every two military-age men—seems notably high, but even at that it is somewhat low when compared to existing estimates of military service by men across the entire state. Most studies of Texas in the Civil War estimate the number who served at somewhere between a low of 58 percent and a high of 76 percent.12 So, unless Harrison County was far less supportive of the Confederacy than were other Texas counties—
and there is no reason to think that it was—information in my database suggests that existing estimates of the proportion of the state’s militaryage men who entered the service are too high. Only when the age group is limited to young men does the proportion who served rise to previously estimated levels. Table 1, which shows a break down of the militaryage population in terms of men who were in their teens, twenties, thirties, and forties in 1861, shows that only those who were in their twenties had a rate of service notably above 50 percent.

Turning to demographic characteristics, Table 2 reveals findings that, although not surprising, are worth confirming. First, an examination of

Table 1
Record of Military Service during the Civil War by Military-Age Men in Harrison County in 1860

Age and Number in 1860

Number and Percent who served

13-46                                 1,728

844             49%

13-18(Teens in 1861)            422                 

223             53%

19-28(Twenties in 1861)        654

387             59%

29-38(Thirties in 1861)          391

171             44%

39-46(Forties in 1861)           261

63               24%

Table 2

Demographic Characteristics of Soldiers and Non-Soldiers in Harrison County, 1860

  Medium Age Native of Lower South Native od Upper South Native of Free State or Foreign Born Married
Soldeiers 23 63% 31% 6% 34%
Non-Soldiers 28 54% 31% 15% 45%

age shows that the men who served were approximately five years younger than those who did not. Soldiers from the military-age group had a median age of twenty-three in 1860, whereas the median for nonsoldiers was twenty-eight. Second, a comparison of the birthplaces of those who served and those who did not also confirmed a pattern that might be expected. Among those who entered the military, 63 percent were natives of states in the lower South, 31 percent were from upper South states, and only 6 percent were from the free states and foreign nations. Among those who did not serve, 54 percent were from the lower South, 31 percent from the upper South, and 15 percent from the free states and foreign nations. Clearly, men born in the lower South felt the strongest commitment to the Confederacy, and those born in free states and foreign nations felt the least. Third, unmarried men were more likely to serve than those who were married. Only 34 percent of the soldiers were married, whereas 45 percent of the non-soliders had wives. To repeat, these findings are not surprising, but it is always worthwhile to confirm the expected.


Moving from demographic characteristics to economic information on the military-age white male population of Harrison County, brings us to the much-controverted “rich man’s war, poor man’s fight?” question.13 The first problem, of course, is the definition of “rich” and “poor.” One key in differentiating classes in antebellum Texas was slaveownership, so one definition of “rich” is “slaveowner,” Ownership of a single slave, however, did not make a person truly “rich,” so a combination of slaveownership and wealthholding in general is probably more revealing. I will consider both definitions.

Table 3 shows that men who were slaveholders or members of slaveholding families entered military service at a notably higher proportion than did nonslaveholders. Of 770 slaveholders, 438 (57 percent) served, as compared to 406 (42 percent) of 958 nonslaveholders. Thus, if slave ownership is used to define “rich,” it is clear that members of Harrison County’s wealthier class did not leave the fight to the poor.

A combination of slaveholding and wealthholding in general provides four “classes” whose participation in the war may be analyzed in some detail. The wealthiest class was composed of men from families that owned slaves and also had total wealth (a combination of real and personal property) of $16,000 or more, a number that placed those families above the mean wealthholding for slaveholders in East Texas. The second wealthiest class consisted of men from families that owned slaves and had a wealthholding of less than $16,000. Nonslaveholding families were divided into those with total wealthholdings of $1,500 or more (the mean wealthholding for nonslaveholding families in East Texas) and those with less than $1,500 in wealth.14 Table 3 shows the military service of men in these four classes. There is a regular progression in terms of military service from the wealthiest to the poorest class as 59 percent of the former served as compared to 40 percent of the latter. This table also indicates that men from the wealthiest class entered the war earlier than others. Those from slaveholding families worth $16,000 or more were the only group to have a larger percentage of those who served enter the war in 1861 rather than in 1862 (46 percent compared to 41 percent), whereas the majority in the three other classes joined in 1862. Please note also that regardless of the high percentages of soldiers from the wealthier classes, the largest group in absolute numbers (287) came from the poorest class. These statistics strongly support the conclusion that for residents of Harrison County the Civil War was a “rich man’s war” and a “rich man’s fight” and, when absolute numbers are considered, a “poor man’s fight” as well.

Table 3
Service According to Economic Class:Military-Age Men of Harrison County in 1860

Economic Class

Total Population

Number & Percent Who Served

Year Entered Confederate Service






438            57%

38%                              47%                           13%



406            42%

32%                              57%                            9%

Slaveholder  $16,000 or More


191            59%

46%                              41%                           11%

Slaveholder  Less than $16,000


247            56%

31%                              52%                           14%

Non-slaveholder  $1,500 or More


119            50%

25%                              62 %                          11%

Non-slaveholder  Less than $1,500


287            40%

35%                              56%                             8%


Table 4
Experience of Harrison County Soldiers During the Civil War

Types of Experience



Served Without Serious Illness or Wound



Killed/Fatally Wounded in Action 



Died of Disease



Died while Prisoner of War



Suffered Serious Illness/Survived*






Prisoner of War/Survived*



Discharged After Passage of Conscription Act



Received Disability Discharge









No Record of Military Experience Excerpt that Served



*A few individuals were counted twice in these categories because they both were wounded and prisoners or war, or they both suffered illnesses and were prisoners of war.

A final, and critically important, question about the demographic impact of the war on Harrison County’s white male population is: What happened to those who entered military service between 1861 and 1865? Table 4 shows that only about one-quarter (26 percent) can be documented as having served much of the war without suffering serious illnesses or wounds. At least one-half lost their lives or had experiences involving serious physical or psychological suffering. A little more than one-fifth (21 percent) died in the army—6 percent in battle, 11 percent from disease, and 4 percent in prisoner-of-war camps. Another 14 percent suffered serious illnesses or wounds but survived and continued to serve, and 13 percent spent time as prisoners of war, some on two separate occasions. Five percent received disability discharges, most of which resulted from illness. (So, 53 percent died or faced serious physical or psychological suffering.) Approximately 6 percent received discharges without disabilities, most of them for being over or under age at the time conscription began in April 1862. Only 2 percent were listed as deserters, although a higher percentage may have left the army in 1865 as defeat became imminent. (Service records that would reveal an increasing desertion rate are often unavailable for late 1864 and early 1865.) The military experiences of 13 percent fell into miscellaneous categories such as transferring to a non-Texas regiment or simply could not be documented.

In conclusion, then, several important suggestions about the impact of the Civil War on the white male population of Texas may be drawn from this quantitative study of Harrison County’s white men aged thirteen to forty-six in 1860. First, the percentage of military-age men in Texas who actually served may have been somewhat smaller than is generally believed, closer perhaps to one-half than to two-thirds. Second, those who served probably were younger and more likely to be from the Lower South than those who did not serve. Third, married men likely were a decided minority among soldiers. Fourth, wealthy slaveholders and their sons did not leave the fighting to men from poorer families. The war was not a “rich man’s war and poor man’s fight” in the sense that the rich did not fight. They did—and so did the poor. Fifth, among those who entered military service, two of every ten did not return, and nearly three in ten of those who survived suffered serious illnesses, wounds, or imprisonment. This casualty rate supports as especially important suggestion, I think, about the demographic impact of the Civil War on Texas. Consider this: If white males aged thirteen to forty-six across all of Texas in 1860 served in the Confederate Army at the same rate as did soldiers from Harrison County (49 percent) and died at the same rate (21 percent), the losses appear disastrously large—more than 12,000 men, most of them in their twenties and thirties. To appreciate the magnitude of such losses, think about this: If Texas had entered a war in 2001 with its population as reported in the 2000 census and had comparable percentages of its military-age males serve and die, total deaths would have been more than 540,000 men. It is not easy to contemplate the meaning of such a loss—in many ways—but almost certainly the deaths of so many young men would reduce productivity and growth for years to come. Moreover, do not forget that thousands of soldiers survived the war but came home weakened by wounds, disease, and imprisonment. Josiah Perry Alford, the young man whose photograph you saw when I began this presentation, suffered a thigh wound at the Battle of Chickamauga and had his left leg amputated on the field without benefit of anesthetic. He survived, came home, married in 1883, and lived productively, serving as the County Clerk of Harrison County for more than ten years. However, he died in 1897 at the relatively young age of 53, having, in the words of his widow “suffered from [the wound and amputation] as long as he lived.”14 So the war had a massive demographic impact on Texas in the form of both the terrible losses suffered during the fighting and the postwar problems faced by many veterans; yet, Texas grew rapidly after 1865. For example, the state’s population rose from 604,215 in 1860 to 1,591,749 in 1880.15 Cotton production rose from 431,463 bales in 1860 to 805,284 bales by 1880.16 Some 8,000 miles of railroad tracks were built in the state between 1870 and 1890. Texas’s post-war growth and productivity is probably explained in part by the likelihood that half of the its military-age white males did not serve and therefore did not suffer the often-ruinous impact of war, in part by the fact that a sizable minority of the state’s adult male population, the African-Americans who were slaves, suffered few wartime losses in the army, and in part by the fact that Texas had the good fortune to receive a huge influx of immigrants from the older southern states after 1865. Where that left much of the rest of the South?states such as Mississippi that probably had a higher percentage of their military-age white population serve, had at least as high a casualty rate among those who served, and then lost many of those who survived when they migrated to Texas—where that left states such as Mississippi is a matter that is not pleasant to contemplate. Look how far the investigation of a simple question “who fought for the Confederacy” can take you—all the way to wondering about the lingering demographic impact of the Civil War on the entire South!


1. Randolph B. Campbell, A Southern Community in Crisis: Harrison County, Texas, 1850–1880 (Austin: Texas State Historical Association, 1983), 24–27, 52,190-191; United States Bureau of the Census, Agriculture of the United States in 1860; Compiled from the Original Returns of the Eighth Census (Washington:Government Printing Office, 1864), 140–151, 240–242. (The cotton crop produced in San Augustine County in 1859 appears to have been the second largest in the state because it was reported incorrectly at 31,342 bales. The actual crop was 3,142 bales, far fewer than Harrison County produced.) Walter L. Buenger, Secession and the Union in Texas (Austin: University of Texas Press, 1984), 164, shows that it was not uncommon for Texas counties to support secession by 95 percent or more of the total vote cast.

2. Campbell, Southern Community in Crisis, 200.

3. Simpson, Marshall Guards, passim; Campbell, Southern Community in Crisis, 201, 210, 216–217; Sifakis, Compendium of the Confederate Armies, 106-108.

4. Campbell, Southern Community in Crisis, 201–203, 205–206, 210–211, 217; Sifakis, Compendium of the Confederate Armies, 47-48; Hale, The Third Texas Cavalry in the Civil War.

5. Sifakis, Compendium of the Confederate Armies, 70–72, 88–90, 128–130. Service by Harrison County men in these regiments raised primarily in other counties will be documented in detail below.

6. Index to the Compiled Service Records of Confederate Soldiers Who Served in Organizations from the State of Texas, Records of the Department of War, RG109, National Archives, Washington, DC; Compiled Service Records of Confederate Soldiers Who Served in Organizations from the State of Texas, Records of the Department of War, RG109, National Archives, DC. The search for Harrison County soliders was aided greatly by a privately printed source, ?Confederate Soldiers That Served in Units Formed in Harrison County, Texas,? the work of Jimmy and Patsy Oliphant of Shreveport, Louisiana. Also, Janet B. Hewett, editor, Texas Confederate Soldiers, 1861–1865 (2 vols., Wilmington, NC: Broadfoot Publishing, 1997), the most extensive listing of Civil War soldiers from Texas available, proved extremely useful in identifying men from Harrison County by unit.

7. H. P. N. Gammel, comp., The Laws of Texas, 1822–1897 (10 vols.; Austin: Gammel Book Co., 1898–1902), V: 455–465; Muster Roll of the Harrison County Regiment, 1862, Archives Division, Texas State Library, Austin.

8. Gammel, comp., Laws of Texas, V: 675–676; Confederate Indigent Families List, 1863-1865, Archives Division, Texas State Library, Austin.

9. H. P. N. Gammel, comp., The Laws of Texas, Supplementary Volume to the Original Ten Volumes, 1822–1897 (Austin: Gammel Book Company, 1902), 182-185; Confederate Pension Application Files, 1899–1975, Archives Division, Texas State Library, Austin.

10. Vertical Files on Harrison County Families, Harrison County Historical Museum Library, Old Courthouse, Marshall, Texas. Files with varying amounts of information exist for hundreds of families from the antbellum era.

11. The great majority of Harrison County soldiers served in what might be termed “regular” Confederate Army units, but some, mostly older men, enlisted in State Troop outfits that were organized in late 1863 and disbanded six months later without ever leaving Texas. Harrison County men were found in three companies of one of these units, the “Texas 1st Cavalry Battalion, State Troops.” It may be stretching the point to say that these men served in the Confederate Army, but in an effort to be as inclusive as possible, they were counted as such. Sifakis, Compendium of the Confederate Armies, 38.

12. Robert A., Calvert and Arnoldo De León, The History of Texas (2nd ed.,Wheeling, IL: Harlan Davidson, 1996), 134, gives the statistic 68,500 to 90,000. Stephen B. Oates, “Texas Under the Secessionists,” Southwest Historical Quarterly, 67 (October 1963), 187, set the number at 88,000. The largest number, 90,000, was given by Governor Francis R. Lubbock as the total of Texans in service in 1863. Ralph A. Wooster, Texas and Texans in the Civil War (Austin: Eakin Press, 1995), 32, 95. According to Wooster, this meant that “nearly all adult males” were in military service. The male population of Texas in 1860 is found in United States Bureau of the Census, Population of the United States in 1860; Compiled from the Original Returns of the Eighth Census (Washington: Government Printing Office, 1864), 474–477. (The census reported only aggregate totals for males aged 10 to 14 and 40–49. To determine the number aged 13–14, I assumed that each year of
age accounted for an equal part of the total and therefore took 40 percent—accounting for two of the five years—of the total. In the case of males aged 40-46, I made the same assumption and took 70 percent of the total reported for that age group.)

13. A recent study, David Williams, Rich Man?s War: Class, Caste, and Confederate Defeat in the Lower Chattahoochee Valley (Athens: University of Georgia Press, 1998), indicates the continuing interest in this question.

14. Context on wealthholding in East Texas is provided by Randolph B. Campbell and Richard G. Lowe, Wealth and Power in Antebellum Texas (College Station: Texas A&M University Press, 1977), 64 (Table 22).

15. Application of Mrs. Addie Alford, 1926, Confederate Pension Application Files.

16. Bureau of the Census, Population of the United States in 1860, 483; United States Bureau of the Census, The Statistics of the Population of the United States; Compiled from the Original Returns of the Ninth Census (June 1, 1870) (Washington: Government Printing Office, 1872), 63; United States Bureau of the Census, Statistics of the Population of the United States at the Tenth Census (June 1, 1880) (Washington: Government Printing Office, 1883), 443.

17. Bureau of the Census, Agriculture of the United States in 1860, 140–151; United States Bureau of the Census, The Statistics of the Wealth and Industry of the Untied States ... from the Original Returns of the Ninth Census, 1870 Washington: Government Printing Office, 1872), 252; United States Bureau of the Census, Report on the Productions of Agriculture as Returned at the Tenth Census (June 1, 1880) (Washington: Government Printing Office, 1883), 242.