Agriculture contraction and political response

The 1873 through 1896 agricultural and economic contractions were abrupt interruptions to Europe and North America’s Second Industrial Revolutions and the beginning of a prolonged British economic contraction. Various factors account for the crisis, and unjustified optimism in the emerging railroad industry is a leading explanation. Railroads were pivotal in the North’s Civil War victory, and after the Conflict, large-scale railroad construction encouraged by railroad land grants were associated with over-optimistic construction (Chandler, Reference Chandler1977; Gordan, Reference Gordan2015, pp. 132–142; Levy, Reference Levy2011). Railroad expansion continued with the 1873 Coinage Act (Kindelberger, Reference Kindelberger1996, p. 32), and when the Jay Cook & Company was unable to make payment on its Northern Pacific Railway debt, the financial crisis deepened, and the US economy entered crisis (Lamereaux, Reference Lamereaux1985). Despite its origin, the 1873 US agricultural contraction had various effects on rural farmers and workers in non-agricultural sectors.

For a generation after the Civil War, southern agriculture contracted and reflected the South’s deteriorating human capital and physical infrastructure (Woodward, Reference Woodward1951, pp. 175–204; Brinkley, Reference Brinkley1997). Deteriorating post-war economic conditions were associated with a period of political and economic populism, which led to the formation of various political movements. The Grange movement began in 1867 when President Andrew Johnson’s Agriculture Department’s secretary—Oliver Kelley—went to the South to investigate conditions to improve Southern economic and agricultural conditions (Woodward, Reference Woodward1951, pp. 32–34, 82–83; Chandler, Reference Chandler1977, p. 230; Cochrane, Reference Cochrane1979, pp. 95–97; Brands, Reference Brands2010, pp. 480–482). By 1873, the Grange coalesced behind the national Grange movement to promote railroad rate regulations that promoted agricultural interests. By 1877 in Munn vs. Illinois, the US Supreme Court ruled that grain houses were a private utility in the public interest and could be regulated under federal law, which the National Grange supported because it set a maximum price that railroads could charge in shipping rates (Cronon, Reference Cronon1991, pp. 138–142). The Grange movement was also social and went on to promote women’s suffrage, affect senate elections, and promoted temperance within agriculture.

The Greenback movement led by Ohio Democrat—George Pendleton—advocated that the US government continue the 1863 issuance of large Greenback debt to fund the North’s Civil War liquidity demands, which would have increased the money supply. Easy monetary policy redistributed purchasing power from large eastern banks to small western farmers (Kindahl, Reference Kindahl, Fogel and Engerman1971, pp. 469–470; Woodward, Reference Woodward1951, pp. 81–85; Brands, Reference Brands2010, pp. 482–483). By 1873, the US public was polarized over the appropriate currency, and farmers appealed to Congress for the widespread issuance of Greenbacks with unlimited silver coinage, which inflated the currency and allowed farmers to repay their mortgages with depreciated currency. As a reaction to the 1873 Coinage Act, the 1878 Bland-Allison Act reduced specie and the money supply, which required the U.S. Treasury to purchase and circulate silver dollars that traded simultaneously with gold, creating a bimetallic currency. Although Pendleton’s Plan remained popular among debtors—such as farmers—it was not adopted, and the Greenback movement failed because it lacked the political support and patronage shared by Democrats and Republicans.

The Populist Party was a third late 19^th century political movement that began among farmer alliances that also supported free and unlimited silver coinage. The Populists influenced 1890 local and state elections to put James B. Weaver in office but disintegrated in the early 20^th century (Woodward, Reference Woodward1951, pp. 242–263; Cronon, Reference Cronon1991, pp. 360–365; Brands, Reference Brands2010, pp. 491–506; Levy, Reference Levy2011). Subsequently, the Grange, Greenback, and Populist movements are three political movements related to agriculture at the end of the 19^th and early 20^th centuries, whose policies sought to change the relative bargaining power between agriculture and commercial interests that were designed to increase agricultural wealth and improve farmer’s living conditions. To the degree these economic and political events affected agriculture and net nutrition, farmer BMIs, height, and weight may be affected differently by race and gender between the agricultural and non-agricultural sectors. Subsequently, this study partitions individuals in the agricultural and non-agricultural sectors, genders, and race to evaluate net nutritional variation by socioeconomic status, gender, and race at the end of the 19^th century (Schneider, Reference Schneider2023, p. 12).

Margo and Steckel (Reference Margo and Steckel1983) first reported a white US male antebellum stature diminution during the 19^th century’s 2^nd and 3^rd quarters, which called into question the prevailing view that early US industrialization created broad-based economic growth (Komlos, Reference Komlos1998, p. 779). Komlos (Reference Komlos1987) also finds that white statures decreased during the 19^th century’s second and third quarters, a pattern known as the antebellum paradox. Various studies confirm the result (Craig, Reference Craig, Komlos and Kelly2016; Fogel, Reference Fogel, Engerman and Gallman1986, pp. 462–463; Fogel, Reference Fogel2000, pp. 139–142); however, the proposition does not account for women and non-white populations or minorities (Schneider, Reference Schneider2023, p. 12). Steckel (Reference Steckel2000) and Coelho and McQuire (Reference Coelho and McGuire2000) debate the relative merits vs. disease to explain the decline.

A considerable literature demonstrates that height is inversely related to urbanization, and the US urbanized during the 19^th century. Despite urbanization’s harmful effects, 19^th century households migrated to and remained in urban areas because urban areas’ net benefits remained positive. Carson (Reference Carson2008, pp. 366–368), Zehetmayer (Reference Zehetmayer2011), and Zehetmayer (Reference Zehetmayer2013) show that 19^th century urban statures were short compared to rural statures. This urban-stature relationship was noticed early (Fogel et al., Reference Fogel, Engerman, Floud, Steckel, Trussell, Wachter and Villaflor1979; Sokoloff and Villaflor, Reference Sokoloff and Villaflor1982), and multiple studies show a negative net urban effect (Margo and Steckel, Reference Margo and Steckel1983; Steckel and Haurin, Reference Steckel, Haurin and Komlos1994). Urban external effects were adversely affected by disease and higher relative food prices. These urban agglomeration effects may have been related to race. Higgs (Reference Higgs1977, pp. 33–35) indicates that urban African-American’s net nutrition may have been better because of more progressive urban institutions, better medical care, and urban areas may have allowed blacks greater consumption and health investments than rural areas when rural blacks were exposed to greater rural isolation that increased the likelihood of white-on-black violence. Nonetheless, urban locations provided positive effects from higher incomes and wealth that allowed some to benefit, yet the overall effect was negative. Subsequently, a considerable part of the antebellum paradox by occupation may be related to 19^th century urbanization experienced differently by race and gender.

Data

Height and weight data used in this study are part of an extensive effort to collect physical descriptions using 19^th and early 20^th century US prison records. Military and prison records are two sources used to study net nutritional conditions, and military records were an early source for stature studies (Fogel et al., Reference Fogel, Engerman, Trussell, Floud, Pope and Wimmer1978; Fogel et al., Reference Fogel, Engerman, Floud, Steckel, Trussell, Wachter and Villaflor1979). However, military records over-represent individuals classified as white, and underrepresent females and non-Europeans. Military records were also drawn from males of European ancestry, whereas prison records include women and various ethnic groups (Schneider, Reference Schneider2023, p. 12). In addition, military enlistment standards may have varied with conscription needs that may have been related during active military periods, and early 19^th century military needs may have sampled individuals in higher socioeconomic groups. Prison records complement military records to augment these military record shortcomings. For example, prisons include females and minorities, creating a more diverse sample. Prison records are not, however, above scrutiny and may disproportionately include individuals from lower socio-economic groups who turned to crime for survival. Because physical measures within prisons were used to identify individuals and in case they escaped and were recaptured, prison records are valuable and reliable sources to measure late 19^th and early 20^th century US net nutrition.

Each state prison was contacted on multiple occasions, and available and affordable prison records were entered into a master data set. State prisons used in this study are Arizona, Colorado, Idaho, Illinois, Kentucky, Mississippi, Missouri, Montana, Nebraska, New Mexico, Oregon, Pennsylvania’s East and West Prisons, Philadelphia, Tennessee, Texas, Utah, and Washington. Physical descriptions and characteristics were recorded at the time of entry, subsequently, represent pre-incarceration conditions. Accurate physical descriptions were important because they had legal implications in case inmates escaped and were recaptured.

Race and gender are two characteristics that help identify individuals within prisons. Prisoners of African and European ancestry were the two most prominent racial groups, and individuals of African ancestry were recorded as negro, light, medium, and dark black. Individuals with European ancestry were recorded as light, medium, and dark. This European classification system is further supported because individuals claiming European birth were also recorded with the same light, medium, and dark classifications. Individuals of combined African and European ancestry were recorded as ‘mulattos,’ however, are described as ‘mixed-race’ in the results that follow. The Arizona and Montana prisons were the only institutions that, for at least a time, included both photographs and written complexion descriptions, and it is clear from these photographs that individuals reporting African and European ancestry are consistent with complexion descriptions used by enumerators to classify blacks and whites, where black refers to black-African and white refers to white European. There were also individuals with Mexican, Asian, and American Indian complexions in the sample. Gender was recorded as male and female; however, US state prisons did not consistently record women’s pregnancy status.

There are international and domestic nativities within prison records that reflect migration flows that drew immigrants to the United States (Ferrie, Reference Ferrie1999). International migrants are from Africa, Asia, Australia, Europe, Great Britain, Latin America, and Mexico. Domestic nativities are separated into Northeast, Middle Atlantic, Great Lakes, Plains, Southeast, Southwest, and Far West nativities (Carlino and Sill (Reference Carlino and Sill2001). Northeast nativity includes Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont. Middle Atlantic nativity includes Washington DC, Delaware, Maryland, New Jersey, New York, and Pennsylvania. The Great Lakes includes Illinois, Indiana, Michigan, Ohio, and Wisconsin. Plains nativity includes Iowa, Kansas, Minnesota, Missouri, Nebraska, North Dakota, and South Dakota. Southeast nativity includes Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Virginia, and West Virginia. Southwest nativity includes Arizona, New Mexico, Oklahoma, and Texas. Far West nativity includes California, Colorado, Idaho, Montana, Nevada, Oregon, Utah, Washington, and Wyoming.

Occupations are the primary means of classifying socioeconomic status, which varied by gender. The most common female occupations were domestic laborers, such as household domestic labor and household servants. Women found some opportunities in skilled labor. However, their occupations—such as midwives, nurses, and tailoresses—served other women (Goldin, Reference Goldin1990; Burnette, Reference Burnette, Whaples and Parker2013, pp. 306–307). Enumerators also recorded pre-incarceration occupations and are classified here into five separate occupation groups. White-collar workers are bankers, administrators, and physicians. Skilled workers are blacksmiths, carpenters, and craftsmen. Farmers are farmers, ranchers, and dairymen. Unskilled workers are cooks, miners, and laborers. There are also workers with no listed occupation or are not decipherable, which are classified with no occupations.

Prison samples are younger than the general population (Gottfredson and Hirschi, Reference Gottfredson and Hirschi1990; Patterson, Reference Patterson2005, p. 43). In both historical and contemporary populations, crime is committed by the young, and 95 percent of the prison population consisted of individuals younger than age 50 (Table 1). Whites within prisons were the most common racial group, and individuals of African and mixed race are the second largest population within prisons. Blacks within prisons are a larger proportion of the prison population relative to the general population (Haines, Reference Haines, Engerman and Gallman2000; Steckel, Reference Steckel, Haines and Steckel2000). The South is the most common residence within the sample, followed by the Middle Atlantic and Plains. While populations are concentrated in the South, Northeast, and Middle Atlantic, eight of the 18 prison facilities are in the West, and the West constitutes the largest geographic region for unskilled workers, and unskilled workers are the most prominent occupation group. Farmers within prisons are a smaller occupation group compared to the general population (Rosenbloom, Reference Rosenbloom2002, p. 88; Church et al., Reference Church, Thomas, Tudor-Locke, Katzmarzyk, Earnest, Rodarte, Martin, Blair and Bouchard2011; Gordan, Reference Gordan2015, pp. 53, 254–258). Most individuals were born in the 1880s and received in the 1910s.

Table 1.

Farm and Non-Farm Late 19^th and Early 20^th Century Characteristics

Source: Arizona State Library, Archives and Public Records, 1700 W. Washington, Phoenix, AZ 85007; Colorado State Archives, 1313 Sherman Street, Room 120, Denver, CO 80203; California State Archives, 1020 O Street, Sacramento, CA 954814; Idaho State Archives, 2205 Old Penitentiary Road, Boise, Idaho 83712; Illinois State Archives, Margaret Cross Norton Building, Capital Complex, Springfield, IL 62756; Kentucky Department for Libraries and Archives, 300 Coffee Tree Road, Frankfort, KY 40602; Maryland State Archives, 350 Rowe Building, Annapolis, MD 21401; Missouri State Archives, 600 West Main Street, Jefferson City, MO 65102; William F. Winter Archives and History Building, 200 North St., Jackson, MS 39201; Montana State Archives, 225 North Roberts, Helena, MT, 59620; Nebraska State Historical Society, 1500 R Street, Lincoln, Nebraska, 68501; New Mexico State Records and Archives, 1205 Camino Carlos Rey, Santa Fe, NM 87507; Ohio Archives Library, 800 E. 17^th Avenue, Columbus, OH43211; Oregon State Archives, 800 Summer Street, Salem, OR 97310; Pennsylvania Historical and Museum Commission, 350 North Street, Harrisburg, PA 17120; Philadelphia City Archives, 3101 Market Street, Philadelphia, PA 19104; Tennessee State Library and Archives, 403 7^th Avenue North, Nashville, TN 37243 and Texas State Library and Archives Commission, 1201 Brazos St., Austin TX 78701; Utah State Archives, 346 South Rio Grande Street, Salt Lake City, UT 84101; Washington State Archives, 1129 Washington Street Southeast, Olympia, WA 98504.

Because there were few historical institutions and practices that collected height and weight under controlled randomized conditions, cross-sectional data reflect the purposes they were collected, which is common in historical height and weight studies. These selection concerns were recognized from height study’s beginning, and it is difficult to identify any sample that reflects the true “general population’s” average height and net nutrition (Fogel et al., Reference Fogel, Engerman, Trussell, Floud, Pope and Wimmer1978; Sokoloff and Villaflor, Reference Sokoloff and Villaflor1982; Fogel, Reference Fogel, Engerman and Gallman1986). US black inmates were taller than blacks in slave manifests. White prisoners were mostly shorter than US Civil War recruits, but it is difficult to identify the reason for the difference. While younger soldiers and inmates were in their early 20s, there were proportionately more older inmates than soldiers in their later ages, who were shorter as their heights diminished with age (Huang et al., Reference Huang, Lei, Ridder, Strauss and Zhao2013). Civil War recruits also had proportionately more rural farm workers, who were taller and heavier because of their close proximity to rural net nutrition. White male prison statures were comparable to Union Army’ Midwest and North Central nativities, and white passport applicants were from higher socioeconomic groups who benefited from better diets and net nutrition and did not experience the dire disease effects as the general population. So, while selection is always a concern, it was recognized early, and US prisoner heights and net nutrition are comparable to similar US military unskilled workers.

Comparative net nutritional conditions by gender and race

We now consider late 19^th and early 20^th century net nutrition variation by socioeconomic status, gender, and race. To evaluate late 19^th and early 20^th century current and cumulative net nutrition, body mass, height, and weight are regressed on demographic, socioeconomic, nativity, and geographic characteristics.

Body mass index

(1) $${BM{I_i} = \alpha + {\theta _c}Centimeter{s_i} + \sum\limits_{r = 1}^5 {{\theta _r}Complexio{n_i} + \sum\limits_{a = 14}^{80s} {{\theta _a}Ag{e_i} + \sum\limits_{n = 1}^{14} {{\theta _n}Nativit{y_i}} } }}$$

$${+ \sum\limits_{l = 1}^{17} {{\theta _l}{\mathop{\rm Re}\nolimits} sidenc{e_i}} + \sum\limits_{t = 1840}^{20} {{\theta _t}Obervation\,Yea{r_i} + {\theta _u}Urbanizatio{n_i} + \,{\varepsilon _i}}}$$

Centimeters

(2) $${Centimeter{s_i} = \alpha + \sum\limits_{r = 1}^5 {{\theta _r}Complexio{n_i} + \sum\limits_{a = 14}^{80s} {{\theta _a}Ag{e_i} + \sum\limits_{n = 1}^{14} {{\theta _n}Nativit{y_i}} } }}$$

$${ + \sum\limits_{l = 1}^{17} {{\theta _l}{\mathop{\rm Re}\nolimits} sidenc{e_i}} + \sum\limits_{t = 1840}^{20} {{\theta _t}Birth\,Yea{r_i} + {\theta _u}Urbanizatio{n_i} + {\varepsilon _i}}}$$

Kilograms

(3) $${Ki\log ra{m_i} = \alpha + {\theta _c}Centimeter{s_i} + \sum\limits_{r = 1}^5 {{\theta _r}Complexio{n_i} + \sum\limits_{a = 14}^{80s} {{\theta _a}Ag{e_i} + \sum\limits_{n = 1}^{14} {{\theta _n}Nativit{y_i}} } }}$$

$${ + \sum\limits_{l = 1}^{17} {{\theta _l}{\mathop{\rm Re}\nolimits} sidenc{e_i}} + \sum\limits_{t = 1840}^{20} {{\theta _t}Obervation\,Yea{r_i} + {\theta _u}Urbanizatio{n_i} + {\varepsilon _i}}}$$

Height in centimeters is included in BMI models to account for the inverse relationship between BMI and weight models to account for the positive relationship between weight and height (Carson, Reference Carson2009a; Carson, Reference Carson2012; Carson, Reference Carson2015a; Komlos and Carson, Reference Komlos and Carson2017). Black, mixed-race, Mexican, Asian, and Native American dummy variables are included for complexions to determine net nutrition variation by race. Annual youth age dummy variables are included for early stature growth, while adult decade age dummy variables are included to account for net nutrition variation at older ages. International nativity dummy variables are included for Africa, Asia, Australia, Great Britain, Canada, Europe, Latin America, and Mexico nativities. There are two ways to interpret BMI, height, and weight variation over time. Measured in the current period, BMIs and weight reflect current net nutrition by diverse cohorts at the time of measurement. Measured by birth year, stature reflects a cohort’s cumulative net nutrition variation since birth (Carson, Reference Carson2019, p. 32). Subsequently, birth decade dummy variables are included in height regressions, and observation period dummy variables are included in BMI and weight models.

Three paths of inquiry are considered when evaluating late 19^th and early 20^th century body mass, height, and weight by social class. First, the antebellum paradox is the pattern where white US male average statures stagnated during the 19^th century’s second and third quarters (Margo and Steckel, Reference Margo and Steckel1983; Komlos Reference Komlos1987; Craig, Reference Craig, Komlos and Kelly2016). However, little is known regarding the antebellum paradox for non-whites, women, and African-Americans (Schneider, Reference Schneider2023, p. 23). To the degree farmer’s net nutrition was affected, their body mass, stature, and weight should have decreased compared to workers in non-agricultural occupations between 1870 and 1900 because their living standards decreased relative to non-farmers. However, farmer BMIs and weight increased between 1870 and 1900, and their height was little different than the mid-1870s (Table 3; Figs. 1–3). Assessing trend stature variation over time is with bubble figures, where the circle sample size is decade proportion to the total sample size. Time coefficients are weighted by each decade’s proportion in sample space. The same method is applied to remaining figures in this study. Moreover, farmer’s net nutrition improved relative to non-farmers, and non-farmers height was significantly lower between 1870 and 1900. Before and after the War, farmers and agricultural workers were consistently taller than non-farmers, with greater body mass and heavier weights (Gordan, Reference Gordan2015), indicating that despite political hyperbole, the Grangers, Greenbacks, and Populist movements had little effect on lower socioeconomic status net nutritional conditions.

Figure 1.

Body Mass Index Variation over time by Farmers vs. Non-Farmers. Source: See Tables 1 and 3.

Notes: Coefficients weighted by sample size within each decade. Hollow circles weight each decade to the sample

Figure 2.

Height Variation over time by Farmers vs. Non-Farmers. Source: See Tables 1 and 2.

Notes: Coefficients weighted by sample size within each decade. Hollow circles weight each decade to the sample

Table 2.

US Prison Height Comparison to Existing Literature

Sources: Steckel (Reference Steckel1979, Weighted Average for Ages 25 through 49, Table 2, p. 368); Margo and Steckel (Reference Margo and Steckel1983, Table 1, pp. 169–170); Komlos (Reference Komlos1987, Table 2, p. 901); Haines, Craig, and Weis, Reference Haines, Craig and Weiss2003, Table 5, p. 400; Zehetmayer (Reference Zehetmayer2011, Table 2 pp. 318–319); Sunder (Reference Sunder2013, Table 2, p. 251); Carson, height in present study.

Table 3.

Late 19^th and Early 20^th Century Farm and non-Farm Body Mass, Height, and Weight

Source: See Table 1.

Notes: *** Significant at .01; **Significant at .05; * significant at .10.

Figure 3.

Weight Variation over time by Farmers vs. Non-Farmers.

Source: See Tables 1 and 2. Notes: Coefficients weighted by sample size within each decade. Hollow circles weight each decade to the sample

While individual time coefficients reflect net nutrition over time, they do not, collectively measure birth and observation effects over time. Time series F-tests between unrestricted and restricted models indicate a measurable association between net nutrition and its variation over time. A joint test on farmer’s BMs with time variables is F(17, 21,587) = 3.21, P = .000. The non-farm joint BMI test is F(20,180,068) = 29.78, P = .000. The farmers’ joint stature time test is F(23, 21,582) = 3.98, P = .000, while non-farmers’ joint stature test is F(27, 180,062) = 22.20, P = .000. Farmer’s joint weight-time test is F(17,21,587) = 3.21, P = .000. Non-farmer’s joint time test is 29.71, P = .000, indicating that farm and non-farm net nutrition varied over time individually and collectively, and farmers did better than non-farmers during the post-bellum period when Grangers, Greenbacks, and Populists advocated pro-agricultural policies. Subsequently, farmer statures and cumulative net nutrition improved after 1875, BMI and weight increased with the 1873 and 1893 contractions, and the difference in farm minus non-farm net nutrition favored rural agricultural conditions (Figs. 1–3).

Second, net nutritional conditions also varied by gender, both within the household and within the economy (Oren, Reference Oren1973; Carson, Reference Carson2018; Carson, Reference Carson2022a). Household resources are shared resources (Oren, Reference Oren1973, pp. 107 and 110), and household income and wealth mask individual net nutritional variation within the household—particularly for mothers—who suppress their personal consumption during periods of dietary stress to reallocate nutrition to children, creating material and net nutritional inequality within the household. Female average BMIs were high in the early 1870s and decreased considerably in the late 1870s to remain constant until the early 20^th century (Table 4). Figure 4 indicates that male body mass index values remained constant around 30 throughout the late 19^th and early 20^th centuries. The female-male BMI difference followed the 1870 female body mass decrease and remained lower until the early 19^th century. Gender-related statures were less plastic than body mass and weight values and remained constant (Figs. 4–6), and male weights decreased relative to female weights, which remained high until the early 20^th century (Fig. 6). Subsequently, between 1860 and 1890, female net nutrition improved relative to men, and current male net nutrition varied less over the post-bellum period then females.

Table 4.

Late 19^th and Early 20^th Century Body Mass, Height, and Weight by Gender

Source: See Table 1.

Notes: *** Significant at .01; **Significant at .05; * significant at .10.

Figure 4.

Body Mass Index Variation over time by Gender. Source: See Tables 1 and 3.

Notes: Coefficients weighted by sample size within each decade. Hollow circles weight each decade to the sample

Figure 5.

Height Variation over time by Gender. Source: See Tables 1 and 3.

Notes: Coefficients weighted by sample size within each decade. Hollow circles weight each decade to the sample

Figure 6.

Weight Variation over time by Gender. Source: See Tables 1 and 3.

Notes: Coefficients weighted by sample size within each decade. Hollow circles weight each decade to the sample

Third, Steckel (Reference Steckel1979) was the first to show that fairer-complexioned individuals with European ancestry were consistently taller than darker-complexioned individuals with African ancestry. Bodenhorn (Reference Bodenhorn1999) finds that 19^th century whites and mixed-race individuals were taller than blacks and suggests the stature difference is due to social preferences that disproportionately favored individuals with fairer complexions (Bodenhorn, Reference Bodenhorn1999, pp. 983 and 994; Bodenhorn, Reference Bodenhorn2002, pp. 21 and 43–44). However, if taller urban mixed-race net nutrition persisted because of social preferences, white weights, and body mass should have been greater than darker complexioned blacks. In fact, the opposite is true, and black and mixed-race individuals had greater weights and higher BMIs than their white counterparts. Johnson (Reference Johnson1941, pp. 256–257) and Fogel and Engerman (Reference Fogel and Engerman1974, p. 132) show that mixed-race individuals were more common in urban locations. Net nutrition by birth and residence illustrate that northeastern blacks were shorter (Carson, Reference Carson2008; Carson, Reference Carson2009b), and Higgs (Reference Higgs1977, pp. 33–35) indicates that rural black net nutrition may have been lower if rural Jim Crow policies and racial intimidation prevailed in rural locations. Because there are urban-racial agglomeration effects, greater urban mixed-race populations may have created better urban black and mixed-race net nutritional conditions, and part of the BMI, height, and weight differences by race may have biological origins (Carson, Reference Carson2015a; Carson, Reference Carson2015b).

Other patterns are consistent with expectations. Greater access to regional nutrition was associated with taller statures and heavier weights. For example, net nutrition varied regionally, and the South was agriculturally more productive in corn, pork, and beef than other regions. In 1860, average Southern corn production was 34.03 bushels of corn per person compared to the North’s 9.25 bushels per person. The South’s average pork production was 1.27 swine per capita per annum, compared to the North’s .65 swine per capita (Hilliard, Reference Hilliard1972, Tables 3, and 6). The South produced 3.16 times as much corn per capita as the North, 96 percent more cattle, and 90.58 percent more pork than the North. Cattle and dairy are compliments in production; however, because of temperature differentials and poor dairy storage in the South, individuals in the South consumed less milk, and milk is related to stature growth (Baten and Murray, Reference Baten and Murray2000, pp. 361, 364–367; Wiley, Reference Wiley2005). Subsequently, Southern net nutrition was higher than the North, and Northeastern and Middle Atlantic net nutrition was lower than elsewhere within the US.

Table 5.

Farm, Non-Farm Body Mass, Height, and Weight Decomposition by Agricultural Status

Source: See Tables 1–4.

Table 6.

Male-Female Body Mass, Height, and Weight Decompositions

Source: See Tables 1–4.

Urban–rural BMI, height, and weight decompositions by occupation and gender

Characteristic coefficients illustrate individual net nutrition variation. They do not, however, indicate collective net nutrition variation between gender and race for collective returns by characteristics. Oaxaca-Binder decompositions are a statistical technique that partitions dependent variable differences into returns to characteristics and mean return characteristics.

(4) $${{\gamma _h} = {\theta _{oh}} + {\theta _{1h}}{X_h}}$$

(5) $${{\gamma _l} = {\theta _{ol}} + {\theta _{1l}}{X_l}}$$

To isolate 19^th and early 20^th century net nutrition by combined characteristics, let γ_h and γ_l be high and low individual’s BMI, height, and weight returns by demographic, socioeconomic status, and residential characteristics.

(6) $${\Delta \gamma = {\gamma _h} - {\gamma _l} = {\alpha _h} + {\beta _h}{X_h} - {\alpha _l} - {\beta _l}{X_l}}$$

High and low response variable gaps separate net nutritional conditions into structural and compositional characteristics, and structural differences are explained by differences across characteristics, while composition effects illustrate net nutrition variation with average characteristics.

Decompositions partition dependent variable differences into returns to characteristics and average characteristics. Adding $ - {\beta _h}{X_l} + {\beta _h}{X_l}$ to Equation 6 is high returns to characteristics observed at low characteristics decomposition, and adding $ - {\beta _l}{X_h} + {\beta _l}{X_h}$ to Equation 6 is a low returns to characteristics at high returns to average characteristics.

(7) $${{\gamma _h} - {\gamma _l} = \left( {{\alpha _h} - {\alpha _h}} \right) + \left( {{\beta _h} - {\beta _l}} \right){X_l} + \left( {{X_h} - {X_l}} \right){\beta _h}}$$

(8) $${{\gamma _h} - {\gamma _l} = \left( {{\alpha _h} - {\alpha _h}} \right) + \left( {{\beta _h} - {\beta _l}} \right){X_h} + \left( {{X_h} - {X_l}} \right){\beta _l}}$$

Equations 7 and 8 first right-hand side components are autonomous net nutrition values independent of returns or average characteristics. The second component is the share of dependent net nutritional structural differences due to returns to characteristics. The third component is the dependent net nutritional difference share due to returns to average compositional characteristics. Equation 7 is the dependent variable differences observed at low average characteristics and high returns to characteristics. Equation 8 is the dependent variable differences at high average characteristics and low returns to characteristics.

Table 4 partitions farm and non-farm BMIs, stature, and weight into structural and composition differences by height, demographic, and urban status. Overall, non-farmer BMI and weight returns to height were greater than farmers, indicating that non-farmers, who had short statures, had greater returns to current net nutrition from cumulative net nutrition. Non-farmer BMI returns to residence, ages, and nativity were greater than farmers. For BMI, height, and weight, returns to characteristics were greater than returns to average characteristics.

Table 5 partitions male and female BMIs, stature, and weight into structural and composition differences by height, demographic, and urban status. Panels A through C are segregated into BMI, height, and weight decompositions. Autonomous BMI component differences were nearly offset by females’ greater rate of return to stature (Table 4, Panel A). Women also had higher BMI returns associated with residence, age, and nativity. Male BMI returns were higher for blacks and observation period, and males were consistently taller than females with nativity, residence, observation period, and ages. Males had greater weights associated with returns to complexion and observation periods that were offset by female’s weight composition. Females had greater weight returns associated with height and nativity.