At the annual John Whitmer Historical Association meeting in September, Craig Foster announced that he and Newell Bringhurst will be editing an anthology on polygamy. Two of the 15 or so essays will take opposing views on whether teen marriage was normal in the 19th century. Squaring off will be an extended version of Todd Compton’s Sunstone West presentation and a paper co-authored by Craig Foster, Greg Smith, and myself. My role is to be the stat man, while Greg is an expert on Nauvoo plural marriage, and Craig is a accomplished historian and has mastered the literature on marriage trends. Craig and Greg are more prolific authors than myself and I summarized some of their work at the height of the Romney campaign here on the FAIR blog. I think this gave an early picture of what might happen if the three of us combined skills.
There will be somewhat of a wait before our book chapter will be published and I feel for those church members and investigators who are exposed to presentist anti-Mormon propaganda and don’t know where to look for accurate information. While I don’t like giving attention to such sites, sometimes they do more damage if left unchecked. I do not want to start a debate with the sponsors of such sites and forums. How they react to my constructive criticism is up to them. They can ignore me and leave their articles in their poorly researched and analyzed state, or they can attack me by labeling me as a pedophile, or they can work to improve their content. I would actually encourage them to do the latter, though I do not feel obligated to do their homework for them.
One article about 19th century marriage statistics and Joseph Smith’s plural marriages that is widely used by critics is found at i4m.com. It is poorly researched, but provides a. reference section that gives to the unwary that it is well documented. The few sources appear to have been cobbled together by some anti-Mormon message board poster and backs up almost nothing in the article’s main text.
Where can one find accurate information on 19th century marital trends? I recommend the following:
J. David Hacker. ”Rethinking the ‘Early’ Decline of Marital Fertility in the United States” Demography Volume 40, Number 4, November 2003, pp. 605-620
Michael R. Haines. “Long-term marriage patterns in the United States from colonial times to the present” History of the Family 1996, Volume 1, Number 1, pp. 15–39
Catherine Fitch and Steven Ruggles. 2000. “Historical Trends in Marriage Formation.” In Linda Waite, Christine Bachrach, Michelle Hindin, Elizabeth Thomson, and Arland Thornton, eds., Ties that Bind: Perspectives on Marriage and Cohabitation. Hawthorne: Aldine de Gruyter, pp. 59-88. http://tinyurl.com/yhc9sc5
The Foster, Keller, Smith paper will extend the methods for data collecting, analyzing, and reporting found in these works. One important extension is that we will look at marriage cohorts in addition to birth cohorts, the latter being more typical in the demographics literature. A marriage cohort is a set of females married in a year’s time frame, which can then be broken down by age. A birth cohort is a set of females that were born the same year, which can then be broken down by the age at marriage (though it was typical for 7-8% of a birth cohort that reached 50 to have never married).
Marriage cohorts had lower average marital age than birth cohorts because a larger amount of the total population resides at the younger ages. Visualize a demographic pyramid with a wide base. Marriage cohorts therefore have a higher percentage of teens than birth cohorts. These breakdowns more closely approximate marriage market dynamics and provide a better comparison to the age profile of Joseph Smith’s wives.
In demographic literature, the two types of statistical breakdowns are called age-independent (for birth cohorts) and age dependent (for marriage cohorts). When data is collected from US Census it is most usually reported age-independently. When county records are summarized age-dependently. When good records about both birth and marriage are available (genealogical databases) generally age-independent reporting is preferred. So age-independent breakdowns dominate demographics literature.
It is possible that my method that I developed to convert age-independent Census data to age-dependent data will be of interest to the wider academic community (outside of Mormonism) . I won’t go into details here, but main tools for the job are the widely used Coale-McNeil model and a generalized extension presented in this article:
Ryuichi Kaneko. “Elaboration of the Coale-McNeil Nuptiality Model as The Generalized Log Gamma Distribution: A New Identity and Empirical Enhancements” Demographic Research 2003 Volume 9 pp. 223-262 http://www.demographic-research.org/volumes/vol9/10/9-10.pdf
The Coale-McNeil model is also used in Haines (1996) article above. With the appropriate assumptions and extensions the Coale-McNeil model is extremely useful for estimating marriage trends for 1800-1840 and for providing a age breakdown if the mean marital age is known. Furthermore it makes it easier to adjust for age population dynamics (due to high birth and death rates) that make the actual marital age lower than the age-independent census data does.
Now with the backing of this recent scholarship let me dissect some of i4m’s accusations.
i4m: “There is no documentation to support the idea that marriage at fourteen was ‘approaching eligibility.’”
This should probably be changed to i4m has not found any documentation. Yet such documentation is easily found. Coale’s model has established a statistical benchmark for the minimum age of eligibility (which turns out to be the mean minus 1.73 times the standard deviation). Haines (1996) above tabulates estimates from Sanderson for years 1800-1920 for the minimum age and it ranges from 14.0-14.3. I get a little lower using better data and a maximum likelihood estimation technique (~13.5 in 1800).
i4m “Actually, marriages even two years later, at the age of sixteen, occurred occasionally but infrequently in Helen Mar’s culture. Thus, girls marrying at fourteen, even fifteen, were very much out of the ordinary. Sixteen was comparatively rare, but not unheard of.”
Here i4m is describing marriage at 14, 15, and 16 in Helen Mar Kimball’s culture, but have not quantified anything. Suppose we take HMK’s culture to be 1840 white America (more refinements could be made based on region married in, region born in, region parents born in, rural or urban, etc.). Then based on my latest estimates if one were to attend 100 random weddings involving first time brides in 1840: 1.9 brides would be 14 or younger, 3.8 would be 15, and 6.7 would be 16.
i4m:”American women began to marry in their late teens;”
This statement is false. From above 12.3% married before turning 17. By Coale’s benchmark, which he described as the “the earliest age of a significant number of first marriages,” and noted that this age “depends on age laws, the onset of menarche, and traditional community standards. As noted earlier, empirically this age was around 13.5 or 14.0 in the mid 19th century.
i4m: “around different parts of the United States the average age of marriage varied from nineteen to twenty-three.”
This might be the most defensible number or characterization that i4m makes. For 1880, Hacker (2003) writes that “SMAM [an age-independent measure of the mean] ranged from 21.4 in the Mountain and Pacific regions to 24.6 in New England.” He calculated the national mean to be 23.3. Nationwide a reasonable compromise presented in Haines (1996) is that the marital age was 21.0, but as i4m notes Sanderson estimated 19.0. I use the conservative compromise in my estimates but 19.0 sounds about right for some of the more frontier regions of the country.
However i4m misuses the mean in their analysis. In the Coale-McNeil model 60% of a birth cohort have been married by the mean age. This is standard regardless of census year data used. After adjusting for population increase the mean age drops in 1880 to 22.4 and 38% of first time brides were teenagers. Note that I am switching to age-dependent statistics. For 1840, I estimate teenage brides in 43% of all [first time] weddings. The most frequent marrying age was 18. This is a far cry from marriage beginning to occur in the late teens as i4m portrays.
i4m: “In the United States the average age of menarche (first menstruation) dropped from 16.5 in 1840 to 12.9 in 1950.”
No papers are cited to support this claim. One should take pre-1900 numbers with a large dose of salt. Peter Laslett wrote (p. 215) “almost nothing is known about age of sexual maturity for any society before the 20th century.” http://tinyurl.com/yfhnzrf
Here is a part of a brief survey that Craig Foster has compiled:
J. B. Post. “Ages at Menarche and Menopause: Some Medieval Authorities.” Population Studies 25:1 (March 1971): 86. claims that medieval manuscripts placed menarche around the age of twelve to fourteen.
EliseDe la Rochebrochard. “Les âges à la puberté des filles et des garçons à partir d’une enquête sur la sexualité des adolescents,” Population 54:6 (NOV-DEC 1999): 938.
Age at menarche in France
1750 – average of almost 16 years
1850 – a little over 15 years
1900 – 14 years
2000 – 12.5 years
Joseph Jacobs and Maurice Fishberg. “Menstruation,” JewishEncyclopedia.com
Menarche statistics during mid to late-19th century:
Jewish girls – 13 years
“Slavonian” girls – later
Magyar peasant girls – 15-16
Jewish girls – 14-15
Slovak girls – 16-17
Both Jewish and non-Jewish girls began to menstruate between ages 14 and 17.
New York City
Average age for Jewish Girls – 12.7 years
American-born – 12.1 years
Foreign-born – 13.2 years
In the Laslett link I gave also gives numbers in the 16-17 range for 19th century Europe. For America my current best guess for 1850 would be around 15.5, but I won’t waste time arguing why. The exact mean age, even if it could be precisely determined is not all that important as I will now show.
Laslett (p. 232) also conjectures menarche ages have a normal distribution with a standard deviation between 1.7-1.9. The Coale-McNeil model also assumes age of eligibility is normally distributed. So even if we were to take i4m 16.4′s number as the 1840 mean and 1.7 as a standard deviation, 17% of all females aged 14.7 would have reached the age of puberty. This provides plenty of eligible females into the marriage market to account for the 1.9% [age dependent] or so that get married before age 15 reported earlier. There is no evidence that a statistically significant number of pre-pubescent girls were getting married in 19th century America.
i4m: “The mean age of first marriages in colonial America was between 19.8 years to 23.7, most women were married during the age period of peak fecundity (fertility).”
Studies on mid to late 19th century Mormon data shows the most fertile married cohort was the 15-19 age group. I could see no evidence that the younger brides in that group had a longer interval to their first birth or infant mortality rates were higher. Note that 19th century Mormon cohorts average marriage age was around 19.0 . Yet still no evidence a significant number of teens were marrying before menarche.
See D. L. Anderton and L. L. Bean. “Birth spacing and fertility limitation: a behavioral analysis of a nineteenth century frontier population.” Demography. 1985 May 22(2): 169-83.
i4m: “The psychological sexual maturity of Helen Mar Kimball in today’s average age of menarche (first menstruation) would put her psychological age of sexual maturity at the time of the marriage of Joseph Smith at 9.1 years old. (16.5 years-12.8 years =3.7 years) (12.8 years-3.7 years=9.1 years)”
Most theories of marriage timing note the increase in the mean marital age over time suggest that teens are psychological less mature now than back then. The phenomenon has been called “extended childhood” in the literature.
“The fact is Helen Mar Kimball’s sexual development was still far from complete. Her psychological sexual maturity was not competent for procreation.”
Even if sexual relations could be proved (even Todd Compton, IIRC, argues this marriage was “dynastic” and likely not consummated); i4m has presented no evidence that Helen Mar wasn’t one of the 17 percent (using i4m’s mean and Laslett’s standard deviation) at the 14.7 age mark that had already reached menarche. I take the general US trend to not marry pre-pubescent girls (sometimes their was engagement understandings in place) is evidence that Helen wouldn’t even been considered for marriage by her parents. Helen remarked (hat tip Cal Robinson) “I had grown up very fast and my father often took me out with him and for this reason was taken to be older than I was.”
So i4m’s content on marriage statistics in the 19th century is not very good. In declaring it a myth that it was normal for girls aged 12-14 to marry they are have created propaganda that can easily be falsified by readily available census data. One example of how this propaganda has perpetuated is found in Brian Mackert’s defense of the awful “Search for Truth” video that FAIR reviewed. Mackert borrowed i4m’s material wholesale, even though I think he is competent enough to do his own independent research. I do not wish to create drama with such critics, but rather hope those really looking for the truth will be able to find it here.