Some recent volleys have been fired in the on-going culture wars between the faithful Mormon community and their anti- and ex-Mormon counterparts. It is not quite fair to contrast a general authority’s polished speech and some off-hand message board remarks backed by shoddy research. Bruce C. Hafen’s remarks were summarized in this Mormon Times article, but I take responsibility for applying them to my recent experiences in researching marital ages of 19th century wives alongside those of Joseph Smith and the Deseret era saints. Some excerpts from MT:
“Readers have no way of knowing which critical claims have already been discredited, and the anti-Mormon sponsors are certainly not going to tell them right there on the site,” he said.
The democratization of ideas sometimes confuses the reader as to what is true and what is not, as all ideas are presented horizontally and as fact, thus positioning the blogger’s flippant opinion alongside the scholar’s well-researched dissertation.
Last week on a message board someone took note of Dr. Gregory Smith’s article that reports statistics from the 1850 Census and from pre-exodus Nauvoo that shows that age differences between spouses were substantial back then. There was some thoughtful discussion taking issue with GLS’s plea to avoid presentism and some brought up aspects of Joseph’s plural marriages that would have been unacceptable even for that time (like bigamy, which I have actually blogged on recently.)
Then the discussion went backwards, when one poster, despite the statistics to the contrary claimed “”the age of marriage in 19th Century America was probably later than it is now and teenage girls that did marry usually did so to boys more their own age and not men 20+ plus years older.” Even though this forum does not allow any defense of the Church I thought that posting some objective numbers would be helpful.
As far as average age, 1980 was close to 1840, but there has been a lot of movement since.
1840 SMAM 22.4
1980 SMAM 22.9
2008 SMAM 26.6
SMAM = Singulate Mean Age At (first) Marriage 
10% of 2008 brides were teens compared to 43% in 1840. 
1850 grooms aged 34-38: wife averaged 10.1 years younger, 19% were teens
Joseph Smith: wives averaged 6.7 years younger, 30% were teens.
The poster making the claims I addressed, helpfully linked to some items thought to call the statistics I presented in question. Otherwise, the ad hominem reaction I received brought Hafen’s remarks into focus for me. In a sense, it doesn’t matter how well I research and publish statistics in forthcoming scholarly venues on 19th century nuptiality. Anti-Mormon sponsors will hardly ever acknowledge that a criticism has been addressed or seriously engage empirical that does not support their paradigm. Wanting to not reciprocate, below I will address one of the statistical claims made in my interlocuter’s supporting links.
Claim: “ In Utah (1850s to 1890s), the average age of a 2nd wife was 17 (husband average age early 30s) and the average age of a third wife was 19 (husband average age mid to late 30s). The average age in the USA for a first marriage in the late 19th century was about 22.”
I want to be the first to acknowledge when an ex-mormon researcher gets something right. The numbers given for USA average is in the right ballpark for female first marriages. The given typical ages for husbands depending on marriage scenario are also probably defensible, even if they lack context. For example they could have addressed how serial monogamists or older single men in their 30s married back then. To return to my earlier point, using 1850 “marrinyr” data from IPUMS, men aged 34-38 married women 10+/-6 years younger than themselves. Using all men lessens the gap to 5+/-6. I am reporting mean and standard deviation to the nearest integer. Men in their 40s were 13+/-10 years older than their wives (11% teens).
Let’s move on to testing the claims about wives’ mean ages being 5 years below the national average during the plural marriage era. The table below analyzes the 1880 census. It shows that Utah’s overall average and its Coale-McNeil minimum age was very close to that of the West.
1880 10% Census Sample 
|AREA||SMAM||Min Age||AD Mean||14||15||16||17||18||19|
|11: New England Division||25.1||14.3||24.2||0.3%||1.3%||3.5%||7.7%||13.8%||21.5%|
|12: Middle Atlantic Division||24.5||14.4||23.4||0.3%||1.4%||4.1%||9.0%||16.2%||25.2%|
|21: East North Central Division||23.6||14.4||22.5||0.4%||2.0%||5.9%||12.6%||21.9%||32.6%|
|22: West North Central Division||22.5||14.2||21.7||0.7%||3.0%||8.3%||17.0%||27.9%||39.8%|
|31: South Atlantic Division||23.4||13.6||22.1||1.2%||3.9%||9.1%||17.0%||26.6%||37.1%|
|32: East South Central Division||22.8||13.4||21.5||1.8%||5.5%||12.1%||21.4%||32.2%||43.3%|
|33: West South Central Division||21.2||13.4||20.4||2.4%||7.8%||17.1%||29.1%||42.0%||54.1%|
|41: Mountain Division||20.7||13.1||20.3||3.5%||9.7%||19.5%||31.4%||43.8%||55.3%|
|42: Pacific Division||22.9||13.9||22.0||1.0%||3.6%||8.9%||17.2%||27.4%||38.4%|
Many scholarly papers  have made use of the massive Utah Population Database. Regarding marital ages depending on wife order, it has been found that the mean of the ages of the first and second wife of a polygamous man is roughly the same as that as the average age of all monogamous wives. Put more simply, in a mixed polygynous/monogamous the society the ages of each sub-group track each other. Over time, first wives can be up to a year younger than monogamous wives and second wives up to a year or so older than monogamous wives. Contrary to the claim above, third wives were typically older than 23. In short, the 1880 census numbers above support the average ages being around 20.5 and not 17. On the other hand, Kathryn Daynes  estimates that during the Mormon Reformation in Manti average marital ages got much lower (~16.5 years). However the situation soon corrected itself and returned to above 20.
Bean and Mineau  divided Utah men into 3 birth cohorts. The last one (1840-1859) is of the most interest since it was the most active group around 1880.
|Mean Age||Mean Age Difference||St. Dev. Age Diff.|
 SMAM calculated using Hajnal’s method on raw data extracted from IPUMS. 1840 SMAM estimated using procedure outlined in previous blog entry and comments.
 % take into account age structure. More information about my methodology can be found in the link immediately above. Some adjustments have been made. 1) I assume a more conservative Coale-McNeil minimum age for 1800-1840 of 13.8. 2) I have added constraints in my numerical optimization routine to not overestimate the marriage rate at 14.5 years (to prevent minimum age from being underestimated). This sometimes results in an increase of about 0.3 years. 3) After 1940 I use a general log-gamma distribution on the first pass to get a tighter fit than the Coale-McNeil distribution. The result is that teen percentage in 2008 is actually closer to 8.5% rather than 10% as reported above.
 Census Division Map at http://www.census.gov/geo/www/us_regdiv.pdf
Min Age = Coale-McNeil minimum age of eligibility. Equal to AI Mean – 1.73 X standard deviation
AD Mean = Age Dependent average (takes population pyramid into account, closer match to marriage year cohort, county records )
% are cumulative percentages of brides in marriage cohort married by the end of the designated age. For instance 21.5% were married as teens in New England.
I should note that the chart shows a worst case comparison for Utah by not allowing the optimization routine to underestimate Min Age. If Utah had been processed like the other regions its Min Age increases to 14.0.
 Some papers I have found to be useful are:
G. P. Mineau, L. L. Bean, M. Skolnick, “Mormon Demographic History II: The Family Life Cycle and Natural Fertility,” Population Studies, Vol. 33, No. 3 (Nov., 1979), pp. 429-446
Lee L. Bean, Geraldine Mineau, Douglas Anderton, “ Residence and Religious Effects on Declining Family Size: An Historical Analysis of the Utah Population,” Review of Religious Research, Vol. 25, No. 2 (Dec., 1983), pp. 91-101
 Kathryn M. Daynes, More Wives than One: Transformation of the Mormon Marriage System, 1840–1910 (Urbana, Illinois: University of Illinois Press, 2001), 107.
 L. L. Bean and G. P. Mineau , “The Polygyny-Fertility Hypothesis: a Re-evaluation,” Population Studies, Vol. 40, No. 1 (Mar., 1986), pp. 67-81 see the table on p. 72. The other birth cohorts had higher age differences as plural marriage was phased in.