Because, yes, I have a grasshopper intellect which jumps from one topic to another without resting too deeply on any of them. (Alternately, I’m interested in lots of stuff).
I’ve been doing a lot of writing and thinking about the pulp era, and about the current misconceptions of it. One of those misconceptions is that it was an unusually violent time period, with things like the Valentine’s Day Massacre a regular occurrence.
It wasn’t, of course–not unusually violent. But how to prove that? Fortunately, Douglas Lee Eckberg’s “Estimates of Early Twentieth-Century U.S. Homicide Rates: an Econometric Forecasting Approach” (Demography v32n1, Feb 1995, 1-16) does the job.
The gist of the article is that previous studies of homicide rates in the U.S. in the early twentieth century, which tend to claim that there was a surge in homicides in that time, are wrong for a number of reasons, specifically
improving accuracy in reporting, and changes in the composition of the registration area. Within ‘comparable areas’ the homicide rate had not increased notably, but as the registration area grew it took in states with ‘appreciably’ higher homicide rates….
At least some of the early increase in violence was an artifact of policing practices, which changed from underreporting homicides to ‘over charging’ offenses by ‘one or two degrees’…in Philadelphia much of the apparent increase was caused by a transitory policy of counting deaths by automobile accident as homicides…in addition, a ‘dark figure’ of unreported murders decreased greatly. As a result, researchers now disagree about the existence of a ‘homicidal crime wave.’
The crude data seems to indicate a surge beginning in 1904, at around 1.5 per 100,000, peaking at around 9.5 per 100,000 in 1933. But the data before 1933 was misleading because it “disproportionately included states with low rates of homicide” and “during the first decade…homicides were underreported within the registration area, perhaps extensively.”
Not only that, but “during the first decade of registration…both murders and suicides often were tabulated as ‘accidents’ because insufficient information was forwarded by local authorities.” ‘Accidents’ included categories like “traumatisms,” gunshot wounds, and “other external violence.”
Also, not every state contributed to the Mortality Statistics from the beginning. From 1900-1905 only Connecticut, DC, Indiana, Massachusetts, Maine, Michigan New Hampshire, New Jersey, Rhode Island, and Vermont did. No Virginia until 1913, no Illinois until 1918, no Arizona until 1926, no Texas until 1933.
So, basically, there was no real surge of violence. The homicide rate increased (6.4 per 100,000 in 1900, 7.9 in 1910, 7.8 in 1920, 9.2 in 1930), but it wasn’t a surge by any means.
Now, 9.2 seems like a lot. We’re at 4.8 today (down from 7.9 in 1984), which puts us at 34th on the list. We must have been pretty violent back then, right?
Well, no. One of the standard works on historical homicide rates is T.R. Gurr’s “Historical Trends in Violent Crime: Europe and the United States,” which is fascinating reading. For example:
Go farther back, and things get worse. There are problems with the reporting of violent crimes, of course: “no distinction was drawn between [homicide and manslaughter] in English criminal procedure until the sixteenth century,” and country-wide statistics weren’t gathered in England until 1805.
As it happens, England was a pit of crime centuries ago. One study
shows convincingly that murderous brawls and violent deaths at the hands of robbers were everyday occurrences in medieval England. The average annual homicide rates for five rural counties, studied at scattered intervals between 1202 and 1276, ranged from 9 per 100,000 population in Norfolk to 23 in Kent, compared with a contemporary rate of about 2 per 100,000 in all of England and Wales. Most of these violent deaths resulted from fights among neighbors, while between 10 and 20 percent were attributed to robbers or bandits. Knives, axes, cudgels, and other implements found in every agricultural community were the typical instruments of death. Usually several assailants were charged in each homicide, another item of evidence that most deaths resulted from brawling among small groups rather than one-on-one attacks.
The medieval English population was more than 90 percent rural. Fourteenth-century London had only 35,000-50,000 people, Oxford about 7,000. Estimates of urban homicide rates for this period vary enormously. In the thirteenth century…the London rate was about 12 per 100,000 and in Bristol 4 per 100,000. On the other hand Hanawalt estimates that London in the first half of the next century had homicide rates of 36 to 52 per 100,000…and in Oxford…an extraordinarily high rate of about 110 homicides per 100,000, a rate to which scholars contributed no more (as victims or assailants) than might be expected from their proportions in the population. Analysis of the details shows that virtually all victims and assailants in Oxford were males and, except for scholars, were almost always of low status. Scarcely any homicides occurred within families and at least one-third involved “strangers,” such as people having no fixed residence in Oxford. Unlike the rural pattern, very few homicides in Oxford resulted from robbery or burglary….
Homicide rates in Elizabethan England were noticeably lower than those for the medieval centuries, but still far higher than contemporary figures. Spontaneous violence was still common among rural people…few of the killings investigated at assizes during this time resulted from calculated violence: ‘rather, they occurred during acts of sudden, unpremeditated aggression and results from attacks with a variety of knives and blunt instruments. Fatal quarrels could originate in almost any context–at work, in drink or at play.” Handguns, known as “pocket dags,” were in limited use by this time and about 7 percent of violent deaths involved firearms.
So, Americans, take comfort. Even if it seems like we’re violent now, we’re not. The English were much worse.