It looks as if, after correcting for current imprisonment, the number of black men that are truly missing is not nearly as large as has been suggested. Here’s a simple way of looking at it: the number of black infants born in a year is about 600,000 [ 661,701 in 1990, 587,781 in 1995, 601,346 in 2000, 583,759 in 2005]. A little more than half of those are boys: call it 300,000 . The average number of black male victims of homicide in recent years is a little over 7000 a year [average of 7267 per year over the period 1995-2007] . Rates were some higher in the 1980s: 9767 in 1980, 7891 in 1985, 11489 in 1990.
Which would mean that something like 2.3% of black men eventually became victims of homicide in recent years, perhaps as many as 3% during the peak crime years in the 70s thru the mid-90s.
Total deaths of black men by homicide since 1970 have to be fewer than 400,000. If a couple of million black men are missing – a couple of million more than those in currently in prison – homicide can’t be the primary explanation.
What about some other kind of mortality that disproportionately strikes black men? Well, there is one, cardiovascular disease. According to the CDC, it has a larger effect on average lifespan than homicide Average lifespan is 76.5 for white men, 71.8 for black men. Higher homicide rates among black men explain just under a year of that difference – more is explained by higher rates of cardiovascular disease. But of course cardiovascular deaths generally occur in mid-life or later.
About 250,000 black men are on active service in the Army. They don’t show up in the Census tables of the ‘civilian noninstitutionalized population’, but they’re certainly not missing.
Blacks have had significantly higher AIDS-related mortality than whites. At the worst of the epidemic, in 1991, before the effective antiviral cocktails, HIV was the leading cause of death among young black men [25-44], more common than homicide. This risk is much smaller today. So how many missing black men died of AIDS? Total deaths due to AIDS among blacks are bout 250,000: most were men, but a significant number (perhaps a third?) were among women. So AIDS may explain 160,000-200,000 missing black men.
I could see about a real shortage of about half a million black men (50 and under) compared to the number expected if blacks had the same mortality rates as whites, and another ~600,000 imprisoned [state and federal]. Any number much higher than that must be due to guys escaping being counted by the Feds, or alien abduction.
West Hunter 
“I don’t have an address,” he said. “I don’t live anywhere. I go from house to house, to friend to friend, to family member to family member. What I’ve been doing for the past four weeks is wearing out my welcome with everybody who knows me. Some places I don’t think I’m allowed to go back. I don’t know. I don’t do nothing.”
http://www.dailydot.com/lol/charles-ramsey-cleveland-hero-broke/
I know this is motivated by the example of homicide, but before breaking it down by type of death, you should start by checking if the census numbers are compatible with observed all-cause mortality.
All data is for blacks in the years 1999-2010. For the 15-19 group, there are about 1500 more deaths among males than females each year. For 20-24, 2k more deaths. For 25-34, almost 3.5k. So from 14 to 35, the sex difference should change by 7k, that is, the sex ratio should change by just one point.
oops… not 1 point, but more than 2 points.
They are missing, but not dead, not as we know death anyway.
(This faintly reminds me of a Black Panther poster 30 years ago, showing that many prominent members had all mysteriously died. The implication was that the police had executed them for being uppity. Closer examination of the deaths showed a more prosaic reason. Death by colleague, not death by cop. Traffic accidents and other mishaps accounted for most of the others).
Thanks for sorting all this out.
Let’s move to another subject.
I wonder how many murders never come to the attention of the cops. I mean, with normal people, their relatives and friends will quickly know something is wrong and get the authorities involved. But with criminals, transients, drug addicts, or other people on the margins of society? If they vanish, the people they know maybe assume they had to split for some reason, or they don’t like involving the cops, or don’t care. Of course the body might turn up, but if its disposed of well enough it might not. For example its not exactly rocket science to put it in a car trunk and find a deserted cornfield or forest, dig a hole, and bury it.
The people who kill people like that are usually the same kind of people and don’t spend lots of time driving way outside their urban conglomeration for sophisticated corpse disposal.
“I wonder how many murders never come to the attention of the cops. I mean, with normal people, their relatives and friends will quickly know something is wrong and get the authorities involved. But with criminals, transients, drug addicts, or other people on the margins of society?”
Criminals aren’t really marginal in the sense you mean in the sub-culture they are a part of – transients yes.
Too many movies
For a crude estimate of the sex ratios you might start with a stationery population approximation. If Sm & Sf are the male and female survival probabilities then if r is the initial sex ratio at birth and a and b are the lower and upper bounds of an age range then r (S (a b) Sm(x) dx )/( S (a b) Sf(x) dx ) is the expected sex ratio between ages a and b for a stationery population of males and females. The Census Bureau publishes life tables distinct by sex for US blacks which you can use to estimate the integrals.
Stationery population is a poor approximation for populations heavily affected by migration but presumably the US black population as a whole is not too greatly affected by migration. However black population in a particular region might be subject to considerable in and out migration. Otherwise the more temporal variation there is in birthrates and deathrates the worse will be a stationery population approximation.
The Census Bureau attempts to estimate the amount of undercount but this is more an art than a science. Because payments to states and localities for many Federal programs often depend on Census Bureau population numbers there has been a lot of litigation over just how or whether to estimate population undercounts.
The Census Bureau publishes descriptions of how their life tables are constructed including any allowance they make for uncounted exposure.
Well, as I indicated in a previous thread, I never meant to imply that a large fraction of the “missing” urban black males had died by homicide, merely that they’re generally dead or imprisoned, with a large fraction of the deaths probably due to “lifestyle” factors rather than just the normal differences in male/female mortality.
Obviously, cultural and genetic factors cause different racial/ethnic groups to follow different mortality curves. For example, despite their relative poverty, Hispanics tend to have lower mortality and morbidity rates than whites, consituting the so-called “Hispanic health paradox.” A similar, opposite effect is certainly true for blacks as well.
But that’s why it’s useful to separately examine America’s major cities (250K+) vs. the rest of the country. The first dataset contains about 30% of American blacks, 30% of Hispanics, and 30% of Asians, but only 10% of non-Hispanic whites, and the serious crime rates are generally about 3x-4x the national average.
As it happens, the relative percentage of “missing” black males is roughly 10x greater in that urban dataset than in the remainder of the country, which shows the huge relative impact of imprisonment and unusual death patterns as opposed to normal black mortality trajectories. Basically, urban black females outnumber males by about 25%, while the figure for the rest of the country is closer to 2%.
One large factor is the relocation impact of imprisonment. Major cities have lots of jails, but state and federal prisons are almost always located elsewhere, so most imprisoned blacks are shifted out of the urban into the non-urban Census category.
I strongly suspect that urban blacks also are much more likely to make “lifestyle” choices that severely reduce their life-expectancy, having far higher rates of drug and alcohol abuse and “risky behaviors.” Deciding what fraction of the various mortality categories have been influenced by such behaviors may be difficult, except in the case of a few small items such as drug overdoses and AIDS deaths. Perhaps there’s some data on black mortality categories stratified by city that could be aggregated and compared to the national average in order to estimate some of that difference.
It’s also certainly possible that a significant number of the “missing” urban black males just fell between the cracks of the Census and the annual ACS surveys, but I’m just not sure how to estimate that sort of error in the official statistics.
Incidentally, I took a closer look at the BJS reports on incarceration, and it appears that the figures are based on a year-end time-slice rather than year-round totals, meaning that the short-term nature of most jail sentences doesn’t really impact the daily average. In 2011, it appears that about 2.3M total Americans were imprisoned at any given point in time, and since about 40% of them were black, over 900K blacks are probably incarcerated at this very moment. Given that urban areas have such high crime rates compared to the rest of the country, I’d say that between one-third and one-half of the “missing” urban black males are currently behind bars somewhere.
“I strongly suspect that urban blacks also are much more likely to…”
I think gang culture itself magnifies this.
“As it happens, the relative percentage of “missing” black males is roughly 10x greater in that urban dataset than in the remainder of the country, which shows the huge relative impact of imprisonment and unusual death patterns as opposed to normal black mortality trajectories.”
Does this not contradict your other research which suggests that the percentage of blacks in an area predicts the homicide rate? What you seem to be saying is that the missing black male problem is 10x greater in the urban data set than it is for non urban black communities. But if the overall rate of predicting homicide based on the percentage of blacks in area is consistent and linear, then how do we have evidence to believe that the rate of homicide against black males is much much greater in urban areas simply because they are urban?
AIDS and a violent lifestyle consumes many black male lives. True. But why is it that a black project in a city has 10x more of these factors against men primarily (not both men and women) than an equally poor black neighborhood in North Carolina?
Or did you use the same data set for your other research for the homicide rate correlation for black percentage composition? That was also only looking at the major cities?
I’m not entirely sure I understand your question. As I emphasized, my crime correlation analysis was entirely restricted to major cities and I ignored the rest of the country. One reason is that cities almost invariably have much higher crime rates than rural areas or suburbs, so including the latter would severely distort the results (unfairly penalizing a heavily-urbanized group such as blacks).
Remember, even if one-quarter of urban black males are “missing” the other three-quarters are not, and they presumably have something to do with the very high crime rates. Since the “missing” blacks—being imprisoned or dead—are probably disportionately criminal or violent, I also pointed out that if they were still around, the urban crime rates and correlations would surely be enormously higher.
Finally, my paper avoided suggesting any particular theoretical explanation for these empirical findings, partly because I don’t really have anything original to offer. But individuals of various ideologies are welcome to fill this hole.
@Ron
‘Penalize’ blacks? Either the data is there or it is not. It is not a penalty to report the truth. What would the data look like if the whole info was included?
Cities are disfunctional socially. Most of the blacks I know who escape the cities do better because they are a cut above anyway.
> The people who kill people like that are usually the same kind of people and don’t spend lots of time driving way outside their urban conglomeration for sophisticated corpse disposal.
Well now. Such people are awfully rough, dumb (mostly), and out-there — but that doesn’t mean they sneeze at 12 years of prison. If they do a drive-by or similar assassination on somebody, then maybe they don’t care about the corpse, though they might well care even then.
But I would imagine they get into plenty of impromptu altercations at places where they are quite likely to have trouble with the law — or the ‘anti-law’, or both — if they leave the body lying around. Accordingly, I would expect the number of undocumented murders to be high.
If you are some thug, why file a missing person report about your friend? If someone wants a ransom they’re going to contact you. If your friend skipped town he’s not looking to be found. If he overdosed in some alley or crackhouse, too late.
Now, maybe the census comes looking for him seven years later: he was around before, and his death isn’t on record, so where’s he at now? But how can they reasonably sort it out, when multitudes of living people just straight-up refuse to be counted, according to the census worker on the other thread. Almost all of those people were once documented too, at least as kids, and now they are off the record, probably forever in many cases.
> and now they are off the record, probably forever in many cases
Eh, maybe that’s not true. If they eventually die nonviolently, I guess somebody will typically be reporting that fact somewhere?
Or, might they go into the hospital in some terminal state, whether caused by violence or not, and just refuse to give their true ID? In that case they will have gone off-record forever.
Note that in a stationery population with sex ratio r at birth and life expectancy em for males and ef for females the aggregate population sex ratio is r em/ef . You can use the Census Bureau life tables for black male and females to get a crude guess at what the aggregate sex ratio for the US black population should be based on a stationery population model.
Ron – Amerindians have generally even lower mortality than Hispanics. In addition to the Census Bureau websites you might wish to check the Society of Actuaries and American Academy of Actuaries websites for addition data.
RS – I’ve read that in a place like Detroit less than 10% of homicides ever result in an arrest. So I would guess that the typical murderer in a place like Detroit isn’t much worried about getting caught. The US generally has pretty good reporting of deaths. Any undertaker or funeral home which failed to file death certificates with the county health department would certainly lose their license it this were discovered and be subject to severe penalties.
In times past it might take several years for a death certificate to get from the county health department into the national database but this is no longer the case.
On the other hand the reporter LeDuff stumbled on a cadaver lying around in Detroit and I’ve read that the Detroit coroner’s office has a backlog of several humdred cadavers.
The country is big but the Government has such a comprehensive database that I doubt anyone can go “missing”. The perpetrators of the Boston Marathon bombing and other cases were found in hours. The fact that Greg and Unz wonder about the whereabouts of hundreds of thousands “missing” men must mean something. Certainly not that they are weak in statistics and demography, but that the answer is not straigthforward.
One methodology would be to look at the registration of births seventy or eighty years ago in a racially mixed sample, and track them against completed death certificates.
Trying to get the correct exposure to construct census life tables has always been a problem. The Census Bureau estimates the exposure undercount in constructing their tables but these estimates are pretty uncertain. Migration is the biggest problem, unreported births less important but significant particularly in past times when birth registration was much less complete. Unreported deaths is a minor problem though possibly not in Detroit.
Most places i’ve known about single moms on welfare get less money if they admit to having a live-in boyfriend – so they don’t admit it. I doubt the boyfriends are going to be signing any forms at her house. Is it possible to compare if white or hispanic underclass men are strangely missing from the census also? If so then the missing difference in the male/female gap between populations may simply be an artifact of the different proportions within each population who are part of the underclass?
Actually not even just underclass really – the percentage of children born outside wedlock within might be a closer fit.
Completely off topic, but I am interested in finding papers on the scientific testing of smallpox vaccines or even a careful analysis of any potential of lessening in virulence due to selection for lower virulence.
Black males, illegal immigrants and so cannot just drop out of society and disappear. Not in America. The Government may not count nor bother them, providing an illusion that they are invisible. Simply, the Government has zero interest in feral black males.
Abortion. They never existed in the first place, and there are incentives to inflate the population numbers.
Unfortunately, I need to correct some of the black gender demographics I’d casually quoted a few days ago based on a quick glance at one of my spreadsheets followed by a back-of-the-envelope estimate. I realized that the figures were so extreme that I decided to go back and take a closer look. Here are the corrected numbers, all from the 2010 Census.
The three American cities with the largest black populations are NYC, Chicago, and Philadelphia. Across these cities, adult black women outnumber adult black men by 34%.
Next, if we just consider the twenty-odd American cities with black populations of 150,000 or more, adult black women outnumber adult black men by about 29%. These cities contain approximately 23% of the total black population.
Meanwhile, if we consider the rest of America excluding these most heavily black cities, adult black women outnumber adult black men by about 13%. I’m sure the figure would be much higher in most larger cities and much lower in suburbs, towns, or rural areas.
Finally, the overall national figures show adult black women outnumbering men by around 17%.
I still think this large difference in black gender ratios in different locations is due to the imprisonment and excess mortality factors I’d previously suggested (perhaps reinforced by unknown degree of urban Census error), but it is obviously much less extreme than the differences I had previously provided. My apologies for the error.
The Census Bureau gives for 2007 a life expectancy of 70 years for black males and 76.8 for black females. If the sex ratio at birth is 1.02 this gives 1.02(70)/76.8 = .93 for the aggregate population sex ratio for a stationary population.
Perhaps after the clarification, we can move on now? Entirely off topic, here is a late addition to the lactose tolerance literature which probably will seem redundant to our distinguished hosts. http://www.nature.com/news/archaeology-the-milk-revolution-1.13471
“As Middle Eastern Neolithic cultures moved into Europe, their farming and herding technologies helped them to out-compete the local hunter-gatherers. And as the southerners pushed north, says Gerbault, the LP allele ‘surfed’ the wave of migration.”
Combined with the genetics and looking at their very nice map
http://www.nature.com/news/500020a-i2-0-jpg-7.11661?article=1.13471
it seems to me much more plausible that the neolithic farmers expanded rapidly during the climactic optimum but as that came to an end and their population declined other populations along the periphery rebounded: whether Iberian archers, Funnelbeaker milk drinkers or Steppe wanderers, with the rebounding populations having a higher proportion of hunter-gatherer descent.
I don’t want to be too cute here, but it really this is like the search for missing matter in the universe – dark matter and/or dark energy. We know there must be something out there but we can’t see it.
Mr. Unz thinks that the missing blacks in the cities are real whereas others suggest that there is a population of real enough black men who for one reason or another are invisible to conventional perceptions and measurements
This is the point where I would like to propose the key to the conundrum – but I can’t. Mr. Cochran seems to have exhausted all the obvious and not so obvious explanations that are amenable to analysis from available public data sources. I think for further insight, someone will have to do some active research.
If there really is a current of black men that flows beneath our sight they must still come into view on occasion. When a black man comes to the attention of the police we need to determine if he was on the public radar screen in the previous year. When the cops arrest someone for murder they could determine if he were counted in the previous census. If not he would have been in the invisible undercounted black male urban population. I can speculate as to how such a study would come out – but again I really don’t know.
One problem is that there is no institutional incentive for the cops to bother. A case can be made that the FBI and local police have a bias in their statistics gathering such that race issues are obscured. I think there might be resistance to active research that might uncover unwanted race info.
Just use the virial theorem!
> RS – I’ve read that in a place like Detroit less than 10% of homicides ever result in an arrest. So I would guess that the typical murderer in a place like Detroit isn’t much worried about getting caught.
Caught by whom? What about other gangsters? That’s what I meant by ‘the anti-law’.
Plus, Detroit sounds pretty singular. (Too singular for me to want to go check.) Apparently Chicago is plenty bloody, but I doubt it is as anarchic.
And 0.10 * (12 to 25 years or whatever) in the can is still no joke, even if it is 10x closer to being a joke.
Extensive undercount of black males in many cities seems quite plausible to me. Patrick – The main priority of cops in places like Detroit is staying alive not sociological/demographic research.
I find this post about “Missing Black Men” fascinating.
The sex ratio thing, and the matriarchal nature, among the black African-American community is something talked about and repeated by a lot of American black women. Black women in the USA are disproportionally single, despite sometimes being preferred by affirmative action (hence, “financially successful”) and part of the artificial middle class created for them in the USA.
A couple of articles I’ve seen is trumping the drum of them “dating outside of their race” towards white American and doing interracial stuff, but it’s a bit too late since immigration is flooding the (Southern) USA with mestizo Hispanics and some Asians (Asians are the smallest group in the USA).
The competition isn’t just fierce, it’s also a numbers game (local trends, exceptions, proportions, generalities and statistics). By sheer numbers interracial stuff isn’t that common (most white Americans are married and have children with each other) but when it does happen (the rare and occasions), there’s preferences. Asian women and Hispanic women tend to be liked more by white American men (at least, in the South).