October 13, 2005

Jarring, Revolting, Abhorrent and True

Approximately 43% of the dramatically decreasing American crime rate can be attributed to the increased number of police (accounting for 10%), and the increased number of people imprisoned (accounting for about 33%).

The remainder is attributable to legal abortion.

The longitudinal data is in, has been carefully analysed, and it’s well established that American crime rates have been “dropping like a stone since 1991”. The year 1991 marks roughly 18 years since the US Supreme Court upheld women’s right to abortion in Roe v Wade.

To put this simply: the population required to preserve, or increase, the crime rates of the past does not exist, because a significant percentage of today’s would-be criminals were never born. "Abortion is one of the greatest crime-lowering factors in American history.”

Let’s be very blunt: if legal abortion is ever significantly restricted or even banned (a highly improbable scenario, in Australia, however, America seems perpetually on the precipice), we can sit back and wait 20 years for a “catastrophic rise in crime”.

Where have all the criminals gone?” is chapter four in the book Freakonomics, by Levitt and Dunbur. This is not the first time the evidence has been raised, but on this occasion it can’t be sweep away as too distasteful for polite conversation, after all, Freakonomics is currently the best selling nonfiction book in the US.

In The Weekend Australian (8 – 9 October 2005) Caroline Overington provides that increasingly rare beast in MSM concise, clear reporting, about a complex matter, covering all key points in a dispassionate and impartial manner. No hysterics, no asinine opinions, no hyperbole, pure reporting.

Ignore the erroneous headline, the work of a clumsy and uncomfortable sub-editor, it would seem. Crime rates seen as a matter of life and death

Post Script – meanwhile, Craig contributes this thought - “Call me old fashioned, but I'd rather execute the criminals after they have committed the crime...”


  1. whilst your point is rather compelling there could be a plethora of other things that have helped reduce the crime rate that we just don't know about...

  2. Wow - that is an amazing idea. It may well be true as well. I didn't quite follow the reasoning though - is it purely because 1991 is 18 years since RvW? That doesn't seem a strong argument.

    I imagine I will have to read the book - Freakonomics is on my list anyway.

    You wont belive this, but I just finished reading Rudolph Hoess autogiography - Commandant of Auschwitz - last night. Spooky coincidence or what?

  3. I can't see any causality between the two things at all, only that the abortion rate is up and the crime rate is down.

    But so is the number of people buying buggy whips.

    I've recenty seen people try and claim that the US crime rate has dropped significantly since the playstation was introduced, and that therefore violent computer games are good for you, but the problem is that we don't have a control group to test against and there are numerous other factors involved that aren't/can't be included in the results.

  4. What I find even more remarkable than the crimerate/abortion trend is Caz's discovery that global warming is inversely proportional to the number of pirates on earth.

    Now, piracy is a severe crime. However, piracy on the high seas has fallen remarkably in the last few centuries, which is surely even more significant than the fall in US crime rates.

    This means that in order to save the planet from global warming, we should presumably try to ensure that as many abortions as possible occur each year. Only when every fertile young woman sucks at least a few of her unborn infants into bloody oblivion can we hope to save this planet for the future generations that we'd probably have if they weren't all aborted first. Hurrah! :)

  5. “coffee and cigarettes” – other than economic prosperity and the aging population, I struggle to think of the hidden “plethora” of other things that may have contributed to crime rates decreasing, year after year, over a period of nearly a quarter of a century. In other words, where is the plethora of “other things hiding, and how is it that we know nothing at all about them? It’s got me beat.

    “Rodeo Clown” – I presume the authors address the causal links thoroughly in their book, but we’ll all have to read the whole chapter for ourselves – I can’t speculate on the detail, as I haven’t read it yet either. I don’t think they are making the claims lightly, or without having evaluated all other possible reasons.

    Craig – As I said, above, I think we’ll all need to read the chapter to understand how they have reached their conclusions and established causality, but I gather this is not a flippant or superficial conclusion. Keep in mind, they have assessed data over a period of decades, that is, abortion in the US has been legal since 1991 – the author’s did not look at a single year and fast forward 18 years, they had 24 years of data upon which to draw, and a crime rate that has continued to plummet for 24 years. They do discuss and quantify all other factors contributing to the drop in crime, and it just happens that abortion is the single largest contributor (by more than 50%), but not the only reason.

    Hugo – excellent pick-up on the possible linkages between global warming, pirates, abortion, and falling crime rates. I did not even think of this myself, so thank you for pointing it out. However, you seem to have it a bit arse-end-about, in that it’s the deficit of pirates causing global warming, so your deductions and conclusion are exactly wrong.

  6. So Caz are you pro choice are what?
    BTW are you an Anglican? I like to know what religion folks identify as.

  7. I'm reading Feakonomics at the moment, and the stats appear to be legit. I'll post about it sometime in the near future

  8. Craig - look forward to your review & opinion. Have not even bought the book yet, so I will rely on you to do the hard yards for me!

  9. olympiada - no, athiest. I do not identify with any man-made religion, and there are no other kind.

    I am firmly in favour of safe, affordable, and appropriate health care for all women, at all stages of their lives, with full protection and privacy from public, social and political scrutiny. Ditto for men.

  10. So, I'm guessing if you went back in time and were faced with the option of killing Hitler as a baby, you *would*?

    Now *that* is the way to prevent crime through abortion.

    And one other thing...

    When you use statistics to try and draw conclusions, it is nice to have some sort of method. For example, where is your control sample? Where is the sample of the population where abortion wasn't an option and an analysis of that crime rate? Basically what I see is a person taking two statistics that may or may not be even remotely correlated and drawing a conclusion based on the fact that the two lines go the same direction.

    The Pirate/Global warming analogy is a good one.

  11. Strange how some readers seem to think that I made up this research all on my own, despite pointing out and linking to a book review, which names the authors, and provides a brief outline of their research, analysis and conclusions. This is a controversial and challenging piece of research, and it warrants sensible attention.

    The alternative, it would seem, is total censorship of unpalatable facts and ignorance of multiple view points, not just about abortion, but in relation to any "sensitive" subject, which would rule out just about everything really, since there's always at least one person who is "a bit touchy", or just a bit twitchy, about any given topic.

    So, we all just shut up and play dumb?

    Ignorance and stupidity are not viable options. Censorship, used for the purpose of keeping people ignorant of truth and facts, is abhorrent.

    Ooops early forgot - pirates & global warming Jason? Hmm, yes, very serious piece of research that one. Excellent to see that you recognized the scholarly quality and similarities between that study and the one on "where have all the criminals gone”
    Although, alas, another, normally outstanding, blogger has challenged the pirates and suggests that the cause is actually the deficit of bushrangers. You would appreciate that I reject his hypothesis out of hand - absurd, churlish, unfounded suggestion. We need pirates, damn it, not bushrangers!

  12. Easy now. No one is suggesting censorship of any ideas (or at least I'm not). However, I am suggesting that this idea is likely completely unverifiable.

    You are correct that I have not read the book, but only the article you linked. I am absolutely certain that the author made lots of careful studies and that the lines all go up, but there is no way to definitively demonstrate that one variable is correlated to the other. Why is that? Because there is no control sample. Now, if they had an equal population sample with which to compare crime/abortion rates, then you might be onto something.

    Obviously the Pirate example is taking this to an absurd level. On the other hand, I'd really be interested to see if auto deaths have dropped steadily since 1991, 18 years after Roe v. Wade. I'm guessing they have, but I doubt that it's because the bad drivers have been aborted.

    Sure, the book is likely great fun. I will put it on my list of reads. But I think statistics like this are best taken with a grain of salt.

    And finally, no, we should never shut up and play dumb. We also should never accept a study as fact unless it can be completely proven. *That's* science.

  13. Jason – Craig has read the book; you can read his quick review of the chapter in question on his site - http://tinyurl.com/ad65v.

    Refer also to one of my previous posts - http://tinyurl.com/7s9to - “Tassie and other tidbits”; but the relevant bit is at the bottom of the post, about the fact that more than 50% of all findings are eventually found to be untrue.

    This is not a complete response, due to lack of time & energy, and the fact that the skinny blog design makes comments look ten times longer than they really are…

    Likely unverifiable? Hmm, interesting comment, but I would need to read the chapter myself before I could jump to any conclusion that they are pissing in the dark.

    Your auto-deaths example is a reasonably good one, but, of course, there are far more parameters – all quantifiable – surrounding car accident death statistics, so logically, I can’t see it as an equivalent to the example in the book. Also, where is your control group for auto-deaths? There is none, by the standard you seem to be suggesting, and there never will be, because it would be immoral to put a group of people behind a wheel without seat belts, drunk-ish, only let them ride on bad roads, tell them to drive at high speed, and only let them have very old cars, with none of the benefits of current design and technology. But wait, that’s not true, is it - people do each of these things of their own will, or because they have no option, therefore, there are defacto control groups and defacto statistics. Again, I haven’t read the chapter, but are there defacto control groups for abortion and crime – yes, you can compare cohorts and compare decisions and outcomes and so on. Indeed, for all social phenomena, there are, in fact, defacto, or naturally occurring control groups. There would be no fields of research into a squillion different things if this was not the case. So I don’t think your point about control groups stands up. In fact, it falls apart very easily.

    BTW – I would be very surprised if any correlation was found between road deaths and abortion rates in America, or rather, a very weak correlation, if any, not the 50% plus suggested for abortion / crime rates. But then again, I don’t know if their road statistics have improved in the way they have in Australia, so, I’ll retract all comment, as I don’t know how, or if, America has the same approach to road accident prevention that we do.

    Apart from that, not everything requires a control group, as control groups are only relevant to some types or areas of enquiry.

    Two things struck me about this particular study, the first being that this is not a new idea, as it was apparently raised a few years ago, but was promptly buried. My guess is that will happen again, despite the wider publicity. Personally I like these type of ideas, that is, backed up with substance, not yet another entire book about someone’s unsubstantiated or highly selective opinions. I like these ideas because they give us a real prod – and how rare is that, these days? I can honestly say that any connection between abortion rates and crime rates had never, ever, occurred to me. Personally, I find it difficult to get my head around, but only because the concept, even in theory, is quite uncomfortable.

    The second thing that immediately came to mind was whether the statistics and the conclusions are unique to America. Would similar findings be replicated in other, comparable countries? My thought is “no”, but that could well be wrong. I only think “no” because I’m making unfounded assumptions that, perhaps, in the US, pregnancy terminations are significantly skewed, rather than, for example being dispersed across a wide range of age, economic, religious and social groups – in Australia, for example, I think you would find a relatively wide spread. In America? Perhaps it isn’t, perhaps that’s part of what leads to fairly compelling “evidence” on the link with dropping crime rates? Craig’s quick review suggests that this is the case.

    We should never accept things until they are “proven” – yes, sure, but again, I refer you to my previous post. Most things are eventually disproved, except for the goodness of broccoli, of course (also covered in a previous post, as chance would have it).

    “Completely proven” – your emphasis. Remarkably little has ever been “completely” proven. I always find it extraordinary how much we know, until I remember how very, very, very, little we know for sure.

    I’m hoping that other professionals tackle this one, and do their bit to disprove the abortion / crime link, I’m hoping it doesn’t get swept into the basement, never to see the light of day again. I think it’s an intriguing area of study, very much warranting others, with differing views to jump into the fray. I suspect it may not happen though, which is a pity, because if this study and it’s conclusions are so wafer thin, so lacking in substance, it should be easy for a few dozen other economists or social scientists to shoot it down, and provide us with evidence of the "real" reasons for the dramatic decline in the American crime rates, particularly murder. Preventing murder and other violent crimes, as well as lesser crimes, is an enormously important and demanding area for study and understanding. No hypothesis should be untouchable (and I know you are not suggesting such), and it has been a very long time since any new light, or new ideas, were thrown into matters of crime prevention. This would be a good starting point for refreshed debate, new insights, new ideas. Every minor reduction in crime is one less victim of crime.

    BTW – Hitler was a politician, not a criminal. Politicians are most commonly criminal in retrospect, and even this outcome is entirely dependent on how the winds of history blow. Of course, there are exceptions to this “rule”. I can think of a few current day “criminal” rulers, for example, who are widely shunned & condemned, but who will never be tried in a court, or bought to account.

  14. Caz, re this issue being raised a while back, I remember reading about it a couple of years ago.

    It actually made plenty of sense to me at the time, with the rationale being that a high proportion of criminals come from single-parent families, principally with absent fathers and little appropriate role modelling by either parent. Mum too busy working, and the guilt that goes with single parenting often leading to lax discipline. Add to that the stresses that go with living on or under the poverty line, and you have the makings of an argument.

    Of the women that I know who've had abortions, one was in a violently abusive relationship, another was a single parent (she subsequently had 2 more kids and so has 3 kids to 3 different fathers. She is still married to the 3rd) and another had to abort an ectopic pregnancy.

    How would the children of the first girl fare with a father who smacks their mother around? Especially when he already has 2 kids to another woman he used to beat up on, also.

    When children are brought into a world where they have no secure, respectful home life, where discipline and restraint are all over the shop, and there is never enough money for even a tank of petrol to drive down to the beach for the day, and you're stressing over the rent, and the electricity bills and food, it's easy to let things slip by in a child's education.

    That's my 2cents worth, anyway.

    And for the record, I've never had an abortion. :) Never wanted one, although single parenting gets you down sometimes. As was pointed out to me the other day, anyone can be a mum or dad, it takes bloody hard work to be a good parent.

  15. nilknarf - there are times, not often, but definitely times, when I think we all get that creeping feeling that SOME people should be BANNED from being a parent, or prevented from repeating the exercise, if it's already too late to stop them.

    As I tend to say: any idiot can breed, it takes no talent, no skill, no intelligence, and takes barely a few moments out of a person's life time; it's not clever and it's not an achievement. (Notwithstanding the heart-break for people who have fertility problems, and no disrespect to them.) Breeding is easy; parenting is a life-time committment. Your children are always your "babies", no matter their age. Parenting is not some temporary thing that ends at some point.