More problems with Kleck's DGU estimate

"Eugene Volokh" writes:

I should say that I agree with some of your criticisms of the Kleck
& Gertz results, and of the 1.5 million count arrived at by the
NSPOF study;

In case anyone remains who finds the Kleck estimate credible, let me
make a couple more observations:

On page 170 Kleck "generously" estimates that there are about 550,000
gun crimes each year. According to his survey, in about 18% of his
2.5M DGUs, the offender was armed with a gun. That's about 450,000
gun crimes. Apparently we are supposed to believe that in 90% of gun
crimes the victim gets to use a gun for defence. This could perhaps
help explain the 0.5M DGUs by people from non-gun-owning households --
they merely take the criminal's gun away and use that ;-)

Kleck has no trouble believing that defensive gun use is much more
common than criminal gun use because "there are far more gun-owning
crime victims than there are gun-owning criminals". Presumably he
would have no trouble believing that there were more many criminal
victimizations than crimes committed because there are more crime
victims than criminals. Of course, what determines the relatively
frequency of criminal and defensive gun use is the fraction of
criminals that use guns, versus the fraction of victims that use guns.
Common sense suggests that since criminals are more likely to be
involved in a violent crime than victims, they have a greater
incentive to go armed. They can also choose to only commit crimes
when armed. Consequently it is reasonable to expect that criminal gun
use will be more frequent than defensive gun use.

46% of Kleck's DGUs were by women. As Kleck admits (p178) because men
are much more likely to own guns and much more likely to be crime
victims, the true percentage should be much less. True to form, Kleck
explains this anomaly by supposing that there are millions more DGUs
by men that his survey did not pick up. An alternative explanation is
that DGUs that are made up do not have to follow actual gun ownership
or crime patterns.

but I was wondering what you thought about the NCVS
point I raised again a few days ago. To my knowledge, waiting for
respondents to volunteer information is generally considered rather
bad survey practice; and we saw that with the rape statistics
shifting to a direct question changed the total by about a factor of
2.5 or 3, if I recall correctly.

I have even been told --- entirely outside the defensive gun use
context --- that the trick is cuing as often as possible: Asking the
question directly, several times, in subtly different ways, to
trigger people's memories (and perhaps willingness to respond).

This was considered in the NCVS redesign -- it asks the screening
questions in several different ways. This would seem to be better
than just having one screening question, as Kleck's survey does.

Am I mistaken about this understanding of general survey
practice? If I am correct, doesn't it quite severely undercut the
80,000 estimate (while of course not proving the 2.5 million

There are three ways it could be a problem.

Firstly, respondents may forget to mention a burglary when asked
directly about it, but they might recall using a gun to defend against
a burglar when asked about using a gun for defence.

However, the survey already asks about such crimes in several
different ways, so the difference made by asking yet another cueing
question would be small.

Secondly, respondents may remember about the crime but forget they
used a gun to defend, but be reminded by a question that specifically
asked about guns.

I don't see this as particularly likely. Once you are recalling one
detail of an incident, other details come to mind.

Thirdly, respondents may be a little reluctant to mention the gun use
but a direct question may overcome this reluctance.

This seems possible. I can't guess at how much of a differnece it
would make. It would be interesting to do a survey that followed NCVS
methodology except for asking a specific question about gun use as
well as a general question about defensive actions.

In any case, I believe that a bigger problem is that there are
probably quite a few DGUs by criminals who are not going to
participate in victimization surveys, and by people who just do not
which to recount their DGUs to strangers. So, while the NCVS provides
the best available estimate, it probably undercounts, perhaps by as
much as a factor of two.

Eugene Volokh:

pointed out by Tim only a few have addressed gun use in self defense
against a crime. The NCS, the Police Foundation and Kleck.

Dr. Paul H. Blackman writes:

No, several have, but not the NCVS, which measures victimization.

It does indeed measure victimization. But this does not prevent it
from measuring gun use in self defence against a crime. Are you
attempting to deny the very existence of the NCVS estimate of 80,000

Eugene Volokh:

In sample size
for instance and response rate the NCS outperforms these surveys

Dr. Paul H. Blackman writes:

The sample size is lovely; it's kind of a shame it wasn't a
survey to measure the protective use of guns.

That's not its sole purpose, but it is one of the things that it was
designed to measure. You only have to look at the questions and the
categorization of the responses to see that.

Eugene Volokh:

Tim has shown several examples of 'problems' with Suter, not in the least
with Suter's selective use of data to 'prove a point'.

Dr. Paul H. Blackman writes:

Gee, and I thought Suter's selective use of data to prove a
point -- specifically, citing a few high-crime/low-gun and a
few low-crime/high-gun jurisdictions

You forgot to mention that he claimed they were representative.

-- was just a cutesy
imitation of what Zimring and the "gun control" movement have
been doing without criticism for three decades or more,
designed to show what would happen if pro-gunners attempted
the "science" of anti-gunners.

I have provided a specific example of Suter's deliberately deceptive
practices. Practices which you apparently condone. If you want to
claim that Zimring has done something similar, you need to provide
some evidence.

Incidentally, I understand that Killias and Kleck each
submitted affidavits in Canada regarding the court challenge
to their new law. But, under the Canadian system, for the
affidavit to count, the expert has to defend his statements
from examination/cross-examination -- presumably with the
lawyers questioning being assisted by some other expert
(Killias to help the lawyers questioning Kleck and Kleck
helping the lawyers questioning Killias). Under the
circumstances, Killias dropped out rather than defend his
affidavit from a lawyer assisted by Kleck. It's probably
just modesty, a professorial fear of performing in public.

More likely the trip across the Atlantic made it a bit inconvenient.
You seem to be insinuating that he was afraid of some question that
Kleck might ask. Feel free to share this "killer question" with us.
If it exists, of course.

I believe that it has already been explained here how Kleck's critique
of Killias is erroneous.


More like this

Dr. Paul H. Blackman writes: I was curious about the suggestion that hardly anyone could possibly still believe the Kleck data now that NSPOF had become the 15th or 16th such survey in the same general category. Then you seem to have misunderstood. Kleck's estimate (not his data - I have no…
"Eugene Volokh" writes: Please, please, let's take special care to be polite in these exchanges. This is a sensitive subject, but even when we think the other person is dead wrong, it's better to say this in a subtler way. OK, I'll do my best to be polite. I won't say anything in reply to the ill…
kebarnes writes: Are Kleck's numbers concerning the self-reporting of robbery and burglary incidences from this survey out of line with the comparable NCVS results, for instance? Rs to Kleck's survey reported that 5.5% (274/4977 Rs) had been a burglary victim within the past year, and 2.5% (124/…
"Eugene Volokh" writes: but I was wondering what you thought about the NCVS point I raised again a few days ago. To my knowledge, waiting for respondents to volunteer information is generally considered rather bad survey practice; and we saw that with the rape statistics shifting to a direct…