wrote in message
oups.com...
Riley Geary wrote:

Year Reg MC Fatalities/10,000 Reg MC (NCSA)
1993 38000 7.6
1994 36000 7.8
1995 36000 7.8
1996 36000 7.8
1997 37000 5.4
1998 37000 9.5
1999 39000 9.7
2000 43000 9.3
2001 48000 11.8
2002 51000 12.9
2003 54000 12.6

The other problem here is that the LA-DPS data conveniently begins
with
1997, a year with an anomalously low number of motorcycle fatalities
in
Louisiana. Was this simply an innocent choice of both the data range
and
data source by the author of section V, or was it deliberately chosen
to
mischaracterize the increase in motorcycle fatalities per 10,000
registered
motorcycles as being much higher than it really was following repeal
of
Louisiana's MHL?

It was both innocent and deliberate.

Oh really? The more I dig into the "Evaluation of the Repeal of Motorcycle
Helmet Laws in Kentucky and Louisiana" report at
http://www.nhtsa.dot.gov/people/inju.../kentuky-la03/ the
more fraudulent the whole study appears to be, so I'll have to eventually
provide a more comprehensive critique of it; but first...

It was the year before the repeal,
versus the year of the repeal. Totally logical. But if you want to
average 1993-1997, then average 1998-2003, that's fine too.

And why would we want to do that when the MHL wasn't repealed until August,
1999? The 1998 data must be included on the other side of the ledger, and
the 1999 data should probably not be a part of either data set since it was
a transitional year. Now this obviously won't change the fact that
motorcycle fatality rates have indeed risen much faster in Louisiana than in
most other states recently, but your evident inability to get even the
simplest things straight is becoming increasingly annoying. Do we really
need to fact-check every piece of data you wish to bring to our attention?


7.3 per 10K with MHL versus 11.0 per 10K after the repeal. So it's only
a 51% increase in fatalities.

And again, it's the fatality *rate* per 10K registered motorcycles we're
talking about here, not fatalities. But the real point (assuming there
really is a point, given the natural variation inherent in the statistics of
small numbers) is that the fatality rate didn't suddenly jump as a response
to the repeal of the MHL, but had already been on the upswing for the
year-and-a-half prior to repeal, and actuall *fell* during the first full
year following repeal!

Well, that last part isn't quite true either. It turns out NHTSA has not
been consistent in the presentation of its motorcycle fatality rate per 10K
registered motorcycles data. Starting with the 1996 Fact Sheet, it began
calculating the fatality rate based on registration data from the previous
year rather than the current year, with the further complication that the
1998 Fact Sheet erroneously repeated 1996 registration data instead of using
the 1997 numbers, and the 2000, 2001, and 2003 Fact Sheets apparently used
both previous year fatalities and registrations. Sorting through all this
obfuscation, and assuming for the moment that 1996 and 1997 registration
numbers are essentially equivalent (which certainly seems to be true for
Louisiana at least), we have the following corrected fatality rates for
Louisiana:

Year Reg MC Fatalities/10,000 Reg MC (NCSA)
1993 38000 7.6
1994 36000 7.8
1995 36000 7.8
1996 37000 7.6
1997 37000? 5.4
1998 39000 9.0
1999 43000 8.8
2000 48000 11.9
2001 51000 12.4
2002 54000 12.2


Sorry Riley. There is just no way you can twist the data to prove what
you're trying to prove.

No, I'm the one trying to untwist the data so you can actually understand
it, but I suspect that may be a rather hopeless cause by this point.

And it doesn't matter anyway. The states did
the right thing by repealing the MHLs, but don't kid yourself into
thinking that they did it without knowing that there'd be consequences
in the fatality rates.


Sure there are likely to be consequences in terms of increased fatalities
whenever a MHL is repealed, but that's primarily due to the increase in
motorcycling by aging boomers such a repeal facilitates, not to any real
change in the underlying fatality rate with respect to either registrations
or vehicle miles travelled. As a little sneak preview of my forthcoming
critique on the Kentucky/Louisiana study, I'll simply point out that
Louisiana is the *only one* of the five states repealing their MHL's to see
a significant long-term increase in fatality rates for motorcyclists thus
far. It could be argued that Kentucky (1998) and Texas (1997) still have
marginally higher fatality rates than they had in the last few years prior
to repeal, at least as of 2002, but Florida (2000) has been essentially
flat, and Arkansas (1997) actually has a substantially *lower* fatality rate
now than it had prior to repeal.

Likewise, Guy had earlier posted CDC data indicating states without a MHL
generally had lower fatality rates than states with a MHL, and my recent
research into the subject comfirms that finding--with the 25 continuously
non-MHL states experiencing a slight decline over the past decade (from
about 5.3 to 5.0) and the 20 continuously MHL states (plus DC) experiencing
a more pronounced increase since the mid-to-late-1990's (from about 5.5 to
6.5), so who is really kidding whom about fatality rates?

Riley Geary