On Sunday, August 16, 2015 at 7:08:22 PM UTC-4, ceg wrote:
On Sun, 16 Aug 2015 13:42:28 -0700, trader_4 wrote:

As I and others have said, it could be that other causes of accidents eg
drunk driving, have been going DOWN. We know the number of deaths due to
drunk driving have been cut by half. It's reasonable to assume that
there are also a lot more non-fatal accidents that have also been
eliminated. It could be changes in what gets reported and what
doesn't. Were the standards of reporting, the methods the same in all
states, over all those years? It seems the census folks have concerns
about something there, with the warning about year to year comparisons.

It could be a *lot* of things, I agree.
Hence the paradox.


Nice edit job, where you ignored where I showed you that you're
continued statement that accidents have been going down for years
is wrong and also where the census bureau folks that you cite say
that trying to compare data from year to year "should be done with caution".

Looks like you're not interested in the actual facts, just repeating
the alleged "paradox"

Here is what I posted again:

For someone so concerned about what's going on, seems you haven't really
spent much time looking at the data yourself, even though you dumped
the unanalyzed, raw data links on us. From your very first link:

http://www.census.gov/compendia/stat...es/12s1103.pdf

In 1995 there were 10.7 mil accidents, in 2009 there were 10.8 mil.
That isn't going down, down, down.

They also state:

"Data are estimated. Year-to-year comparisons should be made with caution."

Which may explain why with the number mostly steady at about 10.7 mil
from 1995 to 2009, there is a one time huge jump up to 13.4 mil in 2000.
In other words, given that disclaimer, we really don't know the accuracy
of the data set.