If I could elaborate for just a minute, you have to take in context, number one, that the improvement in 2006 versus 2005 represented about a 30% reduction in main-track accidents. Granted, if you compare year-to-date 2007 numbers versus 2006 numbers, you would think, if you don't understand the data, that there may be an issue.
The reality is 2006's numbers were phenomenal numbers. If you compare 2007's numbers to 2004 and 2005, they are 40% better. In 2006 we had a very mild winter, and as much as I understand your comments about weather, weather plays a huge impact when it comes to steel wheels, steel rails. If we have extremely adverse weather conditions, the likelihood of an incident occurring when it comes to rail failure increases. We can't change the metallurgical components of steel; that's what happens.
If you look at the trend now, from March 1 through to yesterday, and I'll tell you through the first quarter, if we compare 2006--keeping in mind that was a phenomenal quarter--the improvements exceeded much more than 30%. Annually, we improved 30%.
If I go back and look at the first quarter of 2006, we had a total of 67 TSB accidents on the CN railway compared to 76 in 2007, and the gap is growing closer. Today, through April 24 or 25, we've had 85 in 2006 versus 89 in 2007. As I've stated before, 2007's numbers are still a 40% improvement over 2004 and 2005.
While if you don't understand the data, it could raise a concern--and you may think the focus is not there that was in 2006--you have to take into account that 2006, number one, was phenomenal. The year 2006 had a very admirable winter in that we had temperatures not even coming close to what we just experienced this year. So the trend or the concern is going to dissipate; the momentum is going to continue.
The focus is still there. The numbers? We still feel very confident that 2007 will generate the same types of improvements year over year versus 2006. It's simply not fair to take a look at the first three months of 2006--when there are so many differences between 2006 and 2007--and make your assessment based on just that.