UNC’s Noreen MacDonald has a very nice manuscript in the upcoming issue of the Journal of the American Planning Association:
From their abstract:
We found that 75% of parents driving their children less than 2 miles to school said they did this for convenience and to save time. Nearly half of parents driving their children less than 2 miles did not allow their child to walk to school without adult supervision. Accompanying a child on a walk to school greatly increases the time the household devotes to such a trip. Few Safe Routes to School programs effectively address issues of parental convenience and time constraints. 
So here we have it. The good thing about Noreen and her co-author is that they won’t allow the interpretation here to turn into some working-mom-blame (you think I kid; I have heard public health people say that moms entering the workforce have contributed to childhood obesity because more meals are prepared away from home and the children are outside walking less. This may be, but nobody’s blaming working fathers for this, now are they? Let’s think about how this problem is framed.)
One of the things about the manuscript that makes me wonder: 2 miles is a long way for a young kid to walk–so yeah, it’s going to take some time. I wonder if they were to go finer-scaled–five blocks or so away from school–whether they would get some new insights on why those parents are or are not walking their kids to school.
 N. C. McDonald and A. E. Aalborg. Why parents drive children to school: Implications for safe routes to school programs. Journal of the American Planning Association, 75(3):331–342, 2009.