Originally Posted by
mr_bill
Thank you for the nice link.
Originally Posted by ACS
The 2009 ACS question related to means of transportation asked respondents in the workforce, “How did this person usually get to work LAST WEEK?” (see Figure 1, Question 31). Although commutes may involve multiple transportation modes (for example, driving to a train station and then taking a train), respondents are restricted to indicating the single travel mode used for the longest distance.
The link leaves me with a comment and a question. First, that's a lame way to collect and process data. It would be pretty straightforward to simply ask the length of the segments and do the arithmetic to show the true amount of each mode, either by time or by distance. Imagine a community in which everyone drives twenty percent of their commute to catch a train for fifty percent and then use a bike-share for the final thirty percent. According to the ACS, that's a community that is 100% public transit. In reality, that mythical community is far from that. This problem is only going to get worse as more regions get bike-share and multi-modal trips get more common.
My question is: Has it ever been thus? The one ACS I filled out several years ago asked about trip segments. I don't remember if there was a separate commuting question, but it asked the purpose of each segment. The link was from 2009, so I assume they have been doing it this way since at least then. Was it done differently prior to 2009?