Measuring Thanksgiving Travel Volumes in the Midwest

Thanksgiving is a major travel holiday. This month on the blog, we’re going to examine how – and if – Thanksgiving travel appears in enplanement data in our region.

We took a look at six airports: Willard, Peoria, O’Hare, Indianapolis, Springfield, and Bloomington. Using monthly enplanement data from the Bureau of Transportation Statistics from 2003 to 2017, we determined where November ranked in monthly enplanement, for each year and at each airport.

Table: November Enplanement Ranking, 2003-2017Download table data for November Enplanement Ranking, 2003-2017.

Source: Bureau of Transportation Statistics. TranStats. Flights. All Carriers, Selected Airports. Monthly data, 2003-2017. <https://www.transtats.bts.gov/Data_Elements.aspx?Data=2>. (Retrieved 30 October 2018).

It’s immediately apparent that, although Thanksgiving weekend may be a major travel holiday, November does not consistently top the rankings for most enplanements at any of the five surveyed airports.

At Willard, November had the greatest number of enplanements once, in 2017, and the second-greatest number in 2015. In Peoria, November had the second-greatest number of enplanements in 2007, and again in 2016. November also had the second-greatest number of enplanements twice in Bloomington, in 2007 and again in 2015. In contrast, the highest November ever got in O’Hare’s rankings was fourth-largest number of enplanements, in 2003. It had the second-highest number of enplanements once at Indianapolis, in 2004, and was never higher in the rankings than sixth in Springfield.

Table: November Enplanement Distribution, 2003-2017Download table data for November Enplanement Distribution, 2003-2017.

Source: Bureau of Transportation Statistics. TranStats. Flights. All Carriers, Selected Airports. Monthly data, 2003-2017. <https://www.transtats.bts.gov/Data_Elements.aspx?Data=2>. (Retrieved 30 October 2018).

Next, we looked at a simplified breakdown of the same rankings data: whether November’s ranking in a given year at a given airport was “high” (first through sixth in enplanements) or “low” (seventh through twelfth in enplanements). November breaks into the “high” half at least twice at every airport during the study period. But in five years of the 15-year study period (2005, 2008, 2009, 2012, and 2013), November falls in the “low” half of the rankings at all five airports. In the 75-result dataset in the table above, November is in the “high” half only 17 times. The year with the greatest volume of Thanksgiving travel in the region seems to have been 2007, with November in the “high” half of the rankings for four of six analyzed airports.

However, this array of results is not a reason to conclude that the volume of Thanksgiving travel is smaller than we thought. Enplanement data, particularly monthly enplanement data, leaves out content and detail.

The most obvious limitation is that enplanement data only covers air travel: trips for the holiday by car, bus, train, or other modes are not covered in this dataset. Also, enplanements cover departures, not arrivals. An assessment of arrival data, and whether the region sees greater numbers of travelers leaving in a given timeframe than arriving, would add interesting depth to this study, but is not possible with an enplanement-only dataset.

Another major consideration is that monthly enplanement data is, well, monthly. The tables above look at enplanements for the entire month of November, not just travel related to Thanksgiving. This is both a granularity challenge and a motivation challenge.

If weekly enplanement data was available, it would better allow us to gauge Thanksgiving’s specific impact on November travel. We could isolate the fourth week of November and only look at enplanements occurring then, without the dataset also including enplanements in the first three weeks, which almost certainly have nothing to do with Thanksgiving. But this brings us to the motivation challenge. Even if we could isolate data from the fourth week, we could not be certain that every single enplanement occurring that week was Thanksgiving-related without surveying travelers, which, as we discussed back in March 2018, is not exactly feasible. We could say that travel is occurring in that time period, but we could only assume that it’s related to the holiday.

So let’s assume that, when people talk about Thanksgiving travel being some of the heaviest of the year, they’re referring to the week, or even just a few days, immediately around the holiday. And let’s assume that that’s true, that if we were able to see weekly or monthly enplanement data, we would see a spike in enplanements on the relevant dates. If both of those things are true, why would November mostly still get lost in the middle or bottom of the list of monthly enplanements, as we saw above?

One possible explanation could be that, even if November sees a significant spike in enplanements in the days around Thanksgiving, enplanements for the rest of the month are so low that even the holiday boost doesn’t consistently drag November far up the list.

Another is that the spike in enplanements may be present but not as significant as one might expect, and a large segment of the individuals who choose to travel for the holiday choose to do it via car, bus, or train, and do not appear in this dataset at all.

A third explanation deals more with how we discuss travel volume in relation to holidays and other events that might have an impact: whether high volume is assessed in terms of days, weekends, weeks, or whole months. We would posit that other travel-heavy periods of time (e.g., December holidays, spring and fall school breaks, summer vacations) have more varying calendars and may entail longer trips, so all related trips would not be loaded on a small span of days or a single week the way that Thanksgiving-related trips are assumed to be.

So what can we conclude from this discussion?

We can probably continue to assume that people do travel a lot around Thanksgiving, even if this dataset doesn’t provide unequivocal support for the idea. We can conclude that this dataset is not ideal to gauge the degree to which trips are concentrated in periods of time smaller than a month (but that it would be better for looking at travel-heavy months where the volume of trips is distributed across the entire time period). We can continue to keep in mind that enplanement data is mode-limited and only pertains to air travel, and that even if we talk about a shorter, more specific timeframe, we can’t with total certainty attribute a holiday as the reason for a trip, even if the trip happens during that timeframe.

Travel volume data, for air travel and for other modes, has applications in transportation and infrastructure planning. Knowing when high traffic volumes are expected, and what those volumes are, can inform the capacity needed at passenger terminals. It can also highlight periods in which there is significant demand for first-mile/last-mile transportation services, or services that move a traveler from their home to an airport, train station, or bus station, or from the airport, train station, or bus station to their destination. Monthly enplanements may not be the most apt to assess Thanksgiving travel volumes, but can provide insight on other travel and transportation planning questions.

Related Documents

Leave a Reply

Please be respectful. All fields are required, and all comments are subject to moderation. URLs are not allowed.