Analyzing the effect of removing the Monroe Blue Line stop on transit ridership in Chicagoland
Monroe Blue Line and the CTA’s Station Spacing Problem
If you take public transit, you know how aggravating it can be when you’re in a rush to get to work, school, or a social event and the train or bus you’re on seems to be stopping every thirty seconds. “Why are we stopping again?” you think to yourself, “there do not need to be this many stops.” You’re not alone. Countless others experience this same problem every day. That said, while you might experience this in New York, or London, or Paris, you are perhaps most likely to experience this in Chicago.
Chicago has a problem: bus and ‘L’ (Chicago’s metro system) stops are too close together. This is readily apparent to all transit riders, with buses sometimes stopping multiple times on one block. It is not just anecdotal either; Chicago has some of the shortest distances between bus stops of any (major) city in the United States (Pandey et al., 2021) and the median line speed of CTA buses has slipped below nine, yes nine (!), miles per hour.
The L is no different; stops much closer together than other systems in North America. There are numerous places on the system where stations can feel redundant, particularly within Chicago’s core ‘Loop’ business district. The loop of track the neighborhood gets its name from is one of the worst offenders, with stations every two or three blocks the whole way around. Similarly, the Red and Blue Lines, which run straight through the heart of the city have stations incredibly close together, with some station entrances as close as half a block away! This brings us to our big question: is it really worth it to have stations so close together?
Trade-Offs: Convenience vs. Speed
The more a bus or train has to stop, the longer it will take to get from A to B, and the slower it will be able to take people around the city. Imagine a bus that stops every five feet; it would spend hardly any of its journey actually moving, and when it would move, it would either be accelerating or braking, not cruising at the speed limit. That said, the other extreme (too few stops) can be just as bad. Imagine a bus that only stops twice across a ten-mile route, flying by anywhere that anyone would actually want to go. Clearly then, there is some happy medium, a compromise that provides enough people enough access to the places they want to go, without slowing the service down so much as to make it practically useless. It is worth asking then if the CTA has found this happy medium, or not (probably not).
On the Blue Line, there are three stations (Washington, Monroe, and Jackson) within just six blocks. Worse yet, the blocks are actually half-block length, so the distance between the bookending stations, Jackson and Washington, ends up being just six hundred (600) meters, roughly a seven-minute walk. Finally, the cherry on top comes from the fact that, like many stations on the system, each of the three has multiple entrances and exits, placed at the ends of the stations. This means that one entrance to Monroe is less than one block (!) from an entrance to Jackson! The interactive map below shows the current L system centered on the Loop. Note that the distances between Washington and Monroe and Monroe and Jackson tie for the shortest on the network, along with Monroe and Jackson on the Red Line and Clark/Lake and State/Lake on the Loop.
L System Interactive Map
Further, as close as these three Blue Line stations appear on paper, they feel far closer in real life. This is due to the fact that the Dearborn Subway, the actual tunnel in which Washington, Monroe, and Jackson lie, is actually one long platform, stretching from Washington Street in the north, to Van Buren Street in the south, with no barriers between (nominally) separate stations. Should they want to, passengers are free to walk along the platform from one of these stations to the next, and it is not uncommon to hear jokes about running from station to station and beating the train.
I hope it’s plainly obvious by now that the Monroe Blue Line’s existence warrants some (to put it mildly) skepticism. Why does it have to exist when there are two other stations within literal shouting distance? Well, maybe it doesn’t…
A Note on Station Removal
Before examining what removing Monroe might look like, it is worth discussing the complexity that comes with removing any transit station. First, there are obviously a massive number of factors we could consider. Removing a transit station is never painless, and it often means picking winners and losers. Longer travel times for some might mean losing access to employment or convenient access to their social network. For some, such as those with disabilities, it may have serious adverse impacts on their health or ability to get around independently at all, and these concerns should not be taken lightly.
With that said, planners must also consider who benefits from any proposed change. It is not difficult in the least to imagine how swifter access to the Illinois Medical District, the state’s largest healthcare center, might benefit the many resident’s on the Northwest Side, many of whom are disabled or elderly. Similarly, quicker access to O’Hare, the second largest center of employment in Chicago, could yield greater economic opportunity for residents from all over.
Evidently, to analyze every effect closing Monroe might have would be borderline impossible. A good start, however, would be examining the underlying claim here; would removing Monroe really save anyone time at all?
Approach to the Problem
If Monroe disappeared tomorrow, the CTA Blue Line would be faster, no way around it. But at least some people would now have to walk further to access the Blue Line. To explore the impact removing Monroe would have, I analyzed travel times across Chicagoland, for the current transit network as well as for our hypothetical Monroe-less scenario. I then used these travel times to predict how transit ridership would be affected. Would people respond to these changes by switching to driving? Would faster travel times attract new riders? Or would the network simply absorb the change with minimal disruption?
To model ridership for both scenarios, I used a trip-based modelling approach that has long been the standard in transportation planning.1 The most common trip-based model is the four-step model, and this is what I used as the basis for my analysis. The first step is Trip Generation, which predicts how many trips people in a given zone are likely to make, sometimes based on the characteristics of that zone (zonal model) and sometimes on the characteristics of their household (categorical analysis model). The next step, Trip Distribution predicts how these trips will be distributed throughout the region, i.e., it predicts where each trip will end or start based on the kind of trip in question.
Fortunately, for this analysis I was able to use extremely high-quality trip generation and distribution model data from CMAP’s “Go To 2050” transportation plan (2017). In the end, I had a roster of all daily trips in the Chicago area, with their origin and destination transportation analysis zones (TAZs). I also had the purpose and time period for each trip as well. In total number of daily trips in the region is staggeringly large, coming in at just over 31 million!
With travel time matrices calculated for both the current network and the Monroe-less scenario, I could then estimate how ridership would change. The mode choice model I implemented is known as a pivot-point model, which is a simplified (but still powerful) approach to analyzing small changes to an existing transportation system. Rather than attempting to predict absolute ridership from scratch, the pivot-point method starts with the existing distribution of trips across different modes and calculates how changes in travel times would shift the probability of choosing one mode over another. The basic logic is straightforward: if transit becomes faster relative to driving, more people will choose transit; if it becomes slower, some riders will switch to driving. The magnitude of these shifts depends on how sensitive travelers are to changes in travel time, which varies by trip purpose and whether the destination is in the central business district.
The model uses mode choice coefficients calibrated by CMAP specifically for the Chicago region from their “On To 2040” model (CMAP, 2010), which reflect how people actually respond to differences in travel time between transit and driving. For each trip in my dataset, I calculated the utility of transit in both scenarios (with and without Monroe) as well as the utility of driving, then used a logit model to determine the probability of choosing transit in each case. The change in probability multiplied by the number of existing transit trips gives the estimated change in ridership for that origin-destination pair. Some trips would see transit become more attractive as through-riders benefit from faster service, while others would see transit become less attractive as former Monroe users face longer access times to Washington or Jackson stations. Summing these individual changes across all 1.05 million transit trip records in my dataset, representing 1.4 million daily person trips, I could estimate the net impact on regional transit ridership. The fourth and final step of the traditional four-step model, Trip Assignment, was not necessary for this analysis since I was only concerned with whether people would continue to use transit at all, not which specific routes they would take.
Results
After countless hours of work, I was thrilled when my model was finally done running. So what were the results? My model found that removing the Monroe Blue Line station barely affected overall transit ridership.
Removing Monroe Blue Line station would reduce transit ridership along the Blue Line by exactly three trips, despite the large number of trips that pass through Monroe. This represents a change of negative 0.009 percent, so small it would be completely invisible in real-world ridership data, which fluctuates by far more than this with the weather, special events, or simple day-to-day variation. For all practical purposes, Monroe could disappear tomorrow and nobody would notice.
No News is Good News
Given that removing Monroe would have essentially no impact on transit ridership (according to my model), does it make any sense to remove the station? Yes actually. While removing Monroe might not improve ridership, it would not decrease it significantly at all, and there are plenty of reasons for a transit agency to prefer to have fewer stations outside of this that might justify removal.
Keeping an extra station open means more infrastructure to maintain, and therefore higher costs. At the moment, the CTA is repairing some stairs at Monroe, just one example of a maintenance cost that could be avoided. It also means higher personnel costs for two reasons. First, and most obviously, the agency has to hire workers who work at the station. Second, increased end-end line times for transit vehicle means increase driver-hour needs. When a station is removed and the line speeds up, the amount of time vehicles need to be running, and therefore staffed, goes down. So removing the station could save the CTA money. Beyond money, fewer stations means less complexity. The tracks at Monroe would still exist, as would the platform, but getting rid of service at the station would mean less unpredictability and therefore fewer delays.
Of course, there are other arguments in favor of keeping the station, as discussed above. The point here is not to give a comprehensive evaluation of the arguments for or against removing the station, but to suggest this: because ridership would not significantly change, the CTA should evaluate these other arguments and make a reasonable decision about whether to remove Monroe. Maybe someday soon we will live to see a world without Monroe, but if we do, don’t expect ridership on the Blue Line to change all that much.
Author’s Note
This article is quite long, and I have tried my best to limit it to what I believe is interesting. I have broken this article down into sections with useful names which should make it easier to digest. I have included a (lengthier than I was expecting) section that details the process I went through in working on this project, as I believe this gives some interesting insight into what a project like this actually entails. I realize, however that this is likely the most dull part, so please do feel free to skip it. You can access the code I wrote for this project here. The data page here contains links where you can download the input data I used, as well as my outputs. Finally, you can find links to all the relevant sources in the references section below.
This project is the culmination of a semester’s worth of work in my Intro to GIS class (yes, the scope of this project was a bit absurd from the outset) and it taught me a lot about GIS and programming along the way. I hope you enjoy it!
References
Chicago Transit Authority. (2025). Monthly Ridership Report. https://www.transitchicago.com/assets/1/6/Monthly_Ridership_2025-10.pdf
City of Chicago. (2024a, July 12). CTA – “L” (Rail) Lines. Cityofchicago.org. https://data.cityofchicago.org/Transportation/CTA-L-Rail-Lines/xbyr-jnvx/about_data
City of Chicago. (2024b, August 9). CTA – “L” (Rail) Stations. Cityofchicago.org.
CMAP. (2010). 2040 Travel Model Documentation. N/A (No Longer Available). https://www.cmap.illinois.gov.
CMAP. (2025). Travel Demand Model Data (c24q4). Illinois.gov. https://datahub.cmap.illinois.gov/documents/090fa5e7e1df41b3957ade1e80aafb71/explore
CMAP-REPOS. (2025). GitHub – CMAP-REPOS/ONTO2050-indicators: ON TO 2050 indicators help measure the progress of plan implementation. GitHub. https://github.com/CMAP-REPOS/ONTO2050-indicators
Flickr, & Thomas, J. (2025, December 11). 14 – Jeffery Express. Flickr; 14 – Jeffery Express | Jamaal Thomas | Flickr. https://www.flickr.com/photos/cta5760/3459710489/in/faves-203941433@N04/
Miller, P., & Flickr. (2025, December 11). Chicago. Flickr; Chicago | Peter Miller | Flickr. https://flic.kr/p/23KwCLu
Pandey, A., Lehe, L., & Monzer, D. (2021). Distributions of Bus Stop Spacings in the United States. Findings. https://doi.org/10.32866/001c.27373
Trip Generation Zones – Zones (2017). (2017). Arcgis.com. https://cmap-cmapgis.opendata.arcgis.com/datasets/901627cb9e3149ff8f67663998f0602c_0/explore?location=41.903633%2C-87.928194%2C9.02
Wolf, E., & Flickr. (2025, December 11). Illinois Medical District Station. Flickr; Illinois Medical District Station | Erik Wolf | Flickr. https://flic.kr/p/2jijRFU
Notes
- Many planning agencies are moving to other approaches such as agent-based models or tour-based models. ↩︎
Leave a Reply