News & Politics » Feature

How To Fix the El

The CTA’s Red and Purple Modernization project is a rare chance to rethink Chicago’s rail system. Here’s what we should do.



The el is a mess and urgently needs fixing. But how best to go about it?

Here are some design features on which we can surely all agree. One: The stations shouldn't smell like outhouses. Two: When you remove the emergency shoring, the viaducts shouldn't collapse. Three: The trains shouldn't be crowded and torturously slow.

The Chicago Transit Authority has begun an effort to address some of these problems. It's launched the Red and Purple Modernization (RPM) project to rebuild the deteriorating elevated north of Belmont, most of which was built before 1922.

But more than repairs are called for now. It's time to rethink how service in the city's most important rail corridor works. Unless we use imagination and act decisively, a once-in-a-century chance to transform it will slip away.

The bulk of the money for RPM will come from the federal government, which provides most funding for mass transit capital improvement projects. State and local government must also pitch in. At the moment, all three are strapped, so this part of the job has its challenging aspects. But we'll let Rahm worry about that.

Assuming we do get the money, how should we spend it? Here's where things get knotty.

Let's consider the situation of four riders, based on an analysis I originally did with my boss, Cecil Adams, for a Straight Dope Chicago column last summer.

Aaron lives in Edgewater and boards the Red Line at Granville. He doesn't look forward to it for two reasons. The first is the sheer number of people Red Line trains carry—close to a quarter million riders on an average workday. Although there's still breathing room at Granville, rush-hour trains are often packed by the time they reach Belmont, and they stay that way into the Loop.

The other problem is how long the trip takes, partly because the Red Line's many stations are as little as two blocks apart, and the train stops at all of them. The ride from Howard to Roosevelt takes 42 minutes according to the published schedule, and sometimes considerably longer.

As Aaron bumps southward on the Red Line, he glumly watches a Purple Line Express cruise past. Everybody has a seat. Why the special treatment for suburbanites? he wonders.

We turn now to Bea, a Wilmette resident who rides the Purple Line. Bea indeed has a seat, which isn't surprising, since she boarded at Linden, the first stop. Crowding is never an issue at Linden or at the seven stops in Evanston. At a time when ridership on the north side has risen sharply, the number of people boarding at Linden has fallen by almost two-thirds—from 2,900 on an average workday in 1982 to about 1,000 now.

Why? Because the service sucks. During rush hour the trains depart every eight to 15 minutes. (On the Red Line, it's every three to seven.) Once the Purple Line Express gets to Belmont, it makes every stop from there to the Loop. The trip from Linden to Adams and Wabash officially takes 52 minutes. Metra trains make it from Wilmette to the Ogilvie Transportation Center in 34.

Stopped at Armitage, Bea looks up to see the Red Line trundle past on its way into the subway. She knows it'll arrive in the Loop sooner than she will, but if she'd transferred at Belmont, she'd have spent the last four miles of her commute smashed against a door. In fact the only reason the Purple Line stops in the city at all is to pick up the overflow from the Red and Brown lines. Some express service, she grouses.

Meanwhile, Cesar is trying to board the Red Line in the subway at North and Clybourn. The once-desolate neighborhood has perked up since he began catching the train here 25 years ago, but back then at least he could get on easily. Now, in the morning rush, the cars are usually packed when they arrive, and many more people are trying to wedge themselves in—4,500 board here on a typical weekday versus 1,600 in 1986. On the worst days Cesar has to let a train or two pass before he can find room.

Dee, who lives on the far south side, near 133rd and Indiana, has a more basic problem: the el doesn't go anywhere near her. To get to her job on the near north side, she rides a bus 4½ miles to the Red Line terminal at 95th, then takes the train north. The trip is an hour and 20 minutes.

Can the CTA realistically make the commute easier for all these people? We think so. But let's start with a few observations.

The el at the moment has three sets of problems. The first, evident to all who board the Red or Purple lines at any but the handful of renovated stops north of Addison, is general decrepitude—crumbling concrete, foul smells, leaking roofs.

Conceptually if not financially, this is easy to fix. The CTA has a lot of practice, having completed major rehabs of the Green, Pink, and Brown lines over the past couple decades, plus numerous smaller repair projects.

The second problem is that the el doesn't provide convenient service to large parts of the city, including some neighborhoods where residents desperately need better access to jobs. Here, too, the solution is conceptually simple: build more tracks. The CTA board has already approved preliminary plans to extend the Red Line to 130th Street, a project Rahm Emanuel endorsed during his campaign.

The rail system's last set of problems is more difficult. They're the kind of problems most transit systems would be happy to have—the challenges of growth. Ridership on the trains has boomed in recent years. Last year almost 174 million people passed through the turnstiles, the most since 1967. (And in 1967 Chicago's population was 25 percent greater than it is today.) The city is becoming more dependent on trains and less on buses. Since 1992, weekday el ridership has increased nearly 25 percent, while bus ridership has dropped by roughly a sixth.

But the growth has been skewed, concentrated on just three lines: the Howard branch of the Red Line, the O'Hare branch of the Blue Line, and the Brown Line. Traffic on these routes is at or near historical highs. In 1976 they accounted for 42 percent of passengers entering rail stations outside downtown; today it's 59 percent.

Nowhere is the trend more evident than on the Red Line, the city's backbone rail service. In 2010 the Howard branch surpassed the pre-World War II peak of 38.5 million entering riders recorded in 1927. In contrast, traffic on the Dan Ryan branch, though up modestly over the past decade, is substantially down since 1980.

Serving the north and south halves of Chicago with the same rail line is a tough challenge. On the north side the issue is mainly one of logistics—figuring out the most efficient way to squeeze growing numbers of people into a system with limited capacity. On the south side, the story's different. There's plenty of capacity; the real need is extending rail service to low-income neighborhoods where residents headed for jobs on the north side now endure a punishingly long commute.

That brings us to the mundane but maddeningly complicated business of train scheduling. Those unacquainted with the subtleties of railroading may imagine there's a simple way to improve service: more trains more often!

It's not that easy. The problem is what's known among transit geeks as "load balancing." On an average weekday more than twice as many riders enter the Howard branch of the Red Line as the Dan Ryan branch—115,000 versus 51,000. If the CTA simply sent more Red Line trains south from Howard in the morning to pick up north-side crowds, they'd unload downtown, then run largely empty down to 95th Street.

Extending the line to 130th Street will make matters worse. The trip from Howard to 130th and back will take more than two and a half hours. Most trains will have time to make just one run before the busiest part of rush hour is over. Upshot: too many trains on the south end of town, too many waiting passengers up north.

The trick, therefore, is to add service on the north side without throwing the rest of the system out of whack. The obvious way to do that is to carry more passengers on Brown Line and Purple Line trains, which circle the Loop and head back north. In 2009 the CTA completed a massive station-rebuilding project to accommodate eight-car trains on the Brown Line, which was previously limited to six-car trains. That increased capacity by a third, but created another problem we'll talk about later. Right now we need to discuss the rush-hour incarnation of the Purple Line, known in former days as the Evanston Express.

The el has been running up to Evanston since 1908, but express service in roughly the form we know it dates from 1949, when the CTA, then only a couple years old, implemented a systemwide reorganization of Chicago rapid transit operation. Until 1957 the Evanston Express stopped at Morse, Loyola, Wilson, Chicago, and the Merchandise Mart before rounding the Loop and returning north. City stops between Howard and Merchandise Mart had been eliminated by the late 1970s, then began to be added back 13 years later as ridership in Lincoln Park and Lakeview steadily increased. In 1997 the Purple Line Express began making all stops from Belmont south, leading to complaints that it was no longer much of an express.

The inelegance of this improvised solution bothered us at the Straight Dope. Knowing the RPM project was in the works, we began consulting with a collection of Chicago transit experts, most of whom had done work for the CTA in various capacities—the sachems, we called them—in hopes of devising a better way. Although the scheme we came up with had its primitive aspects, we worked it out in sufficient detail to draw three conclusions: It would improve train service for a broad swath of Chicagoans on both sides of town; it was, as far as we could tell, feasible; and rail transit planning was a remarkably intricate business. Little did we know.

See the diagram to the right for the Straight Dope plan.


The Red and Purple lines would operate in tandem, with the Red Line the local and the Purple Line the express, on the model of New York subways. On the far north side the Purple Express would make a few stops to let Red Line local riders change to a faster train; then the lines would merge and enter downtown via the State Street subway.

OK, maybe we got a little obsessive. But we figured this out in considerable detail:

• Rush-hour Purple Line trains, all with eight cars, would depart Linden every five minutes at peak, making all Evanston stops.

• Red Line trains would also operate every five minutes. The schedule would be arranged so that Red and Purple Line Express trains departed alternately from Howard, 2½ minutes apart.

• The Purple Line would stop at the busiest stations between Howard and Addison, namely Morse, Loyola, Bryn Mawr, and Wilson. From Granville to Sheridan the trains would run on the outer tracks. (The Bryn Mawr and Wilson stations would have to be rebuilt with "island" platforms to permit both locals and expresses to stop.) Red Line riders boarding at the northernmost local stations—Jarvis, Granville, and Thorndale—could transfer to Purple Line trains at the next express stop, Morse or Bryn Mawr.

• At Addison, Purple Line trains would switch to the inner tracks and make all current Red Line stops from there south, entering the Loop via the subway.

• The last Purple Line stop would be Roosevelt. South of that point the trains would exit the subway, stop on an unused stretch of the elevated, then head back north.

• Red Line trains would continue to the south side, but the line would be extended to 130th Street rather than terminating at 95th.

This system, we argued, would improve the commute for CTA riders in these ways:

Evanston riders: Purple Line trains would make more stops south of Howard, which would slows things down, but enter the Loop via the subway rather than the circuitous elevated, which would speed them up. Net impact on running time: zero. However, the trains would run twice as often, shortening the overall trip by as much as five to eight minutes. In addition, Purple Line trains would serve the Red Line stops between Addison and Roosevelt, which are by far the most heavily used in the system. Fewer Purple Line riders would have to change trains.

Most north-side riders: At the busiest stations, served by both the Red and Purple lines, trains would arrive every 2½ minutes at peak versus three minutes now.

North-side riders between Howard and Bryn Mawr: Service would be five minutes faster for those boarding at express stops and 2½ minutes faster for those boarding at local stops and transferring to an express. (The Purple Line would overtake one Red Line train en route.)

North-siders boarding at Berwyn, Argyle, Lawrence, and Sheridan: For these people, we were obliged to admit, service would be slower—trains every five minutes at peak rather than the current three. (You could transfer to an express at Wilson, but normally it wouldn't pass a local and you wouldn't save any time.)

South-side riders: Trains would operate at the same frequency as before, but terminate at 130th Street rather than 95th, shortening the bus ride for many on the far south side.

We were sufficiently taken with our handiwork that we sent it to the CTA for comment last May. Some weeks later, we received the following reply:

"Several items in the proposal are similar to items in the vision study [preliminary planning based on public comments] CTA currently is undertaking on the Red Line North and the Purple Line. There are differences; however, because CTA is analyzing recent feedback and analysis, we cannot at this time get into specifics."

We thought: Similar, eh? This we gotta see.

We didn't have long to wait. Earlier this year the CTA released preliminary design alternatives for the RPM project. The task now was to solicit public opinion on this range of rehab approaches. After that the CTA would begin drafting an environmental impact study in which the choices would be winnowed, culminating in the selection of a preferred alternative for approval by the Federal Transit Administration, ideally in 2012. Approval in hand, the city could then seek funding for engineering and construction. How long might that take? Not to pass the buck, but you'll have to ask Rahm.

In late January we went to an RPM open house the CTA held at Senn High School. In a hall off the main lobby, agency staff stood near display boards and answered questions about the project. Six alternatives were shown, ranging from "no action" (which nonetheless cost $280 million for patchwork fixes) to a complete reconstruction of the el north of Belmont for an estimated $4.2 billion.

Some of the options were intriguing. A three-track modernizing plan would replace the two express tracks with one—Purple Line Express trains would head southbound into the city during the morning rush and northbound in the afternoon.

An underground alternative was more daring still. The current four-track elevated would be replaced with a two-track subway under Sheffield, Sheridan, and Broadway between Belmont and Loyola. Advantages: easier construction (the elevated could continue to operate while the subway was built) and faster service, since there'd be fewer stops and curves.

All the alternatives except the cheapest, we noted, had similarities to our ideas:

• The Red Line would be extended to 130th Street—at any rate that's what one of the display boards showed.

• Purple Line platforms would be extended to accommodate eight-car trains.

• Except for the subway alternative, transfer stations would be added to the Purple Line Express.

• Some stops might be closed or consolidated. We'd considered this as well, mainly on the Purple Line—in fact at one point we wanted to cut back all service in Evanston north of Davis Street, on the grounds that few used it and sending trains up there was a waste. But the transit sachems talked us out of this. The CTA was bolder, proposing to close stops on both the Red and Purple lines, although when this drew a predictably negative reaction agency officials backed off a few days later, telling reporters they were just throwing out ideas.

But the CTA had ignored a key feature of our scheme: running the Purple Line into the subway. CTA staff in attendance told me they weren't necessarily opposed to that idea, but felt it was beyond the project's scope—the RPM project only dealt with matters north of Belmont.

We considered this a narrow view. The el was a system; you couldn't change one part without affecting others. But we got the impression the agency was planning to rebuild much of the most heavily used stretch of its rail system without thinking through how it would work.

Take the two-track subway alternative, for example. That would mean the end of the Purple Line Express, although Evanston service would be retained. How then did the CTA plan to operate its trains? Would trains run all the way from Linden to 130th, a distance of some 30 miles?

The CTA engineers we spoke to at the open house didn't know—train operation wasn't within their purview. They'd been asked to look into ways to rebuild an old rail line, and a subway offered some practical advantages. But as for how the resultant service might operate—well, that was a job for another day.

We were never able to get much of an answer to this or any other question about the kind of service the CTA expected to offer after rebuilding the Red and Purple lines. Mike Connelly, the agency's vice president of service planning and scheduling, told me in an interview that the RPM Project "is not a major operational study; it's about rebuilding infrastructure."

The RPM planners were making allowances for some service changes, such as adding transfer stops to the Purple Line Express. But that was as far as they planned to take it. "This study does not include any stations south of Belmont," a CTA spokesperson said in answer to follow-up questions. "All potential service patterns are being considered within the area of study in a way that will not preclude future service options."

And what might those future service options be? For example, what did the CTA think about rerouting the Purple Line Express through the State Street subway? That idea was "not a part of this study," the spokesperson said.

We worried that the CTA, by focusing so specifically on infrastructure, might spend years and billions without resolving the fundamental problems of inefficient service and slow, crowded trains.

We raised these issues with the CTA

sachems—specifically, three ex-CTA planning officials with more than 60 years of transit experience among them. They preferred to remain anonymous lest they appear to be criticizing their successors, but they too were concerned about RPM's direction. They felt the project ought to encompass the entire Red/Purple run, from Linden to 130th Street. As part of the process, they thought, the CTA should work with city and suburban planners to coordinate transportation and land use and protect future transit options.

Regarding the Purple Line, they offered this analysis (see diagram to the right).

This is current north-side service. As the sachems pointed out, because the Purple Line operates as a local after it gets to Belmont, it loses on the second half of the trip what it gains on the first half, and so gets downtown just two minutes faster than the local. Obvious question: What's the point of the express?

Adding a couple transfer stops to the Purple Line at Loyola and Wilson—as proposed in several RPM alternatives—would make the "express" trip slower still if the Purple Line continued to enter the Loop via the elevated (see left)


The sachems had a better plan, which was similar to the one we'd come up with at the Straight Dope (bottom right diagram).


Points of interest:

• The Purple Line entered the Loop via the State Street subway.

• The sachems objected to our idea that the Purple Line should simply exit the subway south of Roosevelt and pause on the elevated tracks before heading back north. They offered technical arguments (the tracks were on an incline), but basically, we suspect, the clumsiness of our solution offended their sense of aesthetics. Instead, they felt the trains should terminate at a station. While they reached no consensus on location, sentiment seemed to favor building a new stop at Cermak on the Green Line, in a fast-growing area near McCormick Place that could stand some train service.

• The main thing was, the Purple Line Express got downtown six minutes faster than now.

Doubtless some will now ask: You're going to all this trouble for six minutes?

This is a short-sighted view. Consider the chart to the right.


Pretty simple story. City ridership, as pointed out earlier, is up; suburban ridership is down. Entering traffic on the Purple and Yellow lines has fallen by a third—since 1987 these two lines between them have lost two million annual riders. Why? It's not as though the north suburbs have gone to the dogs. Downtown Evanston in particular has boomed. But service frequency in the suburbs and everywhere else was sharply cut back during the 1990s, when el ridership fell to the lowest level since the 1930s.

City traffic eventually recovered and service was mostly restored. In contrast, in the suburbs it remains pathetic. Today, for much of the day Purple and Yellow line trains run every ten to 15 minutes. As previously indicated, a few Purple Line Express trains operate every seven to eight minutes during rush hour, taking 52 minutes to get downtown from Linden.

Evidently that was enough to persuade people to find alternative ways of commuting. Today, fewer than 10,000 people board in Evanston on an average weekday, and fewer than 2,700 in Skokie—setting aside city boardings on express trains, the Purple and Yellow lines have the lowest ridership of any el branch. Granted, to assume ridership would return to its former levels if service were better entails a leap of faith; an argument can be made for sharply curtailing or even abandoning suburban operation. Since that's politically unlikely, the most sensible course is to plan for a service that more people might actually use.

Or so both we and the CTA sachems thought. We asked: Are you saying that between us we've developed a practical plan?

Not yet, said the sachems. The scheme had a technical problem—and here's where we truly began to understand how complicated the railroad business was.

If Red and Purple line trains left Howard every 2½ minutes (as Cecil had originally proposed, and which had been the working assumption up till now), the system would choke. The pinch point, we learned, was the Howard station itself. The main impediment was the two-car Yellow Line, which drops passengers at the same platform as the Red and Purple lines before shuttling back to Dempster. Getting the Yellow Line out from underfoot to let the other trains pass requires complicated switching. Attempting it with departures every 2½ minutes would lead to intolerable delays.

You're kidding, we said. You're telling us the iceberg that's going to sink this scheme is that every ten minutes the CTA has to make room for a two-car train?

Not exactly, we discovered; at least it wasn't the only iceberg. We'll spare you the details of a very long discussion, but we came up with a solution that met all objections. Here's the gist:

(1) Another pinch point was "Clark Junction," the spot just north of Belmont where the Brown Line splits off from the Red and Purple lines. Northbound Brown Line trains briefly block all the tracks when making the turn. The resulting congestion means the closest practical spacing between Red and Purple line trains is every three minutes, not every 2½.

(2) However, if trains departed Howard every three minutes, as it seemed clear they must, the two lines together wouldn't have enough capacity for all the passengers boarding south of Belmont.

(3) One solution was adding more Brown Line trains. However, you couldn't appreciably increase the frequency of departures from the terminal at Kimball, because Kimball yard has been at capacity since Brown Line trains were lengthened from six cars to eight, and expansion in the densely built-up Albany Park neighborhood would be a major project.

(4) Yard expansion wasn't necessarily the most cost-effective solution anyway, because more Brown Line trains weren't needed north of Belmont—they were needed from Belmont south, where the busiest stations were. A better approach was to "short-turn" some northbound Brown Line trains at Belmont—that is, send them immediately back downtown rather than out to Kimball. This could be accomplished on a short stretch of main line track no longer needed by the Purple Line Express.

(5) That was just a temporary fix, though. The long-term solution, inevitable if ridership continued to grow, was to build a combination "flyover" (up-and-over crossing, as with an expressway interchange) and turnback track at Clark Junction. That would eliminate the bottleneck and allow extra Brown Line service from Belmont south.

So there you have it. What would all this get us? Faster, more frequent service for most riders boarding north of Belmont; more trains and less crowding from Belmont south; rail service extended to the far south side.

Would the plan work? The sachems thought it was worthy of serious study—and as far as they were concerned, that was the point. The purpose of the exercise wasn't to prove that reconfiguring the Purple Line was the only way to make the el work better. Rather, it showed the importance of looking at the big picture, and thinking about future improvements now.

We put the question to the CTA: If funding were available, would the agency be open to a study in the near future of the entire Red/Purple corridor, from Linden to 130th? Here's the response we got:

"This project is one part of the overall plan to repair, upgrade and extend the entire Red Line. CTA continually examines ways the agency can improve the service we provide customers. The RPM study is intended to bring the capital assets in the project area to a state of good repair. It only includes the area between Linden and Belmont."

Comments (104)

Showing 1-25 of 104

Add a comment

Add a comment