MIAMI — United Airlines’ Washington Dulles hub faces an uncertain future. Structural shifts in the Washington DC air travel market and the declining economics of 50-seat regional jets (RJs) have crippled its once robust domestic operation. Simultaneously, its niche in the combined United route network has shrunk for both domestic and international destinations due to the strength of the legacy Continental hub 212 miles up I-95 in Newark.

Screen Shot 2015-11-03 at 2.12.23 AMThe high-level picture for the Dulles hub is worrisome. Back in summer 2011 (the week of June 23), the hub was at roughly 300 daily departures to 115 unique destinations, including roughly 55 daily departures on turboprop aircraft and 95 on 50-seat RJs. By the summer of 2015 (week of June 20), the hub had shrunk to roughly 230 daily departures to 104 destinations, with just 20 daily turboprop departures and roughly 50 on the 50-seat RJs. The capacity figures are not as startling because United has added a lot of mainline/70-seat RJ flying but the overall trend is still painful. And while United both added and removed destinations over the interceding four year period, the additions tended to be low frequency vacation flights while the reductions were a mix of long haul and domestic high frequency flights. The overall changes in destination profile for the Dulles hub are as follow:

Structural changes in the Washington D.C. market have made Dulles a less attractive hub

The past five years have witnessed several major shifts in the Washington D.C. air travel market, each of which worsened the relative value of Dulles for passengers in the region. The first of these major shifts was the Delta-US Airways slot swap, which gave US Airways (now American) 42 additional daily slot pairs at Reagan. The major impact to United of the slot swap itself was indirect, as only 3 new destinations that US Airways added (Jacksonville, Ottawa, and Savannah) and 1 additional frequency to Hartford actually overlapped with United flights offered at Dulles.

But the indirect effect of the slot swap was to incentivize more passengers in the DC area to shift loyalty from United to US Airways. Moreover, US Airways and Delta were also required to divest 8 slot pairs at Washington Reagan, which were auctioned off to JetBlue and increased LCC competition at the airport. JetBlue used the slots to add multiple daily frequencies to Boston, Fort Lauderdale, and Orlando as well as launch new daily service to Tampa, all of which competed with United at Dulles.

Even more damaging was the increase in beyond perimeter exemptions at Reagan in May of 2012, when 4 new perimeter exceptions were granted and the law was changed to allow existing carriers to shift four slot pairs from within perimeter to beyond perimeter routes. These shifts resulted in new flights/frequencies from Reagan to Portland by Alaska Airlines, Austin by Southwest, San Juan by JetBlue, San Francisco by United and Virgin America, Los Angeles by American (and eventually from pre-merger US Airways which initially added San Diego), and Salt Lake City by Delta. These new beyond perimeter routes broke or weakened the O&D dominance for Dulles on these routes; traffic and revenue which United held the lion’s share of. In fact, Delta’s second daily Salt Lake City – Reagan flight was a major factor in United eliminating Salt Lake City service

Less than a year later, US Airways and American announced their merger, which immediately provided an incremental boost to an already strong US Airways by strengthening the beyond perimeter portfolio at Reagan (adding two daily flights to LAX, the most important beyond perimeter destination) and combining operations at Dulles and Baltimore-Washington to build more critical mass at those two spokes for DC O&D. But the larger effect was once again divestments, which increased LCC competition at DCA sharply. The new American was required to divest 44 slot pairs (basically the entire pre-merger operation for American at DCA), which were given to JetBlue, Southwest, and Virgin America.

This injected a ton of competition to big domestic origin and destination (O&D) airports from the preferred airport for the DC market, further weakening the O&D that United could pull to IAD. JetBlue won 12 slot pairs, using them to add or grow service to Charleston (SC), Hartford, Nassau, Tampa, Fort Myers, and West Palm Beach. Southwest was the big winner, as it purchased 27 slot pairs and deployed them for service to Chicago Midway, Nashville, New Orleans, Tampa Bay, Akron, Dallas Love, Indianapolis, St. Louis, and Columbus. Finally, Virgin America won 5 slot pairs, which it used to launch service to Dallas Love using gates that it also won in the US Airways-American merger divestment lottery.

Assessing the O&D data for these markets

In order to get a complete picture of the structural shifts in the DC market, we took a look at the O&D demand and average fare metrics, as well as some derived statistics using masflight’s excellent tool to assess how the market changed between 2011 and 2014. The pool of markets we analyzed included markets that gained new nonstops to Reagan due to the slot swap, the resultant divestitures, the divestitures after the US Airways-American merger, and the additional beyond perimeter exceptions granted in 2012. We looked at Q3 figures for each year, and that data is presented in the table below.

*Note* – The Chicago (ORD + MDW), Cleveland (CLE + CAK), Dallas (DAL + DFW), Houston (HOU + IAH), Los Angeles (LAX + LGB + SNA + BUR + ONT), Miami (MIA + FLL + PBI), and San Francisco (SFO + OAK + SJC) data look at city, not airport markets for the purposes of this analysis

*Note* – For the Dallas market only we looked at Q4 data instead of Q3 data because Southwest and Virgin America’s nonstop services to Dallas Love weren’t launched until October 2014.

Source: masflight
Source: masflight

At a high level, the data is illustrative of the broader trends in the Washington D.C. air travel market. For most of the markets, O&D demand at Dulles is stagnant or down, whereas traffic has pretty much exploded at Reagan. The one outlier to this trend is Omaha, where both Delta and Frontier had service (since eliminated) in 2011. Across the board, the share of Dulles in terms of overall traffic in the DC market has declined, and in the beyond-perimeter markets, it has also lost its once robust fare premium to places like San Francisco or Salt Lake City.

United’s restructuring didn’t help either

Accompanying a downturn in the airport’s broader fortunes was a shift in the role of Dulles in United’s network. Prior to the United-Continental merger, Dulles was the primary European gateway and connecting complex to Northeast for United. But today, Newark, which has stronger O&D to most domestic and international destinations serves those flows more and more, so IAD is retrenching down to focus O&D. Fleet shifts have also hampered the Dulles hub, in particular the drawdown of turboprop and 50-seat RJ flying.

Indeed most of the domestic destinations lost since 2011 were props to small cities in the Northeast. The lack of props means that there is no viable way to serve places like State College, Beckley, or Altoona, so these cities are lost from the Dulles network. This creates a cyclical reaction, as the loss of these routes lowers feed for United’s domestic ops, which in turn reduces viability for some larger domestic destinations as they are now more reliant on O&D. Dulles domestic ops were also dependent on 50-seat RJs as a way to offer frequency and capture O&D. But the economics of these have collapsed in recent and United has shifted to a model of larger aircraft with lower frequency, which also means less O&D.

Dulles is caught in a vicious cycle

Hubs are driven by a mix of O&D and connectivity. But at Dulles, both of these have been hit for United. Structural changes have pushed down O&D and the number of markets served, both of which reduce connectivity. On the domestic side, every lost destination reduces connectivity, which means the O&D threshold for remaining destinations is higher for sustainability purposes. But that is being hurt by structural changes in the market, and so the overall domestic destination portfolio for United at Dulles declines. Next you add international into the analysis.

Connections over Newark are now more convenient (or always were). Each connection that flows over Newark means that the O&D threshold for international flights at Dulles increases. When aggregated, this means that some international destinations are no longer sustainable or must be downgauged to better serve O&D traffic The loss of international markets to feed means even more domestic destinations become superfluous. And the net result is a much smaller hub, domestically and abroad.

There is more shrinking on the horizon

On the whole, there is still some capacity that is likely to be pruned by United. But Dulles isn’t Cleveland, Cincinnati, or Pittsburgh. It isn’t even Phoenix. Washington D.C. is a massive global city that throws off tons of international O&D traffic. This gives the operation a staying power that other somewhat redundant hubs don’t have. Over time, we expect United’s hub at Dulles to converge on an international O&D focused operation with service to key business domestic destinations to retain corporate contracts dependent on DC and offer some feed for the international routes.

In other words, United’s Dulles operation will look a lot like that of the new American at New York JFK. This does mean that the small cities – Charleston, Albany, Columbia, Charlottesville, Greensboro, and the like – will likely be dropped. So overall, this works out to between 150-175 daily departures, mostly on mainline aircraft with a similar international destination profile to today give or take a Manchester. In the long run, you’ll likely see more international routes added as the DC metro grows, as well as a few targeted domestic routes added if the opportunity presents itself.