Thursday, December 26, 2013

Silver Service and Florida Intrastate Travel

Just before Christmas, the National Association of Railroad Passengers (NARP) delivered an early gift to those of us who like to look into the various statistics behind Amtrak with the 2013 ridership data. Being a bit of a data junkie, I immediately took a dive into it and came up with some rather surprising information regarding the distribution of travel within the state of Florida. The perception which I had always had of the two Florida serving trains, the Silver Star and the Silver Meteor, were that they were, to a significant degree, used, or intended at least, for travel from the Northeast to Florida and back again. The truth, however, is quite different: A significant and possibly the greatest single segment of travel on these two trains is actually Florida intrastate travel.

Struck by the fact that Florida city pairs, all to Tampa, were the primary city pairs for ridership on the Silver Star, I decided to look deeper into the ridership according to distance recorded by each station stop in Florida. As NARP has presented the data, the total number of boardings and alightings is given, as well as a percentage breakdown by distance, for every hundred miles of travel. I took this information, put it into a spreadsheet, and then calculated the distance to both Jacksonville, the most northerly Florida station, and the nearest out of Florida station, Jesup, Georgia. This information was either readily available from the station data sheet when Jacksonville was one of the top city pairs or was calculated from the current Silver Service timetable and the distance to the next station on the route which did have that information readily available. I then proceeded to sum up all of the boardings and alightings which existed in bands which were indisputably within the state of Florida. This method resulted in a certain degree of undercounting: There are no boardings or alightings counted from Jacksonville because Jesup lay only 92 miles away; similarly, ridership between Tampa and Jacksonville is not counted, despite being a top ridership city pair, because Jacksonville and Jesup lay within the same distance band at 203 and 299 miles distance respectively.

That conservative underestimate gives us a figure of at least 357,162 boardings and alightings, which translates into 178,581 passengers solely traveling within the state of Florida. With a combined ridership of 770,586 for the Silver Star and Silver Meteor, this means that at least 23% of their combined ridership is from travel solely within the state of Florida. In fact, at least 42.8% of all Florida traffic is within the state of Florida. Considered as a separate service, Florida certainly wouldn’t be one of the top performing routes, ranking between #36 Washington-Lynchburg and #37 Piedmont (though I suspect ridership levels on par with the Palmetto if all Florida intrastate travel were counted), but it would be a respectable performance level nonetheless, and all the more so for how poor the current service is for intrastate travel. The fastest trip between Miami and Orlando, one of the major rail corridors, is 5 hours and 3 minutes (Southbound, North is 5:45), aboard the Meteor, arriving at Orlando at 6:55pm and Miami at 1:23pm. This compares with a driving time of three and a half hours, assuming no major delays. Between Tampa and Orlando, though the train takes an extra hour than unobstructed traffic, and is probably on par with normal freeway travel before consideration of the last mile (2:03 vs 1:18), it suffers from similarly poor timings and only a single frequency in the midday and evening.

Obviously, this inclines one to think that All Aboard Florida will do quite well when they begin service, especially if future extensions are built to Jacksonville and Tampa. It also implies that the lack of an intrastate Florida train has been a severe mistake by Amtrak and the state of Florida. There is definitely a market to be served, yet all that they are offered is a pair of long distance trains which run in close succession to each other.

Certainly there appears to be a will to spend, and to spend heavily, for passenger rail service in Florida. Twice billions were appropriated for building a high speed rail service, though twice again cancelled by the governor, and currently a billion dollars, 25% from the state and 25% from local counties, is being used to purchase and construct the SunRail commuter line in Orlando. Further back, between 1982 and 1984, the state of Florida funded a once daily train between Miami and Tampa, canceling it when it failed to maintain a 60% operating ratio as required under the enabling Florida law. This funding requirement is not as ambitious as it may nowadays seem; Amtrak’s accounting systems were different back then and under them the Pacific Surfliner had farebox recovery levels of 59.1% and 76.4% for 1982-1983 and 1983-1984 respectively (page 24). It does however, seem to be an experiment foredoomed to failure by lack of frequencies and the short timeframe in which to build up ridership. At an expense of only $2.1 million over those two years, it also does seem like quite the odd penny to pinch.

The failure of the Silver Palm may have soured the state of Florida, unreasonably in my opinion, on supporting an intercity train, but it should not have done so for Amtrak. Just a few years earlier, for about the same million dollar per year cost as the Silver Palm, Amtrak added, at their own expense, an additional round trip between Los Angeles and San Diego; this despite the fact that there were already three state supported round trips on the route (and three more that were not supported by the state). Ridership in the waning months of the second year showed increases of 66% over the previous year and it is reasonable to suggest that it could have continued to increase to the point of no longer requiring a subsidy.

Indeed, I would go so far as to suggest that strong consideration should have been made to truncating one of the Silver Service trains at Jacksonville, using its equipment for a Florida intercity service, while transferring its sleepers to the surviving train. This would have freed up four sets of equipment for use within the state of Florida while consolidating certain costs for the sleepers onto just one train, possibly allowing for greater efficiencies. Given the example of the second Silver Palm, which ran from New York to Miami, it seems likely that this would have been a ridership positive move.

First running November of 1996, ridership on the Silver Palm appeared to completely cannibalize that of the Silver Star and Silver Meteor. In a year of general increase, the Silver Palm began with 188,000 riders while the Star and Meteor collectively lost 174,000 riders (page 37). With how important intrastate ridership is to the contemporary Silver Service, it’s no wonder that the Silver Palm, which had extremely poor timing for state service, did no great miracles in increasing ridership levels. Southbound it left Jacksonville at 1:56 in the morning, skipping Orlando except for a bus connection (though at that hour, it’s doubtful there would have been many riders bound for DisneyWorld), and arriving at Tampa at 6:47 in the morning. Five minutes later it left for a 12:07 arrival into Miami.  At 5pm it left Miami, reaching Tampa at 10:13, skipping Orlando once more, and arriving at Jacksonville at 2:33am.

These were terrible times for an intercity service to run and they are worsened by the fact that the Silver Palm originated in New York, making all of its southbound times somewhat theoretical and dependent on keeping good time for nine hundred miles earlier; something that those stuck relying on the Empire Builder have known is easier said than done these past few years! Unsurprisingly, when the Silver Palm, now without sleeper or diner service and renamed the Palmetto, was cut back to Savannah, Georgia, there was only a minor change in ridership upon the Silver Service routes (combined in Amtrak’s ridership figures).

Why this is so is fairly simple of course: People want to travel only a few hours and they want to do so at reasonable times of the day, with an emphasis on day. Incredibly few intercity trips of any type are taken in the middle of the night or the earliest hours of morning; even commuter traffic is relatively rare at this time. Train stops at these times will not be well patronized, as one can clearly see in Ohio. Furthermore, these trips to be taken by train, they need to be competitive with other modes of travel, such as by air. This is, of course, ignoring the ability that speed has to induce ridership; many more trips were taken between Los Angeles and Chicago when air service began than were ever taken by train between the two cities.

If we look at NARP’s data, for all long distance trains, 31.2% of the trips taken were under three hundred miles in duration and an additional 18.8% were less than 500 miles in length and only 14.9% were over a thousand miles in duration. To a certain extent the distance which passengers travel is inflated thanks to the fact that on busier corridors, such as New York to Washington and New York to Albany, the long distance trains run receive or depart only, not allowing trips to be booked within those distances and requiring passengers to use corridor trains instead.

Now, it is suggested by some that the advantage of the long distance train is that it allows multiple of these corridors to be undertaken by a single train with the added bonus of some long distance passengers to help subsidize the travel. This comes at the downside, however, of giving many communities and even important corridors poor hours of service, especially when there is a misguided focus on creating a trip which best serves an end to end run rather than the more typical and desirable journeys in between (which may involve one end, but not both). As well, as I mentioned earlier, the long distances introduce significantly more potential for major delays to passengers as well as increased risk of train cancellation due to work on a portion of the line. Foamers may joke about how they should pay extra for the “pleasure” of riding longer in a delayed train, but passengers who are stuck waiting several hours, especially with Amtrak’s employee culture of refusing to give information about delays, for the single daily train which serves them will have rather a different view of the situation.

There is also the factor of equipment utilization. The current Silver Star and Silver Meteor use, not counting maintenance reserves and protect equipment, sixteen locomotives and seventy-six cars of various types. The Pacific Surfliner uses half that figure. Could not significant gains have been made in both ridership and revenue by having only a single train from New York to Miami while the rest of the equipment provides multiple frequencies? Certainly it could not match the frequencies of the Surfliner, not when Tampa to Orlando crawls along at 25 miles per hour (though that figure could easily have been improved with improvements to the line), but several frequencies, each conducive to more riders, could reasonably have been made, even across the whole extent of the state from Miami to Jacksonville.

Now, of course, it is far too late for Amtrak to consider such a thing. All Aboard Florida has committed to providing a significant number of frequencies with significantly improved service between Miami and Orlando, predicting three million riders and nearly twice as much revenue as the Silver Star and Silver Meteor combined, despite using less equipment. But Florida is hardly the only state where intrastate ridership makes up such a major portion of total existing ridership. Even if Amtrak is unwilling to shorten or cancel long distance routes in order to create spare equipment for more frequent corridor routings, it should still examine the potential for new corridor service that the existing long distance trains have shown may be popular that it might proactively offer this service to the states. It really makes no sense for Amtrak to be so incredibly passive and to be reliant upon the various states and local communities taking the initiative every time. It should be proactively offering plans for improved service and keeping those plans updated so that a change in the political weather does not risk losing its wind before anything is capable of starting.

Wednesday, December 25, 2013

Merry Christmas

IN THE 5199th year of the creation of the world, from the time when in the beginning God created heaven and earth; from the flood, the 2957th year; from the birth of Abraham, the 2015th year; from Moses and the going-out of the people of Israel from Egypt, the 1510th year; from the anointing of David as king, the 1032nd year; in the 65th week according to the prophecy of Daniel; in the 194th Olympiad; from the founding of the city of Rome, the 752nd year; in the 42nd year of the rule of Octavian Augustus, when the whole world was at peace, in the sixth age of the world: Jesus Christ, the eternal God and Son of the eternal Father, desiring to sanctify the world by His most merciful coming, having been conceived by the Holy Ghost, and nine months having passed since His conception was born in Bethlehem of Juda of the Virgin Mary, having become man.
THE NATIVITY of Our Lord Jesus Christ according to the flesh.