Why the ARTA-RTC report numbers are so big

The Rails-to-Trails Conservancy report – commissioned and paid for by Adirondack Recreational Trail Advocates and released last summer – is flawed in many areas, but I’ll focus on the summary of rail-trail surveys on page 54.

The first table (see PDF) provides statistics on six trails that are the basis for ARTA’s projections regarding trail usage and income. The report claims these trails are “similar to the proposed Adirondack Rail Trail.” These claimed similarities are crucial to the validity of the report’s conclusions. As it turns out, only three of these trials are even vaguely similar to the proposed trail. The other three should be eliminated from the sample. They are:

-Heritage Rail Trail (394,823 annual uses, 21 miles) – Begins in downtown York, Pa., and proceeds south through rural woods and farmlands to the Maryland state line. The York-Hanover metropolitan area population is 398,416. (See more at Torrey C. Brown Trail below.)

-Great Allegheny Passage (612,991 uses, 150 miles) – Longest rail trail east of the Mississippi River. Southern terminus is Cumberland, Md. (metro population 103,299). Northern terminus is Pittsburgh, Pa., city center (2.3 million metro). Though the trail does cross some rugged and heavily wooded country, long stretches of it are crisscrossed and paralleled by numerous public roads and highways. Public access points on the Adirondack route are relatively few, and the nearest million-plus metro area, Montreal, is two-and-a-half hours away.

-Torrey C. Brown Trail (800,000 uses; 21 miles) – Trail is wholly within Baltimore County, Md. Northern terminus connects to the Heritage Rail Trail. Southern end is in Cockeysville, a Baltimore suburb. The Baltimore metro population is 2.7 million, part of the larger Baltimore-Washington combined statistical area’s 8.4 million residents. This trail, together with the Heritage Rail Trail, makes a single recreational corridor between the Baltimore metro area and York, Pa.

Inclusion of these three trails greatly inflates the average “annual user visits,” which is fundamental to ARTA-RTC’s projections, and by itself should invalidate the report. (The Virginia Creeper Trail could easily join this group of inappropriate examples but at least is rural at its east end.)

Note: The following analysis has been verified for accuracy by two college math instructors, one of whom is an ARTA member.

The summary table (see PDF) from page 54 of the report has eight columns and seven rows. Let’s refer to the columns as A to H and the rows as 1 to 7. Looking at the numbers in this table, we can see that the result of (B x C x F) + (D x E x F) = “Total annual expenditures,” column “H.”

Look at line 7, the averages. The figure in column “B” is wrong. It should be $16.35, as it is in the “daily expenditure” table on page 54. Column “C” is not calculated but comes to 75 percent if we use the average of the five knowns for the “NA” in line 5. Column “D” is $61.83. Column “E,” after correcting the erroneous “0.05 percent” in line 6 to 0.005 percent, is 20 percent. Column “F” is 354,135. Using the (B x C x F) + (D x E x F) formula, which gave us the totals in column “H,” results in $8,721,814. The lower figure of $7,085,077 at the bottom of column “H” is a simple average of the six amounts in that column.

Excluding the three urban-influenced trails results in total annual expenditures of $3,237,458. A simple average of those three column “H” totals is $2,938,328. These numbers are probably much more representative of what an Adirondack trail would bring in. But ARTA-RTC claims an Adirondack trail would earn at least six times as much. Even though the “percent overnight” column averages 20 percent for all six trails (22 percent for three), the report assumes overnight visitation of 69 percent for the Adirondack route.

Column “G” of the summary table is “Avg non-local user expenditures/day.” For every trail in the summary table, it is the sum of columns “B” and “D.” Thus ARTA-RTC assumes that EVERY non-local user stays overnight and EVERY non-local user spends what a local user spends in addition to what an overnight user spends. The average of the expenditures in column “G” is shown as $81.02. It is also shown as $86.02 on page 8. The average is actually $81.52, but we will use ARTA-RTC’s $81.02 for simplicity’s sake.

Look at the third table on page 54, “Estimated annual trail user visits.” ARTA-RTC takes the 354,135 average “annual user visits” from the summary table, rounds it to 354,000 and multiplies that by 69 percent. Where they get that 69 percent is unexplained. This table puts average non-local visits at 244,260 (69 percent of total). ARTA-RTC takes $81.02 from column “G” of the summary table and multiplies it by 244,260. Thus they get their astonishing income of $19.8 million. Applying this formula, expressed as F x G x 69 percent, to the Ghost Town Trail results in $4,779,266 – quadruple the total expenditures shown for that trail. But the problem is not just two formulas. The mysterious 69 percent interestingly yields almost the same result in the formula for the example trails as it does in the formula for the Adirondack trail. The 69 percent is an arbitrary figure that bloats the projected income regardless of what formula is used.

I spoke with Tom Sexton, director of RTC’s Northeast Regional Office, about their selection of example trails. I asked him why – with well over a thousand rail-trails under their purview – did RTC choose six trails that were so unlike the one proposed by ARTA. He explained that the number of trials for which RTC has user surveys is still comparatively small, “120 to 150,” and many of those are entirely in urban or suburban areas. He said they selected the trails that were “most similar” to ARTA’s idea. But to say trails terminating in Pittsburgh or Baltimore or even York, Pa., are similar to an Adirondack trail is an unjustifiable stretch.

Later, I questioned Carl Knoch, lead author of the ARTA-RTC report. I asked him why determining total annual expenditures on the proposed Adirondack trail utilized an equation different from the one that was used to determine income for the example trails. I also questioned the use of 69 percent as the percentage staying overnight. He responded that they had “assumed that most of the users would be non-local” and would be staying overnight because of the distance they would have to travel to get to the trail. I asked him how “non-local” was defined. He said that for this trail it would be “more than an hour’s drive.” He then added that “the Pine Creek trail is most similar to the Adirondack trail” in its remoteness. I pointed out that even the Pine Creek trail has an overnight usage of only 26 percent.

And this doesn’t even address the trails ARTA-RTC chose to determine construction costs (page 34), one of which uses a much cheaper surfacing material and one of which hasn’t been built yet. Excluding those two from their projections results in costs of $30,962 more per mile than ARTA-RTC claims as a representative average.

It should be clear that any decision-maker taking this report’s conclusions at face value risks being misled.

Phil Gallos lives in Saranac Lake.