The Price of Skiing

Last year I was a senior Economics major at Colby College.  This meant that I had to do a final project that involved a lot of my own research and said something new about what I found.  I decided to do my project on skiing.  Specifically, I wanted to see if there was any relationship between the quantifiable factors that describe a given cross-country ski center and the price that that ski center charges.  Without getting too deep into the details of the project, this article seeks to describe what I did and what I found.

I started off with a sample of cross-country ski centers.  I decided to use all the centers I could find in New England and New York.  I then dropped the ones that run without charging for trail passes.  This left me with 80 different centers across the northeast.  I then came up with any variable I felt may play a role in the pricing of a center.  There were 12 variables I came up with.  I collected data on all 12 for all 80 ski centers.  There were two types of variables.  The first group were normal ones that lend themselves to being quantified, including total kms, average annual snowfall, average local income, other centers within 20 miles, other centers within 100 miles, population within 20 miles, and population within 100 miles.

The second group of variables accounts for things that are harder to quantify.  They are just answers to yes or no questions, such as, “is there a retail shop?”  If yes, then that center has a “1” in that category.  If no, the center has a “0”.  There were five of these “amenity variables”.  They were: is there a high-quality grooming machine (Pisten Bully, Bombardier, etc. ), is there a retail shop, are there lessons offered, is there on-site lodging, and is there a downhill resort nearby.  With those twelve statistics as well as the price of each center established, I was ready to dive into the analysis.

The first thing I did was add two more variables that were manipulations of the variables I already had.  One was the population within 20 miles multiplied by the local income per capita.  This was an attempt to have a variable that represents the total amount of money within 20 miles.  The second manipulation was a variable that I called “quality factor.”  It was the five amenity variables added together.  So if a center has a Pisten Bully, offers rentals, lessons, and lodging, and has a downhill resort in town, then that center gets a 5 for the “quality factor”.

The next part is kind of fun, in a sort of “I’m a lame economics nerd” kind of way.  You can try it yourself and see what you come up with.  Which of these variables do you think determine the price?  All of them?  Some of them?  None of them?  There’s a lot more that goes into pricing, so maybe this whole thing won’t work.  Or maybe, all of these variables help to perfectly determine price for all the centers.

As it turns out, three variables are pretty good at determining price.  I’d suggest you try again to guess what they are.  Here’s the answer.  The most important determinants of price are: total kilometers (classic+skate), number of centers within 20 miles, and the “quality factor” that I described.  Those three are consistently significant, and all the others are consistently terrible.  After some further analysis, I found that having a downhill resort in town actually had no effect, so the final equation I came up with used three independent variables, including total kms, centers within 20 miles, and a quality factor that is lessons+retail+grooming+lodging.  That equation looks like this:

Price=8.683+.024 x Total Km+.534 x Centers w/in 20 miles+1.025 x Quality Factor

This means that a center can charge (or rather, does charge) $8.68 even if they offer nothing.  For each extra kilometer of skiing they have, classic or skate, they can charge another 2.4 cents.  Then if they count up all the other centers within 20 miles of their own, they can charge an extra 53.4 cents for each one.  Then if they have a big grooming machine, lodging, lessons, or a retail shop, they can charge an extra dollar and 2.5 cents.

This all makes sense.  The number of kilometers has a direct effect on both how nice it is to ski at a place and also how much work it takes to maintain the trails.  Both of these should drive the price up.  The number of centers within 20 miles seems to work as a good measure for how popular skiing is in that area and how good the local terrain and snow might be.  Where skiing seems to be best in the Northeast, there are many ski centers-places like Stowe, VT and Jackson, NH.  So in those places, prices can be higher.  Finally, for each of the 4 amenities I mentioned, there is a value there that makes the ski center more accessible and higher-quality.  The services also increase the amount of work it takes to run the center.  This will increase the price as well.

The most important question to ask now is whether or not the equation worked.  Sure, there were a few significant variables and things seem to make sense, but how good is the equation at actually predicting the price?  The answer, predictably, is that the equation is ok, but not perfect.  The range of prices is from 8 to 22 dollars.  The equation is off by an average of $2.30.  It predicted over a quarter of the center’s prices within a dollar, almost half within 2 dollars, and almost three quarters within 3 dollars.

The final question to answer here is what does all of this mean?  First, this formula is probably a pretty good resource for ski centers to use.  Second, it shows how centers can change their prices if it changes what they offer.  Finally, and most importantly, it shows a lot about the cross-country skiing experience.  It shows what is important (kms, lessons, lodging, grooming, etc.) and what is not (snowfall, local income, population, etc.)  Take from that what you will.  My most important conclusion comes from two findings.  First, the centers within 20 miles is important.

It’s not competition that drives price down though, it’s an indication of a good skiing culture and good conditions.  Where the people like to ski, there are ski areas.  And where people like to ski, the prices can be higher.  The second important finding is that income and population do not matter.  It does not matter if there are few people with low per-capita income.  If the people like to ski, they will ski.  Just as Charlie Yerrick, director of ski operations at Trapp Family Lodge, told me in an email, “[Price is determined more by the local culture than the income.  People who like skiing ski whether they have money or not.”

You can download my entire paper in PDF form using the link below.

MTB Thesis

Matt Briggs hails from Eastern Massachusetts where he started skiing as a freshman at Concord-Carlisle High School and joined the CSU club team as a junior. Matt was a 4-year carnival skier at Colby. His career highlights include two top-40s at nationals in 2008, and ten top-10s and three top-5s in EISA carnival racing, as well as a win and three 2nds in Eastern Cups. He is happy to continue skiing in New England with the Craftsbury Green Racing Project. This year he hopes to qualify for the US squad for U23 World Championships. Matt is also a part owner of a company, lobsterjoke.com, which is the world’s leading lobster-related joke company.