It’s a conundrum that’s puzzled me since I first began looking into the cruise industry in Atlantic Canada: why do we rely solely on numbers from the cruise lines for our economic impact calculations?
Having read the work of cruise-skeptic Ross Klein, a professor at Memorial University in Newfoundland, I knew that those numbers — how many passengers and crew members disembark in any given port and how much they spend while ashore — were dubious. I knew from speaking to officials at the Port of Sydney that all our calculations for the value of the industry to our local economy are based on these cruise line numbers because they’re “the only numbers we have.” I knew that the consultants who did the due diligence on the second berth for Sydney also relied on industry numbers (and they still had a job making the second berth seem economically viable).
No one, it seemed, worried that we were making local spending decisions — involving millions of dollars that could be spent on other infrastructure — based on “industry standard” figures. Why wasn’t somebody trying to find some other numbers?
And then, this morning, this:
Acadia University economics prof Burç Kayahan, with funding from Memorial University’s Collaborative Allied Research in Economics (CARE) initiative, has done a study of:
…the economic impact of cruise tourism in Atlantic Canada by taking indirect and external factors that are associated with costs and benefits of this type of tourism. The…study covers the four main ports in Atlantic Canada: Saint John, Halifax, Charlottetown and St. John’s.
Now, before you get upset about being left off that list, let Kayahan explain why they didn’t include Sydney. He told me in an email:
Originally, we wanted to include Sydney in our study as well, however, once we began working on our surveying schedule it became apparent that our surveying costs would exceed our budget if [we] surveyed all of the five ports. Unlike the Cruise Association study, we did not want to focus our surveying time frame into a couple of months, so you can imagine how costly it becomes to survey through the entire cruise season in a given port.
Moreover, ships that come to Atlantic Canada frequently visit multiple ports (including Sydney) so we thought that Sydney would be implicitly included in the study as well, although it is logical to question whether a passenger visiting Halifax or Charlottetown spend[s] the same amount while visiting Sydney (or Cornerbrook for example).
Personally, I think that this study is applicable to us simply because it shows there is a gap between industry spending estimates and estimates based on local research. Of course, Kayahan had a pretty good idea that was the case from the outset:
Cruise associations report impact estimates in the millions of dollars, however there is a substantial body of literature that point out deficiencies of such self-reported statistics by cruise associations.
Belly full, purse closed
Kayahan’s study — “Economic Impact of Cruise Tourism in Atlantic Canada” — actually begins with a review of that body of literature, and it sent me dashing all over the interwebs in search of more detail. For example, it states:
- Average visitor expenditures are often based on theoretical expectations. Cruise passengers spend significantly less than overnight tourists even after excluding overnight expenditures. (Larsen et. al., 2013)
- 20-40% of the passengers do not leave the ship during a stop-over. (Stavanger, 2012)
I found both those facts in the wonderfully named “Belly full, purse closed,” a study by Svein Larsen, a psychology professor at the University of Bergen in Norway. Larsen (et.al.) conducted a comprehensive survey of tourism in Western Norway between 2010 and 2012, quizzing 8,000 tourists — 1,300 of whom were cruise ship passengers — about their onshore spending.
According to Physics.org:
Larsen’s research shows that the average cruise tourist on average spends about NOK 300 [CAN$47] a day onshore. Between 20 and 40 per cent don’t even leave the ship. Half of those who leave the ship spend less than NOK 250 [CAN$39] onshore.
In comparison, the average camping or hostel tourist spends twice this amount. The average family tourist, who stays in a hotel, tops the spending list with an average spend of NOK 1,000 [CAN$156] a day. This comes on top of paying for the hotel room.
Larsen had some thoughts as to why cruise ship passengers spent so much less:
The psychology professor believes that the low spend per cruise tourist is down to budget cruise offers on an all-inclusive package. A six-day cruise of the Norwegian fjords may cost as little as €220 [CAN$335]. The same journey by air travel and individually arranged accommodation will usually set you back several times that amount.
Taking a cruise is the new mass tourism, Larsen believes. – The tourists are fed onboard and even do their shopping onboard.
Also, the cruise industry needs to make money. It is not unusual for cruise tourists to be let offboard for a maximum of eight hours per onshore visit.
Kayahan also cites work by Klein that I’ve written about before, work related to the environmental impact of the cruise industry, which is “generally omitted/ignored” in economic impact calculations and the “market power of the cruise industry” and the “distribution of economic surplus between cruise lines and ports/local producers.”
Goal-oriented
To redress some of the deficiencies in the cruise industry numbers, Kayahan set himself the following two goals:
1. Conduct an independent study to estimate the economic impact of CT [Cruise Tourism] in Atlantic Canada in 2016:
- Average passenger/crew spending per port
- Total visitor spending using on-shore visits
- Cruise line spending in Atlantic Canada
- Direct expenditures generated by the CI [Cruise Industry]
- Total economic impacts using a regional I-O [Input-Output] model
2. Analyze other research questions: Determinants of passenger/crew spending and propensity to return as a regular (i.e. air/land) tourist, etc.
The second goal falls under the category of “future research;” this report focuses on the first goal, to accomplish which, Kayahan conducted visitor surveys — 3,050 in total — with cruise ship passengers and crew between April and October 2016 in Halifax, Saint John, Charlottetown and St. John’s, gathering information on demographics and visitation and spending patterns.
This information was compared to the industry numbers contained in the 2016 Atlantic Canada Cruise Association (ACCA) Economic Impact report.
Big spenders?
Kayahan estimated average per capita expenditures using a generalized linear regression model which he explains in detail in the report but which I am not Statistics-literate enough to a) understand or b) explain to you.
That said, it is all spelled out for those of you who do understand and can critique it, so that’s a plus.
What his calculations showed was a gap (sometimes quite startling) between industry estimates of passenger and crew spending and estimates based on his own surveys:
I am now going to explain the difference between an absolute difference and a relative difference because I just spent 20 minutes learning it and I’m not letting this newfound knowledge go to waste. The absolute difference between Cruise Association Estimates for crew spending in St. John’s and Kayahan’s estimated spending, for example, is $73.99-$20.79 or $53.20.
The relative difference between the two is a ratio, expressed as a percentage. First I’ll tell you how it’s calculated then I’ll tell you why anyone would bother:
You take $73.99 and subtract $20.79, divide the answer by $20.79 then multiply by 100. The result is 256%, the relative difference. (Go ahead, try it, it’s fun.)
The reason to note the relative difference as well as the absolute difference is that the absolute number doesn’t convey the magnitude of the difference.
As you can see, the “relative differences” between industry and study figures for spending range from 25% for passenger spending in Saint John, NB to 256% for crew spending in St. John’s, Newfoundland. That latter discrepancy was large enough to cause St. John’s Mayor Danny Breen to declare he didn’t believe it, to question the study’s methodology (I guess he knows more about statistical modeling than I do) and to announce the city was doing its own study.
Talk about hitting a nerve.
Here’s some other interesting results:
And here are Kayahan’s estimates for total direct visitor spending at the ports versus the industry estimates:
Where’s our study?
The study concludes the average cruise ship passenger in Halifax, Charlottetown and Saint John spends $60 while ashore while the average crew member spends $45 in those cities. In St. John’s, the figures are half those in the other ports.
Where is Sydney? Closer to Halifax/Charlottetown/Saint John or to St. John’s?
Even if we assume passengers spend $60 on average here, that’s still lower than the cruise industry’s estimate for 2016, which I found in that 2016 ACCA Economic Impact report:
But there are a lot of puzzling implications here: for one thing, the industry numbers suggest passengers in Sydney spend less than crew members in the other four ports. Is that possible?
Kayahan, understandably, was reluctant to draw conclusions for Sydney, given that none of the research was done here. But I think it’s not too far a reach to say this study shows why we need a study of our own. To have a real discussion about the economic impact of the industry, we need real numbers. When I suggested as much to Kayahan, he replied:
The primary reason we were interested in doing our study in the first place was to have an independent local study that would estimate the passenger spending across the ports using a consistent methodology since the academic literature frequently points to the flaws associated with the “industry standard” numbers.
Let’s hope some enterprising economist will do the same for Sydney.
Note: I haven’t even touched on the figures for overall economic impact in the Atlantic Region and per province. I hope to do that next week. If you read the report and have any thoughts on it, please email me!
[signoff]
