Monday, August 30, 2010

Probability And Logic: Tourism occupancy and spend

A good friend of mine from university days went on to become a government statistician. I once spent an afternoon lounging in a villa pool, having him explain to me the meaning of certain mathematical symbols. I can't say I was significantly the wiser as a result, but I had at least tried to attain some understanding.

All mathematicians, like chess players, are mad. This stems from having a world view determined by logic or equations and having to deal with the rest of us who are determinedly illogical. Even the act of getting up in the morning is an equation, where E equals the effort of shifting yourself from under the duvet, T is the time it takes to have a shave and eat a bowl of muesli and x (or maybe x-squared) is the probability of being hit by a meteorite when you step out the front door.

So when statisticians present us with their figures, we are inclined to believe it is all the work of crackpots. (And, yes, I know there is a difference between maths and stats, but you'll just have to accept they're the same for now.) Equally, because none of us have the faintest idea how they ever arrive at the figures, we dismiss them as mathematical mumbo-jumbo.

There have been some good statistics lately. For this July, by comparison with the same month last year, hotel occupancy rates in the Balearics have risen by 10.1% and tourism spend has also increased - by 12.5%. We might dispute the former, but at least we can appreciate that the calculation shouldn't be that difficult. The latter, though ... well of course we say it's crap because we saw a restaurant with hardly a soul in it the other day. We resort to our anecdotes and subjectivity, because we have no other way of questioning the men with wild, staring eyes and electric grey hair who inhabit the statistical other world.

And so it was also with the editorial in "The Bulletin". Who was spending all the money, it asked. "I don't really know and nor do the people who have complied the report," came its own answer. It's not for the paper to try and find out; it is only a newspaper after all. The answer, one of them, lies with the mysteries of the statistician's science, such as regression modelling, which for the paper would mean ever more of a regression into a Janet and John mode of journalism or a stock photo of Naomi Campbell.

One thing Mallorca does quite well, along with beaches, cold lager and pouring oil on bread, is the study of tourism. The university in Palma is world class when it comes to research into the economics and statistics of the industry. It's this academic rigour which translates itself into governmental research and statistics. We might not believe the figures, and governments are not unknown to massage figures, but the science, be it academic or governmental, is generally robust. Academics, for example, can't just pluck numbers out of thin air, as their published work is subject to independent peer review.

In Palma they have found out who spends the money. There is research there into the use of statistical modelling to ascertain - by age, by nationality, by professional grouping - who spends what. What you get with statistics is the attempt to give as realistic a snapshot as possible within the parameters of the probability models applied. They are never completely accurate, but nor are they fiction. Were they, and some might say the latest occupancy and spend figures are made-up propaganda to give a feel-good lift to the tourism and wider economy, then why have previous figures been negative? But because we don't understand the methods, and because academics and governments are generally lousy at explaining them in simple terms, if at all, we fall back on our anecdotes and consider these to be the truth, when they are nothing of the sort.

There is a problem, though, with the figures in that they are too general; they don't, for example, distinguish between different resorts, hence the not unreasonable scepticism we may feel. One is reluctant to say that July's figures indicate a real recovery - and the predictions before the weather in the UK went belly up was that August's wouldn't be as good - but they are cause for some hope.

There is another, simpler way of looking at the July figures. 10.1% rise in occupancy; 12.5% rise in spend. The two may not match, but they are not so far apart. Logic might suggest that one would lead to the other, though for the statistician the 2.4% difference would - in all probability - not be logical.


Any comments to andrew@thealcudiaguide.com please.

No comments: