A right of passage for many parents is taking their kids to Disney World. I was recently there with my kids and was truly stunned by the number of large tourist groups from Brazil. As far as I can tell, it may have been the start of mid winter school vacation for Brazilian High Schools or Universities, so the timing makes sense, but the sheer quantity was a good reminder that while not all purchasing decisions involve the exchange rate or necessarily change when it moves, tourism is really quite sensitive to exchange rates.
The US dollar bought nearly 4 reals in late 2002 and now it is well under 2 (granted the peak was brief, but there was a 4 year span when it was above 2.7 and it is 50% less than that now). This led to two thoughts: 1. it must be fun to be a tourist in the US right now if you come from a non-Dollar country, and 2. I have sincere doubts that the balance of payments data is catching all this.
The tourism flows themselves are generally thought to be tracked reasonably well, but the massive flows of people buying up everything in sight (often with cash) at stores from Manhattan to Disney World is a good reminder that the services and retail sectors – not just manufacturing – stand to make large gains in sales from a depreciation AND that we may seriously undercount these transactions. If the undercounting were constant over the range of the exchange rate that would be fine, but it seems that there must be more undercounting when the dollar is weak and so many people are buying as much as they can carry while on trips to the US. This will tend to understate the responsiveness of certain sectors to the exchange rate.
As a side note, I went straight from Disney to the NBER Summer Institute, where not only can you hear a great deal of interesting research, but you can typically get answers to various questions such as “How well do we track such purchases in the Balance of Payments”. The unofficial consensus: not very well. A number of people argued that government statistics try to capture some of this activity (especially by tracking purchases by foreign credit cards at large flagship stores in Manhattan), but that large amounts of transactions slip through. One person I asked pointed out that he was from Europe, had bought a fair amount on his trip to Boston for the NBER, and since he had a US based credit card, and purchased the goods in the US, there was no way anyone could have tracked his purchases as “exports from the US”.
Others – my guess is particularly Brad Setser and a few others who like to dig deeply into BoP data – may have a more formal estimate, but the view from Cindarella’s castle at Disney suggested this is an interesting issue.