What are the cheapest countries to visit? See the Travel Cost Index.

There is a lot of interest on blogs and discussion forums about the cheapest nations in the world to visit, but I found no good hard data on the internet.  This is a bizarre hole in the world’s collective knowledge (at least as revealed by Google) because it is easy to predict what countries are the cheapest to travel in for Americans using readily available data and a bit of economic theory.  All you need are two things:  the market exchange rate and the prices in each nation.

In 2008, a Big Mac cost $3.57 in the US and £2.29 in the United Kingdom.  Does that mean that the UK is cheaper than the US?  No, because the market exchange rate was $2 = £1 and that makes the UK seem more expensive than the US because a tourist exchanging a thousand dollars would only get five hundred pounds.

In reality, you need both prices and the exchange rate to know if Britain is a cheap place to travel or an expensive place.

What is the Purchasing Power Parity (PPP) exchange rate and how is the Travel Cost Index calculated?

The best available measure of relative prices in different countries is called the purchasing power parity (PPP) exchange rate.  PPP is simply what you get when you divide the prices of goods in one currency by the prices of the exact same goods in another currency.  PPP is the price ratio that tells you how much higher prices are in one country than in another.  For example, suppose the only price we care about is the price of a hamburger.   If an identical hamburger costs $2 in the US and €4 in France, then a dollar is worth €4/$2 = 2€/$.  That means the purchasing power parity of a dollar is two euros.  PPP tells you how many dollars it takes to buy the same goods in euros and vice versa.

By itself, PPP tells us nothing about how expensive it is to visit Paris because it doesn’t tell how many euros you can actually exchange for one dollar.  That is the market exchange rate.  If the market exchange rate equals the PPP exchange rate, then the cost of traveling in France is exactly the same as the cost of travel in the US. For example, if the two exchange rates are equal, as in the above example, then 2€/$/2€/$=1 and both countries are equally expensive. When the two exchange rates differ, the percent difference tells you how much cheaper (or more expensive) another country will be.

That ratio is the Travel Cost Index (also known as the “real exchange rate“): the PPP exchange rate divided by the market exchange rate.  For example, if the market exchange rate value of the dollar is more than the PPP exchange rate value, then France will be less expensive than America.  For example, at the end of 2017, one dollar bought 1.20 euros, and the price ratio (PPP) between France and America was 1.08 (meaning prices were 8% higher in euros than in dollars).  If you divide the two you discover that (1.08€/$)/(1.2€/$)=0.9 or ninety percent.  This means that France was about 10% cheaper than the US according to the World Bank for 2017.  Of course, this is just an average and Paris is going to be more expensive than most parts of France, but the same is true of New York City compared with the rest of the USA.

The Economist Magazine has used price of Big Macs around the world to actually create what they call the Big Mac index since 1986.  This is often interpreted as a measure of PPP, but it is actually a kind of travel price index or at least it would give useful information about what countries are cheap to visit if you like to eat at McDonalds everywhere you go (like President Trump) and you don’t buy anything else!

According to the Big Mac travel cost index, Switzerland costs 18% more than America whereas Qatar costs 40.9% less.  But that is just using one price, the price of a Big Mac, to adjust for cost of living in different countries.  The World Bank estimates PPP cost of living using an index of hundreds of prices which makes it much more representative of overall costs.  Even though the World Bank PPP is much better than the Big Mac PPP, the two data sets produce fairly similar rankings of the costs to travel in different countries.

This World Bank map gives an approximate value of how expensive each nation was in 2019:

(Click here for an interactive version)

The table below gives the numerical data for 2017.  The cheapest country in the world in 2017 was Egypt which is right next-door one of the most expensive, Israel.  In Egypt, it only took 21 cents to buy a dollar’s worth of goods and services on average.  That is like the whole country being on sale, almost 80% off!  The very most expensive was Iceland where it took $1.32 to buy what normally costs only $1 in the US.  The index says that 49 cents in Mexico buys as much as a dollar in the US which means that Mexico was approximately half price compared to travel in the US.

(Note: The following table is very long, so you can click here to scroll to the end and continue reading.)

What are flaws in the Travel Cost Index?

The Travel Cost Index is only an approximate estimate of relative price levels because, first of all, PPP doesn’t just focus on the kinds of things that tourists buy. It measures a much broader ‘basket’ that tries to encompass all goods and services, some of which, like “200 types of equipment goods and about 15 construction projects” are completely irrelevant to tourists whereas a lot of the things that are important to tourists like tour packages are probably not measured in the PPP index at all.  The PPP index only includes 3,000 consumer goods and services, so it leaves out a lot of things, and the kinds of things travellers buy are poorly represented.

Some countries are going to be cheaper than the Travel Cost Index indicates because some places like Thailand have an extremely efficient tourism infrastructure which brings down prices for tourism-related purchases relative to everything else and other countries like Iraq (probably) have not developed travel infrastructure that is as efficient for handling international visitors from around the world.

Another problem is that some regions in every country have cheaper prices for travelers than other regions and if you are in an expensive region, like New York City in the US, you’ll need a very different budget than you would need in a cheap region, like West Virginia. The PPP index tries to show average prices, and that means that the prices in rural areas is going to typically be lower than the PPP index (at least for the kind of purchases that rural people routinely buy), and costs in large cities are typically going to be higher.  Some international firms like UBS have created proprietary cost of living indexes for the biggest cities in the world, but it is hard to get up-to-date data from them.

To estimate true costs, a traveller should add about 3% because  foreign exchange transactions costs usually average about 3% which just makes international travel more expensive than domestic travel. No traveler can get the published market exchange rate due to those transactions costs and so every country is really going to be about 3% more expensive on average than the Travel Cost Index shows.

Finally, the Travel Cost Index uses World Bank Data which is generally about a year out of date.  This is one reason it doesn’t agree with the newer OECD data, (and I’m not sure what accounts for the rest of the discrepancy).  PPP is obviously not an exact science as explained above.  In any case, the travel cost index can be made more accurate by updating it using newer price information.

PPP update formula

To update old PPP data, just get the increases in inflation, and the change in exchange rates since the old PPP data was calculated.  This is how real exchange rates are calculated although most of the “real exchange rates” merely use the market exchange rate of an arbitrary year as baseline and it is more accurate to use PPP as a baseline.

Then calculate how much the foreign exchange rate has changed (in percentage) and subtract it from how much higher their inflation rate has been compared to the US rate over the same time period. For example, if the foreign inflation rate has averaged 10% over the past year and the US inflation rate has been 2%, that means that their prices have gone up 8 percentage points more than in the US. Then if their exchange rate (the value of their currency per dollar) has gone up 5%, the net change in prices is 3%, meaning that their country is 3% more expensive than it was a year ago.

10% – 2% – 5%  = 3%

10% = How much their prices rose

2% = How much American prices rose

5% = How much the exchange rate value of the dollar rose

This is how PPP is updated by professional economists most of the time.  Because it is expensive to gather actual PPP prices in every nation in the world, and because we have inflation data which gives similar information, the baseline PPP data is ordinarily just updated with this formula.

Vox created a “Vacation Index” which merely used this calculation to show how much cheaper each country got over the prior year instead of using the PPP measurements as a baseline.  That produces an interesting index as shown for May 2016 (below), but any change in the past twelve months is usually going to be very small relative to the Travel Cost Index, although it is somewhat significant in rare examples like when countries suffer a major currency devaluation such as in Argentina and Russia at the top of the graph.

The Vox index in the graph just shows how much the Travel Cost Index changed over the past year in different countries.  This is a poor measure of overall travel costs because it ignores the PPP baseline.  For example, it misleads readers into thinking that the United Kingdom is particularly cheap and Turkey is expensive when the reality is just the opposite.  This chart merely shows that the United Kingdom had gotten 3% cheaper than it had been the previous year and Turkey had just gotten 4% more expensive than it had been the previous year, but the UK started out a lot more expensive than Turkey (and it has been more expensive for at least a century) so the relatively small changes over the prior year weren’t significant compared with the baseline PPP.  Despite falling prices in the UK and rising prices in Turkey, the UK still remained almost twice as expensive as Turkey overall according to World Bank data above.

Of course, my index using World Bank data is also imperfect for the reasons mentioned above, but it is the most comprehensive there is.  The editors at VOX must have determined that their index isn’t very useful because they abandoned it after only a couple years of updates.

2019 is a good time for Americans to travel

Most countries look relatively cheap compared with America right now because the US dollar is particularly high right now relative to our biggest trading partners.  The multilateral, real, trade-weighted value of the dollar has not been this high in about 15 years.

I don’t put as much faith in the Big Mac Index, but it also shows that the US dollar is particularly strong now.

This graph shows the value of foreign currencies in dollars.  That is the inverse of the value of the dollar, so a falling line here means the value of foreign currencies was falling which means that the US dollar is doing the opposite–it has been rising.

If you want to get even more wonkish…

Another way to present this kind of data graphically can be seen on the following Gapminder graph which shows market exchange rate incomes on the vertical axis versus the PPP incomes on the horizontal axis.  (Click on the link to see more detail.)

The black line shows approximately where PPP is equal to exchange rates and nations below the line are relatively cheap whereas nations above the line are relatively expensive.  Poorer nations (on the left side) tend to be cheaper overall than richer nations (on the right) mainly because labor is so much cheaper in poor nations.  The red outliers on the right side of the graph are relatively cheap relative to the other high-GDP countries (Note: they are Saudi Arabia and mostly the oil exporters and this doesn’t make sense to me!) and the little red island nations which are outliers just to the left of the middle of the graph are relatively expensive compared with other circles nearby although only one of them is actually above the line.

Purchasing Power Parity (PPP) would be the same as market exchange rates if there were perfect competition and the law of one price were true.  In reality, most production is non-tradeable  because they are too expensive to be shipped abroad.  Non-tradeable production includes real estate and services like hair cuts and taxi rides.  Goods and services that are not tradeable are not subject to arbitrage and so the law of one price does not apply.  This is why developing countries usually have lower travel costs.  They often only cost between 1/4 and 1/2 as much as the more expensive developed countries and it is mainly because services (labor) and real estate are so much cheaper in poor countries.

Generally tourists buy a lot of labor in hotels, restaurants, taxis, etc., and because wages are way lower in poor countries, services tend to be much cheaper.

For tradeable goods, the law of one price should apply in the long run in theory, but that theory doesn’t work very well in the short run mainly due to the massive scale of financial flows which largely determines spot exchange rates and which have little to do with long-run fundamentals.

This graph shows how much more the market exchange rate fluctuates relative to the PPP exchange rate for Canada/US:

Canada is cheaper than the US when the market exchange rate (red) is above the PPP exchange rate and more expensive than the US when reversed.  The Travel Cost Index is the ratio of these two numbers.

According to the Bank for International Settlements, the foreign exchange market is by far the biggest market in the world in volume.

The massive flows in foreign exchange speculation results in market exchange rates that can fluctuate much more wildly than any changes in the prices of tradable goods, GDP, or anything else in the real economy. Large speculative movements in market exchange rates can persist for years.

Although tradable goods like Ipads should be close to the same price everywhere, even they can sell for radically different prices around the world.  According to the Ipad index, the price varied in 2013 from $473 in Malaysia, to over $1,000 in Argentina.  This is due to differences in transactions costs such as:

1. Taxation (particularly tariffs). This is usually what explains most of the highest prices.
2. Retail industry efficiency.  The USA has one of the most efficient retail industries in the world which helps keep retail prices lower in the US than in many other countries.
3. Transportation costs & information costs.  Transportation is insignificant for an Ipad, but more important for heavy or bulky goods and if international traders don’t have information about potential arbitrage, they won’t be able to drive down price differentials.
4. Regulations that enable multinational price discrimination schemes like the bans on importing pharmaceuticals into the US at Canadian prices or intellectual property rights which ban the sale of DVD movies intended for Asia in the US market or the importation into the US of textbooks intended for Africa.
5. “Sticky prices”.  Because market exchange rates fluctuate much more than other prices, an Ipad could be imported at $500 and sit on a shelf with a constant price while the exchange rate doubles which would dramatically change the deal for foreign tourists.