International Trade Theory and Policy
by Steven M. Suranovic

Trade 20-1

Measuring Protectionism: Average Tariff Rates Around the World

One method used to measure the degree of protectionism within an economy is the average tariff rate. Since tariffs generally reduce imports of foreign products, the higher the tariff, the greater the protection afforded to the country's import-competing industries. At one time, tariffs were perhaps the most commonly applied trade policy. Many countries used tariffs as a primary source of funds for their government budgets. However, as trade liberalization advanced in the second half of the twentieth century, many other types of non-tariff barriers became more prominent.

The table below provides a list of average tariff rates in selected countries around the world. These rates were all taken from the WTO's trade policy review summaries. More details about the trade policies of these countries can be found at the WTO's website at: http://www.wto.org/english/tratop_e/tpr_e/tp_rep_e.htm.

Generally speaking, average tariff rates are less than 20% in most countries, although they are often quite a bit higher for agricultural commodities. In the most developed countries, average tariffs are less than 10%, and often less than 5%. On average, less developed countries maintain higher tariff barriers, but, for many countries that have recently joined the WTO, tariffs have recently been reduced substantially to gain entry.

 

Average Tariff Rates

Japan (2000) 6.5%
European Union (2002)
Industrial Goods
Agriculture


6.4%
16.1%
Norway (2000) 8.1%
Canada (2000) 7.1%
Brazil (2000) 13.7%
Mexico (2002) 16.5%
Chile (1997) 11.0%
El Salvador (1996) 10.1%
Cyprus (1997)
Overall
with EU
Agriculture


16.4%
7.2%
37.6%
Morocco (1996) 23.5%
Bahrain (2000) 7.7%
Malawi (2002) 14.0%
India (2002) 32.0%
Pakistan (2002) 20.4%
Zambia (1996) 13.6%
Malaysia (2001) 9.2%

Problems Using Average Tariffs as a Measure of Protection

The first problem with using average tariffs as a measure of protection in a country is that there are several different ways to calculate an average tariff rate and each method can give a very different impression about the level of protection.

Most of the tariffs above are calculated as a simple average. To calculate this rate, one simply adds up all of the tariff rates and divides by the number of import categories. One problem with this method arises if a country has most of its trade in a few categories with zero tariffs, but has high tariffs in many import categories in which it would never find advantageous to import. In this case the average tariff may overstate the degree of protection in the economy.

This problem can be avoided, to a certain extent, if one calculates the trade-weighted average tariff. This measure weights each tariff by the share of total imports in that import category. Thus, if a country has most of its imports in a category with very low tariffs, but has many import categories with high tariffs but virtually no imports, then the trade-weighted average tariff would indicate a low level of protection. The standard way of calculating this tariff rate is to divide total tariff revenue by the total value of imports. Since this data is regularly reported by many countries this is a common way to report average tariffs. To illustrate the difference, Canada is listed above with a simple average tariff of 7.1%. However, Canada's trade-weighted average, in contrast, is a mere 0.9%.

However, the trade-weighted average tariff is not without flaws. As an example, suppose a country has relatively little trade because it has prohibitive tariffs (i.e. tariffs set so high as to eliminate imports) in many import categories. If it has some trade in a few import categories with relatively low tariffs, then the trade-weighted average tariff would be relatively low. After all, there would be no tariff revenue in the categories with prohibitive tariffs. In this case, a low average tariff could be reported for a highly protectionist country. Note also that, in this case, the simple average tariff would register a higher average tariff and might be a better indicator of the level of protection in the economy.

Of course the best way to overstate the degree of protection is to use the average tariff rate on dutiable imports. This alternative measure, which is sometimes reported, only considers categories in which a tariff is actually levied and ignores all categories in which the tariff is set to zero. Since many countries today have many categories of goods with zero tariffs applied, this measure would give a higher estimate of average tariffs than most of the other measures.(1)

The second major problem with using average tariff rates to measure the degree of protection is that tariffs are not the only trade policy used by countries. Countries also implement quotas, import licenses, voluntary export restraints, export taxes, export subsidies, government procurement policies, domestic content rules, and much more. In addition, there are a variety of domestic regulations which, for large economies at least, can and do have an impact on trade flows. None of these regulations, restrictions or impediments to trade, affecting both imports and exports, would be captured using any of the average tariff measures. Nevertheless these non-tariff barriers can have a much greater effect upon trade flows than tariffs themselves.

The Ideal Measure of Protectionism

Ideally, what we would like to measure is the degree to which a government's policies (both domestic and trade policies) affect the flow of goods and services (on both the import and export side) between itself and the rest of the world. Thus, we might imagine an index of protectionism (IP) defined as follows:

Where the numerator represents the sum of all exports and imports across all N trade categories given the current set of trade policies, and the denominator represents the sum of all exports and imports that would obtain if the government employed a set of domestic policies that had no impact on trade of goods and services with the rest of the world. If IP = 1, it would indicate that current government policies are completely non-restrictive and the economy could be characterized as being in a pure state of "free trade." If IP = 0, then government policies would be so restrictive as to force the economy into a state of isolation or autarky.

If we could calculate and compare the index across many countries, then we could say that countries with a smaller value were more protectionist than countries with a higher value. We could also monitor changes in the index over time for a particular country. Increases in the index value would indicate trade liberalization, while decreases in the index would indicate growing protectionism.

The problem with this index, however, is that although it is easy to define, it would be virtually impossible to measure. At least, I know of no way of doing so without making extreme leaps of faith. Nevertheless, the index definition is useful as a way of indicating how far from ideal are any traditional measures of protection such as average tariff rates.


Endnotes:

1. It is often claimed that average tariffs in the US were raised to almost 60% by the Smoot-Hawley tariff act of 1930. This figure, although correct, represents the average tariff on dutiable imports only. Thus, the figure somewhat overstates the true degree of protection. In comparison, the trade-weighted average tariff in subsequent years rose only as high as 24.8% in 1932, after which tariff rates fell.

International Trade Theory and Policy - Chapter 20-1: Last Updated on 6/14/06

PREVIOUSNEXT