While governments often wish to restrict certain goods from crossing borders, their means and will to do so are often too meagre to discourage business-minded traders. Smuggling is thus prevalent and can result in violence, distorted competition, and loss of tariff revenue (Naim 2005).
In China, tariff evasion is a severe problem that authorities have been trying to tackle for many years. The combination of a bumpy tariff schedule and corrupt customs agents make its border porous (Fisman and Wei 2004). FedEx warns on its website (2010) that “customs officials still have wide discretion concerning the category in which an import is placed [and have] the flexibility […] to "negotiate" duties.” As Ray Fisman explained in Economic Gangsters and on this site, frozen chickens from Hong Kong are simply declared as turkeys to avoid 20% tariffs.
While a recent literature has extensively documented the role of high tariffs in determining tariff evasion (see Javorcik and Narciso 2008 and Mishra et al. 2008), very little research has tried to identify other possible determinants.
In new research (Rotunno and Vézina 2010), we show that an important ingredient for smuggling is the presence of international networks. Rauch and Trindade (2002) show that ethnic Chinese networks, notably through interpersonal relationships known as guanxi, act as trade catalysts by enforcing contracts and providing market information. Overseas Chinese networks may therefore be behind China’s tariff evasion. They might know who is willing to buy smuggled goods, which customs agents are corrupt, how to package the goods to disguise them, and how to fill export declarations appropriately. Moreover, their mutual trust could help overcome the risk of hold-up. To examine whether this is the case we first need a measure of this “invisible” trade.
Capturing smuggling in official statistics
In theory, what one country reports as imports should be equal to what its partner reports as exports. In practice, however, this is rarely the case. As noted by Bhagwati (1964), smuggling may be behind this discrepancy. Smuggled goods may be undervalued or misreported in the importing country, often with the complicity of corrupt customs officials, resulting in missing imports. Smuggling into China can thus be captured by the (log) difference between exports declared by exporting countries and imports declared by China.
But missing trade is a noisy measure that captures much more than smuggling. Import values include cost-insurance and freight (cif) costs whereas export values are free on board (fob), so the difference in reports also includes trade costs. It is also noisy because of exchange-rate miscalculations, lax custom statisticians, and indirect trade confusing reports. But, as Fisman (2009) reminds us, while the trade gap cannot be used to quantify smuggling precisely, it is still relevant for identifying correlations and uncovering the causes of illicit flows.
Without a doubt, asymmetric trade policies such as tariffs allow us to observe smuggling in missing trade. This is illustrated in Figure 1. Missing trade is white noise when there are no prohibitive trade barriers. But when import restrictions are high, i.e. when their tariff is above the 95th percentile (20%), missing imports are almost strictly positive, in other words, missing from the importer’s reports. A positive correlation between missing imports and tariffs is thus a clear indication of tariff evasion.
Figure 1. Missing imports in China
Do Chinese networks increase tariff evasion?
If Chinese networks increase tariff evasion, they should increase the correlation between tariffs and missing imports.
Using trade data on 5,000 products from 130 countries, we regress Chinese missing imports on their corresponding tariffs and the latter’s interaction with Chinese overseas populations in the exporting country (or the Chinese-born share of the population), using data from the Global Origin Migrant Database. Including country fixed effects controls for all trading partner characteristics that could affect the level of missing trade, such as GDP per capita or corruption.
As shown in Figure 2, we find that the tariff semi-elasticity of missing imports indeed increases with the number of overseas Chinese in the exporting country. In trade with the US, where there are more than 1.5 million Chinese-born residents, an increase in tariff from 10% to 20% would increase missing imports by almost 40%. However, in trade with France, where only about 40,000 Chinese-born migrants live, the same tariff change would increase evasion by less than 25%.
Figure 2. Tariff evasion and Chinese networks
Furthermore, as some may argue that the gap in trade values is too noisy a measure to capture smuggling, we show that the results also hold when missing imports are measured in quantities rather than values.
The role of product differentiation
Previous research showed that tariff evasion was more accentuated in differentiated product, both in India (Mishra et al. 2008) and Eastern Europe (Javorcik and Narciso 2008). The explanation is that differentiation increases the difficulty in ascertaining prices and classifications and hence in detecting misreporting or under-invoicing.
We show that this result also holds for Chinese imports but that Chinese networks expertise is actually more useful when tariff evasion is hardest, i.e. for goods sold on organised markets or listed in trade publications.
Does corruption in the trading partner matter?
Tariff evasion does not only occur in China but in many other countries around the world, even more so in the most corrupt ones (Berger and Nitsch 2008, Fisman and Wei 2009). We thus check whether Chinese communities increase tariff evasion in their host countries on imports from China.
Our results, summarised in Figure 3, indicate that tariff evasion is highest when corruption is high and when Chinese communities are biggest, as predicted.
Figure 3. The impact of corruption and Chinese networks on tariff evasion
Is there evidence of misreporting in similar categories?
Using data on trade between Hong Kong and China, Fisman and Wei (2004) show systematic evidence of misreporting, such as chickens being passed as turkeys, as missing imports were negatively correlated with tariffs on similar goods.
We thus check whether such misreporting practices are also common in other countries and if Chinese networks operate through this mechanism. We find that when the tariff spread is above 10 percentage points, tariffs on similar goods have a negative impact on missing imports that increases with both Chinese networks and corruption. Hence, Chinese networks and corruption do increase misreporting.
However, this test does not capture all misreporting. In the Philippines, smuggled rice has been declared as mung beans to avoid a 50% tariff. But rice and beans would not come up as similar in our classification. We leave the identification of such misreporting to future research.
Underinvoicing and misreporting in similar goods are just two strategies used by economic gangsters’ networks to dodge trade restrictions. Smuggling can take many other forms, from transhipment via third countries to double invoicing, i.e. underinvoicing goods and charging for fake services to complete the sell. Shedding further light on this underground but widespread phenomenon will be crucial for the design of effective policy responses.
Berger, Helge and Volker Nitsch (2008), "Gotcha! A Profile of Smuggling in International Trade", CESifo Working Paper 2475, Munich.
Bhagwati, J (1964), “On the Underinvoicing of Imports”, Oxford Bulletin of Economics and Statistics, 26:389-397.
FedEx (2010), China country profile.
Fisman, Raymond and Shang-Jin Wei (2004), "Tax Rates and Tax Evasion: Evidence from "Missing Imports" in China", Journal of Political Economy, University of Chicago Press, 112(2):471-500, April.
Fisman, Raymond, and Edward Miguel (2008), Economic Gangsters, Princeton University Press.
Fisman, Raymond J and Shang-Jin Wei (2009), “The Smuggling of Art, and the Art of Smuggling: Uncovering the Illicit Trade in Cultural Property and Antiques”, American Economic Journal: Applied Economics, 1(3), July.
Fisman, Raymond (2009), “Measuring Tariff Evasion and Smuggling”, NBER Reporter: Research Summary 2009 Number 3.
Javorcik, Beata S and Gaia Narciso (2008), "Differentiated products and evasion of import tariffs", Journal of International Economics, Elsevier, 76(2):208-222, December.
Mishra, Prachi, Arvind Subramanian, and Petia Topalova (2008), "Tariffs, Enforcement, and Customs Evasion: Evidence from India", Journal of Public Economics, Elsevier, 92(10-11):1907-25, October.
Naim, Moises (2005), Illicit: How Smugglers, Traffickers, and Copycats are Hijacking the Global Economy, Doubleday.
Philippine Daily Inquirer (2010), “2 firms charged with smuggling for misdeclaring rice as mung beans”, Manila, 5 August.
Rauch, James E and Vitor Trindade (2002), "Ethnic Chinese Networks In International Trade", The Review of Economics and Statistics, MIT Press, vol. 84(1):116-130, February.
Rotunno, Lorenzo and Pierre-Louis Vézina (2010), “Chinese networks and tariff evasion“, IHEID Working Paper 20-2010