Skip to main content
SpringerLink
Log in

Measuring the usage of preferential tariffs in the world

  • Original Paper
  • Published:
Review of World Economics Aims and scope Submit manuscript

Abstract

The preference utilization ratio, i.e., the share of preferential imports out of total imports, has been a popular indicator for measuring the usage of preferential tariffs vis-à-vis tariffs on a most-favored-nation basis. A crucial shortcoming of this measure is the data requirements, particularly for data on imports classified by tariff schemes, which are not available in most countries. This study proposes another measure for preference utilization, termed the “tariff exemption ratio.” This measure is a good proxy for the value of offered preferences by each importing country to the rest of the world. Importantly, it can be computed by employing only publicly available data, such as those provided by the World Development Indicators, for its computations. We can thus calculate this measure for many countries for an international comparison. Our finding is that tariff exemption ratios differ widely across countries.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2

Similar content being viewed by others

Notes

  1. Some variations exist in the measure of preference utilization. One issue is whether or not to exclude trade in products that are ineligible for preference schemes in the denominator. This point is discussed later.

  2. Our strategy is similar to Pritchett and Sethi (1994), which compared the ratio of import tax revenues to imports (called “collected rates”) with the official, statutory tariff rates at a product level for Jamaica, Kenya, and Pakistan. They demonstrated that significant gaps existed between the collected rates and the official rates. Our tariff exemption ratio is based on the comparison between aggregated tariff revenues under collected rates and counterfactual aggregated tariff revenues under official rates (MFN rates).

  3. In the academic literature, revenue data are also employed to examine the relationship between tax revenues and trade liberalization in some studies (Khattry and Mohan 2002; Agbeyegbe et al. 2006; Baunsgaard and Keen 2010; Hisali 2012).

  4. Although we will discuss the coverage of products included in the computation later, we categorize products into products with zero MFN rates and those with positive MFN rates, the latter of which should be further classified into products that are ineligible to and those that are eligible to preference schemes. We define products that have positive MFN rates and are not covered by preference schemes as “ineligible to preference schemes.” Namely, for the sake of clarity, we treat products with zero MFN rates separately from those ineligible to preference schemes. Indeed, as is discussed later, these two have different implications for our tariff exemption ratio. In addition, eligibility in the actual practice is determined according to not only products but also trading partner countries. However, for simplicity, we develop our measure in Sect. 2 as if only two countries exist in the world.

  5. Note that the calculation of the preference utilization ratio in the literature usually goes with a couple of the following procedures: first, the preference utilization ratio is often measured on the bilateral basis. Second, products ineligible to preference schemes and those with zero MFN rates are excluded in the computation of the preference utilization ratio. For more details, see Keck and Lendle (2012).

  6. One may think that the sum of imports multiplied by preferential rates or applied tariff rates could also yield valuable information. However, actually, these rates cannot be easily calculated. In particular, the publicly available data on applied tariff rates do not take GSP rates into account.

  7. In order to more precisely analyze how MFN or preferential rates affect the difference between these two measures, it is necessary to take into account that I M i and I P i are functions of t M i and t P i . Then, we need to introduce some theoretical model on firms’ utilization of preference schemes, as in Demidova and Krishna (2008) and Cherkashin et al. (2015).

  8. For more details on the equivalence between Eqs. (2) and (5), see “Appendix 1”.

  9. These trade measures are recently invoked in not only developed countries but also developing countries. The activation of those measures reduces imports in targeted products and thus their weights in the weighted average of MFN rates. As a result, the impact of these trade measures on tariff exemption ratios might be rather small.

  10. As of 2009, Thailand has RTAs with 14 countries, including Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, the Philippines, Singapore, Vietnam, China, India, Japan, Australia, and New Zealand.

  11. The WDI metadata indicate “Customs and other import duties are all levies collected on goods that are entering the country or services delivered by nonresidents to residents. They include levies imposed for revenue or protection purposes and determined on a specific or ad valorem basis as long as they are restricted to imported goods or services.”

  12. However, as being mentioned later, MFN rates reported by WDI are quantitatively different from those reported by WITS. Even the number of countries where the data on MFN rates are available is different between those two data sources.

  13. Various reasons may underlie such differences in total imports. For example, according to the IMF website, figures for total imports are not necessarily the same between IFS and UN Comtrade since these are based on different data collection systems with different aims, procedures, timetables, and sources for updates and maintenance. These differences result in gaps in total imports across data sources.

  14. Some studies found that taxes that require frequent interactions between tax authorities and individuals, particularly customs and import duties, are negatively affected by corruption (Thornton 2008; Imam and Jacobs 2014). Therefore, one may say that the tariff exemption ratio has some bias in countries with serious corruption, e.g., developing countries. However, misclassification or unfair classification of products does not yield a gap between the tariff exemption ratio and the preference utilization ratio if imports and import duties are exactly recorded under that classification.

  15. The monthly data may be easily obtained in the case of imports, compared with the case of customs duties. Thus, one may compute imports for fiscal years by manually aggregating the data of monthly imports. This paper, however, does not do this treatment.

  16. There are three more issues in the data on customs duties. First, if the duty paid in a year is refunded in the next year, the tariff exemption ratio does not necessarily indicate the use of preference in that year. Second, some countries do not consolidate central and local government finance data into one account and present only central government budgetary accounts. Nevertheless, in the case of customs and other import duties, even central government budgetary accounts are likely to show the whole picture. Third, when we use the data on total imports reported in the US dollars, we have to convert customs duties, which are typically reported in local currency, into the US dollars. In this paper, we used the annual average of official exchange rates for this conversion. Therefore, the large fluctuation of exchange rates within a year may create a possible bias in the data on customs duties in terms of US dollars, and thus in the tariff exemption ratio.

  17. As shown in Table 5 in “Appendix 2”, too large positive value of the tariff exemption ratio for Namibia is due to the fact that customs duties are negative. Such negative figures might be revised in later years in the WDI. The negative tariff exemption ratios in the rest of the cases are obtained because either import duties are too large or imports are too small; we are not sure which explains the anomalies.

  18. Using figures for the preference utilization ratio in Keck and Lendle (2012), we can check the differences between our tariff exemption ratio and the preference utilization ratio in various countries. Specifically, Table 1 in Keck and Lendle (2012) showed imports according to tariff schemes (i.e., MFN, RTA, and GSP) in Australia, Canada, and the US (and EU) in 2008. The preference utilization ratio is 0.18 in Australia, 0.70 in Canada, and 0.55 in the US. On the other hand, tariff exemption ratios in 2008 were 0.20 in Australia, 0.75 in Canada, and 0.56 in the US. We could say that the difference between the two measures is small for these countries.

  19. In order to minimize errors from the use of data in specific years, we also compute tariff exemption ratios by employing the latest three-year average of each variable. Since the data are available for only 1 or 2 years in some countries, the number of countries for which we can compute the ratios is reduced to 77. We also compute statistics for the same set of countries employing data from the latest single year. As a result, the mean values are almost identical between these two cases (0.58 for a three-year average or 0.57 for a single year).

  20. As mentioned before, EU countries are not included. Thus, Europe in our sample includes, for example, Switzerland, Norway, and Russia.

  21. MERCOSUR includes Brazil, Paraguay, Uruguay, and Venezuela. AFTA includes Cambodia, Indonesia, Lao PDR, Malaysia, Singapore, and Thailand. NAFTA includes Canada, Mexico, and the US. COMESA includes Democratic Republic of Congo, Egypt, Arab Rep., Kenya, Mauritius, Rwanda, Seychelles, Uganda, and Zambia. SAFTA includes Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka.

  22. This database is also publicly and freely available after registration.

  23. As mentioned in the previous section, the original source of tariff data in the WDI is also WITS, but WDI data are constructed by aggregation at an SITC five-digit level.

  24. Further basic statistics on the difference in tariff exemption ratios between cases of WDI and WITS are available in Table 5 in “Appendix 2”.

  25. In particular, the WDI database provides data on imports only from 2005. Thus, if we employ other data sources such as UN Comtrade, we can extend the sample years. For some countries, we can compute the tariff exemption ratio up to 1988, as data on the weighted average of MFN rates have been available in the WDI since that time. In this paper, to maintain the consistency of data sources with other tables and figures, we do not mix data sources on imports to determine values for such a longer period but instead show those since 2005.

  26. Tariff exemption ratios are likely to change from 2006 to 2007 or from 2011 to 2012 due to changes in HS versions, i.e., from HS2002 to HS2007 or from HS2007 to HS2012. Changes in the tariff line structure affect aggregated values of the MFN rates. Thus, a change in weighted averages of MFN rates through a change of HS versions will affect tariff exemption ratios.

References

  • Agbeyegbe, T., Stotsky, J. G., & Woldemariam, A. (2006). Trade liberalization, exchange rate changes, and tax revenue in sub-Saharan Africa. Journal of Asian Economics, 17, 261–284.

    Article  Google Scholar 

  • Baunsgaard, T., & Keen, M. (2010). Tax revenue and (or?) trade liberalization. Journal of Public Economics, 94, 563–577.

    Article  Google Scholar 

  • Bureau, J., Chakir, R., & Gallezot, J. (2007). The utilisation of trade preferences for developing countries in the agri-food sector. Journal of Agricultural Economics, 58(2), 175–198.

    Article  Google Scholar 

  • Cherkashin, I., Demidova, S., Kee, H., & Krishna, K. (2015). Firm heterogeneity and costly trade: A new estimation strategy and policy experiments. Journal of International Economics, 96(1), 18–36.

    Article  Google Scholar 

  • Demidova, S., & Krishna, K. (2008). Firm heterogeneity and firm behavior with conditional policies. Economics Letters, 98(2), 122–128.

    Article  Google Scholar 

  • Francois, J., Hoekman, B., & Manchin, M. (2006). Preference erosion and multilateral trade liberalization. World Bank Economic Review, 20(2), 197–216.

    Article  Google Scholar 

  • Hakobyan, S. (2015). Accounting for underutilization of trade preference programs: U.S. generalized system of preferences. Canadian Journal of Economics, 48(2), 408–436.

    Article  Google Scholar 

  • Hisali, E. (2012). Trade policy reform and international trade tax revenue in Uganda. Economic Modelling, 29, 2144–2154.

    Article  Google Scholar 

  • Imam, P. A., & Jacobs, D. (2014). Effect of corruption on tax revenues in the Middle East. Review of Middle East Economics and Finance, 10(1), 1–24.

    Article  Google Scholar 

  • Keck, A., & Lendle, A. (2012). New evidence on preference utilization, world trade organization (Staff Working Paper ERSD-2012-12).

  • Khattry, B., & Mohan, J. R. (2002). Fiscal faux pas?: An analysis of the revenue implications of trade liberalization. World Development, 30(8), 1431–1444.

    Article  Google Scholar 

  • Manchin, M. (2006). Preference utilisation and tariff reduction in EU imports from ACP countries. The World Economy, 29(9), 1243–1266.

    Article  Google Scholar 

  • Pritchett, L., & Sethi, G. (1994). Tariff rates, tariff revenue, and tariff reform: Some new facts. World Bank Economic Review, 8, 1–16.

    Article  Google Scholar 

  • Thornton, J. (2008). Corruption and the composition of tax revenue in Middle East and African Economies. South African Journal of Economics, 76(2), 316–320.

    Article  Google Scholar 

  • Yu, M., & Tian, W. (2012). China’s processing trade: A firm-level analysis. In H. McMay & L. Song (Eds.), Rebalancing and sustaining growth in China (pp. 111–148). Canberra: Australian National University E-press.

    Google Scholar 

Download references

Acknowledgements

We would like to thank two anonymous referees, Tatsuo Hatta, Hiroshi Ohashi, Hiroshi Mukunoki, Arata Kuno, Kozo Kiyota, and the seminar participants at the Asian Growth Research Institute, JETRO Bangkok, Hokkaido University, and the Japan Society of International Economics. This work was supported by JSPS KAKENHI Grant No. 26705002.

Author information

Authors and Affiliations

  1. Inter-disciplinary Studies Center, Institute of Developing Economies, Wakaba 3-2-2, Mihamaku, Chiba, Chiba Prefecture, 261-8545, Japan

    Kazunobu Hayakawa

  2. Faculty of Economics, Keio University, Tokyo, Japan

    Fukunari Kimura

  3. Science and Technology Development Program, Thailand Development Research Institute, Bangkok, Thailand

    Nuttawut Laksanapanyakul

Authors
  1. Kazunobu Hayakawa
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Fukunari Kimura
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Nuttawut Laksanapanyakul
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Kazunobu Hayakawa.

Appendices

Appendix 1: Equivalence of Eqs. (2) and (5)

In this appendix, we demonstrate that Eq. (2) is equivalent to Eq. (5). The weighted average of MFN rates can be defined as follows:

$$\begin{aligned} & Weighted\,Average of\,MFN\,Rates \equiv \mathop \sum \limits_{i} \left\{ {\left( {\frac{{I_{i}^{M} + I_{i}^{P} }}{{\mathop \sum \nolimits_{k} \left( {I_{k}^{M} + I_{k}^{P} } \right)}}} \right) \cdot t_{i}^{M} } \right\} \\ & \quad = \mathop \sum \limits_{i} \left\{ {\frac{{\left( {I_{i}^{M} + I_{i}^{P} } \right) \cdot t_{i}^{M} }}{{\mathop \sum \nolimits_{k} \left( {I_{k}^{M} + I_{k}^{P} } \right)}}} \right\} = \frac{{\mathop \sum \nolimits_{i} \left\{ {t_{i}^{M} \left( {I_{i}^{M} + I_{i}^{P} } \right)} \right\}}}{{\mathop \sum \nolimits_{k} \left( {I_{k}^{M} + I_{k}^{P} } \right)}}. \\ \end{aligned}$$

Therefore, the denominator in the second term in Eq. (5) can be summarized as follows:

$$\begin{aligned} & \left( {Weighted \,Average\, of \,MFN\, Rates} \right) \times \left( {Total\, Imports} \right) \\ & \quad = \frac{{\mathop \sum \nolimits_{i} \left\{ {t_{i}^{M} \left( {I_{i}^{M} + I_{i}^{P} } \right)} \right\}}}{{\mathop \sum \nolimits_{k} \left( {I_{k}^{M} + I_{k}^{P} } \right)}} \times \mathop \sum \limits_{k} \left( {I_{k}^{M} + I_{k}^{P} } \right) = \mathop \sum \limits_{i} t_{i}^{M} \left( {I_{i}^{M} + I_{i}^{P} } \right) \\ \end{aligned}$$

This term is exactly the same as the denominator in the second term in Eq. (2). Furthermore, by definition, the numerator in the second term in Eq. (2) is the same as “Total Revenues from Import Duties,” which is the numerator in the second term in Eq. (5). As a result, Eq. (5) is equivalent to Eq. (2).

Appendix 2: Several additional tables

See Tables 4 and 5.

Table 4 Outliers in tariff exemption ratios (million USD)
Full size table
Table 5 Difference in tariff exemption ratios according to data sources
Full size table

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Hayakawa, K., Kimura, F. & Laksanapanyakul, N. Measuring the usage of preferential tariffs in the world. Rev World Econ 154, 705–723 (2018). https://doi.org/10.1007/s10290-018-0321-9

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10290-018-0321-9

Keywords

JEL Classification

Search

Navigation