Abstract
We explore the effect of European Union (EU) non-reciprocal preferential schemes and the compliance costs they entail on the agricultural import flows from beneficiary countries. Since such costs are heterogeneous and mostly unobservable, we gauge their influence by some estimated proxies, and specify a gravity model that allows for a different preferential margin impact according to the costs level. For a large sample of developing countries in 2002, we find that the costs of compliance seem to play a role in making the schemes work: the lower the costs, the greater the impact of the preferential margins. Moreover, the margin effect seems different across different regimes.
This chapter is largely based on: Agostino, M., Demaria, F., and Trivieri, F., (2010). Non-reciprocal trade preferences and the role of compliance costs in the agricultural sector: exports to the EU. J Agric Econ 61(3), 652–679.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Since we focus on 2002 and learning effects are likely to affect the utilisation of any schemes, even more so when beneficiary countries are Least Developed Countries (LDC), we do not consider the EBA initiative, which has been in force only from 2001. For a detailed description of the EU non-reciprocal preferences see Chapter 2 as well as Bureau et al. (2007a); OECD (2005); Persson and Wilhelmsson (2007).
- 2.
It is worth mentioning that other obstacles, not directly connected to the scheme requirements, may also hinder the capability of exporting developing countries to access the EU market (e.g., high qualitative standards imposed by private importers, weak infrastructures and institutions, corruption, etc). In estimating our measure of compliance costs, we try to control for them.
- 3.
- 4.
As Low et al. (2005, p. 9) claim: “Since producers use different technologies, it may be convenient for some to use the preference and satisfy the requirements, while the origin rules may make it less convenient for others.” Furthermore, Manchin (2006, p. 1251) points out that “if the value of the preferences does not exceed those of the costs of getting the preferences a trader should have no incentive to ask for them.”
- 5.
A non-parametric measure of the compliance costs has been proposed by Anson et al. (2005), while Carrère and de Melo (2004) rely on non-parametric as well as parametric estimates. Francois et al. (2006) use a sample splitting method to identify a threshold at which the benefits from the preferential regime are offset by the relative compliance costs. See Bureau et al. (2007a) for a review of the methods used to assess the costs of compliance.
- 6.
On the use of this estimator, see Carrère and de Melo (2004).
- 7.
Since we lack data on exporting firms, the Tobit that we adopt models the decision of a hypothetical representative (aggregate) exporter that makes the collective choice to utilize the preference at the line level. Such a decision depends on an underlying latent variable, which represents the utility of utilisation. Higher residuals, being associated to higher utility of utilisation, mean a higher propensity to utilize the scheme; thus, they express lower compliance costs.
- 8.
Since we use very disaggregated data on agricultural lines, the source country GDP might be an unreliable measure of the comparative advantage in the specific sector. When we have tried, though, to retrieve figures on the national production (source FAOSTAT database), we could not obtain data at the HS6 digit level for many lines. As mentioned in the results section, we have also used the agricultural GDP obtaining the same qualitative results as those based on the total GDP, but in smaller samples.
- 9.
We do not employ preferential imports as dependent variable of our gravity equation (as Nilsson and Matsson 2009, do) to allow the possibility of comparing our findings with those obtained by several other empirical studies which investigate the impact of NRPTA on total imports, using only preferential margins within a gravity framework (for instance, Aiello and Demaria 2009; Aiello and Cardamone Chapter 7 of this book ; Cardamone 2011).
- 10.
Incidentally, one could argue that frequent and large values of IMP could influence the COSTS values if scale and learning economies were relevant. To shed light on the latter issue, yet, we would need information (that we lack) on the number and average size of the shipments towards the EU, as large flows could reach the EU through several shipments and vice versa, limited export volumes could enter the EU market through few deliveries.
- 11.
According to these authors, trade flows between two countries depend not just on the barriers between them, but also on the barriers between them and the rest of the world. Therefore the exporter specific effects are intended to capture the policy attitude of a country towards all its trading partners.
- 12.
We have also estimated (10.1) by Ordinary Least Square (OLS), accounting for trading partners’ fixed effects via the inclusion of exporter dummies (v i ’s). This estimator has been traditionally employed in the literature that uses the gravity model to evaluate the impact of non-reciprocal preferential trade policies. In order to account for the existence of zero trade flows, we have added one to our dependent variable to ensure that the logarithm is well-defined. The OLS results confirm those obtained when adopting non-linear estimators.
- 13.
Indeed, as illustrated in the data section, we do not limit our sample to those countries for which positive flows are recorded, because preferences may influence not only the existing trade volume between two countries, but may also open new trade flow routes. Disregarding zero trade observations would translate into disregarding emerging new trade relationships, and this could lead to underestimate the effect of preferences on developing countries exports (on this issue see: Aiello et al. 2010; Helpman et al. 2008; Piermartini and The 2005).
- 14.
At the same time, since multiplicative interaction models are symmetric, our model allows for the possibility that the effects of COSTS vary according to the values of the MARGIN measure. Further, it worth underlying that, as the Poisson model is linear in the log of the expected value, we apply (10.2) and (10.3) to the ln[E(IMP)]. If we wanted to compute the marginal effect of our regressors on E(IMP), we should account for the fact that the Poisson model is not linear in the expected value of the dependent variable (see Norton et al. 2004).
- 15.
Available at: www.cepii.fr/anglaisgraph/bdd/distances.htm.
- 16.
The fact that our measures of costs are retrieved from independent regressions for each preferential regime may explain why some variables display a different sign for different schemes. For instance, in the GSP-Drugs case, the positive sign of the distance variable may be due to the small group of countries eligible to this specific regime (Andean and Central American countries, and Pakistan). Indeed, being the recipients a limited number of countries, almost all relatively close to each other, the distance regressor does not vary much in the sample.
- 17.
The polity score ranges from −10 (high autocracy) to +10 (high democracy), and is drawn from POLITY IV database (available at www.systemicpeace.org/polity/polity4.htm). The telecommunication availability (WDI 2008) is measured as telephone lines per 1,000 people. We have also tried to control for the availability of physical infrastructures (such as railways and road kilometers, WDI (2008), encountering convergence problems in our estimations, possibly due to the presence of many missing values and/or a high correlation between them and the gross domestic product variable.
- 18.
Detailed results of these further robustness checks are available from the authors.
- 19.
Nonetheless, as shown by the F-tests reported, the interaction term is always jointly significant with the MARGIN regressor. The divergence between individual and joint significance may be interpreted as a symptom of multicollinearity (see Brambor et al. 2006) induced by the inclusion of the interaction term. As Brambor et al. (2006, p. 70) draw attention to, “even if there really is high multicollinearity and this leads to large standard errors on the model parameters, it is important to remember that these standard errors are never in any sense “too” large – they are always the “correct” standard errors. High multicollinearity simply means that there is not enough information in the data to estimate the model parameters accurately and the standard errors rightfully reflect this.”
- 20.
On the other way round, the preferences influence is expected to increase for lower costs of compliance. It is not obvious, however, to find that – as the costs of compliance decrease – the impact of the margins on exports is positive and statistically significant. In other words, as Bureau et al. (2007a) highlight, low utilisation costs are not synonymous of trade creation. A preferential scheme may be used because exporters find it convenient to comply with requirements, but exporters may not significantly increase their export volumes.
- 21.
Similarly, the COSTS individual coefficient represents its estimated marginal effect when the MARGIN is zero. Besides, it is worth recalling that the Poisson regression models the log of the expected dependent variable, which is usually a count variable, as a (linear) function of the explanatory variables. Thus, using the approximation properties of the log function, an estimated coefficient may be interpreted as the percentage change in the (expected value of the) dependent variable for a one unit change in the explanatory variable, given that the other predictor variables in the model are held constant. As a consequence, if a regressor is taken in log, the associated parameter may be interpreted as an elasticity.
- 22.
The only exception is represented by the GSP case when using Model 3 specification. We do not emphasize this result, yet, as the Tobit model did not converge when including exporter fixed-effects, hence we had to estimate it by including only the margin variable.
- 23.
To ensure that the Cotonou results are not driven by some extreme values, observations laying in the first and last percentile of the costs distribution have been omitted from the estimations. Results obtained without trimming the cost distribution are analogous, the Cotonou effect being only slightly higher.
- 24.
The output obtained when using Models 1 and 3 is analogous and available on request.
- 25.
To give an example, we consider grapes fresh (HS6 080610) as a raw commodity, grapes dried (HS6 080620) as slightly processed, wine of fresh grapes (HS6 220410) as highly processed.
- 26.
When adopting a Negative Binomial model, the main finding of this chapter is confirmed also if we use an alternative measure of the degree of processing: a dichotomous variable coded 1 for processed products and zero otherwise, on the base of the FAO classification (employed also by Bureau et al. 2007a).
References
Aiello F, Agostino MR, Cardamone P (2010) Evaluating the impact of non-reciprocal trade preferences using gravity models. Appl Economics 42:3745–3760.
Aiello F, Demaria F (2009) Do trade preferential agreements enhance the exports of developing countries? Evidence from the EU GSP. PRIN PUE&PIEC 2007 Working Paper n. 2009/18 (available via http://www.ecostat.unical.it/anania/PUE&PIEC%20Working%20Papers.htm)
Anderson JE, van Wincoop E (2003) Gravity with gravitas: A solution to the border puzzle. Am Econ Rev 63:881–892.
Anson J, Cadot O, Estevadeordal A, de Melo J, Suwa-Eisenmann A, Tumurchudur B (2005) Rules of origin in north–south preferential trading arrangements with an application to NAFTA. Rev of International Economics 13(3):501–517.
Brambor T, Clarck W, Golder M (2006) Understanding interaction models: improving empirical analyses. Political Anal 14:63–82.
Brenton P (2003) Integrating the least developed countries into the world trading system: the current impact of EU preferences under Everything But Arms. World Bank Policy Research Working Paper n. 3018.
Brenton P, Manchin M (2003) Making EU trade agreements work: the role of rules of origin. World Econ 26(5):755–769.
Bureau JC, Chakir R, Gallezot J (2007a) The utilisation of trade preferences for developing countries in the agri-food sector. J of Agric Economics 58(1):175–198.
Bureau JC, Disdier AC, Ramos P (2007b) A comparison of the barriers faced by Latin American and ACP countries' exports of tropical products. ICTSD (International Center for Trade and Sustainable Development) http://ictsd.org/i/publications/3086/. Accessed 2007.
Candau F, Jean S (2009) What are EU trade preferences worth for sub-Saharan Africa and other developing countries? In: Hoekman B, Martin W and Primo Braga CA (eds) Trade preference erosion: measurement and policy response. Palgrave-McMillan and The World Bank, Washington DC
Cardamone P (2007) A survey of the assessments of the effectiveness of Preferential Trade Agreements using gravity models. International Economics 60(4):421–473.
Cardamone P (2011) The effect on monthly fruit imports of preferential trade agreements granted by the European Union. Eur Rev of Agric Economics, forthcoming.
Carrère C, de Melo J (2004) Are different rules of origin equally costly? Estimates from NAFTA. CEPR Discussion Paper 4437 http://publi.cerdi.org/ed/2004/2004.12.pdf. Accessed 2004.
Cipollina M, Salvatici L (2010) The impact of European Union agricultural preferences. J of Econ Policy Reform 13:87–106.
Coe DT, Subramanian A, Tamirisa NT, Bhavnani R (2002) The missing globalization puzzle. IMF Working Papers 02/171 http://www.imf.org/external/pubs/cat/longres.cfm?sk¥16104.0. Accessed 2002
Felbermayr GJ, Kohler W (2006) Exploring the intensive and extensive margins of world trade. Rev of World Economics 142(4):642–674.
Francois J, Hoekman B, Manchin M (2006) Preference erosion and multilateral trade liberalization. The World Bank Econ Rev 20:197–216
Gallezot J (2003) Real access to the European Union’s agricultural market. http://www.tcd.ie/iiis/policycoherence/index.php/iiis/eu_agricultural_trade_policy/agricultural_protection_measures. Accessed 2003.
Gallezot J (2005a) Database on European agricultural tariffs, DBTAR. TradeAg Project Working Paper 05/07 http://tradeag.vitamib.com/. Accessed 2005.
Gallezot J (2005b) Database of EU preferential trade, TRADEPREF. TradeAg Project Working Paper 05/08 http://tradeag.vitamib.com/. Accessed 2005.
Gallezot J, Bureau JC (2005) The utilisation of trade preferences: The case of agricultural and food products in the EU and the US. OECD, Paris.
Helpman E, Melitz M, Rubinstein Y (2008) Estimating trade flows: trading partners and trading volumes. Q J of Economics 123(2):441–487.
IEG (2006) Assessing world bank support for trade, 1987–2004. World Bank.
Inama S (2003) Trade preferences and the world trade organization negotiations on market access. J of World Trade 37(5):959–976.
Low P, Piermartini R, Richtering J (2005) Multilateral solution to the erosion of non-reciprocal preferences in Nama. WTO Staff Working Paper, ERSD-2005-05, http://www.wto.org/english/res_e/reser_e/wpaps_e.htm. Accessed 2005.
Manchin M (2006) Preference utilisation and tariff reduction in EU imports from ACP countries. The World Economy 29:1243–1266
Nilsson L, Matsson N (2009) Truths and myths about the openness of EU trade policy and the use of EU trade preferences. Working Paper http://trade.ec.eu.europa.eu/doclib/docs/2009/july/tradoc_143993.pdf. Accessed 2009
Norton EC, Wang H, Ai C (2004) Computing interaction effects and standard errors in logit and probit models. Stata J 4(2):154–167.
OECD (2005) Preferential trading arrangements in agricultural and food markets. The case of the European Union and the United States. Paris, ISBN 92-64-00932-9.
Persson M, Wilhelmsson F (2007) Assessing the effects of EU trade preferences for developing countries. In: Bourdet Y, Gullstrand J, Olofsdotter K (eds) The European Union and developing countries. Trade, aid and growth in an integrating world. Edward Elgar Publishing
Piermartini R, The R (2005) Demystifying modelling methods for trade policy. World Trade Organization, Discussion Paper 10, http://www.wto.org/english/res_e/reser_e/discussion_pa-pers_e.htm. Accessed 2005
Santos-Silva JMC, Tenreyro S (2006) The log of gravity. Rev of Economics and Statistics 88:641–658.
WDI. World Development Indicators (Washington, DC: World Bank, 2008) CD-ROM.
Acknowledgments
Financial support received by the “New Issues in Agricultural, Food and Bio-energy Trade (AGFOODTRADE)” (Small and Medium-scale Focused Research Project, Grant Agreement no. 212036) research project, funded by the European Commission, and by the Italian Ministry of Education, University and Research (Scientific Research Program of National Relevance 2007 on “European Union policies, economic and trade integration processes and WTO negotiations” – PUE&PIEC) is gratefully acknowledged. We are indebted to Giovanni Anania, Luca De Benedictis and Luca Salvatici for their precious observations and suggestions, which reshaped the original manuscript. Special thanks go also to Paola Cardamone, Anne Célia Disdier and Sophie Droguè for their useful comments. Any remaining errors are solely our responsibility.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2011 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Agostino, M., Demaria, F., Trivieri, F. (2011). Trade Impact of European Union Preferences: The Role of Compliance Costs. In: De Benedictis, L., Salvatici, L. (eds) The Trade Impact of European Union Preferential Policies. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-16564-1_10
Download citation
DOI: https://doi.org/10.1007/978-3-642-16564-1_10
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-16563-4
Online ISBN: 978-3-642-16564-1
eBook Packages: Business and EconomicsEconomics and Finance (R0)