VoxEU Column EU policies International trade

Trade preferences need predictability

Uncertainty over the future course of trade policy really matters. Developing countries obtain non-reciprocal tariff reductions from many industrialised countries through Generalised Systems of Preferences, but the uncertainty surrounding the continuation of preferential treatment is likely to undermine the schemes’ effectiveness. This column studies the 2014 reform of the EU’s Generalised System of Preferences, which removed the uncertainty on preferences for a subset of beneficiary countries. It finds that the removal of uncertainty led to an increase in EU imports from affected countries by about 45% without any changes in other market access conditions. This average trade impact is driven by country-sector pairs that are most exposed to preferences uncertainty prior to the reform, for which trade increased by over 70%.

Uncertainty about future conditions can be a major factor hampering economic activity. Most recently, the COVID-19 pandemic has forcefully brought to our attention the detrimental effects of uncertainty and its likely adverse effects on world output (Baker et al. 2020, Caggiano et al. 2020). In the realm of trade policy, tensions caused by tariff hikes and retaliations between major trade partners, as well as the prolonged Brexit negotiations, have been sources of an unprecedented rise in uncertainty, resulting in stock market volatility, lower trade volumes, and general concerns about investment and growth (Baker et al. 2019, Caldara et al. 2020, Crowley et al. 2018). 

Concerns about the detrimental effects of policy uncertainty surrounding in particular tariffs and trade had been raised before the exceptional turmoil of the last two years. For instance, Handley and Limão (2015, 2017) show how the elimination of uncertainty on tariffs entailed by Portugal’s accession to the EU and China’s accession to the WTO, respectively, explains a large fraction of the subsequent increase in exports. Uncertainty matters for firms making irreversible investment decisions to enter a foreign market, expand their product scope, or to develop a new technology. Domestic firms can also be affected by trade policy uncertainty through changes in prices, employment and productivity.

Policy uncertainty in GSP schemes and the 2014 EU GSP reform

A specific source of uncertainty arises in the context of non-reciprocal preferences schemes. Generalised Systems of Preferences (GSP) schemes were created to stimulate export-led growth in developing countries. To this end, donor countries unilaterally apply tariffs that are lower than most-favoured nation (MFN) rates (i.e. preferences) on imports from beneficiaries of the schemes.1 Yet a significant flaw of GSP schemes is the uncertainty that surrounds these preferences: 

  1. The non-reciprocity of the schemes confers donors some discretion over choosing countries and products eligible for reduced tariffs, as well as the power to revoke the schemes altogether (Grossman and Sykes 2005).
  2. The schemes typically require periodic renewals (Hakobyan 2020). 
  3. Some schemes feature mechanisms to revoke the preferential tariffs from countries or sectors that become internationally competitive (Hakobyan 2017). 

These aspects of uncertainty about future trading conditions are likely to undermine the development scope and the effectiveness of GSP schemes altogether, as they might induce underinvestment in products eligible for the preferential regime.

We show that removing the uncertainty surrounding GSP preferences leads to a substantial increase in trade by beneficiary countries (Borchert and Di Ubaldo 2020). To estimate this effect, we exploit a particular feature of a reform with which the EU amended its GSP scheme in 2014. Amongst other things, the reform changed the conditions for ‘graduation’ – that is, the loss of preferential treatment for a country-sector when it becomes ‘too competitive’ (i.e. when its import share exceeds a certain threshold2 it is deemed no longer in need of preferential treatment and will therefore have its preferential tariff status removed). This lingering prospect of graduation could act as a barrier, deflecting GSP beneficiaries’ trade away from the threshold. However, in 2014 the EU eliminated the threat of competitiveness-related graduations for some countries, namely, the beneficiaries of the GSP+ arrangement (a sub-scheme of the main GSP). GSP+ members were thereby given the certainty of their preferential access to the EU, regardless of the size of their exports to the EU, relative to those of other members. 

We find that the 2014 GSP reform led, on average, to an increase in EU imports from GSP+ countries by 45%. In order to isolate the role of uncertainty removal from other aspects of the reform, we consider the concomitant changes in tariffs applied on products shipped by GSP+ countries; however, we find a 35% increase in EU imports also for those products with no change in tariffs. In the absence of any observable change in market access conditions, and subject to a triple difference-in-difference estimation design, we attribute the estimated trade impact to the removal of uncertainty.

Moreover, the average impact on trade hides considerable heterogeneity. That is because country-sector pairs with pre-reform import shares closer to the graduation threshold are more starkly affected by uncertainty over future trading conditions. In these cases, exports may have been held back in order not to lose their GSP status. Indeed, we find that imports from country-sector pairs closer to the threshold (within 7.5 percentage points of it) increased by about 71%.3

It could have been conceivable that the post-reform increase in trade to the EU came at the expense of other export destinations, i.e. a kind of trade diversion. Yet we find no evidence of beneficiaries simply having re-directed their exports from alternative destinations; rather, GSP+ countries appear to have taken advantage of more secure (and occasionally better) access to EU markets by actually increasing their overall exporting activity. 

Potential implications 

Given how GSP+ countries benefited from the 2014 reform, one might ask whether similar gains could potentially be reaped by economies in the standard GSP scheme, for which preference uncertainty continues to prevail. As an illustrative example, Figure 1 shows the import shares of the four largest Indian sectors over the period preceding the 2014 EU GSP reform.4 The import shares grew over time but seem to be hovering just below the 15% threshold, which would have triggered the loss of preferential access to the EU. India is a standard GSP beneficiary and as such was not affected by the reform described above. Yet in light of our empirical results obtained for GSP+ countries, India and other standard GSP beneficiaries could potentially benefit a great deal from uncertainty removal. 

Figure 1 EU import shares from India by top sectors, 2004-2013



Our findings demonstrate that trade policy uncertainty adversely affects trade flows to a significant and quantitatively important extent. For exporters in developing countries to take full advantage of the benefits offered by non-reciprocal preference schemes, there needs to be a sufficiently high degree of stability in trading conditions. Based upon our results, we conjecture that beneficiary countries would benefit from further reform that eliminated discretionary elements from GSP schemes. 


Baker S, N Bloom and S Davis (2019), “The extraordinary rise in trade policy uncertainty”, VoxEU.org, 17 September.

Baker S, N Bloom, S Davis, and S Terry (2020), “COVID-induced economic uncertainty and its consequences”, VoxEU.org, 13 April.

Borchert, I and M Di Ubaldo (2020), “Go ahead and trade: the effect of uncertainty removal in the EU's GSP scheme”, ESRI Working Paper 691.

Caggiano G, E Castelnuovo, R. Kima (2020), “The global effects of Covid-19-induced uncertainty”, Economics Letters 194.

Caldara, D, M Iacoviello, P Molligo, A Prestipino and A and Raffo (2020), “The economic effects of trade policy uncertainty”, Journal of Monetary Economics 109: 38-59.

Frazer, G and J Van Biesebroeck (2010), “Trade growth under the African Growth and Opportunity Act”, The Review of Economics and Statistics 92(1): 128-144.

Grossman, G M and A O Sykes (2005), “A preference for development: the law and economics of GSP”, World Trade Review 4(1): 41-67.

Crowley, M A, O Exton and L Han (2018), “Renegotiation of Trade Agreements and Firm Exporting Decisions: Evidence from the Impact of Brexit on UK Exports”, Cambridge Working Papers in Economics 1839.

Hakobyan, S (2020), “GSP expiration and declining exports from developing countries”, Canadian Journal of Economics/Revue canadienne d'économique 53(3): 1132-1161. 

Hakobyan, S (2017), “Export competitiveness of developing countries and US trade policy”, The World Economy 40(7): 1405-1429.

Handley, K and N Limão (2017), “Policy uncertainty, trade, and welfare: Theory and evidence for China and the United States”, American Economic Review 107(9): 2731-83.

Handley, K and N Limão (2015), “Trade and investment under policy uncertainty: theory and firm evidence”, American Economic Journal: Economic Policy 7(4): 189-222.

Thelle, M H, T Jeppesen, C Gjodesen-Lund and J Van Biesebroeck (2015), “Assessment of economic benefits generated by the EU Trade Regimes towards developing countries”, European Commission, DG-DEVCO.


1 The relevance of GSP schemes is substantial; for instance, the share of EU imports that originate from beneficiaries of its GSP exceeded 57% in 2009, whereas imports from members of EU free trade agreements accounted for only 16% of EU imports, and trade partners without any agreement made up the remaining 27%. Recent studies have found that awarding non-reciprocal trade preferences has a positive effect on exports from eligible countries and products (Frazer and Van Biesebroeck, 2010; Thelle et al., 2015).

2 This threshold was initially set at 15% (12.5% for textiles), and then revised upwards in 2014 (17.5%, 15% for textiles), and in 2015 (57%, 47.2% for textiles, 17.5% for minerals, live plants, animal and vegetable oils) upon the exit of large beneficiaries from the scheme, notably China in 2015. 

3 This is the marginal effect computed as eβ-1, where β is the estimated coefficient on the triple interaction term of a GSP+ import flow near the threshold that graduated, after the reform relative to before.

4 We chose to focus on India as this country is currently the largest GSP beneficiary and, therefore, features the largest number of sectors close to the graduation threshold.

735 Reads