Growing non-tariff and overall protection

Saileshsingh Gunessee, Chris Milner, Zhaohui Niu 19 June 2018



Trade liberalisation, in particular implemented following multilateral trade negotiations and growing membership of the WTO, but also associated with regional, bilateral, and unilateral trade reforms, is expected to have ushered in an era of increased globalisation. Indeed, several people have written about how globalisation has achieved new heights, talking of the ‘death distance’, a ‘flat world’, and the ‘great convergence’ (Caincross 2001, Friedman, 2007, Baldwin 2016). Yet a nagging question remains. What has happened to overall trade protection in the wake of the tariff reduction driven by this apparent trade liberalisation? After all, overall trade protection is more than tariffs, with non-tariff measures (NTMs) being as important. In light of this, some have pointed to the vexing fact that NTMs may not have fallen as one would have expected but – quite the contrary – have actually increased (Bacchetta and Beverelli2012). Indeed, the Trade Analysis and Information System (TRAINS) database report that about 2,852 product lines were subject to one NTM type in 2015, compared to 1,456 product lines in 1997. 

To get a complete picture of overall trade protection we therefore need to account for NTMs. However, the task can be daunting given the challenges of measuring NTMs beyond mere incidence and having measures that are comparable to tariffs. In a recent study we took up the challenge of measuring overall trade protection levels, where we estimate the ad-valorem equivalents (AVEs) of NTMs at the product level for several countries (Niu et al. 2018a). This builds on the seminal work of Kee et al. (2009) adding a time element, where we use a recently assembled database on the incidence of NTMs, namely, UNCTAD’s Multi-Agency Support Team(MAST). In short, this exercise enabled us to measure and track AVEs of NTMs and overall protection levels over time, with the data available here.

With this new measure, we can address several questions of topical interests. With the downward trend in tariffs, how have NTMs changed and what has been the change in overall trade protection over time? As implied by the anecdotal evidence outlined above, have NTMs and overall protection in fact increased? What happened to trade protection levels in the aftermath of the Global Crisis? Finally, is there a trade policy substitution between NTMs and tariffs?   

NTMs, tariffs, and overall protection 

Figure 1 depicts the ‘import-weighted’ average of tariff-equivalents of NTMs, tariffs, and overall trade protection levels for countries and products for the period 1997-2015. It reveals how non-tariff measures are evidently higher than tariffs through the period. In fact, tariff rates fell over the period from 10% in 1997 to 4% in 2015, a reflection of the often-heralded trade liberalisation. However, NTM trade protection was 22% in 1997 and grew as high as 51% in 2009. NTMs, while already an important source of trade protection at the beginning of the period, grew even more important over time. More importantly, overall trade protection has also increased, mirroring the rise in NTMs, albeit with some fluctuations. In short, the so-called trade liberalisation is in fact tariff liberalisation only offset by growing non-tariff trade protection. Indeed, overall trade protectionism has been rising over the last decade or so. 

The post-2008 period shows a sharp increase in NTMs and overall protection, despite a drop in average tariffs. To be precise, AVEs of NTMs rose on average from 25% to 51% from 2006 to 2009, with overall protection rising from 34% to 58% and average tariffs falling from 10% to 8%. This is indicative of a rise in trade protectionism in the immediate aftermath of the financial crisis, driven solely by NTMs. This observation is contrary to earlier findings of a ‘no pervasive increase in protectionism’ post-financial crisis (Kee et al. 2013). 

When reflecting on incidence of the core NTMs, the category ‘technical measures’ is the most applied followed by ‘quantity control’ and, to a lesser degree, ‘price control’ and ‘monopolistic measures’. This is reflected across countries and products as well (Niu et al. 2018a).    

Figure 1 Average AVE of NTMs, tariffs and overall protection, 1997-2015

Data source: Niu et al. (2018a)

Figure 2 Average AVE of NTMs across sectors, 1997-2015

Data source: Niu et al. (2018a)

NTM protection across sectors

How uniform is the pattern of changing protection identified above across manufacturing and agriculture? Figure 2 offers some clues. It shows the average AVEs of NTMs for agriculture and manufacturing. NTMs are generally higher for agriculture than for the manufacturing sector. While non-tariff protectionism in agriculture has been fluctuating, with a sharp rise post-2008, the manufacturing sector has experienced a steady increase in the average AVE of NTMs. Within manufacturing, the most NTM-protected activities are textiles, footwear, machinery and electrical equipment, and rubber and plastics (Niu et al. 2018). All in all, agricultural products have remained highly protected, while manufacturing products have become more protected. 

The fact that by 2015 textiles figure prominently as one of the most ‘protected’ sectors suggests that developed countries have found new ways to restrict trade in textiles to replace the quantity restrictions that were removed up to 2004 with the dismantling of the Multi-Fiber Agreement (MFA). Further, inspection of the TRAINS database reveals that ‘technical measures’ are now the predominant core-NTM applied to textiles (Niu et al. 2018a).     

Trade protection across country groupings

The evolution of trade protection can also be studied across countries grouped by regions and income types. The two panels in Figure 3 show country groupings along regional and income lines. From the charts, it is evident that average tariff levels have been relatively stable or moderately declining over time in most regions and for all groupings. By contrast, the upward path of overall trade protection has been driven in general by changes in NTM protection. It can be observed that, save for the Sub-Saharan Africa region, trade protection was higher in all regions and income groups at the end period compared to the start. By 2015, trade protection is estimated to be about 60% in both low-income and OECD economies (Niu et al. 2018a). North America shows a consistent trend of rising protectionism over the period, while most regions and income groups exhibit a fluctuating trend. The post-crisis protectionist tendencies are again reflected in most charts, with both NTM and overall trade protection rising sharply after 2006 and up to 2009.

Figure 3 Average AVE of NTMs across country groupings, 1997-2015 (decimal percentage)

Source: Figures 2 and 3 in Niu et al. (2018)

Trade policy substitution

The possibility of substitution of one type of trade policy measure by another is well grounded in both the theoretical and empirical literature on trade policy. It is natural to explore whether NTMs have increased as a result of tariff reductions or whether NTMs are unrelated to such tariff cuts. Therefore, in follow-up work to that outlined in this column so far, we formally investigate whether NTMs overall have substituted for tariffs over the recent period (Niu et al. 2018b). The findings suggest overall trade policy substitution, with higher substitution in the case of sectoral products with above-average tariff cuts as a result of the Uruguay Round. Though the relationship may appear small, in absolute terms the results indicate fairly complete substitution between the two trade policy instruments, given much higher levels of NTM than tariff protection. 


Trade protectionism has been rising over the last two decades, despite the perception of falling protection due to tariff cuts over this same period. In the aftermath of the 2008 financial crisis, non-tariff measures and overall trade protection peaked in 2009, but remained much higher at the end than at the start of the period. NTM protection does vary across sectors and countries, but the evolution of overall trade protection over this period has broadly been driven by changes in NTM protection, with tariff levels remaining stable or falling a little. 

Our AVE estimates confirm the anecdotal evidence that has suggested increasing rather than declining overall trade protection, mainly due to the greater use of non-tariff measures (Bacchetta and Beverelli 2012). Indeed, though not directly comparable, our above findings tally with the reported incidence of trade-impeding policy interventions in the Global Dynamics database and reports (e.g. Evenett and Fritz 2017) of the Global Trade Alert (see For instance, the number of new trade-impeding interventions mirrors what we find – a rise from 2009 to 2015 (from 274 cases to 648). 

Given the growing dominance of non-tariff protection over tariff protection, its clear policymakers need to pay careful attention to NTMs during trade negotiations, whether bilaterally or multilaterally involving the WTO. Multilateral agencies such as the IMF and the World Bank, as they tie aid allocation to trade reforms, should also consider non-tariff measures as part of trade protection indicators.


Bacchetta, M and C Beverelli (2012), “Non-tariff measures and the WTO”,, 31 July.

Baldwin, R (2016), The Great Convergence: Information Technology and the New Globalization, The Belknap Press of Harvard University Press. 

Bown, C (2012), “Import protection update: Antidumping, safeguards, and temporary trade barriers through 2011”,, 18 August.

Caincross, F (2001), The Death of Distance: How the Communications Revolution Is Changing our Lives, Harvard Business Review Press.

Evenett, S J and J Fritz (2017), Will Awe Trump Rules? The 21st Global Trade Alert Report, CEPR Press.

Friedman, T L (2007), The World is Flat: The Globalized World in the Twenty-First Century, Penguin. 

Kee, H L, C Neagu and A Nicita (2013), “Is protectionism on the rise? Assessing national trade policies during the crisis of 2008”, Review of Economics and Statistics 95(1): 342-346.

Kee, H L, A Nicita and M Olarreaga (2009). “Estimating trade restrictiveness indices”, Economic Journal 119(534): 172-199.

Niu, Z, C Liu, S Gunessee and C Milner (2018a), “Non-tariff and overall protection: evidence across countries and over time”, GEP Research Paper 2018/05, forthcoming in Review of World Economics.

Niu, Z, C Milner, S Gunessee and C Liu (2018b), “Are non-tariff measures and tariffs substitutes? Some panel data evidence”, GEP Research Paper 2018/07.



Topics:  International trade

Tags:  protectionism, non-tariff measures, NTMs, trade liberalisation

Assistant Professor in Economics, Nottingham University Business School China

Emeritus Professor of International Economics, University of Nottingham

Assistant Professor in Economics, Beihang University


CEPR Policy Research