In recent weeks official bodies such as the World Trade Organisation and the European Commission as well as leading private sector associations – the International Chamber of Commerce (ICC) and the so-called B20 group of business leaders – have made strong statements concerning rising protectionism in the run up to the G20 summit in Los Cabos, Mexico. On the basis of most extensive update to the Global Trade Alert (GTA) database, that was conducted in preparation for this, the eleventh GTA report, they were right to do so.1
There has been a steady stream of protectionist measures introduced since the last G20 summit – at least 110 measures have been implemented, 89 of which were imposed by G20 members. This report demonstrates that the amount of protectionism in 2010 and 2011 was considerably higher than previously thought. An additional 226 protectionist measures were found in those two years, representing a 36% increase on the number of beggar-thy-neighbour policies implemented during 2010 and 2011.
Such protectionism translates into lost commercial opportunities, threatened jobs, and slower economic recovery. In very tangible terms, if the information available now had been known just six months ago – in November 2011 when our last report was published – then the number of times China’s commercial interests have been hurt by foreign protectionism in the three years following the November 2008 summit would have been increased by 105. Over the same time frame the US’ commercial interests were harmed 107 more times than previously thought. The understatement in previous reports of the frequency of harm done to many G20 countries’ commercial interests is of the order of at least 20%. Ultimately, what this means is that the world trading system did not settle down to low levels of protectionism after the spike in beggar-thy-neighbour policies in 2009.
What is more, the evidence presented in this report casts doubts on the strength of international restraints on the resort to protectionism by governments, in particular by G20 governments. There are two pieces of compelling evidence here.
- First, the share of the worldwide totals of protectionism implemented by the G20 countries has risen year-in and year-out. In 2009 60% of protectionism was implemented by G20 governments – that percentage has risen in the year to date in 2012 to 79%. Findings such as these cast the repeated G20 commitments to eschew protectionism in a particularly bad light. Some observers of the G20 have noted that these commitments have been demoted in the respective summit declarations and the GTA’s evidence reveals just how little priority the G20 countries have actually given to maintaining an open world trading system.
- Second, while there has been a sustained increase in the use of trade defence measures since the last G20 summit, resort to the traditional forms of protectionism that are relatively-speaking better regulated by the WTO account never exceeded 42% of measures implemented in any recent year. During the crisis era, then, governments have circumvented tougher WTO rules and used beggar-thy-neighbour policies subject to less demanding or no binding multilateral trade rules. Much of that discrimination is pretty non-transparent – that is, it is murky protectionism.
This finding does not imply that the WTO rules are useless, rather so long as they remain incomplete that circumvention is to be expected. If anything, the policy implication is that more far-reaching WTO rules are ultimately needed, even if there is little apparent appetite among governments for expanding the remit of multilateral trade rules at this time. It is probably safer to conclude that the WTO rules have altered the composition rather than the amount of protectionism in recent years.
Since official international initiatives amount to a weak bulwark against protectionism, any restraint is likely to have domestic sources. For sure, it would be desirable for G20 governments to start doing what they said they would do at their first summit in November 2008 – that is, refrain from protectionism – and having failed to do so to date, to unwind the protectionism that has been put in place. Moreover, peer pressure could and should be employed to rein those G20 countries that have engaged in extensive discrimination against trading partners.
Short of a major change of heart, the G20 is unlikely to deliver on these recommendations – and expectations should be moderated accordingly. The emphasis, then, must be on winning the argument for maintaining open borders in each major trading nation. Here business associations, consumer groups, and the media – supported by information provided by international organisations – should be at the fore of making the case against protectionism. The hard work in fighting protectionism is at the national level and not in writing reports for international summits. Information has its role, but it is not enough to limit the damage done to the relatively liberal world trading system created in the post-war era.
The remainder of this Executive Summary describes in more detail some of the key insights in this, the eleventh GTA report.
Not only is there no let up in the resort to protectionism – if anything 2010 and 2011 saw much more protectionism than previously thought
Figure 1 sheds light on the rate at which protectionist measures were imposed since the first crisis-era G20 summit in November 2008. By comparing the data from this report with the last one, there is a substantial upward revision in the totals for every quarter since Q4 2009. The immediate implication is that the levels of protectionism in 2010 and 2011 were higher than previously thought.
This finding helps make the point that evidence on protectionism takes time to assemble and instant assessments almost certainly underestimate the true amount of protectionism. Another implication is that the falling off of the total number of protectionist measures per quarter reported for 2011 and 2012 should not automatically be attributed to greater restraint on the part of governments. Similar initial findings of lower quarterly totals were found in earlier GTA reports. Those totals were revised substantially up over time.
Figure 1. After reporting lags the total number of protectionist measures implemented during 2010 and 2011 converged to the 100-120 range per quarter. Q4 2008 seems more anomalous as time goes on – protectionism jumped up in 2009 and has not returned to 2008 levels.
A sharp jump up in the number of protectionist measures was seen in Q1 2009, with over 170 measures imposed in that quarter alone. After that, the next four quarters saw more than 115 measures implemented. Each subsequent quarter from Q2 2010 to Q3 2011 – with the exception of Q3 2010 – has rapidly trended towards totals of 100-120 measures. Quarter-by-quarter governments have consistently added to the stock of crisis-era protectionist measures – Q1 2009 may have been exceptionally awful but it was hardly the only quarter when open borders came under threat.
Figure 1 also contains some good news. Quite a few – though certainly not a majority – of protectionist measures implemented from Q1 2009 to Q1 2010 are no longer in force. Still, approximately 100 protectionist measures implemented in each of those quarters remain in force. For reference, this total is double the number of protectionist measures implemented in Q4 2008 that are still in force.
The G20 is responsible for the bulk of crisis-era protectionism
Together the G20 countries have implemented 877 almost certainly discriminatory measures and 136 likely discriminatory measures since November 2008. Of the total (1013), 272 were trade defence measures. This implies that 73% of all protectionist measures implemented worldwide during the crisis era were not antidumping, countervailing duties, or safeguards actions.
Figure 2. The share of worldwide protectionist measures implemented by the G20 has risen every year since 2009.
Taken together the G20 countries are responsible for 66% of all protectionist measures taken since the first crisis-era G20 summit in November 2008. Of the protectionist measures still in force the G20 countries are responsible for 69% of the worldwide total. In 2009 approximately 60% of all protectionist measures were taken by G20 countries. By 2011 that percentage had risen to 75%. In the year to date, that percentage has risen further to 79%, see Figure 2.
Figure 3. The composition of state measures taken by each G20 country since November 2008, ranked by descending order in the proportion of protectionist measures implemented.
Countries may differ in the rate at which they announce state measures, so a simple count of the total number of protectionist measures may be misleading. One alternative is to compute for each G20 country the proportion of the total number of implemented measures that are almost certainly discriminatory (coded red in the GTA database), likely to be discriminatory (coded amber), and liberalising or transparency-improving (coded green) measures. Having sorted the G20 countries in descending order of the number of discriminatory measures (red plus amber), the results are portrayed in Figure 3. Compared to a ranking based on the total number of discriminatory measures implemented, Japan, Turkey, India, and Saudi Arabia move up the list of offending G20 nations. Argentina, the EU 27, and France remain highly ranked. This figure gives one way to compare the resort to protectionism across the G20 countries.
In terms of both the number of discriminatory measures imposed and the proportion of discriminatory measures imposed , the G20 country that has refrained from protectionism the most is Mexico, the host of the forthcoming summit.
G20 countries frequently in the lists of the top 10 worst offenders
The GTA does not calculate the amount of commerce affected by each state measure2, nor the associated welfare impact. To do so for 2430 state measures would be exceptionally resource intensive. Instead four intermediate metrics of harm done by a jurisdiction are reported: the number of almost certainly (coded red) discriminatory measures implemented, the number of tariff lines affected by almost certainly discriminatory measures, the number of sectors affected by almost certainly discriminatory measures, and the number of trading partners harmed by a jurisdiction’s almost certainly discriminatory measures. All the trading jurisdictions in the GTA database are ranked in descending order on these four metrics and the top 10 offenders on each category are reported in Table 1. Looking across the top 10 lists it is striking how often G20 members are mentioned.
In terms of discriminatory measures imposed, the EU 27 is the worst offender, in terms of tariff lines affected Vietnam is worst (due to repeated, transparent competitive devaluations), in terms of sectors affected Argentina tops the list (a combination of many protectionist measures), and in terms of trading partners harmed China tops the list (in part due to its extensive export management policies through selective VAT rebates for exporters). China and Argentina are represented in the top 10 list of offenders in all four categories of harm. Germany, India, and Russia are listed in three of the four top 10 lists of worst offenders.3
Table 1. Which countries have inflicted the most harm since November 2008?
|Rank||Metric, Country in specified rank, Number|
|Ranked by number of (almost certainly) discriminatory measures imposed||Ranked by the number of tariff lines (product categories) affected by (almost certainly) discriminatory measures||Ranked by the number of sectors affected by (almost certainly) discriminatory measures||Ranked by the number of trading partners affected by (almost certainly) discriminatory measures|
|1||EU27 (302)||Viet Nam (931)||Argentina (63)||China (193)|
|2||Russian Federation (169)||Venezuela (786)||Algeria (62)||EU27 (187)|
|3||Argentina (141)||Kazakhstan (732)||EU27 (57)||Netherlands (163)|
|4||India (74)||China (701)||China (52)||Germany (155)
|5||UK (67)||EU27 (656)||
|6||Germany (64)||Nigeria (599)||India (153)
|7||France (61)||Algeria (476)||Germany (44)|
|8||China (60)||Argentina (467)||
|9||Italy (56)||Russian Federation (446)|
|10||Brazil (54)||India (401)||Ghana (41)||Argentina (151)|
Note: There is no single metric to evaluate harm. Different policy measures affect different numbers of products, economic sectors, and trading partners. GTA reports four measures of harm.
Governments have circumvented tougher WTO rules
The traditional forms of protectionism – tariff increases and trade defence instruments – account for less than 37% of the worldwide total of discriminatory measures implemented since November 2008. Only in 2010 did that percentage breach 40% and then not by much (see Figure 4). In terms of measures still in force the respective percentage is 38.9%. In short, non-traditional forms of protection still dominate crisis-era protectionism. The fact that so many of the top 10 most used protectionist instruments are subject to weak or no WTO rules confirms an earlier GTA finding – namely, that governments under pressure during the crisis era have circumvented the more constraining binding multilateral trade rules (see Table 2.7 in the next chapter).
Figure 4. The circumvention of tougher WTO rules is a recurring feature of crisis-era protectionism.
Evenett, Simon J and Johannes Fritz (2011), “’Jumbo’ Protectionism and the Trade Coverage of Crisis-Era Protectionism”, in Simon J Evenett (ed.), Unequal Compliance: The Sixth GTA Report, June.
Evenett, Simon J and Johannes Fritz (2012), Die negativen Auswirkungen des krisenbezogenen Handelsprotektionismus auf die Schweizer Aussenwirtschaft. February. This study, which as the title suggest is written in German, is summarised in the second section of this Report.
Evenett, Simon J, Johannes Fritz, and Yang Chun Jing (2012), “Beyond Dollar Exchange Rate Targeting: China’s Crisis-Era Export Management Regime”, Forthcoming, Oxford Review of Economic Policy.
1Chapter 2 of the 11th GTA report provides an extensive overview of the key features of crisis-era protectionism.
2Some evidence on the amount of trade affected by certain (typically far-reaching) discriminatory measures can be found in Evenett and Fritz (2011, 2012), and Evenett et al. (2012). The common punch-line here is that the total amount of trade affected is far greater than estimates based on tariff increases and trade defence measures imposed, an approach taken by in some monitoring reports on protectionism.
3The results for Germany aggregate the (many) measures taken by the central government in Berlin and the (fewer) measures taken by the European Commission on behalf of all of the EU member states. This approach was applied consistently to each EU member state. At the request of some users, Table 1 reports information on both individual EU member states and the total amount of measures taken by all the member states plus the European Commission (listed as EU27).