Antidumping protection: Good for bad firms but bad for good firms

Hylke Vandenbussche, 3 October 2008



While for the past 20 years the world has seen a drastic fall in tariff barriers over the past twenty years, protection is still around – albeit in a different form. The fall in tariffs has coincided with a spectacular increase in the number of antidumping measures. Antidumping duties are now the most frequently used instrument of trade protection.

In principle, antidumping duties can only be applied when the foreign trade partner is charging lower prices abroad than in its own market. This is regarded as an “unfair” dumping practice that needs to be regulated. However, current antidumping rules are not well equipped to distinguish between “fair” and “unfair” trade. When foreign producers produce goods more cheaply, their prices are bound to be lower, especially when they export to a large market like the US or the EU where they are likely to face more competition than in their own domestic markets. What appears to be unfair trade may well be an indication of foreign comparative advantage.

Due to the poor distinction between “fair” and “unfair” trade, one cannot exclude that antidumping protection can be “abused” to shelter uncompetitive domestic industries from more efficient rather than “unfair” foreign importers (Shin 1998).

The recent availability of firm-level datasets now makes it possible to verify whether antidumping protects “inefficient” industries. Moreover, we can assess the efficiency of domestic firms that receive protection, which usually lasts around five years. I do so in recent work co-authored with Jozef Konings (Konings and Vandenbussche, 2008).

Does antidumping protect inefficient firms?

Analysing over 4,000 EU producers involved in EU antidumping cases in the mid-1990s, we compare the productivity of firms that filed for and obtained antidumping protection to the productivity of similar firms that neither applied nor obtained protection. The average protected firm is much less productive than a similar firm prior to protection, which suggests that use of antidumping measures may have little to do with “unfair” practices by foreign firms. Antidumping duties protect less efficient EU industries, substantiating charges that antidumping is used for industrial policy purposes. Interestingly, firms that file for protection but do not get it appear to have higher productivity levels prior to filing than the firms that file for and receive protection. This seems to suggest that, amongst firms that file for protection, the European Commission grants it only to the least efficient producers. The productivity ranking of firms prior to filing is illustrated in Table 1 below. In the left margin we distinguish three groups of firms: 1) matched firms (no filing, no protection), termination firms (file but no protection) and protected firms (file and protection affirmed). The first column gives the average total factor productivity in each group prior to filing.

Table 1. Firms’ total factor productivity and trade protection

  Productivity before filing Productivity after filing
Matched control group    
Mean 2.23 2.32
Median 1.43 1.53
Standard Deviation 2.55 2.63
Termination cases    
Mean 1.46 1.43
Median 1.14 1.18
Standard Deviation 1.51 1.13
Affirmative AD cases    
Mean 1.32 1.55
Median 1.10 1.23
Standard Deviation 1.05 8.65

Antidumping protection’s impact on productivity

Can antidumping protection improve the efficiency of the protected firms? Can temporary protection turn around the fortune of EU firms by closing the efficiency gap described above?

Our results show that the productivity of the average domestic firm improves during antidumping protection. However, the productivity increase is never sufficient to close the efficiency gap with similar firms belonging to other sectors. The productivity ranking between the various groups of firms as they occur at the end of the protection period is illustrated in column 2 of Table 1. The productivity of matched firms remains far higher than productivity in the groups of firms that filed for protection. Productivity gains only raise protected firms to the level of firms that requested but did not receive protection. Based on these results, one can question the desirability of protection. In the absence of protection, some firms from the filing industries likely would have exited. The resources freed by their exit would be reallocated towards more efficient sectors in the economy, resulting in a larger productivity increase.

Antidumping: Good for bad firms, but bad for good firms!

Average effects can hide underlying firm heterogeneity. The initial firm level productivity distribution of protected firms is shown in Figure 1 and reveals another interesting result. The horizontal axis is firm efficiency, normalised from zero to one. There are clearly substantial differences between protected firms in terms of initial productivity. The productivity distribution is skewed to the left suggesting that the majority of firms have an efficiency level, measured by their distance from 1, well below the efficiency of the best firm in their industry.

Figure 1. Distribution of good and bad firms

Allowing for firm heterogeneity in the regression analysis suggests that the effect of antidumping protection on firm-level productivity, while positive on average, is negative for highly efficient firms Hence it appears that antidumping protection is good for bad firms but bad for good firms!

There are several plausible explanations. The threat of exit is one element that clearly differs across firms with different efficiency levels. The most efficient firms will face little or no threat of exit while the least efficient may be much more pressured to take action during protection to try and avoid exit. This may explain why the least efficient firms increase productivity during protection more than the most efficient domestic firms. Another possible explanation could be related to the correlation between efficiency and exporting. More efficient firms are more likely to export (Melitz, 2003; Melitz and Ottaviano, 2005). When the EU imposes antidumping protection against its trade partners, its own exporters may face retaliation abroad (Vandenbussche and Zanardi, 2008). This retaliation will reduce the sales of highly efficient EU exporters, which could explain the loss in productivity of highly efficient EU firms during protection. These possible explanations merit further investigation

Firms’ responses to trade policy

We do not understand how firms respond to trade policy changes very well, although interesting results are emerging. Do firms change their product mix (Bernard et al., 2006)? Do they change the skill composition of workers (Monfort et al., 2008)? Do they increase R&D investment? Do they engage in higher value-added activities (Schott, 2008)? Does protection lower domestic firms’ exit probability?

Antidumping protection seems to have become a tool of industrial policy aimed at fostering the interests of inefficient industries. Therefore, policymakers should at least know how effective the duty instrument is with respect to their objectives e.g. fostering the interests of domestic firms. At present, very little is known about within firm-level responses to trade policy, and this clearly merits further analysis.


J. Konings and H. Vandenbussche (2008),”Heterogeneous Responses of Firms to Trade Protection”, Journal of International Economics, forthcoming. CEPR Discussion Paper 6724

M. Melitz (2003). “The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity”, Econometrica, vol. 71(6), pp. 1695-1725.

M. Melitz and G. Ottaviano (2005), “Market Size, Trade and Productivity “, NBER working paper 11393, June.

A.B. Bernard, S. Redding and P. Schott (2006),”Multi-Product Firms and Product Switching”, revised version of NBER-paper 9789.

P. Monfort, H. Vandenbussche and E. Forlani, (2008), “Chinese Import Competition and Skill-Upgrading in the Textiles industry”, LICOS-discussion paper, KULeuven.

P. Schott (2008), “The Relative Sophistication of Chinese Exports”,Economic Policy, 53, 5-49.

H. J. Shin (1998),”Possible Instances of Predatory Pricing in Recent US Antidumping cases” in R. Lawrence (ed.), Brookings Trade Forum 1998, Brookings Institute Press, pp. 81-88.

H. Vandenbussche and M. Zanardi (2008), “What Explains the Proliferation of Antidumping Laws?”, Economic Policy, January

Topics: International trade
Tags: antidumping, heterogeneous firms, total factor productivity

Hylke Vandenbussche

Professor of International Economics, Université Catholique de Louvain-la-Neuve; and Research Fellow, CEPR

Vox Talks