Illustration EconPol Opinion Series

Trade Wars in a ‘Winner-takes-all’ Economy

Daniel Gros

Trump’s trade follies have attracted the attention of policy makers and markets across the world. This is only natural given that the US has switched from being a major underwriter of the global multilateral trading system to its number one enemy. But dramatic changes like this rarely happen without some deeper undercurrent that enables erratic politicians to overturn long-established structures and mechanisms. A closer look at the key issue between the US and China reveals that today’s ‘trade war’ is taking place in a different economic framework than in the past.

Current trade tensions originated in the steel sector. This is an ‘old economy’ industry par excellence, which is now plagued by enormous excess capacity, mostly in China. Excess capacity in the steel sector is a recurrent phenomenon and has always created frictions. The second Bush administration also imposed steep tariffs on imports of steel, but relented when a WTO panel ruled against the US. This is an example of how multilateral institutions manage a standard conflict. Trade hawks in the Trump administration view this ruling against the US as a loss. But most economists agree that not being allowed to impose unilateral tariffs on steel was good for the US economy, which does not gain from taxing a major input for so many other industries.

The real problem surfaced when Trump specifically targeted China with high tariffs, initially on USD 50 billion in imports under section 301 of the US trade law, which empowers the President to act if a US industry has sustained an injury due to unjustified actions by a foreign government. The key accusation levelled at China is that it forces major foreign high tech enterprises to reveal their intellectual property if they want to be present in the country. Why is this issue key for the US?
Consider the case of a dominant player in social networks or in search engines. The cost of entering a new market is essentially zero as the existing software can easily serve many more millions of users. All that needs to be done is to translate the interface in yet another language. In other words, new markets mostly mean higher profits for the ‘winners’. Requiring such firms to reveal their software would destroy their business model.  

In a competitive industry, by contrast, the cost of producing and selling more abroad is fairly close to the price, meaning that there is little excess profit to be lost. The gains of opening new markets are relatively limited in competitive industries as a result. Lobbying by potential exporters for better access to countries with high tariffs has largely been muted in the past, which is why India’s protectionism has never encountered much attention. But the owners of ‘winning’ intellectual property forego very large profits if a big market like China remains effectively closed.

This difference in profit opportunities explains why trade conflicts become more acute in a ‘winner takes all’ economy. In such an economy, trade policy is all about re-distributing rents; jobs and the interests of consumers come second. Under competitive conditions, by contrast, the overall productivity enhancing aspect of trade can outweigh the distribution of rents across countries.

Of course, ‘winner’ projects emerge out of a fiercely competitive process. Venture capitalists know that for every one winner who pays back handsomely they are likely to have lost money on dozens of ‘losers’, i.e. new ventures that looked promising, but did not come out on top. If venture capitalists know that a winner will be heavily taxed, they invest less in innovation. This is why it has long been known in economics that it is difficult to establish the optimal degree of protection of intellectual property and of the taxation of monopoly profits, as the advantage of making intellectual property widely available today has to be weighed against the incentive to invest in tomorrow’s knowledge.

But these intellectual arguments are lost when there are a few well-established global players that dominate an entire industry and enjoy (near-) monopoly rents.

In China, a bid to grab the rents of foreign multinationals has motivated the demand that they reveal their intellectual property if they want to enter the domestic market.

But why has Europe exerted so little pressure on the same US internet giants to reveal their intellectual property? This is not only due to a stronger respect of IPRs, but also to different incentives. In China, either shutting out or accessing the intellectual property of the high-tech foreign firms tends to benefit national champions, which are large enough to make the most of this knowledge. In Europe, by contrast, there are no ‘European’ champions, making it less attractive for EU authorities to follow the Chinese example.

In Europe, the temptation to grab part of these monopoly profits has taken the form of calls for what has inappropriately been called a ‘tax on the internet’. Proponents of this tax argue that profits should be taxed where they are made. The implicit argument is that profits are made where there are consumers. But this is an arbitrary criterion. The US firms can legitimately claim that their ‘European’ profits are just a return on their intellectual property, which can formally be localized anywhere, preferably in a low-tax jurisdiction. It is thus unlikely that any European ‘tax on the web’ will yield substantial revenues.
A priori, the EU and the US have a shared interest in protecting the intellectual property of their multinational firms operating in China. But the degree of interest is very different. European exporters and investors operate mainly in the manufacturing sector, where profit margins are low. Even ‘high-tech’ German car companies have single-digit profit margins. By contrast, the margins for the biggest US high-tech firms like Google or Amazon range between 50-80 %. This explains why European complaints have been much more muted.

This is also the reason why the looming ‘trade war’ promises to be asymmetric: the US is home to all the dominant tech firms with the fattest profit margins. The US will struggle to find allies against China, given that European and Japanese intellectual property is mostly in more competitive industries.  

Trade wars in the old economy might be an easy win for a country with a large trade deficit. But a war designed to force the rest of the world to open up, so that dominant domestic firms can earn higher rents will be a different kind of conflict.

The trade wars of the colonial era were usually fought for the benefit of (and sometimes even by) monopolist trading companies. Their intellectual property usually consisted of fiercely guarded maps and knowledge of trading routes.

Today, the US government is essentially lining up its diplomatic guns behind its internet giants, while Europe and China are baying for their monopoly profits. This is a zero-sum game, which can only turn negative sum through the collateral damage that it causes to the global trading system.

Daniel Gros
Director CEPS


Gros, Daniel, " Trade wars in a `winner takes all’ economy", EconPol Europe Opinion 7, May 2018