The Davos lie

As I write these words, the great, the good and the self-important are trudging around in the Alpine slush, sporting their best parkas and a variety of silly hats, and opining about the state of the world. Davos is back. If there's one thing that people agree about in Davos, it's that globalisation is a Good Thing. And indeed, so it is, if the alternative is the autarky of the inter-war period. No one can deny that exporting has been the key to growth in China and other developing countries in recent years, or that this growth has raised hundreds of millions of our fellow human beings out of poverty. If you are even slightly cosmopolitan in your ethical outlook, you should want this to continue. But it always makes sense to ask whether you can have too much of a good thing. Standard international trade theory teaches us that trade increases overall incomes, but that not everyone benefits: instead, trade helps some groups in society, and hurts others. The textbooks then make the point that, since overall incomes have increased, the losers could be compensated by the winners, leaving everyone better off.1 If this were the way the world actually worked, then we would not have to worry about the income distribution effects of globalisation; but of course that is not the way the world works. ‘The Davos lie’ is how Larry Summers characterised the argument in 2007.2 Who are the winners from globalisation, and who are the losers? Let's consider international trade. According to the Swedish economists Eli Heckscher and Bertlin Ohlin, who were writing in the early twentieth century, countries with lots of land (and where land is consequently cheap) should export commodities whose production requires a lot of land; countries with lots of unskilled labour (and where unskilled labour is consequently cheap) should export commodities whose production requires a lot of unskilled labour; and so on.3 Countries with not much land (and where land is consequently expensive) should import commodities whose production requires a lot of land; countries without much unskilled labour (and where unskilled wages are consequently high) should import goods whose production requires lots of unskilled labour; and so forth. In the context of today's globalised economy, one implication of this ‘Heckscher-Ohlin theorem’ is that developing economies with lots of cheap unskilled labour should export textiles and other labour-intensive manufactured goods to rich economies where wages are high. A second implication is that labour-intensive industries should go into decline in rich countries. A third implication is that this should lower the demand for unskilled workers, hence lowering unskilled wages and increasing inequality. As is well known, many Western societies have become more unequal over the past two decades, although this is not uniformly true. During the 1980s and 1990s, the consensus was that this growing inequality was due not to international trade, but to technological change that was systematically favouring skilled over unskilled workers. One strike against the Heckscher-Ohlin version of events was that the skill premium (the gap between skilled and unskilled wages) was rising in developing economies such as Mexico: if Northern inequality were being increased as a result of rising exports of unskilled-labour-intensive goods from the South, then those same exports should be driving up unskilled Southern wages, and lowering Southern inequality. That is not in fact what happened. More recently, however, the debate has swung back towards the view that trade is important in explaining rising inequality, not only in rich countries, but potentially in developing economies such as Mexico as well. For one thing, Mexico is competing against countries like China, not just against countries like the United States. For another, the outsourcing activities of multinational firms may be reallocating labour tasks between countries in ways that traditional Heckscher-Ohlin trade theory did not take account of. For example, very low skill service jobs, and high skilled jobs involving abstract tasks, may be difficult to outsource, but middle-ranking routine tasks may be much more easily so. And indeed there seems to have been a ‘hollowing out’ of the income distribution in recent years: wages at the top have been pulling away from average wages, while in some cases low wages are converging somewhat on the average. I happen to think that inequality matters for its own sake, but even if you don't agree with that value judgement you should still care about inequality, since it matters politically as well. If trade is lowering the living standards of substantial sections of the community, then they will object to this. If they are sufficiently powerful, an anti-globalisation backlash is predictable. These are not mere theoretical speculations. During the late nineteenth century, for example, a move towards more liberal trade policy in Western Europe came to a shuddering halt in the late 1870s as cheap grain started to flow in from the Ukraine and the New World, hurting peasant farmers and aristocratic landlords.4 And this is why the Davos lie is so important to true believers in completely unfettered international markets: acknowledging that trade (or other dimensions of globalisation, such as migration) could create losers as well as winners might leave the door open to such an anti-globalisation backlash in the future as well. Unfortunately for Davos, globalisation's losers are becoming increasingly hostile to trade (and immigration) regardless. Over the past decade, political scientists and economists have amassed a considerable body of survey evidence which shows that ordinary people's attitudes towards globalisation are exactly what Heckscher-Ohlin economics would predict. The paper which sparked off this literature was published in 2001 by a political scientist, Princeton's Kenneth Scheve, and an economist, Dartmouth's Matthew Slaughter.5 Scheve and Slaughter were interested in whether Americans’ attitudes towards globalisation were better explained by the sector they worked in, or by their class (that is, by their skill level). The results were unambiguous: class was a much better predictor of people's attitudes towards trade than the sector they worked in. Specifically, the better educated and higher skilled were much more likely to support free trade than less well educated and lower paid workers. By contrast, it mattered little for trade attitudes whether workers were employed in export or in import competing sectors. Sectoral considerations were only important for home owners, the value of whose houses depended on the fortunes of the particular industries operating in their county of residence. The finding that blue-collar workers are more hostile to trade than the better educated and higher skilled is exactly what Heckscher-Ohlin theory would predict, but one could argue that on its own, this finding proves nothing. Might the correlation be driven by completely different mechanisms? Might it be, for example, that the better educated are more likely to support free trade not because they are more likely to benefit from it, but because they are better informed about its virtues? If this were the case, then presumably the better educated would be more likely to support free trade everywhere, not just in the United States. This is where an international perspective turns out to be particularly useful. If economics and self-interest are driving the correlation, in Heckscher-Ohlin manner, then the correlation should vary across countries in a predictable way. It is in the richest countries, where unskilled wages are the highest, and skills are most abundant, that blue-collar workers should be most hostile to free trade, and skilled workers should be most in favour of it. The correlation should be weaker in poorer countries, and reverse itself in the poorest: in very poor countries, where unskilled wages are low and skills are scarce, it should be the unskilled who are in favour of free trade, since it provides them with jobs in export-oriented industries. In such countries, by contrast, skilled workers should want to maintain the privileges which being sheltered from international competition affords them. And this is precisely what international survey evidence finds.6 Far more important than such academic studies is the fact that such attitudes are now beginning to influence politics in several rich countries. An early example came in 2005, with France's referendum on the so-called European Constitutional Treaty. Although the reforms the treaty was supposed to introduce were largely procedural, the debate ended up being largely about the nature of the European integration process, seen as a regional example of globalisation more generally. Left-wing opponents of the treaty pointed to the outsourcing of French jobs to Eastern Europe, and denounced the plans by Frits Bolkestein, an EU commissioner, to create a Europe-wide market for services. Fears were raised about the impact of competition from Polish plumbers on the living standards of blue-collar French workers, while it was argued that unfair competition from low-tax, light-regulation economies would lead to ‘social dumping’, which would ultimately place the French welfare state under threat. These arguments proved decisive, and in a major shock for Europe's political establishment the referendum was rejected by a margin of 55 per cent to 45 per cent. A few days later Dutch voters rejected the Constitutional Treaty by an even bigger margin. What is particularly striking about the French results is the way that voters divided along class lines. Only 35 per cent of professionals voted against the Treaty, while the figure was 53 per cent for middle management, 67 per cent for clerical workers, and an overwhelming 79 per cent for blue-collar workers. A similar class divide was apparent in the 2008 Irish referendum on the successor to the Constitutional Treaty, namely the Lisbon Treaty. In affluent areas of Dublin's southside, over 60 per cent of voters supported the treaty, while more than 60 per cent of voters in working-class areas opposed it. There are at least two ways of interpreting such patterns. The first would hold that well educated voters are more politically sophisticated, and better able to understand the issues involved in a complex amendment to the institutional underpinnings of the European Union. The second interpretation is that, on the contrary, both rich and poor are capable of correctly discerning where their economic interests lie, and vote accordingly. The argument would be that globalisation generally, and European integration more narrowly, has overwhelmingly favoured skilled workers, at least in affluent countries such as France, Ireland and the Netherlands. Unskilled workers, by contrast, feel under threat from Romanian (or Asian) competition, or immigration from Eastern Europe and further afield. Subsequent analysis of survey data carried out on behalf of the Irish government found some support for both of these positions. On the one hand, there was clear evidence that the more information people had about the contents of the treaty, the more likely they were to vote in favour of it. On the other hand, concerns about low wages were an important factor leading people to vote no – so long as they were in the labour force. Similarly, university education led people to vote yes, but only in the case where they were in the labour force. If education mattered for voting behaviour only through its allowing people to understand complex issues, it should lead to people voting yes whether or not they are active in the labour market. The fact that education and concerns about lower wages matter only for workers and the unemployed – that is to say, for people whose living standards are directly affected by European integration, and globalisation more generally – suggests that economic interests were playing an important role in these referenda as well. Economists can tut-tut all they want about working-class people refusing to buy into the benefits of globalisation, but as social scientists we surely need to think about the predictable political consequences of economic policies. Too much globalisation, without domestic safety nets and other policies that can adequately protect globalisation's losers, will inevitably invite a political backlash. Indeed, it is already upon us

Argument (complex analysis) · Development economics · Economics · Globalization · Political economy · Political science · Poverty · Sociology · Global Financial Crisis and Policies · Law