Globalisation and the redistribution of income

The labour share of total income has fallen in most developed countries over the last few decades. In colloquial terms, given the size of the cake, the portion given to the workers has shrunk over time. This trend has coincided in time with an increase in economic globalisation. Are both trends related? Can we blame the shrinking size of the cake received by workers to the rise in globalisation?

Globalisation may have affected the redistribution of income through different channels. The most direct one is the fact that workers compete now in a global arena, not just at a national level. There are several indirect channels but one of the most extensively studied is the declining weight of trade unions in developed economies.

Let us start with the first channel. The main effects of globalisation have been the increase in trade possibilities and the capacity to take advantage of the relative comparative advantages of each country. Some economists, see for example Baldwin (2012) and Basco and Mestieri (2013),(1) distinguish between two types of globalisation: (i) the first globalisation (from the end of the Second World War to the late 1980s) was characterised by a reduction in tariffs and (ii) the second globalisation (since the early 1990s), which was the result of the information technologies revolution and has reduced communication costs. Both types of globalisation have made trade between countries easier. However, one important difference between these two globalisations is that the second helped to break the value supply chain. The typical exchange of the first globalisation, in which the production process started and finished within each country, is the sale of textiles by England to France in exchange for wine. A typical example of the second globalisation is the iPhone, which is designed in the USA and made in China using components produced in different Asian countries. Moreover, the internet has changed the nature of tradeable goods. One example is the fact that many companies have moved their accounting unit or customer service department to other countries, or that some hospitals offshore the reading of their x-rays. These are services which, in all cases, are tradeable thanks to the internet revolution. In summary, the second globalisation has enlarged the basket of goods and services that can be traded and has made it possible for different stages of the production process to be carried out in different countries.

After analysing the particular features of each globalisation, we now need to examine whether their effects on wages are similar. Note that, in both cases, firms make more profits because they sell their product to a larger market and they can also reduce their production costs by offshoring production. For example, England could specialise in the production of textiles and thereby increase its earnings, and Apple can sell its products to all markets and produce each component for the iPhone in the cheapest country. However, although the size of the cake is larger in both cases, we do not know a priori the trend in the labour share. To answer this question, we need to identify the type of worker affected by each type of globalisation. In the first, the goods sold required a relative low-skill workforce, so that the workers most affected by foreign competition had lower qualifications. In contrast, in the second globalisation the traded goods require middle-skill workers, so that those workers with intermediate qualifications are hardest hit, to which we should also add the drawback that any new jobs they find tend to be lower qualified and with lower wages. In other words, globalisation has resulted in the type of products (or part of the production) made in the advanced countries requiring increasingly fewer workers and more capital. Consequently, both globalisations have similar qualitative effects but the second globalisation affects more workers and tends to exacerbate the decline in the labour share.

To check the validity of these hypotheses, we need to look at the figures. Several empirical studies (see the references in Basco and Mestieri, 2013) show how the globalisation process has caused «wage polarisation» (the relative wage of middle-skill workers has declined with respect to both high-skill and low-skill workers during the second globalisation). In other words, middle-skill workers have gone on to jobs with low qualification requirements and wages. This evidence in wage distribution suggests, but does not necessary entail, a change in the labour share. The best way to settle this question is by analysing whether those industries most exposed to foreign competition have suffered a larger decrease in their labour share. Elsby et al. (2013)(2) have carried out this exercise for the USA and confirm the hypothesis: the labour share has effectively fallen further in those industries where international competition has increased the most. Therefore, this confirms the idea that offshoring part of the production process has been a significant factor in the secular drop in the labour share. Given that globalisation is an ongoing process and competition at a global level will get stronger in the future, these findings mean that the best policy to ensure jobs remain in developed countries is by increasing worker productivity. As developed countries have a comparative advantage in more highly qualified work, investment in education would reinforce this advantage and help to relocate workers who have lost their jobs due to globalisation in less skilled employment. Another possible measure would be to enhance infrastructures to improve competitiveness, with the result that the wage gap compared with other undeveloped countries would have to be larger for companies to move their production there.

As will be remembered, the second indirect channel through which globalisation is claimed to have affected the redistribution of income has been the gradual loss of trade unions' influence in most developed countries. A formal analysis of the origins of this relative decrease in importance of trade unions is beyond the scope of this Dossier; we just assume that this is a fact and that globalisation, since it forces rich countries to compete with countries with less trade union protection, is one of the causes. However, we are interested in examining how the role played by trade unions affects how the cake is distributed. The initial premise is as follows: if trade unions lose presence, the relative bargaining power of firms increases and, consequently, workers receive a smaller share of the firm's production. Hence we could infer that the reduction in the relative weight of trade unions is associated with a reduction in wages. Once again, empirical evidence (in this case, the 2012 OECD report) should be used to confirm whether there is a direct relationship between the change in the percentage of workers affected by collective bargaining agreements and the change in the labour share. In the second diagram contained in this article no correlation can be observed. Note, for example, that wages have fallen sharply in Italy and Sweden but the change in collective bargaining coverage has been small. One possible criticism of this analysis would be that such effects cannot be observed between countries but only between different industries in the same country. Following the logic that trade union power differs from industry to industry, Elsby et al. (2013) analysed whether industries in the USA where trade union influence had shrunk more had also seen a greater reduction in the labour share, and they did not find any differential effect. Therefore, all the evidence suggests that the trade unions' loss of influence does not lie behind the reduction in the labour share.

In conclusion, to date we have benefitted from the positive effects of globalisation, such as the increase in products or being able to buy them at more affordable prices (i.e., a larger cake). However, this has been accompanied by a significant change in the distribution of income, particularly affecting the middle classes. Given this situation, education policy and competitiveness measures that improve the quality of production are crucial for the population as a whole to benefit from the positive effects of globalisation without this impairing the distribution of income.

Sergi Basco

European Unit, Research Department, "la Caixa"