Labour’s share is falling everywhere

One of the long-accepted stylised facts in economics is the constancy of labour’s share in national income. But that is not valid anymore. Labour’s share (wages and other benefits to workers) in GDP has been falling since the early 1990s in most advanced countries. Further, this has been the case in many emerging economies (though not all) as well at a time when these countries were growing at a fast rate.

The gap in per capita income between the developed and many of the developing countries narrowed and millions were pulled above the poverty line. Hence, it cannot be considered a simple case of developed country labour losing and developing country labour gaining as a result of competition from cheap third world labour, following globalisation of trade, migration and capital movements. Though this is currently the popular political narrative in the US and Europe, elevating people such as Donald Trump and Marine Le Pen to prominence, the story is more complex.

Analytically speaking, the falling share of labour in GDP is a reflection of real wage of labour growing slower than the average productivity of labour.

The traditional trade theory model suggests that when trade opens up between a high-wage developed country and a low-wage developing country, the wage rate (and hence labour’s share since the quantities of labour and capital are assumed to be fixed and fully employed in such models) would go down in the former and up in the latter. This model fails to represent the current global reality in several respects, in particular because (unlike in the model) labour-saving technological progress is taking place at a fast pace and capital, along with technology, is moving freely across the globe. Hence, we cannot expect the results of the traditional model to hold in today’s world.

In addition, in the real world, there are many countries at different levels of development with different wage rates. So, if, for example, trade opens up between the US (highest wage), Mexico (medium wage) and China (lowest wage), it is possible that the Mexican wage rate would go down as a result of competition from Chinese labour while the Chinese wage rate goes up and the US wage rate falls.

So, globalisation may have different effects on the wage rate and labour’s share on different countries within the labour-abundant developing world, even if we retain all the assumptions of the traditional model. The same may be the case within the capital-abundant developed world as it consists of countries with different prices of labour and capital.

A recent IMF empirical study ( IMF World Economic Outlook , April 2017), using a sample of 49 countries (31 advanced and 18 emerging), focusses on three major drivers of change in labour share: technological change (cheapening of machines relative to labour leading to substitution of capital for labour); global integration in trade (import competition, offshoring of activities, participation in global value chains), and finance (international mobility of FDI and portfolio funds); and unionisation and other policies/institutions/regulation (like corporate tax rate, competition policy). (Incidentally, India is not included in the sample and the impact of immigration of workers such as low-skilled Mexican labour and high-skill Indian H-1B engineers into the US are left out of the analysis.)

These drivers are not independent of each other. For instance, rapid improvement in transportation, communication and information technologies has facilitated global integration in trade and finance. The power of unions has gone down due to the possibility (or threat) of capital moving abroad and corporate taxes have been cut by many countries to attract foreign investment. Thus, it is difficult to separate the effects of these various forces. Subject to this caveat, the IMF study tries to estimate the relative significance of these drivers in explaining the change in labour’s share for both the developed and the emerging economies.

Technological progress has made machines cheaper and cost-effective to replace routine labour in many industries. This is found to be the most important factor in reducing labour’s share in the developed countries. Globalisation has played a much smaller role. The more labour-intensive parts of the production process (which cannot be replaced by automation) are being shifted to the developing countries. Even then the technology used by firms in the emerging economies for participating in the global value chains is typically more capital-intensive than the prevailing technology there. Thus, technological progress together with globalisation has served to increase capital’s share and reduce labour’s share in both the developed and the developing world.

Though this is the broad aggregate picture, considerable inter-country variation exists, depending on the exposure to automated technologies and the extent of participation in the global value chains. The roles of weakened trade unions and lower corporate taxes to lure investment have not been found to be significant in explaining declining labour share.

Another interesting question revolves around the income distribution within the broad class of ‘labour’. Here the finding is that the middle-skill workers have been the biggest losers in the developed countries as automation has replaced them while globalisation has also shifted these jobs to the developing world. These displaced middle-skill workers have been forced to take up lower-wage low-skill jobs. The high-skill workers ( IT professionals, doctors, lawyers, Wall Street bankers, company CEOs) have gained, even when the overall labour share has shrunk. In fact, if all the compensations of these people (specially the value of their bonuses and stock options) are included as part of labour income, it is doubtful (as economist Dean Baker shows) whether the labour share so defined has gone down at all.

But, then, it remains an open question whether the Wall Street bankers and company CEOs should be considered a part of ‘capital’ or ‘labour’ and the bonuses and the stock options of such people should be included in capital or labour income. Even many of the doctors, lawyers and accountants having their own practising firms may be earning profits, instead of wages or salaries. The enormous earnings of top film stars or athletes represent more of ‘rents’ (arising out of scarcity value) than wages.

Finally, we should not forget that in fast-growing developing countries with surplus labour (such as China and India), rising real wages along with rising employment may well be associated with falling labour share to the extent the rise in wages is less than the rise in labour productivity. In such a situation labour as a class is gaining in absolute terms, though losing in a relative sense. Opinions may differ on whether that should be a matter of major concern.

The writer was a professor of economics at IIM-Calcutta

COMMENT NOW