Ask the expert: US wage stagnation

Listen to this article


We have recently witnessed a flurry of comment in the US on the long-running stagnation of wages. Many believe that the future livelihood of the “middle class” is also at risk, writes Jagdish Bhagwati, professor at Columbia University and senior fellow at the Council on Foreign Relations, in the Financial Times.

Prof Bhagwati rejects the view that globalisation has reduced the bargaining ability of workers and thus put a downward pressure on wages. “The argument is not relevant when employers and workers are in a competitive market and workers must be paid the going wage”, he writes.

“The culprit is not globalisation but labour-saving technical change that puts pressure on the wages of the unskilled. Technical change prompts continual economies in the use of unskilled labour...There are assembly lines today, but they are without workers; they are managed by computers in a glass cage above, with highly skilled engineers in charge.”

Do you agree? Prof Bhagwati answers your questions below.

Jagdish Bhagwati has suggested that it is not globalisation but labour-saving technology that puts pressure on the wages of the unskilled. Certainly the data do not support the argument for globalisation (see below), and I would like to suggest that the main factor affecting the labour share of output in the US has been currency intervention.

Rapidly growing economies naturally have rising real exchange rates. This has been impeded by currency intervention, notably by China, while so far sterilising the monetary consequences to prevent inflation in China.

This has depressed the relative prices of goods and services where China has a comparative advantage, notably manufactured goods. These, as I showed in a response to Martin Wolf, have very high capital/output ratios relative to non-traded goods and services. The impact on relative prices has had several interesting results:

• The US and the UK have been able to have the lowest investment ratios and faster GDP growth rates over the past 15 years. But only at the expense of large and growing trade deficits.

• Their economies have been able to operate with narrowing output gaps, without increasing inflationary pressure. As profit margins move with output gaps, this has allowed them to trend upwards - as they have been doing since the late 1970s. (For work on the connection between the labour/profit share and inflation see Inflation Dynamics and the Labour Share in the UK by Nicoletta Batini, Brian Jackson and Stephen Nickell, published by the Bank of England External MPC unit discussion paper no. 2 November, 2000.)

• Low inflation and falling unemployment are the ideal conditions for high financial profits, notably from consumer lending, particularly with the associated increase in house prices. Financial profits have been the major factor in rising profit margins.

• The swing away from manufacturing profits has hit the major source of union supported median wages. Wage data suggest that low and high wage incomes have risen relative to the median, with high wages in finance being readily observed.

• Claims that China’s entry onto the world scene has caused profits to rise by increasing labour supply relative to capital does not fit the data. The labour share did not rise sharply ab initio only to fall away over time, and real interest rates have been on a falling rather than rising trend. (The entry of China seems to have boosted the supply of savings rather more than the demand for capital.)
Andrew Smithers, founder of Smithers & Co and contributor to the FT’s Economists’ forum

I cannot agree with Prof Bhagwati’s conclusions only because his assumptions are not so dressed up as we are used to in his writings. Nevertheless we had had in our history other globalised stages, in our today’s globalisation phenomenon we cannot dichotomise technology from globalisation and afterwards their separated impact on labour wages. Our present globalisation is only on this stage because has the technology to support and strengthen her. Separate these two realities ex-ante it’s to subvert the future conclusions of this or other similar debates on the area. Here we can not use the ceteris paribus hypothesis, once both are part not the some body but part of the same hand.
Pedro Conceição Parreira, Portugal

Jagdish Bhagwati: You are right that I have not been able to provide the “dressed up” (ie extended) arguments that I provide when I am not confined to 800 words. But you will find fuller arguments in the Afterword to my new edition of the Globalization book.

But let me take issue with your notion that technology and globalisation cannot be treated independently. We are always using the technique of ceteris paribus and of using partial derivatives to isolte effects of different factors. I see no particular reason to think otherwise here.

Of course, I have not considered how, in a globalised world economy, techniques may diffuse from the rich to the poor countries, for example. Nor did I consider how technical change itself may be stimulated by trading with the poor countries.

On the latter question, let me just say that if trade with poor countries makes people feel that wages are going to be capped, the search for new techniques will be biased in favour of using more unskilled labour, not less, and that would help and not harm our workers’ wages!

If Tom Friedman is right and the world is (now) flat, what is to prevent most Americans from getting reduced paychecks over the next 25 years as the rest of the world’s workers catch up?
James L Foote, Miami, US

Jagdish Bhagwati: Tom Friedman is getting it wrong, I am afraid. Long ago, in an article in The New Republic, I identified several factors which were leading to intensification of competition: people with access to similar knowhow, integration of financial markets so that interest rates were not so far apart, etc.

But from this I inferred that there would be pressures to reduce cost differences (to flatten the world, in Friedman’s language) so as to moderate the fierce competition! So, today, we see the Democrats in the US Congress (AFL-CIO, ICFTU etc) trying to raise standards in the poor countries, almost entirely because they wish to raise the cost of production of their rivals there, while masking their demands in terms of altruism and empathy for the workers in these countries!

I haver already argued in my FT article that globalisation is not the real culprit in pressuring our workers’ wages; so if I am right, you need have no fears that low wages due to trade with poor countries is our destiny.

Your analysis notes a muted affect of globalisation on wages in the US. Does this suggest that those countries with more stringent labour market regulations (such as the northern Europeans) are also unlikely to face pressure on wages and thus carry on as normal?
James Sloan, London, UK

Jagdish Bhagwati: In answer to another reader below, I argued that the labour market regulation that makes firing staff more difficult is the real problem when globalisation produces volatility, and hence the necessity to close down a plant and have layoffs if you suddenly lose comparative advantage.

Even very efficient firms will then invest outside their country: a situation not uncommon in France and Germany. This also puts pressure on wages as firms will hire, if at all, only after discounting for the probability of expensive layoffs.

Also, volatility means high turnover where labour laws are more relaxed. So, firms are unlikely to invest in the human capital of their unskilled workers. Unions therefore have to now provide the human capital, which led to a rising-pay profile, in other ways which will involve giving their workers the necessary augmentation of their “portfolio of assets”.

Just think of what the earlier trade unions did: when I was at Oxford many decades ago, I recall Ruskin College doing exactly that.

Unionists in Denmark openly admit that their bargaining position is weakened by globalisation. It is the same in the German debate about the Hartz labour reforms. How do you square this type of evidence with the statistical evidence you mention? Another point to make is that globalisation and technological change are not necessarily independent of each other. Mechanisation can also be a response to competitive pressures. That is definitely the case in Denmark’s remaining heavy industry. Is it considered in the econometric studies you mention?
Jacob Gulmann, Copenhagen, Denmark

Jagdish Bhagwati: I think that the bargaining model applies better to the European case than to the US on which I concentrated in my article. But, even then, if wages are too high, I do not see why the exchange rate cannot compensate for it and restore competitiveness, as Paul Krugman has argued. What you argue has more salience when the exchange rate is fixed, as within the EU.

The French and German problem with globalisation is compounded by the fact that globalisation today involves volatility of comparative advantage, that I call kaleidoscopic comparative advantage. If you cannot fire workers when volatility requires that you do, you are unlikely to invest in a country which imposes high cost on firing workers.

So, while French and German firms are excellent, these labour laws are not attuned to the times; they will force firms to invest elsewhere.

When do you think this period of intensified technological change began and where have the benefits of this change manifested themselves? What consolation do low/middle income groups have? Sooner or later, the positive effect of ongoing change on wages should begin to kick in. When will that be? In a hypothetical situation where all technological progress stops, can we expect to see a subsequent rocketing of wages?
Alexander van Kemenade, Beijing

Jagdish Bhagwati: There is a fair amount of evidence produced by economists, like Alan Kruger of Princeton, about change in general and its role in explaining wage stagnation.

I was myself advancing a hypothesis about the nature and incidence of change that would explain the prolonged stagnation. I find that the theoretical argument appeals to a few economists I have tried it out on. I now plan to explore it systematically at the empirical level also; after that, I should be better able to answer your important question

How can you be so convinced that globalisation is playing no role in wage negotiations in the US? Globalisation in the last decade has been an unprecedented economic phenomenon such that it must surely have an impact on the competitive market and influence the going wage.
Gwil William, Birkenhead, England

Jagdish Bhagwati: I cannot disagree with the the distressed tone of your question. I have thought pretty hard about this question over the last decade and looked at the empirical and empirical analyses and come to the conclusion that globalisation does not seem to be a plausible cause of the stagnation of workers’ wages in the rich countries. I am open to persuasion otherwise; one always should be.

Shouldn’t globalisation also rightly claim a role in the stagnation of wages? With technological change, more tasks can be automated, and unskilled workers would be at the losing end. Globalisation exacerbates the impact, as firms have the option of shifting their factories and production of services to other countries. As Professor Bhagwati has also pointed out, foreigners have also been pouring FDI into the US, but such investment pertains to the production of higher value-added goods and services, like, R&D and marketing, which requires a highly skilled and adaptable workforce.
Siang Meng Tan, London, UK

Jagdish Bhagwati: You raise a very interesting question. In my analysis, I was arguing that the evidence strongly suggested that change was putting pressure on wages and that trade with the poor countries was either not a significant factor in adding to that pressure or that, as per my own empirical analysis, it had actually been an independent factor in reducing that pressure. (The latter analysis is in my paper titled Play It Again Sam: Yet Another Look at Trade and Wages.)

You ask instead the question: compare the effect of technical change in a closed economy to that in an open economy: is the adverse effect of any given unskilled-labour-economising change on wages amplified or dampened in the latter case? Pascal Lamy also asked me that question, and many think that an open economy amplifies the adverse effect. I have little doubt that one can model either outcome.

So I would simply caution that the answer is not necessarily amplification of the adverse effect. To illustrate the benign possibility from another area, remember that the incidence of famines was diminished in the Middle Ages when markets were integrated instead of segmented: food moved from the “surplus” areas to the “deficit” areas.

I agree with your analysis. The question is what is the most appropriate policy response to this trend? As Larry Summers wrote, “bad ideas flourish when there are no good ones”. How should one respond to this trend to avoid the bad ideas becoming policy?
JP Landman, Johannesburg, South Africa

Jagdish Bhagwati: I am glad that you find my analysis persuasive. If we are right, then two consequences follow. First, the politicians in the rich countries must draw in the union leaders and others frightened about globalisation into dialogue and try hard to convince them that their fears, while perfectly legitimate, are at best much exaggerated and, at worst, mistaken.

This is not going to be easy. A Russian proverb says that fear has big eyes; I usually add that it also has deaf ears. But there is no alternative to using the megaphone.

President Clinton did not really do this well: he believed in Nafta and in the Uruguay Round; but he never really sat down with John Sweeney of AFL-CIO and other sceptics to confront their fears; hence he had a pretty rough time of it in selling trade.

Second, we need policy and institutional changes to help the workers to cope with the effects of such rapid and deep technical change. I have dealt with many such changes in the Afterword to the new edition of my book, In Defense of Globalization, due out in June/July 2007.

But let me illustrate with one idea. If workers are going to be displaced continually, and this is also going to be something that the volatility of comparative advantage means today (an aspect of globalisation which is most critical today), the unions can no longer meaningfully define security for workers in terms of specific jobs; it has to be now defined in terms of workers themselves.

Giving them education while they work, through weekend classes, through short-term courses at special colleges the way the executives are given short courses in business schools etc are the way this can be, and should be, done. The worker’s portfolio of assets, as it were, needs to be augmented and diversified.

Do you think globalisation can affect wage distribution in a more indirect way; not necessarily by reducing the bargaining power of workers, but by changing the make-up of the industrial and skill base due to the relocation of industry to China, et al? Do you agree that the replacement of industry by the service sector (outside the financial centres of the world) generally results in lower skilled jobs and thus lower wages? Furthermore, is there a danger that with the demise of industry, there will be an associated demise of the managerial and technical middle-class (so affecting wage rates and income distribution)?
Gary Coathup, Evesham, UK

Jagdish Bhagwati: Your comments suggest that there may be mechanisms by which the changed composition of activity, thanks to openness, may result in different future technological effects that affect the future division of labour, the gains from trade and the distribution of income.

Obviously this can happen. Lod Kaldor’s Selective Employment tax was based precisely on such a view, that if the UK had more manufacturing production, it would also have more change. But much econometric analysis disproved such an assumption. I find it hard to believe that anything definitive can be said about these kinds of effects. But I do think you are right to remind me that your kind of connections between globalisation, technical change and distribution of income may be relevant.

Doesn’t this analysis ignore the externalities that a business operates in, for example, its regulatory environment? If workers overseas can be paid less because their government permits their exploitation or the environment’s exploitation cost-free, then globalisation is a race to the bottom. So globalisation is only workable if there is a baseline set of externalities that nations can agree to maintain, for example: the environment and the workplace.
Ted Sabety, New York, US

Jagdish Bhagwati: Environmental neglect, as in China, reflects a lack of democracy: the countervailing NGOs and citizens working through the voting mechanism are simply unavailable. So, the growth rate is exaggerated, I would say by about 3 per cent points, and also the growth is likely to be unsustainable.

But this has nothing to do with China’s competitiveness. A nation’s competitiveness depends on its exchange rate and macroeconomic policies, as my student Paul Krugman pointed out years ago in a scathing attack on commentators like Robert Kuttner, who thought otherwise.

As for a race to the bottom, I demonstrate in Chapter 10 of my book, In Defense of Globalization, there is little evidence of such a race.

Abstracting from the issue of skilled v unskilled labour, profits as a percentage of GDP are at a 50-year high in the US, and multi-decade highs in Germany. To what do you attribute this?
Eric Lonergan, UK

Jagdish Bhagwati: The share of total wages in national income, which used to be pretty constant for decades, has certainly recorded a small decline, even after we consider the fact that, when fringe benefits are taken into account, the deadline is definitely less.

I have no better explanation of this phenomenon on than what I produced in my FT article: rapid, continuous and deeply unskilled-labour-saving technical change seems to me to be the most compelling explanation.

But, if I am correct, that ought to be self-correcting in the long run as the productivity-enhancing effects of the technical change finally pull up the overall demand for labour in a growing economy.

I agree that wage stagnation is caused by the technological change that seriously affects work time and working costs. My question is: why can’t US wages benefit from such technological change? Your native country, India, is taking advantage of technology to create wealth and high levels of GDP. Lets not forget that the changes in technology are, after all, responsible for globalisation.
Emmanuel Kaliontzakis, MBA, New York, US

Jagdish Bhagwati: India, and indeed China and other rapidly developing countries such as the East Asian economies, have profited from the ability to cheaply absorb technology, as embodied in equipment from technologically innovative rich countries. This was made possible as they began to generate export earnings, which increased the attractiveness of investment, matched by savings rates that far exceed those in the rich countries.

So, whereas we celebrate as a miracle a sustained 4 per cent growth rate in the rich countries, this would be considered a failure in developing countries where an economic miracles usually implies two-digit growth rates! As long as India and China were inward-looking on trade and inward investment, they recorded low growth rates.

After the early 1980s, they changed their policies to profiting from closer integration into the world economy and they have been recording high growth rates, with considerable impact on their poverty.

If it is technology that is causing US middle class wages to stagnate, please explain why companies are moving labour intensive jobs, such as call centres and engineering, to low wage countries such as India. As an engineer whose company is cutting wages here in the US while greatly expanding in Asia, I find your technology argument a bit of a smoke screen.
Michael Johnson, Minneapolis, US

Jagdish Bhagwati: Of course, the possibility of online trade in services is augmenting trade possibilities as previously non-traded services are becoming tradeable and also internationally traded. But at least three things need to be noted.

First, it is the rich countries that wanted such trade to be freed up because we expected, quite correctly, that we would export huge amounts of high-value, skilled services such as lawyers, architects, accountants, management consultants etc and import some of the relatively low-skilled activities like call centres. This is exactly what has been happening. The balance of trade is hugely in our favour.

Second, the fear that we would wind up importing even most of skilled online services is unfounded. There was much alarm on this when a Mass General Hospital radiologist outsourced reading x-rays digitally to India. But the facts show that not one radiologist has lost his job or found his earnings reduced. The bulk of radiological reading of x-rays is now sent to India and Australia where the time zone difference enables the x-rays to be read there when our radiologists are sleeping or enjoying their long weekends.

Besides, the strict enforcement of required credentials by professional bodies like the American Economic Association has also limited what can be transacted online in such professional services.

Third, the Indian experience shows that even simpler jobs like call-answer services are subject to rapidly rising costs; wages have been rising quickly in recent years. Only a small fraction of India’s huge population that can enter college does so. Of this, only a further small fraction studies English. Of that, a further fraction can speak English. Of that, again, only a small fraction can speak English in a way which can be understood in US and UK.

So, India’s size (and China’s size) which spread the fear of their taking over everything, skilled and unskilled, is not justified.


Jagdish Bhagwati: Technology, not globalisation, drives wages down

In depth: America’s anxious middle

Copyright The Financial Times Limited 2017. All rights reserved. You may share using our article tools. Please don't copy articles from and redistribute by email or post to the web.