Innovation and Prosperity in an Age of Transition

A foreword to the New Thinking on Innovation series

Published: November 21, 2017

Author: Michael Spence


he post-World War II period of unprecedented growth and prosperity has brought home lessons, many of which are relevant today and likely will be in the future. Several things stand out after studying the growth patterns and challenges in a wide range of developed and developing countries for the past 15 years.

The global economy has been a key enabler of growth, war recovery and, in particular, progress in developing countries. No one foresaw the scope of the potential for poverty and reduction and prosperity that actually occurred, when the architecture was put in place. 

But, in the last few decades, that architecture has had difficulty keeping up with the realities.  Previously small and relatively poor countries are now major players. Digital technology has vastly expanded the ability to manage complexity in global supply chains at low cost. Trade data, built on the assumption that something is produced in country A and consumed in country B, has very little to do with the reality in many sectors. And, perhaps most importantly, trade in services (broadly defined to include intangibles and intellectual property [IP]), initially a tiny fraction of global trade, may become the dominant part in the future.

While there are common ingredients across countries in successful growth recipes, there are also striking differences. Local conditions, history, governance and path dependence make each case to some extent idiosyncratic. Unfortunately, this valid proposition is sometimes taken as a rationale for not learning from the vast array of experience of other countries.

Digital technology has vastly expanded the ability to manage complexity in global supply chains at low cost.

A subset of countries endowed with substantial natural resource wealth have struggled — to varying degrees of success — with the complex challenge of exploiting that wealth for the benefit of their citizens while creating a broad and diverse range of employment opportunities.  It is awfully easy to create a structure in which the tradable sector is dominated by natural resources and a disproportionate share of employment lies in the non-tradable service side of the economy. 

Economic and social progress and well-being have been most pronounced in countries that viewed people as their most important investable resource. In the end, they are the source of innovation, creativity and dynamism. Of course complementary tangible and intangible assets are required to exploit this potential. 

While trade in goods — and more recently and increasingly in services — has played a crucial role in postwar growth and prosperity and an extended period of productivity and income growth has been fuelled, in part, by specialization, there are reasons, related to convergence and digital technology, to suspect that while this engine of productivity growth will not disappear, it may not be as powerful in the future as it has been in the past 70 years. And, in any case, its contours and shape will likely change dramatically.

Downward trends in productivity growth are widespread across advanced economies, yet the press is full of daily stories of breakthroughs in various fields of science and technology. This juxtaposition presents a puzzle, at least in the minds of most concerned citizens. How can you have robots that can see, have fine motor skills and can assemble electronics, or machines that learn to do things that humans do, but do not know exactly how we do them (all breakthroughs of the past 10 years), and not see effects in productivity? 

Compared to technologies of the past, modern technologies may not be as powerful in driving productivity . (Photo: Everett Historical /

There are a number of factors, not mutually exclusive, that may help frame the issue. 

One is time lags. It took more than 40 years in the “computer” age before anything plausible showed up in productivity data, and that was only after the internet became generally accessible in the 1990s. In the more distant past, science and mathematics was advancing nicely for several centuries before the Industrial Revolution in Britain. 

A second perspective is what might be called the Robert Gordon view.1 Modern technologies may just not be as powerful in driving productivity as those of the past. It is hard to think of an innovation that has the impact of electricity. It is worth taking seriously because you do not want to fight a battle you cannot win. My view is that at this stage, it is impossible to know whether digital technologies (given a long enough run measured in decades) will have impacts of comparable magnitude to those of the past. But there is nothing in economic theory or history that suggests that innovation produces a steady non-cyclical stream of productivity-enhancing technologies. 

In addition, there is some evidence that productivity in research and development is declining.  An interesting recent paper by Bloom et al. (2017) pursues this line of enquiry. For a brief glimpse, Moore’s law2 is still operating in the semi-conductor area, with enormous paradigm-changing effects, artificial intelligence being an example as it is reliant on huge databases, fast networks and large amounts of computing power. But the effort measured in engineering talent required to achieve each successive doubling is also increasing exponentially.

A third possibility, one that I have suggested, is that a substantial fraction of the creative talent in developed economies (where incomes are already high) has gone and continues to go into innovations that are not targeted primarily at enhancing productivity and income growth. By far the largest federal research budget in the United States is the National Institutes of Health (US$32 billion), whose target is advances in biomedical science. It funds the research of tens of thousands of the top research talent in the country. There may be secondary spillover effects on economic performance as conventionally measured, but it certainly is not the primary goal.  There is nothing wrong with this. Society does not have a singular goal (something like income or wealth). Through market choices and government/collective choice mechanisms, we may have been busy doing something else. And there are many other examples, such as social media.  Value creation is not the same thing as rising incomes, although they are not completely orthogonal.

Value creation is not the same thing as rising incomes, although they are not completely orthogonal.

In reviewing performance and opportunities going forward, it is important to devote some effort to clarifying economic and social goals and ways of measuring them. This is not at all to dismiss concerns about downward productivity trends and their effect on growth in incomes, but rather to put it in perspective. If, as seems likely, achieving inclusive growth patterns, improving health and longevity, enhancing social interaction or evolving to more sustainable growth patterns are societal priorities, it is reasonable to expect that both talent and innovation will show up in multiple places in pursuit of multiple goals, and there may be some diminution in GDP growth as a result.

I do not have any doubt that if Canada somehow found a way to devote most of its vast creative talent to productivity enhancement in the conventional sense, it would yield impressive results. But not necessarily better ones. This dilemma should not proscribe a quest to understand productivity trends and their link to other outcomes, such as income and well-being in Canada, and, ideally, lead to strategies to improve productivity and welfare outcomes.

That said, it is useful to search for ways to unlock underutilized resources that can help expand the scope of innovation. Key among these is the policy environment, at home and abroad, within which innovation occurs. The central thesis of the essays collected in this report is that in Canada this environment can be improved.

There is, however, a major set of challenges, ones that have been at least two or three decades in the making. Growth patterns in developed economies have been characterized by job and income polarization. This pattern has been documented across most of the Organisation for Economic Co-operation and Development countries; the patterns are not identical country to country. Roughly middle-income jobs have been declining and upper- and lower-income jobs have been increasing. As a result, in part, the income distributions have flattened out, declining in the middle and rising at both ends.

Although it is not the whole story, globalization (historically) and digital technology (both historically and prospectively) are correctly viewed as major contributing factors. Countries mitigated these forces and the trends to varying degrees via income redistribution and the provision of very high quality public services, including health, education and social security.  Canada and the Nordic countries have done it better than, say, the United States and the United Kingdom. This shows in the income distributions, but also in differences in essential public services and social security systems.

The current pattern of rapid structural evolution is being called the fourth industrial revolution. (Photo: wi6995 /

But even these admirable mitigating policies and structures do not completely solve the problem. Perhaps the best way to put it is to ask the question, what is the future of work for various segments of the population — in the face of increasingly powerful digital technologies, in the automation and disintermediation of blue- and white-collar jobs, in additive manufacturing and in machine learning and artificial (terrible term) intelligence? It appears that our economies have been and still are in a pattern of rapid structural evolution. It is being called the fourth industrial revolution. I am not sure what the second and third are.

A subset of routine jobs is being removed from the economy, while other jobs are being created.  Digital technologies are expanding the scope of what is “routine” — formerly “codifiable,” and increasingly “machine learnable.” People with jobs in this expanding “routine” category are being sideswiped by the economic and structural trends. Repositioning their skills and human capital to thrive in the “new” economy is an investment that may or may not have a positive return, and the private and social returns to these investments may be quite different.

Huge rents accrue to those who create and successfully commercialize the IP that drives the fourth industrial revolution, thus creating a nexus of issues around wealth accumulation and its distribution, and the quality of work.

These are major challenges. Innovative societies are going to meet them more effectively.  Indeed, institutional and social innovation is going to be an essential ingredient in effective adaptation, a core implication of the many recommendations in this report.

Let me offer a few thoughts on trade, related again to digital technology. Globalization can be understood in terms of cross-border flows of goods and services, capital, information, technology and ideas, and people. Of these, the least mobile are people. In the postwar period, global supply chains have been organized with increasing efficiency and granularity, quite rationally around utilizing the least mobile resource, people or human capital. In some areas, such as manufacturing, this may change. Digital technologies are likely to reduce the labour content of many manufacturing processes enough that labour availability and cost will no longer be the decisive factor. That is in process now. As a result, manufacturing becomes mobile, and labour is no longer the decisive factor. What will happen as a result is that manufacturing will move either toward centres of innovation and excellence in the respective technologies and industries or toward the final market. Probably we will see both. The latter implies relocalization. But it is easy to see a likely trend in which centres of innovation become even more decisive factors in attracting economic activity and jobs.

Trade, except for trade in natural resources, will continue to be organized around finding and utilizing valuable pools of human capital, which is still relatively immobile, but it will be trade in services, in ideas, in IP and in brands and brand images. Some of those ideas and IP will be embedded in physical products. But what are being shipped around are really the embedded results of innovation.

These are major changes. No one can see the future in clear precise outline, especially out 10 years or more. Generally, I think there is a tendency to think that what is possible to imagine or foresee will happen faster than it actually does. But it seems reasonable to operate on the premise that societies that will navigate successfully through a period of profound structural change and thrive in terms of growth, quality of work and adaptability to change are those in which creativity is fully unleashed and innovation is deeply embedded in the culture.

1 Gordon (2016) situates recent developments in information and communications technology in the larger cycle of economic progress and finds they do not match up to the impacts of five previous great inventions — electricity, urban sanitation, chemicals and pharmaceuticals, the internal combustion engine, and modern communication.

2 The number of transistors in a dense integrated circuit doubles about every two years.

Works Cited

Bloom, Nicholas, Charles I. Jones, John Van Reenen and Michael Webb. 2017. “Are Ideas Getting Harder to Find?” NBER Working Paper 23782, September.

Gordon, Robert. 2016. The Rise and Fall of American Growth: The U.S. Standard of Living since the Civil War. Princeton, NJ: Princeton University Press.

The opinions expressed in this article/multimedia are those of the author(s) and do not necessarily reflect the views of CIGI or its Board of Directors.

In the Series