Can Innovation beat Inequality? Only if direct measures to address inequality are met by investments in long-term productivity growth will economies become more equitable.

Inequality is the “defining challenge” of our time, according to Barack Obama. InCapital in the 21st Century, Thomas Piketty captured the world’s attention by explaining why the challenge is so big – and probably getting bigger. Whilst his economic case has been extensively reviewed and mostly endorsed, albeit with questions raised by Chris Giles on his data methodology, the same cannot be said of the book’s underlying assumptions. These are: first, economies will stagnate; second, capital will top talent; third, capitalists don’t innovate. In short: Piketty is deeply sceptical about innovation as a growth driver and equalizing force in this century. Has innovation lost its punch?

Economies are not doomed to stagnation

As the economist Paul Krugman writes, “Productivity is not everything, but in the long run it is almost everything”. If productivity growth is weak, economic growth is going to be weak, too. According to Piketty this is what we are seeing. Pointing to historical tax records he argues that no country at the technological frontier has ever grown at more than 1.5 percent for a lengthy period of time. Overall, growth accelerated in the Industrial Revolution, peaked in the twentieth century, and is now returning to “much lower levels”.

Piketty bases his prediction on estimates which see population growth at near zero in the second half of this century. He also cites the Northwestern University economist Robert Gordon who argues that recent waves of innovation had less impact on productivity than others: the invention of electricity, the internal combustion engine and running water between 1870 and 1900 had longer and more radical effects on productivity than the computer and internet revolution, which started in the 1960s and peaked in the late 1990s. As Tyler Cowen put it, “we picked all the low-hanging fruits”.

In consequence, Piketty suggests a long-term scenario of 1.2 percent growth in advanced economies. Even this rate is in his view optimistic, requiring “significant technological progress, especially in the area of clean energy”. Other leading economists including Larry Summers who coined the term secular stagnation, Tyler Cowen who writes about the great stagnation, or Mohamed El-Erian who introduced the new normal present equally lacklustre outlooks.

As prominent as they are, these views shouldn’t be accepted as orthodoxy. In The Second Machine Age, Eric Brynjolfsson and Andrew McAfee argue that information technology, like electricity, is a general purpose technology. It spreads through most industries and speeds up the march of economic progress – but only with time: electricity was invented in the outgoing 19th century but big productivity gains kicked in decades later. With the advent of real artificial intelligence and ubiquitous connectivity information technology may now be at a similar point. The digital revolution is not over, it is just starting.

What makes Eric Brynjolfsson and Andrew McAfee so sure is a fundamentally different concept of innovation: along with new growth theorists like Brian Arthur orPaul Romer, they reject the idea of innovation waves that get “used up”. To them innovation is combinatorial: “it is not coming up with something big and new, but instead combining things that already exist”. We are not emptying a finite pool of ideas. We are filling an indefinite one. But to identify the most valuable ideas we need many pairs of eyes and a lot of computing power. And that is exactly what the two breakthroughs above may give us.

Capital needs talent

Piketty’s second prediction is a capital-labour elasticity of substitution which is greater than one. In normal language that means in this century people will be able to accumulate enormous amounts of capital without reducing returns to zero. Even if individual returns decrease at some point, “the most likely outcome is that the volume effect will outweigh the price effect”, so that the amount of capital relative to national income continues to grow. This is great news if I earn my income from capital. If I earn my income from labour it means my income may be at risk.

This is Piketty’s most crucial idea, as it means that over a long period of time more and more income will be accumulated at the top. It also breaks with the assumption that human capital becomes more important as economies become more sophisticated. Clearly skill levels have increased markedly over the past two centuries, but the stock of industrial, financial, and real estate capital did as well. We have not “magically gone from a civilization based on capital, inheritance, and kinship to one based on human capital and talent”. The shift from “capitalism” to “talentism” is not happening.

But what if Brynjolfsson and McAfee are right and progress is not dead? Most economists argued in the past this would serve us all: whilst individual workers may suffer, new and usually better uses are found for the workforce as a whole. John Clark already wrote in 1915: “the well-being of workers requires that progress should go on, and it cannot do without temporary displacement of laborers”. History largely supports this claim: according to the McKinsey Global Institute, all but one 10-year rolling pe­riod since 1929 has recorded increases in both US productivity and employment.

Does that mean income inequality has gone down or at least wage earners are better off in absolute terms? Not necessarily. In the US, the gap between average income and median hourly wages has grown since the 1970s. James Surowieckioffers an explanation why: in 1960, the country’s biggest employer, General Motors, was also the best-paying. Today, the country’s biggest employers are retailers and fast-food chains, almost all thriving on low pay and low prices. As technology improves, new jobs are generated for the workforce – but not always better ones.

The reason is an inherent skill-bias in technological change: the digital revolution dramatically lowered the demand for workers who carry out routine tasks, regardless of whether they are cognitive or manual. This leads to what Daron Acemoglu and David Autor call job polarization: a collapse in demand for middle-income jobs, while non-routine cognitive jobs such as financial analysis and non-routine manual jobs – like retail and fast-food –hold up relatively well. Those whose skills are substitutedby smart machines lose out. Those whose skills are augmented by smart machines, allowing them to achieve better results at lower cost, win. And those who own these machines win, too.

To avoid this trap, Brynjolfsson and McAfee suggest we must learn to complement rather than compete with machines. “The value of human inputs will grow, not shrink”, they predict, and with the right skills we will achieve things that could never have been accomplished by unaugmented humans or machines alone. We may not move from capitalism to talentism but without investing in talent, we will end up in a race we cannot win individually, and which slows societies down collectively.

A starting point must be basic education: research shows that in the US, children between 4 and 5 from the poorest fifth of homes are almost two years behind those from the richest, and students from the highest social-class are twice as likely to go to college as those from the lowest. Technological progress, with the skills it demands, will only be an equalizing force if quality education is accessible to all.

Capital drives innovation – sometimes

Piketty’s last assumption is the least explicit and the most problematic: capitalists seek profit, not progress. They claim larger slices of the cake yet they don’t make the cake itself any bigger. This view flows from the low growth assumption along with the assumption that labour can be replaced by capital at low cost. It also is a consequence of Piketty seeing technological progress as exogenous, an external force, opposed to Brynjolfsson and McAfee who see it as endogenous. Piketty dedicates just half a page of his 700-page tome to technological change, under the title The Caprices of Technology.

This doesn’t mean there is no innovation in Piketty’s world; it means innovation does not create growth. Clayton Christensen explains this in a more nuanced framework. The Harvard Business School professor distinguishes two types of innovation:Empowering innovations yield new products and services, create jobs to build, sell, distribute and service them, and use capital to expand capacity and inventory.Efficiency innovations reduce the cost of making and distributing existing products and services, almost always reduce the net number of jobs, and liberate capital on balance sheets.

In a healthy economy, Christensen argues, the two innovations operate in tandem. Empowering innovations are essential for growth because they create new consumption. Efficiency innovations are critical as they liberate capital for empowering innovations. What troubles the US economy today is that capital is being reinvested into yet more efficiency innovations.

Christensen blames this on the efficiency focused finance doctrines which emerged after the Second World War, when capital was scarce. “Those doctrines were appropriate to the circumstances when first articulated”, he explains, “but we’ve never taught our apprentices that when capital is abundant … the same rules are the wrong rules”. “Continuing to measure the efficiency of capital”, he continues, “prevents investment in empowering innovations that would create the new growth we need”.

Others go further, speaking of vested interests and political influences. Luigi Zingales warns that lobbying often promotes the interests of existing businesses rather than market competition. Acemoglu and Robinson describe how economic inequality translates into political inequality and from there back into even more economic inequality as the rich stack the cards in their favour. Brynjolfsson and McAfee agree such a scenario is plausible and worry that growing income gaps could hamper progress in the long term.

These views turn Piketty’s argument on its head: the problem is not capital accumulation but capital use. Piketty argues that capitalists focus on the slice of the cake because the cake doesn’t expand – Christensen and others argue the cake doesn’t get bigger because capitalists just focus on their slice.

This is more than semantics. The crucial difference lies in the policy recommendation: Piketty wants a global wealth tax to bring up effective demand: letting the stock of capital diminish until the capital return equals growth would allow for a permanently higher level of per capita consumption, and thus for a better social state.

Christiansen wants to focus on innovation: “it’s true that some of the richest Americans have been making money with money – investing in efficiency innovations rather than investing to create jobs”, he argues, but if we just redistribute wealth, most of it will just replace consumption with consumption. Instead, “we must give the wealthiest an incentive to invest for the long term”.

In the end, both measures are critical. Piketty started a debate on wealth taxation which focuses too much on the numerator and too little on the denominator of the capital-income ratio: technological progress. Christensen pays too little attention to the social implications of skill-biased technological change which are often reproduced across generations. If one truly wishes to found a more just and rational social order, investing in “empowering innovation” won’t be enough.

Since the mid-1970s it has been widely accepted that moderate inequality is the price for a dynamic economy. Now a point has come where a dynamic economy is threatened by rising inequality. But putting all hopes into demand-side adjustments misses the point. Only if direct measures to address inequality are met by investments in long-term productivity growth will economies become more equitable.

Published: Forum Blog, June 4, 2014

Image: Workers assemble Android-based tablets from imported components at the Surtab factory in the Sonapi Industrial Park of Port-au-Prince March 11, 2014. REUTERS/Marie Arago

About Sebastian Buckup