Clearer Rules and Norms about the Ownership of Intangibles

A time-honored way to encourage investment in intangibles is to make rules to allow people and businesses to own them. The classic examples are patents and copyrights; indeed, such rules even get their own clause in the US Constitution. This mitigates the problem of spillovers by making it illegal for anyone but the rights-holder to use the asset without permission.

One option for a government committed to reducing the spillovers of intangible assets is to tighten and extend these laws. It could allow broader, longer-lasting patents, IP rights over intangibles like design, and a tolerance of noncompete clauses (which help firms reduce the spillovers of training, by making it harder for trained workers to leave for other firms). At the margin, it could also make it cheaper to obtain IP protection.

All of this would certainly increase companies’ ability to appropriate the intangible investment they make. But there would be a high price to pay. Strong IP rights reduce the opportunities for other firms to realize the synergies between intangibles; so while it may increase the incentive to invest, it reduces the productivity gain from investments. As Bronwyn Hall has pointed out, interlocking suites of interlocking patents sometime act as a barrier to competition (Hall, Helmers, and Graevenitz 2015). In some cases, the potential productivity gain from synergies is sufficiently strong that there is a good case for weakening rather than strengthening IP rights—such as software patents, or telecoms, where there are a lot of interlocking patents that make it too hard for an innovating firm to bargain with all the relevant patent-holders. A further danger of strengthening IP laws is that it’s possible to do so in uneven and partial ways that favor incumbent rights-holders and patent trolls (both of which groups often devote significant resources to lobbying), while doing little to encourage new intangible investment.

There is, however, a good case for clearer IP rights. Well-run patent offices are equipped to reject vague patents that create uncertainty. Clear legal processes give IP owners confidence that their rights do what they think they do and discourage vexatious lawsuits and the kind of legal shopping-around that has brought so many patent cases to the notoriously troll-friendly Eastern District of Texas court.

There is more to rules around intangibles than intellectual property law itself: markets and norms matter too.

Let’s consider markets first. As we saw in chapter 4, part of the reason tangible assets are less likely to be sunk costs is that they are tradable and people know what they are worth. Valuing a patent or a copyright will probably never be as easy as valuing a used van, but establishing markets like the Digital Copyright Exchange proposed in the UK by Ian Hargreaves in 2011 may help this process; patent pools, in which firms coinvest in research and agree to share the resulting rights, have been used in a variety of industries since the early twentieth century.

Given sufficient advances in technology and infrastructure, these kinds of markets and institutions need not be limited to major intangible assets like patents or copyrights. They may also be applicable to the tiny elements of user-generated data that collectively make up the vastly valuable databases and networks of firms like Google and Facebook. Jaron Lanier, the philosopher and computer scientist, called for a system that would allow the creators of user-generated content—that is, you and me every time we interact online or often offline—to charge a very small fee for the use of our data. Establishing these kinds of exchanges is a major undertaking, requiring significant coordination between rights-holders, content platforms, collection agencies, and governments. But it may be worth the effort: efficient markets and platforms for exchanging IP will be economically valuable in an intangible economy.

Because intangibles often have valuable synergies, they rely on norms, rules, and standards about how to bring them together. Some of these norms are technical, like technological protocols that allow different pieces of software to interact; some of them are professional norms, like the phasing of rounds of venture capital; some of them are regulatory, like rules about what data websites can gather and what they can do with it, or the regulations that govern relationships between firms (for example, between platforms like YouTube and the owners of copyrighted video content). Most of them are underpinned by social consensus about how things should work (the belief among developers that software should be compatible, not proprietary and closed off, for example, or the beliefs about the balance between personal privacy and the rights of businesses that inform data protection law).

To maximize effective intangible investment, an economy needs carefully thought-out rules, based on informed and somewhat stable social consensus. This, in turn, requires investment (developing standards is not free) and social capital. Divided, fractious, or incurious societies, those with insufficient social capital, are more likely to have unstable and ever-changing views on what the rules on things like privacy should be trying to achieve—and when it comes to encouraging investment, stability may be more important than the precise norm that is adopted. Establishing and enforcing these norms is costly: they require appropriately funded patent offices and regulators, and they require government to expend political capital to regulate fairly (rather than simply implementing the will of the most expensive lobbyists).

Helping Ideas Combine: Maximizing the Benefits of Synergies

Good public policy should be just as assiduous about creating the conditions for knowledge to spread, mingle, and fructify as it is about creating property rights for those who invest in intangibles.

Despite frequent predictions that the Internet will lead to the death of distance, for the time being the spillovers between intangibles happen in specific places where people congregate, especially in cities. This makes good urban planning and land-use policies extremely important. There is of course a vast literature on what constitutes good policy for cities, but in the context of intangibles, there are two important principles.

On the one hand, city rules should not make it hard to build new workplaces and housing. Cities should have freedom to grow to make the most of the ever-increasing synergies arising from intangibles.

On the other hand, cities need to be connected and livable. Synergies are more likely to be realized if people meet each other and interact than if urban life is atomized and siloed. Getting this right involves striking a balance; it takes a combination of Jane Jacobs–style liberalism, tolerating messy and diverse areas rather than building multilane highways through them, and of some benign planning, providing enough infrastructure for people to get around and places for them to meet. The kinds of cities that attract what Richard Florida called the “creative class,” or the “innovation districts” that Bruce Katz observed emerging across the United States, involve a mixture of judicious planning and organic growth.

There are inevitably tensions in this kind of policy. In intangible-intensive cities like New York and London, liberalizing planning rules to allow more housing to be built is criticized for causing the destruction of important public spaces and cultural venues where people congregate. Good development, especially in an age of intangibles, involves providing both for the basics of housing and transport and for conviviality.

A skeptical reader might at this point ask what is new about the need for good town planning and land-use rules. After all, liberalizing planning rules is one of those policies that academic economists have been demanding for decades, but that is generally thwarted by other factors, from the desire to reduce sprawl and unsightly development to existing homeowners’ defensiveness about the value of their properties. It is true that this is a well-trodden issue. But the significance of the rise of intangibles is that with each year that goes by, the economic cost of bad urban policy, of protecting greenbelts, of restricting building heights, or of protecting lot sizes, will go up: the more reliant the economy is on intangibles, the more the economy will lose out by restricting the opportunities for intangibles to cross-fertilize.

To get a sense of the changing costs of planning restrictions consider the geographer Christian Hilber’s (2016) dramatic example from London. If you go to King Henry VIII’s Mound, Richmond Park, ten miles southwest of St. Paul’s Cathedral in London, you will find an avenue of trees, planted in the early 1700s, that creates a “keyhole” view of the dome of St. Paul’s. That view has been the same since 1710. How has it survived so long? It is protected by planning regulations, namely the London View Management Framework.1 These regulations prohibit tall buildings in the sight line between Richmond Park and St. Paul’s. They also don’t allow tall buildings behind St. Paul’s, which the planners have decided would constitute an unacceptable backdrop to that view. As the LVMF (2012, para. 175) says: “In determining applications [for new buildings], it is essential that development in the background of the view is subordinate to the Cathedral and that the clear sky background profile of the upper part of the dome remains.” As Hilbers says, “while this view is certainly enjoyable for those living nearby or for hikers, it arguably imposes an astronomic and ever growing economic ‘opportunity cost.’ . . . The protected vista, through limiting supply, raises housing costs of all Londoners and adversely affects the capital’s productivity.”

But creating the infrastructure for spillovers is not just about physical space. Currently, the most effective collaborations happen face-to-face, despite the dizzying variety of digital technologies for socializing and collaboration, from Skype and e-mail to Facebook and Slack. But just because the widely predicted death of distance has not come to pass, doesn’t mean that it will never do so. It seems very likely that at some point people will discover better ways to interact meaningfully with one another at a distance using IT, as new applications develop and the workforce becomes populated by people who grew up with online social lives and hobbies.

The question of how people use technology to boost what some call “collective intelligence” has a long history: it lies behind the famous “mother of all demos,” the 1968 presentation in which Douglas Engelbart demonstrated the world’s first instances of videoconferencing, dynamic file linking, revision control, and electronic collaboration. Collective intelligence is intimately entwined with the development of Internet phenomena like Wikipedia, and it continues to evolve in the form of platforms like Slack and GitHub.

An economy that can develop technologies and ways of working that replicate the social power of face-to-face interaction at a distance will be transformed, particularly when it comes to land use. Freedom from the tyranny of ever more expensive city center property is a big economic win. So while the death of distance may be a long shot, the kind of thing that is promised long before it is delivered, the economic rewards are very large.

There are a few things governments could do to help this along the way. The government of Foo, our intangible-friendly state, could follow the example of DARPA in the 1960s and 1970s and fund experimental development into the use of technologies to foster collective intelligence and effective collaboration. (Indeed, the European Union currently funds several research programs into this under the Horizon 2020 program; in the United States, programs by not-for-profit organizations such as the MacArthur Foundation Research Network on Opening Governance play a similar role.) More ambitiously, Foo might experiment with tools for distance working and collaboration in its own business. This could involve making government departments lead users of distance working tools; it could also involve using digital collaboration tools to run public consultations, democratic deliberations, and other engagement exercises that only governments tend to do.

A Financial Architecture for Intangible Investment

We saw in chapter 8 that financial markets are designed to meet the needs of businesses that invest in tangible assets, not in intangible ones. Changing how financial markets work is not easy, but most governments already do it to some extent, whether by offering government loan guarantees, tax breaks for certain types of finance like venture capital, or, most significantly, by treating debt interest but not the cost of equity financing as a tax-deductible expense. What would a country do differently if it wanted to create the conditions for intangible-intensive businesses to get the capital they need to thrive?

First of all, governments should encourage new forms of debt finance that make it easier for companies to borrow against intellectual property—intangible assets to which property rights can be attached. Government cannot usually make financial innovation happen, but it can make it easier. As we saw, the governments of Singapore and Malaysia have put in place schemes to encourage intangible-backed loans, partly by subsidizing the loans and partly by instructing government bodies responsible for IP (such as patent offices) to work with banks to reduce legal and technical barriers.

In the longer term, the government should create the conditions for a shift from debt to equity financing. As we have seen, it is often hard to raise debt financing against intangible investments because they are sunk: the bank can take a charge or lien on your fleet of vans or your office in the event you can’t repay your loans, but it’s harder to do the same with a proprietary process or a brand. Because companies can claim tax relief on interest payments but not on the cost of equity, debt is cheaper than equity for any given level of risk. As intangible investment becomes more important, this distortion will hold back investment more and more.

Creating a tax credit for equity financing—that is, reducing a company’s tax liability by an amount reflecting the cost of equity—is one way to correct this; another is to tax debt interest payments but lower overall tax rates to compensate. It is a proposal with a respectable pedigree: a credit exists in Belgium, and various versions of the credit were analyzed and recommended by Nobel laureate James Mirrlees in his canonical 2011 review of the UK tax system (Mirrlees et al. 2011). Governments should not be under any illusions about the difficulty of this kind of plan: it is the equivalent of open-heart surgery on a central part of the corporate tax system, and it will be opposed by any number of vested interests whose business models depend on cheap debt. And such a plan would require the emergence of new institutions for providing equity financing to small and medium-size businesses, which would also take a lot of time to emerge. But as the importance of intangibles grows, the rewards for making the change, in terms of increased investment and productivity, will grow ever greater.

We would expect to see public equity investment dominated more by institutions, some of whom would commit to taking longer-term stakes in intangible-rich companies, enabling greater investment. Government has a few roles that it can play here. First, it can remove regulations that discourage blockholding (these include disclosure requirements, rules on what information companies may provide blockholders, and rules about which shareholders may vote with borrowed stock). Second, it can reexamine standards of financial accounting to identify better ways of reflecting intangible investments (following the lead of the designers of California’s planned Long-Term Stock Exchange or of accounting scholar Baruch Lev’s reform agenda set out in The End of Accounting).

There may also be a different strategy available to those governments fortunate enough to run sovereign wealth funds or large, endowed state pension funds. As we have seen, the largest institutional investors may be able to invest broadly across an ecosystem, knowing that they can benefit from spillovers of intangible investments even if an individual company they have backed does not. These larger national funds could be deployed to invest in particular ecosystems (in the way that Fidelity is reported to have invested across Elon Musk’s intangible-intensive business empire).

Alongside these regulatory changes, we might cautiously hope for a cultural shift among the managers of large companies and institutional investors. The UK’s Purposeful Company project (Big Innovation Centre 2017) and the international initiative Focusing Capital on the Long Term have both argued for managers and large shareholders to be more willing to make long-term investments, particularly in intangible investments like R&D and organizational and human capital. Skeptics might argue that fine words and good intentions are not enough to change the behavior of big businesses. But in combination with other policy measures, they may prove effective; it certainly seems that the behavior of those companies that are willing to make large, long-term investments in intangibles is at least partly a matter of culture.

While we might expect to see venture capital develop further in an increasingly intangible economy, it is not clear that governments can or should do much more to promote it than they already do. As Josh Lerner showed in The Boulevard of Broken Dreams (2012), once tax breaks or subsidies for venture capital get beyond a certain level, they tend to encourage dumb investment (since the tax gain on its own is enough for the investors to profit); since the entire point of venture capital is smart investment, very large tax breaks are self-defeating. For a country to grow its venture capital sector, time and favorable framework conditions are more important than additional subsidies.

Solving the Intangible Investment Gap

So far we have suggested that governments should mitigate the underinvestment problem posed by an intangible economy in three ways: strengthening ownership claims to intangibles, where possible; creating the conditions where businesses can make the most of intangible spillovers and synergies; and encouraging financial reform so that companies face fewer incentives to underinvest. These are all worth doing, but it seems unlikely that they will entirely solve the underinvestment problem. After all, the underlying incentives for companies to hold back from investments with high spillovers will still remain. And to the extent that intangibles look set to become more and more important to the economy each year, the underinvestment problem will get worse and worse.

There are two actors in the economy that have an interest in making intangible investments, despite the risk of the benefits spilling over.

The first is the small number of large, dominant firms that seem to have an ability to not only gain from their own investments but also to appropriate the benefits of other firms’ investments. This is one interpretation of what firms like Google or Facebook are doing when they back moonshot-style R&D programs, or when they spend liberally on supporting “start-up ecosystems” in major cities—if you are a big and diverse enough firm, these kinds of investments may be in your enlightened self-interest.

The second is the government and other public interest bodies such as large not-for-profit foundations, both of which are meant to take a wider perspective.

It is theoretically possible that large, dominant firms might over time invest more and more in intangibles, making up the shortfall from the rest of the business sector. This would effectively be a return to the days of large, Bell Labs-style investments. It may even be that some of the same underlying dynamics encourage them: just as the public-good research of Bell Labs was in some ways a quid pro quo for the US government’s willingness to tolerate AT&T’s telecoms monopoly, perhaps big tech companies of the future that enjoy effective monopolies due to networks will be encouraged to invest in R&D and other intangibles as part of their license to operate. But, on the whole, this seems unlikely: the relationship between government and business in most developed countries has changed so much since the 1960s and 1970s that it is hard to imagine this kind of corporatism being re-created on a significant scale, and it is hard to imagine that it would not have other negative effects that would lower productivity.2 (A smaller scale version of this might emerge, however, if more companies follow the pattern of Microsoft, which generated great personal wealth for its founders, who then went on to fund intangible investments for the public good. Examples are Bill Gates, whose Gates Foundation funds research into tropical diseases, and Nathan Myhrvold, who backs nuclear and geoengineering research. But this seems unlikely to make up the difference.)

This leaves government as an investor of last resort. It is hard to escape the conclusion that if intangible investment is harder for businesses to fund, and if it is becoming more important to the economy, then unless we are prepared to see a shortfall in investment, the role of government as an investor will have to grow.

This ought not to be a totally alien concept in developed countries, whose governments already make significant investments in intangibles that firms use, especially in the form of public R&D and subsidized training. In the UK, about a third of all R&D, and a much greater proportion of early-stage R&D, is funded by the government. But anyone with a passing familiarity with public sector software projects will know that not all government-funded intangible investment works well. How then can a country increase publicly funded investment in intangibles without leading to widespread malinvestment? There are a few practical options.

Public R&D Funding. The first is to increase government spending on R&D: spending more on university research, public research institutes, or research undertaken by businesses. Paying for research is one of the least ideologically controversial types of investment a government can make to promote growth: it is popular with Jeremy Corbyn and Bernie Sanders on the left, Peter Thiel on the right, and a significant number of politicians and pundits in between. The rationale harks back to one of our four S’s of intangibles: spillovers. Because returns on R&D are not always captured by the person or business investing in it, businesses do less R&D than is optimal for the economy as a whole, and therefore government has a legitimate role in stepping in, either funding research in universities or institutes or paying firms to do R&D with grants or tax breaks. In total, in 2013 OECD countries spent about $40bn on publicly funded R&D and another $30bn on R&D tax breaks (Appelt et al. 2016).

The evidence for the economic benefits of public research turns out to be harder to evaluate than you might think, but such evidence as we possess looks quite positive. Research by one of the authors together with Alan Hughes, Peter Goodridge, and Gavin Wallis suggests that extra investment by the UK government in research in universities increases national productivity by 20 percent (Haskel et al. 2015). (There were substantial swings in government support for universities over the 1990s and 2000s, and those ups and downs are well correlated with productivity ups and downs, with around a three-year lag.)

As we have pointed out, correlation does not prove causation. For example, many universities are in economically fortunate areas. But does this mean having a good university raises local economic fortunes? Or do rich areas open universities? One needs a strategy to identify the causal link, if there is one, from university spending to local prosperity.

One clever way to get at the answer to this question of linkage is by studying more or less an experiment arising from a unique custom of US university finance. The economists Shawn Kantor and Alexander Whalley (2014) pointed out that many universities in the United States spend a fairly fixed amount (about 4 percent) of their endowments each year. (This practice has a name, the Bengen rule, after the financial adviser who calculated that this was a sustainable annual drawdown or spending rate from an endowment or pension fund.) So, when stock markets boom or collapse, university spending tends to rise or fall with the market values of their endowments, independently of local economic conditions. Kantor and Whalley looked at whether there was a correlation between changes in university spending per capita (caused by these shocks) and local economic conditions (measured as nonuniversity wages in the local area).

Using a sample of 135 colleges and universities located in 85 US counties, they found that, when there was a stock market boom leading to higher drawdowns, the increases in university activity (mostly increased research output) did indeed raise local incomes. So there is a spillover link from university research to local economic success, and it lasts for a long time (at least five years, based on their data), but on average it is modest. Interestingly, the strength of the link varies. The links are larger if (a) the university is research-intensive and (b) conditions in the local area are more conducive to absorbing that research. Those conditions are that firms in the local area are more high skilled and are technologically closer to the university’s research (e.g., they cite university patents).

One thing we can infer from this is that science policy can be a complement with regional policy and not a substitute for it. The benefits to providing science funding to a university in a disadvantaged town will be substantially muted unless that local community has the capacity to absorb the results of the increased research (for example, high-skilled workers and local industries that can make use of research outputs).

There is a lively debate on what specific methods—from funding public research to mission-oriented programs, competitions, tax credits for private R&D—work best. (In the UK, the mix is roughly £1 on R&D grants to businesses to £3 in tax breaks for business R&D to £10 of publicly funded academic science.) But the idea that we would expect to see more public money spent on R&D in one form or another seems a logical consequence of the growing importance of intangible investment.

R&D is not the only sort of intangible that governments can fund, though. The public sector has historically played an important, if largely unheralded, role in funding the other intangibles needed to bring products to market. Sometimes this happens through tax breaks or direct funding. The government of Singapore subsidizes business investment in a range of intangibles through what is in effect an intangibles tax break, the Productivity and Innovation Tax Credit,3 which covers design, automation of processes, training, and the acquisition and development of various sorts of intellectual property, alongside R&D. Some governments provide cheap or free advice on production methodologies (such as the UK’s recently discontinued Manufacturing Advisory Service or the US Manufacturing Extension Partnership)—this is in practice a publicly provided organizational development or design investment. Governments are also financing intangibles when they fund the arts, to the extent that this benefits those parts of the economy dependent on design, expression, or aesthetic creativity. Nesta research suggested that over 10 percent of the economy of the UK could be classed as “creative” in this sense (Higgs, Cunningham, and Bakhshi 2008), and that publicly subsidized art in the UK made a significant contribution to the commercial creative industries.

Public Procurement. Another way government can in practice fund intangible investment is by using the lever of procurement. When the US military funded the development of the semiconductor industry in the 1950s, they did not just fund R&D. By acting as a lead customer (often paying on a cost-plus basis), they effectively funded America’s businesses to invest in the intangibles needed to produce and sell chips, an investment that proved valuable when the businesses expanded into commercial markets. The Taiwanese government’s support for its nascent semiconductor industry in the 1970s and 1980s (particular through its technology agency ITRI) worked similarly: ITRI did not just invest in R&D, it incubated companies like UMC and TSMC, investing in the intangibles they needed to run semiconductor foundries effectively and link them to the global semiconductor supply chain. The success rate of industrial policy in supporting infant industries is an open question; but to the extent that it works, it is an example of government investment in non-R&D public intangibles.

The innovation scholar David Mowery has studied whether “smart” public procurement works in the United States and whether the romance about US examples like DARPA, the Small Business Innovation Research program, or the Apollo and Manhattan programs is misplaced (Henderson, Newell, and Mowery 2011). He observes that the development of the US IT sector is a major success story for procurement. In the 1950s the US military procured a lot of software and semiconductors, and this demand helped Texas Instruments and other firms invest not just in R&D, but also in the other processes necessary to make and sell semiconductors. He notes that one provision of this program was that more than one supplier was needed, and this made the sharing of information and standards in the industry a common practice. By the late 1960s, however, the military was buying a very small share of IT products, and private sector demand had become very important. Fortunately, the military projects were, in fact, highly complementary to private sector needs. Over time things have changed, and the software industry has matured to a point where the military now buys software from the private sector. An attempt by the US military in the 1980s to make its own software was a failure. So ultimately the success was in developing something that was very complementary to private sector needs, and the private sector just took over the lead.

While the development of the US IT industry is an example of procurement effectively encouraging intangible investment, there have also been some failures. In the 1970s and 1980s the US Air Force invested in the development of computer-aided machine tools, but their efforts were surpassed by those of Japanese businesses. Civilian nuclear power also benefited from swathes of defense funding and procurement spending, but if anything the needs of the defense sector (such as powering submarines and creating plutonium for atomic weapons) hindered rather than helped the evolution of effective nuclear technologies.

All this shows that using procurement to encourage intangible investment is not a free lunch. There are four things governments must believe they can get right if they want this strategy to work.

First of all, there is the question of scale. Policy wonks around the world often look enviously at DARPA, the US defense innovation agency, which spends around $3 billion a year on a mixture of innovative research and challenges and has played an important role in the development of technologies from the computer mouse to the driverless car. But part of the reason DARPA works is that it is backed up by the United States’ $600 billion defense procurement budget, one of the principles of which is to maintain the technological superiority of the American military.

Second, there must be a sufficient level of political commitment. Using procurement to encourage innovation involves a risk of failure. If a government cannot tolerate this and constantly tries to mitigate the risk of failure, innovation is unlikely to result. One of the reasons that defense procurement has in the past been good at encouraging innovation is that it has typically been insulated from many political pressures, freeing recipients of defense funding to take more risks.

Third, there is an inherent tension between the normal incentives of procurement—getting good value for money—and the risks and mindsets involved in promoting innovation. It is not just a question of personnel, although that is important—the kinds of officials who are good at getting value for money will often be different types of people from those who are good at fostering wild breakthroughs. More worrying is that when value-for-money procurement fails, innovation is often used as an excuse (“we lost money, but we were trying something new!”). This runs the risk that doing too much innovation procurement creates a cover for standard procurement failure.

The final question for any government considering using innovation to foster procurement is Clint Eastwood’s: “Do you feel lucky?” It is very hard to know what the real odds of success in innovation procurement are partly because survivorship bias is great (How many failed attempts to use procurement to foster innovation do we simply not know about?), and partly because what made it work is so unclear (To what extent was fostering innovation in semiconductors or data communications good luck? How easy would it be to pick the next winner?).

Training and Education. We might also foresee a growing public role in financing particular sorts of training and education. Governments’ involvement in training relates mainly to funding the education of young people (which has many effects and purposes, one of which is to improve the productivity of citizens as workers) and providing some subsidies for industrial training programs, such as (in some countries) apprenticeships.

Paying for citizens to go to school for longer was, for much of twentieth century, an important way that governments increased productivity; the economists Claudia Goldin and Lawrence Katz documented the vital role of education in the economic growth of the United States, pointing out, for example, that while 62 percent of the 1930 US birth cohort graduated from high school, 85 percent of the 1975 cohort did (Goldin and Katz 2008). Robert Gordon and Tyler Cowen have argued that there are diminishing returns here—children and young people can only spend so long in school or college—and that this will prove to be a major brake on US economic growth in the future (Gordon 2016; Cowen 2011).

Working out how to defy these diminishing returns has proved challenging. Goldin and Katz suggest more targeted support at all stages of education to increase the supply of educated workers: more very early stage support, lower class sizes for middle schools, and more support for college. And, of course, some people are spending longer in school, as more occupations demand degrees or even postgraduate qualifications. But the challenge of trying to fit more education into the finite number of years of a person’s youth remains.

Others have argued that the answer lies in changing not how much we teach, but what we teach. In recent years, it has become fashionable to argue that particular types of education may be unusually valuable: consider, for example, the fashion for teaching coding in schools, or for encouraging children and students to learn collaborative problem-solving skills, both of which, it is argued, will give them skills that will be particularly useful in the economy of the future.

However, we should be somewhat skeptical about our ability to predict what skills the economy of the future needs and our ability to teach them. Perhaps in twenty years’ time, coding will, for the most part, be automated. Perhaps collaborative problem solving cannot be inculcated by changing the curriculum.

But there is an alternative that may solve both the problem of when to teach people and the question of what to teach them: increasing the amount of training people receive in adulthood. Adult education has always been the Cinderella of the educational system, starved of prestige and of public funding. But its usefulness in an increasingly intangible-rich economy seems clear.

First of all, adult education by definition need not delay people from entering the workforce; being able to invest in a person’s education throughout a person’s life makes many more decades available. Second, the availability of adult education reduces the problem of trying to guess what skills will be valuable in twenty or thirty years’ time. For all the excellent research done on the skills needed by the economy of the future, predicting a couple of decades out reminds one of Sam Goldwyn’s advice: “never make predictions—especially about the future.” But if people have the opportunity to acquire more skills during their working lives, prediction becomes less important: adult education provides people with option value. It may also help mitigate some of the problems of inequality we described in chapter 6: to the extent that the growth of intangibles disadvantages those with poor skills and makes some skills obsolete, the availability of training offers a way of redressing the imbalance.

However, anyone planning to expand adult education faces a strategic problem: how to deliver it effectively. Schools, universities, and further education/community colleges are well established organizations with long track records. They may not be perfect, but they have evolved and improved over many years, and society has evolved alongside them, such that going to school and into further or higher education is an expected part of most people’s life-course, at least in the developed world. Adult education is less of a known quantity. Moreover, it seems likely that new technologies ought to make educating adults easier: digital technologies should offer ways of teaching more cheaply and conveniently. (MOOCs—Massive Open Online Courses—seem so far not to have lived up to their initial promise, but the field is still less than two decades old, and it is too early to say whether better versions could be substantially more effective.) What is needed is significant investment in innovating how we deliver adult education, to identify new models that work cost-effectively and at scale. Even if these forms of education end up being paid for by the adult students themselves rather than by taxpayers, the research to develop new models that work seems like a worthy goal of public policy.

Government funding can also help reduce coordination problems that may hold businesses back from investing. Suppose there are big economic gains to be had from developing self-driving cars and reconfiguring our cities around them (fewer car accidents, more productive commutes, freeing up parking spaces for redevelopment, and so on). But realizing these benefits requires a lot of investments to be made together (driverless car technology, urban design, new insurance policies, and so on); it may well be that no company is willing to make investments on its own unless it knows that others will make complementary ones. In this case, it is possible that government investment may not only be useful in itself (by funding high-spillover investments that others would not make), but would also encourage wider investment by increasing the likelihood that others would make complementary investments. The role of government in making these kinds of “test-bed” investments will increase in an intangible economy.