The Future of Capitalism

by Paul Collier

  To create equivalent influences in America and Britain we must waive farewell to the symbols of cognitive privilege. The word ‘degree’ needs to be defanged: lathes and Latin America can become more glamorous than three more years of the classroom. Germany has done this well, but the leader is Switzerland. Vocational training in Switzerland is serious: courses are typically three to four years long, and firms are closely involved because they pay half of the costs, which are considerable. It is also popular: 60 per cent of young people choose vocational courses, partly because they are paid while studying on them, but also because such training is an accepted avenue to top jobs.* The achievement is all the more remarkable because this world-beating vocational training coexists with a university in the global top-ten: the cognitive paths do not have to be weakened in order for the non-cognitive paths to thrive.

  Vocational training needs enhanced kudos, not only for those who take the courses, but for those who give them. Teaching cognitive skills provides easy kudos: we have titles like ‘professor’, and belong to a ‘university’. Vocational training is currently too fragmented to offer such easy kudos. Perhaps the many vocational courses need to be given the common enhanced status of meeting a vital national purpose: a National Skill Service in which all staff can take pride.

  Securing the job horizon

  Once in a productive job, how much security of employment should a worker have? Workers take on long-term obligations, such as mortgages, and so need as much job security as possible. In contrast, firms face periodic shocks to the demand for their products and so will want as much flexibility as possible. The compromise they reach will depend upon their relative bargaining power, but this in turn is heavily influenced by government policy. At one extreme, exemplified by France, governments legislate to make job security a requirement of employment. At the other extreme, exemplified by America in the 1920s, governments legislated to restrict unions. In between, sector-by-sector differences in the bargaining power of workers produce a patchwork. Every professor, however pedestrian, has lifetime job security: otherwise we might get anxious and this could interfere with our ability to think great thoughts (doubtless, other professors will come up with further justifications). Meanwhile, my award-winning actor nephew, working in a sector saturated with job-seekers, looks forward to a lifetime of transience.

  In rethinking employment rights, ideology is not going to help: while the ideologues of the left abhor a market for labour, those of the right sanctify it. The most common free-market critique is that minimum wages cause unemployment. Unemployment is the most salient indication that something is amiss, but it is not always the most important. A labour market has two distinct functions. The one that matters for unemployment is to pair up job-seekers with a particular skill with the jobs that firms create for those skills: what is going on is matching. But the function that matters for mass prosperity is the creation of those skills: investment. There is an inherent tension between these two. Being able to make binding commitments can make the investment more viable. The training that a worker needs in order to acquire a skill is costly, and someone has to pay it. To the extent that the worker pays for it, she worries whether the firm will employ her on a higher salary for long enough that her investment in training is worthwhile. But to the extent that the firm pays for it, it worries that, once trained, the worker will quit and take a better-paid job with another firm. Guaranteed job security can give the worker the confidence to overcome that first worry. The unemployment generated as a side effect of wage controls can give the firm the confidence to overcome the second one, so between them they are likely to increase investments in training. But guaranteed job security and wage controls discourage firms from hiring workers, and so impede the matching function of the labour market. That is why it is better to solve the firm’s investment problem not by using high unemployment to discourage workers from quitting, but by paying for the training through a government-imposed levy.

  But workers need job security not just to recover their investments in skill, but because they take on commitments that anticipate their future salary. This ability to take on commitments such as raising children, or buying a home, is beneficial to society, so job security is socially valuable. It may be more efficient for the firm to adjust to the need to pay the worker during periods of slack demand than for the worker to bear the risk of being laid off. If the firm has to keep the worker, it may train her in several tasks so that when the demand for one task drops it can switch her to another.

  There has, however, to be a limit to such security; while firms should be able to cope with temporary fluctuations, they cannot adjust to a large, permanent drop in demand without shedding labour. At the limit, the firm goes bankrupt. Yet the fact that job loss is unavoidable in no way mitigates the cost to the worker. For this class of shock, we need an entity larger than the firm – the state. Nobel Laureate Jean Tirole has proposed a smart way for government to induce firms to retain workers through market troughs, while still enabling them to shed employees when faced by a permanent contraction. This is to impose a charge on labour-shedding to reflect the extra costs to the state of welfare payments and retraining.

  The governments that have reputedly best responded to such job shocks are those of Denmark and Sweden, which developed the concept of flexicurity. The policy is closely related to the challenge of reviving broken cities: if an industry has collapsed, it will have hit some specific locations hard and its workers will need retraining. Janesville is a rare study of retraining programmes in an American town hit by the closure of its major plant.17 It reveals that the retraining was a decisive failure. Those among the redundant who took the programme were less likely to get work than those who didn’t and, if they did find work, earned less than those who didn’t retrain. Why did the programme fail so resoundingly? I think that three key things were neglected. Moreover, the neglect went right back to schooling: the men made redundant had never been taught things that are basic for modern learning. The neglect then continued through their long period of employment at the plant. Not having been faced with the prospect of the penalties of redundancy proposed by Tirole, the firm had had no incentive to equip the men with a broader range of skills that could have made them more employable. But above all, retraining was not co-ordinated with any targeted stimulus designed to attract a new industry to the town. Instead, the cluster effect triggered a downward spiral in which the closure of the plant led to corresponding contraction among the other local employers so that there were few jobs for retrained workers to chase. The experience recounted in Janesville suggests that, without such a high-profile co-ordinated effort, retraining is a snare offering the illusion of hope. But most likely, even with better education, a wider endowment of prior skills and a big push to form a replacement cluster, redundant workers will hesitate to sink their newly needed savings into retraining. Two professors at Chicago’s Business School, Luigi Zingales and Raghuram Rajan, have proposed that all workers should be given a lifetime credit that they draw on for retraining as needed.*

  The incipient robotics revolution, and whatever further technological revolutions lie beyond it, will require many people to retrain. Robotics is, I think, unlikely to reduce the need for work – our wants are probably insatiable. But it will change the composition of tasks for which workers will be needed. This is the essence of a valuable insight. Think of the typical job as made up of a series of tasks. Even the most seemingly routine job invariably involves moments that require judgement, the ability to interact with other people, and some non-routine action. Robotics will eliminate some tasks, and in doing so will sharply reduce the cost of the output currently produced during a working day. By redeploying to the remaining tasks that do not lend themselves to robotics, and to new tasks that reliance on robotics create, the typical worker can become much more productive.18 Because different jobs have very different compositions of robotic-suited and unsuited tasks, the skill composition of work is likely to keep cha
nging substantially; periodically, people will need to retrain to be able to perform new packages of tasks. Just as Parisian waiters earn more than their counterparts in London, tomorrow’s workers will earn more than today’s, but only if, like those Parisian waiters, they learn different skills. A corollary is that one of the highly labour-intensive sectors that will need to expand massively is the training sector.

  Security in retirement

  I would like to retire, but please not yet. But I already know the incomes I will receive from my state and university pensions: I am secure until death. Not so, many others.

  Risks can easily be pooled, and for most types of risk, if they are pooled they evaporate. The reason for caution in pooling risks is ‘moral hazard’. In some situations, once the risk is shared, everyone takes greater risks: because we all have fire insurance we are more careless. But one risk borne by many pensioners involves no moral hazard whatsoever: it is the risk involved in all defined-contribution pension schemes. Virtually all firms have decided that defined benefit schemes such as my own are ruinously expensive. My own scheme, for British universities, bears this out; it has accumulated the largest deficit in a pension fund ever recorded. Fortunately for me, this will not affect my own entitlement, which will be borne by the next generation of academics and by students who will pay higher fees. They will be heartened to know that I am truly grateful.*

  Meanwhile, everyone else has been shunted into the defined contribution pension schemes. Here they find themselves bearing three risks. One is that the entire pension fund into which they are contributing may perform worse than other funds; in contrast to a defined benefit scheme, the shortfall is no longer the liability of their employer. Another is that their choice of investments within the fund may do worse than the average choices of other employees. Finally, on the day on which they retire, when their benefit is cashed out, the market may have dropped below its long-term average: stock markets are sometimes highly volatile. As a result of these three risks, two workers with the same history of pension contributions can end up with considerably different pensions.

  While the defined benefit schemes such as my own are too generous, shifting all the risk on to society, the defined contribution schemes are needlessly exposing people to avoidable risks just when they are least able to bear them. They have shifted from pooling risks so that they evaporate, to dumping them on individuals at a time when they are vulnerable. This is an eminently fixable error of design.

  But the people facing the most serious retirement insecurities are those whose working life is spent shuffling between the firms from hell. They do not even accumulate entitlements to defined contributions. Dumped on to society when they are too old to work, they become society’s liability. Again, this is a market failure: their employers have been allowed to cut their employment costs excessively in not making adequate payments into a pension scheme. As with minimum wage laws, French policy looks to be superior to the Anglo-Saxon model: the high contributions required from employers ensure that as long as people work they build up an adequate entitlement to a pension. That proviso of course implies that the economy must be run in such a way as to generate enough productive jobs for everyone. This is the critical benchmark that must be met by training programmes; mopping up the unemployed with lousy jobs is a failure, not a substitute.

  Belonging to society

  While I have emphasized family, workplace and nation as the cornerstones of belonging, in all healthy societies there is also a dense web of networked groups to which people become attached. Robert Putnam’s celebrated book Bowling Alone lamented the decline of these forms of belonging in America. Such attachments encourage people into habits of acknowledging reciprocal obligations, as well as countering isolation and its corollaries of loss of self-esteem, and depression. The decline in America is neither inevitable nor universal across the West. In Germany, formally registered civil society groups, vereine, are common and increasing. Half of all Germans belong to at least one such club, and the number of clubs has increased by a third in the past twenty years. The proportion of Germans taking part in such groups is around triple that of southern Europe.19

  CURBING THE HAVING-​IT-​ALL’S

  The rise of the new educated class has certainly widened social inequality. But most of the behaviours that have made it so successful have not been at the expense of the rest of society. Their strategies would be better emulated than curbed. But some aspects of the educated class’s success are at the expense of others: zero-sum housing demand; zero-sum work; and zero-sum social behaviour.

  Housing: homes versus assets

  People have two motives for buying a house. For most people it is a home; for some it is an asset. In the Britain of 1950, half of the entire housing stock was owned as an asset and rented to people needing a home. Only 30 per cent of people actually owned their home. One of the triumphs of social democracy was to transform this situation. By 1980 the private rented sector had shrunk drastically, to only 10 per cent, while owner-occupation had nearly doubled. In the early 1980s a further twist in public policy raised owner-occupation to a high point of 70 per cent, by enabling tenants in social housing to buy their homes at a discount.

  This increase from 30 to 70 per cent was a cumulative triumph of public policies. Owning a home enhances the sense of belonging, and that, as I have suggested, is a vital social good. Belonging is the foundation for reciprocal obligations. Home ownership also gives people a greater sense of having a stake in society, and inclines them to be more prudent: psychologists have discovered that, once people have something, they become highly averse to losing it. And owning a home anchors people. A street in Oxford was once divided halfway along between rental and owner-occupation; the divide is still visible because of the height of the trees – only owners planted them.

  Four public policies kept house prices affordable for families on median income. A building programme run by local government increased supply; restraints on net immigration limited the rate of increase in households; curbs on buy-to-let restricted the pure asset-demand for housing; and curbs on the ratio of mortgages to income restrained what people could bid. The asset-transfer to tenants in social housing complemented these policies, by enabling families whose incomes were below-median to own their house.

  From the late 1980s this progress began to unravel. Home ownership has already fallen to 60 per cent and is still declining; young families can no longer afford to buy a home. Over the past twenty years, the price of the average house has jumped from 3.6 times average earnings, to 7.6 times. This is not surprising: all four of the policies that had restrained house prices have been reversed. Local government building programmes were stopped in the hope that private firms would replace them (they didn’t, partly because acquiring land with planning permission was far more difficult for them than for local government). Immigration controls were relaxed, becoming the main driver of household growth. The rules that had curbed buy-to-let were replaced by those that encouraged it, unleashing a huge new asset-demand for housing. Buy-to-let properties have doubled to around 20 per cent of the housing stock. Finally, the curbs on mortgage finance were lifted, giving way to the lending frenzy that seized the banks in a bonus-hungry race over the cliff. That is why there was an explosion of house prices. Nor, as new below-median income families formed, was there any equivalent to the asset-transfer programme.

  As a result of high prices and unrestricted credit, the people who wanted housing as an asset were able to outbid the people who wanted housing as a home, who typically were young families. Twenty years ago, over half of young families took out a mortgage; now it is around a third. Those squeezed out were not the high-skilled assortative maters, but those in the less-educated class. Their inability to buy a home, and the diminishing prospect of ever being able to do so, is central to the new anxieties. But who are the people who outbid them? With house prices rising, everyone wanted to buy a house: the people who were able to do so were those who c
ould borrow the most. The winners in this race have been the older members of the educated class, and smart people who exploited the borrow-and-let opportunity to the hilt. A spectacular case was a couple of teachers who quit their jobs and accumulated a vast housing empire. The affluent and the smart have benefited from a double bonanza: being better able to borrow than young families, they can charge rents that exceed their interest payments. On top of this, as house prices have risen, they have accrued huge capital appreciation.

  So, what can be done about it? Again, ideology is a menace. Those on the left want to return to the rent controls of the 1940s; as then, this would freeze people into the home they are currently renting, reducing job mobility. Those on the right want to increase finance for first-time house purchase; by further fuelling demand, this would jack prices up yet further. Yet addressing this problem is not difficult, because we know what worked: the same policies would work again.

  It makes sense to increase supply, and the most credible way of doing so is to break the planning log-jam. Local governments are best placed to plan new building programmes, while execution can be in partnership with commercial developers. Local authorities could plan build-to-buy, instead of build-to-rent. But an increase in the supply of housing needs to be gradual: a quantum increase would risk crashing house prices, plunging many young home owners into negative equity. Correspondingly, it makes sense to curb household growth by restoring restrictions on immigration. The credit frenzy unleashed by financial deregulation did not usher in nirvana – it ended in the regulatory disgrace of a bank run. The sight of depositors besieging the branches of Northern Rock was the first such spectacle in Britain for 150 years. As with a house building programme, change will need to be gradual, but its direction is unambiguous: we need to return to ceilings on the ratios of mortgages to income and of mortgages to deposits. It also makes sense to curb buy-to-let. The public benefit from home ownership warrants giving priority to those who want a house-as-home, over those who want a house-as-asset.