Much of the structure of income inequality arises from the structure and behaviour of firms. If the labour share falls then more money is being distributed to the owners of capital, who in general are likely to be better off than workers. If firms reward a small number of their most senior employees much more than the bulk of their employees then earnings inequality will be high. If some firms are very much more successful than others, perhaps because they are much more innovative and/or perhaps because they have secured some form of monopoly position, then their employees and shareholders may do especially well, perhaps at the expense of consumers. If firms as a whole are not becoming more productive then wage growth is likely to stall.
A panellist’s introduction, by Richard Blundell
It seems likely that all of these forces have been at play in the UK, and internationally, over recent decades. There is evidence of a falling labour share – although that is less evident in the UK than in the US, and perhaps less than in other OECD economies. Within-firm wage inequality has increased as senior managers and more skilled employees have taken a greater share. We have seen a rise of ‘superstar’ firms, especially in the tech sector where there are often very high returns to innovation which can often become returns to something akin to a monopoly position in markets (which have features of natural monopoly). We have certainly seen a dramatic lack of productivity growth among the majority of firms for at least a decade and probably longer, resulting in wage stagnation.
The conclusion must be that if we want to influence inequality, we need to take seriously policy towards firms. That is likely to mean a robust policy on competition, which adapts to take account of the new technology firms. It may mean new rules on corporate governance and ownership. It will certainly mean a focus on increasing productivity throughout the economy by promoting innovation, investment, competition, and effective partnerships between public and private sectors to ensure public R&D and investment are best targeted and used. Policy towards firms has not adapted over recent decades to dramatic changes in technology, market structure and globalisation. The clear argument of all the papers commissioned for this part of the review is that overhauling these policies will be fundamental to tackling growing inequalities.
The chapter, by Jan De Loecker, Tim Obermeier and John Van Reenen, takes a fresh look at firms and inequality in the UK, establishing some of the main trends in ‘firm’ inequality and examining what lies behind them. Using unique new data sources, and a detailed comparison with other developed economies, they note that the UK looks broadly comparable to the US across many dimensions. Similar patterns are there on average in other EU countries too. The authors use this evidence to argue that it is unlikely that differences in institutions alone can explain changes in firm inequalities, more likely that common trends in technology and globalisation play a key role.
Among some of their most striking findings is the strong rise in the inequality of productivity across firms in the UK, and the slowdown in productivity across all firms since the Global Financial Crisis. The top 5% of firms in the UK, ‘frontier firms’, saw productivity grow by about 60% between 1996 and 2010 whereas for the remaining firms productivity grew by only just over 10%. Perhaps most shocking is that median productivity has not increased over this period at all. This has a direct bearing on equality across people as the authors show that larger more productive firms typically pay better and the distribution of wage inequality between firms mirrors the pattern of productivity.
The growth in the level and distribution of firm markups – the excess of price over marginal cost – may also impact on inequalities across people as increasing markups may reflect a larger share of productivity growth and rents going to the owners of capital rather than to employees. The authors find a large increase in the aggregate markup in the UK together with an increase in markup inequality, documenting essentially zero change in markups for firms at the median while firms at the 90th percentile have had a 40% increase.
It is partly the link between markups and the labour share that has focused attention on rising markups. The authors note though that this will depend on the level and distribution of fixed costs. Not only have fixed costs changed over time, they have also become harder to measure, as is clearly the case for intangible investments. Perhaps even more importantly, changes in self-employment and how self-employment income is counted, as well as changes in non-wage compensation, look to be playing a major role in calculations of the labour share for the UK. Consequently, even though the labour share in the UK appears more stable than in the US and other economies, the authors argue that there is some fall in the labour share between 1981 and 2019, especially when corrected for self-employment and non-wage compensation.
To understand the growth in high-performing superstar firms, the authors point primarily to changing technology and globalisation. Even if the powerful positions of these firms were gained primarily through the forces of technological change and globalisation, the chapter points out that the potential for increasing monopoly power in consumer markets as well as increasing monopsony in the market for labour (and other inputs) can allow such firms to increase their markups over consumer prices and markdowns over the marginal product of labour and other inputs. Outsourcing has also increased, potentially increasing large firms’ power in outsourced low-skilled occupations such as cleaners, caterers, security guards and drivers. The decline in trade unions is likely to have also played a role: since unions are located predominantly in large firms, their weakening could have helped big firms become more profitable.
As the broad trends in productivity, wage inequality and markups are similar across countries, the authors do not see institutional change, such as changing competition policy and labour market regulation, as the main drivers of these firm inequalities. Rather they point to changes in technology and globalisation. Consequently, they argue for a focus on increasing firm productivity through better policies on innovation, skills and investment, noting that improving the human capital of workers (especially at the middle and bottom end of the distribution) is the surest way to raise wages and reduce inequality. They do though see a clear role for modernising antitrust and improving the power of labour – specifically, changing competition law to account for the increase in digital technology, and changing labour market regulation to adapt to the growth in self-employment, outsourcing and gig workers.
In his commentary, Jan Eeckhout asks the question, ‘Does market power offer an alternative (and complementary) explanation to technological change for the rise in inequality?’. He focuses on the labour share and notes that the UK is the one notable exception where the labour share has remained fairly stable. As discussed above, the chapter authors have produced new evidence which partly reconciles these differences between the UK and other economies, but even so the labour share in the UK does not appear to have fallen as dramatically as in the US. The commentary then goes on to highlight the low share of manufacturing in the UK, an outlier relative to almost any other developed economy. Since the labour share is observed to decline most in manufacturing and less in other sectors, this could help explain the patterns in the UK. Moreover, the high representation of services could explain the dichotomy between rising market power and a more stable labour share. Eeckhout also points out that top incomes tend to be part of overhead costs, and if overhead in the UK consists disproportionately of labour then we may see a declining variable labour share, but a constant labour share once overhead labour is included. If market power is the cause, what should be the response? In the face of the importance of new digital technologies, the commentary argues that breaking up firms is not necessarily the best option. Instead, regulation such as interoperability would be preferred. Such a policy can maintain the advantages of scale economies but engenders competition between firms operating with this technology which keeps the scale advantages.
Philippe Aghion and Rachel Griffith, in their commentary, look in more detail at the relationship between innovation and inequality. Their main focus is on the impact of firm-level innovation on income inequality. They argue that although the expansion of ‘superstar’ firms can initially boost productivity, the long-term effect of this market expansion has been to discourage innovation and entry by other firms, hence the decline in total factor productivity (TFP) growth. They suggest that a lack of appropriate competition policy in the US, and in particular the absence of a good regulation of mergers and acquisitions, has facilitated this expansion of superstar firms to many local markets and product lines, and therefore it has fostered the decline in aggregate productivity growth. In terms of policy, they argue for a US eco-system of innovation together with a Danish system to protect individuals who lose their jobs and to insure all individuals against big macro shocks. They also suggest the need to closely analyse the organisation and functioning of lobbying.
Mariana Mazzucato, in her commentary, makes the case that the contribution of the state for innovation and firm productivity has been understated with important consequences for the overall distribution of income and wealth. She argues for tilting the playing field, not levelling it, to penalise rent-seeking, reward practices that foster long-termism and sustainability, and ensure a fairer distribution of that co-created wealth. She also suggests that mission-oriented policies can help to shift the status quo market-fixing dynamic towards market co-shaping. Market failure analysis cannot address issues around new technologies and sectors that did not previously exist. Markets themselves should be viewed as outcomes of the interactions between both public and private actors. The idea is that for many of the great new innovations, the government/state is important along the entire innovation chain from basic research to product development. It is when government investments are bold, strategic and mission-oriented that the most crowding-in has happened. In her view, missions exemplify a more proactive approach to policy than market fixing suggests. They do not fix existing markets but create new markets.
In his piece for the review, Colin Mayer examines the nature of the UK’s financial, ownership and governance arrangements. He makes the case that the UK has moved to a highly centralised banking sector that provides predominantly short-term working capital and, as a result, small and medium-sized enterprises (SMEs) are dependent on equity sources to fund their growth and expansion. The result is a high level of regional disparities in financing and governance, and the replacement of long-term relationships between investors and firms with short-term, transactional engagements. Those engagements have become increasingly focused on shareholder returns at the expense of the interests of other parties, with adverse consequences for aggregate productivity, diffusion of productivity gains between firms, inequality within as well as between firms, and regional disparities. The focus is on the concentration of activity in the South East and the lack of diffusion, noting that the UK has one of the highest levels of divergence in labour productivity across its regions among major industrialised countries. The argument is that this divergence is down to the way small firms are financed and the lack of local banks with local knowledge – in particular, that over the course of 150 years, the British financial system shifted completely from a locally based relationship banking and then stock market economy to a highly concentrated banking and institutional investment system headquartered in London. Mayer makes the case that the purpose of a business is to promote not only the interests of shareholders but also those of its customers, societies and the natural world in which it operates. The objective of a firm should be to ‘produce profitable solutions to the problems of people and planet, not profiting from producing problems for either’. The implication is that attention needs to be given to the laws and regulations, particularly those pertaining to finance, ownership and investment, that underpin the UK financial, ownership and governance system.
We have seen how the market power of firms can have an important impact on many inequalities between people that concern society. These appear to be particularly accentuated for tech companies. In his paper for the review, Jean Tirole digs deeper into these concerns and takes a detailed look at the challenges for competition policy and regulation in the digital age. He argues that public policies could be much improved within existing laws. The network externalities and fixed costs that characterise many new technological innovations have resulted in large markups and limited competition. High markups may reflect excessively high prices (or high fees for advertisers/sellers), low innovation and the abuse of a dominant position. Alternatively, they may reflect the high costs of providing services which eventually fall on the consumer of the services. To investigate further, Tirole poses the question, ‘Are platform profits in line with investment costs, or do platforms enjoy “supranormal profits” or “ex-ante rents”?’. The nature of service being provided, the large set-up costs involved and the, often, long periods of negligible profits with the prospect of large future returns make the analysis much harder than in the standard textbook monopoly case. But there are key changes to policies and institutions that could and should be made.
Tirole notes that tech companies exhibit natural-monopoly characteristics much like those that were evident in the public utility companies of the 20th century. When it comes to policy though there are new challenges. Fixed costs associated with set-up are hard to measure and verify, making it difficult to calculate a reasonable rate of return which can cover the initial costs as well as the risk involved in the success of the original investment in innovation. With tech industries based on service provision and a rapidly changing technology, it is particularly hard to separate an underlying stable essential facility that could be regulated. These companies are typically global too, requiring coordination in policymaking. There is need for a more dynamic approach to examining competition – competition for the market rather than competition in the market. The idea is to keep incumbents on their toes, innovating to keep rivals out. Of course, policy should ensure the actual entry of competition and not the takeover of entrants by incumbents. Innovating firms are often acquired by incumbents, typically in the early stages of product development. This can be good for overall innovation, but it could be that an acquiring company will buy an entrepreneurial start-up specifically to shut down the target company’s activities and projects before they develop into a competitive threat, so-called ‘killer acquisitions’. Low initial profits with the prospect of long-term gains mean competition policy should be agile and proactive, relying less on current turnover for merger and acquisition approvals.
Data play a central role in tech companies and in modern technologies. Data holdings by large tech companies allow personalised advertising, generating much more revenue for the advertiser. There is a strong case to limit self-preferencing, for preserving multihoming and to limit exclusivity agreements. The service provided might often appear free to the consumer but the individual data are highly valuable and are used to charge a fee for use on the other side of the platform. This is as true for search engines as it is for healthcare apps. Alternatives are possible: a limited amount of data could be collected, or there could be compensation through micropayments, or data could be licensed, perhaps available as a ‘public good’.
There is increasing pressure to develop an active industrial policy to address some of the concerns raised from the slow and uneven productivity among firms. Tirole lays out the case for the development of a transparent code of conduct for industrial policy which, among other recommendations, requires independent experts, dissemination of results and strengthening universities to bring them closer to the start-up world, with monitoring by an independent agency being an overarching requirement.