The CMA's research on competition and UK labour markets
A speech delivered by Sarah Cardell, Chief Executive of the Competition and Markets Authority, at Durham University.
Introduction
I am delighted to announce the publication today of the first major research report by the CMA’s Microeconomics Unit: Competition and market power in UK labour markets.
I’m also pleased to be launching this work here in Durham, where the Microeconomics Unit benefits from a highly productive partnership with Durham University. Engagement with a range of external experts is hugely valuable to the CMA’s research activities, and indeed, much of our broader work. We are grateful for those relationships, which strengthen our capabilities and enhance the impact we can deliver for people, businesses and the UK economy. We look forward to continuing this close collaboration between the Microeconomics Unit and Durham, as well as many other academics across the UK.
This is important new research on the UK labour market and certainly some of the first of its kind. It’s also different from much of the CMA’s work, reflecting the broader purpose of our Microeconomics Unit to produce microeconomic research that informs not only the CMA’s own work but also wider government and policy thinking.
As well as unpacking the findings of the report, I want to outline why we consider labour markets such a timely and pertinent area of exploration, and what the findings might mean for the CMA and others. I’ll also share some thoughts on the CMA’s wider portfolio of work relating to labour markets which we have identified as a strategic priority in our Annual Plan. First, though, let me provide some context on the Microeconomics Unit itself, and its important contribution to economic research and evidence-based policy development.
The CMA’s Microeconomics Unit
The benefits of establishing a ‘microeconomic sibling’ for the Bank of England’s macroeconomic research function were well-argued in the UK government-commissioned 2021 Penrose report on competition policy. The Microeconomics Unit was duly established in 2022 and now has 17 very talented staff based in its office in Darlington.
The Microeconomics Unit helps the CMA keep pace with emerging issues and the latest methods in industrial economics, to evolve the analysis conducted in our core casework. In addition to informing the CMA’s casework, the Microeconomics Unit has also been designed to be an independent, open-access, collaborative centre of microeconomic research expertise for government as a whole. It’s important to bear this in mind as you read this report because not all of the findings will directly inform the CMA’s own work and much of the analysis will be of wider policy interest.
Competition and market power in UK labour markets is an excellent first flagship report for the Microeconomics Unit. We know the government values the research expertise of the unit on such a topical area. However, this is by no means the only output it is producing. Very soon, the Unit will publish research on investment and innovation and in May we expect to publish the third edition of our State of Competition report, which surveys the degree and drivers of product market power across the UK. We are also working on topical research on ownership networks in the UK, drivers of productivity and growth, and the results of the next Consumer Protection Survey. This labour markets report is a particularly important and timely piece of work, however. Let me explain why.
Competition authorities’ interest in labour markets
This report adds to the robust body of evidence to support the benefits of well-functioning labour markets, widely recognised as an important driver of economic growth. Where labour markets work well, workers are able to access the right jobs for them and firms find the workers they need in the easiest, most efficient way. The amount each worker can produce rises, more suitable workers contribute more in a given hour, which in turn boosts revenues and wages. Governments have been historically interested in improving how labour markets work to boost participation and grow the economy – 2 critically important goals for the UK in the current economic climate. It is therefore in the shared interests of workers, firms and governments to ensure that the labour market works well for everyone.
However, we also know that frictions will always exist in the labour market. David Card, Nobel Prize winner in 2021 for his work on labour markets, has said that “the time has come to recognise that many – or even most – firms have some wage-setting power.” From a competition law perspective, this wage-setting power may be illegal, if it involves collusion between firms.
It is well-established in economic theory (and supported by empirical research) that employer market power can lead to lower wages. End-consumers may also be left worse off if firms with labour market power restrict the number of workers they hire, potentially reducing output and resulting in higher prices if the firm has market power in the product market as well.
Despite this, labour markets are an area in which competition authorities have traditionally been less active. We have generally focussed principally on competition in product markets. However, we have the power to take enforcement action against firms which break the law by fixing wages, just as we can act against firms which collude to fix prices.
Defining what we mean by labour market power and concentration
This report provides valuable new analysis and evidence to advance understanding of a range of issues relating to labour markets, not just for competition policy but more broadly. I hope it will provide helpful insights to inform thinking across the academic and policymaking community at a time when we are all seeking ways to stimulate our economy and make people better off.
The title of the report is Competition and market power in UK labour markets. Before I turn to the findings in the report, let me first define what we mean by concentration and market power when it comes to labour markets.
When we say concentration, we mean how many firms there are in a particular market. A low degree of concentration means the industry is closer to a perfect competition scenario, where many firms of more or less equal size share the market. Highly concentrated industries will have a small number of firms with a much higher share of the market. In this report, we’ve used the Herfindahl-Hirschmann index, also known as HHI, to measure concentration in labour markets.
Employer market power is the ability of firms to pay workers less than the value of their contribution to the firm’s output. The report outlines 2 driving causes of this market power:
Firstly, oligopsony, where the labour market is concentrated and there are a small number of employers in a market. Just like monopsony, where there is one buyer in a market, a small number of suitable employers, for example in labour markets where workers have specialist skills, gives market power to employers.
Secondly, the costs of searching for a new job. Searching for new jobs takes time and is costly for workers. The more difficult it is to find out about new opportunities, the more market power employers have. When searching for jobs is costly or time-consuming, workers are less likely to leave. Knowing this, employers can pay them less.
A word of caution, employer market power and labour market concentration are not one and the same. Just because a labour market is concentrated, this does not mean firms can automatically exercise market power. Other factors such as the technology firms use, the mobility of workers and existing employment laws may all influence both concentration and market power.
Report findings
I would like now to take us through some of the key findings of the report. The report has taken data from across the UK to understand how labour markets vary across regions and nations, sectors and types of occupation. I’ll highlight some of our key findings relating to 4 broad areas:
-
first, levels of labour market concentration
-
second, the impact of labour market power on wages
-
third, the prevalence of non-competes
-
and finally, the impact of some recent developments in labour markets
Undoubtedly, it is a source of new evidence in the academic debate around labour markets. While this major research piece advances the understanding of UK labour markets, it does not cover every aspect of this complex area and nor does it mean every finding needs a response, either from the CMA or others.
How concentrated are UK labour markets?
So how concentrated are labour markets in the UK? The report finds that mean labour market concentration in the UK is roughly the same as 20 years ago, despite structural changes to the labour market, for example the boom in financial services, rise of the gig economy and the impact of the COVID-19 pandemic. Mean concentration fell by 25% between 2002 and just before the pandemic. It then showed a sharp but short uptick during the pandemic, back to near 2002 levels. This contrasts with the US where there is a substantial body of research showing that market power in labour markets over there has been increasing over time. Indeed, US competition authorities have begun to look more carefully at competition in labour markets, recognising the rising employer market power over there.
As an aside, I’d note that this highlights another important aspect of this work: it is UK focused. Much of the competition policy debate around labour markets has been driven by research findings about the US economy. It is important that domestic UK policy is informed by evidence related to the UK economy.
How much do workers receive?
One issue the report addresses is the share of revenue received by workers. Where there is labour market power, workers receive less than their full contribution to revenue. This is true for most workers in the UK to some extent. How much market power there is ranges widely, reflecting the diversity of the UK labour market. The share of income workers receive has been rising slightly, since 2008, to around 2 thirds of their contribution to revenues.
Concentration negatively affects wages
Although labour market concentration in the whole economy today appears to be slightly lower than 20 years ago, there are still some industries and areas which are highly concentrated. In these industries employer market power can depress wages. Higher levels of labour market concentration are associated with lower wages, even after comparing similar workers in similar businesses. For comparable workers working in comparable firms, a move from a market with many employers (in the least concentrated 10% of labour markets) to a highly concentrated labour market with few employers (in the most concentrated 10% of labour markets) is associated with a roughly 10% decrease in a worker’s wage.
This matters for regional disparities
The report also shows that labour markets are much more concentrated outside of London and the South-East. For example, labour markets are around 4 times more concentrated in Scotland and Wales than they are in London. This means that, for a given occupation, fewer businesses account for a larger share of jobs within an area most people consider their relevant job market, a so-called Travel-to-Work (TTWA) area. In simple terms, there are more employers per person in London and the South-East than outside this area. In rural areas, we’d expect to find far fewer employers for a given occupation and that’s why we find that these labour markets are more concentrated.
…and for sectoral disparities too
Turning to variations across different sectors, workers in managerial and professional jobs have faced roughly constant concentration, while concentration for workers like plant operatives, skilled trade and care professionals has decreased on average since 2012 as relative labour demand has outstripped supply, sometimes by around half. There is substantial industry variation in labour market concentration. Manufacturing, transport and storage and financial services are particularly concentrated. And in the last 20 years, labour market concentration has risen significantly in administration and support services, utilities, and mining and quarrying.
And in some circumstances this can be affected by collective bargaining
For affected workers in concentrated industries, the data shows that the impact on wages from employer market power is substantial. Unsurprisingly, it also shows that collective bargaining agreements can provide some counter-balance to concentration. It’s important to note though, that this finding only applies within concentrated industries, and the results we have focus only on the private sector.
Non-competes are widespread, even in low-paid jobs
One of the main areas the report has looked at are non-competes. These are restrictions on employees’ ability to work for a rival firm once they leave their current employer. It is important to note here that non-compete arrangements between an employer and an employee do not generally breach UK competition law. They are typically a matter for employment law. This is to be distinguished from no-poach agreements between competing employers which I will come back to later.
The report finds that non-competes between employers and employees are prevalent across the UK economy. A post-term non-compete prevents workers from working for a rival employer for a specific amount of time and in one worker-level survey, roughly 30% of workers claim to have a non-compete agreement. This rises to over 40% in ICT and professional and scientific services. The report finds that many UK firms use employment clauses such as non-competes that restrict worker mobility. Non-competes are typically justified on the basis of enabling investment in workers through training or sharing confidential information, that employers may be less willing to do if employees might take these skills straight to a competitor. On this basis we might expect to only see non-competes only in particular industries or groups of workers where this sort of investment is happening. However, we have found that they are prevalent and found across the economy in all industries and across the whole income distribution. This includes some of the lowest-paid workers, who might be less likely to receive this sort of training or investment.
The gig economy has grown and is around 5% of employment
Turning to some recent developments in labour markets, the report looks at the rise of the gig economy (by which we mean labour services contracted through digital platforms). This has increased in importance over the last few years in the UK, but still hovers at only around 3-5% of total employment, according to the most robust data sources. This compares against 10-12% of workers in wider casual work. Workers in the gig economy often supplement their income with other jobs and sources of income. They also move more across the income distribution, in both directions, than workers in traditional employment. Overall, low-pay jobs are common in the gig economy, with 8% of workers over 25 earning below or at the minimum wage, compared to 5 percent in the traditional economy.
What about hybrid working?
Another recent development is the rise in hybrid working. The number of job postings that offer remote and hybrid working has increased dramatically since the pandemic and appears to have stabilised at about 20%. Hybrid jobs are more common in areas with low labour market concentration and associated with a rise in earnings. Workers value the benefits of hybrid working. By changing the geography of labour markets, hybrid working has the potential to affect employer-employee power.
Is there a link between product market power and labour market power?
The report explores whether the same firms which have market power in product markets also have market power in labour markets. If a firm is in a dominant position, it’s likely to be able to affect the price of its inputs (which include the price it pays for its workforce) as well setting the price of its outputs. The report uses different methodologies to assess this link. We cannot say for certain that firms which have product market power will also have labour market power, without understanding a firm’s market power for all other relevant inputs (with labour being one of these). The nature of this relationship depends on assumptions we make around the competitiveness of other input markets. Together with the academic community, the Microeconomics Unit will continue work to build this knowledge.
The implications of the report, both for the CMA and others
I’d now like to explore some of the implications of the key findings from the report, firstly for policy and academic communities more broadly, and then for the CMA. In that context, it’s worth restating that the purpose of the Microeconomics Unit’s work is somewhat different from most of the work of the CMA.
As I outlined earlier, the Microeconomics Unit is a centre of economic research excellence housed within the CMA. Some of its findings may be of direct relevance to the CMA’s work and some will be of more relevance to wider government and policy thinking. This report is primarily a contribution to the wider policy debate and the evidence base regarding labour markets, and we hope that the research and policy community will use the findings to inform and advance their own work. The paper provides evidence to support policymaking and further academic research in multiple areas. Let me give you just 4 examples, though I am sure many more will emerge.
Well-functioning labour markets mean stronger productivity and higher wages
We know that governments around the world are thinking about how to solve the productivity puzzle. Labour market frictions are certainly partly to blame. There is ample evidence to show that productivity growth and wage growth go hand in hand. The report provides some encouraging insights in this regard. It shows that, at an aggregate level in the UK, labour market concentration is roughly the same now as it was in 2002, and workers’ ‘cut’ of their contribution to value creation, how much goes to them against how much goes to their employer, has risen slightly since 2008.
But the report also shows that labour markets where there is employer market power can see lower wages for workers. Employer market power may also lead to a less efficient economy just as product market power does. The report shows employment rates are lower in more concentrated markets, which could suggest that firms are not hiring enough workers to be as efficient as possible. And while overall concentration levels have fallen slightly, there are some variations across regions and between sectors.
Collective bargaining counteracts the negative effect on wages of market power, but only in concentrated industries
Collective bargaining agreements are shown in the research to push against the downward pressure on wages caused by employer market power in concentrated labour markets. But that does not necessarily mean workers need to join a union to receive higher wages. Most labour markets in the UK are not highly concentrated, so workers are unlikely to be facing employer market power. It’s only in the most concentrated labour markets where collective bargaining can counteract the negative effect of employer market power on wages. And it’s worth re-emphasising that the analysis we have reviewed only relates to the private sector.
Labour markets are continuing to evolve, with the rise of hybrid working and the gig economy
The rise of the gig economy and, more recently, the unprecedented shift brought about by hybrid working, are changing the way we work and transforming labour markets as a whole. Gig economy workers are working across multiple platforms to boost their wages. Whilst evidence from this report shows that these jobs only make up around 3-5% of the labour market, there is potential for this to grow significantly. The report’s survey data also shows the value that many workers attach to hybrid working.
Law and policy on non-compete clauses may need updating
As the labour market evolves, it is possible that employment law too may need updating. The findings in this report show that non-compete clauses are prevalent across the economy, even in low-paying sectors. There is often a legitimate reason for these restrictions in improving firms’ incentives to invest in training for their workforce. However, the evidence in this paper is that the prevalence of non-competes across industries and income levels suggests that not all non-competes protect substantial training or client relationship investments. The widespread prevalence of non-competes across the economy could act as a barrier to job switching. Employment law and policy is obviously a complex and multi-faceted area that goes well beyond the remit of competition authorities, but we hope the evidence in the report helps inform policy makers as they consider the many different factors they are balancing. We note that the UK government has announced that they will legislate to limit post-term non-competes to 3 months in Great Britain. Our evidence supports this direction of travel.
What does this report mean for the CMA’s work?
I’d now like to turn to the CMA’s current and potential future work in labour markets.
We are using our competition law enforcement powers in labour markets
Well-functioning labour markets are of benefit to workers, businesses and the economy more generally. As a competition authority, we are therefore concerned if businesses collude to restrict competition between them in labour markets. And in our most recent annual plans, we have made it a priority to tackle such conduct, reflecting that, particularly during a cost of living crisis, it is incumbent on us to seek to address unlawful behaviour which can directly impact household budgets.
Using our Competition Act powers (the same powers we use to address cartels in product markets), we have opened 2 investigations into suspected anti-competitive conduct relating to rates for workers in the sports and non-sports TV production and broadcasting sectors, and have very recently broadened one of our other existing cartel investigations to cover suspected unlawful no-poaching arrangements in the consumer fragrances industry. And, whilst it would not be appropriate to comment on the specifics of these cases (where we have not yet reached any findings), they demonstrate our commitment to tackling suspected illegal collusion in labour markets.
And firms must comply with competition law
More broadly, last year we published advice to employers on how to avoid competition law breaches in relation to labour markets, making clear that just as we would expect them not to engage in price fixing, bid rigging or information exchange in product markets, so we would expect the same in labour markets. Our advice focused on 3 labour market distortions which may be in breach of the Competition Act: wage fixing, no-poaching and information sharing.
Employee non-competes and collective bargaining are outside the scope of competition law enforced by the CMA
While we are clear in our intent to take action against illegal labour market cartels, this does not mean the CMA will step in to use its competition enforcement powers in every case involving labour relations. For example, more often than not, the competition rules do not have a role to play in regulating the contract of employment between an employee and their employer – so that the type of non-competes that are discussed in our report would not generally infringe competition law and so not fall to be considered by the CMA. Likewise, the competition rules have developed so that they do not impede traditional collective bargaining agreements between employers and employees, which serve to further constructive relationships and mutually beneficial solutions.
But, as the report confirms, we have seen significant shifts in labour markets and the nature of employment, with a rise in the prevalence of self-employed people, especially in the gig economy. These developments potentially give rise to new questions for competition authorities about how and when such activity is or should be caught by the competition rules (for example when a group of self-employed people come together in a form of collective action).
And the CMA will not prioritise cases where self-employed workers and companies using their services reach a genuine collective bargain
We have been reflecting on how we approach and prioritise our work in labour markets, including where there may be a lack of clarity about how the rules apply to a particular category of worker. We will continue to do this as we gain more experience with labour markets, but in the meantime we want to provide reassurance that like other competition authorities internationally, the CMA has no interest in interfering where self-employed workers and their employers come together to reach a genuine collective bargain and does not expect to stand in the way of such behaviour. Our focus will remain on deterring cartel conduct which by its nature is detrimental to well-functioning labour markets.
Conclusions
I am very pleased to have had this opportunity to share the findings of the Competition and market power in UK labour markets report with you. The depth, rigour and relevance of this research not only represents a significant contribution to this field of study, it also bodes extremely well for the potential of the Microeconomics Unit to enrich the CMA’s understanding of complex market dynamics across a range of areas. The CMA will be using the findings of this report to inform our own policy thinking, where labour markets are relevant to our remit. But, given the complexity of analysis and insights in this work, I hope and expect that others will also make use the report to develop their own thinking and research, so that the Microeconomics Unit’s valuable contribution informs wider UK policy development for a broad range of stakeholders. For those interested in reading more, you can find copies of the report available online and I hope many of you will engage further with the Microeconomics Unit and others at the CMA. Your views and expertise are so helpful in advancing our work and your engagement is always valued. I hope that you have found this session useful, and I would welcome any questions.
The CMA would like to thank colleagues in the Microeconomics Unit for producing this report, in particular Jakob Schneebacher, Joel Kariel, Ana Rincon Aznar, Nicole Scholz, Max Read and Renisha Rana.