Category: Economics 11e

Posted on

Will AI make the world less equal?

With businesses increasing their use of AI, this is likely to have significant effects on employment. But how will this affect the distribution of income, both within countries and between countries?

In some ways, AI is likely to increase inequality within countries as it displaces low-skilled workers and enhances the productivity of higher-skilled workers. In other ways, it could reduce inequality by allowing lower-skilled workers to increase their productivity, while displacing some higher-skilled workers and managers through the increased adoption of automated processes.

The effect of AI on the distribution of income between countries will depend crucially on its accessibility. If it is widely available to low-income countries, it could significantly enhance the productivity of small businesses and workers in such countries and help to reduce the income gap with the richer world. If the gains in such countries, however, are largely experienced by multinational companies, whether in mines and plantations, or in labour-intensive industries, such as garment production, few of the gains may accrue to workers and global inequality may increase.

Redistribution within a country

The deployment of AI may result in labour displacement. AI is likely to replace both manual and white-collar jobs that involve straightforward and repetitive tasks. These include: routine clerical work, such as data entry, filing and scheduling; paralegal work, contract drafting and legal research; consulting, business research and market analysis; accounting and bookkeeping; financial trading; proofreading, copy mark-up and translation; graphic design; machine operation; warehouse work, where AI-enabled warehouse robots do many receiving, sorting, stacking, retrieval, carrying and loading tasks (e.g. Amazon’s Sequoia robotic system); basic coding or document sifting; market research and advertising design; call-centre work, such as enquiry handling, sales, telemarketing and customer service; hospitality reception; sales cashiers in supermarkets and stores; analysis of health data and diagnosis. Such jobs can all be performed by AI assistants, AI assisted robots or chat bots.

Women are likely to be disproportionately affected because they perform a higher share of the administrative and service roles most exposed to AI.

Workers displaced by AI may find that they can find employment only in lower-paid jobs. Examples include direct customer-facing roles, such as bar staff, shop assistants, hairdressers and nail and beauty consultants.

Such job displacement by AI is likely to redistribute income from relatively low-skilled labour to capital: a redistribution from wages to profits. This will tend to lead to greater inequality.

AI is also likely to lead to a redistribution of income towards certain types of high-skilled labour that are difficult to replace with AI but which could be enhanced by it. Take the case of skilled traders, such as plumbers, electricians and carpenters. They might be able to use AI in their work to enhance their productivity, through diagnosis, planning, problem-solving, measurement, etc. but the AI would not displace them. Instead, it could increase their incomes by allowing them to do their work more efficiently or effectively and thus increase their output per hour and enhance their hourly reward. Another example is architecture, where AI can automate repetitive tasks and open up new design possibilities, allowing architects to focus on creativity, flexibility, aesthetics, empathy with clients and ethical decision-making.

An important distinction is between disembodied and embodied AI investment. Disembodied AI investment could include AI ‘assistants’, such as ChatGPT and other software that can be used in existing jobs to enhance productivity. Such investment can usually be rolled out relatively quickly. Although the extra productivity may allow some reduction in the number of workers, disembodied AI investment is likely to be less disruptive than embodied AI investment. The latter includes robotics and automation, where workers are replaced by machines. This would require more investment and may be slower to be adopted.

Then there are jobs that will be created by AI. These include prompt engineers, who develop questions and prompt techniques to optimise AI output; health tech experts, who help organisations implement new medical AI products; AI educators, who train people in the uses of AI in the workplace; ethics advisors, who help companies ensure that their uses of AI are aligned with their values, responsibilities and goals; and cybersecurity experts who put systems in place to prevent AI stealing sensitive information. Such jobs may be relatively highly paid.

In other cases, the gains from AI in employment are likely to accrue mainly to the consumer, with probably little change in the incomes of the workers themselves. This is particularly the case in parts of the public sector where wages/salaries are only very loosely related to productivity and where a large part of the work involves providing a personal service. For example, health professionals’ productivity could be enhanced by AI, which could allow faster and more accurate diagnosis, more efficient monitoring and greater accuracy in surgery. The main gainers would be the patients, with probably little change in the incomes of the health professionals themselves. Teachers’ productivity could be improved by allowing more rapid and efficient marking, preparation of materials and record keeping, allowing more time to be spent with students. Again, the main gainers would be the students, with little change in teachers’ incomes. Other jobs in this category include social workers, therapists, solicitors and barristers, HR specialists, senior managers and musicians.

Thus there is likely to be a distribution away from lower-skilled workers to both capital and higher-skilled workers who can use AI, to people who work in new jobs created by AI and to the consumers of certain services.

AI will accelerate productivity growth and, with it, GDP growth, but will probably displace workers faster than new roles emerge. This is likely to increase inequality and be a major challenge for society. Can the labour market adapt? Could the effects be modified if people moved to a four- or three-day week? Will governments introduce statutory limits to weekly working hours? Will training and education adapt to the new demands of employers?

Redistribution between countries

AI threatens to widen the global rich–poor divide. It will give wealthier nations a productivity and innovation edge, which could displace low-skilled jobs in low-income nations. Labour-intensive production could be replaced by automated production, with the capital owned by the multinational companies of just a few countries, such as the USA and China, which between them account for 40% of global corporate AI R&D spending. For some companies, it would make sense to relocate production to rich countries, or certain wealthier developing countries, with better digital infrastructure, advanced data systems and more reliable power supply.

For other companies, however, production might still be based in low-income countries to take advantage of low-cost local materials. But there would still be a redistribution from wages in such countries to the profits of multinationals.

But it is not just in manufacturing where low-income countries are vulnerable to the integration of AI. Several countries, such as India, the Philippines, Mexico and Egypt have seen considerable investment in call centres and IT services for business process outsourcing and customer services. AI now poses a threat to employment in this industry as it has the potential to replace large numbers of workers.

AI-related job losses could exacerbate unemployment and deepen poverty in poorer countries, which, with limited resources, limited training and underdeveloped social protection systems, are less equipped to absorb economic and social shocks. This will further widen the global divide. In the case of embodied AI investment, it may only be possible in low-income countries through multinational investment and could displace many traditional jobs, with much of the benefit going in additional multinational profit.

But it is not all bad news for low-income countries. AI-driven innovations in healthcare, education, and agriculture, if adopted in poor countries, can make a significant contribution to raising living standards and can slow, or even reverse, the widening gap between rich and poor nations. Some of the greatest potential is in small-scale agriculture. Smallholders can boost crop yields though precision farming powered by AI; AI tools can help farmers buy seeds, fertilisers and animals and sell their produce at optimum times and prices; AI-enabled education tools can help farmers learn new techniques.

Articles

Questions

  1. What types of job are most vulnerable to AI?
  2. How will AI change the comparative advantage of low-income countries and what effect will it be likely to have on the pattern of global trade?
  3. Assess alternative policies that governments in high-income countries can adopt to offset the growth in inequality caused by the increasing use of AI.
  4. What policies can governments in low-income countries or aid agencies adopt to offset the growth in inequality within low-income countries and between high- and low-income countries?
  5. How might the growth of AI affect your own approach to career development?
  6. Is AI likely to increase or decrease economic power? Explain.
Posted on

The 4Ps approach to merger policy in the UK

The approach towards mergers remains the most controversial area of competition policy. Some argue that policy makers in both the UK and EU have been too easily persuaded by the arguments put forward by firms and so have allowed too many mergers to proceed. Others claim that the opposite is true and that merger policy has prohibited mergers that should have been allowed to proceed. This, then, has a negative impact on investment, innovation, productivity and growth.

In recent years there has been more specific criticism of merger policy in the UK. The government has indicated that it wants the Competition and Markets Authority (CMA) to be less interventionist and take a more pro-growth approach.

In February 2025, in response to this criticism, the CMA launched its new ‘4 Ps’ approach to merger policy: Pace, Predictability, Proportionality and Process. Various changes to the investigation process have been proposed in the past 12 months using this framework.

Pace. The time taken by the CMA to initially assess a merger before deciding whether a Phase 1 investigation is necessary (i.e. the pre-notification procedure) was reduced from 65 to 40 working days. Also, the target to complete straightforward Phase 1 investigations was reduced from 35 to 25 days.

Predictability. The proposed merger guidelines, published in October 2025, provide more detail on (a) what criteria will be used to measure market shares when applying the ‘share of supply test’ (this is where the combined UK market share of two merging businesses is at least 25%, provided one business has a UK turnover of at least £10 million), and (b) the factors that are likely to lead to the competition authorities concluding that one business has gained ‘material influence over another’. Businesses had complained that there was too much uncertainty about the way the share of supply test and material influence were applied. The CMA is also considering greater alignment with other international regulators over decision making rather than its previous policy of acting independently. All these measures should increase the predictability of the investigation process.

Proportionality. Proportionality refers to the objective of addressing any competition issues in merger cases in a way that places the minimum burden on the businesses involved. To improve proportionality, the CMA has indicated that in future cases it will be more willing to use behavioural remedies – requiring firms to take or desist from certain actions. New draft guidelines identify more situations where the use of behavioural remedies may be appropriate. However, they also show that the CMA still views structural remedies (e.g. preventing the merger or requiring firms to demerge or to sell certain assets) as more effective in many situations. Another important measure to improve proportionality is the introduction of a new ‘wait and see’ approach to global mergers. The CMA will now wait to see if the actions taken by other competition authorities in global cases address any concerns in the UK market before deciding whether to launch a review.

Process. To improve the process, the CMA has announced plans to engage with businesses at a much earlier point in the process. For example, it has pledged to share its provisional thinking in the early stages of an investigation by implementing new ‘teach-in’ sessions and having more regular update meetings. Much earlier meetings that focus on possible remedies will also take place. This may make it possible for the CMA to assess the suitability of more complex remedies during a Phase 1 investigation rather than having to wait for a longer and more costly Phase 2 review. Phase 2 reviews will also no longer be managed by panels of independent experts. This role will now be carried out by the internal CMA board.

Some critics argue that the CMA has not fully considered the potential benefits of mergers in many cases. For example, a merger could (a) have procompetitive effects, known as rivalry enhancing efficiencies (REEs) and/or (b) benefits for consumers outside of the relevant market, known as relevant customer benefits (RCBs). In response to this criticism, the CMA is currently reassessing its approach to including evidence on REEs and RCBs.

The CMA is still currently consulting with interested parties about many of these proposed changes. It will be interesting to see what final decisions are made in the next couple of years.

Articles

CMA documentation

Questions

  1. Of all the mergers considered by the CMA in 2024/25, find out what percentage were formally investigated. How many were blocked from taking place? Do you believe that this indicates that merger policy is too weak or too strong?
  2. What three criteria must be met for a business arrangement to be classed as a ‘relevant merger situation’ by the CMA?
  3. Identify some different methods that one business could use to gain material influence over the way another company operates.
  4. Outline the ‘turnover test’, the ‘share of supply test’ and the ‘hybrid test’.
  5. Discuss the potential advantages of using behavioural remedies as opposed to structural remedies in merger cases. Why has the CMA still preferred the use of structural remedies in most situations?
Posted on

The second-hand market and GDP

This Christmas, more people are considering giving second-hand (or ‘pre-loved’) goods as presents. This allows them to afford better-quality presents and to save money at a time when a large proportion of the population are finding that their finances are stretched. This continues a trend towards buying second-hand products – a trend driven by the rise of various online retailers, such as Vinted and Preloved, and a growing online presence of charity shops, as well as extensive use of established platforms, such as Facebook Marketplace, eBay, Depop, Gumtree and Nextdoor.

Clearly, people gain from buying and selling second-hand items – part of the ‘circular economy’. But what are the implications for gross domestic product (GDP)? After all, GDP is one of the main indicators of the size of an economy, and growth in GDP is probably the most widely-used measure of economic progress. Are second-hand transactions captured in GDP?

If you directly sell your own second-hand items, this does not count towards GDP. There is no new product being made. The items are only counted when they are first produced. Any service you provide to the purchaser (and to yourself) is in a similar category to housework, childcare, DIY and other services that people provide to themselves, household members and friends. But like such services, there is a strong argument that they should be.

Likewise, the environmental benefits (positive externalities) of recycling products, rather than throwing them away or hoarding them, are not counted. In fact, if reusing products causes fewer new products to be made, this would be counted as subtracting from GDP.

If, however, you set up a business by buying and selling second-hand items, the service you provide would contribute towards GDP. What would be counted would the value added to the product – captured through the difference in the purchase and selling prices. In fact, HMRC has warned people that buying and selling second-hand items is taxable, as it counts as self-employment for tax purposes. But it is only this value added that counts. If you buy an item on Vinted, only the value added by Vinted counts towards GDP.

As no production takes place, the purchase of second-hand items adds either nothing to GDP or just the service of a retailer. It is effectively just a transfer of goods and money. If buying second-hand items means that you buy fewer new ones, then that would cause GDP to fall if the response of firms is to produce fewer newer items. However, the person selling the second-hand items will gain revenue, which could be used to buy new items. If that increased production, that would boost GDP. The net effect on GDP of this transfer of goods and money in the second-hand market will be pretty small.

Yet, clearly, the second-hand market provides a welfare gain to both sellers and purchasers – a gain that is likely to grow as the use of second-hand markets increases. At Christmas time, it provides a timely warning of the limitations of using GDP to measure wellbeing.

Articles

Questions

  1. What other items or activities affecting human wellbeing are not counted in GDP?
  2. Name some goods and services that are produced, and hence are included in GDP, but which can be classed as ‘bads’.
  3. For what reasons might a country have a high GDP per capita but a poor average level of wellbeing?
  4. How might GDP figures be adjusted for international comparison purposes?
  5. Would it be possible to adjust GDP figures to take account of externalities in production (negative and positive)? If so, how?
  6. Production involves human costs. To what extent does GDP take this into account?
  7. What is meant by the circular economy? How might you have a ‘circular’ Christmas?
Posted on

AI stock: still under priced or a speculative bubble?

The share prices of various AI-related companies have soared in this past year. Recently, however, they have fallen – in some cases dramatically. Is this a classic case of a bubble that is bursting, or at least deflating?

Take the case of NVIDIA, the world’s most valuable company, with a market capitalisation of around $4.2 trillion (at current share prices). It designs and produces graphics cards and is a major player in AI. From a low of $86.62 April this year, its share price rose to a peak of $212.19 on 29 October. But then began falling as talk grew of an AI bubble. Despite news on 19 November that its 2025 Q3 earnings were up 62% to $57.0bn, beating estimates by 4%, its share price, after a temporary rise, began falling again. By 21 November, it was trading at around $180.

Other AI-related stocks have seen much bigger rises and falls. One of the biggest requirements for an AI revolution is data processing, which uses huge amounts of electricity. Massive data centres are being set up around the world. Several AI-related companies have been building such data centres. Some were initially focused largely on ‘mining’ bitcoin and other cryptocurrencies (see the blog, Trump and the market for crypto). But many are now changing focus to providing processing power for AI.

Take the case of the Canadian company, Bitfarms Ltd. As it says on its site: ‘With access to multiple energy sources and strategic locations, our U.S. data centers support both mining and high-performance computing growth opportunities’. Bitfarms’ share price was around CAD1.78 in early August this year. By 15 October, it had reached CAD9.27 – a 421% increase. It then began falling and by 24 November was CAD3.42 – a decline of over 63%.

Data centres do have huge profit potential as the demand for AI increases. Many analysts are arguing that the current share price of data centres undervalues their potential. But current profits of such companies are still relatively low, or they are currently loss making. This then raises the question of how much the demand for shares, and hence their price, depends on current profits or future potential. And a lot here depends on sentiment.

If people are optimistic, they will buy and this will lead to speculation that drives up the share price. If sentiment then turns and people believe that the share price is overvalued, with future profits too uncertain or less than previously thought, or if they simply believe that the share price has overshot the value that reflects a realistic profit potential, they will sell and this will lead to speculation that drives down the share price

The dot.com bubble of the late 1990s/early 2000s is a case in point. There was a stock market bubble from roughly 1995 to 2001, where speculative investment in internet-based companies caused their stock values to surge, peaking in late 1999/early 2000. There was then a dramatic crash. But then years later, many of these companies’ share prices had risen well above their peak in 2000.

Take the case of Amazon. In June 1997, its share price was $0.08. By mid-December 1999, it had reached $5.65. It then fell, bottoming out at $0.30 in September 2001. The dot-com bubble had burst.

But the potential foreseen in many of these new internet companies was not wrong. After 2001, Amazon’s share price began rising once more. Today, Amazon’s shares are trading at over $200 – the precise value again being driven largely by the company’s performance and potential and by sentiment.

So is the boom in AI-related stock a bubble? Given that the demand for AI is likely to continue growing rapidly, it is likely that the share price of companies providing components and infrastructure for AI is likely to continue growing in the long term. But just how far their share prices will fall in the short term is hard to call. Sentiment is a fickle thing.

Articles

Questions

  1. Using a supply and demand diagram, illustrate how speculation can drive up the share price of a company and then result in it falling.
  2. What is meant by overshooting in a market? What is the role of speculation in this process?
  3. Does a rapid rise in the price of an asset always indicate a bubble? Explain.
  4. What are the arguments for suggesting that markets are/are not experiencing an AI share price bubble? Does it depend of what part of the AI market is being considered?
  5. What is meant by the market capitalisation of a company? Is it a good basis for deciding whether or not a company’s share price is a true reflection of the company’s worth? What other information would you require?
  6. Find out what has been happening to the price of Bitcoin. What factors determine the price of Bitcoin? Do these factors make the price inherently unstable?
Posted on

Age-old problems but new technology: developments in collusion in the algorithmic era

In my previous blog post on this site, I examined how AI-powered pricing tools can act as a ‘double-edged sword’: offering efficiency gains, while also creating opportunities for collusion. I referred to one of the early examples of this, which was the case involving Trod Ltd and GB Eye, where two online poster and frame sellers on Amazon used pricing algorithms to monitor and adjust their prices. However, in this instance there was also an explicit agreement between the firms. As some commentators put it, it was ‘old wine in new bottles‘, meaning a fairly conventional cartel that was simply facilitated through digital tools.

Since then, algorithms have increasingly become part of everyday life and are now embedded in routine business practice.

Some of the effects may have a positive effect on competition. For example, algorithms can help to lower barriers to entry. In some markets, incumbents benefit from long-standing experience, while new firms face significant learning costs and are at a disadvantage. By reducing these learning costs and supporting entry, algorithms could contribute to making collusion harder to sustain.

On the other hand, algorithms could increase the likelihood of collusion. For example, individual algorithms used by competing firms may respond to market conditions in predictable ways, making it easier for firms to collude tacitly over time.

Algorithms can also improve the ability of firms to monitor each other’s prices. This is particularly relevant for multi-product firms. Traditionally, we might expect these markets to be less prone to collusion because co-ordinating across many products is complex. AI can overcome this complexity. In the Sainsbury’s/Asda merger case, for example, the Competition and Markets Authority suggested that the main barrier to reaching and monitoring a pricing agreement was the complexity of pricing across such a wide range of products. However, the CMA also suggested that technological advances could increase its ability to do so in the future.

The ‘hub-and-spoke’ model

One of the other growing concerns is the ability of AI pricing algorithms to facilitate collusion by acting as a ‘hub’ in a ‘hub-and-spoke’ arrangement. In this type of collusion, competing firms (the ‘spokes’) need not communicate directly with one another. Instead, the ‘hub’ helps them to co-ordinate their actions.

While there have been only limited examples of an AI pricing algorithm acting as a hub in practice, what once seemed to be a largely theoretical concern has now become a live enforcement issue.

A very recent example is the RealPage case in the United States. The Department of Justice (DOJ) filed an antitrust lawsuit against RealPage Inc. in August 2024, alleging that RealPage, acting as the ‘hub’, facilitated collusion between landlords (the ‘spokes’).

RealPage provided pricing software to numerous landlords, including the largest landlord in the USA, which manages around 950 000 rental units across the country. These landlords would normally compete independently in setting rental prices, discounts and lease terms to win consumers. However, by feeding competitively sensitive information that would not usually be shared between rivals into RealPage’s system, the software generated pricing recommendations that, according to the DOJ, led to co-ordinated rent increases across competing apartment complexes.

In the RealPage case, the authorities reported that they had access to internal documents and statements from the parties involved, which helped support their allegations. These included references within RealPage to helping landlords ‘avoid the race to the bottom’ and comments from a landlord describing the software as ‘classic price fixing’.

Evidence in these cases really matters because the standard of proof required to establish a hub-and-spoke arrangement is much higher than for traditional cases of explicit collusion. This is because it can be difficult to distinguish between legitimate and anti-competitive communication between retailers and suppliers. Also, proving ‘anti-competitive intent’ is inherently challenging.

Other competition authorities around the world are also turning their attention to these issues. For example, the European Commission recently announced that a number of investigations into algorithmic pricing are underway, signalling a clear shift toward more active scrutiny. As technology continues to advance, it is clear that algorithmic pricing will remain an area where both firms and authorities must move and adapt quickly.

Articles

Questions

  1. In what ways does the RealPage case differ from the earlier Trod Ltd and GB Eye Ltd case? Consider the roles played by the firms, the nature of the alleged co-ordination, and the extent to which pricing algorithms were used to facilitate the conduct.
  2. How might the use of pricing algorithms affect the likelihood of firms colluding, either explicitly or tacitly? Consider ways that algorithms may make collusion easier to sustain but also ways in which they may reduce its likelihood.
  3. Should firms be held liable for anti-competitive outcomes produced by algorithms that ‘self-learn’, even if they did not intend those outcomes? Explain why or why not.