The world today – remarks by Andrew Bailey

Speech

It is a great pleasure to be here in AlUla and share thoughts at this important conference. You have given me the task of setting the scene in terms of the current context of the world economy. In other words, you have unfurled a large canvas, and given me 15 minutes to cover it with paint. Here goes!

I will start by drawing out the key points from the latest update of the IMF World Economic Outlook. The good news is that the world economy has been remarkably resilient in the face of much higher policy uncertainty. Although this uncertainty, including the impact of tariffs, has weighed on the level of activity, and accepting that there is varying momentum of economic activity across countries and sectors, the world economy has shown an impressive ability to adapt to the shifting landscape. Inflation has not risen markedly in the last year, though the cost of living (which is an issue of price levels relative to income levels) remains an important concern in quite a few countries.

Alongside this resilience of the world economy, global financial conditions have been accommodative, despite episodes of volatility and rising sovereign yields. An important part of this story has been equity valuations in the technology sector, and particularly in the AI part.

Overall, market conditions could have been much worse given the backdrop. That they have not been so reflects I think a number of factors at work. First, markets have become cautious in their reactions since not all of the initial announcements of policy shifts have been followed through to the word, and on occasions the impact of the announcement on economies and financial markets has not been as initially predicted. Second, markets are cautious to trade geopolitical risk when some of the traditional safe haven assets are close to the epicentre of the risks themselves and exhibit close correlations to risky assets, thus negating the safe-haven protection. Third, we have seen evidence of fear of missing out, backed by arguments along the lines of this time is different, for instance because of the expected productivity benefits of AI. The net result is a risk of some complacency in financial markets.

The IMF caution in their update that risks to the world economic outlook are tilted to the downside. Four reasons for this can be drawn out. First, there could be a significant escalation of geopolitical tensions. Second, and closely aligned, there could be further disruption to the fragile balance of trade policy. Third, fiscal vulnerabilities could emerge against a context of elevated public debt levels. And fourth, expectations of AI driven productivity gains could be disappointed.

I think this summary from the latest WEO fairly describes the current state of the world economy and the risks. Let me now move on to describe the more structural economic backdrop that conditions both the current situation and where it goes from here. I am going to cover five broad areas, so necessarily it will be brief.

The first is the change in the nature of the economic shocks that we have been facing. These have been larger supply side shocks, going back at least to the global financial crisis and then more recently Covid, the impact of Russia’s appalling invasion of Ukraine, and tariffs.

These have been much larger shocks than those that were seen in the preceding period of time. And, they have been supply-side shocks. On the whole, our macroeconomic frameworks are less well equipped to deal with supply-side than demand-side shocks.

The second broad area of backdrop that I will set out concerns the deeper structural parameters of many of the advanced economies. Over the last fifteen years, the potential growth rate of our economies has declined. For the UK, as an illustration, the decline has been from an average of around 2½ % p.a. over the twenty years before that to around 1½% in the last fifteen years. The largest contribution to that decline has come from productivity growth.

Productivity growth has had a pattern of long cycles since the Industrial Revolution. To borrow from the economist Joseph Schumpeter’s phraseology, industrial development involves change that occurs in “discrete rushes” but “separated by spans of comparative quiet”. The key idea here is that innovation and diffusion are at the heart of the growth process known as creative destruction. Cumulative innovation matters, as do clear property rights, and there is a positive role for public policy and institutions to support innovation.

The destruction point is that new innovation makes former innovation obsolete. A key here is the nature of the innovation which comes in rushes – so-called General Purpose Technology. The essence of GPTs – think steam engines, electricity, ICT/the internet – is that they enable innovation very broadly across our economies.

However, there have also been longish periods between waves of innovation when growth has been slower – the late 19^th century in the UK was such a period. I think for the last fifteen years we have been in such a slower phase, as the growth effects of ICT and the internet matured.

The third broad area of structural economic backdrop comes from the common feature of advanced economies, and some others too, of the average ageing of the population and the falling replacement rate. This creates lower economic growth by reducing labour supply and putting more pressure on fiscal positions. I would add that while the economics of ageing populations has been an issue much discussed and assessed in academic and policy circles, I am not persuaded that the significance of it is properly understood in the wider debate.

The fourth broad area concerns trade and global imbalances. Before Schumpeter gave us the theory of creative destruction, the classical economists gave us the trade-based model of growth. Adam Smith set out how trade facilitated the division of labour which became a basis for supporting technological innovation and growth. A reversal of trade openness has negative growth effects. And those effects are likely to be larger for more open economies, simply because the gains and losses from trade are larger for more open economies.

The fifth – and you will no doubt be pleased to hear last – broad area concerns the financial system. In this respect, I am speaking wearing both of my hats, as Governor of the Bank of England and as Chair of the Financial Stability Board. Over the period since the financial crisis, we have seen profound changes to the financial system, necessarily so given what happened then. The system has undoubtedly become more robust, and so has been able to absorb the big supply side shocks of recent years well. There has been a relative shift in the balance of financial intermediation from banks to non-banks. But the banking system remains a crucial source of credit to support real economic activity and the crucial source of liquidity and funding, including to the non-bank financial world. Banks remain unique in the private sector as the holders of most of the stock of money in the system (the other part is with central banks). Alongside this, there have been profound changes in core government debt markets, the rise of so-called private asset markets, and innovations which seek to broaden the scope of private sector money. These are big changes.

Having covered the canvas with paint, I am going to use the rest of my time to look a little more into the future, focusing on two of the areas I have painted – productivity and imbalances.

I will start with productivity. I mentioned that the creation part of innovation has in the past been associated with General Purpose Technology. The obvious question then is what comes next? What is the next GPT, and when will it arrive on the scene? The best guess is AI and robotics (both separately and in combination).

I am an optimist on the potential for AI and robotics to move the dial on productivity, and thus economic growth. But I like to think I am a realistic optimist. My impression is that we have made more progress so far applying AI to well-defined task-based work, rather than more ambitious goals, which I don’t find surprising^{footnote [1]}.

Also, growth via innovation and productivity enhancement takes time, it isn’t a quick fix. This is a lesson of economic history, which we can see, for instance, in the introduction of steam engines, electricity, and more recently ICT.

An important question is how will AI and robotics influence the labour market and jobs? Recent work by my colleague Edward Egan has used four channels through which the effects may be seen:

• Productivity augmentation, increasing productivity by automating repetitive tasks, freeing labour up for other higher-value activities. If firms use these gains to expand production, this can increase the demand for labour in non-automated tasks.
• Displacement automation which will reduce the demand for labour in certain jobs
• Reinstatement via new tasks, where as seen in the past technology innovation creates new tasks that could not have been imagined before.
• Compositional reallocation, such that even if aggregate employment doesn’t change much, AI is likely to reallocate jobs between sectors. Some industries might shrink, others grow, and affected workers will need to retrain to adapt their skills.

Clearly, the overall effect on employment will depend on the mix of these channels, which is as yet highly uncertain.

We may see displacement indirectly. For example, in the UK, in the last three years new online vacancies in the most AI-exposed roles have decreased by more than twice as much as in the least exposed group. But, on the positive side, there has been an observed significant increase in new tasks such as integrating AI tools into firms workflow processes^{footnote [2]}.

Two points to conclude on productivity and AI. First, education and training in AI skills will be critical. Second, we shouldn’t resort to oversimplified conclusions on the employment effects.

I will finish on imbalances in the world economy. In the period since the financial crisis, the headwinds to growth have made it harder to achieve domestic consensus to support international co-operation and openness. While it is true that openness supports growth and has reduced global poverty, it has had distributional consequences in economies, and there has been an undermining of domestic cohesion in many countries, which has created opposition to economic openness.

The effectiveness of the international financial systems depends on national support and license – it cannot operate in isolation. The goals of international co-operation must sit alongside domestic national policy objectives, but there also must be scope for the international goals to shape those domestic objectives. It cannot be a one-way street and this principle must apply to all participants. It follows that there is a natural tension between economic globalisation and domestic objectives and that we must robustly define and tackle excessive imbalances.

The international financial system must be robust to many states of the world. This requires a considerable degree of flexibility in the design and operation of the system. Three lessons from history stand out for me here.

First, in the current context of slower growth, we must focus on what is needed to raise potential growth rates, and thus on the role of innovation including obviously AI and robotics. But we must not forget the contribution that economic openness will make to that growth, the important lesson from Adam Smith.

Second, today we continue to face the challenge of adjusting the system to a more
multi-polar world, and where the shifting of the poles reveals tensions. One of the lessons of economic history is that such shifts in polarity inevitably strain the operation of the system. We have to be prepared for this.

But, my third lesson is that the record since Bretton Woods suggests that the International Financial Institutions – the IMF, the World Bank et al – have been pretty good at institutional recalibration.

One final point here. The question has been raised of whether the future lies more in so-called variable geometry, partnerships of the willing and aligned if you like. There may well be a role in other fields of public policy – indeed to some degree there always has been. But, I would argue that this does not apply in the world of international finance where the benefits come precisely because national borders are not restrictive to activity. This is a strength of the system, but we have to balance this strength with effective tools that can assess and manage the resulting risks to monetary and financial stability. This is our job, and we must do it in a global institutional framework. For the Financial Stability Board, this reinforces the importance of international standards to support global financial stability and a level playing field on which firms can compete, all of this supported by active surveillance for emerging risks and vulnerabilities.

Let me end on a personal note. We, of course, need strong collective and individual leadership. An important part of that is the International Monetary and Financial Committee. The IMFC advises the IMF on the international monetary and financial system. We are fortunate to have Minister Aljadaan leading us as chair of the IMFC.

Thank you.

I would like to thank Sarah Breeden, Edward Egan, Karen Jude, Andrea Rosen and James Talbot for their comments and help in the preparation of these remarks.