The first section explains some basic ideas, and those familiar with arbitrage and the natural real rate of interest can skip to the austerity section.
Basics
Longer term interest rates, rates on borrowing for a number of years, are what helps determine the cost of government debt, the cost of long term borrowing by firms and the cost of mortgages. A lot of popular discussion links these longer term rates to the short term interest rate set by central banks. The reason is arbitrage. Someone who buys a longer term asset with a fixed return over five years, say, could instead hold that money in a variable rate account, and arbitrage means that rates will move until lenders are indifferent between the two, so longer term rates and short rates are linked.
By that logic, interest rates on a 5 year bond, say, are just equal to expectations about how the central bank will set short term interest rates over the next five years. It’s not that simple because of liquidity and uncertainty. Having money in a variable rate account normally gives you instant access, while investing in a longer term asset will either tie your money up for a long time or to a potential loss by selling it before term, so lenders want some compensation for this. In addition, what is going to happen to short rates is very uncertain, so most lenders want compensation for taking the risk of betting on a particular forecast.
Guessing how independent central banks will set interest rates in the future is not easy, but the logic behind their actions are pretty clear. Most aim to hit an inflation target, so they will set interest rates at whatever level is required to either keep inflation at target, or get inflation back to target. In crude terms, interest rates act to control aggregate demand, and keeping inflation steady involves getting demand matched up with available supply.
A critical question is therefore what level of interest rates will stabilise inflation in a few years time. That depends on what will happen to aggregate demand, but it also depends on how demand is influenced by interest rates. Demand is mainly determined by real rather than nominal interest rates, where the real rate is the nominal rate less expected inflation, so we want to know what real interest rate will match aggregate demand to supply. That level is often called the natural real interest rate. Judging by historical data the natural real interest rate seems to vary a lot over time. [1]
Most measures of real interest rates fell steadily from 1980 until at least 2000. Above is an example based on US rates (source). While you could argue that rates were unusually high in the 1980s because central banks were squeezing demand to bring inflation down, that doesn’t explain the continuing fall in the 1990s.
We need two more conceptual ideas before going further. The first is the importance of the supply and demand for savings. If, say, a technological revolution produces a large increase in borrowing by firms, but there is no increase in the supply of savings to meet it, real interest rates will rise to encourage more saving and discourage some borrowing. While I tend to discount the idea that the supply of particular assets, like an individual government’s debt, is the critical factor determining the interest rate on that debt [2], at the aggregate level the supply and demand for assets does matter.
The second idea is the distinction between safe assets, where the person borrowing is almost certain to pay back the loan, and other investments where the loan may be only partially paid back. Equities are the main example of the latter, because not only do aggregate share prices move around but any individual company may, in the extreme, go bust. The difference between the two types, returns on safe assets and the return on equity or capital more generally, is called the ‘equity risk premium’, but that gap may not just be about risk. In particular the overall supply and demand for either type of asset can influence this risk premium.
The Financial Crisis and Austerity Period
For more detail on this and the next section I strongly recommend this fascinating discussion between Paul Krugman and Ricardo Caballero. Caballero argues that after 2000 there was a continued fall in the real interest rate on safe assets, much greater than anything happening to the return on capital, and he described this (at the time) as a shortage of safe assets. (This study suggests that the return on capital in the US actually started rising in the 90s.) In other words rather than there being too much government debt around (as many politicians and journalists claimed) there was actually too little.
Was this shortage one reason that the private sector tried to make its own safe assets with mortgage backed securities, the failure of which was the spark that led to the Global Financial Crisis (GFC)? Caballero suggests it was a factor. It would be horribly ironic if the GFC, the impact of which was often misattributed to excessive government borrowing, might itself have been partly caused by governments borrowing too little! But what Caballero’s argument certainly tells us is that attempts in most countries to cut government deficits from 2010 because government debt was dangerously high were especially wrong.
The original sin of austerity was of course to cut government spending in a recession, going against basic macroeconomic teaching understood since Keynes. We now know that policy, as well as delaying and probably diminishing the recovery from the financial crisis, also probably failed to even reduce the government debt to GDP ratio. The argument that economists like myself made was that fiscal consolidation, if required, should wait until the recovery from the recession was all but complete. However, if there was in fact a shortage of safe assets in the form of government debt before the financial crisis, it may well be that the large amount of debt issued as a result of that crisis was a necessary correction for that shortage. Even the excuse for austerity was wrong.
Because safe assets were in short supply, austerity meant that interest rates remained extremely low for a decade after 2010. Government debt was increasing, but this may have simply replaced the hole created by the demise of mortgage backed securities. (For a period a lot of European government debt also stopped being safe.) [3] What is unforgivable over that period is that governments kept public investment low, despite the fact that it was so cheap to borrow. [4] We are suffering the results of that colossal policy mistake in most countries today. The one country that didn’t make that mistake was China, which is the reason they have a state of the art railway network today.
Rising interest rates today
Those days of very low short and longer term interest rates seem to be over. The first point to note is that this is essentially an international phenomenon. Political journalists like to focus on domestic issues when talking about this, but you will not find an explanation for rising longer term interest rates across the globe in anything the UK government has or will do. [5].
One story that follows from the post-GFC discussion above is that we no longer have a shortage of safe assets, and indeed there may now be a glut. The Eurozone crisis ended as a result of the switch in policy by Dragi and the European Central Bank, so that supply is back. In addition the pandemic led to a step increase in the amount of government debt worldwide. Finally, in the US in particular, budget deficits have been large and (unlike in the UK) there are no plans to bring them down.
Another part of the explanation for higher longer term rates that is often discussed is inflationary pressure caused by the US attack on Iran. However what is noticeable about longer term interest rates is that they are higher at all durations. Why should higher oil prices today be adding to inflationary pressure in 20 years time? One answer to that is a point I have noted before, which is the rise of right wing populist and even fascist governments. [6]
Take the worst case scenario, which is that traditionally right wing parties are either being overshadowed by right wing populist parties or are morphing into such parties. (Of course we might get both, as we have in the UK.) That means that periods of right wing populist government become almost inevitable. It is unlikely but not impossible that such governments would actually default on their debt. What is more likely is that they would end, by one means or another, independent central banks and allow inflation to stay permanently and significantly above today’s target levels. For any given level of the real natural rate, that means higher nominal interest rates in the future.
A third factor that may generate more persistent inflationary pressure is the AI boom. Most people believe that the current AI revolution will greatly improve future productivity growth. [7] Prospective higher productivity in the medium term can lead to increases in aggregate demand in the shorter term through two routes. First firms will increase investment in AI. We are already seeing that in the US with the development of AI itself, but that will spread to investment in these models across many other sectors. Second, higher future productivity means higher expected future incomes, and people’s consumption today depends in part on expected future income. (In effect they will spend more today because of a strong stock market.) Just as with a future natural resource boom, demand will increase before supply does, putting additional pressure on inflation and therefore raising the natural rate of interest.
Finally the AI boom itself tends to increase the return on capital, which other things being equal will raise longer term interest rates across the board. Higher future productivity raises the long term natural rate of interest according to standard economic models. This is why we are seeing stock markets remaining strong despite rising interest rates. In addition the investment required to implement AI improvements requires additional corporate borrowing, pushing up longer term interest rates. [8] At the moment that borrowing is concentrated in the tech companies developing AI, but later it may spread more widely as firms adapt to integrate AI into their production methods.
All these stories, with the possible exception of higher oil prices, imply that higher longer term interest rates are here to stay, unless a financial crisis forces central banks to lower short term rates (see [6]). To the extent that this is the result of anticipated productivity gains, it can be viewed as a positive rather than negative development.
If that is the case, what should policymakers do about it? In terms of the distribution from borrowers to lenders not much directly. We may end up judging the era of cheap borrowing after the GFC, rather than conditions today, as an historical oddity. As far as fiscal policy is concerned, responsible governments will at the very least need to stop debt to GDP rising as a result of higher interest rates on government debt. As higher interest rates particularly benefit wealth holders, it makes sense to raise taxes on those who save rather than those who borrow. Cutting public spending makes much less sense as long as private consumption continues to increase..
[1] As aggregate demand varies around supply, then the natural real rate will be approximately the average of real interest rates over a number of years
[2] For the same reason that a sudden shortage of, say, VW cars is unlikely to lead to a large increase in their price.
[3] Another story to explain the low natural real rate of interest popular at the time was secular stagnation (low corporate borrowing because of limited investment opportunities) or a glut of savings associated with China in particular.
[4] Even Ken Rogoff, whose work helped justify austerity, said he didn’t agree with cuts to public investment, which in the UK were critical in stalling the economic recovery.
[5] For some time the rate on UK 10 year government debt has been at the top of the range among major economies. It’s hard to put much stress on domestic political/fiscal factors to explain this when the interest rate on 10 year government bonds in France is 1% below the UK. One possible explanation is the Bank of England unwinding Quantitative Easing. If you really want to relate relatively high rates to UK fiscal decisions, I think the best line to follow is the reduced scope of action created by both a (quite understandable) political unwillingness to cut spending and a commitment not to increase the major tax rates, which reduces the chance of fiscal policy helping to reduce demand thereby putting all the burden on interest rates.
[6] An additional longer term pressure on inflation is climate change. With significant global warming now inevitable, disruption to supplies of food in particular will become more frequent.
[7] Note that this doesn’t necessarily mean that the high equity price of the leading AI companies is justified. You can believe that AI will greatly improve future labour productivity but also believe that the companies producing the AI will find it difficult to capture much of that through monopoly profits. The likelihood of a future productivity boom and who gets the benefits of that boom are separable questions. One quite possible source of a financial crisis is a sudden realisation that AI will not lead to big increases in the profits of those doing the development of these models, making the debt of some of these companies vulnerable.
[8] As noted above, higher future income may encourage a reduction in current saving, whereas what may be required is a rise in current savings relative to income to finance additional investment.
