When I interview candidates – I like to ask them to talk me through how the central bank controls the economy using interest rates. A good candidate will normally answer by saying that when inflation and growth are too low, the central bank will lower interest rates. Lower interest rates mean that households consume more and save less, stimulating demand. Businesses will find that more potential projects have expected returns higher than the rate of interest on loans, and thus investment will pick up. My next question normally catches people pretty well off guard. What evidence do we have that this works? How do we know that households save less and consumer more in response to a fall in interest rates? It’s not an unreasonable question. If I were interviewing someone to be a physics teacher – I would expect them to know about Rutherford’s experiments to determine the nature of the atom. As I’m interviewing market participants who are responsible for setting interest rates, making predictions about interest rates, and understanding the nature of the relationship between interest rates and the macroeconomy – It’s reasonable to expect them to have some kind of grasp of the evidence, isn’t it?
Well actually no – it isn’t – for many reasons. The first is one of ontology. Economics, at least as stated by my interview question, is many layers of abstraction removed from what one could conceivably run lab style experiments on. The evidence will be less pithy than, say, Rutherford’s alpha particles being bounced off onto his detectors. But the second is that just amazingly little work has been done by professional economists to uncover just how this causal mechanism of lowering and raising interest rates actually feed through to the actions of households and businesses. This St Louis Fed paper from 2012 takes it as read that the effects of interest rates on economic actors are pretty much undetectable -and the paper it chooses to reference that stance, co-authored by a former Fed Chair no less, says the same in a roundabout why. And this was written in 1995! Have we come a long way from there? If any friendly economists reading have links to send me – I should be glad of them. After all, one day one of my candidates might actually answer the question and then I’ll be in real trouble! But my own amateurish investigations indicate to me that we haven’t.
There are a couple of aspects to this debate that I think particularly are never discussed enough, and I want to air them here that others might engage on them. Firstly, it’s not obvious to me at all that people save because they intend to consume that money in the future. It seems to me instead that people save because they are intending to earn interest income – which is why stocks and shares are so unpopular but savings accounts and houses are. There is a massive industry within finance of taking risky assets and repackaging them for retail consumption so that they look more like conventional savings accounts. For example, one can buy all manner of equity linked instruments that pay a steady coupon if the underlying stocks perform according to certain conditions (a typical contract will entitle to saver to a fixed coupon of an equity index goes up in a given year, but pay nothing if it does not). Most long markets in long dated options are dominated by the weird and wonderful exposures these products throw off for their issuing banks. The broader point is, if people are investing for income not for future consumption, lowering rates could very well raise the desired savings rate.
The behaviour of US savings rates since the recession is not inconsistent with that idea at all.
Another aspect that’s underappreciated is that this behaviour is institutionalised in pension arrangements that pay a percentage of final salary – for whom lowering interest rates simply raises the pension providers liabilities – resulting in a greater need for the provider to put capital into the fund. In other words, we’ve managed to institutionalise lower interest rates leading to higher savings rates- just by the corporate not the household sector. Indeed, the corporate sector has been running substantial positive financial balances with the rest of the economy in the UK for years (i use the UK because Neil puts that nice sectoral balance chart on his blog and i’m too lazy to do my own). There are other reasons for that to be the case, but it’s not inconsistent with the fact that corporates have to effectively save more and more in order to meet their pension liabilities. British Airways is a pension fund that has an airline arm.
Anyway, the point of all of this reflection is that we’ve built an apparatus for monetary policy that takes this understanding of the world – that lower rates stimulates demand and vice versa – as read – or publicly says that it does. I quietly suspect that in private, central bankers are pretty sceptical about their own abilities to understand or forecast economic variables – and the extent to which they can influence them via policy. When looking at the relationship between markets and central bankers, it’s always good to bear in mind that both are looking uncertainly at the real economy trying to work out what’s going on. The fact that one set of actors cares about the outcomes and one doesn’t does not mean that their actions are more effective.