Board Member Tuomas Välimäki
Dinner Speech: Monetary economics and reality: how the monetary policy strategy review relates to IQ, reggae and throwing Frisbee
Bank of Finland and CEPR joint conference
Helsinki, 12 October 2019
Monetary economics and reality: how the monetary policy strategy review relates to IQ, reggae and throwing Frisbee
Ladies and Gentlemen, it’s my pleasure to welcome you all to the CEPR conference dinner at Suomen Pankki.
The theme of this weekend’s conference is monetary economics and reality. These two concepts are like ‘kangaroo ball’ and ‘sun lotion’. They don’t seem to have much in common at the outset, but actually they do – they can be dangerous if used simultaneously.
Well, in reality, monetary economics and reality are not that far from each other. To me monetary policy has always been the combination of these two concepts.
A pre-dinner speech, which is given to an audience after more than 10 hours of intensive participation in a conference, should always include a joke. Now, if it’s an old joke, even better, as you can already start laughing half-way through. The following anecdote somehow relates to monetary economics and reality, or at least to modelling and intuition.
According to Stanley Fischer, Paul Samuelson once said: "I'd rather have Bob Solow than an econometric model, but I'd rather have Bob Solow with an econometric model than without one."
Well actually this wasn’t a joke after all. What we policymakers need is both monetary economics (i.e. theory) and reality, i.e. good intuition about the functioning of the world. One should never rely on pure intuition, particularly during times when we need to take actions that we’ve never taken before. Yet, even our best models are useless if we never subject them to a reality check.
It is precisely this type of thinking that we at the Bank of Finland have behind our call for a review of the ECB’s monetary policy strategy. I simply find it very hard to understand why we shouldn’t periodically assess the validity of the strategy, as well as the models and tools being applied. The more so, if we haven’t done so in sixteen years and the environment we operate in has gone through extreme developments – like the Global Financial Crisis, the European sovereign debt crisis, and the enduring fall in the natural rate of interest, to name just a few.
Furthermore, in reaction to the observed changes, we (the policymakers) have started to apply measures we didn’t use before the global financial crisis, like large-scale private and public sector asset purchases. Also, with our standard toolkit, we have gone into unforeseen calibrations. For example, we’ve not only taken policy rates into negative territory; we’ve held them there already for five years.
Although we have demonstrated that the lower bound was not at zero, we still aren’t totally sure where the effective lower bound lies, or what kind of side effects we might be bringing about if and when we continue this for a number of years (as the market pricing currently indicates). We still need a lot of analysis and research from central banks and the academic community to reach a full understanding of our unconventional policies – this, too, calls for a strategy review.
Moreover, some of our fresh tools, and the new ways we manage the old ones, rely more than before on expectations, and hence also on the behavioural changes their communication may result in.
Since the rational expectations hypothesis we’ve been well equipped for situations where our policy actions rely on beliefs, expectations and behavioural changes. We can simply ignore the process that forms the expectations, as we may deduce the relevant expectations directly from our models, can’t we? Unfortunately not. Growing empirical evidence documents large deviations from the null of full-information rational expectations (Coibion & Gorodnichenko (2012 & 2015)).
These findings have brought about a revival in studying the formation and updating of subjective expectations, partly because subjective expectations shape individuals’ economic choices and ultimately aggregate outcomes, as our keynote speaker Andrei Shleifer has well documented (Gennaioli & Shleifer (2018)).
In times of low interest rates, the management of expectations and central bank communication have become important items in the toolbox of all major central banks. Yet, in the analysis of the impacts of these measures, all decision-makers, including also non-expert households, are still often assumed to behave according to the rational expectations hypothesis.
Theoretical predictions and empirics have not always walked hand-in-hand when it comes to the non-standard monetary policy measures. On one hand, for example, Ben Bernanke once said that “the problem with quantitative easing is that it works in practice, but it doesn’t work in theory.” On the other hand, the forward guidance puzzle, which has been referred to many times in today’s sessions, indicates that expectations management, or forward guidance, is less effective than the underlying models imply. One possible deviation from the standard representative-agent New Keynesian model is to allow for heterogeneous agents with uninsurable income shocks.
This class of models results in a discounted Euler equation and can resolve the limited power of forward guidance (see e.g. Nakamura & Steinsson (2016), Del Negro et al., Kaplan, Moll, Violante (2016), but also Werning (2018), who argues that in this class of models forward guidance can be just as powerful). We are delighted to have among us both professor Werning and professor Violante, who are important contributors to this literature.
Recent theoretical literature, which is strongly represented at our conference, has taken a different detour from the classical model and allowed for deviations from the null of full information and rational expectations to explain the limited effectiveness of expectations-based policies. Examples are the finite planning horizons of agents (Woodford (2018)), level-k thinking (Farhi & Werning (2018)), and behavioural New Keynesian agents (Gabaix (2019)).
Unfortunately, I have not lately had much time to devote to research. However, there is one paper at this conference that I have personally and concretely contributed to.
To test whether limited cognition matters for the effectiveness of conventional and expectations-based policies, D’Acunto, Hoang, and Weber teamed up with Maritta Paloviita from the Bank of Finland to use Finnish data for studying these questions.
They could observe the measures of IQ at the individual level for almost all men in Finland (and this is my personal contribution I just referred to: I took the IQ test in 1989 while doing my military service). They merged the data with inflation expectations and other expectations from the micro data underlying the consumer sentiment data of the European Commission, and also household balance sheet data from Statistics Finland.
They found striking differences according to the cognitive abilities of Finnish males. Individuals with high-IQ have on average absolute forecast errors for inflation that are smaller by a factor of 2.5 than those with low-IQ. [Now, looking at the forecast errors of our inflation forecasts just makes me wonder from which part of the distribution we’ve been hiring our experts.] Their results also indicate that people with high IQ also adjust their consumption plans to their inflation expectations in line with economic theory and they also adjust their behaviour to changes in interest rates both on the upside and on the downside.
None of these results holds true for men with low cognitive abilities. As these men represent a large proportion of the population and also a great share of aggregate income, the result may indeed be relevant for aggregate demand.
These findings suggest limited cognition could be an important explanatory factor for the limited effectiveness of expectations-based policies and should possibly be included in the development of models for policy decisions. Moreover, they may raise the concern that communication-based strategies and policies might result in unintended consequences, such as redistribution from those who do not adjust their economic decisions as a consequence of central bank action to those who do.
Now, according to Gorodnichenko and Weber, about 40% of US citizens assume that the Fed’s inflation aim is 10% or above. Having this in mind, one may ask if we should even bother trying to reach everybody with our communication. Isn’t it enough if the financial markets understand our aim and our reaction function? Or if in addition to the financial markets, the social partners and firms that agree on wages and set prices understand our aim? I actually don’t know – this is truly an open question to me.
Still, at some level I’m also happy with this finding. I have a 20-year-old daughter. She often questions me and my work: “Why are you always fussing around with inflation. It’s so from the ’80s. Don’t you understand that nobody’s interested in inflation anymore?” This makes me happy, as I always have in my mind Alan Greenspan’s definition of price stability. He defined price stability as a situation where “expected changes in the average price level are small enough and gradual enough that they do not materially enter business and household decisions.” It seems that we are there; if nobody cares about inflation, we have price stability.
So, do I then think that we can simply just ignore communication to the general public? No, certainly not; quite the opposite. One of the challenges we, i.e. central banks around the world, face is indeed how to effectively communicate with the ordinary consumer, in terms of both reaching consumers and ensuring that they understand the aims and objectives of our policies.
If we don’t do this, if we don’t convince new generations of the justness of our cause, there will always be some stable genius who starts questioning our actions. Without ex ante communication it can be hard to convince people of the importance of price stability as an ultimate goal – isn’t price stability just something that will always prevail anyway?
Now, how to do this? There is the recent example of the Bank of Jamaica, which taped music videos with famous reggae stars to explain to consumers the basic aims of monetary policy. Is this the way to go?
At the Bank of Finland, we are proud of being among the few active central banks on social media like twitter, in terms of both an institutional account, but even more importantly also through personal accounts at all expert and managerial levels, including our Governor, Olli Rehn. Perhaps we can even pave the way for other countries both inside and outside the euro area in terms of direct communication to ordinary consumers.
Finally, I just explained why I believe we should be active in our communication, not only to the financial markets and the academic community, but also to wider audiences. But we should also adjust how we try to reach these different audiences.
I have a dog and a friend with a PhD in physics. When I throw a Frisbee, my dog can catch it 8 times out of 10 throws. When I started trying that, the ratio was 1 out of ten. We improved the result by repetition. I am sure we all agree that the result wouldn’t have improved much if my friend had devoted hours of his time in explaining air resistance or gravity to my dog. Our experts are free to use DSGE codes when they try to explain the merits of temporary price level targeting to you, but when I try to explain the recent decision of the ECB’s governing council to a TV audience, I’d rather be using plain Finnish.
Simplicity is a virtue in communication. So, to put it simply, dinner is now being served in the room next door. Thank you.