The Kerslake Review of the Treasury

This review, published today, was commissioned by John McDonnell but is entirely independent. Although it is ultimately Lord Kerslake’s review, it is the product of a small panel of which I was a member, and also reflects submitted evidence and meetings of invited experts. I can say that in my area, macroeconomic policy, this external evidence was very influential and let me thank again all those involved. This post just focuses on these macroeconomic aspects of this review of the Treasury. [1]

The obvious place to start is to think how the role of the Treasury has changed in the last two decades. In 1997 setting monetary policy was delegated to the Bank of England. In 2010 the forecasting aspects of fiscal policy were delegated to the OBR. To a government obsessed by cutting the size of the state that might suggest that the Treasury did not need to have a large macroeconomic capacity, But if you think about the major macroeconomic disasters if the last decade, that view is completely misguided.

One way of thinking about these disasters is that they reflect a failure to consider potential risks to the economy, and what might be done to both mitigate those risks and respond to them if they occurred. No one was ever going to predict the exact time and date of the financial crisis, but someone in government should have been thinking about what risks a rapidly expanding banking sector might pose. There were not many who warned about the risks, but enough to warrant a risk analysis. As I have said before, I doubt that this could have avoided a crisis - the banking lobby is too strong - but at least the government would have given some thought about what to do if it happened before it happened.

When it came to austerity, everything would have been relatively unproblematic if the economy had grown at the pace at first expected in 2010, because monetary policy would still have had control. (Interest rates would have been above their lower bound.) But someone should have been focusing on what happens if things turned out to be less rosy, and making sure ministers had to address these risks. At the very least that analysis would have pinpointed the need to change fiscal policy the moment that more pessimistic outcome came to pass, but perhaps also thinking about this risk might have injected a note of caution into policy before this happened. In a secret Treasury that might not have stopped a determined politician, but if this risk analysis had been made public?

Who in government should have been doing this risk analysis? The obvious institution is not the central bank, which can be far too tentative in the area of fiscal policy and too biased on financial policy, but the Treasury. The Treasury, to use a phrase suggested at one of our evidence gathering meetings, should be “the country’s risk manager of last resort”. The Treasury is uniquely capable of getting information from all the parts of government, including the Bank, and putting it together within a consistent macroeconomic framework.

But this isn’t the only reason why the Treasury still needs a strong macroeconomic capacity. It sets the rules by which fiscal and monetary policy operate, and the danger of not having this capacity is that the rules get determined by political whim, or don’t change through inertia. And it also needs the capability to undertake large pieces of complex analysis very quickly, as we have again seen over the last two decades.

What do I mean by capacity? Above all people: people who have the ability to do and understand state of the art macro analysis. If you compare the number of macroeconomists at the Treasury and the Bank there is a huge imbalance which is not conducive to good policy making. It is absurd to think that you need suites of models to set interest rates, but virtually nothing to set monetary and fiscal policy rules and analyse the impact of potential risks to the economy.

None of this is guaranteed to stop the Treasury become obsessed with the deficit and ignoring macro analysis, but the stronger the macro team is in the Treasury the less likely this is to happen. One other way that is often suggested of combating this danger, and which we considered, involves splitting off from the Treasury key aspects including macro policy into a new Economics ministry. My own view, which is similar to that expressed in the report, is that such a split just runs the danger of institutionalising the dominant role of balancing the budget in policy making.

There is one final benefit of enhancing the macro capacity of the Treasury, and that would be to provide the potential to increase openness. I take it as given that greater openness would be a good thing, and also being an essential way of utilising existing expertise around the country. It is far from clear why risk anaysis has to be secret. To take just two examples, the Bank makes a concerted attempt to find out what is being done in UK universities that might be useful to it, and it publishes a regular blog where their economists can flag interesting data and analysis. It would be good if the Treasury had the capacity to do something similar.

[1] There is a great deal more in the report, both about macro policy and issues around devolution, working with other departments, the overall goals of policy and much more. I also feel I need to note one area where I disagree with how Bob talked about the report yesterday (on Peston’s show and to the Guardian). While I’m sure it is true that the Treasury has lost trust as a result of its incorrect pre-referendum short term forecast, by highlighting this in the context of this report you inevitably give the impression that it did something unprofessional. But both assessments were signed off by Charlie Bean. and the Treasury were hardly alone in expecting negative short term impacts from Brexit. Worse still, it risks suggesting that their long term analysis is suspect.

Whatever happened to the government debt doom spiral

A number of people, including the occasional economic journalist, are puzzled about why government debt at 90% of GDP seemed to cause our new Chancellor and the markets so little concern when his predecessor saw it as a portent of impending doom. I always argued that this aspect of austerity had a sell by date, so let me try to explain what is going on.

The 90% figure comes from a piece of empirical work which has been thoroughly examined, and found to be highly problematic. (Others have used rather more emphatic language.) Part of the problem is a lack of basic thinking. Why should the markets worry about buying government debt, beyond the normal assessment of relative returns. The answer is that they worry about not getting their money back because the government defaults.

If a government cannot create the currency that it borrows in, then the risk of default is very real. Typically a large amount of debt will periodically be rolled over (new debt sold to replace debt that is due to be paid back). If that debt cannot be rolled over, then the government will probably be forced to default. Knowing that, potential lenders will worry that other potential lenders will not lend, allowing self fulfilling beliefs to cause default even if the public finances are pretty sound.

The situation is completely different for governments that can create the currency that the debt they sell is denominated in. They will never be forced to default, because they can always pay back debt due with created money. That in turn means that lenders do not need to worry about forced defaults, or what other lenders may think, so this kind of self fulfilling default will not happen.

Of course a government can still choose to default. It may do so if the political costs of raising taxes or cutting spending is greater than the cost of defaulting. But for advanced economies there is an easier option if the burden of the public finances gets too much, which is to start monetising debt. That is what Japan may end up doing, and what others may also do if QE turns out to be permanent. But this is a very different type of concern than the threat of default. And it does not, in the current environment, lead to the emergence of large default premiums and market panics.

How can I be so sure? Because with QE we have had actual money creation, and it has not worried the markets at all. It seems hard to tell a story where markets panic today about the possibility of monetisation in the future, but are quite sanguine about actual monetisation today.

So for economies that issue debt in currency they can create, there is no obvious upper limit anywhere near to current debt/GDP ratios when economies are depressed and inflation is low. Japan shows us that, and we must stop treating Japan as some special case that has no lessons for the rest of us. (How often did we hear of their lost decade in the 1990s that it couldn’t happen anywhere else.)

It was good that the IFS suggested Hammond has a look at Labour’s fiscal rule. As I explained in this post, Hammond’s new ‘rule’ is pretty worthless. But one key part of Labour’s rule that keeps being ignored but is crucial in today's environment is the knockout if interest rates hit their zero lower bound. It is for the reasons described above that this knockout is there and is perfectly safe: when interest rate policy fails you can completely and safely forget the deficit and debt and use fiscal policy to ensure the recovery. It is the basic macro lesson of the last 6 years that is fairly well understood among academic economists but still remains to be learnt by most people who talk about these things. Whether senior economists in the UK Treasury need to learn it or just keep quiet about it for other reasons I do not know.