Rethinking macroeconomic policy: Introduction
Olivier Blanchard 20 April 2015
This year’s IMF conference, “Rethinking Macroeconomic Policy III”, gathered many of the world’s greatest economists to reflect on the state of post-Global Crisis macroeconomics. This column, which presents remarks by the IMF’s Chief Economist Olivier Blanchard, argues that much detailed work has been done. The trenches are being dug, but we still do not have a good sense of their final destination.
On 15 and 16 April 2015, the IMF hosted the third conference on “Rethinking Macroeconomic Policy”. I had initially chosen as the title and subtitle “Rethinking Macroeconomic Policy III. Down in the trenches”. I thought of the first conference in 2011 as having identified the main failings of previous policies, the second conference in 2013 as having identified general directions, and this conference as a progress report.
My subtitle was rejected by one of the co-organisers, namely Larry Summers. He argued that I was far too optimistic, that we were nowhere close to knowing where were going. Arguing with Larry is tough, so I chose an agnostic title, and shifted to “Rethinking Macro Policy III. Progress or confusion?”
Where do I think we are today? I think both Larry and I are right. I do not say this for diplomatic reasons. We are indeed proceeding in the trenches. But where the trenches are eventually going remains unclear. This is the theme I shall develop in my remarks, focusing on macroprudential tools, monetary policy, and fiscal policy.
Let me start with macroprudential tools. If anything, the Crisis has convinced most of us that they have to be part of the basic macro toolkit. And much progress has been made, both in terms of research and in terms of policy. Measures of systemic risk are being developed, with an eye on implied instruments. A recent survey by the US Office of Financial Research identified 31 methods to identify particular dimensions of systemic risk, and there are no doubt many more. Most countries have put in place macroprudential authorities, sometimes at the central bank, such as in the UK, sometimes outside with central bank participation, such as in the US. More and more countries are experimenting with loan-to-value ratios and other tools to affect housing demand and prices.
But where are these trenches going? We do not know the shape of the future financial system, for example the degree to which it will be institution/bank based or market based. Sure, the same uncertainty applies to, say, the high-tech sector, and we do not worry; we just observe, and we shall see where it goes. But finance is different. Policymakers cannot be simple observers, as what the financial system will be depends very much on regulation. And we do not have a good sense of what regulation should be.
Consider the distinction between financial regulation and macroprudential policies, which, for these purposes, I shall define, borrowing from the presentation by Paul Tucker, as “dynamically adjusted financial regulation”.
Take, for example, capital ratios. In principle, variable capital ratios (macroprudential) sound better than fixed ones (financial regulation). But is the tradeoff really that clear? The difficulty of identifying when and how to move the ratios, and the political economy complications associated with such changes, make the answer far from obvious.
The preliminary conclusions from a study by an IMF team looking at banking crises since 1970 in advanced economies are that 90% of them would have been avoided had the capital ratio been 15% (a bit higher than what is coming out of Basel III). Maybe a constant 15% is better than a ratio that varies between, say, 10% and 20%.
Very much the same set of issues arises with respect to cross-border flows. Much work is going on revisiting the effects of the various types of capital flows, and the efficacy of various forms of capital controls and FX intervention. As we heard from the speakers in thesession on capital flows, there is fairly wide agreement that capital flows can be disruptive, and that it may make sense to use FX intervention, and capital controls. But again, the ultimate question, namely the degree of openness of the capital account, remains unanswered. Should some flows be permanently banned, if it were indeed possible? We do not know.
Let me turn to monetary policy. Central banks have experimented and researchers have explored – often in that order. We have learned a lot, namely that the zero lower bound (ZLB) can be reached and is hard to get away from, that financial assets are truly imperfect substitutes, that QE can affect the term premium, and that bank runs happen to non-banks as well. Down in the trenches, a lot has been done and is being done. But again, it is not clear where the trenches will lead.
Take two related issues – the size and the composition of central bank balance sheets. (I shall leave out one of my favourite issues – how to avoid the ZLB in the future, and the optimal rate of inflation; not much progress has been made here but I don’t think the issue has gone away. I shall also leave aside the central issue of coordination of monetary and macroprudential tools.)
I shall focus on the Fed, but the arguments apply to the other central banks just as well. At this time, the balance sheet of the Fed is about $4.5 trillion, roughly five times larger than it was before the Crisis in 2007. T-bills, which accounted for about 25% of the total, are gone from the balance sheet. Nearly all Treasury securities that the Fed holds have a maturity of more than one year, more than half have a maturity of more than five years. Relative to where we are, how should the balance sheet of the future look like?
For the moment, the issue is largely moot. The required direction of movement is clear, and it will take a long time for the balance sheet to adjust to whatever it should be. But should it go back to its pre-Crisis size, or even smaller given the steadily decreasing demand for currency for transaction purposes? Or should it remain larger, with interest paid on the excess reserves that banks would have to hold? I see no argument for a large central bank balance sheet, but the discussion has not taken place yet.
Should the Fed return to intervening only at the short end of the yield curve, or are there good reasons for continuing to intervene along the curve? Put another way, what are the advantages of affecting both the short rate and the term premium on longer maturity bonds? How solid is the argument that, to the extent that the Fed holds T-bills, it is depriving the private sector of precious collateral? I do not know the answer to those questions, yet they are central to where we want to go in the end.
Let me move to a brief discussion of the third pillar – fiscal policy. We have learned many things. Fiscal stimulus can help. Public debt can increase very quickly when the economy tanks, but even more so when contingent – explicit or implicit – liabilities become actual liabilities. The effects of fiscal consolidation have led to a flurry of research on multipliers, on whether and when the direct effects of fiscal consolidation can be partly offset by confidence effects, through decreasing worries about debt sustainability. (There has been surprisingly little work or action where I was hoping to see it, namely on a better design of automatic stabilisers. It is explored in this spring’s Fiscal Monitor (IMF 2015), but so far there have been no actual policy changes.)
Admittedly, navigation by sight may be fine for the time being. The issue of what debt ratio to aim for in the long run is not of the essence when there is a large consensus that it is too large today and the adjustment will be slow in any case – although, even here, Brad DeLong has provocatively argued that current debt ratios are perhaps too low. (I shall not embark on the debate on secular stagnation, and whether we can act under the assumption that r will be less than g for the indefinite future.)
But how to assess what the right goal is for each country? This remains to be done. It has become clear that there is no magic debt-to-GDP number. Depending on the distribution of future growth rates and interest rates, on the extent of implicit and explicit contingent liabilities, one country’s high debt may well be sustainable, while another’s low debt may not. Conceptually and analytically, the right tool is a stochastic debt sustainability analysis (something we already use at the IMF when designing programmes). The task of translating this into simple, understandable goals remains to be done.
In short, while the trenches are being dug, we still do not have a good sense of their final destination. We clearly need to have another conference in two years.
IMF (2015), “Now Is the Time. Fiscal Policies for Sustainable Growth”, Fiscal Monitor, April.
 A video of the panel discussion is available at http://www.imf.org/external/mmedia/view.aspx?vid=4177815578001.