Can we pull off a soft landing?

Update: Gauti Eggertsson picks up on my post, pointing out that his paper with Pierpaolo Benigno actually does try to come up with an estimate for inflation once the flat part of the Phillips curve has been reached (figure 8 in the paper). Their estimate is 2.5%, in which case I’d be much more confident that a soft landing can be achieved. This estimate comes with a large confidence interval, though, so if the pace of inflation at point D in the second chart below is closer to 3-to-4%, for example, it will be a lot harder to avoid a hard landing, to get inflation down.

—

Central bank hiking cycles in the developed world are slowly but surely coming to an end, raising the question of whether they have pulled off a soft landing, defined as a fall in inflation back towards target of around 2% without a meaningful decline in output and rising unemployment. On the face of it, the answer to this question is a resounding no. Interest rates in Europe, the UK and the US are up anywhere from 300 to 450bp in less than a year, driving up bond yields , and pushing yield curve inversions to near record levels. Anyone using these data points to predict what comes next, using historical relationships, will conclude that the wheels are about to come off in developed economies and their financials markets alike. The difficulties in the US regional banking sector is, in this case, simply a canary in the coal mine, warning of bigger shocks to come. The investment implications of such a view are simple; short equities and long short-term government bonds.

This tried-and-tested recession trade might yet prove a winner, but a new paper by Pierpaolo Benigno and Gauti B. Eggertsson makes the encouraging prediction that it is in fact possible for the Fed, and by implication other central banks, to engineer a soft landing. It is interesting to see professor Eggertson presenting this kind of research. Any of my readers who are inclined towards the darker arts of theoretical macroeconomics will know that Eggertson’s work has been instrumental in thinking about how to conduct monetary policy in a liquidity trap, and the effectiveness of fiscal policy at the zero bound. Now he is musing about what happens when the economy gets off the zero bound too. Benigno and Eggertsson (2023) identifies three possible explanations for why policymakers and markets got the transitory-story wrong. First, almost all medium and long-term measures of inflation expectations were low and anchored even as inflation was rising, giving credence to the idea that inflation would come down quickly. Secondly, policymakers struggled to correctly identify the inflationary effect of supply disruptions in product and services markets, due mainly to the sustained damage to global supply chains as a result of Covid lockdowns. Thirdly, policymakers were operating with a an assumption of a very flat Phillips Curve, leading them to the false conclusion that even a relatively large degree of excess demand, a positive output gap, would lead to only a small rise in inflation.

It is the failure of this third assumption that Benigno and Eggertsson (2023) identify as the key reason for the fact that the transitory story missed the target, and the resulting great chase to bring down inflation across the developed world. In a nutshell Benigno and Eggertsson (2023) propose the idea of a non-linear phillips curve. I will spare readers the gory details, but instead attempt to capture the model, and its implications, through two carefully chosen charts from the paper. The first chart below shows the empirical foundation for the non-linear Phillips Curve, and the second plots the theoretical implications in a standard Keynesian AD/AS macro-model. The first offers empirical evidence for two periods, 1960-to-1969 and 2008 to 2022, with significant non-linearities due to vacancy-to-unemployed ratio above 1. Note that we don’t know whether similar non-linearities existed in the periods 1970-to-1987 and 1987-to-2008, because at no point during these periods did the vacancy-to-unemployment ratio rise above 1, in this sample. The second shows why this matters for policymakers.

Figure 4 in Benigno and Eggertsson (2023)

Figure 10 in Benigno and Eggertsson (2023)

The soft landing in this model is achieved by realising that inflation, in the case of a tight labour market, can be brought down quickly through a relatively small sacrifice in output, signified by the move from point C to D in the chart. This sounds great, but it only works as a soft landing if we assume that the inflation rate D is below or at the central bank’s target. If it isn’t, you’re stuck in a difficult environment with a flat phillips curve and above target inflation, which implies a significant sacrifice of output, a large rise in unemployment, to achieve a small decline in inflation. This, in turn, implies that negative supply shocks, such as the one experienced during Covid, could well be a lot more important than assumed by Benigno and Eggertsson (2023). This is because it is exactly such supply shocks, moving the Inv-L supply curve to the left, which can result in inflation being above target, even at point D.

More generally, the model in Benigno and Eggertsson (2023) puts labour supply at the forefront for policymakers. In particular, it is easy to see how negative labour supply shocks—declines in labour force participation—can have huge implications for the trade-off between inflation and output faced by policymakers because such shifts have the ability to shift the economy from the flat to the steep part of the Phillips Curve relatively quickly, especially if the economy is close to “D”, to begin with. The relatively persistent fall in labour force participation in the U.S. during and after Covid, compared for example to Europe, seems to be a key story here, though it is difficult to quantity in relative terms between the US and European experience.

Finally, the soft landing requires policymakers to adequately perceive the non-linear Phillips Curve, which requires detailed and timely data on job openings and labour supply. The US certainly seems to be in a good position here with the Jolts survey, but even in the context of timely and accurate supply-side data, it will still be difficult for policymakers to identify exactly where the kink is, and where the economy is relative to it. In short, even in this model, the risk of over-tightening monetary policy after a long period of getting inflation wrong remains a high risk, especially with inflation rising as far as it has, and given the inherent lags between policy, output and the labour market.

Are we in a new regime?

The question of whether we’ll achieve a soft landing is as much a question of how we define this landing, which in turn depends on the point we’d like to get to, and the state of the global economy after Covid and the war in Ukraine. As such, I suspect we are now living in a world where inflation at point D is in fact still (significantly) above target, which leaves monetary policymakers two choices. They can stick to their mandate and sacrifice a significant degree of output to bring down inflation, the hard landing, or they can accept that the economy is now—due mainly to geopolitical shifts and fiscal policy—one in which the pace of nominal activity has shifted higher, at least for a while, or in the language of the Keynesian AD/AS model, in which inflation is now higher for a given level of real economic output. The ability to do that without a damaging loss of credibility depends on a number of things, chiefly among which is the ability for monetary policymakers to operate close to the kink, or perhaps even on the steep slope of the supply curve, without undue political pressure. This, in turn, depends on fiscal policy and the political economy. We do on occasion still hear about price controls and MMT, which would alleviate the pressure on central banks for the simple reason that they, under such a policy regime, would cease to be independent. But I still think it is safe to assume that even if fiscal policy remains inflationary, politicians will continue to blame central banks for failing to reign in inflation, while at the same time criticising them if they tighten too far too quickly, causing a recession. As such, a soft landing can be pulled off if DM central banks are able, both within their own institutions and vis-a-vis politicians and the public, to fudge their inflation targets. If they can’t I fear that even in the Benigno and Eggertson model, inflation at point D in the chart above is still above target, which forces policymakers to impose a hard landing on the economy to bring down inflation.