You would be forgiven, and probably applauded, for allowing your summer to drift happily by without noticing some angst between a few high-profile economists across the pond. The topic of debate has rarely strayed far from the financial headlines this year, namely, “can the Federal Reserve achieve a soft landing?”. If you have managed to escape these headlines, a soft landing can broadly be described as the central bank bringing down inflation without increasing rates too far and causing a meaningful recession. One way of quantifying what this looks like is a reduction in the current number of job openings without a meaningful jump in the rate of unemployment. This is what the debate centres around after a speech in May from Federal Reserve Governor Christopher Waller made the case for a soft landing being an achievable target.
The theory goes as follows, there has been a good historical relationship between the vacancy rate (number of job vacancies/(number of occupied jobs + number of job vacancies)) and the unemployment rate, that is, as policy tightens to slow demand and push down vacancies, we move along this curve and unemployment increases, this is known as the Beveridge curve (Figure 1).
However, Waller’s argument, which is nicely displayed in Figure 2, is that the pandemic has created a situation in which vacancies are outside the normal range due to matching efficiency in workers not well suited to the jobs available. This means policy can tighten and reduce the number of vacancies from the red dot entitled Mar. 2022 and effectively move straight down to the red dot entitled Jan. 2019 and thus not significantly increase unemployment.
Waller’s argument hinges on layoffs, when layoffs are low as they currently are, his model shows the Beveridge curve to be very steep. Olivier Blanchard, Alex Domash, and Lawrence Summers of the Peterson Institute for International Economics took umbrage with this theory, in what I like to imagine is a kind of 8-mile moment for the world of economics. Their retort hinges on the pandemic structurally impacting the natural level of unemployment due to poor matching efficiency and high labour reallocation, which effectively means people who lost their jobs in sectors such as travel and leisure have retrained into another industry, leaving a structural shortage of workers (case and point being airports over the school summer holidays). Demand has now returned strongly in those sectors but due to large cohorts of people leaving those sectors, many are nervous to return or have moved on, leaving a structural gap where matching efficiency is low. They go on to say that tighter monetary policy will lower demand across the economy, but it will not likely influence the efficiency of matching people to jobs where they are most needed. The outcome of this, in Blanchard, Domash, and Summers’ eyes is that there will be a decrease in vacancies but also a rise in unemployment, presumably because monetary policy is not specific enough to tighten conditions in some industries and not in others. They finish by citing historical empirical evidence that vacancies have never fallen from a peak without a meaningful rise in unemployment. The graph below displays the relationship between vacancy rates and unemployment rates, the blue dots representing the quarterly peak vacancy rate at different points in time and the red dots are two years hence, indicating the trend of rising unemployment as vacancies decrease from their peaks.
Both arguments have their merits and flaws, but on pure theory I would have to side with Blanchard, Domash, and Summers who shall be henceforth known as PIIE (Peterson Institute for International Economics) for brevity and hilarity. Stepping back from the x=y world of economics for a second, I struggle to envisage this goldilocks scenario where some employers are drastically reducing the number of job postings due to a large decline in consumer demand, but other sectors are not making meaningful layoffs in response to the same issue, especially as sticky wage inflation beds in creating lower profit margins across the board. Arguments have been made by Fed officials that tech sector hiring freezes are an early example of this working, but at the same time there have been headlines in early September of Goldman Sachs planning to lay off staff imminently and JP Morgan reported that investment banking revenues may be 50% lower than last year. The bankers and tech employees will be fine, this is simply an illustration of how complex the issue is and how cyclical or stable different industries are. As always it will be the lower income end of the spectrum where the pinch is felt the most.
Where I struggle with the PIIE argument is their justification in using historical evidence given that we are outside of the historical range in the relationship between vacancy rates and unemployment. Covid had huge implications for the labour market which we are still learning about and adapting to. Waller’s argument is based on a theoretical model which uses certain assumptions, they may be right, but they will likely be wrong to some lesser or greater degree. In his ever timely memo, Howard Marks (Co-founder of Oaktree Capital) recently wrote about the dangers of forecasting and how the input assumptions on which models are based may work to some degree in normal times, but they cannot account for the unforeseen risks which move markets significantly and where forecasting would actually be most valuable. In a way, this plays into Waller’s argument, he is accepting we are outside of normal times and attempting to wade through the murky waters to see what lies ahead, whereas the PIIE theory is using historical evidence which likely will not play out as they expect and which to their credit they do concede. Were Waller not a politician I would have more faith, but people in such high office have a habit of talking their own book, so whilst I am hopeful a soft landing can be achieved, I am fearful it may not be so.