Economy – Forget soft landings – how much of a recession is needed to tame inflation?

We look back at how inflation responded in previous downturns to help gauge the scale of the adjustment in growth and unemployment needed to tame it this time around.

Keith Wade, Chief Economist and Strategist, Schroders

Since the last interest rate move by the Federal Reserve (Fed) on 27 July, investors have taken a more optimistic view of when the central bank can bring monetary tightening to a close. Markets are now pricing in a “Fed pivot” in late 2023, when the central bank is expected to cut interest rates.

Two factors have supported the move.

First, Fed chair Powell has said that US interest rates are now neutral, indicating that the initial adjustment from the ultra-easy pandemic policy is over and that future rate decisions will be taken on a meeting-by-meeting basis depending on the data.

Second, that data show the economy cooling as retail spending and housing slow. The latest GDP figures actually showed the US economy contracted in the first two quarters of the year and although they overstate the weakness of the economy, final demand clearly softened.

However, despite signs of slowdown the likelihood of a more pessimistic outcome on policy, where interest rates have to remain higher for longer, has significantly increased in our view. The obstacle to a Fed pivot is the high level of underlying inflation and the strength of the labour market, as evidenced by the latest employment report which showed a significant increase in payrolls and a further fall in unemployment.

An analysis of past cycles shows that it would be a rare achievement for an economy which is so late in its cycle to bring inflation back to target without a fall in activity, or an outright recession. In our view it would be better if the Fed took a leaf out of the Bank of England’s playbook and acknowledged this, rather than projecting soft landings. 

Cycles and inflation: a look back

At the peak

The rapid rebound in the US since the economy reopened from Covid restrictions last year has taken activity above its long-run trend, as evidenced by a tight labour market and high capacity utilisation rates in industry. At 3.5% the unemployment rate is well below estimates of equilibrium, or the Non-Accelerating Inflation Rate of Unemployment (NAIRU) (the lowest level of unemployment that can occur in the economy before inflation starts to rise). The Congressional Budget Office (CBO) put this at 4.4% in the second quarter.

Meanwhile, CPI inflation has risen to 9.1%, its highest level for 40 years. Relative to previous peaks since 1960, the current position compares with an average CPI inflation rate of 6.1% and an unemployment rate 0.5% below the NAIRU (chart 1).

Source: Schroders

The US is clearly late in its cycle and as signs of slower growth come through, we would argue that the economy probably reached a peak relative to trend in the current quarter.

How much of a slowdown in activity is needed to bring inflation down?

To help answer this we have looked back at previous peaks in the cycle to gauge the effect of the subsequent contraction in output on unemployment and inflation. According to the NBER, there have been nine previous occasions since 1960 where the economy has been at a peak. In each case the economy then went into recession, before troughing out several months later. The last such contraction took place between February and April 2020 – the shortest recession on record.

Prior contractions in the US since 1960 have lasted between six and 18 months and are more typical of what we might expect going forward. Looking at those eight cycles, the average fall in GDP was 1.6% from peak to trough and the unemployment rate rose by 2.5 percentage points (pp), moving from below to above the NAIRU. On the CPI measure, inflation fell by 1.5 pp on average.

There was a wide range of experience, with inflation falling by more than average during the “Great Recession” of 2007-09 and the second “Volcker” recession of 1981-82, when GDP fell 3.8% and 2.5% and inflation by 5.5 and 6.2 pp respectively. The worst outcome in terms of the growth-inflation trade-off was 1973-75 when, despite a contraction of 3.1% in GDP, inflation actually rose 2%, a severe case of stagflation (see chart 2).

So how does this relate to the current position? So far, we have seen that a significant fall in GDP has been needed to bring a major fall in inflation. For example, a 6 pp fall in CPI inflation from current levels to 3% would require a decline in GDP of just over 3 pp based on the two major recessions mentioned above. In terms of the impact on jobs, the unemployment rate would rise by around 4 pp to 7½%.

From this perspective, the Fed’s projected soft landing where growth slows to just below 2% and inflation falls below 3% in 2023 looks like wishful thinking.

However, before we dismiss the Fed’s forecasts completely, we need to dig deeper into the current high CPI inflation rate. Are there reasons to believe that inflation may come down more easily, i.e. with less output loss or a smaller increase in unemployment?

To start, commodity prices have played a significant role in boosting inflation. If we strip these out, then inflation on the Fed’s preferred measure (the core PCE deflator) is running at 5.2%. It would seem that a fall in inflation back to 2% from here would be less onerous. However, the sensitivity of core PCE inflation to changes in GDP is also lower. For example, in the two major recessions cited earlier the impact of GDP on inflation is more than halved, so we would still need a 3 pp fall in GDP to generate a fall in core inflation of just over 2%.

Nonetheless, that would still bring inflation closer to target. Would such a downturn in the US also lead to lower commodity prices, helping to drive headline inflation down further? In the past a US recession could be expected to trigger just such a fall as global demand weakens, but today the outcome would be very dependent on how the world economy adjusts to the potential loss of Russian supply. It is possible that shortages keep oil (and commodity prices in general) elevated, even with a US recession.

So far it looks as though a significant slowdown in GDP will be needed to hit the Fed’s inflation goals. However, our historical comparison does not capture the structural changes in the world economy over the past 60 years. In particular the success in keeping inflation low and stable for a considerable period of time means that inflation expectations remain well anchored.

One of the reasons inflation proved so stubborn during the 1970s and early 80s was the pick-up in wages which followed the initial spike in inflation. Subsequent second round effects kept inflation high as wages and costs rose. To a large extent this reflected a lack of belief in the ability of the authorities to bring inflation down.

Chart 3 shows that inflation expectations (both short and medium term) were elevated in the late 1970s and when combined with strong trade unions and labour bargaining power, it was not surprising that pay accelerated and the economy entered a wage price spiral. Consequently, unemployment had to rise significantly to bring wage growth down.

Today the picture is different, although short-run inflation expectations and wage growth have picked up with the tight labour market, medium-term price expectations remain stable. Short-run expectations, which tend to be sensitive to the price of gasoline, have risen, but over five years households expect inflation to be close to target. If sustained, this bodes well for the labour market adjustment; unemployment need not rise as much if wage growth is contained.

Greater central bank credibility and possibly lower commodity prices could help bring inflation down faster than in the past and at less cost in terms of output and employment. However, the fundamental problem remains: the US economy and much of the world is late cycle and overheating.

Monetary policy is a pretty blunt instrument in these circumstances, with central banks being forced to tighten until unemployment rises and sufficient slack is created. In our view, this would point to a fall in GDP of around 2% from peak to trough, less than in the Great Recession or Volcker era, but still significant and more than the current consensus of economists. 

Fed should take a leaf out of the Bank of England’s book and forget the soft landing

To achieve this the Fed will have to tighten further and take interest rates above their current view of neutral. Rates will be higher for longer, but that does not mean tightening relentlessly until unemployment is 6 or 7%, for example. The lags from higher rates to the economy mean that the Fed should proceed cautiously as the full impact is not felt for many months later. In this respect there is scope for a Fed pivot toward the end of next year, with rates likely to be easing as the economy falls into recession.

Although the Fed’s options are limited, it could take a leaf out of the Bank of England’s (BoE) book. The BoE has taken considerable flack for forecasting a significant recession in the UK with inflation only moving slowly toward target. However, no one could argue that they have not warned people, giving households and businesses a signal to what is ahead.

In this respect it would be helpful if chair Powell and the Fed stopped projecting a US soft landing. A look back at history shows that such forecasts only give false hope and create a further misallocation of resources. Politically this is difficult, but the earlier households and firms can start to make the inevitable adjustments the better.