Fed Terminal Rate: Powell vs Volcker

Summary

• A new NBER working paper co-authored by Larry Summers implies even a Volcker-era interest rate hike may not be enough to return US inflation to 2%.
• It argues former Fed Chair Paul Volcker’s success in tackling inflation in the 1980s is overstated in official CPI data. In reality, doubling the Fed Funds rate to 20% only dropped core inflation 4.9pp.
• Jerome Powell requires a similar level of disinflation today, i.e., US inflation is as severe as it has ever been. But changing consumer habits and less effective monetary policy mean achieving disinflation may be harder than in the 1980s.

Introduction

The current high US inflation comes from many factors. These include expansionary fiscal policy, accelerating wage growth, rising food prices and record-breaking housing inflation. None are likely to abate soon.

But there is good news. The US Consumer Price Index (CPI) – a basket of goods and services representing US consumer spending (and the most popular measure of inflation and deflation) – appears to be near its peak, having risen only slightly in May to 8.6%.

That is far below the official March 1980 peak of 14.8%. Back then, Fed Chairman Paul Volcker doubled the main interest rate (the Fed Funds rate or FFR) to 20% in two years to drop CPI by 12.3pp. The move caused two back-to-back recessions. But with the April 2022 CPI much lower, and decreasing, can we escape similar macroeconomic consequences?

Not necessarily, according to a new NBER working paper. When correcting for differences in how CPI was previously measured, the inflationary pressures the Fed faces now are much higher – like those it faced 40 years ago. And the disinflation it engineered then is the same size as what we need today to return core CPI to 2%. However, the authors argue achieving this now may be harder as consumers spend more on services rather than transitory goods.

How Have CPI Measures Changed Over Time?

Since 1983, CPI has undergone numerous methodological changes, from quality adjustments on used cars to using geometric means to calculate price changes for subcomponents. This makes comparing official CPI figures across time inaccurate (Chart 1).

In response, the Bureau of Labor Statistics (BLS) publishes a CPI for all Urban Consumer Research Series (R-CPI-U-RS). It answers the question: ‘what would have been the measured rate of inflation from 1978 forward had the methods currently used in calculating the CPI-U been in use since 1978?’ The paper then augments this series with a similar methodology pre-1979.

The difference between the official and time-consistent CPI series is most evident pre-2000s, and especially during The Great Inflation. For example, using today’s methods, headline CPI from 1978-83 peaked at 11.6%, rather than 14.8%, and ended at 3.4% instead of 2.5%. For core CPI, the peak was 9.1%, far lower than the official 13.6% figure.

Housing Inflation

According to the authors, the disparities between the two series from 1960-83 boil down to how housing inflation was measured.

Housing is both a large, fixed investment and a good that provides a service, shelter, that is consumed daily. Today, we use the owners’ equivalent rent (OER) to measure housing, which is inferred from rental prices.

Between 1953 and 1983, however, OER was known as ‘homeownership’. It broadly captured changes in the expenses of homeowners mainly determined by house prices and mortgage rates.

By inferring OER from rental prices, the BLS can strip away the investment aspect of housing to isolate owner-occupiers’ consumption of residential services. The homeownership component did not distinguish between the two, however, which garnered two important criticisms.

First, it created a mechanical relationship between monetary policy and homeownership costs in the CPI basket. Shelter costs would move very tightly with the FFR because the FFR is closely connected to mortgage rates (Chart 2).

Second, consumers either paid for the home purchase price or the mortgage payments, but not both. Including both led to a larger share of housing in the consumption basket. As a result, homeownership received 26.1% of the weight of the overall CPI index in December 1982. After the change, one month later, the OER received only 13.5%.

The strong positive co-movement between housing inflation and the FFR, alongside the considerable weight housing was given in the overall CPI basket, explains why official inflation today does not equal official inflation figures pre-1983.

Why Does It Matter?

The paper and BLS (post-1978) adjust official CPI figures with the latest methodology. Doing so shows the current inflation rate is nearer the peak of other cycles than the official CPI data suggests (Chart 3).

This is especially true for core CPI (which excludes volatile food and energy components), as housing accounts for a larger proportion of the overall basket. Core CPI was higher than today only two other times: 1975 and 1978-82. Clearly, official CPI data understates the scale of the current inflationary problem relative to the past.

Moreover, the adjustments highlight a worrying discovery. CPI responded far less to monetary policy during the 1960s and 1970s than the official CPI statistics implied. For example, Volcker’s extremely hawkish stance only reduced core CPI by 4.9pp, instead of the 10.6pp official statistics show. Remarkably, this is only slightly more than required to return current core CPI to its 2% target (Chart 4). In other words, official CPI data also overstates how effective monetary policy is at curbing inflation.

Has Disinflation Become Harder to Achieve?

Quantitatively, the amount of disinflation required today is like the 1980s. It could, however, prove harder to achieve. The reason, the authors argue, is because of changing consumer spending habits. Specifically, consumers are purchasing fewer transitory goods, such as food and apparel, and more ‘price sticky’ services (Chart 5).

Housing inflation, which is particularly sticky, is one such example. The OER component had a weight of 14.5% in 1983. This has now risen to 24.3% of overall CPI and 30.6% of core CPI in 2022. In April 2022, housing added as much as 2.7pp to headline inflation, far above average (Chart 6), and is expected to continue to rise throughout 2022.

As a result of changing consumption habits, the authors estimate that the average frequency of price adjustment has increased from 4.5 months in the late 1940s to 7.3 months in 2022. This price adjustment can be thought of as the frequency at which businesses change their prices to consumers.

A Note on Real Rates

There is concern over whether negative real rates (interest rates adjusted for inflation) could increase inflation further by stimulating investment. Economist Olivier Blanchard explains why he worries about this, assuming real rates are as low as in the 1970s. But if we use the adjusted level of inflation, the real FFR is now more negative than during the 1970s (Chart 7). So his analysis may be overly conservative.

Bottom Line

Volcker’s Fed took a far more hawkish tone on inflation in the 1970s and 80s than current Chair Jerome Powell. Last week, we explained why the fiscal nature of this inflationary shock may be behind the Fed’s reluctance to raise rates as significantly this time around. Summers and co., however, show why this reluctance may do little to curb inflation meaningfully.

There are a few takeaways from this paper:

• The inflationary pressure we face today is not far off the most substantial we have faced in a generation.

• The primary tool we have to tackle inflation is less effective than official CPI data implies.

• Changing consumer habits have increased prices in areas of the economy that make this bout of inflation less likely to fade as quickly as previous ones.