Sunday | 25 Oct 2020
08 Oct 2020

Measuring price stability in Covid times


The recent fall in both headline and core inflation in the euro area has increased pressure on the European Central Bank (ECB) to adopt further measures to stimulate the economy. With headline inflation turning negative for two months, additional measures seem to be warranted. However, inflation data based on consumer prices should not be the only guide for policy in the kind of turbulent times we are currently experiencing. A broader price index such as the GDP deflator provides important additional information for policy decision.[1] This broad indicator is increasing at a rate of around 2%. There seems to be no danger of broad-based deflation.

How to measure price stability?

The main aim of the ECB is to preserve price stability, which it has defined as an inflation rate of “below, but close to 2%”. To calculate the inflation rate, the ECB uses the harmonised index of consumer prices (HICP). This is harmonised across countries and covers the consumption baskets of representative households. Changes in HICP captures changes in purchasing power that can affect aggregate demand and consumption in particular. For this reason, it is a key variable for public perceptions of price developments and can be more easily understood than the more abstract concept of the GDP deflator. This is why it has been adopted by the ECB as the reference metric. Likewise, measures of expected inflation-based consumer prices vary widely, both in the form of surveys and those derived from inflation-protected bonds (which are based on measures of the consumer price indices (CPI), rather than the GDP deflator).

The GDP deflator is an alternative measure of inflation. It measures the difference between nominal and real GDP and, unlike the HICP, captures changes in prices related to production and income developments throughout the entire economy. The GDP deflator incorporates not only the prices of domestically produced consumer goods and services, but also other categories of prices, such as the prices of capital goods, including government spending. Crucially, and unlike the HICP, it excludes taxes and the prices of imports. It is not affected, therefore, by changes in tax policies and input price developments. In recent years, this latter aspect has become important in a context of volatile commodity prices.

These two different measures of inflation (CPI and GDP deflator) have a slightly different economic meaning and their relevance can change over time. The main concern is that measures of inflation based on consumer prices are more relevant in a context where there are developments in demand, and in particular households’ propensity to spend and their purchasing power. This was the case during the ‘Great Moderation’ in the early 2000s when development in private debt was largely disregarded. With the advent of the financial crisis, high levels of government and corporate debt became the key issue for macroeconomic policy. Debt sustainability for governments and corporates depends more on the growth of their revenues, i.e. nominal GDP, than on consumers’ purchasing power as measured by the CPI.

The Covid-19 pandemic differs from previous crises in that it has had a strong impact on consumption patterns, with demand disappearing for some goods and services and being temporarily in excess for others. Moreover, goods or services requiring close contact have changed in nature and/or have become subject to restrictions. All these aspects are not necessarily reflected in the baskets used to compute the price indices. Indeed, recent research[2] points to the fact that the Covid-19 pandemic has led to changes in consumer expenditure patterns that can introduce significant bias in the measurement of inflation. The results suggest that the official CPI, measured in its standard composition, understates the actual inflation rate. In addition, as part of the response measures to Covid, many governments have lowered VAT rates, contributing to a fall in consumer prices. In most countries – Germany is one example – the VAT reduction is temporary, implying that measured CPI inflation will increase again next year when the tax cut is reversed. A forward-looking central bank should look beyond this temporary effect.

Figure 1, which shows the annualised growth rates of the headline HICP, the core index and the GDP deflator by quarter[3] over the past two years, indicates three very different states of inflation. Since the beginning of the year, HICP inflation has been falling drastically and even turned negative in the last observation. By contrast core inflation is still positive but on a declining trend. The GDP deflator is on an increasing pattern, despite the recession, and above the target. Even before the pandemic, the GDP deflator had been much closer to the ECB’s 2% target than the core CPI. This suggests that, pre-Covid, the broader, underlying price trends have been less deflationary than often assumed.

If one takes a longer time perspective (see Figure 2), it is apparent that the HICP headline inflation rate has been very volatile over the past few years, and much more so than other measures of inflation. This is mainly due to shifts in energy prices. It is not the first time the GDP deflator provides a different signal than consumer prices-based measures. For example, in early 2011, the ECB was concerned about an increase in inflation, mainly induced by higher oil prices. The GDP deflator, however, was weaker. If the ECB had given some weight to this indicator, it might have avoided what later turned out to have been a too-tight policy stance.[4]

In the present context of high debt, volatile oil prices combined with temporary but significant changes in consumption, the information coming from consumer price measures is very loud. In these circumstances, policymakers should take into account developments in the GDP deflator, which for the time being do not point to additional easing measures. We do not argue that the ECB should redefine its policy target in terms of the GDP deflator, but it should consider this variable when setting policy.

[1] Taylor (1993) showed that the policy of an inflation-targeting central bank could be described by a simple rule in which the central bank reacts to ongoing inflation. In the original specification, the proper measure of inflation is the broadest possible price index, namely the GDP deflator. See also Bernanke (2015)

[2] See for instance Cavallo (2020) and Seiler (2020)

[3] A major drawback of the GDP deflator is that it is available only quarterly and with a lag.

[4] See Alcidi et al. (2016)