Galina Potjagailo, Boromeus Wanengkirtyo and Jenny Lam
CPI inflation in the UK has markedly increased over the last year, reaching 10.1% in September. The aggregate increase reflects potentially different dynamics across disaggregated prices, from which CPI inflation is constructed. How much of the increase has been broad-based across a wide range of prices? We assess this through a measure of ‘underlying inflation’ that captures comovement across many disaggregated prices – energy, food and other (‘core’) price items. We observe a substantial rise in underlying inflation, hence many prices have increased jointly. Broad-based energy price increases have been the main driver of underlying inflation. Furthermore, about a quarter is due to core price items which reflect more persistent inflation.
A rise in CPI inflation can reflect strong price rises concentrated in some items that are generally more volatile or that are affected by sector-specific shocks. Or it can be due to many prices increasing simultaneously due to aggregate shocks or spill-overs via so-called ‘second-round effects’. The recent rise in UK inflation was initially fuelled by two large external shocks: supply bottlenecks along global value chains due to the Covid-19 (Covid) pandemic (for instance, with microchips and used cars) and soaring energy and food prices related to Russia’s invasion of Ukraine. Energy and food prices tend to be volatile and are set internationally, thus reflecting external factors and changes in the UK’s terms of trade, the price received for exports relative to the price paid for imports. However, broader domestic price increases via ‘second-round effects’ can follow because energy prices affect the costs of many other goods and services through their role as indispensable input in production and transport. And food and energy are salient consumer goods on which households rely over-proportionally when forming inflation expectations. Also, firms’ inflation expectations affect inflation via their forward-looking cost optimisation and these effects tend to be stronger in expansions.
A simple metric indicates that since mid-2021 an increasing share of CPI components – most recently 90% – has reached inflation rates above their past averages, as shown in Chart 1. And a first principal component reflecting static comovement among CPI items shows a substantial gradual rise. However, these metrics do not capture dynamic interdependencies between price items via which second-round effects from broad-based food and energy price increases and aggregate shocks can permeate into aggregate inflation. In the following, we track the breadth of price increases in the UK using a more refined statistical method.
Chart 1: Simple measures of comovement among price categories
An underlying inflation measure (UIM) for the UK
We measure the comovement across 438 monthly CPI items over the period 2006 to 2022 using a dynamic factor model, based on the framework developed at the Federal Reserve Board. The inflation rate of each item is disentangled into: i) a common component that is shared with the inflation rates of all other disaggregate price items; and ii) idiosyncratic fluctuations that capture fluctuations over time specific to that item and measurement error. In a second step, the common components of all individual price items are aggregated into the UIM using the item weights in the CPI basket, which reflects households’ average consumption patterns.
The UIM tends to reflect slow-moving changes in inflation. This is shown in Chart 2, which decomposes annualised headline CPI inflation into the underlying inflation (red bars), and idiosyncratic components (yellow). Historically, underlying inflation picks up the more stable component of inflation and slow-moving changes more closely related to service prices, stripping out month-to-month fluctuations. Most transitory changes that last two to three years and often relate to goods prices are instead captured by the idiosyncratic component.
Chart 2: Underlying inflation, 2006–22
Underlying inflation has picked up substantially over the last year
In September, the historic underlying inflation measure (red bars) has reached 5.8%, the highest level observed over the sample period, having increased by 3.6 percentage points in cumulated terms since early 2021. This is close to where an alternative measure, the average inflation in the lowest-volatility quantile currently stands, as discussed in a recent speech by Catherine L. Mann. It is also comparable to the Federal Reserve’s estimates of US underlying inflation (6.0% for September).
We also account for instabilities and unusual comovement across many prices due to the Covid shock and the Russia-Ukraine war not seen in the data before. For instance, the two shocks could have induced stronger linkages across prices by affecting global value chains, or comovement may have increased once inflation rates reached higher levels. At the same time, the large size of recent shocks might have induced instability into the model, which these additional components would pick up. These effects are captured by adding two threshold effects in the parameter estimation in March 2020 and February 2022. The Covid component captures comovement across prices between March 2020 and February 2021 that differs from historical norms. Similarly, the Russia-Ukraine war component captures unusual changes in comovement from February 2022 onwards (we assume that there is no additional unusual comovement stemming from Covid from that point onward).
The effects from Covid (green bars) initially weighed on inflation. Since the fourth quarter of 2021, it has turned positive, which likely reflects goods price pressures from global supply constraints and increases in services prices as the economy reopened. Since March 2022, the effects from the Russia-Ukraine war (blue bars) have increasingly boosted inflation. The effects of these components should fade out relatively quickly as the two shocks subside – we therefore view them as another type of erratic component rather than as part of underlying inflation. Nonetheless, there is a risk that these shocks might have contributed to a more persistent shift towards stronger price comovement, for instance because the degree of price spill-overs across items can increase at high inflation rates. Also, the pass-through of input costs and wages to prices might have changed, and recent firm-level evidence from the Decision Maker Panel suggests that supply constraints and recruitment difficulties are increasingly affecting firms’ price setting in the UK.
Broad-based energy price increases might begin to propagate to domestic price pressures
Measures of underlying inflation typically only focus on core price items, which are less affected by external cost pressures. However, this overlooks that energy and food items, despite their volatility, can partially reflect aggregate shocks, or spill over to other prices via production networks or expectations. And some ‘core’ goods items are still susceptible to idiosyncratic movements, such as with used car prices over the pandemic. In the following, we reweight the common components of parts of the basket only and we distinguish between the part of broad-based inflation that is due to food and energy prices moving in sync with the rest of the basket (‘Food/Energy contribution’), and the part which reflects underlying inflation in core goods and services (‘Core contribution’).
The bulk of the increase in underlying inflation is due to broad-based energy price increases, suggesting that energy prices have increasingly comoved with other prices. Chart 3 decomposes the cumulated increase in UIM, the Covid and Russia-Ukraine war effects since March 2021 (overall 8.8 percentage points). Almost two thirds of this increase came from broad-based increases in energy items’ prices (5 percentage points), and a much smaller contribution of 1.3 percentage points came from broad-based increases in food prices. This reflects that energy prices have increased much more than food prices, and have been moving more strongly in sync with other prices compared to food items. These broad-based increases in energy and food items contributed not only to the more erratic Covid and Russia-Ukraine war effects, but also to the UIM. The effects on underlying inflation should, in principle, decay once the external shocks behind energy and food price spikes subside. However, the fact that these items have moved jointly with many other UK price items indicates that the external shocks might begin to propagate to domestic price pressures.
Chart 3: Cumulative change in inflation since March 2021
Almost one quarter of the increase in broad-based inflation is due to core item inflation that picks up more persistent, structural drivers of inflation. Underlying inflation among core items is generally very stable, remaining close to 1.9% until 2019 before slightly declining in 2020. However, it has increased by 2.5 percentage points since early 2021, reaching 3.9% in September. It has continued rising over the recent months, even as published UK core inflation itself has stabilised. Hence, the contribution of core underlying inflation relative to idiosyncratic movements has been growing. This suggests that inflation in the UK is broadening across price items and increasingly reaching core item prices.
Conclusions
The rise in inflation rates has put central banks across many countries in front of difficult trade-offs. The UIM provides a tool to assess how broad-based the increase in inflation has been, and whether increases in volatile price items that reflect external shocks, such as energy and food, are becoming embedded across many prices. Our finding of a rise in underlying inflation among core items suggests that the inflation in the UK is partially driven by broad-based increases in prices that are typically rather stable. Over the past, shifts in this component have been quite persistent, so it could plausibly remain elevated. The precise link between the breadth of price increases and inflation persistence in a high inflation environment remains an open question relevant for central banks.
Galina Potjagailo and Boromeus Wanengkirtyo work in the Bank’s Structural Economics Division and Jenny Lam works in the Bank’s Advanced Analytics Division.
If you want to get in touch, please email us at bankunderground@bankofengland.co.uk or leave a comment below.
Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.
