In pursuing their policy objectives, monetary policymakers in many advanced economies have set interest rates at unprecedentedly low levels. These levels are understood to be close to their ‘effective lower bound’ – a point where they cannot be lowered any further. By lowering nominal interest rates, central banks have sought to reduce also the real rate of interest (the nominal interest rate adjusted for any anticipated inflation developments), stimulating investment and consumption.1 But when nominal interest rates reach their lower bound, the main factor that can further reduce the real rate of interest is a higher rate of expected inflation. In line with this, policymakers have stressed the role of inflation expectations in the effective transmission of asset purchases and other non-conventional policies. The current context of the Covid-19 pandemic has increased the risk of inflation levels that are too low, highlighting the need to avoid a further drop in inflation expectations as part of the process of macroeconomic stabilisation.2
The stabilising effect of higher inflation expectations hinges crucially on the extent to which consumers would act on their inflation beliefs and, in particular, whether they would increase current consumption when inflation is expected to be higher in the future. However, as reviewed in a paper by Candia et al. (2020) at the Jackson Hole Economic Symposium in August 2020, it is fair to say that the evidence on whether or not households would behave in this way has, to date, been very mixed.3 One important study by Bachmann et al. (2015) concluded with a cautionary lesson for central banks; higher inflation expectations may even reduce current consumption when interest rates are constrained by the lower bound. Similarly, Coibion et al. (2019) report experimental evidence for Dutch households showing that consumers who revise their inflation expectations upwards tend to reduce their spending on durables, at least in the short term.
In this column, we review new evidence on this important issue from our recent study (Duca-Radu et al. 2020). The evidence relates to the euro area and exploits the data-rich Joint Harmonised EU Programme of Consumer Surveys which provides quantitative data on consumers’ individual perceptions of (and expectations for) inflation, as well as consumers’ readiness to spend. The survey provides, on average, 26,440 individual responses on a monthly basis. These answers are highly representative of the populations across the euro area. The study documents that an increase in consumers’ inflation expectations relative to their own perceptions of current inflation is indeed associated with a positive response in the readiness to spend. This finding – which is observed with striking consistency across demographic and economic groups and individual euro area countries – supports the view that consumers’ inflation expectations can play a role in macroeconomic stabilisation.
Readiness to spend and the expected change in inflation
When studying the relationship between spending and inflation expectations, we start from a very simple observation. The decisions about spending that consumers make based on their current expectations of inflation may be equally conditioned by their most recent experience of inflation. Why would consumers’ readiness to spend be influenced by their recent experience of inflation? One possible reason is that if consumers perceive current inflation to be higher, they may perceive their real income or their real wages to be lower.
In line with the above observation, we place at the centre of our analysis the difference between consumers’ expectation of future inflation and their own perception of current inflation. As a result, according to our approach, for higher expectations to help stimulate consumption, they must rise relative to a consumer’s own perception of inflation. This approach also brings important practical advantages for empirical analysis of individual consumer data. Most importantly, by focusing on the deviation of expected future inflation from perceptions about current inflation, it is possible to control for unobserved factors at the individual level. As discussed in, for example, Arioli et al. (2017), such unobserved factors include consumer sentiment about the state of the economy (e.g. optimistic or pessimistic beliefs), as well as emotional survey responses. However, to the extent that both perceptions and expectations of inflation are equally influenced by such unobserved factors, their effect will be cancelled out using the proposed difference.
A first look at the data shows that the proposed approach brings a dramatic reassessment of the link between expected inflation and readiness to spend. Figure 1 portrays two scatter plots linking inflation expectations and consumers’ readiness to spend. In the graphical analysis, the consumers’ readiness to spend takes values from one to three, with ‘one’ indicating that they are not ready to spend and ‘three’ indicating that they are ready to spend. To help visualise the relationships, the chart aggregates the individual survey replies by averaging across respondents in each euro area country for each month between January 2003 and December 2016. The relationship between readiness to spend and inflation expectations alone (which may be influenced by unobserved factors and consumer-specific characteristics) is, on average, negative and also moderately unstable (left-hand panel). However, the data clearly reveal that when normalising inflation expectations around perception about current inflation, i.e. looking at the expected change in inflation (right-hand panel), a much more stable and positive relationship with the readiness to spend is detected.
Figure 1 Readiness to spend, inflation expectations and the expected change in inflation
Note: Binned scatterplots of country aggregates weighted by individual weights at one moment in time identified by month and year, using 100 bins. Readiness to spend is measured on the vertical axis. In the left-hand chart, the horizontal axis depicts expected inflation only while in the right-hand chart the horizontal axis depicts the expected rate of inflation minus the perceived rate of current inflation, i.e. the expected change in inflation.
This graphical evidence is supported by evidence from a more formal empirical model that is described in detail in Duca-Radu et al. (2020). One potential concern in such an analysis is that the strong positive correlation observed in the right-hand panel of Figure 1 reflects reverse causation running from consumption to the expected change in inflation, and not vice versa. For example, if respondents expect their own spending to correlate with other consumers’ spending, they might expect the overall economic situation to improve. As a result, they may also then expect stronger inflation in the future. To address this concern, the model explicitly takes into account respondents’ expectations about their own future financial situation as well as their expectations about the overall future general economic situation. As a result, the response of spending to inflation expectations can be estimated while holding such additional factors constant.
Estimates using this model reveal very strong statistical evidence for the positive relationship highlighted in the right-hand panel of Chart 1. The results imply a 0.26 percentage point increase in the probability of being ready to spend in response to a one percentage point increase in the expected change in inflation during normal times (when interest rates are not constrained), and a higher 0.33 percentage point increase when interest rates are constrained by an effective lower bound. This latter value points to an even stronger potential stabilisation role of higher inflation expectations relative to current perceptions about inflation when the lower bound is constraining nominal interest rates. Equally, it points to the potentially stronger destabilising effects of a drop in expected inflation relative to current inflation perceptions in such an environment. Translating such micro evidence to more aggregate macroeconomic conclusions is challenging but our analysis suggests that the response is economically important. Holding perceptions about current inflation constant, a gradual 2.0 percentage point expected increase in inflation (e.g. from inflation that is too low at 0.0% to a level of 2.0%) is shown to increase euro area consumption by approximately 0.28% in cumulative terms over a three-year period when the effective lower bound is constraining policy, and by 0.22% when it is not.
Results hold widely across the population and across countries
More disaggregated analysis reveals that the positive relationship between consumption and the expected change in inflation holds widely across nearly all euro area countries. There are, however, some cross-country differences in the strength of the relationship. For example, the positive spending response is stronger for consumers in countries with higher financial literacy scores – a result which suggests that the stabilising benefits might be augmented by policies aimed at increasing financial literacy. The response is also stronger for consumers in countries where people tend to save a lot and are thus more likely to have a larger stock of accumulated liquid assets (such as cash savings). Findings across different demographic groups also reveal a consistently positive spending response, though employed consumers and consumers with higher incomes or higher levels of education tend to react more strongly.
Consumers’ degree of economic pessimism or optimism and their knowledge about official inflation statistics also seem to have an impact on the strength of the spending response. Figure 2 shows that more optimistic beliefs about one’s own future financial situation, or about the economy more generally, are associated with a stronger spending response. Figure 3 depicts how the spending response differs according to the accuracy of consumers’ inflation expectations (where accuracy is assessed by how well consumers’ inflation expectations can predict the official inflation statistics). The chart shows that consumers with the most accurate expectations (i.e. with an absolute prediction error below two percentage points) exhibit the strongest response. Nevertheless, even the spending of consumers with very large absolute prediction errors (above ten percentage points) responds positively to beliefs about future inflation. Such results suggest that the stabilising benefits of higher inflation expectations might be augmented by policies aimed at increasing the public’s knowledge about official inflation. Equally, the stabilising benefits may be enhanced if beliefs about higher future inflation are associated with greater consumer optimism about the economic outlook.
Figure 2 Spending response to inflation beliefs of optimistic and pessimistic consumers
Note: ELB refers to the effective lower bound. The chart shows the effect of a unit increase in the expected change in inflation on the probability that consumers are ready to spend given current conditions. The spending probability is measured on a scale of 0 to 100. Consumers are considered optimistic when they expect their own financial situation or the general economic situation to either get a little or a lot better; otherwise they are considered pessimistic.
Figure 3 Spending response to inflation beliefs by forecast accuracy
Note: ELB refers to the effective lower bound. The chart shows the effect of a unit increase in the expected change in inflation or inflation expectation on the probability that consumers are ready to spend given current conditions. The spending probability is measured on a scale of 0 to 100.
Overall, the empirical evidence for the euro area tends to support the potential economic stabilisation benefits of higher inflation expectations when interest rates are constrained. Importantly, our study has not addressed key questions related to the transmission of central bank policies and policy communication to consumers’ inflation expectations. Some recent studies (e.g. Haldane and McMahon 2018) have pointed to the potential benefits of communication strategies that are targeted at specific groups. Other studies (such as Candia et al. 2020) have emphasised a risk of unintended effects, especially in cases where consumers interpret a rise in expected inflation as ‘bad news’ for their future economic situation. In the case of the euro area, a new Consumer Expectations Survey4 – currently being piloted across six euro area countries – will offer a useful research tool to further enrich knowledge on these important issues.
Authors’ note: This column first appeared as a Research Bulletin of the European Central Bank. The authors gratefully acknowledge comments from Michael Ehrmann, Alberto Martin and Louise Sagar. The views expressed here are those of the authors and do not necessarily represent the views of the European Central Bank, the Eurosystem, or the European Commission.
Arioli, R, C Bates, H Dieden, I Duca, R Friz, C Gayer, G Kenny, A Meyler and I Pavlova (2017), “EU consumers quantitative inflation perceptions and expectations: An evaluation”, ECB Occasional Paper 186, April.
Bachmann, R, T O Berg and E R Sims (2015), “Inflation expectations and readiness to spend: Cross-sectional evidence”, American Economic Journal: Economic Policy 7(1): 1-35.
Candia, B, O Coibion and Y Gorodnichenko (2020), “Communication and the beliefs of economic agents”, paper presented at the Jackson Hole Economic Symposium, August.
Coibion, O, D Georgarakos, Y Gorodnichenko and M van Rooij (2019), “How does consumption respond to news about inflation? Field evidence from a randomized control trial”, NBER Working Papers 26106.
Duca-Radu, I, G Kenny and A Reuter (2020), “Inflation expectations, consumption and the lower bound: Micro evidence from a large multi-country survey”, Journal of Monetary Economics, forthcoming.
Haldane, A and M McMahon (2018), “Central Bank Communications and the General Public,” American Economic Association Papers and Proceedings 108: 578-583.
Lane, P (2020), “Monetary policy in a pandemic: ensuring favourable financing conditions”, speech at the Economics Department and IM-TCD, Trinity College Dublin, 26 November.
Schnabel, I (2020), “Covid-19 and monetary policy: Reinforcing prevailing challenges”, speech at the Bank of Finland Monetary Policy webinar: "New Challenges to Monetary Policy Strategies", 24 November.
1 Schnabel (2020) raises the question of whether the inflation expectations of households and firms may be more relevant than expectations of financial markets in shaping macroeconomic outcomes.
2 See, for example, Lane (2020), who emphasised “Tolerating a longer phase of even lower inflation than originally envisaged would be costly and risky. First, it would imply a weaker recovery of consumption and investment, as a result of higher expected real interest rates”.
3 See also Duca-Radu et al. (2020) for a further review of the evidence.
4 See the Consumer Expectations Survey.