Howdoinflationexpectationsimpact consumerbehaviour?

How do inflation expectations impactconsumer behaviour?∗

Ioana A. Duca†, Geoff Kenny‡, Andreas Reuter§

August 19, 2016

AbstractThis paper investigates empirically the relationship between con-

sumer inflation expectations and spending using individual consumerlevel data from the European Union Joint Harmonised Business andConsumer Survey. We document that for the Euro Area this relation-ship is positive, that is, higher expected inflation which reduces realinterest rates is associated with an increase in the consumer′s like-lihood of spending. More specifically, for a 1 percentage point (pp)increase in inflation expectations we find a 0.16 to 0.33 pp increasein the probability that a consumer will spend in the current period.This relationship is stronger when the effective lower bound is bind-ing. Country analysis corroborates the pooled results. All countries inthe sample except one exhibit a positive, although heterogeneous, re-lationship between consumer inflation expectations and the likelihoodof spending. Finally, using these estimated probabilities, we indirectlyestimate the impact of a gradual increase in inflation expectations onactual real consumption, and find that this impact is also positive inline with economic theory.

JEL classification: D12, D84, E21, E31, E52Keywords: Consumer inflation expectations, Consumption, Micro data

∗Preliminary draft, please do not circulate. The views expressed in the paper belong tothe authors and are not necessarily shared by the European Central Bank or the EuropeanCommission.†European Central Bank‡European Central Bank§European Commision DG-ECFIN

1

1 IntroductionWith the effective lower bound binding in several economies around theworld, the relationship between consumer inflation expectations and aggre-gate demand plays an important role in policy making. According to main-stream economic theory, under sticky nominal rates, an increase in inflationexpectations should lower real interest rates (due to the so called FisherEffect) and as a result boost consumption or aggregate demand by loweringconsumers′ incentives to save. In an effective lower bound environment, wheninterest rates are bounded from below, this relationship becomes even moreproeminent as central banks are deprived of the use of their main conventionalpolicy instrument, the short-term interest rate. A large theoretical literature,among which Krugman et al. (1998), Eggertsson and Woodford (2003), Junget al. (2005), Eggertsson (2006) emphasizes the stabilization role of infla-tion expectations at the effective lower bound. Yet, empirical evidence onthe inflation expectations - consumption relationship is still scarce and whatstudies exist have brought forward conflicting conclusions with respect to thenature and direction of the relationship: Bachmann et al. (2015), Ichiue andNishiguchi (2015), D′Acunto et al. (2015), Burke and Ozdagli (2013).

The foundation for empirically investigating the consumer inflation ex-pectations - consumption relationship is having the right data. Aggregatedtime series data would not necessarily do the trick as through aggregation alot of information is lost and the heterogeneity of consumer behaviour cannotbe taken into account. See, for instance, Figure 1 where we plot consumerinflation expectations, consumer readiness to spend indicator and real totalconsumption growth, all in aggregate terms. Moreover, investigating thisrelationship in the more recent effective lower bound (ELB) period wouldalso be difficult with aggregate time series data since there are only few ob-servations available. In contrast, microeconomic data can help identify thisrelationship and how it may change over time due to constraints such asthe effective lower bound. This paper benefits from a very rich consumersurvey dataset which has been generated in the framework of the EU JointHarmonised Business and Consumer Survey Programme1.

1The programme is administered by the European Commission (EC). Its consumersurvey is the largest of its kind, covering the 28 European Union (EU) member states, aswell as four of the five candidate countries, with the number of respondents amounts to upto 41060 each round each month. For comparison, the University of Michigan Survey ofConsumers interviews only 500 consumers, while the Federal Reserve Bank of New YorkSurvey of Consumer Expectations has 1200 households in the sample.

2

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

2.5

3

-20

-15

-10

-5

0

5

10

15

20

25

30

35

2003 2005 2006 2008 2009 2011 2012 2014

Consumer Inflation Expectations (lhs) Readiness to spend (lhs)

Real total consumption growth (rhs)

Figure 1: Consumer inflation expectations, real total consumption and readi-ness to spendNote: Quarterly data, 2003Q4 - 2015Q3. Consumer Inflation Expectations - price trendsover next 12 months (balanced statistics); Readiness to spend - major purchases at present,percent positive replies, source: DG-ECFIN Consumer Survey; Real total consumption -household consumption expenditure, Eurostat

To investigate the inflation expectations - consumption relationship inthe Euro Area (EA), we follow a two-steps approach. First, based on a veryrich dataset provided by the EU Consumer Survey, we study the relationshipbetween inflation expectations and the propensity to spend of the EA con-sumer and for a large collection of its constituent countries. To the best ofour knowledge, this is the first paper to provide evidence of this relationshipfor the EA. The granularity of the survey provides an ideal micro-informationset to investigate this relationship. Since 2003, the Consumer Survey includesspecific quantitative questions about consumers′ perceptions about currentinflation and their expectations for inflation over the next 12 months. More-over, consumers are asked other questions referring to their financial situa-tion, the general economic situation, their savings behaviour and intentionswith regard to major consumer purchases and these replies can be directlymatched to replies about inflation expectations. In addition, the replies canbe broken down across several important dimensions, such as gender, educa-tional attainment, employment status, income level, etc. In a second step,as the survey only provides information about the intention or readiness tospend, we estimate a bi-variate VAR to model the interaction between the

3

probabilities from our micro-level analysis and actual consumption. This al-lows us to estimate indirectly the impact of changes in inflation expectationson real consumption at the aggregate level and conduct scenario analysis toexamine how consumption will respond when inflation expectations rise orfall, both when the lower bound is binding and when it is not.

We find that EA consumers behave in line with macro-economic theory,that is, when they anticipate an increase in inflation, consumers decide alsoto increase their current spending. There are four main results that supportthis conclusion. First, pooled EA analysis shows that for 1 pp increase ininflation expectations the likelihood of spending increases by 0.16 pp to 0.33pp, depending on the economic regime. This result holds under differentspecifications in which we exploit the granularity of our dataset and con-trol for demographics, expectations about the individual situation and alsoabout the overall economic situation. To account for the heterogeneity ofthe economies that constitute the EA, we also include country dummies. Weintroduce time dummies and macro controls to take account of aggregateand country specific macroeconomic developments. Second, as our econo-metric model distinguishes between two regimes, outside the ELB and atthe ELB2, we find that the relationship between consumer inflation expec-tations and likelihood to spend is stronger at the ELB. This result is robustacross all model specifications. Third, individual country results confirmthe pooled results. With only one exception, all countries in the sampleexhibit a positive relationship between consumer inflation expectations andthe likelihood of spending. Indeed, country heterogeneity also shows up inthe results as the average marginal effects, although generally positive, canalso differ across countries. Fourth, the VAR analysis reveals that given ascenario in which consumer inflation expectations increase by 0.25 pp overfour consecutive quarters, there would be a 0.3 to 0.5 cumulative increasein the annual growth of real consumption growth rate over a three yearshorizon compared to a baseline scenario where inflation expectations remainunchanged. Papers by Ichiue and Nishiguchi (2015), D′Acunto et al. (2015)find that also Japanese and German consumers respectively behave in linewith theory, while Bachmann et al. (2015) and Burke and Ozdagli (2013)have contrary conclusions for durables consumption in the US. Also for USconsumers, Armantier et al. (2015) verify whether consumers act on theirbeliefs about future inflation in an investment decision by combining surveydata with a financially incentivised experiment and find evidence that most

2We define a ELB dummy, which takes value 1 from June 2014 to July 2015 (end ofour sample) and 0 otherwise. June 2014 the rate of the deposit facility became negative.

4

respondents make their choice in accordance with economic theory.

The remainder of the paper is structured as follows. In section 2 we pro-vide important details about the dataset that we use and shortly illustrateour methodology to exploit fully the micro dataset. Section 3 includes allthe results of our probabilistic analysis of the micro level data in which wedetermine a consumer inflation expectation - propensity to spend relation-ship, as well as the second-step analysis where we translate the results forthe consumer propensity to spend into an impact on actual consumption.Country specific results are also detailed in this section. Section 4 concludes.

2 Data and methodology

2.1 DataThrough its level of detail and the focus of the questionnaire, the EU Con-sumer Survey provides the ideal micro-information set to study the relation-ship between inflation expectations and the likelihood to spend for the EAconsumer. The survey is carried out at a monthly frequency and is coveringall European Union economies, as well as four of the five candidate coun-tries, although in this paper we focus only on EA countries3. Thus, eachmonth we benefit of a sample of 26, 440 consumers, subject to adjustmentdepending on the response rate. The sample is designed to be representativeof the population in each country. Its size varies across countries accord-ing to the heterogeneity of their economies and it is generally positivelycorrelated with the country population size. Each month there is a new sam-ple of consumers that are interviewed, so we actually work with a repeatedcross-section. The vast majority of the surveys in the euro area countries isconducted by computer-assisted telephone interviews (CATI). 4 Most of thequestions in the survey are qualitative and refer to the consumer′s financialsituation, the general economic situation, their savings behaviour and inten-tions with regard to major consumer purchases. Since 2003, the ConsumerSurvey includes specific quantitative questions about consumers′ perceptionsabout current inflation and their expectations for inflation over the next 12

3Actually, our sample does not include Ireland due to data availability.4Only in three countries (Germany, Latvia, Slovakia), interviews take place in a face

to face (F2F) setting. Two countries apply mixed modes which combine CATI (Austria)or CATI and F2F (Lithuania) with web interviews. The households to be interviewed aredetermined by random sampling or quota sampling from a frame which, in most cases, iseither the country’s telephone directory or its population register.

5

months. In addition, the replies can be broken down across several impor-tant dimensions (e.g. gender, educational attainment, employment status,income level etc.) and thus allow us ensure our results are not driven byspecific sources of heterogeneity. The sample employed in this paper coversthe period between May 2003 and July 2015.

The EU Consumer Survey data indicates that EA consumers hold veryheterogeneous opinions about inflation expectations and perceptions depend-ing on their gender, age, education, income or employment status. Inflationexpectations are higher for females, the unemployed, consumers aged be-low 50, with low income and holding only primary or secondary education(see Table 1). Inflation perceptions follow the same pattern, though theyare persistently higher than expectations, although the gap between the twohas tended to narrow over the sample. Also, both consumers′ expectationsand perceptions of price changes are persistently higher than actual inflationdevelopments, measured by the HICP. This positive difference might be ex-plained in several ways: the survey questions are open ended with a genericreference to consumer prices and provide no guidance for the respondent indetermining the inflation rate, unusual replies are not probed, respondentsare asked not about an objective index but assumably their own subjectiveinflation experience and hence they are likely taking as reference a differentindividual basket of goods. Nevertheless, the size of the difference has nar-rowed considerably throughout time reflecting at least two possible issues:the substantially higher inflation perceptions at the beginning of the sam-ple might have been due to the introduction of the euro notes which waspartly corrected in subsequent years as mentioned in Biau et al. (2010); con-sumers have become more informed and more confident about the objectiveand actions of the European Central Bank. Disregarding this persistent pos-itive difference, both expectations and perceptions co-move quite stronglywith actual inflation (see Figure 2),and such a strong co-movement certainlyprovides strong grounds to use this dataset for investigating the consumerinflation expectations - spending relationship.

The main questions that we are using from the questionnaire are:Q51: By how many per cent do you think that consumer prices have gone up-/down over the past 12 months? Consumer prices have increased by __,__%/ decreased by __,__%.Q61: By how many per cent do you expect consumer prices to go up/downin the next 12 months? Consumer prices will increase by __,__%/ decreaseby __,__%.Q8: In the view of the general economic situation, do you think that now itis the right moment for people to make major purchases such as furniture,

6

Inflation expectations Inflation perceptionsMean Median Mean Median

GenderMale 6.7 3.0 9.3 5.0Female 8.5 5.0 11.9 7.0Age16-29 8.1 5.0 10.8 5.030-49 7.8 4.0 10.9 5.050-64 7.6 3.5 10.4 5.065+ 7.0 3.0 10.2 5.0EducationPrimary 7.8 3.0 11.8 5.5Secondary 8.0 4.0 10.8 5.5Further 6.8 3.0 8.8 5.0Income1st quart 8.8 5.0 11.9 6.02nd quart 7.6 4.0 10.5 5.03rd quart 7.3 3.0 9.9 5.04th quart 7.0 3.0 9.4 5.0Employment statusUnemployed 9.5 5.0 12.7 8.0Employed 7.5 3.5 10.3 5.0Euro Area 5.5 2.0 9.8 4.0

Table 1: Mean and median inflation expectations and perceptions over 2003- 2015

electrical/electronic devices, etc.? Survey respondents can answer: i) yes, itis the right moment now; ii) it is neither the right moment nor the wrongmoment; iii) no, it is not the right moment now; iv) don′t know.Questions Q51 and Q61 are quantitative and the answers are expressed inpercentage points, reflecting consumer inflation perceptions and consumerinflation expectations respectively. Q8 is qualitative and shows whether theconsumer would be willing to spend at the current time on durables givena certain macroeconomic context, we will refer to it throughout the paperas the so called ”readiness to spend” of the consumer. In addition to thequestions listed above we use also information conveyed by questions askingabout demographic characteristics, expected and current consumer financialsituation, expected general economic developments, expected unemploymentsituation.

7

-5

0

5

10

15

20

25

2003 2005 2007 2009 2011 2013 2015

Mean Inflation Expectations Mean Inflation Perceptions

HICP

Figure 2: Mean inflation expectations and perceptions vs HICPNote: Individual and country weights used for aggregation. Time period covered: May2003 - July 2015.

When investigating the relationship between consumer inflation expecta-tions and its likelihood to spend, we are using a measure that is innovativein the related literature which we call the expected change in inflation. Thismeasure is simply the individual difference between inflation expectation andperception. The reason for this choice is that it allows us to control for thestrong variation in the level of perceived inflation across consumers. In otherwords, when changing their spending intentions, consumers do not take intoaccount only expected future inflation, but they also consider expected infla-tion relative to the perceived current inflation rate. D′Acunto et al. (2015)also separately controls for consumers′ perceptions of past inflation and findsthat marginal effects would be virtually identical over several specificationsif inflation perceptions would not be included. This finding is yet anotherargument in support of our measure. Bachmann et al. (2015) control forthe current official inflation rate, which by definition is common across allconsumers. A first look at the data (see Figure 3) indicates a positive rela-tionship between the expected change in inflation computed in this way andthe average readiness to spend.

8

11.

52

2.5

3R

eadi

ness

to s

pend

-30 -20 -10 0 10Expected change in inflation

Figure 3: Scatterplot readiness to spend vs inflation expectationsNote: One dot represents a country aggregate (weighted by individual weights) at onemoment in time (identified by month and year). Readiness to spend is coded 1 for notbeing the right moment to spend, 2 for being neither the right moment nor the wrongmoment and 3 for being the right moment to spend.

2.2 MethodologyOur data dictates our modelling strategy. The discrete nature of the spendingattitudes that are retrieved from the survey combined with the fact thatwe observe a repeated cross-section recommend the use of a discrete choicemodel. In this paper, we employ the ordered logit model. Thus, what wemodel in this paper is not the relationship between inflation expectations andaggregate spending, but rather between the expected change in inflation andthe likelihood to spend at the individual level. There is a natural orderingin our dependent variable, the consumer readiness to spend. As it representsthe answer to the question whether it is a good moment to spend, it can thenbe ordered into being more or less ready to spend. Choosing an alternativeover another depends on a latent variable (i.e. some continuous measure ofreadiness to spend) which is not observable and that can be modeled as:

y∗it = Xitβ + εit (1)

where i is consumer i and t is time, y∗it is the latent variables, Xit is a vectorof controls that will be explained in detail in the next section, βit a vector ofcoefficients and εit is the error term.Each alternative can then be defined in relation to the latent variable definedin equation 1:

9

yit =

1 if y∗it < α12 if α1 ≤ y∗it < α23 if y∗it ≥ α2

(2)

Each alternative response has a probability Pr attached:

Pr (yit = j) = Pr (αj−1 < y∗it ≤ αj) = Pr (αj−1 < Xitβ + εit ≤ αj) == Pr (αj−1 −Xitβ < εit ≤ αj −Xitβ) =

= F (αj −Xitβ)− F (αj−1 −Xitβ) (3)

where j is alternative j and F is a function that satisfiesF (−∞) = 0,F (+∞) = 1 and dF (x)

dx> 0. The probabilities of all alternatives must sum

to 1. We model F through a logit function which ensures that the estimatestake values between 0 and 1, i.e. the domain of admissible values for a prob-ability. Alternatively, we could have used a probit function. However, inpractice, the probit and logit models generally yield very similar predictedprobabilities and marginal effects (see, e.g.Davidson and MacKinnon (2004)).

We use maximum likelihood to estimate the parameters of these proba-bility functions, including the thresholds of the latent variable depending onwhich the consumer chooses one survey reply over the other. Nevertheless,parameters β are of limited interest, instead we are interested how the prob-ability of each alternative changes with a change in our controls, i.e. we willfocus on the marginal effects measuring the impact of a change in a givencontrol on our estimated probabilities:

∂Pr (yit = j)Xk

i

= [f (αj−1 −Xitβ)− f (αj −Xitβ)] βk (4)

where k is regressor k and f = F ′, in our case the probability densityfunction of the logistic distribution.

3 Empirical resultsThis section reports all our empirical findings following the two step approachdescribed in the introduction. Thus, in subsection 3.1 we carefully explainour micro-data model specifications and show the EA results concerning theinflation expectations and the propensity to consume relationship, and inaddition we discuss what our model implies for the role of other factors inthe consumption decision. Subsection 3.2 shows how we translate our microdata conclusions about the consumer inflation expectations impact on the

10

propensity to consume to the impact on actual consumption. Subsection 3.3shows country specific results concerning the impact of consumer inflationsexpectations and the propensity to spend, thus it is based solely on the microdataset.

3.1 Model specifications and overall resultsOne important challenge for empirical analysis is whether any identified rela-tionship between inflation expectations and consumption can be interpretedas causal effect of inflation expectations on consumption. With macro data,such a problem of endogeneity is particularly severe because aggregate infla-tion expectations and aggregate consumption are determined simultaneouslyand therefore it is not possible to distinguish the causal effect. With mi-cro data, the situation is much improved in particular because individualexpectations about how aggregate prices will evolve can have an impact onindividual spending, whereas it is less plausible to expect that expectationsabout the aggregate price level would be driven by a consumers own indi-vidual spending intentions. In addition, to ensure that what we capture issolely the effect of inflation expectations on spending, we control for a seriesof covariates that can be correlated both with spending and inflation expec-tations. We also include interaction terms that allow the effect of inflationexpectations to vary depending on certain characteristics and we distinguishbetween ELB and non-ELB regimes. A quick summary of how we model thelatent variable can be found in the equation below:

y∗it = β0 + β1ELB + β2∆πeit + β3∆πe

itELB +Xitγ + εit (5)

where ∆πeit is the expected change in inflation, ELB is a dummy variable

taking value 1 from June 2014 to July 2015, Xit is a vector of controls, εit isthe error term and β1, β2, β3, γ represent parameters and vector of parame-ters respectively.

In the estimation process we gradually control for several potential de-terminants that could simultaneously affect both inflation expectations andreadiness to spend and we allow several interactions. First, we control for arich set of consumer characteristics: age, gender, education, employment sta-tus and income; which we wrap up together under the group ”Demographics”.We have already seen in section 2.1 that there is significant heterogeneity ininflation expectations and perceptions in relation to consumer characteris-tics. Souleles (2004) shows that variations in inflation expectations depend

11

on consumer demographics. The same characteristics may determine differ-ent purchasing propensities and we would want to ensure that any impactof inflation expectations on spending - if it is to be interpreted as structural- is not simply driven by these differences. Second, we consider equally im-portant to control for individual expectations of the general economic andunemployment situation, e.g. in a booming economic environment, con-sumers may increase spending on expectations of rising inflation or simplydue to the favourable economic context. Likewise, controlling for the in-dividual expected financial or current debt situation is important as mostoften consumer′s personal situation has a stronger impact on his/her be-haviour compared to the general economic developments, e.g. provided thata consumer expects that his/her financial situation may deteriorate, his/herconsumption plans will probably decrease even though he/she expects thatthe economic situation will get a lot better and inflation will increase. Thesecontrols are grouped under ”Expectations and financial status”. Third, wecontrol for a number of pairwise interactions between the expected changein inflation and the expected financial situation, debt status, employmentstatus, education and respectively income. Figure 4 shows in a series ofscatterplots that holding the same expectation with respect to the changein inflation, consumers′ readiness to spend is different depending on theireducation, income, the expected financial situation or employment status.Therefore, by introducing interaction terms in our model specification wecapture this heterogeneity in the inflation expectations - readiness to spendrelationship. Fourth, we introduce annual time dummies to control for aggre-gate macroeconomic developments. Fifth, to account for the heterogeneityof the economies that constitute the EA, we also include country dummies5.Sixth, we also include country specific and EA macro aggregates, by drawingon information sources outside the survey. In particular, we control for dis-posable income, lending rates and uncertainty6, which we hold as simultane-ously affecting consumer inflation expectations and their spending attitudes.

We find that there is a robust positive effect of our measure of the ex-pected change in inflation on the probability of being ready to spend acrossall specifications. Table 2 reports average marginal effects of a one unit in-crease in our measure of expected change in inflation, across all specifications

5We actually check the scatterplots of the expected change in inflation and consumerreadiness to spend at country levels and indeed we find heterogeneity among EA economies.

6We proxy uncertainty by the Survey of Professional Forecasters GDP uncertainty,although we have also estimated our model using two other uncertainty proxies: theVSTOXX and unemployment levels, with similar results.

12

Rubric

www.ecb.europa.eu ©

Heterogeneity in the relationship brought in by consumer education,

income, expected financial situation or employment status.

How do inflation expectations impact consumer behaviour? 9

Note: One dot represents a simple average a particular category of consumers (a defined by education, income, expected

financial situation and employment status) at one moment in time (identified by month and year).

Step 1: Consumer inflation expectations and readiness to spend – micro evidence

Figure 4: Scatterplots of expected change in inflation vs readiness to spenddifferentiating by consumer education, income, expected financial situationand employment statusNote: One dot represents a simple average of a particular category of consumers (asdefined by education, income, expected financial situation and employment status) atone moment in time (identified by month and year).

that we estimate7. The average marginal effects are based on the orderedlogit estimation and correspond to the alternative of being ready to spend8,i.e. show the impact on the probability of being ready to spend. Outsidethe ELB average marginal effects range from 0.16 to 0.29 pp increase in theprobability of being ready to spend for 1 pp rise in the expected change in

7We have also run a separate set of ordered logit regressions when instead of the ex-pected change in inflation variable we use inflation expectations and add inflation per-ceptions as a separate control. Results generally confirm the ones reported in this paperwith all marginal effects of inflation expectations expectationsestimated to be positive,although they are smaller. Moreover, consistent with our estimates, the average marginaleffect of inflation perceptions exhibits a negative sign, most likely capturing the real in-come impact of higher inflation on spending. Note that these results are not reported inthis paper.

8With the ordered logit model one can separately estimate the probabilities of eachalternative, i.e. being a good moment to spend, not a good moment to spend, neithergood nor bad moment to spend, and the same for the marginal effects.

13

inflation, while at the ELB they range between 0.24 to 0.33 pp. As D′Acuntoet al. (2015) we find that adding demographics and other consumer expecta-tions significantly improves the Pseudo R2. In addition, in our case, addingcountry dummies to account for the EA heterogeneity almost doubles thePseudo R2 and diminishes the average marginal effects, although they remainhighly significant and positive. Average marginal effects of the specificationswhich include country dummies are the smallest in the range. Nevertheless,given all specifications, EA consumers behave in agreement with standardeconomic theory, where an increase in inflation expectations leads to an in-crease in individual spending intentions.

In terms of the sign of the impact, our results are consistent with re-sults reported by D′Acunto et al. (2015) although our analysis suggests thatthe impact on spending probabilities is considerably smaller. These authorsdemonstrate that German consumers9 behave in line with theory and anincrease in inflation expectations leads to a 6 to 9 percentage points (pp)increase in the probability that consumers are ready to spend. Our resultsare also in accordance with Ichiue and Nishiguchi (2015). They use micro-data for Japan, which in contrast with other economies, has experienced aprolonged period of near zero interest rates. Thus, the authors argue, even inexpectation of higher inflation, consumers are less likely to expect a similarsimultaneous movement in nominal rates. Their results show that consumerswith higher inflation expectations tend to increase current consumption rel-ative to future spending. However, our results are at odds with findings ofBachmann et al. (2015). Using the Michigan Survey, the authors find thatthe effect of higher inflation expectations for US consumers is very close tozero and statistically not significant during normal times, while in periodswhen the ELB is binding, it is shown to be negative (i.e. higher inflationwhich reduces real interest rates is associated with a drop in consumption).One practical observation that we make in relation to these results: theyare based on 67,855 observations covering a time span of 24 years10. Af-ter a simple calculus, every month the authors are left on average with asample of 195 consumers out of the 500 who are interviewed. This is a con-sequence of the fact that for these baseline results only a subsample of firstinterviews11 is used, month-household observations that are larger than 20

9They use data provided by the market research firm GfK, which conduct the consumersurvey for Germany on behalf of the European Commission. Nevertheless, they have onlyqualitative data about inflation expectations and perceptions.

10In our paper we benefit of a sample of 1,793,110 over approximately 12 years, whichamounts to 6427 observations per month.

11The Michigan Survey has a rotating panel structure, i.e. about 40% of the respondents

14

percent12 are excluded. As a consequence, one could reasonably doubt therepresentativeness of this reduced sample for the US economy. Also withrespect to US consumers, Burke and Ozdagli (2013) finds that consumers donot increase their spending on large home appliances and electronics in re-sponse to an increase in inflation expectation, but they do increase spendingon non-durable goods by 1.1% for 1 percentage point (pp) increase in infla-tion expectation and they are more likely to purchase a car (17% over thebaseline purchase risk). These conclusions are based on panel survey datafrom the New York Fed/RAND-American Life Panel household expectationssurvey. This dataset contains detailed actual consumer spending, includingboth durables and non-durables. This result is not necessarily at odds withthe results reported in this paper as the question that we interpret as theconsumer readiness to spend might for an economist strictly refer to durablegoods, while the consumer interpretation might differ and point to a largerset of goods. Actually, the contemporaneous correlation of the aggregatedindex of the readiness to spend with real total consumption (in logs) is 45%and reaches 60% with the fourth forward lag of real consumption (also inlogs). Also for the US, Armantier et al. (2015) verify whether consumers acton their beliefs about inflation expectation by combining survey data with afinancially incentivised experiment about investment. Basically, respondentsare asked each round to choose between two investments, one that dependson the inflation rate and another with a fixed rate. Each round the fixedreturn investment changes (increasing or decreasing) such that in the endone can establish what is the inflation expectations threshold for which re-spondents consider it is worth switching from one investment opportunityto the other. They find evidence that most repondents make their choicein accordance with economic theory, and the ones that do not, have lowereducation and financial and numerical literacy.

The average marginal effects at the ELB are higher than the ones out-side the ELB for each of the model specifications that we have employed.Moreover, the coefficient of the interaction term ”ELB x expected changein inflation” is positive (equal to 0.00083) and statistically significant at 1%level. This is actually in line with what one would expect: in a ELB en-vironment nominal interest rates are bounded from below (of course thisbound can be breached by negative interest rates, although if this happensone would not expect to get too far below zero), thus the change in the real

are interviewed also in the next round.12We also perform an exercise in which we estimate our model specifications on a reduced

sample, in which we eliminate statistical outliers, i.e. above 3 standard deviations and wefind that marginal effects are even stronger in this case.

15

interest rates depends almost solely on developments in inflation expecta-tions, making the latter even more relevant to the spending decision thanotherwise. The results support theories which argue in favour of using theinflation expectations channel as a way to stimulate aggregate consumption.

16

Averagemargina

leffe

cts

ELB=

0EL

B=1

Dem

ograph

icsEx

pectations

andfin

ancial

status

Interactions

Tim

edu

mmies

Cou

ntry

dummies

Macro

aggregates

Pseudo

R2

Observatio

ns

0.00260***

0.00302***

0.0037

2,240,624

(2.11e-05)

(8.86e-05)

0.00250***

0.00269***

X0.0107

1,927,382

(2.41e-05)

(9.46e-05)

0.00268***

0.00280***

XX

0.0432

1,793,108

(2.51e-05)

(9.40e-05)

0.00293***

0.00305***

XX

X0.0434

1,793,108

(2.73e-05)

(9.43e-05)

0.00290***

0.00325***

XX

XX

0.0456

1,793,108

(2.71e-05)

(0.000102)

0.00166***

0.00239***

XX

XX

X0.0788

1,793,108

(2.62e-05)

(9.79e-05)

0.00163***

0.00241***

XX

XX

XX

0.0794

1,793,108

(2.63e-05)

(9.81e-05)

Table2:

Prop

ensit

yto

spend:

averagemargina

leffe

cts,

Euro

Area

Note:

Stan

dard

errors

inpa

rentheses,

***p<

0.01,**

p<0.05,*p<

0.1.

The

tableshow

stheaveragemargina

leff

ectof

aun

itincrease

inthe

expe

cted

chan

gein

infla

tionon

theprob

ability

that

consum

ersareread

yto

spendgivencurrentcond

ition

s,estim

ated

bytheorderedlogitmod

el.

”Dem

ograph

ics”

includ

esage,

gend

er,e

ducatio

n,em

ploy

mentstatus,incom

e;”E

xpectatio

nsan

dfin

ancial

status

”includ

esexpe

ctations

ofindividu

alfin

ancial

situa

tion,

generalecon

omic

andun

employmentsit

uatio

nan

dconsum

ercurrentfin

ancial

status,i.e

.debtor

orno

n-debtor;”Interactions

”includ

espa

irwise

interactions

asfollo

ws:

expe

cted

chan

gein

infla

tion_

theexpe

cted

finan

cial

situa

tion,

expe

cted

chan

gein

infla

tion_

debt

status,

expe

cted

chan

gein

infla

tion_

employmentstatus,e

xpectedchan

gein

infla

tion_

income,

expe

cted

chan

gein

infla

tion_education;

”Tim

edu

mmies”

includ

esyear

dummies2004

to2015;

”Cou

ntry

dummies”;

”Macro

aggregates

”:SP

FGDP

uncertainty,

lend

ingrates,

logdisposab

leincome;

ELB

dummytakesvalue1from

June

2014

toJu

ly2015.

17

Tables 3 and 4 show in detail the average marginal effects associated withall controls that we have used in the estimation. All are statistically signifi-cant, except the marginal effect associated with the SPF macro uncertaintymeasure. Also, they are generally similar across the ELB and outside ELBregimes. Being older decreases the propensity to spend, while being a fe-male as opposed to being a man, having a higher level of education, higherincome, being employed as opposed to unemployed, or a non-debtor insteadof a debtor, all increase the consumer propensity to spend (see Table 3). A1 pp increase in lending rates decreases the probability of consumers beingready to spend by approximately 0.60 pp, approximately two and half timesmore than the effect of a 1 pp change in inflation expectations. This indicatesthat while inflation expectations are important for the consumption decision,nominal interest rates weigh even more. 1 percent increase in the aggregatedisposable income increases the probability to spend by approximately 11.4pp. This is also intuitive: with more money to spend the probability ofspending increases. Table 4 instead shows the average marginal effects forindividual expectations about individual or aggregate developments. Averagemarginal effects of the expectations of a better individual financial situationare always higher than expectations for a better general economic situationshowing that individual situation prevails over the general economic situa-tion in the consumption decision. Nevertheless, expecting that the generaleconomic situation gets a lot better relative to getting a lot worse increasesthe propensity to spend by approximately 13.70 pp, while expecting thatunemployment levels will increase sharply as opposed to falling, decreasesthe probability to spend by approximately 9 pp. Importantly, the impact ofthese expectations on the consumer probability to spend cannot be directlycompared with the impact of a 1 pp change in inflation expectations, thelatter is a continuous variable while the former are discrete variables. Evenso, such results highlight also the importance of structural policies aimed atincreasing prospects for long-term growth and lowering structural unemploy-ment in driving consumers′ willingness to spend.

18

Variables ELB=0 ELB=1Expected change in inflation 0.00163*** 0.00241***

(0.0000263) (0.0000981)ELB 0.0138***

(0.00171)SPF GDP uncertainty -0.00293 -0.00302

(0.00184) (0.0019)Lending rates -0.00598*** -0.00617***

(0.000138) (0.000144)Log disposable income 0.111*** 0.114***

(0.00433) (0.00448)Age (30-49) -0.0166*** -0.0172***

(0.000742) (0.000767)Age (50-64) -0.00544*** -0.00560***

(0.000794) (0.000818)Age (65+) -0.00182** -0.00187**

(0.000881) (0.000908)Gender (Female) -0.0154*** -0.0159***

(0.000486) (0.000505)Education (Secondary) 0.0253*** 0.0261***

(0.000604) (0.000632)Education (Further) 0.0371*** 0.0383***

(0.000723) (0.000759)Income (2nd Quartile) 0.0203*** 0.0211***

(0.000697) (0.000728)Income (3rd Quartile) 0.0348*** 0.0360***

(0.000721) (0.000758)Income(4th Quartile) 0.0609*** 0.0629***

(0.000757) (0.000812)Employment status (Employed) 0.0326*** 0.0337***

(0.000982) (0.00103)Debt status (non-debtor) 0.0351*** 0.0364***

(0.00101) (0.00106)Observations 1,793,108 1,793,108

Table 3: Full specification: average marginal effectsNote: Standard errors in parentheses, *** p<0.01, ** p<0.05, * p<0.1. The table shows theaverage marginal effect estimated with an ordered logit model. The results show the averagemarginal effect of a one unit increase in inflation expectation on the probability that con-sumers are ready to spend. For the discrete variables, the reported marginal effect shows thediscrete change from a base alternative, e.g. expected general economic situation gets a lotworse to another alternative, e.g. the expected general economic situation get a lot better.The ordered logit regression includes all groups of controls: ”Demographics”,”Other expec-tations and current financial status”, ”Interactions”, ”Time dummies”, ”Country dummies”,”Macro-aggregates”. 19

Variables ELB=0 ELB=1Expected financial situation (a littleworse)

0.0436*** 0.0455***(0.000993) (0.00106)

Expected financial situation (the same) 0.0772*** 0.0803***(0.000966) (0.00106)

Expected financial situation (a littlebetter)

0.0924*** 0.0960***(0.00116) (0.00127)

Expected financial situation (a lot better) 0.117*** 0.121***(0.0024) (0.00251)

Expected general economic situation (alittle worse)

0.0357*** 0.0371***(0.000817) (0.000863)

Expected general economic situation (thesame)

0.0507*** 0.0526***(0.000837) (0.000897)

Expected general economic situation (alittle better)

0.0878*** 0.0908***(0.000976) (0.00107)

Expected general economic situation (alot better)

0.136*** 0.140***(0.0028) (0.0029)

Expected general unemployment situation(fall slightly)

0.00833*** 0.00855***(0.00279) (0.00286)

Expected general unemployment situation(the same)

-0.0241*** -0.0248***(0.00276) (0.00283)

Expected general unemployment situation(increase slightly)

-0.0459*** -0.0473***(0.00276) (0.00284)

Expected general unemployment situation(increase sharply)

-0.0905*** -0.0936***(0.00279) (0.00289)

Observations 1,793,108 1,793,108

Table 4: Full specification: average marginal effects - continuation of Table3Note: Standard errors in parentheses, *** p<0.01, ** p<0.05, * p<0.1. The table shows theaverage marginal effect estimated with an ordered logit model. The results show the averagemarginal effect of a one unit increase in inflation expectation on the probability that con-sumers are ready to spend. For the discrete variables, the reported marginal effect shows thediscrete change from a base alternative, e.g. expected general economic situation gets a lotworse to another alternative, e.g. the expected general economic situation get a lot better.The ordered logit regression includes all groups of controls: ”Demographics”,”Other expec-tations and current financial status”, ”Interactions”, ”Time dummies”, ”Country dummies”,”Macro-aggregates”.

20

3.2 Impact on real consumption: VAR analysisIn the previous section we have estimated what is the impact of consumerinflation expectations on the propensity to spend, nevertheless, the questionof whether or not this relationship translates into an effect on actual con-sumption remains unanswered. For this purpose, we use a bi-variate VAR tomodel the relationship between aggregate real consumption13 and the averageestimated consumer propensity to spend from our ordered logit model. Webuild a quarterly the average estimated measure of the propensity to spend,based on the fitted probabilities obtained at the previous step, weighted byindividual consumer weights14.This measure reflects a probability and at thesame time it summarizes all the micro and macro level information that wehave included in our ordered logit specification. Therefore, we believe, abivariate VAR based solely on real total consumption and the propensityto spend measure is the most appropriate model and a multivariate analy-sis with other relevant macroeconomic controls in the VAR is not necessary.Figure 5 shows the impulse response functions based on a Cholesky decompo-sition, with the propensity to spend variable ordered first. Thus, we assumethat in the first period consumer propensity to spend does not react to ashock in consumption. Impulse response functions behave as expected: fol-lowing a shock, the propensity to spend increases in the first period and thenit slowly decays, while the propensity to spend does not react to a shock inconsumption. Consumption slowly increases after a shock in the probabilityto spend up until the sixth quarter, while it slowly decays afterwards, and itincreases and then slowly decreases after a consumption shock.

Based on this VAR, we implement a scenario of 1.0 pp expected increasein inflation implemented gradually as a 0.25 pp increase that takes place overfour consecutive quarters. To do so, we first translate the 0.25 pp increasein inflation expectations in terms of changes in the consumer propensity tospend. We follow Cameron and Trivedi (2010) and manually calculate thatfor 0.25 pp increase in inflation expectations the average marginal effect onthe propensity to spend would be at the ELB 0.2418 pp and outside the ELB0.1632 pp. You might notice that the figures that we report here are veryclose to the ones reported in Table 2, which we approximate with a 1 ppincrease in inflation expectations. This is however expected when working

13Source: Eurostat. Individual consumption expenditure Euro Area changing compo-sition, world concept, Households and non-profit institutions serving households, Euro,chain linked volumes, calendar and seasonally adjusted.

14These weights are based on the representativeness of a consumer in total populationand therefore control for variation in sample size across countries.

21

0

0.05

0.1

0.15

0.2

0.25

0.3

1 2 3

An

nu

al g

row

th r

ate

s d

iffe

ren

ce

Time (years)

ZLB

non-ZLB

Figure 5: Bivariate VAR: impulse response functionsImpulse response functions based on a Cholesky decomposition of a bivariateVAR(1) including consumer aggregate propensity to spend and log real total con-sumption (quarterly frequency), in this order.

with non-linear functions such as the logit, and it is basically saying that theeffect on the consumer propensity to spend is almost the same when increas-ing inflation expectations by 0.25 pp or 1 pp in one step.

Using a conditional forecast, we estimate two scenarios. First, a ELB sce-nario, where we fix the path of the aggregate consumer propensity to spendto increase by 0.2418 in each of the four consecutive quarters and then toremain constant for the next eight quarters. Second, an outside ELB sce-nario, where the only difference is the increase in the consumer propensityto spend, which we fix to 0.1632 in each of the four quarters. Figure 6 showsthe difference in the annual real consumption growth rates relative to a base-line scenario where inflation expectations are kept constant throughout theforecast horizon (12 quarters). As expected, the impact at the ELB wouldbe higher than outside the ELB. The peak of the difference in the annualgrowth rates would be reached in the second year amounting to 0.25 pp atthe ELB and 0.17 pp outside the ELB. We see this approach as a more rig-orous one, compared to previous attempts in the literature. Bachmann et al.(2015) also use a bivariate VAR in which they include the aggregate index for

22

buying conditions, which is actually the fraction of people saying that nowit is good moment to buy durable goods minus those reporting that now it isa bad moment to buy, and the HP-filtered natural logarithm of real durableconsumption expenditures. They then show impulse response functions forwhich they calibrate the size of the innovation corresponding to the aggregateindex such that it corresponds to the marginal effect of a 1pp point increasein inflation expectations as computed based on their micro-data analysis.Note however that what is computed from the micro-dataset is the marginaleffect on the probability of being ready to spend and not on the aggregatedindex. They of course find that the impact on almost zero outside the ELBand about -0.1% at the ELB, in line with their estimated marginal effects. Inorder to estimate the impact on real consumption, D′Acunto et al. (2015) per-form a ”back-of-the-envelope” calculation as they call it, and simply regressthe natural logarithm of real durable consumption expenditure on the endof quarter value of the average durable purchasing propensity and quarterlydummies. They find 4.8% higher real durable consumption if all Germanswould expect higher inflation as opposed to prices not changing. This ismuch higher than the effect that we obtain for the euro area. However, theresults go in the same direction and the differences should not be overstatedgiven that they measure a different scenario (an increase compared to 1 ppincrease, on durables instead of total consumption) and that their resultsconcern only Germany.

3.3 Country resultsCountry results15 confirm aggregate EA results: all countries except one,show a positive relationship between inflations expectation and the propen-sity to spend and the relationship becomes stronger at the ELB, see Figure7. The exception is Malta for which we find average marginal effects of -0.13outside the ELB and an even stronger negative one of -0.24 at the ELB.This is the only exception that we find among the EA countries. Of course,among the countries which exhibit a positive marginal effect, we do findheterogeneity as the range for the effects stands between 0.02 and 0.60 pp.Most countries show effects around the EA estimate of 0.16 pp outside theELB and 0.24 at the ELB. Spain and Portugal seem to have a weak, closeto zero relationship of inflation expectations and consumption, meaning thatfor consumers in these countries inflation expectations do not matter in the

15At country level we do not report results for Estonia, as information for the consumerinflation perceptions was available only at the beginning of the sample.

23

0

0.05

0.1

0.15

0.2

0.25

0.3

1 2 3

An

nu

al g

row

th r

ate

s d

iffe

ren

ce

Time (years)

ELB

non-ELB

Figure 6: Impact on real consumption of a gradual increase in consumerinflation expectations compared to a scenario of constant consumer inflationexpectationsThe difference in the annual growth rates results from conditional forecasts based on aVAR(1) which includes a quarterly aggregate measure of the consumer propensity to spendand the natural logarithm of real total consumption, in this order. We implement twoscenarios, which we then compare to a baseline scenario, . The first scenario, assumesfour consecutive quarters increase of 0.2418 in consumer propensity to spend at the ELB(equivalent to 0.25 pp increase in inflation expectations) and eight quarters following thelast increase constant consumer propensity to spend. The second scenario is similar exceptthat a 0.1632 pp increase is assumed outside the ELB. The baseline scenario assumes thatconsumer inflation expectations remain constant throughout the 12 quarters.

consumption decision. Finland exhibits a marginal effect close to 0.60 bothoutside and at the ELB, making the Finish consumers the most sensitive toinflation expectations impact on the consumption decision. For consumers inGermany and Slovakia the relationship between inflation expectations andconsumption has become significantly stronger at the ELB, with averagemarginal effects standing between 0.50 and 0.6016. German consumers havebeen experiencing deposit rates below 1% since mid-2012, which have beendecreasing to reach levels around 0.3% in July 2015 (which is the end of oursample). Just like what Ichiue and Nishiguchi (2015) reports for the Japaneseconsumers, it is possible that once confronted with an extensive period of lowand close to zero deposit interest rates German consumers have become moreaware of the importance of inflation expectations when deciding between cur-

16D′Acunto et al. (2015) have indeed reported average marginal effects between 6 and9 pp for an increase in inflation, nevertheless these are not directly comparable with ourresults as their marginal effects are based only on qualitative information and reflect theeffect of an increase in inflation as oposed to prices remaining at the same levels.

24

rent consumption relative to saving and consuming in the future. Also notethat the period that we have chosen to define the ELB17 coincides almostperfectly with this period. Finish consumers, on the other hand, have con-tinued to experience deposit rates around 1% during the ELB period.

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

AT BE CY DE EL ES FI FR IT LT LU LV MT NL PT SI SK

ZLB=0

ZLB=1

Figure 7: Country average marginal effectsThis figure presents average marginal effects based on the ordered logit specificationwhich includes the following groups of controls: ”Demographics”,”Other expec-tations and current financial status”, ”Interactions”, ”Time dummies”, ”Countrydummies”, ”Macro-aggregates”. The ”Macro-aggregates” group does not includeSPF GDP uncertainty, as we did not have country specific information. For SIwe report results based on an ordered logit specification which excludes ”Macro-aggregates” due to convergence issues. All average marginal effects are statisticallysignificant at 1% or 5% level, except average marginal effects at the ELB for Greeceand Portugal.

4 ConclusionIn this paper we analyse the relationship between consumer inflation expec-tations and consumption. Although this relationship stands at the basis of

17We have chosen June 2014 as the beginning of the ELB period, as at that time therate of the deposit facility became negative.

25

most standard macroeconomic models, there are very few papers that chal-lenge this relationship empirically. We investigate this relationship for theEuro Area and for most of its constituent countries. To the best of ourknowledge, we are the first to provide such a comprehensive view on thiscentral issue for understanding Euro Area consumer behaviour. We benefitfrom a very rich micro dataset provided by the EU Consumer Survey whichprovides information about consumers expectations concerning different eco-nomic areas and developments of the individual financial and employmentsituation, and which can be broken down by a detailed set of demographiccharacteristics. Most important for our analysis, the survey includes quan-titative consumer expectations and perceptions and gives information aboutthe consumer intentions to spend. Our conclusions are based on 1,793,108observations18, which were carefully collected to reflect EA population overapproximately a 12 years period, from May 2003 to July 2015. As the sur-vey does not include information over actual consumption, we perform atwo-steps analysis. First, we estimate the relationship between consumer in-flation expectations and the propensity to spend based on the survey data.To do so, we use an innovative measure of the consumer′s expected changein inflation, which reflects simply the difference between consumers expecta-tions about future inflation and their individual perceptions about currentinflation. We find this measure extremely relevant as it reflects the fact thatwhen changing their spending intentions, consumers do not only take intoaccount expected future inflation, but they consider expected inflation levelsrelative to the currently perceived level of inflation. The measure is availableat an individual level which makes it even more attractive to use. In a secondstep, using macro level information, we translate this relationship to a re-lationship between consumer inflation expectations and actual consumptionusing a simple VAR framework.

Euro Area consumers behave in line with economic theory and when theyexpect higher inflation, all other factors held constant, they adjust positivelytheir intention to spend at the current moment. The result is robust acrossseveral specifications, in which we gradually control for demographics, otherconsumer expectations, financial situation, interactions of consumer expectedchange in inflation with other controls, time dummies, country dummies andmacro-aggregates. Our pooled results suggest that for a 1 pp expected in-crease in inflation the consumer probability to spend increases between 0.16pp and 0.33 pp. This is confirmed by country level results, where for all coun-

18This the number of observations that we are left with after eliminating those belongingto consumers that did not reply to the full question set that we use in this analysis.

26

tries except Malta we find a positive relationship between consumer inflationexpectations and propensity to spend, though there is some notable hetero-geneity. This result comes to complement existing literature on consumerbehaviour using survey data: Bachmann et al. (2015), Ichiue and Nishiguchi(2015), D′Acunto et al. (2015), Burke and Ozdagli (2013), Armantier et al.(2015). While our results for the EA and most of its constituent countriesconfirm previous findings for Germany and Japan, they differ to recent find-ings for the US, like the ones in Bachmann et al. (2015). Nonetheless, evenfor the US, Armantier et al. (2015) finds that consumers act upon their ex-pectations and in line with theory in an investment decision, while Bachmannet al. (2015) and Burke and Ozdagli (2013) conclude they do not in a durablesconsumption decision.

Another important result of our study is that the relationship betweenconsumer inflation expectations and consumption becomes stronger at theELB. It seems that once confronted with a lower bound on nominal rates,consumers become even more aware of the current and future real value oftheir money. In expectation of higher inflation, real interest rates can be-come even negative in the proximity of the ELB and thus the consumers′willingness to consume now instead of saving for future consumption be-comes stronger. We see this both in our pooled results and our countryspecific results. When comparing a scenario in which inflation expectationsincrease by 0.25 pp for four consecutive quarters relative to a scenario whereinflation expectations remain unchanged, we find that annual average realconsumption growth rate is higher with 0.25 pp at the ELB and with 0.17pp outside the ELB.

Overall, from a monetary policy perspective, our micro analysis of con-sumer spending behaviour provides strong support to central bank concernsabout a drop in inflation expectations because such developments have thepotential to weaken aggregate demand further by reducing consumers′ readi-ness to spend. Moreover, perceptions about weak overall aggregate demandlinked to weak consumption may equally restrain firms′ incentives to invest.Our results also highlight the importance of non-standard monetary policymeasures in helping to stabilise inflation expectations and thereby providingsupport to economic recovery.

27

ReferencesArmantier, O., Bruine de Bruin, W., Topa, G., Klaauw, W. and Zafar, B.(2015). Inflation expectations and behavior: Do survey respondents acton their beliefs?, International Economic Review 56(2): 505–536.

Bachmann, R., Berg, T. O. and Sims, E. R. (2015). Inflation expectations andreadiness to spend: cross-sectional evidence, American Economic Journal:Economic Policy 7(1): 1–35.

Biau, O., Dieden, H., Ferrucci, G., Friz, R. and Lindén, S. (2010). Consumers′quantitative inflation perceptions and expectations in the euro area: anevaluation, Technical report, Mimeo.

Burke, M. A. and Ozdagli, A. (2013). Household inflation expectations andconsumer spending: evidence from panel data.

Cameron, A. C. and Trivedi, P. K. (2010). Microeconometrics using stata,Rev. ed. Stata press College Station, TX.

D′Acunto, F., Hoang, D. and Weber, M. (2015). Inflation expectations andconsumption expenditure, Chicago Booth Global Markets Working PaperSeries .

Davidson, R. and MacKinnon, J. G. (2004). Econometric theory and methods,Oxford University Press New York.

Eggertsson, G. B. (2006). The deflation bias and committing to being irre-sponsible, Journal of money, credit, and Banking 38(2): 283–321.

Eggertsson, G. B. and Woodford, M. (2003). Zero bound on interest ratesand optimal monetary policy, Brookings Papers on Economic Activity(2003(1)): 139–233.

Ichiue, H. and Nishiguchi, S. (2015). Inflation expectations and con-sumer spending at the zero bound: Micro evidence, Economic Inquiry53(2): 1086–1107.

Jung, T., Teranishi, Y. and Watanabe, T. (2005). Optimal monetary policyat the zero-interest-rate bound, Journal of Money, credit, and Banking37(5): 813–835.

Krugman, P. R., Dominquez, K. M. and Rogoff, K. (1998). It’s baaack:Japan’s slump and the return of the liquidity trap, Brookings Papers onEconomic Activity (1998(2)): 137–205.

28

Souleles, N. S. (2004). Expectations, heterogeneous forecast errors, and con-sumption: Micro evidence from the michigan consumer sentiment surveys,Journal of Money, Credit, and Banking 36(1): 39–72.

29