Mobile Edition
Print
Get the Mag
Weekly UpdatesINTRODUCTION
The existence of divergences between economic agents' subjective opinions concerning current price developments and statistical inflation indicators creates a challenge for monetary policy makers. In particular, a gap between the two variables may lower the credibility of monetary policy, influencing expectations about future price changes and distorting decisions on prices and wages.
The empirical literature suggests that a rapid increase in prices of frequently purchased products was one of the principal factors explaining increased consumers' inflation perception in the euro area member states after the launch of the new currency. A similar phenomenon where the price increases of frequently bought goods were significantly higher than the overall inflation rate characterised the period of Polish accession to the European Union (EU). The phenomenon was mainly due to the rapid growth of food prices, which exceeded consumer price index (CPI) inflation by 2.8 percentage points on average in the year of accession, 2004. Fluctuations of relative prices of this kind offer a good basis for analysing the impact of price changes of frequently bought goods and services on Polish consumers' inflation perception, which is the aim of this study. In order to do this I calculate different proxies of price changes of frequently purchased products and examine their correlation with survey-based measures of consumers' inflation perception.
The paper is structured as follows: the next section reviews the empirical literature on the impact of price changes of frequently bought items on consumer inflation perception in the Economic and Monetary Union member states after the launch of the euro; the subsequent section describes the data on Polish consumers' inflation perceptions and price changes of frequently purchased goods and services used in the study and then the later section presents the results of the analyses; and the last section concludes.
LITERATURE REVIEW
The empirical literature on the impact of price changes of products frequently bought on consumer inflation perceptions that have grown significantly since the launch of the euro in 2002. The influence of the common European currency on consumer inflation perceptions led many researchers to examine the reasons for a sizeable gap between price changes subjectively perceived by consumers and official inflation measures. (1) Qualitative survey data collected by the European Commission suggest that the increase in the perceived inflation began in early 1999; see ECB (2002, 2003, 2005, 2007) for details. Although it corresponds to the birth of the Economic and Monetary Union, it seems that the performance of survey data in 1999-2001 had to do mainly with the increase in inflation itself from approximately 1% in the first half of 1999 to 2.9% at the beginning of 2001. However, in 2002, when the euro was put into circulation, there was a dramatic increase in the price changes perceived by consumers, in spite of the stabilisation of the Harmonised Index of Consumer Prices (HICP) annual growth trend around 2.2%. Since then, inflation perception has moderated, gradually reducing the gap with actual price dynamics. Existing studies examining the euro-related jump in the perceived inflation indicate that European consumers' focus on changes in prices of frequently bought items (such as food, petrol, etc.) constitutes one of the possible explanations of the inflation perception bias.
Dziuda and Mastrobuoni (2006) develop and estimate a model of imperfect information, in which consumers are fully rational but, after the cash changeover, remember and compare prices with some noise. Model results are in line with the observed phenomenon consisting of prices of cheaper and more frequently purchased goods being subject to higher growth after the changeover relative to prices of other goods. As the model predicts, this might have happened due to imperfect price transparency, especially among smaller shops.
ECB (2003) introduces a price index of the 'out-of-pocket' expenditure. It represents 35-40% of the all euro area HICP items and includes non-durable goods and daily consumer services, that is, food, beverages, tobacco, nondurable household goods, transport services, fuel, postal services, hotels, restaurants, cafes and hairdressing. Even if a price index of this kind does not correlate much more strongly with inflation perception than the overall HICP inflation on the whole sample, (2) the jump in the subjective assessment of current price changes at the beginning of 2002 could be partly explained by the 'out-of-pocket' measure of inflation. (3)
Experiences of some individual euro-area economies provide even clearer indication of a substantial weight attached by consumers to price developments in the goods and services they buy more frequently. Del Giovane and Sabbatini (2004, 2005) show that the rise in perceived inflation in the period of the currency changeover can be largely explained by relatively large increases in prices of frequently purchased products, combined with the media's unusually extensive coverage of price developments and the perception of a substantial loss of purchasing power, especially on the part of the least well-off households. The selected group of frequently purchased products contains food products, tobacco products, everyday household goods such as detergents, soaps, etc., newspapers, some services including municipal transport and railways, postal and banking services, restaurants and coffee shops, recreational and cultural services, and fuels, accounting for around 40% of the consumer basket used to derive the official measure of inflation.
Spanish experiences are similar. As shown by Alvarez Gonzalez et al. (2004), inflation perceived by consumers seems to be driven by price changes of frequently consumed goods and services, such as foodstuffs, beverages and tobacco, energy, urban public transport, restaurants, bars and cafes, communication, cultural services, newspapers. These items account for 50% of the CPI consumer basket.
Fluch and Stix (2005) show that, after the currency changeover, prices of frequently bought items in Austria did rise more sharply than prices of other goods in the market basket and that a higher, unweighted, share of goods became more expensive. Moreover, price rises seemed to be perceived by consumers more strongly than price reductions. These two factors account for a part of the gap between actual and perceived inflation. In a similar study Stix (2006) finds that Austrian consumers confronted with prices of frequently purchased goods, which rose faster than the overall inflation, perceived a higher rate of inflation. He also identifies additional factors contributing to the inflation perceptions' divergence from official inflation measures in the euro area. The first one is a substantial role of inflation expectations prior to the launch of the euro. More precisely, a psychological phenomenon of the 'expectancy confirmation in spite of the disconfirming evidence' may be noted: persons who believed that prices would increase after the cash changeover have later perceived a significantly higher rate of inflation than other consumers who did not expect price increases. (4) A similar effect was identified in Finland, where consumers having advance beliefs of a euro-induced increase in prices paid more attention to prices of goods and services that did in fact rise, for example, daily food purchases, while ignoring commodities, whose prices declined (Kurri, 2006). The second factor disconnecting official inflation indices and consumers' perceptions in Austria highlighted by Stix (2006) is the use of the old currency for price comparisons. Individuals who mentally convert euro prices into Austrian schilling prices and thus compare actual prices with prices prior to 2002, especially those who do so imprecisely, perceive higher price growth.
DESCRIPTION OF DATA
As a measure of Polish consumer inflation perception I use a range of indicators, based on the qualitative survey results concerning price developments in the course of the previous 12 months. (5) The survey question, included in the European Commission Consumer Survey conducted in Poland by GfK Polonia since 2001, has the following form: 'In your opinion, is the price level now compared to that twelve months ago: (1) much higher; (2) moderately higher; (3) a little higher; (4) about the same; (5) lower; (6) difficult to say'. Responses to this question (see Figure la) may be quantified with the use of a probability method consistent with the canonical Carlson and Parkin (1975) approach; see Lyziak (2005) or Lyziak and Stanislawska (2006) for details. However, due to the fact that this method assumes backward unbiasedness in terms of CPI inflation, the quantification procedure imposes a relatively high correlation between its outcomes and CPI figures (Figure lb). Alternatively, assessing respondents' opinions, one may use balance statistics, defined as differences between weighted or unweighted proportions of respondents (Figure 1c-f). They do not measure perceived inflation directly, but at the same time they are not influenced by the assumptions imposed in quantification algorithms. For this reason in the present study I use balance statistics, employing four different measures of this kind. The first two are unweighted statistics: [BS.sub.1] is the difference between proportions of respondents noticing increase in prices and their decrease, whereas [BS.sub.2] is the difference between proportions of respondents noticing increase in prices and their stabilisation or decrease. The third balance statistics, [BS.sub.3], is a weighted one, frequently used--for example, Del Giovane and Sabbatini (2004, 2005), ECB (2002, 2003, 2005)--attaching weight 1 to the proportion of respondents perceiving that the prices now are much higher than 12 months ago, 1/2 to those claiming that prices are moderately higher, 0 to those declaring that prices are a little higher, -1/2 to the fraction of respondents declaring stabilisation of prices and -1 to those noticing their fall. The fourth balance statistics, [BS.sub.4], summarises the survey results in the way consistent with the normal distribution of perceived inflation, as assumed in the probability quantification method. In this method perceived inflation is a product of balance statistics [BS.sub.4] and a scaling factor, that is, a range of implied perceived price changes of the respondents claiming that prices are about the same relative to its level 12 months ago; see: Lyziak (2005), Lyziak and Stanistawska (2006).
FEED