Skip to content
Licensed Unlicensed Requires Authentication Published by De Gruyter January 29, 2020

Pro-rich inflation in Europe: Implications for the measurement of inequality

  • Eren Gürer and Alfons Weichenrieder EMAIL logo
From the journal German Economic Review

Abstract

This paper studies the distributional consequences of a systematic variation in expenditure shares and prices. Using European Union Household Budget Surveys and Harmonized Index of Consumer Prices data, we construct household-specific price indices and reveal the existence of a pro-rich inflation in Europe. Over the period 2001–15, the consumption bundles of the poorest deciles in 25 European countries have, on average, become 11.2 percentage points more expensive than those of the richest deciles. We find that ignoring the differential inflation across the distribution underestimates the change in the Gini (based on consumption expenditure) by almost up to 0.04 points. Cross-country heterogeneity in this change is large enough to alter the inequality ranking of numerous countries. The average inflation effect we detect is almost as large as the change in the standard Gini measure over the period of interest.

JEL Classification: D31

Acknowledgment

We thank Eurostat for granting access to the microdata on European Household Budget Surveys and Harmonized Index of Consumer Prices, Ariana Gilbert-Mongelli and Jeremy Edwards as well as the participants of the 11th RGS Doctoral Conference, the 2018 ZEW Public Finance Conference, the 2018 Verein für Socialpolitik annual conference, and the 74th IIPF Annual Congress for providing valuable comments. We are grateful for the advice by two referees. This paper is part of the research program of the LOEWE Center ‘Sustainable Architecture for Finance in Europe’ (SAFE). None of the conclusions of this paper may be attributed to Eurostat, the European Commission or to any of the national statistical authorities whose date have been used.

Appendix A

A.1 Data preparation

A.1.1 Construction of consistent categories from HBSs

As mentioned in Section 3, HBSs contain consumption expenditure data in many aggregation levels which are represented by the number of digits in the variable codes, ranging from the 2- to 5-digits. For example, the variable representing the consumption expenditure on “Food and Non-alcoholic Beverages” is 2-digit. Sub-categories of the 2-digit level of aggregation, naturally, include 3-digit categories (e. g., food, non-alcoholic beverages). Although it would be ideal for the sake of precision, it is not possible to employ categories more disaggregated than the 3-digit ones used because there is a significant amount of missing price data.

A serious flaw in 3-digit categories is that in approximately 9 % of the observations, 3-digit categories do not add up to their corresponding 2-digit aggregate. If, for example, the sum of the 3-digit category “Food” and the 3-digit category “Non-alcoholic Beverages” do not add up to their 2-digit aggregate “Food and Non-alcoholic Beverages”, it is natural to expect that sum of all 3-digit categories would not add up to the total consumption expenditure. One approach would be to work with only the twelve 2-digit categories of the survey. However, because using the most disaggregated categories whenever possible increases precision, we deal with this issue by scaling up the 3-digit categories proportionately such that they will add up to their 2-digit aggregate.

In the HICP data, unfortunately, prices are not fully available on the 3-digit level. Therefore, we investigate every 3-digit category to see the extent of missing price data. If the number of missing data points is too large such that an imputation could create meaningless results, we use the 2-digit aggregate for that particular strand. For example, just like HBS data, the HICP data splits the 2-digit aggregate “Education” into 3-digit categories such as “Primary Education”, “Secondary Education”, etc., and reports prices both on the 2-digit and 3-digit level. If price data on the 3-digit level is missing in a significant number of country-year observations, we collapse that strand to its 2-digit aggregate and only use the 2-digit level in the analysis. If there are relatively few missing observations, then we impute them. Details of the imputation procedure are provided in the next section. Table 4 presents the 30 expenditure categories and their codes. Table 5 contains the list of 25 EU countries analyzed, along with their abbreviations.

Table 4

Expenditure Categories and Variable Codes.

Variable CodeExpenditure Category
00All Consumer Goods
011Food
012Non-alcoholic beverages
021Alcoholic beverages
022Tobacco
031Clothing
032Footwear
041Actual rentals of housing
042Imputed rentals of housing
043Maintenance and repair of the dwelling
044Water supply and misc. services
045Electricity, gas and other fuels
051Furniture and furnishings, carpets and other floor coverings
052Household textiles
053Household appliances
054Glassware, tableware and household utensils
055Tools and equipment for house and garden
056Goods and services for routine household maintenance
061Medical products, appliances and equipment
062Out-patient services
063Hospital services
071Purchase of vehicles
072Operation of personal transport and equipment
073Transport services
081Postal services
082Telephone and telefax services and equipment
09Recreation and culture
10Education
111Catering services
112Accommodation services
12Misc. goods and services

Table 5

List of EUCountries and Their Abbreviations.

AbbreviationCountryAbbreviationCountry
BEBelgiumITItaly
BGBulgariaLTLithuania
CYCyprusLULuxembourg
CZCzech RepublicLVLatvia
DEGermanyMTMalta
DKDenmarkPLPoland
EEEstoniaPTPortugal
ELGreeceRORomania
ESSpainSESweden
FIFinlandSISlovenia
FRFranceSKSlovakia
HUHungaryUKUnited Kingdom
IEIreland

Table 6

Simple Mean of “Imputed Rentals of Housing” by Decile and Country.

Decile12345678910
Average13,4 %14,5 %15,2 %15,7 %15,5 %15,6 %15,5 %15,3 %14,8 %12,7 %
BE11,8 %15,5 %16,9 %16,2 %16,0 %14,9 %14,7 %13,9 %12,0 %8,1 %
BG18,5 %17,7 %18,6 %19,4 %19,6 %23,1 %25,7 %27,9 %30,3 %27,6 %
CY20,7 %20,0 %20,0 %20,02 %17,5 %18,0 %17,3 %17,0 %15,2 %13,9 %
CZ0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %
DE5,2 %9,2 %12,2 %14,2 %15,7 %15,9 %16,2 %16,5 %15,4 %11,0 %
DK4,6 %9,7 %9,7 %11,0 %11,3 %12,2 %11,5 %12,3 %11,6 %10,5 %
EE19,2 %18,3 %16,6 %17,1 %16,8 %15,8 %14,2 %12,1 %10,5 %9,3 %
EL17,7 %14,8 %16,9 %17,7 %17,2 %17,0 %17,6 %16,2 %16,9 %15,6 %
ES21,6 %22,8 %22,8 %22,3 %22,6 %21,5 %20,5 %19,8 %19,5 %16,5 %
FI11,3 %13,7 %16,9 %16,7 %18,5 %18,6 %18,2 %16,4 %15,7 %15,1 %
FR9,2 %12,9 %12,9 %13,7 %14,3 %14,6 %15,2 %14,2 %15,3 %13,9 %
HU14,9 %17,4 %18,9 %19,8 %20,1 %20,1 %20,6 %20,0 %20,0 %16,2 %
IE9,9 %12,0 %15,7 %15,6 %15,7 %17,0 %16,6 %16,3 %18,3 %16,5 %
IT21,6 %23,8 %23,7 %23,7 %23,7 %23,0 %23,2 %21,9 %20,3 %14,9 %
LT19,1 %15,8 %16,0 %16,7 %16,4 %18,4 %20,6 %21,1 %21,3 %15,5 %
LU18,2 %21,9 %21,9 %24,2 %23,5 %23,5 %22,5 %20,7 %21,9 %19,8 %
LV13,3 %12,0 %10,9 %10,9 %8,9 %8,4 %9,6 %8,3 %7,0 %6,5 %
MT0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %
PL19,7 %18,3 %17,4 %17,3 %16,4 %16,0 %14,9 %14,2 %12,5 %8,8 %
PT21,5 %22,1 %22,8 %22,1 %20,5 %18,7 %18,9 %17,1 %15,5 %14,6 %
RO15,3 %16,3 %18,1 %19,3 %20,8 %22,8 %14,2 %25,5 %27,1 %27,3 %
SE7,3 %14,3 %14,5 %17,2 %17,8 %16,3 %16,2 %17,3 %15,3 %11,6 %
SI18,7 %20,0 %20,8 %22,1 %20,1 %19,7 %19,1 %17,4 %16,9 %14,6 %
SK14,1 %14,7 %14,0 %14,5 %14,2 %14,4 %14,1 %14,1 %14,2 %10,3 %
UK0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %0,00 %
  1. Note: Imputed rentals are not reported in CZ, MT and UK.

A.1.2 Harmonised index of consumer price data

While preparing the HICP data, we encountered two main challenges. First, as it was mentioned in the previous section, there are missing country-year price observations after constructing 30 expenditure categories. We therefore imputed missing country-year observations as follows:

  1. 3-digit category “Hospital Services” (063) data is missing for Slovenia in 2001; it is proxied using Slovenia’s 2002 price.

  2. 3-digit category “Hospital Services” (063) data is missing for Estonia between 2001 and 2003; these data are proxied using Estonia’s 2004 price.

  3. 3-digit category “Hospital Services” (063) data is missing for Hungary between 2001 and 2006; these data are proxied using Hungary’s 2007 price.

  4. 3-digit category “Hospital Services” (063) data is missing for Slovakia and Romania between 2001 and 2015; they are proxied using the price of their 2-digit aggregate “Health” in the respective country and year.

  5. 3-digit category “Telephone and Telefax Services and Equipment” (082) data is missing for Latvia in 2014; it is proxied using Latvia’s 2013 price.

It is important to note that the fraction of proxied country-year price observations is a mere 0.002 %. Moreover, the mean expenditure fraction of main problematic 3-digit category “Hospital Services” (063) across countries is about 0.02 %. Therefore, we are confident that proxying missing price observations does not have a serious impact on our results.

A final challenge to deal with is the 3-digit category “Imputed Rentals of Housing” (042). Naturally, HICP does not provide any information on prices of this category. One immediate resolution for this problem would be proxying the prices of this category with prices of the 3-digit category “Actual Rentals of Housing” (041). However, by definition, values in “Imputed Rentals of Housing” (042) do not imply an actual expenditure; values purely represent the rental price of the dwelling as if it is consumed by its owner. Therefore, a price increase would not imply a decline in the real expenditure of the household who owns the dwelling. In order to neutralize the effect of this category, we assume that the price of “Imputed Rentals of Housing” (042) has not changed with respect to the base year. The fraction “Imputed Rentals of Housing” (042) in total expenditure across deciles is not large enough to explain away our findings. Table 6 presents the fractions. For this reason, we believe, this assumption is not biasing our results. Finally, one should note that imputed rentals would not be available for three countries.

Appendix B

B.1 A simple model using Stone-Geary preferences

We construct a simple model based on the stylized fact that poorer households spend a higher fraction of their income on necessities than on luxuries. Moreover, we provide a simple intuition for this fact. Following that logic, the model implies that richer households can more easily substitute away certain goods in response to price increases.

Consider a representative household with a Stone-Geary utility function in an economy which consists of two goods: necessities (qn) and luxuries (ql). The household maximizes

(7)U=βnlnqnγn+βlln(ql)

over qn and ql such that the budget constraint pnqn+ql=y is satisfied. In this specification, γn indicates the subsistence parameter of necessities (subsistence parameter of luxuries is assumed to be zero). Let pn be the price of necessities relative to luxuries and y is the nominal income of the household. Finally, βn and βl are preference parameters. We assume βn>0, βl>0 and βn+βl=1.

Typical FOCs yield the following demand functions:

(8)qn=γn+βnpnypnγn
(9)ql=βl(ypnγn)
These demand functions are simple and intuitive. A household first sets aside enough income to purchase a subsistence level of necessities, then allocates the rest of its income, depending on prices and preference parameters. Following this, we derive the optimal expenditure share of necessities in total demand.

(10)sn=γnpn+βnypnγnγnpn+βnypnγn+βl(ypnγn)pl

Given βl>0, an increase in income causes a higher increase in denominator due to higher scaling factor (βn+βl>βn). Hence, sny<0. Intuitively, each household has to spend on necessities at least as much as the subsistence level, which is a constant. As the subsistence level-income ratio decreases in income, richer households spend a lower fraction on necessities, given that luxury goods are at least marginally desirable (βl>0).

Another implication of the model is that richer households have a higher ability to substitute away certain goods in response to a price increase. Let the price elasticity of demand of necessities be as follows:

(11)ϵn=qn/qnpn/pn=qnpnqnpn=βnypnγn+βn(ypnγn)=βnypnγn+βnyβnpnγn

Note that ϵn<1. Hence, ϵny>0, which means that the price elasticity of demand for necessities is increasing in income. The intuition behind this result is as simple as the previous one. The household can only substitute the expenditures that is left after setting aside the subsistence level. Given that poorer households are left with a lower amount after paying for the subsistence, their ability to substitute is lower compared to richer households.

Appendix C

C.1 Effective inflation rates by country

Figure 8 Effective Inflation Rates of Poorest and Richest Deciles over Time by Country.
Figure 8 Effective Inflation Rates of Poorest and Richest Deciles over Time by Country.
Figure 8 Effective Inflation Rates of Poorest and Richest Deciles over Time by Country.
Figure 8 Effective Inflation Rates of Poorest and Richest Deciles over Time by Country.
Figure 8

Effective Inflation Rates of Poorest and Richest Deciles over Time by Country.

Appendix D

D.1 Inequality analysis

Table 7

Baseline Variance of Log Expenditure Measures by Country.

CountryVariance of Log ExpenditureRankAdjusted Variance of Log ExpenditureRank
Portugal0.37110.3713
Estonia0.36920.4221
Malta0.34930.3942
United Kingdom0.31440.3684
Latvia0.31250.3605
Italy0.30360.30210
Cyprus0.28470.3506
Lithuania0.28280.3297
Greece0.28090.3069
France0.273100.28712
Spain0.262110.27413
Finland0.261120.29611
Romania0.254130.3088
Luxembourg0.250140.25817
Poland0.243150.27014
Ireland0.238160.26816
Germany0.231170.24019
Bulgaria0.226180.26915
Belgium0.215190.22221
Slovenia0.206200.25118
Sweden0.198210.21522
Hungary0.196220.23420
Denmark0.168230.18724
Slovakia0.155240.18125
Czech Republic0.146250.19123

Figure 9 Absolute Change in Variance of Log Expenditure after applying Household-specific Price Indices, 2001–15.
Figure 9

Absolute Change in Variance of Log Expenditure after applying Household-specific Price Indices, 2001–15.

References

Amble, N., and K. Stewart. 1994. “Experimental Price Index for Elderly Consumers.” Monthly Lab. Rev. 117:11. Search in Google Scholar

Argente, D., and M. Lee. 2017. “Cost of Living Inequality During the Great Recession“. Kilts Center for Marketing at Chicago Booth – Nielsen Dataset Paper Series 1-032. Available at SSRN: https://ssrn.com/abstract=2567357. Search in Google Scholar

Arndt, C., S. Jones, and V. Salvucci. 2015. “When do Relative Prices Matter for Measuring Income Inequality? The Case of Food Prices in Mozambique.” Journal of Economic Inequality 13(3): 449. 10.35188/UNU-WIDER/2014/850-6Search in Google Scholar

Atkinson, A. B. 1970. “On the Measurement of Inequality.” Journal of Economic Theory 2(3): 244–263. 10.1016/0022-0531(70)90039-6Search in Google Scholar

Atkinson, A. B., T. Piketty, and E. Saez. 2011. “Top Incomes in the Long Run of History.” Journal of Economic Literature 49(1): 3–71. 10.3386/w15408Search in Google Scholar

Beck, U. 2015. Keep it Real: Measuring Real Inequality Using Survey Data from Developing Countries. (No. 2015/133) WIDER Working Paper. 10.35188/UNU-WIDER/2015/022-5Search in Google Scholar

BDEW. 2019. BDEW-Strompreisanalyse Juli 2019. Berlin: Haushalte und Industrie, Bundesverband der Energie- und Wasserwirtschaft e.V.. 23 July 2019. Search in Google Scholar

Cage, R. A., T. Garner, and J. Ruiz-Castillo. 2002. Constructing Household Specific Consumer Price Indexes: An Analysis of Different Techniques and Methods. Washington DC: US Bureau of Labor Statistics. BLS Working Paper 354. Search in Google Scholar

Cravino, J., and A. A. Levchenko. 2017. “The Distributional Consequences of Large Devaluations.” American Economic Review 107(11): 3477–3509. 10.3386/w23409Search in Google Scholar

Crawford, I., and Z. Smith. 2002. “Distributional Aspects of Inflation.” IFS Commentary 90. Search in Google Scholar

Engel, E. 1857. “Die Produktions- und Konsumptionsverhältnisse des Königreichs Sachsen.” Zeitschrift des Statistischen Bureaus des Königlich Sächsischen Ministeriums des Innern 8:1–54. Search in Google Scholar

Eurostat. 2017. Gini Coefficient of Equivalised Disposable Income. Retrieved from. http://ec.europa.eu/eurostat/data/database on 14.02.2018. Search in Google Scholar

Faber, B., and T. Fally. 2017. Firm Heterogeneity in Consumption Baskets: Evidence from Home and Store Scanner Data. NBER Working Paper 23101. 10.3386/w23101Search in Google Scholar

Garner, T. I., D. S. Johnson, and M. F. Kokoski. 1996. “An Experimental Consumer Price Index for the Poor.” Monthly Lab. Rev. 119:32. Search in Google Scholar

Garner, T. I., J. Ruiz-Castillo, and M. Sastre. 2003. “The Influence of Demographics and Household-specific Price Indices on Consumption-based Inequality and Welfare: A Comparison of Spain and the United States.” Southern Economic Journal 70(1): 22–48. Search in Google Scholar

Goni, E., H. Lopez, and L. Serven. 2006. Getting Real About Inequality: Evidence from Brazil, Colombia, Mexico and Peru. Washington DC: World Bank. Policy Research Working Paper WPS3815. Search in Google Scholar

Hagemann, R. P. 1982. “The Variability of Inflation Rates Across Household Types.” Journal of Money, Credit and Banking 14(4): 494–510. 10.2307/1991657Search in Google Scholar

Hobijn, B., and D. Lagakos. 2005. “Inflation Inequality in the United States.” Review of Income and Wealth 51(4): 581–606. 10.1111/j.1475-4991.2005.00170.xSearch in Google Scholar

ILO/IMF/OECD/UNECE/Eurostat/The World Bank. 2004. Consumer Price Index Manual: Theory and Practice. Geneva: International Labour Office. Search in Google Scholar

Jaravel, X. 2017. The Unequal Gains from Product Innovations: Evidence from the US Retail Sector. Harvard University. Working Paper. 10.1093/qje/qjy031Search in Google Scholar

Leicester, A., C. O’Dea, and Z. Oldfield. 2008. “The Inflation Experience of Older Households.” IFS Commentary 106. 10.1920/co.ifs.2009.0111Search in Google Scholar

Michael, R. T. 1979. “Variation Across Households in the Rate of Inflation.” Journal of Money, Credit and Banking 11(1): 32–46. 10.3386/w0074Search in Google Scholar

Muellbauer, J. 1974. “Prices and Inequality: the United Kingdom Experience.” Economic Journal 84(333): 32–55. 10.2307/2230482Search in Google Scholar

Murphy, E., E. Garvey, and S. O’Neill. 2008. Cost of Living for Different Social Groups in Ireland, Combat Poverty Agency. Dublin. Working Paper Series, 08/03. Search in Google Scholar

Neuhoff, K., S. Bach, J. Diekmann, M. Beznoska, and T. El-Laboudy. 2013. “Distributional Effects of Energy Transition: Impacts of Renewable Electricity Support in Germany.” Economics of Energy and Environmental Policy 2(1): 41–54. 10.5547/2160-5890.2.1.3Search in Google Scholar

OECD. 2006. The Distributional Effects of Environmental Policy. Paris: OECD. Search in Google Scholar

Oosthuizen, M. 2013. Inflation Inequality in South Africa. University of Cape Town, Development Policy Research Unit. Working Papers 13158. Search in Google Scholar

Piketty, T., and E. Saez. 2014. “Inequality in the Long Run.” Science 344(6186): 838–843. 10.1126/science.1251936Search in Google Scholar

Schultze, C. L. 2003. “The Consumer Price Index: Conceptual Issues and Practical Suggestions.” Journal of Economic Perspectives 17(1): 3–22. 10.1257/089533003321164921Search in Google Scholar

Sturm, J. E., U. Fritsche, M. Graff, M. Lamla, S. Lein, V. Nitsch, D. Liechti, and D. Triet. 2009. “The Euro and Prices: Changeover-related Inflation and Price Convergence in the Euro Area” (No. 381). Directorate General Economic and Financial Affairs (DG ECFIN), European Commission. Search in Google Scholar

Published Online: 2020-01-29
Published in Print: 2020-04-28

© 2020 Walter de Gruyter GmbH, Berlin/Boston

Downloaded on 4.2.2023 from https://www.degruyter.com/document/doi/10.1515/ger-2018-0146/html
Scroll Up Arrow