Policy Research Working Paper 11011 Losing to Inflation Unmasking the Unequal Price Effect of Inflation across Households Muhammad Nasir Shabana Kishwar Moritz Meyer Oscar Barriga-Cabanillas Poverty and Equity Global Department December 2024 Policy Research Working Paper 11011 Abstract Standard measures of inflation, such as changes in the Con- prices in Pakistan, this paper estimates decile-wise inflation sumer Price Index, do not account for the different effects of to explore the heterogeneity of inflation rates across the inflation across income groups emerging from variations in income distribution. The findings suggest that relatively consumption patterns and differences in price trends across poorer households experience higher inflation overall. The consumption categories. This could understate the adverse poorest households, on average, experienced inflation rates welfare effects of inflation for poor households. For instance, that were one percentage point higher than those faced by previous episodes of rising food prices suggested that poor the wealthiest households. This paper also suggests policy households experience higher inflation rates because food recommendations to mitigate the higher inflation effects constitutes a relatively larger share of their consumption on the poor. expenditures. In light of recent increases in energy and food This paper is a product of the Poverty and Equity Global Department. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted atobarriga@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Losing to Inflation: Unmasking the Unequal Price Effect of Inflation across Households∗ Muhammad Nasir Shabana Kishwar IBA Karachi IBA Karachi Moritz Meyer Oscar Barriga-Cabanillas The World Bank Group The World Bank Group Keywords: Inflation, inequality, Pakistan JEL classification: D31, I32, D60 ∗ Muhammad Nasir and Shabana Kishwar are Economics Consultants in the Poverty and Equity Global Practice at the World Bank; and Moritz Meyer is Senior Economist in the Poverty and Equity Global Practice at the World Bank. Aroub Farooq; Erwin Knippenberg and Mohamed Boly have provided comments. We declare that we have no relevant or material financial interests that relate to the research described in this paper. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank Group or any affiliated organizations, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. 1. INTRODUCTION Policy makers often use a macroeconomic lens to examine the consequences of high and volatile inflation. They do have valid reasons to worry about the macroeconomic effects of inflationary episodes. Inflation (i) discourages current investment by creating uncertainty on future returns; (ii) leads to investment in less efficient or unproductive assets (e.g. real estate, gold, etc.) by eroding trust in national currency as a store of value; (iii) discourages savings and therefore economic growth by reducing the real return on financial assets; and (iv) erodes external competitiveness by appreciating the real exchange rate, making the country’s exports more expensive and thereby undermining efforts to improve the trade balance. What is missing from the debate, especially in the context of Pakistan, is the microeconomic effects (firm and household level) of high inflation. Inflation erodes a household's real purchasing power and has adverse effects on household welfare. The typical measure of inflation (e.g., Consumer Price Index or CPI) is based on bundles of goods and services consumed by an average household and, therefore, assumes that the effects of price changes across different income groups are homogenous. It does not account for inflation inequality emerging from differential effects of price changes due to varying consumption patterns across income groups as well as differences in price trends across consumption categories. 1 For instance, households from the bottom of the distribution tend to allocate a higher budget share to food expenditures. If the prices of food items increase faster than other expenditure categories, low-income households will face higher effective inflation rates. These households also belong to a fixed-income group and hence cannot offset the effect of price increases through an increase in income. Hence, when inflation increases, poor households have very little room to react because of a high propensity to consume, and budget constraints are more binding for the poor. This is likely to force them to reduce their consumption (in terms of level) or adjust the composition, which comes at significant welfare losses. It is, therefore, important to assess how inflation behaves at the household level by examining the heterogeneity of consumption patterns between relatively poorer and richer households, differential price trends for different categories of goods and services, and as a consequence, disparities in inflation rates across the welfare distribution. 1.1. RECENT INFLATION IN PAKISTAN Inflation in Pakistan has been on the rise for the last few years. Apart from the dip in January 2021, when the YOY inflation was 5.7 percent, the country has mostly experienced high and volatile inflation since 2019 (See Figure 1). This has eroded the real purchasing power of households (Figure A1 in the Appendix). 1 Although the use of separate CPIs for urban and rural areas captures the regional differences due to assigning of different weights across consumption categories (see Table A1 in the Appendix), these do not capture the variations in consumption patterns across welfare groups. 2 Figure 1: Inflation (YOY) in Pakistan 16.0 14.0 12.0 10.0 Inflation 8.0 6.0 4.0 2.0 0.0 Apr_17 Apr_18 Apr_19 Apr_20 Apr_21 Apr_22 Oct_16 Jan_17 Oct_17 Jan_18 Oct_18 Jan_19 Oct_19 Jan_20 Oct_20 Jan_21 Oct_21 Jan_22 Jul_16 Jul_17 Jul_18 Jul_19 Jul_20 Jul_21 Source: Pakistan Bureau of Statistics (PBS) Although the core inflation has also been gradually rising, the most recent increase in headline inflation from August 2021 onward is primarily driven by food and energy prices. While the food prices (y-o-y) for urban areas increased slightly from 15.3 percent in May 2021 to 15.5 percent in May 2022, these increased from 12.8 percent to 19.0 percent for rural areas during the same period. Figure 2 confirms the elevated levels and volatile nature of food inflation in Pakistan. Figure 2: Food Inflation in Rural and Urban Areas in Pakistan 30 25 20 15 Inflation 10 5 0 Apr_17 Apr_18 Apr_19 Apr_20 Apr_21 Apr_22 Oct_16 Jan_17 Oct_17 Jan_18 Oct_18 Jan_19 Oct_19 Jan_20 Oct_20 Jan_21 Oct_21 Jan_22 Jul_16 Jul_17 Jul_18 Jul_19 Jul_20 Jul_21 -5 Food Urban YoY Food Rural YoY Source: Pakistan Bureau of Statistics (PBS) 3 Energy inflation reached 23.3 percent (y-o-y) in the urban areas and 22.7 percent in rural areas in the first 11 months of fiscal year 2022 compared to 3.1 percent in urban areas and 5.6 percent in rural areas during the same period in fiscal year 2021. The increase in global commodity prices, accompanied by the rupee depreciation, resulted in an increase of domestic prices of petroleum products in Pakistan. In addition, the increase in electricity tariffs by the government also resulted in elevated energy prices. Figure 3 summarizes the recent evolution of energy prices in Pakistan. Figure 3: Energy Inflation in Rural and Urban Areas in Pakistan 50 40 30 Energy Inflation 20 10 0 -10 Energy Urban (YoY) Energy Rural (YoY) Source: Inflation Monitor, State Bank of Pakistan (SBP) Figures 1 to 3 provide the macro situation of inflation in the country. But do this inflation and its components affect all households homogenously regardless of which income groups they belong to? This requires estimating the household level and, subsequently, decile-specific inflation rates in Pakistan. How this is done is discussed in the next section. 1.2. LITERATURE The literature on the relationship between inflation and household welfare is fast growing. Several studies have found the cumulative impact of inflation dispersion across the welfare distribution and other regional and demographic dimensions to have resulted in large disparities over a longer time horizon (Baez et al, 2021). The inflation heterogeneity and its subsequent impact on poverty and inequality in developed and developing countries are well documented (see, for instance, Kaplan and Schulhofer-Wohl, 2017; Oosthuizen, 2013; Arndt et al., 2015). Inflation is typically higher for lower-income households (Kaplan and Schulhofer-Wohl, 2017; Tavares, 2021), and the burden on this segment of the income distribution is mainly driven by substantial increases in the prices of food and non-alcoholic beverages (Baez et al., 2021). The potential reasons for these adverse welfare effects of rising inflation for poor households include 4 but are not limited to, differences in consumption bundles, limited resources for dealing with rising prices (e.g., substitution across goods), absence of indexed wages broadly due to informal employment, and lower financial inclusion resulting in holding savings and wealth in cash (Kahn 1997; Erosa and Ventura, 2002; Burdick and Fisher, 2007; Cysne et al., 2005; Areosa and Areosa, 2016). 2. DATA AND METHODOLOGY Estimating household-level inflation rates depends on combining data on consumer prices with household expenditures. The monthly data on the Consumer Price Index (CPI) are collected and compiled by the Pakistan Bureau of Statistics (PBS). The current base year for price statistics is 2015-16. The data on household expenditure have been collected by the PBS through the Household Integrated Economic Survey (HIES). The latest HIES available is that of 2018-19. The HIES data (2018-19) covered 24,809 households in four provinces. The total number of expenditure items covered in the survey is 283. The expenditure items are grouped into 12 main categories according to the Classification of Individual Consumption According to Purpose (COICOP). These items across the 12 commodity groups include Food and Non-Alcoholic Beverages (98); Alcoholic Beverages, Tobacco (8); Clothing and Footwear (13); Housing, Water, Electricity, Gas and other Fuels (24); Furnishing, Household Equipment, and Routine Maintenance (41); Health (6); Transport (14); Communication (5); Recreation & Culture (14); Education (3); Restaurants and Hotels (29); Miscellaneous Goods and Services (28). The urban CPI collects price data on 356 items, whereas the rural CPI covers 244 for price information. The Laspeyres formula computes urban and rural CPIs using weights from the Household Integrated Income and Consumption Survey (HIICS 2015/16). 2 These regional CPI are then used to compile the national CPI. That is, the weighted geometric mean is computed to obtain the national CPI, using the proportion of urban and rural consumption received through HIICS 2015/16 as their respective weights. It is important to mention that the commodity groups in the CPI are consistent with COICOP categories. The issue, however, is that this CPI assumes the same weight of a COICOP category for all households (See Table A1 in the Appendix). For instance, the national CPI assumes that each household spends 34.58 percent of its budget on food and non-alcoholic beverages. The urban CPI assumes this to be 40.87 percent for all urban households while the rural CPI assume this to be 30.42 percent for all rural households. In reality, however, and we will see in the next section, the weights for these commodity groups (or COICOP categories) vary across income groups. The fact that commodity groups in the CPI are consistent with COICOP categories allows us to combine the two datasets at the commodity groups level (COICOP categories) to estimate the household-level, and subsequently decile-wise, inflation rates in Pakistan. The household-specific inflation rate for each household, i, is calculated as follows: 2 The weights for commodity groups obtained from HIICS are different for rural and urban CPIs as shown in Table A1 in the appendix. 5 =12 ℎ = ∑=1 ℎ ∗ (1) Where ℎ is the y-o-y inflation rate for household ℎ, in month and year . Similarly, c=1….12 represents the 12 COICOP categories. is the share of each COICOP category for household ℎ in particular months and years. These shares are obtained from the unit-level records collected for each household through the HIES (2018-19). is y-o-y inflation for the same month and year . Multiplying the share of each COICOP category at the household level with y-o-y inflation provides the inflation of each COICOP category at the household level. Aggregating the household-level inflation rates of all COICOP categories estimates the inflation rate at the household level. Next, the households are put in their specific deciles constructed on the basis of per adult equivalent household consumption expenditures. The deciles are also adjusted for population sample weights. The decile-specific inflation at the household level is estimated by using the following formula: ℎ = � (2) Here d (=1…10) represents deciles, and N is the total number of households in each decile. This shows that decile-wise inflation is the sum of all households’ inflation divided by the total number of households in that specific decile. Using these two equations, the decile-wise monthly inflation series is obtained from November 2018 to May 2022. 3 The decile-specific inflation rate for each COICOP category is also estimated by employing the above formula. However, this is done separately for each COICOP category. For instance, to estimate the inflation for, say, food and non-alcoholic beverages in the first decile, the sum of household level inflation of food and non-alcoholic beverages in the first decile is divided by the total number of households in this decile. This is done for all the COICOP categories and for all deciles. Subsequently, this provides the components of inflation for each decile (as shown in Figure 5). 2.1. Key Assumptions The calculation and interpretation of decile-disaggregated inflation is based on certain underlying assumptions. Firstly, the inequality of inflation impact exhibits through three channels: its effect on consumption through prices, the change in the asset position of a household, and the differential passthrough to labor income. The methodology of this paper restricts addressing the first channel only. While data limitations prevent us from an examination of the latter two, we believe that the price channel is the most important in the context of lower-income groups in Pakistan due to low levels of asset ownership and a high degree of labor informality, while the labor channel 3 With slight modification, these same equations are used to estimate quantile-wise inflation as well. 6 certainly does matter, the wage adjustment process in this context is dependent on multiple factors not captured in household surveys. Secondly, our interpretation of trends over time assumes that the composition of consumption baskets, and hence the COICOP category weights, remained constant over time and across changes in income levels (zero elasticity). Thirdly, the calculation does not differentiate between purchased and home-produced consumption items, as well as price differentiation stemming from spatial differences, such as urban and crop-producing regions, or income-dependent behavioral differences, for example, price hunting. This is a limitation, but future iterations and work on disaggregating inflation impacts will seek to model this nuance explicitly in the estimation strategy. 3. RESULTS: THE DISTRIBUTIONAL IMPACT OF INFLATION Descriptive statistics show significant heterogeneity in consumption baskets of Pakistani households across income groups. Data compiled from the HIES (2018-19) for the 12 COICOP categories reveals that households from the bottom of the income distribution spend more than two-thirds of their budget on food, housing, and utilities (Table 1). More specifically, households in the first decile allocate around half of their budget for food and non-alcoholic beverages. The top decile, in contrast, spends around 28.4 percent on food and non-alcoholic beverages. Similarly, poor families allocate less than 1 percent for education compared to 7.26 percent by families in the richest decile. Similarly, the budget shares increase across deciles for COICOP categories for housing, utilities, and transport. For instance, the top decile spends more than double (9 percent) on transport compared to spending by the lowest decile (4.2 percent). Overall, the poorest families (decile 1) spend 72 percent of their budget on basic necessities such as food, housing, health, and education. The fact that the consumption structure varies across the welfare distribution implies that high inflation rates for some categories affect some households more than others, depending on their consumption patterns. This inflation heterogeneity depends on the sources of inflation. For instance, keeping in view Table 1, poor households face higher inflation than the rich when it is caused by food inflation. On the other hand, inflation, driven by energy prices and especially fuel prices, affects richer households relatively more. Table 1: Decile-Wise Budget Shares of COICOP Categories COICOP Categories D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 Food and Non-Alcoholic Beverages 49.2 46.9 46.0 44.6 43.5 42.3 40.6 38.4 35.3 28.4 Alcoholic Beverages and Tobacco 1.6 1.5 1.4 1.3 1.2 1.2 1.1 0.9 0.8 0.6 Clothing and Footwear 9.4 9.1 9.0 8.8 8.6 8.3 8.1 7.8 7.5 6.6 Housing, water, electricity, Gas, etc. 18.4 19.0 19.4 20.4 20.7 21.1 22.0 23.5 24.6 27.4 Furnishing, household equipment, etc. 3.2 3.4 3.4 3.3 3.4 3.4 3.3 3.3 3.5 3.9 Health 3.6 3.7 3.5 3.5 3.4 3.5 3.5 3.3 3.4 3.1 Transport 4.2 4.7 5.5 5.8 6.2 6.6 7.0 7.3 8.0 9.0 7 Communication 1.5 1.6 1.6 1.7 1.7 1.7 1.8 1.8 2.1 2.4 Recreation and Culture 1.2 1.4 1.5 1.5 1.5 1.6 1.6 1.6 1.6 1.4 Education 0.7 1.1 1.3 1.7 1.9 2.5 3.1 3.8 4.5 7.3 Restaurants and Hotels 1.6 1.9 1.9 1.9 2.0 2.9 2.2 2.3 2.6 3.3 Miscellaneous 5.2 5.4 5.4 5.4 5.6 5.5 5.5 5.8 5.9 6.5 100 100 100 100 100 100 100 100 100 100 Source: Authors own calculation from PBS data. Note: D=Decile The evolution of inflation experienced by the top and bottom deciles between November 2018 and May 2022 is shown in Figure 4. The figure suggests that poor families experience higher inflation overall. Figure 4: Inflation Inequality across Income Deciles 18 16 14 12 Inflation 10 8 6 4 2 0 Jul_19 Jul_20 Jul_21 May_19 May_20 May_21 May_22 Nov_18 Sep_19 Nov_19 Sep_20 Nov_20 Sep_21 Nov_21 Mar_19 Mar_20 Mar_21 Mar_22 Jan_19 Jan_20 Jan_21 Jan_22 Decile 1 (Poorest) Decile 10 (Richest) YoY_Inflation Source: Authors’ calculations from PBS data. The poorest households (decile 1), on average, experienced one percentage point higher inflation rates than the richest households (decile 10) over this period. The maximum difference of 3.50 percentage points between these two deciles was observed in July 2020. The highest levels of inflation inequality were observed between April 2020 and January 2021, when inflation was on the decline due to COVID-19. This gap can be explained by the fact that during this period, the average energy inflation was -2.4 percent and 0.5 percent for urban and rural regions, respectively (Figure 3). 4 The food inflation, however, was in double digits during the same period (Figure 2). Since food has the highest budget share for the poorest of the households, while utilities and transport are used the most by the top decile, the gap between these extreme 4 This was due to reduced oil and utility consumption resulting from the fall in transportation and industrial activities caused by COVID lockdowns. 8 income groups widened. Another possible explanation comes from the fact that 4 in 5 poor live in rural areas. Hence, these differences in inflation rates are heavily influenced by spatial price trends. This is evident from Figure 2 as well which shows that rural areas experience higher inflation than urban areas, also because of perishable food. It is also important to note that, while the top decile mostly coincided with the national inflation rates, the bottom decile consistently remained above this inflation. Two points call for attention from this illustration. First, there is significant inequality in the price effect of inflation across the income distribution. Second, the inequality is skewed toward poor households, hurting them more. This is especially true for high inflationary periods. Similar conclusions are drawn when the analysis is conducted for different income quantiles (see Figure A2 in the Appendix). To get a clear picture of which expenditure categories drive inflation across the income distribution, we examine the components of inflation across all deciles. Figure 5 shows the contributions of inflation by COICOP categories in each decile for the months of April and May in 2021 (when inflation just started rising after a dip) and 2022 (when inflation was well into double digits). Since poor families spend a major share of their budget on food and non-alcoholic beverages, it is not surprising to see that inflation for poor families is largely driven by this category. For instance, Panel A in Figure 5 shows that the first decile experienced an overall inflation rate of 12.5 percent. However, the major component of this inflation is food and non- alcoholic beverages, with an inflation rate of 7.82 percent. Hence, this category contributes 62.5 percent to the overall inflation faced by the bottom decile. In comparison, the same category explains 43 percent of the overall inflation rate faced by the top income decile. The rest of the panels suggest that, although food price is a major component of inflation in all deciles, the poor households are the ones who are affected the most by the increase in prices of this category. The contribution of this category to the overall inflation faced by the poorest of families varies between 60 – 62 percent. As evident from Figure 5, this contribution gradually decreases as we move up the income ladder. For decile 10, this range varies between 36 – 43 percent. On the other hand, contribution by housing and utilities gradually increases while moving up the income deciles. For instance, the overall inflation faced by decile 1 in April 2021 was 12.5 percent. However, the inflation for the same decile by the housing and utilities was Just 1.79 percent (or 14 percent of the overall inflation for the decile). In contrast, this category explains one-fourth of the overall inflation faced by decile 10 in the same month. The same is the case with the transport category. Its contribution to the overall inflation faced by deciles increases while moving from the first decile to the tenth decile. It is noteworthy, however, that the contribution of transport in the overall inflation becomes more evident in the year 2022 (Panels B and D), even surpassing the contribution of housing and utilities for higher income deciles. This does not come as a surprise since the contribution of energy inflation has been on the rise in recent months due to oil price pressure in the aftermath of events in Ukraine on February 24, 2022. Consequently, the gap between the first and tenth deciles reduced in 2021-2022 because, while the contribution of 9 the food category has remained the same for decile 1, the contribution by the transport category has increased for decile 10 faster than its increase for decile 1. Our discussion above highlights that inflation can impose disproportionate burdens on poor and low-income households when food inflation is higher relative to energy inflation, as has been the case in the years immediately post the COVID pandemic. Furthermore, the poor are more likely to experience inflation more rapidly, leading to depleting savings and substitution towards lower quality food choices. Over the medium term, these differences may compound and have a detrimental effect on human development outcomes. The following section discusses potential policy measures that can be considered to manage inflation while protecting the poor. 10 Figure 5: Contribution to Decile-Wise Inflation Panel A: April 2021 Panel B: April 2022 14 12.5 12.2 12.2 12.0 11.9 13.9 13.9 13.9 13.8 13.8 13.8 13 11.8 11.6 11.4 15 13.7 13.6 13.5 11.1 13.1 12 10.4 14 COICOP Inflation Categories 11 13 COICOP Categories Inflation 10 12 9 11 8 10 9 7 8 6 7 5 6 4 5 3 4 2 3 1 2 1 0 0 Panel C: May 2021 Panel D: May 2022 14 14.2 14.2 14.3 14.2 14.2 14.2 14.1 11.9 15 13.9 13.8 13.5 13 11.8 11.8 11.6 11.6 11.5 11.4 12 11.2 10.9 14 10.4 13 COICOP Categories Inflation 11 COICOP Inflation Categories 12 10 11 9 10 8 9 7 8 6 7 5 6 5 4 4 3 3 2 2 1 1 0 0 11 4. POLICY RECOMMENDATIONS With rising inflation and declining real purchasing power, policy makers have come under pressure to define a policy response that protects especially poor and vulnerable households against rising prices. In Pakistan – like in other countries – policy makers opted for general subsidies on fuel, utilities, and transport, which offered some relief to households. However, given the higher budget shares allocated by the richer households to fuel, utilities, and transport, general fuel subsidies were largely regressive, with a disproportional share of subsidies benefiting relatively richer households. In response to rising fiscal pressure, policy makers reduced general subsidies and switched to targeted social transfers through the social protection program BISP. Yet, by indexing the cash transfers to average inflation rates, the program could not fully mitigate the negative welfare shock arising from higher prices, also because inflation rates to the bottom of the welfare distribution were disproportionally higher. That is, ignoring inflation heterogeneity in deciding public transfers could increase inequalities when, in fact, these are intended to reduce the unequal effects of inflation. The same could be argued for pensions and wages. To rectify such regressive trends, policy makers must avoid measures that impose unsustainable fiscal costs and largely benefit richer households, such as energy relief price measures, and favor progressive policies, such as the reduction of import tariffs on sensitive food items consumed heavily by the poor. Beyond fiscal policy, this paper's findings also call for central banks to monitor differences in inflation patterns and trends between relatively poorer and richer households. Monetary policy should shift focus from welfare maximization of a representative household that is assumed to face average inflation to a set of representatives from the income distribution who face different inflation rates. In other words, inflation inequality should be included in the objective functions of the MP/SBP. Furthermore, episodes of rising inflation with a profound impact on households highlight the need for frequent and representative data collection. The Pakistan Bureau of Statistics collects price data on a weekly (Sensitive Price Index) and monthly basis (Consumer Price Index), but the data is only representative on the national level (and by rural/urban), which could hide a lot of heterogeneity in consumer prices beyond the national average. Increasing the number of markets could support more disaggregated price statistics, which also better reflect the experience of consumers. Moving forward, pro-active and data driven monetary policy making to support price stability objectives is imperative. Lastly, it is important for the government to continue efforts towards the introduction and continuation of targeted safety net measures to protect poorer households from the unequal effects of inflation, especially following shocks that effect food prices. This includes expanding existing programs under the BISP umbrella with clear targeting criteria, and adjusting the benefit levels through regular indexation. 12 REFERENCES Areosa, Waldyr Dutra & Areosa, Marta B.M. (2016). "The inequality channel of monetary transmission," Journal of Macroeconomics, Elsevier, vol. 48(C), pages 214-230. 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. Baez, J. E., Inan, O. K. and Nebiler, M. (2021). “Getting Real: The Uneven Burden of Inflation across Households in Turkey”. Policy Research Working Paper 9869. World Bank Group. Burdick, C. A. and Fisher, T. L.(2007). “Social Security Cost-of-Living Adjustments and the Consumer Price Index”. Social Security Bulletin, Vol. 67, No. 3, pp. 73-88. Cysne, R. P., Maldonado, W. L. & Monteiro, P. K. (2005). "Inflation and income inequality: A shopping-time approach," Journal of Development Economics, Elsevier, vol. 78(2), pages 516- 528, December. Erosa, Andres & Ventura, Gustavo, 2002. "On inflation as a regressive consumption tax," Journal of Monetary Economics, Elsevier, vol. 49(4), pages 761-795, May. Ivanic, M., et al. (2012). Estimating the Short-run Poverty Impacts of the 2010–11 Surge in Food Prices. World Development, 40 (11), pp. 2302-2317 Kahn, S. 1997. “Evidence of Nominal Wage Stickiness from Microdata.” American Economic Review 87 (5): 993-1008. Kaplan, G. and S. Schulhofer-Wohl (2017). Inflation at the household level. Journal of Monetary Economics, 91, pp. 19-38. Laborde, D., et al. (2019). Poverty Impact of Food Price Shocks and Policies. World Bank Policy Research Working Paper 8724 Oosthuizen, M. 2013. Inflation Inequality in South Africa. University of Cape Town, Development Policy Research Unit. Working Papers 13158. Tavares, F. D. A. C. P. (2021). Inflation Heterogeneity and its Impact on Inequality: Evidence from the United States. 13 Appendix Figure A1: CPI Trend Over Time 170 160 150 140 CPI 130 120 110 100 Source: Pakistan Bureau of Statistics (PBS) Table A1: Weights of Commodities Groups In CPI (2015-16 Base) Group# Commodity Groups Average Weights Urban Rural National 1 Food & Non-Alcoholic Beverages 30.42 40.87 34.58 2 Alcoholic Beverage, Tobacco 0.85 1.28 1.02 3 Clothing & Footwear 8.01 9.48 8.60 4 Housing, Water, Electricity, Gas and Other Fuels 27.03 18.49 23.63 5 Furnishing & Household Equipment Maintenance 4.09 4.10 4.10 6 Health 2.31 3.51 2.79 7 Transport 6.14 5.56 5.91 8 Communication 2.35 1.99 2.21 9 Recreation & Culture 1.73 1.38 1.59 10 Education 4.88 2.13 3.79 11 Restaurants & Hotels 7.41 6.19 6.92 12 Miscellaneous Goods & Services 4.77 5.02 4.87 Total 100 100 100 Source: Pakistan Bureau of Statistics (PBS) Note: These weights are based on Household Integrated Income and Consumption Survey (HIICS 2015/16). 14 Figure A2: Inflation Inequality across Income Quantiles 18 16 14 12 Inflation 10 8 6 4 2 0 Jul_19 Jul_20 Jul_21 May_19 May_20 May_21 May_22 Nov_18 Sep_19 Nov_19 Sep_20 Nov_20 Sep_21 Nov_21 Mar_19 Mar_20 Mar_21 Mar_22 Jan_19 Jan_20 Jan_21 Jan_22 Poorest Poor Middle Rich Richest Source: Authors’ calculations from PBS data. 15