THE NEXT URBAN FRONTIER: TWENTY CITIES TO WATCH

THE NEXT URBAN FRONTIER: TWENTY CITIES TO WATCH We identify India s top twenty cities, which together account for just under 1% of India s population, but generate 3.8% of disposable income. We estimate that in 27-8, these cities will fuel just under $1bn of consumption expenditure. Tier II and Tier III cities today may be much smaller in terms of their income and spending weights, but they are growing rapidly. Many of the cities that have registered the fastest economic growth along with the highest asset and financial penetration over the past five years have been outside the 5-million population plus megacities. Sectors heavily geared towards demographics consumer durables, financial services, FMCG and apparel may follow typical patterns where certain Tier II boomtowns will be the next big pockets. At the same time, more specialized consumer markets may find further inroads in smaller niche cities where spending propensities are higher. Consumption patterns in these cities may evolve more quickly, creating deeper addressable markets. These twenty cities will become increasingly important for financial services: savings rates across the cities are nearly double the all-india figure. Megacity savings rates are the highest, with Delhi and Mumbai alone accounting for a fifth of all-india savings. RAJESH SHUKLA ROOPA PURUSHOTHAMAN

RESEARCH TEAM Rajesh Shukla Preeti Kakar Roopa Purushothaman Anindya Roy Saurabh Bandyopadhyay MANY THANKS FOR SUGGESTIONS AND CONTRIBUTION XXXXXXX XXXXXXX FOR PRODUCTION ASSISTANCE Subhash Gopale Future Capital Holdings Limited (FCH), 28 The National Council of Applied Economic Research, 28 ISBN: 81-8883-1- All rights reserved, no part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording and/or otherwise, without the prior written permission of the authors. For further information please contact: rkshukla@ncaer.org aroy@fch.in

FORWARD Nandan Nilekani

PREFACE Suman Bery

PREFACE Sameer Sain

We identify India s top twenty cities and group them into three broad buckets: megacities, which are the largest cities in terms of population and overall consumer markets; boomtowns, which stand out as the next set of big-population cities with high expenditure per household; and niche cities, which are smaller in terms of overall population but still hit well above their weight in spending per household. These twenty cities together account for just under 1% of India s population, but generate 3.8% of disposable income and 21.1% of total spending. We estimate that in 27-8, these cities will fuel just under $1bn of consumption expenditure. We estimate growth for these cities at 11.2%yoy between 24-7. Boomtowns and niche cities today may be much smaller in terms of their income and spending weights, but they are growing rapidly. Many of the cities that have registered the fastest economic growth along with the highest asset and financial penetration over the past five years have been within the boomtown and niche city categories, not the typical megacities. Surat's middle class has more than doubled between 24-5 and 27-8, while the low-income category has shrunk by over a third. In Mumbai, the middle class has grown by 5%, against Delhi where it has only grown by 18%--the slowest of any of our cities. Where has the high income category expanded the fastest? In Lucknow and Jaipur, high-income households have more than doubled, while in Nagpur they have grown by 87%. At less than 4%, the high income segment has grown slowest in Pune and Kanpur. By 215, more than half of households in our top twenty cities will be middle class, while the high-income segment could more than triple. In boomtowns alone, the number of high-income households in 215 could quadruple today s number. Across the twenty cities, 7% of car owners are middle class. In addition, the middle class makes up more than 6% of ownership in the following durables: computers, ACs, washing machines and microwave ovens. The aspirant group edges out the middle class in ownership share of televisions; DVD players; refrigerators; mobile phones; and motorcycles. We present how ownership profiles in middle class households (as well in low-income, aspirant and high income households) differ across cities. In general, the low-income and aspirant categories within boomtowns weigh more heavily in overall asset ownership. For example, aspirants make up over a third of total car owners in boomtowns, against 14.% and 9.5% in megacities and niche cities respectively.

Sectors heavily geared towards demographics consumer durables, financial services, FMCG and apparel may follow typical patterns where the boomtowns will be the next big pockets, and we see some early evidence of this in the household data. At the same time, more specialized consumer markets such as luxury goods, big-ticket durables, entertainment services and housing may find further inroads in niche cities. Athough niche cities may not appear as large as the premier tier of cities on the surface, consumption patterns may evolve more quickly, creating deeper addressable markets. These twenty cities will become increasingly important for financial services: savings rates here are nearly double the all-india figure. Megacity savings rates are the highest, with Delhi and Mumbai alone accounting for a fifth of all-india savings. As much as 29.1% of niche households prefer to keep savings at home.

INSIDE Introduction Trendspotting Across Urban India Megacities Boomtowns Niche Cities Income Distribution Asset Penetration Savings Unusual Expenditure Households are Getting Younger Resurgence of the Joint Family? Insurance Borrowing/Credit 1 3 7 1 12 14 17 22 26 28 31 33 36 Appendix I: Consumer Expenditure by City Appendix II: City Profiles Appendix III: Methodology 39 47 18 Box: Adjusting for cost of living across cities Box: Migration 9 3

1 Eight megacities Mumbai, Delhi, Kolkata, Chennai, Bangalore, Hyderabad, Ahmedabad and Pune garner the lion s share of attention as India s consumption centers. These are the country s largest cities in terms of population (all have populations above five million), and for the most but not entire part they are the largest markets in terms of household income and total consumption expenditure. Beyond this, what is the next set of emerging cities? Here, we come up with a list of twenty of the most dynamic cities, based on the latest NCAER National Survey of Household Income and Expenditure (NSHIE). Following the megacities, we identify two sets of key cities: Boomtowns stand out because of their potential to be the next set of large-population cities with relatively high expenditure per household. In terms of total consumer markets these are the emerging cities that are quickly moving up the ranks as the largest markets following the megacities. Of our twenty cities, this group has younger populations and has posted the fastest growth in disposable income since 2. Cities in this group include Surat, Kanpur, Jaipur, Lucknow, Nagpur, Bhopal and Coimbatore. Niche Cities are somewhat smaller in terms of overall population but still hit well above their weight in spending per household. Household expenditure in these cities is nearly the same as that found in megacities, and niche cities have the highest spending propensity of our three city groups. Faridabad, Amritsar, Ludhiana, Chandigarh and Jalandhar best represent niche cities. These twenty cities together account for 9.9% of India s population, but generate 3.8% of disposable income and 21.1% of total spending. More than half (54.5%) of total urban income is generated by these twenty cities, while megacities alone account for 45.4% of urban expenditure. Boomtowns and niche cities today may be much smaller in terms of their income and spending weights, but they are growing rapidly. Megacities Make Up 7% of India s Population and Generate 2% of the Country s Income Share of total (%) Rural Megacities Urban Niche cities Boomtowns Other cities Total urban All India Households 69.3 8.1.7 1.9 2 3.7 1. Population 7.5 7.4.7 1.8 19.7 29.5 1. Income 56. 24.6 1.7 4.5 13.1 44. 1. Expenditure 63.9 16.4 1.3 3.4 15. 36.1 1. Surplus income Source: NSHIE 32.5 49. 2.9 7.9 7.7 67.5 1. 1 Surat though smaller in population, generates more household income than Hyderabad, Ahmedabad and Pune, and spends more in aggregate than Ahmedabad and Pune.

2 In the following pages, we highlight some of the most interesting spending, savings and demographic patterns found in these transforming cities. The data give comfort that certain trends are developing as expected; however other more counterintuitive trends may change our approach to a uniform urban India. Sectors heavily geared towards demographics consumer durables, financial services, FMCG and apparel may follow typical patterns where the boomtowns will be the next big pockets, and we see some early evidence of this in the household data. At the same time, more specialized consumer markets such as luxury goods, big ticket durables, entertainment services and housing may find further inroads in niche cities. In other words using traditional terminology of boomtowns as Tier II and niche cities as Tier III some sectors should look to Tier III first. Although niche cities may appear as large as the premier tier of cities on the surface, consumption patterns may evolve more quickly, creating deeper addressable markets. A Snapshot of Our Twenty Cities 27-8 Estimates Population (Million) Average household income (Per Annum in 24-5 Rs) Average household expenditure (Per Annum in 24-5 Rs) Mumbai 2.3 459,457 21,14 Delhi 15.5 48,237 25,28 Kolkata 13.8 287,199 174,951 Chennai 6.9 337,59 155,286 Bangalore 6.6 3,678 164,923 Hyderabad 5.7 273,353 149,251 Ahmedabad 4.8 317,856 134,479 Pune 5.1 21,458 126,918 Megacities 78.6 354,9 176,977 Surat 4. 431,21 19,591 Kanpur 2.7 159,761 118,567 Jaipur 3.4 3,374 167,54 Lucknow 2.6 28,393 152,948 Nagpur 2.5 38,625 182,871 Bhopal 2.8 165,21 128,836 Coimbatore 1.7 219,846 152,5 Boomtowns 19.8 283,71 159,47 Faridabad 2.1 252,558 164,457 Amritsar 1.9 267,56 164,54 Ludhiana 1.5 273,211 134,187 Chandigarh 1.1 484,775 212,85 Jalandhar 1.1 296,651 229,335 Niche cities 7.7 34,265 174,287 All 2 cities 16.1 338,58 173,718 Source: NSHIE

3 TRENDSPOTTING ACROSS URBAN INDIA INCOME Across the twenty cities, growth in real household income averaged 7.9% between 22-5. Surat, Nagpur and Ahmedabad registered the fastest growth, at 11.4%, 9.8% and 9.8% respectively. We estimate growth for the top twenty cities at 11.2% between 24-7. At 12.6%, boomtowns appear to be growing the fastest, followed by megacities at 11.% and niche cities at 1.5%. We anticipate the pace of growth across the twenty cities to decelerate gradually to average 1% through 215. Estimated household income levels for Mumbai and Delhi (as well as Chandigarh and Surat) in 27-8 have crossed the 4lakh mark. This makes income per capita in Mumbai and Delhi triple estimates for all-india GDP per capita, and roughly equivalent to China s 27 per capita income levels. Surat is a standout. Surat has nearly the population size of a megacity; it is the youngest city in the entire group (with over 64% of the population below the age of 3); and the city s household income growth tops our urban universe by a comfortable margin. After adjusting for cost of living differences, Surat emerges as the leader in household income levels. The city is already a larger consumer market than Ahmedabad and Pune. INCOME DISTRIBUTION Surat's middle class has more than doubled between 24-5 and 27-8, while the low-income category has shrunk by over a third. In Mumbai, the middle class has grown by 5%, against Delhi where it has only grown by 18%--the slowest of any of our cities. Where has the high income category expanded the fastest? In Lucknow and Jaipur, high-income households have more than doubled, while in Nagpur they have grown by 87%. At less than 4%, the high income segment has grown slowest in Pune and Kanpur. By 215, more than half of households in our top twenty cities will be middle class, while the high-income segment could more than triple. In boomtowns alone, the number of high-income households in 215 could quadruple today s number. Across the twenty cities, 7% of car owners are middle class. In addition, the middle class makes up more than 6% of ownership in the following durables: computers, ACs, washing machines and microwave ovens. The aspirant group edges out the middle class in ownership share of televisions; DVD players; refrigerators; mobile phones; and motorcycles.

4 Annual Household Income Growth Growth rate (%) 21-4 24-7 27-15 Mumbai 8.8 12.7 11.2 Delhi 7.4 1.8 9.5 Kolkata 5.6 8.1 7.1 Chennai 7.3 1.6 9.4 Bangalore 7.8 11.3 1. Hyderabad 6.1 8.8 7.8 Ahmedabad 9.8 14.1 12.5 Pune 6.3 9.1 8. Mega cities 7.5 11. 9.8 Surat 11.4 16.5 14.6 Kanpur 4.8 6.9 6.1 Jaipur 9. 13.1 11.6 Lucknow 8.6 12.4 11. Nagpur 9.8 14.2 12.6 Bhopal 4. 5.7 5.1 Coimbatore 5.4 7.7 6.9 Boomtowns 8.5 12.6 11.7 Faridabad 6.4 9.3 8.2 Amritsar 7.6 11. 9.8 Ludhiana 7.8 11.3 1. Chandigarh 7.4 1.7 9.5 Jalandhar 6.8 9.8 8.7 Niche cities 7.3 1.5 9.3 All 2 cities 7.9 11.2 1 Rest of India 6.1 8.9 7.9 Source: NSHIE India 6.6 9.5 8.4

5 CONSUMER MARKETS The top twenty cities will fuel just under $1bn of consumption expenditure in 27-8, with food spending accounting for just over a third of total spending. Transportation accounts for a fifth of total spending, while household and personal products make up another 15% of overall expenditure. Though still a relatively small share of the consumer basket, communication, education and recreation have posted some of the biggest rises over the past five years. Even with the one of the lowest spending propensities of our twenty cities, Mumbai is the largest spender of the group. Mumbai alone corners a significant share of the twenty-city market in housing and utilities, as well as on apparel. Delhi is the largest spender on healthcare and household/personal products, while both Delhi and Kolkata outpace Mumbai in education and recreation spending. On average, niche household expenditure is nearly the same as that found in megacities, although incomes run 14.2% lower. Niche city households spend on average 24.4% more than megacities on housing, 18.2% more on education, 22.9% more on health, and 27.1% more on unusual social spending. Boomtown households on average spend 12.7% more than megacities on clothing. Overstretched households? 13.3% of households spend more than their reported income, with 8.8% overstretched on routine expenditure alone and 4.5% overstretched due to unusual expenditure (one-off expenditure shocks). Most overstretched households are concentrated in the low-income segment. In this group, unusual expenditure drives a significant number of households into an overstretched position. Composition of Household Expenditure 1% 9% 8% 7% 6% 5% 4% 3% 2.9% 3.1% 2.9% 14.3% 16.% 1.% 5.5% 4.6% 5.2% 5.3% 5.9% 7.4% 8.8% 7.% 7.6% 1.4% 13.1% 8.% 2.3% 21.3% 2% Communication Household and personal products Education and recreation Apparel Health care Housing and utilities Transportation Food, beverages and tobacco 2% 1% 34.6% 31.5% 34.4% % Source: NSHIE Megacities Boomtowns Niche cities

6 SAVINGS Megacities are the biggest savers 1 in our group. As a whole, savings after adjusting to the national accounts figures comes to 5.% of disposable income, nearly double the all-india figure. Mumbai and Delhi alone account for a fifth of all-india household savings. Households primary motivation to save across all cities is to have a buffer against emergencies, followed by a perceived need to save for education and then saving for retirement. However, many times savings are not adequate to cover the reality of economic shock. 24% of people hospitalized in India in a year fall below the poverty line due to hospitalization. Despite high propensities to save, households choose to put money into low-yielding instruments. 65.3% of households across the twenty cities prefer to save by keeping their surplus income as deposits in commercial banks. Even more striking, 18.9% of households simply keep their surplus income at home. As much as 3.9% of niche households prefer to keep savings at home. FINANCIAL SERVICES Boomtowns not megacities have the highest penetration of life insurance and loans outstanding. Nearly half (47.4%) of boomtown households report some type of life insurance coverage, while 18.% of households report outstanding loans. Coimbatore reports the highest credit card penetration and Nagpur reports the highest life insurance penetration of all twenty cities. Niche cities have the highest penetration rate across most durables assets, but they have the lowest financial penetration of the three citygroups. Niche cities are significantly under-penetrated when it comes to credit cards, life insurance and loans outstanding. Niche credit card penetration is just 5.5%, with Amritsar, Ludhiana and Jalandhar below 5% penetration. Niche life insurance penetration is 35.%, with Amritsar having the least penetration of all twenty cities. Unusual expenditure patterns diverge in insured versus uninsured households. Nearly 4% of insured households incur unusual expenditure, against 21.2% of uninsured households. Insured households also spend a third more on unusual expenditure. The composition of unusual expenditure differs markedly between insured and uninsured households. Medical expenses make up 5.5% on uninsured households unusual expenses; in insured households this figure is only 36.1%. 1 In this report, saving is used as a synonym to surplus income. Surplus income=total Household Income-Expenditure (Routine+Unusual).

7 Megacities Megacities are the biggest savers, with Mumbai and Delhi alone accounting for a fifth of all-india savings. Megacity households spend more than households in other citygroups on food, transportation and communication, but less on education and recreation. Interestingly, megacities do not consistently lead in asset/financial penetration. Megacities still dominate the consumer market Megacities make up 73.4% of total population across our twenty cities (boomtowns make up 19.4%, niche cities make up 7.2%), and largely they spend close to their population weight in major market segments. Mumbai and Delhi alone generate 42.7% of income in the twenty cities, and 38.6% of total expenditure. Across all spending categories, megacities make up at least 75% of total spending. Despite lower spending propensities. The average Indian urban household spends just under three-fourths of its income on total expenditure, but this average masks big swings across cities. Interestingly, at 5.% spending propensity (spending as a share of disposable income) is generally lower in megacities, with the exception of Kolkata and Pune. In contrast, Bhopal and Jalandhar spend upwards of 75% of disposable income. The flip side of this is that megacities are the biggest savers 1 in our group. As a whole, savings after adjusting to the national accounts figures comes to 5.% of disposable income, nearly double the all-india figure. Mumbai and Delhi alone account for a fifth of all-india household savings. Megacity savings are 17 times total savings in niche cities and 6.2 times savings in boomtowns. At the household level, megacity savings average 1.4 times boomtown and niche household savings. Part of the savings reflex in megacities may have to do with rapidly rising asset prices and cost-of-living measures, which has led to uncertainty around inflation expectations. Please see the box on p13 for more detail. Megacities Still Dominate the Urban Consumer Market Share of total consumption (%) 1 Megacities Boomtowns Niche cities 8 6 4 2 Food Housing Education Clothing and Footware Durable goods Health Transportation Other Source: NSHIE

8 Even with the one of the lowest spending propensities of our twenty cities, Mumbai is the largest spender of the group. Mumbai s consumer spending is 29.3% higher than Delhi. Mumbai alone corners a significant share of the twenty-city market in housing and utilities, as well as on apparel. Following Mumbai, the southern cities of Bangalore and Chennai spend the most in aggregate on housing and utilities. Mumbai s transport and communication spending is 69.8% higher than that in the next largest market of Delhi. Delhi is the largest spender on healthcare and household/personal products, while both Delhi and Kolkata outpace Mumbai in education and recreation spending. We find only one sector in which megacity households, on average, spend the least of our three citygroups: education and recreation. Households in megacities spend 18% less than niche households on this category. On the other hand, megacity households spend more than both boomtown and niche households in food and beverages, as well as in transportation and communication. Megacity households with primary levels of education earn more than double that of comparable families in niche and boomtowns. At the other end of the spectrum, the income difference is slimmest for graduate-level households. Megacities have the highest share of graduate-level households (51.8%), while 46.2% have at least a high school degree. Megacities also have the highest proportion of both regular wage and salaried workers (51% of all megacity households) and self-employed in non-agriculture households (32%). Over a third of megacity population live in suburbs, or areas outside of the core municipal corporation jurisdiction. While we caution that our core sample is limited, some initial hints of trends may be useful for subsequent study. Chief earners in suburban households tend to be younger on average; average family size is smaller; and suburbia has fewer home owners. Average household income is 11.1% higher in the core. Suburban loan penetration is slightly higher than in the core, although the average borrowing amount is smaller by 11.7%. Suburban households spend more on housing and clothing. Savings Rates Diverge Across India Savings Rate (% of household income) 6 Savings Rate 5 4 3 2 1 All 2 Cities Megacities Boomtowns Niche Cities Urban Rural All India Source: NSHIE

9 Adjusting for cost of living across cities. Throughout this paper, we report nominal income and spending data which gives us comparative information incorporating both prices and levels of spending as one moves across cities in India. However, diverging inflation dynamics across cities affect relative costs of living. We can partly adjust for different costs of living across cities by deflating income using CPI-IW at the city level. When we do this we observe some shifts in the relative rankings of cities: Delhi s and Mumbai s income levels get adjusted downwards by as much as 17% and 12% respectively, while Jalandhar and Amritsar see a boost of 9%. The tables below show both the nominal and adjusted incomes across cities. Asset price inflation also has a significant impact on spending and savings behavior. Arguably, lower spending propensities in megacities could partly be a response to rapid increases in asset prices, which have particularly hit the top tier of urban India. For example, between 21-2 and 24-5 (during which the most recent NSHIE survey was executed), Bangalore saw house prices triple. Similarly gold prices in Mumbai (according to the latest 27 figures from the RBI) have doubled since 21-2. As prices start to correct and inflation expectations become better-anchored, particularly in real estate, we may see more of a boost to consumption in megacities. Average Household Income, 27-8 Rs/houshold Average Household Income, Adjusted for Cost of Living, 27-8 Rs/houshold Chandigarh 484,775 Surat 457,671 Mumbai 459,457 Chandigarh 422,53 Surat 431,21 Mumbai 43,59 Delhi 48,237 Delhi 337,678 Chennai 337,59 Ahmedabad 328,267 Ahmedabad 317,856 Jaipur 325,254 Nagpur 38,625 Jalandhar 323,181 Bangalore 3,678 Chennai 319,759 Jaipur 3,374 Bangalore 32,371 Jalandhar 296,651 Nagpur 298,598 Kolkata 287,199 Amritsar 29,939 Lucknow 28,393 Lucknow 289,2 Hyderabad 273,353 Ludhiana 288,753 Ludhiana 271,514 Hyderabad 275,48 Amritsar 267,56 Kolkata 262,247 Faridabad 252,558 Faridabad 246,129 Coimbatore 219,846 Coimbatore 231,963 Pune 21,458 Pune 191,52 Bhopal 165,21 Kanpur 164,677 Kanpur 159,761 Bhopal 157,847

1 Boomtowns Of the three citygroups, boomtowns are undergoing the most striking shift in income distribution. The group also has the highest share of households who spend more than they earn. Boomtown households spend 12.7% than megacities on clothing. Boomtowns have both the highest life insurance and loan penetration. The boomtown cities couple large populations with high spending per household, resulting in the next pool of big consumer markets. The cities we highlight are meant to be representative, and by no means exhaustive. Some interesting cities that missed the cut include Patna, Indore and Vadodara. These cities are among the top in terms of population (they have larger populations than Bhopal and Coimbatore), but they don t spend enough per household to make our cut. Surat is a standout. Surat has nearly the population size of a megacity; it is the youngest city in entire group (with over 64% of the population below the age of 3); and the city s household income growth tops our urban universe by a comfortable margin. The city is already a larger consumer market than Ahmedabad and Pune. In particular, Surat stands out in apparel spending. It is the largest apparel market after Mumbai, Delhi and Kolkata. Boomtowns have seen the most striking shift in income distribution. The number of high-income households in boomtowns has grown 18.8% annually since 21-2, against 13.7% in the megacities. Today, there are an estimated 9, high-income households and 1.2mn middle-class households across the boomtowns. The income composition of boomtowns now looks similar to that seen in megacities in 21-2. Boomtown households on average spend 12.7% more than megacities on clothing. They also spend slightly more than megacities on education and recreation, and household and personal products. Of the three citygroups, boomtowns allocate the least budget share to foodspend and housing/utilities. On the other hand, boomtowns allocate the most of the three citygroups to apparel; household/personal products; and transportation and communication. More than half (53.6%) of boomtowns unusual expenditure goes towards social spending, while 27.8% goes towards health. Boomtowns have the highest penetration of life insurance and loans outstanding. Nearly half (47.4%) of boomtown households report some type of life insurance coverage, while 18.% of households report outstanding loans. Of the twenty cities, Coimbatore reports the highest credit card penetration, while Nagpur reports the highest life insurance penetration. Coimbatore s attitude towards health stands out. 82.9% of Coimbatore households rank health security (access to health services) as the most important type of household security above life security and economic security far above the rest of the cities. As a share of total expenditure, Coimbatore allocates the most out of all cities to health (12.4%). This city, along with Kanpur, has the highest share of households who borrow for medical expenditure. Still, at 1.2%, Coimbatore has low health insurance penetration. As a group, boomtowns have the highest proportion of overstretched households: more than a fifth of boomtown households spend more than their reported income, compared to 13.8% in niche cities and 11.3% in megacities.

11 Boomtowns Have the Highest Share of Overstretched Households Share of overstretched households (%) 25 2 Routine+Unusual expenditure more than income Routine expenditure more than income BPL 15 1 5 Source: NSHIE Boomtowns Niche cities Megacities Boomtowns Have the Highest Penetration of Life Insurance and Loans Outstanding Life insurance penetration (%) 5 45 4 35 3 25 2 15 1 5 Source: NSHIE Life Insurance Boomtowns Megacities Niche cities Loans

12 Niche Cities Niche household expenditure is nearly the same as that found in megacities, although incomes run lower. In particular, niche households outspend megacity households on housing, education, health and social spending. Despite having high asset penetration rates, niche cities are significantly under-penetrated when it comes to both credit cards and life insurance. Niche cities are smaller than the other two citygroups in terms of overall population. Populations here range from just over a million people in Jalandhar to Faridabad s 2.1 million people. At an estimated 2.6%, annual population growth in these cities is set to be the fastest of the three citygroups over the next decade. All of our niche cities are regionally clustered in the states of Punjab or Haryana. Amritsar, Jalandhar and Ludhiana all sit within Punjab; Chandigarh is the capital of both Punjab and Haryana, while Faridabad is in Haryana. Niche cities have the lowest share of households engaged in wage and salary employment, and a much higher share (relative to the other two citygroups) in casual labor. Of all the cities, Amritsar and Faridabad have the highest share of self-employed households engaged in agriculture (7.9% and 6.9% respectively). These cities will not come close to the other two sets of cities in terms of total markets, but niche cities hit well above their weight in spending per household, even in cases where their income may not match that of our megacities or boomtowns. On average, household expenditure is nearly the same as that found in megacities, although incomes run 14.2% lower. Niche city households spend on average 24.4% more than megacities on housing, 18.2% more on education, 22.9% more on health, and 27.1% more on unusual social spending. Niche Cities Lead in Asset Penetration Rates... Share of households (%) 8 7 Niche Cities Megacities Boomtowns 6 5 4 3 2 1 Television Mobile phone Washing machine Car Computer Source: NSHIE

13 Part of this is mechanical family size is larger, with an average of 5.6 people per household. Niche cities have the highest share of joint families at 46.4%. Still, on a per capita basis, niche cities spend slightly more on healthcare as well as on housing. As a share of total expenditure, niche cities allocate more than the other two citygroups to housing/utilities, healthcare and education and recreation. Income inequality is highest in niche cities, followed by boomtowns and megacities. Differences in spending among various income groups is stark: average household routine spending by the low income segment accounts for only 1.7% of spending by the high income group. In comparison, this figure is 33.1% in megacities. Niche households show the biggest rises in spending as households move into the aspirant segment from the low-income group and into high-income from the middle class. High-income households spend on average 2.9 times more than middle-class households, driven largely by housing and transportation. In megacities, this multiple is less than 1. Niche cities have the highest penetration rate across most assets...their widest lead is in washing machine penetration, where niche cities have a 39.8% penetration rate, against 32.9% in megacities and 22.3% in boomtowns. More than a quarter of niche households own a car, slightly higher than the 24.6% penetration rate in megacities and 15.3% in boomtowns. Niche cities fall short in the following categories: motorcycles, for which boomtowns have the highest penetration; and DVD players, ACs and microwave ovens where megacities have a lead. But the lowest financial penetration of the three citygroups. Niche cities are significantly under-penetrated when it comes to credit cards, life insurance and loans outstanding. Niche credit card penetration is just 5.5%, with Amritsar, Ludhiana and Jalandhar below 5% penetration. Niche life insurance penetration is 35.% (compared to 47.4% in boomtowns), with Amritsar having the least penetration of all twenty cities. Chandigarh s profile is different to the rest of the group: life insurance penetration is 54.5%....But Have the Least Financial Penetration Share 5 of households (%) 5 4 4 Niche Cities Megacities Boomtowns 3 3 2 2 1 1 Source: NSHIE Life Insurance Loans Credit card Life Insurance Loans Credit card

14 Income Distribution By 215 11.% of the top twenty cities could be high-income; 55.3% could be middle class; and less than 7% could fall into the low-income bracket. Delhi and Mumbai alone account for 7% of high-income households across the megacities. Income distribution in the twenty cities has undergone a stark shift over the past decade, away from the low-income segment towards the middle class (households with annual incomes between $6, and $3,). The share of low-income households has fallen to an estimated 15.7% in 27-8 from 32.7% in 1998-99. With more than doubling its 1998-99 population share, the middle class is rapidly closing the gap with the aspirant income segment (where household income is between $3, and $6,). The middle class now accounts for 39.1% of the population across our twenty cities. In niche cities, the middle class makes up more than half (53.3%) of total population, while in boomtowns middle class households are over a quarter (29.3%) of all households. The aspirant segment also a key income group for the takeoff of basic consumer durables accounts for the other sizable chunk of the population, at 4.6% in all twenty cities. Surat's middle class has more than doubled between 24-5 and 27-8, while the low-income category has shrunk by over a third. In Mumbai, the middle class has grown by 5%, against Delhi where it has only grown by 18%--the slowest of any of our cities. Where has the high income category expanded the fastest? In Lucknow and Jaipur, high-income households have more than doubled, while in Nagpur they have grown by 87%. At less than 4%, the high income segment has grown slowest in Pune and Kanpur. Across all income groups the high-income segment (households with annual incomes about $3,) has registered the most rapid growth across all cities. In contrast, the low-income category (those with incomes below $3,) has shrunk by 6.2% annually on average since 21. Since 21, the number of high-income households has more than doubled across the twenty cities. In the megacities, there are an estimated.9 million high-income households in 27-8, accounting for 5.% of total population. 7% of these households are in Delhi and Mumbai alone.

15 The transformation across boomtowns is even more dramatic. The number of high-income households in boomtowns has grown 18.8% annually since 21, against 13.7% in the megacities. Today, there are an estimated 9, high-income households and 1.2 million middle class households across the boomtowns. The income composition of boomtowns now looks similar to that seen in megacities in 21-2. Income Distribution in 215. To get our 27-8 estimates for income distribution, we projected growth in the various income categories assuming 11.2% annual per capita income growth in the twenty cities (11.% growth in megacities; 1.5% in niche cities; and 12.6% in boomtowns). We have also projected out to 215 on the basis of 1% real household income growth. If we see this pace of growth, by 215 12.7% of households could be high-income; 55.3% of all households could be middle class; and less than 7% could fall into the low-income bracket. High-income households may compose as much as 13.6% of megacity population. Household consumption sees the biggest boost as households move into aspirant segments and out of the low-income category: overall spending per household rises by 71.3%. Spending gets a 52.4% boost as household move into the middle class, and a 33.8% boost as household move into the high-income segment. Of the three citygroups, niche cities see the biggest spending jumps as households move into the aspirant and rich segments, while megacities show the biggest spending boost as households move into the middle class. Of the major spending categories, durables spending gets the largest boost on the order of 84.4% as households move into the middle class. Transport spending rises the most as households move into the aspirant and rich segments.

16 More Than Half of Households Will be Middle Class by 215 % of total population 1 Low Income Middle Class Aspirants High Income 8 6 4 2 Source: NSHIE 1998-99 21-2 24-5 27-8 (E) 215-16 (E)... And More Than 15% of Niche Households Could Be High Income Share of total households (%) 1 8 6 High Income Middle class Aspirants Low Income 4 2 27-8 215-16 27-8 215-16 27-8 215-16 Megacities Boomtowns Niche cities Source: NSHIE

17 Assets/Durables Penetration Across the twenty cities, 7% of car owners are middle class. As households move into the middle class, penetration rates for cars jump by 47.ppt while washing machine penetration rates post a 45.3ppt gain. Computer and AC penetration rates rise sharply once households move into the high-income category. Of the household assets covered in the survey, mobile phone and DVD player penetration rates have risen the most since 21. Mobile penetration is 53.3% across all twenty cities, with niche cities outpacing megacities in mobile penetration per household. Niche cities have the highest penetration rate across most assets. Their widest lead is in washing machine penetration, where niche cities have a 39.8% penetration rate, against 32.9% in megacities and 22.3% in boomtowns. More than a quarter of niche households own a car, slightly higher than the 24.6% penetration rate in megacities and 15.3% in boomtowns. Niche cities fall short in the following categories: motorcycles, for which boomtowns have the highest penetration; and DVD players, ACs and microwave ovens where megacities have a lead. Air conditioners have one of the lowest penetration rates of all assets: for example, only 5.5% of megacity homes have air conditioners. We know that penetration rates for air conditioning rise as income grows in hot climates. But beyond this, environmental change in the form of more extreme temperatures and in particular, the acute effects of urban heat islands will further boost demand for air conditioning, particularly in megacities. Once households move into the aspirant segment, penetration rates rise significantly. As households move into the aspirant segment and out of the low-income category, penetration rates rise as much as 4.6ppt for refrigerators; 36.8ppt for mobile phones; and 33.5ppt for televisions. As households move into the middle class, penetration rates for cars jump by 47.ppt while washing machine penetration rates post a 45.3ppt gain. Computer and AC penetration rates rise sharply once households move into the high-income category. Across the twenty cities, 7% of car owners are middle class. In addition, the middle class makes up more than 6% of ownership in following durables: computers, ACs, washing machines and microwave ovens. The aspirant group edges out the middle class in ownership share of televisions; DVD players; refrigerators; mobile phones; and motorcycles. In general, the low-income and aspirant categories within boomtowns weigh more heavily in overall asset ownership. For example, aspirants make up over a third of total car owners in boomtowns, against 14.% and 9.5% in megacities and niche cities respectively. The middle class in niche cities have a larger slice of the ownership pie relative to megacities and boomtowns.

18 Top Twenty Cities: Asset Penetration Rates by Income Group % of households within income group Low Income Aspirants Middle Class High Income Total Cars 4.2 8.1 55. 77.6 23.1 Motorcycles 14. 41.3 48.3 53.2 37.7 Mobile phones 15.8 52.6 77. 88.8 52.7 Televisions 34.5 68. 89.7 96.7 67.8 DVDs 2.4 47.3 62.8 78.1 46.8 Computers 1.2 3.1 17.1 32.3 7.7 Refrigerators 22.7 63.3 9.2 96.8 63.2 Washing Machines 5.8 19.2 64.5 83.3 31.4 Microwave Ovens 1.9 3.3 13.8 26.4 6.7 Air Conditioners 1.2 1.2 1.4 24.2 4.6 Source: NSHIE Middle Class Snapshot: Asset Ownership Profiles Across Different Citygroups Middle Class Penetration Rates (% of middle class households) Megacities Boomtowns Niche Cities Cars 57.48 43.45 46.79 Motorcycles 47.8 6.85 43.66 Mobile phones 76.81 79.5 76.23 Televisions 9.46 87.42 85.78 DVDs 63.13 63.19 59.16 Computers 16.76 22 16.8 Refrigerators 9.2 89.52 91.34 Washing Machines 65.98 55.28 62.3 Microwave Ovens 13.68 14.12 14.3 Air Conditioners 9.95 14.19 9.8 Source: NSHIE

19 Changes in Asset Ownership Patterns Across Cities: Product Distribution by Income Group Share of total ownership (%) 1 8 6 High Income Middle Aspirants Low Income 4 2 All 2 cities Source: NSHIE Mega cities Niche cities Boomtowns All 2 cities Mega cities Niche cities Boomtowns All 2 cities Mega cities Niche cities Boomtowns Refrigerator Washing Machine Microwave Oven Changes in Asset Ownership Patterns Across Cities: Product Distribution by Income Group Share of total ownership (%) 1 8 6 High Income Middle Aspirants Low Income 4 2 All 2cities Mega cities Niche cities Boomtowns All 2cities Mega cities Niche cities Boomtowns All 2cities Mega cities Niche cities TV DVD AC Source: NSHIE Boomtowns

2 Changes in Asset Ownership Patterns Across Cities: Product Distribution by Income Group Share of total ownership (%) 1 8 6 High Income Middle Aspirants Low Income 4 2 All 2cities Mega cities Car Niche cities Boomtowns All 2cities Mega cities Niche cities Motorcycles Boomtowns Source: NSHIE Changes in Asset Ownership Patterns Across Cities: Product Distribution by Income Group Share of total ownership (%) 1 8 6 High Income Middle Aspirants Low Income 4 2 All 2cities Mega cities Niche cities Computer Boomtowns All 2cities Mega cities Niche cities Mobile Phones Boomtowns Source: NSHIE

Asset Penetration Rates Across Cities share of population (%) Car Motorcycle Mobile Phone Color TV DVD Computer Refrigerator Washing Machine Microwave Oven AC Mumbai 29.9 54. 64.1 83.9 78.5 5.6 86.1 36. 6.6 8.2 Delhi 35. 14.8 59.7 76.4 34.8 1.8 82.5 63.4 4.3 6.4 Kolkata 12. 31. 47.5 53.5 36.6 8.4 57.7 19.4 7.8 2.8 Chennai 16.5 4 43.7 7.7 25. 4.2 34.4 28.2 6. 8.5 Bangalore 2.6 26.1 54.3 63. 42.6 11.2 42.2 17.9 5.6 2.1 Hyderabad 16.6 31.1 34.8 41.7 3 4.4 3.4 16..2 1.4 Ahmedabad 37. 5.3 54.7 79.3 7.6 14. 84.5 3.5 3. 8.6 Pune 28.5 6.9 48.3 72.4 42.1 6.4 48.6 22.4 5.7 2.8 Megacities 24.6 37. 53.5 69.3 47.7 7.9 64.2 32.9 5.5 5.5 Surat 24.2 74.8 69.5 91. 9.8 1.1 77.6 28.9 7.3 8.8 Kanpur 5.4 16.3 28.9 38.8 13.8 3.4 47.3 16. 1.7 1.6 Jaipur 23.4 33.5 57.5 59.3 67.3 17. 65.2 34.1 3.6 8.6 Lucknow 8.7 3. 51. 55.9 25.6 6. 56.3 31.9 7.9 6.5 Nagpur 15.5 52.8 41.2 66.2 26. 8.4 64.1 12.8 2.8 2.1 Bhopal 4. 26.5 37.1 49.5 26.1 1.6 45.3 13.7.4.5 Coimbatore 21.2 37.3 35.9 33.9 26.2 1.8 2.9 14.2 1.5 1. Boomtowns 15.3 42. 48.3 6 44.6 6.3 56.7 22.5 4. 4.7 Faridabad 26.7 52.7 44.2 59.5 33.2 4.3 65. 39.2 4.6 3.1 Amritsar 26. 28.1 55.4 69.7 39.2 6.1 65. 4. 2.3 7.1 Ludhiana 25.1 39.1 64.9 78.5 54.5 6.6 58.8 38.5 5.3 3.6 Chandigarh 34.3 21.6 61.4 86.8 38.7 21.5 84.4 52.3 7.4 1.4 Jalandhar 28.6 22.9 68.2 83.9 7. 11.5 58.8 27.9 5.4 3. Niche cities 27.6 34.6 57.8 74.1 45.6 9. 65.8 39.8 4.8 5.3 All 2 cities 23.2 37.7 52.9 68. 47. 7.6 63. 31.5 5.2 5.3 Source: NSHIE future capital research 21

22 Savings Behavior Savings rates in these twenty cities are nearly double the all-india figure. Megacity savings rates are the highest, but it is the boomtowns that have the highest perception of economic security. Delhi and Mumbai alone account for a fifth of all-india savings. As much as 3.9% of niche households prefer to keep savings at home. At 48.7%, savings rates across the twenty cities are nearly double the all-india figure. Megacity savings rates are the highest, with Delhi and Mumbai alone accounting for a fifth of all-india savings. Megacity savings are 17 times total savings in niche cities and 6.2 times savings in boomtowns. At the household level, megacity savings average 1.4 times boomtown and niche household savings. 86.6% of households across the twenty cities earn more than they spend: as a share of total income, savings touch 5.% in megacities, against 42.6% in boomtowns and 43.7% in niche cities. Despite high propensities to save, households choose to put money into low-yielding instruments. 61.8% of households across the twenty cities prefer to save by keeping their surplus income as deposits in commercial banks. Even more striking, 17.9% of households simply keep their surplus income at home. As much as 29.1% of niche households prefer to keep savings at home. For the twenty cities, this works out to roughly Rs3,754bn of savings in a given year. Of surplus income, on average households invested 12.3% of surplus income in financial instruments and 8.7% in physical investment, leaving 79.1% as cash savings. Megacities invest 12.9% in financial instruments, while niche cities and boomtowns invest only 7.9% and 7.8% of surplus income respectively in financial instruments. Investments in physical assets include investments in consumer durables and jewellery. Boomtowns devote a higher share of surplus income to physical assets, on the order of 17.2%, against just 7.5% in megacities. Households in Chandigarh spend the most on average on gold, followed by Surat, Jaipur and Ahmedabad. The survey does not explicitly measure the extent of household real estate investment, but we do know that 78.2% of households own their homes. Generally, renters have lower savings rates relative to homeowners. However, in niche cities we see the opposite trend: savings as a share of income is 27.% for homeowners, compared to 4.4% for renters. Mobilizing savings to buy or build a house is markedly higher in niche cities. Chandigarh stands out because despite a high number of household reporting that savings to buy a house is very important, only 54.4% of households own their homes. This is a wider gap than seen in other comparable cities. 1 Surplus income = Total household income expenditure (routine + unusual). In the report, saving is used frequently as a synonym to surplus income.

23 Households primary motivation to save across all cities is to have a buffer against emergencies, followed by a perceived need to save for education and then saving for retirement. However, many times savings are not adequate to cover the reality of economic shock: 24% of people hospitalized in India in a year fall below the poverty line due to hospitalization. Most households (9.6%) feel that they cannot survive over a year on their current savings. Despite having the highest savings rate of the group, 44.2% of megacity households perceive that they could survive on current savings for less than six months. Nearly three-fourths of Kolkata households and more than half of Mumbai households cite their savings buffer to be less than six months. In contrast, nearly a quarter of Surat households and a fifth on Pune households report that their savings buffer would last over one year. Households are fairly optimistic about employment prospects. 56.2% of all households feel confident that they could find alternate employment within six months if they were to lose their current employment. Hyderabad bucks the trend and is least confident on employment prospects 4.6% of households are pessimistic on employment opportunities. Boomtowns have the highest perception of economic security. Nearly three-fourths of boomtown households are confident about the stability of their major source of income; in niche cities this figure is 49.7%. At the same time, an overwhelming 85.6% of niche households expect to increase their consumption expenditure over the next year. In megacities only 42.7% of households expect to spend more just over a third of households expect to spend the same amount. 19% of Households Prefer to Keep Savings at Home % of households 1 Keep at home Deposit in Co-operative Society Deposit in bank Purchase insurance policies Deposit in post office Others 8 6 4 2 Source: NSHIE Megacities Niche cities Boomtowns All 2 cities

24 Households Invest 21% of Surplus Income, Leaving 79% As Cash Savings % of Surplus Income 9 8 7 6 5 4 3 2 1 Source: NSHIE Megacities Boomtowns Niche cities Savings in Cash Financial Investment Physical Investment Chief Earners Between the Ages of 26-35 Have the Highest Tendency to Keep Savings at Home Share of households across the twenty cities (%) 8 7 Deposit in bank Keep at home Others 6 5 4 3 2 1 Below 25 26-35 36-45 46-55 56-65 Above 66 Source: NSHIE

25 Households Overwhelmingly Draw on Savings to Cope With Shocks % of households who faced economic shock 9 8 Death of chief earner Major sickness of any household member Household/business loss 7 6 5 4 3 2 1 Using own savings Loan from friends/relatives Loan from employer Selling land/jewellery Financial institutions/bank Others Source: NSHIE

26 Unusual Expenditure Just under 3% of all households are hit with some form of unusual expenditure. Younger households spend the most in aggregate on unusual expenditure, driven mainly by medical expenses. Unusual expenditure: Beyond the routine spending that all households incur, in a given year a proportion of households also face unusual expenditure (one-off expenditure shocks). Particularly in the low-income category, unusual expenditure tips a significant number of households into potentially vulnerable positions, where spending outstrips annual income. Within NSHIE, unusual expenditure is captured through five major categories: social spending (occasional but large annual expenditures on weddings, festivals, birth and other social events); medical expenses; travel expenses; education expenses; and others. The largest source of unusual expenditure is social ceremonies, which accounts for 43.5% of total unusual expenses. Following this, 25.4% goes to one-off medical expenses. In niche cities alone, close to 65% of unusual expenditure goes towards social purposes. Just under 3% of all households are hit with some form of unusual expenditure. 4.6% of all middle class households incur unusual expenses, against 23.9% in the low-income group. Overstretched households? 13.3% of households spend more than their reported income, with 8.8% overstretched on routine expenditure alone and 4.5% overstretched due to unusual expenditure. Most vulnerable households are concentrated in the low-income segment. In this group, unusual expenditure drives a significant number of households into vulnerability. Composition of Unusual Expenditure Varies Across Citygroups Share of total (%) 1% Social spending Medical Education Travel Others 8% 6% 4% 2% % Source: NSHIE Megacities Boomtowns Niche cities