CHAPTER - 4 MEASUREMENT OF INCOME INEQUALITY BY GINI, MODIFIED GINI COEFFICIENT AND OTHER METHODS.

CHAPTER - 4 MEASUREMENT OF INCOME INEQUALITY BY GINI, MODIFIED GINI COEFFICIENT AND OTHER METHODS.

CHAPTER-4. MESUREMENT OF INCOME INEQUALITY BY GINI, MODIFIED GINI COEFFICIENT AND OTHER METHODS 4.1 Income Demographics Of India: Indian planning, since its inception, was concerned with the idea of balanced growth. Since the opening up of the Indian economy in 1991-2 and its fairly impressive growth over the past decade, led essentially by a few Indian states, there has been increasing concern about the fact that some states are growing slowly or not at all, and thus falling behind or failing to benefit from the opening up of the Indian economy and the dynamism of certain sectors and regions. In order to compare the income distributions in different regions we take help of the NCAER data and analyse them. In the year 1985-86 National Council of Applied Economic Research (NCAER) initiated MISH (Market Information Survey of Households) to collect the income data. The National Council of Applied Economic Research (NCAER), headquartered in New Delhi, was founded in 1956 through the initiative of a group of a public- and private sector leaders. Designed to be an independent institution, run by a governing body and supported by corporate and individual members, its goal throughout has been to support informed decision making both for public policy and the private sector through rigorous and objective empirical economic research. Since the mid-1980s, the NCAER has conducted repeated surveys of ownership and expenditure pattern of a limited group of consumer goods through MISH. MISH is called the Market Information Survey of Households. Which is primarily designed to assist decision making in the private sector. The income coverage of this survey is now being expanded to make it of greater academic interest. In addition to information on consumer goods, data on household incomes are collected in MISH surveys from all the sample households. One of the by products of these surveys is the distribution of households by income. MISH surveys provide income distribution at a disaggregated level. As the survey of MISH is extensive and large scale it also adopts a cluster approach. Although large numbers of clusters are selected in the sample to control the variations between clusters, the clustering effect is not totally eliminated. 95

The Survey Findings of MISH classified the Households into Five Income Categories. Following are the Annual Income Categories for the year 1985-1999:- Deprived: Annual Income up to 35,000 Aspirers (Lower income): Annual Income 35,001 to 70,000 Seekers (Middle-income group): Annual Income 70,001 to 105,000 Near Rich (Upper Middle-income group): Annual Income 105,001 to 140,000 Sheer Rich: Annual Income 140,001 and above The years for which the survey was conducted by MISH could be organized into following three broad periods:- Pre reform...1985-86 to 1989-90 Post-reform I... 1992-1993 to 1995-96 Post - reform II... 1995-96 to 1998-99 Following two tables reveal the changing income trends during and between the first two periods: Table: 4.1 Distribution of Households by Income, 1985-86 and 1989-90 (in per cent) (Pre Reform Period) Annual Income (Rs.) at 1998-99 prices Income Class 1985-86 1989-80 Rural Urban Total Rural Urban Total <= 35,000 L 42.1 73.6 65.2 37.1 67.3 58.8 35,001-70,000 LM 35.8 21.4 25.2 34.8 23.9 26.9 70,001-105,000 M 15.2 4.0 6.9 17.9 7.1 10.1 105,001-140,000 UM 3.9 0.7 1.5 6.5 1.2 2.7 >140,000 H 3.1 0.3 1.1 3.8 0.5 1.4 Total 100.0 100.0 100.0 100.0 100.0 100.0 L- Low, LM- Lower Middle, M- Middle, UM- Upper Middle, H- High Table: 4.2 Distribution of Households by Income, 1992-93, 1995-96 and 1998-99 (Post Reform Period) Annual Income (Rs.) at 1998-99 Prices Incom Class 1992-93 1995-96 1998-99 Rural Urban Total Rural Urban Total Rural Urban Total <= 35,000 L 38.5 65.5 58.2 27.9 57.2 48.9 19.0 47.9 39.7 35,001-70,000 LM 33.0 2.6 25.4 4.9 29.0 30.7 33.8 34.8 34.5 70,001-105,000 M 16.1 8.2 10.4 20.3 8.6 11.9 22.6 10.4 13.9 105,001-140,000 UM 7.6 2.3 3.7 9.6 3.1 5.0 12.2 3.9 6.2 >140,000 H 4.9 1.4 2.3 7,3 2.0 3.5 12.5 3.0 5.7 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 L- Low, LM- Lower Middle, M- Middle, UM- Upper Middle, H- High 96

From the above two tables we can assess that in both urban and rural areas the income demographics is unrecognizably spreading all over. Low Income Class----- Deprived---------- Shrinking Lower Middle Class Stagnant-----------Negligible Fluctuation Middle Class-------------Upcoming-------- Fast Growth Upper Middle Class---- Rising------------- Spreading Faster High-----------------------Filthy Rich------------- Skipping The above tables reveal that in 1985-86, countrywide almost two thirds of the households were in the low-income class, with barely one percent belonging to the highincome group. The distribution of the income varied widely and was mostly skewed in the lower level of the income scale both in the rural and urban areas. The table 4.1 shows that during the pre-reform period and in the year between 1985-86 and 1989-90, the percentage of the lower income households declined from 65.2 to 58.8. There is a decline of 6.4 percentages in four years, which could be rated as 2.5 percent every year. Whereas in the first post reform period in the year between 1992-93 and 1995-96, the percentage dropped quite sharply from 58.2 to 48.9, which indicates 9.3 percentage points in three years or 5.64 per cent annually. Again if we compare the percentage of the lower income class in the post-reform period in the year 1995-96 and 1998-99, we find that there is a fall of 9.2 percentages in three years at the rate of 6.71 percent per annum. Thus it carries the impression that the lower income group has upgraded at a faster rate during the postreform period. The high-income category shows a mere upgradation of 0.3 points in four years (increasing at a rate of 6.21 percent every year) in the period 1985-90. Post reform period shows a faster upgradation of the high - income households. During the period 1992-96, high-income households increased from 2.3 per cent to 3.6 per cent and further to 5.7 per cent till 1998-99. The rate of increase of these households was 15.02 per cent and 17.7 per cent respectively, during the first and second post-reform periods. 97

In other words we can assess that, the number of low-income households remained stagnant in both 1985-86 and 1989-90. During the period 1985-86 and 1989-90 about 14 million new households were added. Interclass movement of households between different income groups has also been shown in the table. The resultant change in the distribution is so because it has been found that in the low-income group the number of new entrants almost equalled the number that moved out and up the income ladder. Consequently the size of this class remained almost unchanged. In contrast, between 1992-93 and 1995-96 the number of low income households declined by about 9.7 million, which in fact is slightly more than the new additions. Between 1995-96 and 1998-99, low-income households declined by about 12.6 million. Table 4.3 below exhibits the average annual growth rates of various income classes for one pre-reform period and two post-reform periods. Table: 4.3 Average Annual Growth Rates Incom e Class Pre-reform period (1985-86 to 1989-90) Post-reform period -1 (1992-93 to 1995-96) Post-reform period - II (1995-96 to 1998-99) Urban Rural Total Urban Rural Total Urban Rural Total L 0.93-0.20-0.01-7.03-3.03-3.72-10.83-4.51-5.84 LM 3.34 4.19 4.33 5.46 10.22 8.59 0.91 7.71 5.63 M 8.53 17.82 12.72 11.96 3.11 7.01 5.32 7.82 6.63 UM 18.52 16.39 17.83 11.90 12.25 12.06 9.91 8.61 9.33 H 9.557 13.90 10.65 18.22 15.68 17.14 21.50 14.31 18.62 Total 4.14 2.04 2.61 3.50 1.44 2.01 1.69 1.29 1.41 Chart 1 Growth in Income Classes in per cent (Urban) Pre-reform period (1985-86 to 1989-90) Post-reform Period -I (1992-93 to 1995-96) Post-reform Period - II (1995-96 to 1998-99) 98

Chart 2 - Growth in Income Classes in per cent (Rural) Pre-re'orm period (1985-86 to 1989-90) Post-reform Period -I (1992-93 to 1995-96) Post-reform Period - II (1995-96 to 1998-99) It has been observed from the above charts that during the post-reform period the rates of growth in the upper-income categories were higher, while the growth rate for the lower-income category was on the decline. There was a spurt in the growth in the top most income class, while the rate of growth in the lowest income class was actually negative; declining further during 1995-96 to 1998-99 as compared to 1992-93 to 1995-96. Both the urban and rural areas illustrate analogous trends, but the movement in the case of urban areas is more prominent. Table 4.2 reveals that in both the cases of urban and rural, the size of the low-income declined between 1992-93 and 1995-96; but the rate of fall was four percent higher in the case of urban than in the case of the rural. During the period this rate of decline of low-income households further increased by eleven per cent per annum in urban and 4.5 per cent per annum in rural. Similarly, during the periods of 1992-96 and 1995-99, the size of the high-income group grew at 18.22 per cent and 21.50 per cent per annum respectively in the urban areas compared to 15.68 per cent and 14.31 per cent respectively in the case of rural areas. Table : 4.4 Average Annual Growth Rate by Sectors, at 1993-94 prices (in per cent) Period Agriculture Industry Others 1985-86 to 1989-90 3.93 8.10 7.49 1992-93 to 1995-96 2.95 9.36 8.39 1995-96 to 1998-99 4.33 5.01 8.44 From Table 4.4 above it is observed that in the previous years the changes were more manifested in the rural areas. The changes were due to the varying growth rates of 99

the components of GDP. During these three periods, which are taken into consideration, the average annual growth of GDP was almost the same - that is between 6.3 per cent and 6.8 per cent. However, the components of GDP - agriculture, industry and others had varying rates of growth in these three periods. As the rural households are more inclined to the agricultural sector, the entire income of the agricultural sector is credited to rural households. In the first two periods, there was a considerable decline of almost 25 percent in the growth rate of agricultural income, but it increased by around 47 percent in the last two periods. In contrary to the rural economic sources, the urban economic sources mainly emphasised on the industries and other sectors. These two sectors grew more rapidly during the second period. As a consequence, urban incomes grew faster than the rural incomes during this period. Despite the faster growth of population in the urban areas, the average urban household income had a faster growth than the consequent average household income in rural areas. The ratio of average urban income to average rural income, which stood at 1.5 in 1985-86, grew to 1.56 in 1995-96 and further to 1.68 in 1998-99. This probably explains the inconsistency in the growth rates observed during the above three periods. 4.2 Regional Trends Of The Income Distribution: The comparative study of the trends in the income distribution for the four different regions of the country - North, South, East and West reveals that the Eastern region stands out as having exhibited the least change. Between 1990 and 1996, the percentage of low-income households dropped by 2 points i.e. 56 to 54 percent in comparison to the drop of almost 10 points ( i.e. 59 to 49 percent) at the All India level. During the period between 1996 and 1999, the percentage of low-income households dropped by around 4 points, the corresponding drop at the All India level being around 9 points. Middle, upper-middle and high-income groups together shows a marginal change, of 15 per cent to 17 per cent while the other groups exhibited changes of more than 5 points during 1990-96. Thus it is assessed that the pace of change in these income groups was found to be slower in the Eastern region as compared to that in the other regions during 1996-99. The proportion of higher income households increased by around 9 per cent points in the Western region followed by 7 per cent points in the Southern region. The increase was merely 4.5 per cent points for the Northern region. In the Northern region only the top three income groups grew at the rate of less than 10 per cent per annum between 1993 and 1996. 100

The Southern region shows the highest drop in the percent point for the lowincome households among all the regions of the country. During 1990-96, the low-income households dropped from 70 to 53, which further declined to 40 in 1998-99. Simultaneously, the proportion of middle to high-income households doubled between 1990 and 1996 and increased from 17.87 to 24.80 per cent between 1996 and 1999. The West of the country depicts the lowest percentage of low-income households and the highest percentage of high-income households in 1989-90 and continued to do so in 1998-99. Conversely, the North which had a higher percent of low-income households in 1989-90, enhanced its position in 1995-96 and was in the second position till 1998-99. Amongst the urban areas the proportion of high-income households comprising of middle, upper-middle and high-income classes was highest in the West (54.57 Percent) followed by the East (46.48 percent), North (44.90 percent) and South (43.44 percent) during 1998-99. A similar sequence was observed in 1995-96. However, in 1989-90 the East had highest proportion of higher income households followed by the West, the North and the South. The trend can be read from the Table 4.5. 101

Table: 4.5 Region-wise Rural and Urban Changes by Income Class (in per cent) Urban Rural Region Household Growth per Growth per Household Distribution Distribution annum annum 1989-90 1995-96 1998-99 1989-9611995-99 1989-90 1995-96 1998-99 1989-96 1995-99 North L 37.86 29.51 21.61-2.23-7.88 67.11 57.26 46.50-0.28-4.97 LM 34.8C 33.93 33.49 1.48 1.75 22.59 27.20 34.89 5.61 10.67 M 17.31 18.54 20.04 3.08 4.88 8.22 9.41 10.84 4.71 6.78 UM 6.28 9.71 11.14 9.60 6.98 1.36 3.18 4,01 17.93 10.02 H 3.75 8.3C 13.72 16.331 20.85 0.72 2.94 3.76 29.51 10.57 ALL 100.0C 100.00 100.0C 1.91 2.20 100.00 100.00 100.00 2.39 1.86 South L 52.72 33.62 20.43-5.04-14.13 78.39 61.35 49.17-1.72-6.46 LM 29.28 34.60 36.13 5.24) 2.86 17.42 27.11 34.76 10.21 9.40 M 11.64 18.52 22.44 10.611 8.08 3.43 7.51 10.14 16.66 11.28 UM 4.01 7.64 11.04 13,971 14.63 0.56 2.72 3.71 33.43 11.72 H 2.35 5.48 9.96 17.82) 23.72 0.20 1.30 2.21 39.94 20.27 ALL 100.00 100.00 100.00 2 36 1.38 100.00 100.00 100.00 2.38 0.70 East L 27.29 25.05 19.80 0.11) -6.06 62.58 61.94 57.97 1.82-0.99 LM 37.6C 34.49 33.72 26.47 27.38 30.97 2.57 5.47 M 22.69 34.76 25.03 2.33) -8.93 8.84 6.99 6.90-1.93 0.76 UM 8.08 9.22 11.40 3.80 9.04 1.41 2.33 2.28 10.86 0.45 H 4.35 7.48 10.05 11.18 12.10 0.69 1.36 1.89 14.27 12.91 ALL I00.0C 100.00 100.00 1.53 1.60 100.00 100.00 100.00 2.00 1.22 West L 24.48 22.45 13.95 2.80-13.34 59.26 44.56 34.14 1.57-7.35 c f o -'T OO o ' LM 39.31 35.21 31.48 2.40-2.17 31.47 36.79 40.44 5.94 4.49 M 22.69 22.05 23.96 4.94 4.40 7.49 11.27 15.17 10.48 11.80 UM 8.48 11.99 14.96 13.72 9.34 1.30 4.71 6.03 27.97 9.93 H 5.03 8.30 15.65 14.9a 25.46 0.48 2.67 4.21 37.41 17.87 ALL I00.0C 100.00 100.0C 4.30( 1.55 100.00 100.00 100.00 3.22 1.25 Note: L-Low, LM-Lower Middle, M-Middle, UM-Upper Middle, H-High In the urban areas a negative growth rate in low-income households, was observed in all the regions. But the same was more prominent in South and West - pursued by the North. It can be revealed from Chart 2 that the rate of decline was lowest in the East. In the East the rate of decline was more prominent in rural areas; but the rate of decline was comparatively less in urban areas. This can be attributed to the implementation of progressive policies in some of the states, while the other states including the states in the East lagged behind during the period mentioned. 102

Chart: 2: Growth in Income Classes North - Urban North - Rural South-Urban South-Rural 33 4 39.9 %4 7 1 _ m. 11.3 'll 6:5- LM 1.7 I M UM H 1 East-Urban East -Rural 103

Table:4.6 Estimated Distribution of Households by s in some of the Major Cities of India (in Percent) (Rs. per annum) 1985-86 1998-99 Ahmedabad <= 35,000 35.22 4.82 35,001-70,000 39.31 29.40 70,001-105.000 16.98 29.91 105.000-140,000 4.46 17.49 Above 140,000 4.03 18.38 Total 100.00 100.00 Total no. of households 563.00 789.00 Bangalore <= 35,000 36.89 11.28 35,001-70,000 51.10 30.43 70,001-105,000 9.39 24.04 105,000-140,000 1.58 15.96 Above 140.000 1.05 18.30 Total 100.01 100.00 Total no. of households 653.00 940.00 Chennai <= 35,000 59.89 12.71 35,001-70,000 28.57 29.38 70.001-105,000 7.00 24.65 105,000-140,000 3.60 14.42 Above 140,000 0.94 18.84 Total 100.00 100.00 Total no. of households 1.046.00 1,290.00 104

Delhi <= 35,000 16.80 2.18 35,001-70,000 35.50 16.25 70,001-105,000 35.43 22.69 105,000-140,000 6.87 20.65 Above 140,000 5.48 38.24 Total 99.99 100.00 Total no. of households 1,292.00 2,160.00 Hyderabad <= 35,000 14.70 10.14 35,001-70,000 47.87 27.15 70,001-105,000 29.92 28.95 105,000-140,000 5.52 18.04 Above 140,000 1.99 15.72 Total 100.00 100.00 Total no. of households 580.00 1,164.00 Jaipur <= 35,000 37.72 15.91 35,001-70,000 43.36 30.97 70,001-105,000 13.16 30.97 105,000-140,000 4.42 11.93 Above 140,000 1.33 10.23 Total 99.99 100.00 Total no. of households 234.00 352.00 Kanpur <= 35,000 40.76 22.64 35,001-70,000 49.60 35.43 70,001-105,000 8.31 21.17 105,000-140,000 0.88 1.53 Above 140,000 0.44 9.22 Total 99.99 100.00 Total no. of households 388.00 477.00 Kolkata <= 35,000 27.65 7.26 35,001-70,000 34.96 25.07 70,001-105,000 21.26 30.81 105,000-140,000 8.46 17.97 Above 140,000 7.66 18.90 Total 99.99 100.00 Total no. of households 2,003.00 2,577.00 Lucknow <= 35,000 17.21 17.69 35,001-70,000 43.60 32.05 70,001-105,000 17.65 28.46 105,000-140,000 12.46 13.08 105

Above 140,000 9.08 8.72 Total 100.00 100.00 Total no. of households 206.00 390.00 Mumbai <= 35,000 3.94 35,001-70,000 21.39 23.93 70,001-105,000 36.21 24.67 105,000-140,000 16.14 19.87 Above 140,000 22.32 28.47 Total 100.00 100.00 Total no. of households 1,876.00 2,979.00 Nagpur <= 35,000 29.25 15.90 35,001-70,000 55.43 21.02 70,001-105,000 13.02 20.75 105,000-140,000 1.65 19.68 Above 140,000 0.65 22.64 Total 100.00 100.00 Total no. of households 249.00 371.00 Pune <= 35,000 5.25 7.44 35,001-70,000 35.78 13.67 70,001-105,000 38.34 26.30 105,000-140,000 14.16 23.70 Above 140,000 6.47 28.89 Total 100.00 100.00 Total no. of households 455.00 578.00 Gap of a decade shows an enormous upgradation in the economic status in almost all the cities that are taken into consideration. The decline of the lower income in Mumbai in ten years is less than 1 percent, whereas in Pune the percentage of lower income group has gone up from 1985-86 to 1998-99. The percentage of income group above Rs. 140,000 has gone down the ladder in the cities of Lucknow and Mumbai. Percentage of same income group for the rest of the cities has moved up the ladder. The population of Hyderabad has doubled in ten years and the decline in the income group of Rs.35, 000 is observed only 5 percent. Income group of 70,000-105,000 moved up the ladder by only 1 percent. But the population with an annual income of 140,000 and above increased by 14 percent. Thus it is assessed that population growth took place among the affluent group. In Delhi and Mumbai same feature is being observed. In Kolkata the increase in population is nearly 600 and the fall off in first two income levels are 20 and 10 percent respectively. But the elevation in the last two economic statuses is more than 50 percent. 106

Table: 4.7 Cities Deprived Lower Income Middle Income Upper Middle Group Group Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 1985-86 17500 52500 87500 122500 57,500 Ahmedabad Cum. Prop. Person Cum. Prop. Income 35 12 75 50 92 78 96 88 100 100 Bangalore Cum. Prop. Person 36.89 87.99 97.38 98.96 100.01 Cum. Prop. Income 14.32 73.82 92.04 96.33 100 Chennai Cum. Prop. Person 59.89 88.46 95.46 99.06 100 Cum. Prop. Income 27.95 67.95 84.29 96.05 100 Delhi Cum. Prop. Person 16.8 52.3 87.73 94.51 99.99 Cum. Prop. Income 4.23 31.04 75.64 87.58 100 Hyderabad Cum. Prop. Person 14.7 62.57 92.49 98.01 100 Cum. Prop. Income 4.03 43.44 84.48 95.09 100 Jaipur Cum. Prop. Person 37.72 81.08 94.24 98.66 99.99 Cum. Prop. Income 13.64 60.68 84.48 95.67 100 Kanpur Cum. Prop. Person 40.76 90.36 98.67 99.55 99.99 Cum. Prop. Income 16.90 78.58 95.80 98.36 100 Kolkata Cum. Prop. Person 27.65 62.61 83.87 92.33 999 Cum. Prop. Income 7.53 36.11 65.08 81.21 100 Lucknow Cum. Prop. Person 17.21 60.81 78.46 90.92 100 Cum. Prop. Income 4.25 36.53 58.31 79.83 100 Mumbai Cum. Prop. Person 3.94 25.33 61.54 77.68 100 Cum. Prop. Income 0.70 12.10 44.25 64.32 100 Nagpur Cum. Prop. Person 29.25 84.68 97.7 99.35 100 Cum. Prop. Income 10.52 70.33 93.74 97.90 100 Pune Cum. Prop. Person 5.25 41.03 79.38 93.54 100.01 Cum. Prop. Income 1.14 24.39 65.92 87.39 100 Cities Deprived Lower Income Middle Income Upper Middle Group Group Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 1998-99 17500 52500 87500 122500 157,500 Ahmedabad Cum. Prop. Person 4.82 34.22 64.13 81.62 100 Cum. Prop. Income 0.91 17.54 45.73 68.81 100 Bangalore Cum. Prop. Person 11.28 41.71 65.75 81.71 100.01 Cum. Prop. Income 2.26 20.55 44.63 67.01 100 Chennai Cum. Prop. Person 12.71 42.09 66.74 81.16 100 Cum. Prop. Income 2.57 20.39 45.31 65.72 100 Delhi Cum. Prop. Person 12.71 42.09 66.74 81.16 100 Cum. Prop. Income 2.57 20.39 45.31 65.72 100 Hyderabad Cum. Prop. Person 10.14 37.29 66.24 84.28 100 Cum. Prop. Income 2.01 18.17 46.88 71.93 100 Jaipur Cum. Prop. Person 15.91 46.88 77.85 89.78 100.01 Cum. Prop. Income 3.62 24.77 60.03 79.04 100 Kanpur Cum. Prop. Person 22.64 58.07 79.24 90.77 99.99 Cum. Prop. Income 5.68 32.36 58.92 79.18 100 Kolkata Cum. Prop. Person 22.64 58.07 79.24 90.77 99.99 Cum. Prop. Income 5.68 32.36 58.92 79.18 100 Lucknow Cum. Prop. Person 7.26 32.33 63.14 81.11 100.01 Cum. Prop. Income 1.36 15.49 44.42 68.05 100 Mumbai Cum. Prop. Person 3.05 26.98 51.65 71.52 99.99 Cum. Prop. Income 0.51 12.61 33.39 56.83 100 Nagpur Cum. Prop. Person 15.90 36.92 57.67 77.35 99.99 Cum. Prop. Income 3.03 15.06 34.85 61.13 100 Pune Cum. Prop. Person 7.44 21.11 47.41 71.11 100 Cum. Prop. Income 1.23 8.00 29.70 57.08 100 107

Table: 4.8 Percentage of Households in the Low income (L) Category State 1989-90 Rank 1995-96 Rank 1998-99 Rank Kerala 82.83 1 50.99 7 44.25 5 Tamil Nadu 78.70 2 55.70 4 34.14 12 Madhya Pradesh 69.20 3 62.34 2 48.51 3 Orissa 67.04 4 64.14 1 63.30 1 Karnataka 66.22 5 49.75 9 40.72 8 Bihar 62.51 6 58.30 3 57.94 2 Uttar Pradesh 62.38 7 53.35 5 44.60 4 Andhra Pradesh 59.91 8 52.57 6 42.81 6 All India 58.84-48.91-39.66 «Himachal Pradesh 57.07 9 47.23 10 38.80 9 Rajasthan 54.35 10 4035 12 35.44 11 Gujrat 49.35 11 3838 13 24.56 13 Harayana 46.38 12 28.61 16 18.30 16 West Bengal 44.52 13 49.81 8 40.80 7 Assam 44.24 14 44.72 11 38.17 10 Punjab 43.61 15 30.50 15 20.89 15 Maharashtra 41.74 16 33.75 14 23.81 14 The Table: 4.8 above depicts the data on the low-income class trends for each state along with their rankings. In the period between 1996 and 1999 there was a downfall in the proportion of the low-income households. During the period 1990 and 1996 the proportion of low-income households shot up in the states of Assam and West Bengal, whereas the rest of the states followed the same trends. From the table given below it is assessed that in 1990 Kerala had the highest proportion of low-income households; but the proportion declined by almost 35 percent and attained the seventh position in 1996. This further declined by another seven percent to hold the fifth position in 1999. In 1996 and 1999 Orissa observed the highest proportion of low-income households; but it went to the fourth position in 1990. At present Orissa is followed by Bihar, Madhya Pradesh and Uttar Pradesh. These rankings are not only true but are also known facts for Bihar, Madhya Pradesh, Rajasthan and Uttar Pradesh. Uniformity in the decline of low-income households proportion among the states is not being observed. Karnataka, Punjab, Haryana, Kerala and Tamil Nadu show a decline of more than 15 percentage points. During the period 1990-1996 states like Madhya Pradesh, Uttar Pradesh, Andhra Pradesh, Bihar, Orissa and Maharashtra had a fall off of less than 10 percent. At the all India level the proportion of low-income households shot down by more than 9 percentage points during the period 1996-1999. Most of the states followed a similar trend of decline; but only in Bihar and Orissa the decline was less than 1 per cent. 108

Table: 4.9 Percentages of Households in the Lower Middle (LM) Income State 1989-90 Rank 1995-96 Rank 1998-99 Rank Harayana 37.85 1 36.00 3 36.30 6 Assam 35.49 2 37.56 2 42.09 1 Punjab 35.06 3 33.60 6 32.54 11 Maharashtra 34.41 4 38.43 1 39.08 2 Gujrat 34.14 5 34.33 4 35.41 7 Rajasthan 33.92 6 33.76 5 35.32 8 West Bengal 30.23 7 29.33 9 34.30 9 Bihar 28.06 8 27.16 14 28.11 15 Andhra Pradesh 28.06 8 31.34 7 36.67 5 All India 26.95-30.62-34.53 - Himachal Pradesh 26.69 10 27.41 13 31.55 12 Karnataka 23.70 11 28.43 11 31,03 14 Uttar Pradesh 23.64 12 30.74 8 36.82 4 Orissa 23.50 13 24.66 15 24.93 16 Madhya Pradesh 22.18 14 23.02 16 33.28 10 Tamil Nadu 15.66 15 28.55 10 38.37 3 Kerala 12.10 16 28.22 12 31.12 13 Table 4.9 above presents the percentage of households in the lower middle (LM) income class. From the table it is revealed that except the state Punjab all other states had a rise in the proportion of lower middle income households between 1996 and 1999. Almost similar trend is observed between 1990 and 1996 in all states, except the states like Haryana, Punjab, Rajasthan, Bihar and West Bengal. Haryana had the highest proportion of middle-income households in 1990; but the proportion rate decreased by 1.85 per cent in 1996 and thus it moved from the first position to the third position. But thereafter, it attained sixth position among all the states in 1999. In 1999 and thereafter, Assam had the highest proportion of lower middle income and it was followed by Maharashtra, Tamil Nadu and Uttar Pradesh. In most of the states it is found that percentage points have risen highly from 1990 to 1999. During 1996-1999, the proportion of these households gained by more than three per cent at the all India level and it further increased by more than 19 percent during 1990-1999. Table: 4.10 Percentage of Households in the Higher (MH) Income Category State 1989-90 Rank 1995-96 Rank 1998-99 Rank West Bengal 24.24 1 20.86 8 24.90 9 Maharashtra 23.85 2 27.82 3 37.11 4 Punjab 21.34 3 35.95 1 46.57 1 Assam 20.27 4 17.72 10 19.74 12 Gujarat 16.51 5 27.09 4 40.04 3 Himachal Pradesh 16.24 6 25.36 6 29.65 5 Haryana 15.32 7 35.39 2 45.41 2 109

All India 14.21. 20.47-25.81 - Uttar Pradesh 13.98 8 15.91 12 18.59 13 Andhra Pradesh 11.98 9 16.09 11 20.52 11 Rajasthan 11.73 10 25.89 5 29.24 6 Karnataka 10.08 11 21.82 7 28.24 7 Orissa 9.46 12 11.2 16 11.77 16 Bihar 9.44 13 14.54 15 13.95 15 Madhya Pradesh 8.62 14 14.64 14 18.21 14 Tamil Nadu 5.65 15 15.75 13 27.49 8 Kerala 5.08 16 20.79 9 24.63 10 As shown in Table 4.10 above the percentage of households in the Higher Income category (MH) comprising of middle, upper middle and high-income groups increased from 14 to 20 percent at the all India level. However, the individual states show different trends. During the period 1989-90 seven states (West Bengal, Maharashtra, Punjab, Assam, Gujarat, Himachal Pradesh and Haryana) show greater proportion of higher-income households than the all India average of 14.21. The proportion of higher income households was above all-india average in all the three years in the states like Punjab, Haryana, Gujarat, Maharashtra and Himachal Pradesh. Kerala, Karnataka and Rajasthan improved their position in 1995-96 and in 1998-99. Karnataka and Rajasthan retained their old positions while Tamil Nadu moved up the hierarchy. The proportion of MH-category households declined below the national average in case of Kerala. Chart:3 State-wise MH Households, 1998-99 (in per cent) Punjab Gujrat Himachal Pradesh I 29.7 46.6.. 45.4 i 40 137.1 Karnataka All India Kerela Assam Madhya Pradesh Orissa = I 27.5 I 25.8 ------ 1 24.9 ----- 1 24.6 20.5 I 19.7 ] 18.6 I 18.2 14 11.8 0 5 10 15 20 25 30 35 40 45 50 110

As per Table 4.9 above, in 1989-90 West Bengal was on top with the highest percentage of all the three income classes of households which are taken into consideration for the study. But West Bengal thereafter, got a declination in its position by reaching to the eighth position in 1995-96 and it further slipped to the ninth position in 1998-99. The same trend is followed by Assam, which moved from fourth to tenth and then to twelfth position during the same period. This finding is in line with the pace of growth in these states. Orissa and BIMARU states were at the bottom positions during 1998-99. It has been estimated from MISH that at the all India level the household income was Rs. 1,227,653 crore for 1999-2000. The rural areas had a share of 57 per cent in the estimated level. At the all India level, the estimated per household income was Rs.70,082 per annum which was Rs. 1,02,928 per annum for urban households. Among the Union Territories, Chandigarh showed the highest average household income of Rs. 1,69,000 per annum, which was followed by Delhi. Table: 4.11 State Rankings by Average Household Income, 1999-2000 State/UT Rs./annum Rank Chandigarh 169,000 1 Delhi 157,374 2 Punjab 104,875 3 Haryana 98,835 4 Gujrat 94,374 5 Maharashtra 89,709 6 Goa 86,348 7 Tamil Nadu 78,412 8 Himachal Pradesh 77,165 9 Rajasthan 76,426 10 Karnataka 70,948 11 Pondicherry 70,758 12 All India 70,082 - Meghalaya 68,058 13 West Bengal 66,364 14 Kerala 62,752 15 Assam 61,925 16 Andhra Pradesh 61,870 17 Uttar Pradesh 56,364 18 Madhya Pradesh 55,657 19 Bihar 50,259 20 Orissa 41,586 21 Table 4.11 above presents ranking of the states by average household income. Among the bigger states, Punjab, Haryana, Gujarat and Maharashtra are at the top of the hierarchical level. Ill

The following table (Table: 4.12) presents the ranking of the states for urban and rural areas along with the total taking per capita income as an indicator. State/UT Table: 4.12 Per Capita Income of the State (Rural), State (Urban) & State (Rural & Urban), 1999-2000 Rs./ annum Rank Rural Urban Urban&rural Relative Relative State PCI Rs./ to All annum Rank State PCI Rs./ to All annum Rank India PCI India PCI Relative State PCI to All India PCI Chandigarh 27,256 1 2.87 34,509 1 1.78 33,408 1 2.75 Delhi 24,852 2 2.62 29,364 2 1.51 28,885 2 2.38 Punjab 16,540 3 1.74 21,413 7 1.10 18,201 3 1.50 Gujarat 14,574 5 1.54 22,742 6 1.17 17,539 4 1.45 Pondicherry 13,215 6 1.39 18,938 11 0.98 16,912 5 1.39 Tamil Nadu 12,888 7 1.36 24,246 3 1.25 16,790 6 1.38 Maharashtra 11,769 8 1.24 23,747 5 1.22 16,578 7 1.37 Haryana 14,855 4 1.57 18,134 13 0.93 15,841 8 1.31 Goa 11,017 12 1.16 17,440 14 0.90 14,544 9 1.20 Karnataka 11,300 9 1.19 18,394 12 0.95 13,602 10 1.12 Andhra Pradesh 11,033 11 1.16 19,143 10 0.99 13,543 11 1.12 West Bengal 8,792 17 0.93 23,892 4 1.23 13,118 12 1.08 Kerala 10,342 15 1.09 17,372 15 0.90 12,238 13 1.02 All India 9,481-1.00 19,407-1.00 12,128-1.00 Assam 11,109 10 1.17 17,231 16 0.89 11,911 14 0.98 Rajasthan 10,693 14 1.13 15,850 18 0.82 11,905 15 0.98 Himachal Pradesh 10,816 13 1.14 19,881 9 1.02 11,757 16 0.97 Meghalaya 9,248 16 0.98 20,714 8 1.07 11,384 17 0.94 Madhya Pradesh 7,079 18 0.75 14,791 19 0.76 8,758 18 0.72 Uttar Pradesh 6,738 20 0.71 12,257 21 0.63 7,797 19 0.64 Bihar 6,976 19 0.74 12,404 20 0.64 7,736 20 0.64 Orissa 5,704 21 0.60 15,993] 17 0.82 7,123 21 0.59 At the National level PCI is estimated as Rs.12,128 per annum. It is also observed in the table that the combined status of the states and Union Territories like Chandigarh, Delhi, Punjab, Gujarat, Tamil Nadu, Maharashtra and Haryana is nearly 1.3 times the national average. Whereas, the states like Orissa, Bihar, Uttar Pradesh and Madhya Pradesh have much lower PCI in comparison to the national average. When compared with table 4.9 it has been found that PCI has ranked Tamil Nadu and Andhra Pradesh in the higher level of the hierarchy, while Himachal Pradesh and Rajasthan moved down the ladder. As the economic status of the rustic areas is based on their agricultural production, therefore, the states which are more prosperous in the agricultural field are at the top of the hierarchical level. States like Punjab, Haryana, Gujarat, Tamil Nadu and Maharashtra are on the higher rank. 112

The economic condition of the urban areas is influenced by trade and service and therefore their PCIs are also influenced by trade and service. Thus the PCI of Chandigarh, Delhi, Tamil Nadu, West Bengal, Maharashtra, Gujarat and Punjab have higher ranking. And the urban states like the states of Bihar, Madhya Pradesh, Rajasthan and Uttar Pradesh are on the lower rank of the hierarchy. 4.3 Mesurement Of Income Inequality Of Major Indian Cities By Gini And Modified Gini Coefficient: Inequality has been a key theme in research agenda of economists because of (i) the intrinsic ethical attractiveness of equality and (ii) the significant implications of inequality for many macroeconomic variables like growth and poverty. Issues related to inequality can be analysed along several dimensions. So in studying inequality one is always confronted with the issue of how to measure inequality. One of the most commonly used measures of income or consumption inequality is the Gini coefficient. It measures the extent to which the distribution of income or consumption among individuals deviates from equal distribution. A high value of this coefficient signifies a more unequal distribution. Let us consider a population of N individuals with incomes F/, F?,... F\ We can plot the cumulative percentage of total income received cumulative percentages of the population starting from the poorest. Such a plot is called the Lorenz curve. Obviously, for equal incomes, X per cent of the population will always enjoy X per cent of the total income and the Lorenz curve is the diagonal straight line. Otherwise, the curve always lies below this line since the bottom X percentage of population will always have less than X per cent of the total income. Gini coefficient captures the gap between the actual Lorenz curve of the economy and the hypothetical equal distribution line. The Gini formula for inequality measurement is given by: ( b f (y) - M0(y)]f(y)<iy) 2 n

The Gini coefficient is a measure of inequality of a distribution of income. It is defined as a ratio with values between 0 and 1: the numerator is the area between the Lorenz curve of the distribution and the uniform (perfect) distribution line; the denominator is the area under the uniform distribution line. It was developed by the Italian statistician Corrado Gini and published in his 1912 paper "Variabilita e mutabilita" ("Variability and Mutability"). The Gini index is the Gini coefficient expressed as a percentage, and is equal to the Gini coefficient multiplied by 100. (The Gini coefficient is equal to half of the relative mean difference.) The Gini measure is one of the curiosities of statistics involving the correlation of a set of numbers and a weighted sum of the same numbers, insofar as Gini invoked not only the numbers of incomes above a given level, but, in contrast with Pareto, also the aggregate of incomes by those above any given points82. Like the Pareto measure, the properties of the Gini measure remain unaltered when converted to percentage terms from the number of income receivers and individual incomes. The major quality of the Gini measure is that it covers the income distribution down to a much lower income level as compared to the Pareto measure, and is much less confusing in respect of the measure of absolute equality. Both the measures suffer from the limitation that they have no economic interpretation of significant value, nor are they, in their percentage forms, along with the coefficient of concentration derived from the Lorenz curve, capable of taking care of the actual levels of income, when income distributions of very different social groups, at different levels of income are compared for inequality, in terms of their supposed welfare content. They also suffer, in relation to the Lorenz co-efficient of inequality, insofar as they are dependent on the quality of the statistical fit actually obtained33. formula:- Algebraically, the Gini coefficient can be computed using the following G 2NV N N... I X! Yj. Yj ' 1 j=l where Y* is the mean income (y* = (XNj=j Yj)/N) and Yj Yj I is the absolute difference between incomes of individual i and j. It considers difference over all pairs of incomes and there are N2 such pairs of incomes. Hence it can be viewed as one-half of the

mean of absolute values of differences between all income pairs. This measure of inequality has many desirable features. different cities. Cities Now using the above formula we calculate the Gini coefficient for the Table: 4.13 - Income Distribution in 12 Metro Cities 1985-86(in Percent) Deprived Up to 35000 Lower Middle Upper Middle Rich 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.3522 0.3931 0.1698 0.0446 0.0403 Bangalore 0.3689 0.511 0.0939 0.0158 0.0105 Chennai 0.5989 0.2857 0.07 0.036 0.0094 Delhi 0.168 0.355 0.3543 0.0678 0.0548 Hyderabad 0.147 0.4787 0.2992 0.0552 0.0199 Jaipur 0.3772 0.4336 0.1316 0.0442 0.0133 Kanpur 0.4076 0.496 0.0831 0.0088 0.0044 Kolkata 0.2765 0.3496 0.2126 0.0846 0.0766 Lucknow 0.1721 0.436 0.1765 0.1246 0.0908 Mumbai 0.0394 0.2139 0.3621 0.1614 0.2232 Nagpur 0.2925 0.5543 0.1302 0.0165 0.0065 Pune 0.0525 0.3578 0.3835 0.1416 0.0647 Cities Table: 4.14 - Income Distribution In 12 Metro Cities 1994-95 (in Per cent) Lower Middle Deprived Up to 35000 Upper Middle Rich 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.1562 0.4296 0.2469 0.106 0.06 Bangalore 0.2497 0.3456 0.2237 0.116 0.0651 Chennai 0.3131 0.3347 0.1907 0.0924 0.0691 Delhi 0.0655 0.2487 0.2686 0.2013 0.2165 Hyderabad 0.1892 0.3635 0.242 0.1145 0.0906 Jaipur 0.2236 0.3259 0.2939 0.0831 0.0703 Kanpur 0.3636 0.31 0.1772 0.0956 0.0536 Kolkata 0.1813 0.269 0.3022 0.1238 0.1238 Lucknow 0.2619 0.3214 0.2679 0.1101 0.0417 Mumbai 0.0884 0.3631 0.2172 0.1869 0.1441 Nagpur 0.2991 0.2628 0.1662 0.1662 0.1088 Pune 0.1634 0.2106 0.2933 0.189 0.1457

Cities Table :4.15 - Income Distribution In 12 Metro Cities 1995-96 (in per cent) Lower Middle Upper Middle Deprived Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.1135 0.4299 0.2684 0.1268 0.0628 Bangalore 0.2213 0.3483 0.2315 0.1236 0.0753 Chennai 0.2815 0.3524 0.1911 0.0927 0.0815 Delhi 0.0474 0.2355 0.2689 0.2131 0.235 Hyderabad 0.1347 0.3487 0.274 0.1393 0.1033 Jaipur 0.1939 0.3212 0.3091 0.097 0.0788 Kanpur 0.3023 0.3415 0.1932 0.1013 0.0616 Kolkata 0.1217 0.2936 0.3192 0.131 0.1346 Lucknow 0.2237 0.3262 0.2813 0.113 0.0559 Mumbai 0.0547 0.3331 0.2449 0.2086 0.1591 Nagpur 0.2529 0.2238 0.1948 0.1948 0.1366 Pune 0.1285 0.1664 0.3138 0.2155 0.1777 Table :4.16 - Income Distribution In 12 Metro Cities 1996-97 (in per cent) Lower Middle Upper Middle Deprived Rich Cities Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.0874 0.3286 0.3116 0.1706 0.1017 Bangalore 0.1764 0.3457 0.2411 0.1335 0.1034 Chennai 0.2165 0.3501 0.2126 0.1098 0.1111 Delhi 0.0353 0.2048 0.2546 0.2096 0.2957 Hyderabad 0.1268 0.3162 0.2884 0.1553 0.1134 Jaipur 0.1787 0.321 0.277 0.1411 0.0821 Kanpur 0.264 0.3145 0.2029 0.1251 0.0935 Kolkata 0.096 0.275 0.3135 0.1546 0.1609 Lucknow 0.2001 0.3247 0.2501 0.139 0.0861 Mumbai 0.0459 0.2993 0.2516 0.2077 0.1956 Nagpur 0.201 0.2863 0.1964 0.1841 0.1322 Pune 0.103 0.1812 0.2858 0.2157 0.2142 116

Cities Table :4.17 - Income Distribution In 12 Metro Cities 1997-98 (in per cent) Lower Middle Upper Middle Deprived Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.057 0.3342 0.3096 0.1645 0.1347 Bangalore 0.1345 0.3243 0.244 0.1508 0.1464 Chennai 0.1449 0.3268 0.2417 0.1323 0.1543 Delhi 0.0287 0.188 0.2477 0.2118 0.3239 Hyderabad 0.1139 0.3006 0.2927 0.1648 0.128 Jaipur 0.1778 0.3207 0.3149 0.102 0.0845 Kanpur 0.2559 0.3548 0.2043 0.1075 0.0774 Kolkata 0.0859 0.2692 0.3159 0.1621 0.1668 Lucknow 0.1936 0.3236 0.2865 0.122 0.0716 Mumbai 0.0373 0.2775 0.2535 0.203 0.2287 Nagpur 0.1889 0/2222 0.2083 0.1944 0.1861 Pune 0.0914 0.1523 0.2867 0.2294 0.2401 Cities Table :4.18 - Income Distribution In 12 Metro Cities 1998-99 (in per cent) Deprived Lower Middle Upper Middle Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.0482 0.294 0.2991 0.1749 0.1838 Bangalore 0.1128 0.3043 0.2404 0.1596 0.183 Chennai 0.1271 0.2938 0.2465 0.1442 0.1884 Delhi 0.0218 0.1625 0.2269 0.2065 0.3824 Hyderabad 0.1014 0.2715 0.2895 0.1804 0.1572 Jaipur 0.1591 0.3097 0.3097 0.1193 0.1023 Kanpur 0.2264 0.3543 0.2117 0.1153 0.0922 Kolkata 0.0726 0.2507 0.3081 0.1797 0.189 Lucknow 0.1769 0.3205 0.2846 0.1308 0.0872 Mumbai 0.0305 0.2393 0.2467 0.1987 0.2847 Nagpur 0.159 0.2102 0.2075 0.1968 0.2264 Pune 0.0744 0.1367 0.263 0.237 0.2889 117

Cities Table :4.19 - Income Distribution In 12 Metro Cities 1985-86 (Cum. Propn. Of Person) Deprived Lower Middle Upper Middle Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.3522 0.7453 0.9151 0.9597 1 Bangalore 0.3689 0.8799 0.9738 0.9896 1.0001 Chennai 0.5989 0.8846 0.9546 0.9906 1 Delhi 0.168 0.523 0.8773 0.9451 0.9999 Hyderabad 0.147 0.6257 0.9249 0.9801 1 Jaipur 0.3772 0.3108 0.9424 0.9866 0.9999 Kanpur 0.4076 0.9036 0.9867 0.9955 0.9999 Kolkata 0.2765 0.6261 0.8387 0.9233 0.9999 Lucknow 0.1721 0.6081 0.7846 0.9092 1 Mumbai 0.0394 0.2533 0.6154 0.7768 1 Nagpur 0.2925 0.8468 0.977 0.9935 1 Pune 0.0525 0.4103 0.7938 0.9354 1.0001 Cities Table :4.20 - Income Distribution In 12 Metro Cities 1985-86 (Cum. Propn. Of Person) Deprived Lower Middle Upper Middle Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.1152713 0.5012437 0.77911 0.88129 1 Bangalore 0.1431787 0.73817194 0.9204 0.96332 1 Chennai 0.2795202 0.67954821 0.8429 0.96052 1 Delhi 0.0422929 0.31039952 0.75636 0.87584 1 Hyderabad 0.0403336 0.43436865 0.84484 0.95086 1 Jaipur 0.1364146 0.60684966 0.84482 0.95671 1 Kanpur 0.1689674 0.78580608 0.95805 0.98358 1 Kolkata 0.0753427 0.36112701 0.65078 0.81215 1 Lucknow 0.0424729 0.36527641 0.58307 0.79832 1 Mumbai 0.006998 0.12097261 0.44254 0.64321 1 Nagpur 0.1052007 0.7032801 0.93742 0.97896 1 Pune 0.0113713 0.24386493 0.65919 0.87388 1 118

Cities Table: 4.21 - Income Distribution in 12 Metro Cities 1994-95 (Cum. Propn. of person) Deprived Lower Middle Upper Middle Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.1562 0.5858 0.8327 0.9387 0.9987 Bangalore 0.2497 0.5953 0.819 0.935 1.0001 Chennai 0.3131 0.6478 0.8385 0.9309 1 Delhi 0.0655 0.3142 0.5828 0.7841 1.0006 Hyderabad 0.1892 0.5527 0.7947 0.9092 0.9998 Jaipur 0.2236 0.5495 0.8434 0.9265 0.9968 Kanpur 0.3636 0.6736 0.8508 0.9464 1 Kolkata 0.1813 0.4503 0.7525 0.8763 1.0001 Lucknow 0.2619 0.5833 0.8512 0.9613 1.003 Mumbai 0.0884 0.4515 0.6687 0.8556 0.9997 Nagpur 0.2991 0.5619 0.7281 0.8943 1.0031 Pune 0.1634 0.374 0.6673 0.8563 1.002 Table: 4.22 - Income Distribution in 12 Metro Cities 1994-95 (Cum. Propn. of person) Lower Middle Upper Middle Deprived Rich Cities Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.0394295 0.36476082 0.67639 0.86369 1 Bangalore 0.0656604 0.33329446 0.63241 0.84593 1 Chennai 0.0884613 0.37215347 0.64155 0.82429 1 Delhi 0.0118827 0.14723704 0.39088 0.64651 1 Hyderabad 0.0460722 0.31162032 0.60627 0.80144 1 Jaipur 0.0575517 0.30919901 0.68743 0.83715 1 Kanpur 0.1091499 0.38332853 0.6543 0.85519 1 Kolkata 0.0404678 0.22059775 0.55787 0.7513 1 Lucknow 0.0705626 0.33034271 0.69124 0.89888 1 Mumbai 0.0181561 0.24188215 0.46493 0.73364 1 Nagpur 0.07365 0.26778459 0.47241 0.75888 1 Pune 0.0333742 0.1624183 0.46195 0.73217 1 119

Cities Table: 4.23 - Income Distribution in 12 Metro Cities 1995-96 (Cum. Propn. of person) Deprived Lower Middle Upper Middle Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.1135 0.5434 0.8118 0.9386 1.0014 Bangalore 0.2213 0.5696 0.8011 0.9247 1 Chennai 0.2815 0.6339 0.825 0.9177 0.9992 Delhi 0.0474 0.2829 0.5518 0.7649 0.9999 Hyderabad 0.1347 0.4834 0.7574 0.8967 1 Jaipur 0.1939 0.5151 0.8242 0.9212 1 Kanpur 0.3023 0.6438 0.837 0.9383 0.9999 Kolkata 0.1217 0.4153 0.7345 0.8655 1.0001 Lucknow 0.2237 0.5499 0.8312 0.9442 1.0001 Mumbai 0.0547 0.3878 0.6327 0.8413 1.0004 Nagpur 0.2529 0.4767 0.6715 0.8663 1.0029 Pune 0.1285 0.2949 0.6087 0.8242 1.0019 Table: 4.24 - Income Distribution in 12 Metro Cities 1995-96 (Cum. Propn. of person) Cities Deprived Lower Middle Upper Middle Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.0270367 0.33425441 0.65393 0,86536 1 Bangalore 0.0557909 0.31921545 0.61103 0.82915 1 Chennai 0.0765653 0.36411358 0.624 0.8005 1 Delhi 0.0083084 0.13214492 0.36781 0.62928 I Hyderabad 0.0302316 0.26501481 0.57249 0.79134 1 Jaipur 0.0473944 0.28292433 0.66069 0.82665 1 Kanpur 0.0850041 0.37308439 0.64472 0.84411 1 Kolkata 0.0257463 0.21208403 0.54973 0.74372 1 Lucknow 0.0573164 0.30805299 0.66843 0.8711 1 Mumbai 0.010579 0.20384481 0.44066 0.72307 1 Nagpur 0.0563089 0.20579788 0.42266 0.72627 I Pune 0.0242247 0.11833349 0.41412 0.6985 1 120

Cities Table: 4.25 - Income Distribution in 12 Metro Cities 1996-97 (Cum. Propn. of person) Deprived Lower Middle Upper Middle Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.0874 0.416 0.7276 0.8982 0.9999 Bangalore 0.1764 0.5221 0.7632 0.8967 1.0001 Chennai 0.2165 0-5666 0.7792 0.889 1.0001 Delhi 0.0353 02401 0.4947 0.7043 1 Hyderabad 0.1268 0.443 0.7314 0.8867 1.0001 Jaipur 0.1787 0.4997 0.7767 0.9178 0.9999 Kanpur 0.264 0-5785 0.7814 0.9065 1 Kolkata 0.096 0.371 0.6845 0.8391 1 Lucknow 0.2001 0-5248 0.7749 0.9139 1 Mumbai 0.0459 03452 0.5968 0.8045 1.0001 Nagpur 0.201 0.4873 0.6837 0.8678 1 Pune 0.103 02842 0.57 0.7857 0.9999 Cities Table: 4.26 - Income Distribution in 12 Metro Cities 1996-97 (Cum. Propn. of person) Deprived Lower Middle Upper Middle Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.0184361 0.22638007 0.55502 0.80693 1 Bangalore 0.0411755 0.28325669 0.56465 0.78278 1 Chennai 0.0528268 0.30910377 0.56848 0.75602 1 Delhi 0.0058336 0.10736713 0.31774 0.5602 1 Hyderabad 0.0274156 0.23251389 0.54429 0.77933 1 Jaipur 0.0420144 0.26842687 0.59406 0.82628 1 Kanpur 0.0670187 0.30653432 0.56407 0.78638 1 Kolkata 0.0191281 0.13351 0.49584 0.71146 1 Lucknow 0.0479557 0.28140728 0.5811 0.81429 1 Mumbai 0.0084747 0.17425823 0.40653 0.67497 1 Nagpur 0.0444651 0.2344704 0.45171 0.73679 1 Pune 0.0186821 0.11728003 0.37647 0.65034 1 121

Cities Table: 4.27 - Income Distribution In 12 Metro Cities 1997-98 (Cum. Propn. of person) Deprived Lower Middle Upper Middle Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.057 0.3912 0.7008 0.8653 1 Bangalore 0.1345 0.4588 0.7028 0.8536 1 Chennai 0.1449 0.4717 0.7134 0.8457 1 Delhi 0.0287 0.2167 0.4644 0.6762 1.0001 Hyderabad 0.1139 0.4145 0.7072 0.872 1 Jaipur 0.1778 0.4985 0.8134 0.9154 0.9999 Kanpur 0.2559 0.6107 0.815 0.9225 0.9999 Kolkata 0.0859 0.3551 0.671 0.8331 0.9999 Lucknow 0.1936 0.5172 0.8037 0.9257 0.9973 Mumbai 0.0373 0.3148 0.5683 0.7713 1 Nagpur 0.1889 0.4111 0.6194 0.8138 0.9999 Pune 0.0914 0.2437 0.5304 0.7598 0.9999 Cities Table: 4.28 - Income Distribution in 12 Metro Cities 1997-98 (Cum. Propn. of person) Deprived Lower Middle Upper Middle Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.0114656 0.21313916 0.52452 0.75615 1 Bangalore 0.0286134 0.23558695 0.49513 0.7197 1 Chennai 0.0311707 0.24207288 0.50204 0.70126 1 Delhi 0.0046076 0.09515324 0.29398 0.532 1 Hyderabad 0.0238045 0.21227638 0.51814 0.75924 1 Jaipur 0.0424455 0.27212395 0.648 0.81845 1 Kanpur 0.0675038 0.34828141 0.61774 0.81624 1 Kolkata 0.0168138 0.17489088 0.48406 0.70616 1 Lucknow 0.0472737 0.28432593 0.63412 0.84265 1 Mumbai 0.006641 0.15486237 0.38053 0.63353 1 Nagpur 0.0382955 0.17343443 0.38458 0.66045 1 Pune 0.0158998 0.0953814 0.34475 0.62409 1 122

Cities Table: 4.29 - Income Distribution in 12 Metro Cities 1998-99 (Cum. Propn. of person) Deprived Lower Middle Upper Middle Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.0482 0.3422 0.6413 0.8162 1 Bangalore 0.1128 0.4171 0.6575 0.8171 1.0001 Chennai 0.1271 0.4209 0.6674 0.8116 1 Delhi 0.0218 0.3843 0.4112 0.6177 1.0001 Hyderabad 0.1014 0.3729 0.6624 0.8428 1 Jaipur 0.1591 0.4688 0.7785 0.8978 1.0001 Kanpur 0.2264 0.5807 0.7924 0.9077 0.9999 Kolkata 0.0726 0.3233 0.6314 0.8111 1.0001 Lucknow 0.1769 0.4974 0.782 0.9128 1 Mumbai 0.0305 0.2698 0.5165 0.7152 0.9999 Nagpur 0.159 0.3692 0.5767 0.7735 0.9999 Pune 0.0744 0.2111 0.4741 0.7111 1 Cities Table: 4.30 - Income Distribution in 12 Metro Cities 1998-99 (Cum. Propn. of person) Deprived Lower Middle Upper Middle Rich Up to 35000 35000-70,000 70,001-105,000 105,001-140,000 140,001 17500 52500 87500 122500 157500 Ahmedabad 0.0090871 0.17537046 0.45732 0.68813 1 Bangalore 0.0225966 0.20547287 0.44626 0.67007 1 Chennai 0.0256975 0.20390214 0.45309 0.65718 1 Delhi 0.003338 0.07798313 0.2517 0.47303 1 Hyderabad 0.0201151 0.18169014 0.46884 0.71934 1 Jaipur 0.0362208 0.24774047 0.60027 0.79039 1 Kanpur 0.0568173 0.32356263 0.5892 0.79175 1 Kolkata 0.0136361 0.15489942 0.44424 0.68051 1 Lucknow 0.0415083 0.26711718 0.60101 0.81585 1 Mumbai 0.0051389 0.12609729 0.33393 0.56828 1 Nagpur 0.0303302 0.15062091 0.34853 0.61132 1 Pune 0.0122801 0.07996897 0.29702 0.57084 1 123

4.4 Lorenz Curve: In this context we have used Lorenz curve to show the degree of inequality in the distribution of income in the different cities of India by means of joint cumulation, both of the frequencies and of the values of the variable distributed. We have carried out our analysis in terms of income I in the following way:- Let P(x) = cumulative proportion of persons having income < x in total population. = No. of individuals having income < x / Total no. of individuals in the economy and Q(x) = cumulative share of the individual in national income of the economy = income earned by the people having income < x / National income of the economy. Table. 2 B - Lorenz Curve On Income Distribution In 12 Metro Cities Ahmedabad: 1998-99 Cum. Prop, of Person Chennai: 1998-99 Cum. Prop, of Income Cum Prop of Income Cum. Prop, of Income Cum. Prop, of Income 20 40 60 80 100 Cum. Prop, of Person Cum. Prop, of Person 124

Hyderabad: 1998-99 Table. 2 B - Lorenz Curve On Income Distribution In 12 Metro Cities Kolkata: 1998-99 Kanpur:1998-99 20 40 60 80 100 Cum. Prop, of Person Lucknow: 1998 99 Nagpur:1998-99 Cum. Prop, of Income Cum. Prop, of Income Cum. Prop, of Income ium. Prop, of Income 20 40 60 80 Cum. Prop, of Person 100 20 40 60 80 100 Cum Prop, of Person Cum. Prop, of Income co o o o * o <N 125

Table31 Mesurement Of Income Inequality Of Major Indian Cities By Gini Coefficient Gini coeff. - Total Cities 1985-86 1994-95 1995-96 1996-97 1997-98 1998-99 Value Rank Value Rank Value Rank Value Rank Value Rank Value Rank Ahmedabad 0.1749 11 0.1440 3 0.1333 4 0.1292 4 0.1243 4 0.1227 4 Bangalore 0.1453 7 0.1676 9 0.1630 9 0.1560 8 0.1486 7 0.1438 7 Chennai 0.1843 12 0.1817 10 0.1778 11 0.1674 11 0.1523 9 0.1475 9 Delhi 0.1381 5 0.1249 1 0.1184 1 0.1115 1 0.1068 1 0.0995 1 Hyderabad 0.1231 3 0.1574 6 0.1442 6 0.1415 6 0.1384 6 0.1351 6 Jaipur 0.1648 9 0.1594 7 0.1524 7 0.1518 7 0.1493 8 0.1473 8 Kanpur 0.1434 6 0.1888 12 0.1784 12 0.1761 12 0.1716 12 0.1674 12 Kolkata 0.1744 10 0.1541 5 0.1399 5 0.1337 5 0.1306 5 0.1262 5 Lucknow 0.1561 8 0.1624 8 0.1574 8 0.1588 9 0.1542 10 0.1515 11 Mumbai 0.1131 2 0.1383 2 0.1266 2 0.1233 2 0.1199 2 0.1146 3 Nagpur 0.1315 4 0.1846 11 0.1725 10 0.1640 10 0.1587 11 0.1499 10 Pune 0.1117 1 0.1461 4 0.1322 3 0.1287 3 0.1216 3 0.1139 2 4.5 Modified Gini Coefficient: Chakrabarty, R., Et all. have suggested a modification in the Gini coefficient measure. In devising the refinement of the criteria prescribed by Yntema have been broadly kept in mind. They are : (i) an accepted co-efficient must be both independent of the number of persons and the unit of measurement involved in a distribution, (ii) the numerical value of the co-efficient should not be difficult to compute, (iii) the co-efficient must be reasonably adoptable to the peculiarities of the data, (iv) it should admit of reasonable economic interpretation, which is in harmony with the connotation of the word inequality, and (v) the measure should have finite limits. 4.6 Proposed Modification: As already indicated the proposed measure seeks to use the Lorenz curve approach and is built round it. Thus let yt be the mid income of the i-th class and fi be the frequency of the i-th class. Yt's are non-negative. 126

If we write, F(y)=lf(y)dy And (y)dy it is obvious that <D(y) exists, only if pr exists. Here, F(y) varies from 0 to 1, and d>(y) also varies from 0 to 1, provided that the origin is taken to the left of the start of the frequency distribution. Lorenz curve. Now plotting O against F, we would get the curve of concentration or the The Lorenz ratio of inequality is given by, L = 1-2 <t>{y)df (;y) = 1 2E^y)] The peculiarity of this ratio is that the lower and the upper tail ends of the curve have the same weight. Now, it may well happen that two Lorenz curves intersect one another. And if the area under the curves are the same i.e. if the L-ratios of the two curves are identical it would not be possible to compare the inequality coefficients, though it is apparent that one is more unequal than the other. The lacuna can, be traced to the fact that the concept of inequality is not unidimensional. It has at teast two aspects. The lower end of the Lorenz curve has the deprivation or paucity of income aspect while the upper end has the concentration of power aspect. And a measure determining either of them would require appropriate extraneous weights to be super-imposed. If by inequality one means to measure the paucity of income then higher weightage should be given to the lower end of the Lorenz curve. And if, on the other hand, one tries to measure 127

the concentration of power then higher weightage should be given at the upper end. Now in order to measure the paucity of income, we define as f (y) a measure of equality in the range (0,y), Then E ^F(y)) * F(y) = </>(y)d log F (y) may be taken as an over all measure of equality provided such expectation exists. Then the over all measure of inequality is defind as F (y)- <P (>-) 1 rf(y)-hy) 1 - E (4>(y)) = E = P ~df {y). {F(y)j l F(y)J F (y) F(y) Sw 1 is- 1! hj Comparing this measure with the Lorenz ratio of inequality, we see that in this measure 0(y) has weight 1. And since the rate of change of F(y) in the lower range of F(y) values of y is greater than that of <p(y) for the same range, we can say that the measure gives more importance to the lower values of y than to higher ones of y. So the desired criterion is satisfied. If the C-curve, given by V=F(y) against u'= is drawn one would get a curve F(y) like the following: Egalitarian line Here the egalitarian line is one parallel to the y-axis and not the 45 line, as in the Lorenz Curve. The base line is the line of extreme inequality. For different <p (y)s e.g Pareto distribution, log normal distribution etc, the C- curve and the C-ratio are being worked out. For empirical work this C-ratio can be obtained from the decile points. 128

Let us define, <p(& =^ yf (y)dy When is are deciles and^ = J f(y)dy =F{%) jx 1 Thus gi's are the decile points of the f-distribution. If for each 'i' we calculate, K = <b (E.) / it is obvious that K must lie between 0 and 1. Then the ' 10 over all measure of inequality will be given by 1 n V 1 = 10 $ ( j) 10 lu f=l : ^ / = 10 <p (g i )...... C = 1 - ----------------------- l------- = 1-2j 1 this will vary io i between 0 and 1, irrespective of whether 10 is substituted by 100, 1000, or 1,000,000. One principal feature of the statistic is that its sensitivity can be increased to any desired level without affecting its character. For measuring the concentration of power, the C-ratio may be modified by giving higher weightages at the upper tail end. For that purpose we consider an alternative measure of equality as J (/> ( y ) ydy F ( y ) 10' ' ydy where yi is the upper limit of income. The corresponding inequality of concentration of power C' may be defined as C ' = 1 - f v ^ y ^ J F {y) J/1 ydy F(.y)-fl(y) Jo F ( y ) ydy rydy For empirical work we can write, 10 0 (, ) g Then the measure of inequality of concentration of power is given by C' = 1 e'. Before proceeding further, let us note that the criteria (i) (iii) and (v) as suggested by Yntema and mentioned above are satisfied by our coefficient of inequality. As regards the criteria (iv) a little bit of explanation is necessary. It can be easily seen that the Pigou-Dalton criterion, according to 129

which any transfer of income from a poorer person to a richer one must increase inequality, is satisfied both by the Ginni's coefficient and the C coefficient of the present measure. The next stage of sophistication is to make the measure sensitive to: the level of income at which the transfer takes place. It is an accepted fact that a transfer of a given amount of money from a relatively-richer to a relatively poorer person of a higher level of income is not the same thing as the transfer of the same amount of money from a relatively richer person to a relatively poorer person of a lower level of income. As pointed out by Prof. Sen, the Ginni coefficient is rather weak on this point. The new coefficient C passes this test also. For (j) (y) is being weighted by the reciprocal of F(y) and the rate pf increase of F(y) invariably decreases as y increases. Hence, the measure will show greater sensitivity to a transfer of a given amount of money from a man having 1,000 to one having 900, than to the transfer of the same amount of money from a man having 1,000,100 to a man having 1,000,000. The new measure c = \-E rhy)} indeed sensitive to the skew-ness in the F(y) income distribution. Expanding 0 ( y) / F ( y) in a Taylor series around the mean value of / y,//,, we get, <P(y) F(y) df(y) C = l r f )_ i> 1 // [F(m 0 + (y-//i) f(m'o. + 2! 12 'yji 0A + (J-/031 ( 31 f(m') +, df(y)- = 1- (pim) ] "hf F(m) 2\{f(ju')J + h f +. 3\{FQ/)J Where m2, m3, are the 2nd and the 3rd central moments of the income distribution. The effect of skewness on C is determined by the sign of(^(y)/f(y))m. It is not 130

clear at once what the sign of [<p/f)m would be. Whether this new measure C is more sensitive to skewness than the Ginni coefficient depends on the actual magnitudes of (<p{y)lf{y))"' and(<zt>/f)m this is a very complicated exercise. So it can not be said at once which one measure is more sensitive to skewness and what is the nature of skewness for the two measures. The proposed coefficient C also possesses the properties of completeness and transitivity. If C and C2 are two distinct values of C, it must either be that C]>C2 or C <C2 or Ci = C2 ; and if C >C2 and C2>C3, then C >C3; if the transitivity condition is to be satisfied. Further, for Lorenz curves 0,(y), (p2 (y) and <p2 (y), showing different distributions of incomes, it is possible to calculate the corresponding inequality coefficients C ^ and for distributions of the type 0,(y) and <f)2{y) values of C^ and C02 have special properties because in case of intersections, the Ginni coefficient for these two curves may be the same. But and, would be necessarily different, since E = (p(y)l F{y) or <p{^l)li would be different for the two curves. With slabs as broad as deciles, it is impossible that even in such a formulation)/l, 0( 2)/2, etc. would turn out to be the same; but even if it turns out to be the same it can easily be rectified by introducing any desired degree of fineness in the slabs. And thus the vexing problem of intersecting Lorenz curves, that do not fulfil condition (2), can be avoided. The measure is independent of any quirks of the utility function, subject probably to the Atkinson conditions. It is also to be noted that no information is lost in the process of working on the coefficient and given the mean value derived from the 131

original units, if we multiply C with the value of the mean, the measure can be made to take care of the levels of the incomes concerned. This is an important shortcoming of the Lorenz curve approach in its original form. Further, given workable transformation ratios between the units of measurement, the coefficient can be extended to diverse areas of investigation. Cities Table: 4.32- Rank & Estimate of Measure of Inequality. On Income Distribution In 12 Metro Cities - MOI Modified GINI (MOI) 1985-86 1994-95 1995-96 1996-97 1997-98 1998-99 Value Rank Value Rank Value Rank Value Rank Value Rank Value Rank Ahmedabad 0.8482 7 0.7431 6 0.7509 7 0.6736 5 0.6450 5 0.6356 5 Bangalore 0.8644 9 0.7649 9 0.7687 10 0.7178 9 0.6999 7 0.6848 8 Chennai 0.9270 12 0.8093 12 0.7943 11 0.7363 10 0.7007 9 0.6861 9 Delhi 0.8069 6 0.5887 1 0.5594 1 0.5536 1 0.5367 1 0.5352 1 Hyderabad 0.8038 5 0.7277 5 0.6831 4 0.6918 6 0.6994 6 0.6416 6 Jaipur 0.8681 10 0.7691 10 0.7567 8 0.6984 8 0.7269 10 0.6869 10 Kanpur 0.9136 11 0.8079 11 0.8081 12 0.7524 12 0.7959 12 0.7469 12 Kolkata 0.7933 4 0.6985 4 0.7008 5 0.6212 4 0.6268 4 0.6430 7 Lucknow 0.7651 3 0.7626 3 0.7677 9 0.7493 11 0.7571 11 0.6983 11 Mumbai 0.6011 1 0.6541 3 0.6238 2 0.6085 3 0.5945 3 0.5837 3 Nagpur 0.8540 8 0.7443 7 0.7054 6 0.6927 7 0.7000 8 0.6272 4 Pune 0.6841 2 0.6519 2 0.6445 3 0.5790 2 0.5825 2 0.5671 2 Cities Table: 4.33 - Rank & Estimate of Measure of Inequality On Income Distribution in 12 Metro Cities - POC Modified GINI (POC) 1985-86 1994-95 1995-96 1996-97 1997-98 1998-99 Value Rank Value Rank Value Rank Value Rank Value Rank Value Rank Ahmedabad 0.8459 7 0.7185 8 0.7120 7 0.6090 5 0.5754 5 0.5394 5 Bangalore 0.8712 10 0.7413 9 0.7361 10 0.6675 9 0.6317 8 0.5962 8 Chennai 0.9149 12 0.7816 11 0.7705 11 0.6895 10 0.6376 9 0.6016 9 Delhi 0.7729 5 0.4685 1 0.4355 1 0.4188 1 0.3671 1 0.3678 1 Hyderabad 0.7934 6 0.6969 6 0.6302 5 0.6258 7 0.6229 7 0.5562 6 Jaipur 0.8685 9 0.6481 5 0.7269 8 0.6663 8 0.7013 10 0.6388 10 Kanpur 0.9132 11 0.7902 12 0.7788 12 0.7210 12 0.7652 12 0.7133 12 Kolkata 0.7696 4 0.6440 4 0.6290 4 0.5495 4 0.5468 4 0.5476 7 Lucknow 0.7273 3 0.7436 10 0.7359 9 0.7022 11 0.7251 11 0.6661 11 Mumbai 0.4835 1 0.5801 3 0.5380 2 0.5047 3 0.4647 3 0.4495 3 Nagpur 0.8584 8 0.7001 7 0.6393 6 0.6245 6 0.6192 6 0.5226 4 Pune 0.6319 2 0.5758 2 0.5506 3 0.4584 2 0.4560 2 0.4217 2 132

Cities Table: 4.34 - Rank & Estimate Of Measure Of Inequality On Income Distribution In 12 Metro Cities - STD. GIN! C Gini coeff. - Total 1985-86 1994-95 1995-96 1996-97 1997-98 1998-99 Value Rank Value Rank Value Rank Value Rank Value Rank Value Rank Ahmedabad 0.1749 11 0.1440 3 0.1333 4 0.1292 4 0.1243 4 0.1227 4 Bangalore 0.1453 7 0.1676 9 0.1630 9 0.1560 8 0.1486 7 0.1438 7 Chennai 0.1843 12 0.1817 10 0.1778 11 0.1674 11 0.1523 9 0.1475 9 Delhi 0.1381 5 0.1249 1 0.1184 1 0.1115 1 0.1068 1 0.0995 1 Hyderabad 0.1231 3 0.1574 6 0.1442 6 0.1415 6 0.1384 6 0.1351 6 Jaipur 0.1648 9 0.1594 7 0.1524 7 0.1518 7 0.1493 8 0.1473 8 Kanpur 0.1434 6 0.1888 12 0.1784 12 0.1761 12 0.1716 12 0.1674 12 Kolkata 0.1744 10 0.1541 5 0.1399 5 0.1337 5 0.1306 5 0.1262 5 Lucknow 0.1561 8 0.1624 8 0.1574 8 0.1588 9 0.1542 10 0.1515 11 Mumbai 0.1131 2 0.1383 2 0.1266 2 0.1233 2 0.1199 2 0.1146 3 Nagpur 0.1315 4 0.1846 11 0.1725 10 0.1640 10 0.1587 11 0.1499 10 Pune 0.1117 1 0.1461 4 0.1322 3 0.1287 3 0.1216 3 0.1139 2 Table: 4.35 - Rank & Estimate Of Measure Of Inequality On Income Distribution In 12 Metro Cities * MOI Modified GINI ( MOI) Cities 1985-86 1994-95 1995-96 1996-97 1997-98 1998-99 Value Rank Value Rank Value Rank Value Rank Value Rank Value Rank Ahmedabad 0.8482 7 0.7431 6 0.7509 7 0.6736 5 0.6450 5 0.6356 5 Bangalore 0.8644 9 0.7649 9 0.7687 10 0.7178 9 0.6999 7 0.6848 8 Chennai 0.9270 12 0.8093 12 0.7943 11 0.7363 10 0.7007 9 0.6861 9 Delhi 0.8069 6 0.5887 1 0.5594 1 0.5536 1 0.5367 1 0.5352 1 Hyderabad 0.8038 5 0.7277 5 0.6831 4 0.6918 6 0.6994 6 0.6416 6 Jaipur 0.8681 10 0.7691 10 0.7567 8 0.6984 8 0.7269 10 0.6869 10 Kanpur 0.9136 11 0.8079 11 0.8081 12 0.7524 12 0.7959 12 0.7469 12 Kolkata 0.7933 4 0.6985 4 0.7008 5 0.6212 4 D.6268 4 0.6430 7 Lucknow 0.7651 3 0.7626 8 0.7677 9 0.7493 11 0.7571 11 0.6983 11 Mumbai 0.6011 1 0.6541 3 0.6238 2 0.6085 3 0.5945 3 0.5837 3 Nagpur 0.8540 8 0.7443 7 0.7054 6 0.6927 7 0.7000 8 0.6272 4 Pune 0.6841 2 0.6519 2 0.6445 3 0.5790 2 0.5825 2 0.5671 2

Table: 4.35 - Rank & Estimate Of Measure Of Inequality On Income Distribution In 12 Metro Cities - PCO Cities Modified GiNi ( POC ) 1985-86 1994-95 1995-96 1996-97 1997-98 1998-99 Value Rank Value Rank Value Rank Value Rank Value Rank Value Rank Ahmedabad 0.8459 7 0.7185 8 0.7120 7 0.6090 5 0.5754 5 0.5394 5 Bangalore 0.8712 10 0.7413 9 0.7361 10 0.6675 9 0.6317 8 0.5962 8 Chennai 0.9149 12 0.7816 11 0.7705 11 0.6895 10 0.6376 9 0.6016 9 Delhi 0.7729 5 0.4685 1 0.4355 1 0.4188 1 0.3671 1 0.3678 1 Hyderabad 0.7934 6 0.6969 0.6302 5 0.6258 7 0.6229 7 0.5562 Jaipur 0.8685 9 0.6481 5 0.7269 8 0.6663 8 0.7013 10 0.6388 10 Kanpur 0.9132 11 0.7902 12 0.7788 12 0.7210 12 0.7652 12 0.7133 12 Kolkata 0.7696 4 0.6440 4 0.6290 4 0.5495 4 0.5468 4 0.5476 7 Lucknow 0.7273 3 0.7436 10 0.7359 9 0.7022 11 0.7251 11 0.6661 11 Mumbai 0.4835 1 0.5801 3 0.5380 2 0.5047 3 0.4647 3 0.4495 3 Nagpur 0.8584 8 0.7001 7 0.6393 6 0.6245 6 0.6192 6 0.5226 4 Pune 0.6319 2 0.5758 2 0.5506 3 0.4584 2 0.4560 2 0.4217 2 4.7 The Topsis Method: Next we compare the income distribution of all the cities with the help of TOPSIS method. This method is being applied to compare the relative positions of the different cities taking into account the different income groups. As there are different groups with different proportions so if one is interested to compare the positions, one should use the Multi Criterion Decision Method. Procedure for the Topsis result is shown beiow:- In a typical MCDM environment, there are a number of alternatives to be assessed on the basis of their preference order. There are many MCDM techniques available (Hwant and YOON, 1981, Zeleny 1983 and Yoon and Hwang 1995), among which the technique for order preference by similarity to ideal solution (TOPSIS) proposed by Yoon (1980) and Hwang and Yoon (1981) is a very intuitive and effective one. The basic principle employed by TOPSIS is that the best alternative should have the shortest distance from the ideal alternative, it is both intuitive, feasible and important. We now provide a review of the TOPSIS method. 134

4.7.1 The MCDM environment: Suppose that there are all together K alternatives to be assessed and the best alternative is to be selected. Let the alternatives be denoted by Si,...S*. there are also N criteria identified to assess the alternatives, which are denoted by Ci,...Cn. the k-th alternative s value on the n-th criteria is obtained as xk, and we write Sk = (x id,...x kn), 1,...,K, and Cn = (x... xk ), n = 1,...N. 4.7.2 The Ideal Solution: It is both intuitive a feasible to compare each alternative with an ideal alternative to solve the assessment or decision making problem. TPOSIS adopts an intuitive approach to the construction of the best and worst alternatives and calls them the ideal and the negative-ideal alternatives or solutuions. The ideal alternative S+, is formed by taking all the best values attained on each criterion by some alternatives, and can be explicitly denoted by: S+ = (x+i,..., x +N) = (min (x*,},...,min {xkm},max {x^ + i},..., max {x*n})- k k k k and the negative-ideal alternative Scomprises of all the worst criterion values attained by some alternatives, and is denoted by S- = (x.i,..., x.n) = (max {xh], max {xkm},min {xkm + i},..., min {x*n})- k k k k 4.8 The Topsis Procedure: With the above notation and explanation, the TOPSI Sprocedure for assessing the ranking can be described as follows: * Normalise the n-th criterion vector Cn in to TCn: TC, = c / c =U, / C II,...,*» / Cj) = y,...f*).n = l,...,ar, where C, = 2> )! is the Euclidean length or norm of C, so the new criterion vectors have i=l the same unit length and thus units are free and directly comparable. Under the new criterion values, 135

the k-th alternative, Sk, and the ideal and negative ideal solutions S+ and S., are transformed to TSk, TS+ and TS., respectively: TSfc = (tkl,...dkn) = (Xkl/ Ci,..., XkN/] Ci ), k=l,...,k, TS+= (hi... 4n) = M\\Ci\\,...,x^v/ iq4 TS. = (t..., t.n) = (x.,/ Ci,...,x..v/ Cw, * Define the distances of S* and x+ as the weighted Euclidean distance of TSk from TS+: d(st, S,) = W <TS, - TS J = j W. (I, K,f n=1 / C. 2 k= 1,...,K, Where is vector product operator and w is an N-dimensional weight vector whose elements represent the relative importance of the N criteria. Similarly, the distance of Sk from S. is cefined as the weighted Euclidean distance of TSk from TS.: d(,s, SJ = w(tst-ts J = Em,ft. -<- f jikft-x_j\\ C, II)2 n=1 V n=1 k = 1,...,K, * Rank the K alternatives preference order by their relative closeness to the ideal alternative S+, which for the k-th alternative is defined as: r(sk, S+) = d(sk, S+)/[d(Sk, S+) + d(sk, S.)], k = 1,...,K 136

the assessment criterion of TOPSIS is that, the smaller the value of r (Sk,S+) which ranges between 0 and 1, the more preferred is the alternative Sk. 4.9 Choice Of Weights: A reasonably good approach to obtain internal importance weights is to use the entropy concept. It is a criterion for the amount of information (or uncertainity) represented by a discrete probability distribution, pi,... pk and this measure of information was given by Shannon and Weaver (1947) as: k E(p{Pk) = pkln(pk) k=\ where (py=\l\n{k) is a positive constant which guarantees that 0 < E(pi,...,pk) <1. It is noted that the larger the E (pi,...,pk) value, the smaller the variations among the pk s and that 0 entropy means maximum information and 1 minimum information. For the n-th criterion vector Cn = (xjn...xkh) in our MCDM environment, let X = xi +...+ XKn be the total value regarding the criterion. If we view a normalized value Pkn = Xkn/ x for k = 1,...,K as the probability distribution of Cn on the K alternatives, we may similarly define the entropy of Cn as: E(C ) = - 0 k K K pk 1 n(pk) = (pk^ {x:tr l Xn )\b(xkn / Xn), n = 1,...N, an d define the weights as k=1 k=\ Wn = (1- E(Cn))/1;(1 - E(Cj)),n = l,...,n. j=i Taking into account all income categories and the proportional distribution of the Households in the cities listed below, we have followed the Topsis method to rank these cities from the highest to the lowest in the richness scale. 137

Table: 4.37 - Rank & Estimate Of Measure Of Inequality On Income Distribution In 12 Metro Cities - TOPSIS Cities 1985-86 1994-95 To]psis 1995-96 1996-97 1997-98 1998-99 Value Rank Value Rank Value Rank Value Rank Value Rank Value Rank Ahmedabad 0.3342 7 0.3663 6 0.4293 6 0.5283 5 0.6271 4 0.6739 4 Bangalore 0.1778 11 0.2493 9 0.2409 9 0.3012 7 0.4579 7 0.4898 7 Chennai 0.3058 8 0.1686 11 0.1196 11 0.1825 11 0.4322 8 0.4439 8 Delhi 0.5478 3 0.9671 1 0.9444 1 0.9644 1 0.9579 1 0.9537 1 Hyderabad 0.4274 6 0.4004 5 0.4701 5 0.4537 6 0.5015 6 0.5149 6 Jaipur 0.2578 9 0.3097 8 0.3067 8 0.2832 8 0.2740 10 0.2752 10 Kanpur 0.1569 12 0.0840 12 0.0645 12 0.0412 12 0.0139 12 0.0083 12 Kolkata 0.4858 5 0.5291 4 0.5636 4 0.5954 4 0.6164 5 0.6292 5 Lucknow 0.5013 4 0.2044 10 0.2205 10 0.2193 10 0.2194 11 0.2031 11 Mumbai 0.9709 1 0.6749 2 0.7275 2 0.7465 2 0.7945 2 0.8254 2 Nagpur 0.2218 10 0.3340 7 0.3751 7 0.2739 9 0.3628 9 0.3796 9 Pune 0.6481 2 0.6297 3 0.7030 3 0.6807 3 0.7156 3 0.7348 3 Cities 1985-86 (in Percent) 1994-95 (in Percent} 1995-96 (in Per ent) 1996-97 (in Per ent) 1997-98 (in Per cent) 1998-99 (in Percent) Val Rank Rank Rank Rank Rank Valu Rate Ahmedabad 0.150 11 12 12 12 12 0.1882 12 Bangalore 0.209 8 9 7 7 8 0.3768 6 Chennai 0.131 12 4 3 3 6 0.4289 5 Delhi 0.223 5 n 4 4 5 0.3601 8 Hyderabad 0.253 4 n 10 10 10 0.2963 9 Jaipur 0.170 10 10 8 8 7 0.4439 4 Kanpur 0.213 9 i 1 1 2 0.6170 1 Kolkata 0.151 6 6 11 11 11 0.2286 11 Lucknow 0.187 3 8 6 6 4 0.4943 3 Mumbai 0.273 1 7 9 9 9 0.2628 10 Nagpur 0.237 7 3 2 2 1 0.6152 2 Pune 0.276 2 5 5 5 3 0.3751 7 The city of Delhi, which was in the seventh position in 1985-86, achieved the top of the ranking in 1994-95 and continued to remain in the first position till 1998-99. The movement of the households of the different income classes in the hierarchical level was not uniform for all the states. Striking changes in the position of the cities could be seen in the table. It may be observed that the cities like Delhi, Kolkata, Chennai, Ahmedabad and Lucknow upgraded their positions during the period 1986-1999. On the other hand, a decline of 11-4 and 10-5 was witnessed in case of the cities like Nagpur and Bangalore 138