Data on the distribution of income in Malawi

DRAFT EPD INCOEJ DISTRIBTUTION PROJECT DATA ON THE DISTRIBUTION OF INCOME IN MALAWI Jong-goo Park Division Working Paper No. 1980-8 December 1980 Economic and Social Data Division Economic Analysis and Projections Department Development Policy Staff The World Bank Division Working Papers report on work in progress and are circulated for Bank staff use to stimulate discussion and comment. The views and interpretations in a Forking Paper are those of the author and may not be attributed to the World Bank or its affiliated organizations.  DATA ON THE DISTRIBUTION OF INCOME IN MALAWI This paper contains an evaluation of the data on income distribution in Malawi for two complementary household sample surveys conducted in 1968/69. The data from the surveys have been adjusted and combined to derive a more representative distribution by household for the country. Economic and Social Data Division Economic Analysis and Projections Department December 1980  TABLE OF CONTENTS Page I. Introduction ........ ........... ....... * .. . ... 1 II. The 1968 Household Income and FExpenditure Survey for Urban Areas and Agricultural Estates (HIES), . 1 III. The 1968/69 National Sample Survey of Agriculture (NSSA) .................................... 2 IV. Evaluation of the 1968 HIES and the 1968/69 NSSA . 3 V. Data Adjustments and Estimation of Income Distribution . . . . . . . . . . . . . . . . . . . 6 References . . . . . . . . . . . . . . . . . . . . . . 11 Appendix Tables . .......... ................. . ... 12  DATA ON THE DISTRIBUTION OF INCOME IN MALAWI I. Introduction 1. The only sources for nationwide data on income distribution in Malawi are the two complementary household sample surveys conducted in 1968/69: the 1968 Household Income and Expenditure Survey for Urban Areas and Agricultural Estates (HIES) and the 1968/69 National Sample Survey of Agriculture (NSSA). Based on results from these two surveys, Morrisson (1972) derived, in his unpublished monograph, an estimate of income distri- bution by decile "population groups" for 1969. This estimate was published in Jain (1975) and in the Social Indicators Data Sheets of the World Bank. 2. Morrisson (1972) includev non-monetary income which was not covered by NSSA in his estimate of income distribution for smallholder farm households. However, he did not make clear how he derived the distri- bution of non-monetary income by "population group.t' Nor was it clear how he combined the distribution of income for urban and agricultural estate households with that for smallholder farm households. 3. This paper reviews the above two surveys and then derives an estimate of income distribution by household for the whole country for 1967/68. II. The 1968 Household Income and Expenditure Survey for Urban Areas and Agricultural Estates (HIES) 4. The 1968 HIES is the only survey which has covered private house- holds in all urban areas in Malawi. It also covered the agricultural estates  -2- in rural areas. The survey was conducted by the National Statistical Office between January 1968 and February 1969. The main purpose of the HIES was to collect data on household income and expenditures in order to obtain a new weighting scheme to be used in compiling price indices, estimating subsistence consumption in the national accounts and deriving parameters for commodity demand projections. 5. The sample contained 3,250 households in urban areas (6.5% of total private urban households) and 1,250 households (5.6%) at agricultural estates. The households were selected by stratified cluster sampling, using the 1966 census as the sampling frame. First, the whole area was stratified into three subareas: major urban, small urban, and agricultural estate. Clusters of households were then randomly selected from each of the four towns in the major urban areas, from the small urban areas and from the agricultural estates. 6. A household was loosely defined as one or more persons eating together. Household income was defined as the before-tax income of all members of the household, including income both in cash and in kind, trans- fers and production consumed by the producer. Information on income was collected for the previous 12 month period through interviews. III. The 1968/69 National Sample Survey of Agriculture (NSSA) 7. The NSSA was the first agricultural survey carried out on a national scale in the traditional rural areas of Malawi. It was conducted by the National Statistical Office between December 1, 1968 and November 30, 1969. The purpose was to obtain information on cultivated areas, on production and on household income and expenditures as a basis for preparing an agricultural  -3- development plansfor updating the national accounts statistics and for examining patterns of monetary income and expenditures in rural areas. 8. The survey covered all households in the rural areas except those attached to trading centers, estates, missions and government stations. "1 The sample consisted of 5,340 households, 0.6S% of total rural households. The sample households were selected through a multi-stage sampling process using the 1966 population census as the sampling frame. A universe of 4,059 enumeration areas (EAs) was first divided into two strata, one of the 1,153 EAs falling into 35 selected natural areas, and the other, the 2,906 EAs in the remaining 23 districts ("rest of district" area). Second, 410 EAs were chosen from the universe of 4,059 EAs to give at least 5 EAs for each natural area and for each "rest of district" area. Third, one village was randomly selected from each sample EA. Finally, 15 households were selected from each village in the natural area stratum, and 10 households from each village in the "rest of district" stratum. 9. A household was loosely defined as one or more persons eating to- gether. Household income included current before-tax cash income from both farm and non-farm sources. Heads of households were interviewed to obtain information on farm and non-farm income for the previous 12 months. IV. Evaluation of the 1968 HIES and the 1,968/69 NSSA a. Sampling Error and Non-Enumeration Bias (1) The 1968 HIES 10. The relatively large sample size and the stratified sampling /1 The survey covered 97% of the total rural population. The agricultural estates, which are not covered by this survey but are covered by the 1968 HIES, accounted for 2% of the total rural population. The rural households attached to trading centers, missions and government stations accounted for the remaining one percent.  -4- procedure suggest a relatively small sampling error./- Information was collected for 3,890 households out of the total sample of 4,500 which means a relatively high non-enumeration rate of 14 percent. The non-enumeration rate was particularly high (17%) for the agricultural estates. Final results of the survey, however, were adjusted for the differences in the non- enumeration rates of the different sample strata. (2) The 1968/69 NSSA 11. The non-enumeration rate of the 1968/69 NSSA was 5%. Such a low non-enumeration rate is not likely to have introduced significant biases in the survey results. b. Under-Reporting of Income (1) The 1968 RIES 12. As noted before, the reference period for information on income was the past 12 months. This means the recall period for respondents was long, and there is a possibility of response errors. A comparison of total urban household income as derived from the survey with the similar aggregate income from the national accounts shows that the former is 86% of the latter, equivalent to a 14% rate of income under-reporting. /-2 There is also indirect evidence that income was under-reported. In the largest town of Blantyre, even the highest income class (with an average household irLcome of more than K2,000 per year) showed a budget deficit (Appendix, Table A-1), 11 The coefficient of variation for total disposable cash income is 5 percent. /2 Total urban household income as derived from the 1968 HIES was Kwacha (1) 43 maillion, which was compared with the net non-agricultural income of 150 million estimated from national accounts data (see Appendix Table A-7 for details on the estimation of the latter).  -5- (2) The 1968/69 NSSA 13. The report on the 1968/69 NSSA shows that the total current cash revenue for farm households as derived from the NSSA data was 37% lower than that estimated fromr the national accounts data. This, together with the assumption that current farm expenditures were accurately reported in the survey, implies that current farm cash income was under-reported by 44%. A comparison of the survey results with other estimates as shown in the NSSA report indicates that the overall magnitude of reported "other cash incomes" and current cash transfers from abroad was roughly of the right order. Total current cash income including transfers from abroad was, therefore, under- reported by 17% (Table l). Table 1: CURRENT CASH INCOME, SMALLHOLDER FARM HOfUSEHOLDS 1968/69;. (1000 K) NSSA National Accounts A M Estimates (A) Estimates (B) B (1) Farm Cash Receipts 11,182 17,692 63 1. Maize (962) (4,976) 19 2. Groundnuts (2,474) (4,414) 56 3. Other crops (5,466) (6,470) 84 4. Livestock (2,280) (1,832) 124 (2) Farm Cash Expenditures 2,912 2,912 La 100 (3) Farm Cash Income (1-2) 8,270 14,780 56 (4) Other Cash Income 21,506 21,506/a 100 (5) Transfers from Abroad 1,800 1,890 /b 95 (6) TOTAL CASH INCOME (3+4+5) 31,576 382176 83 /a Assumed to be equal to the NSSA estimate. /b Balance. of payments data. Source; Nat4Q9.naL. -,a.I1 X gLAgLi9Jtuei9686 Zomba. Sept. 1970, Tab-lee P and Re.9  c. Exclusion of Non-Cash Income 14. As noted before, the NSSA covered only cash income and expenditures. No information was obtained on such non-cash income as production consumed by farmers, imputed rents from owner-occupied dwellings, and wages and gifts received in kind. National accounts data indicate that non-monetary income in the agricultural sector was three times the monetary income and accounted for 33% of GDP in 1969. 15. The 1968 HIES included non-cash income except that from imputed rents from owner-occupied dwellings. V. Data Adjustments and Estimation of Income Distribution 16. First, we adjusted the survey data on the urban and agricultural estate households and on the rural smallholder households separately. We then combined the adjusted data to derive an estimate of national income distribution for Malawi. a. Urban and Agricultural Estate Households 17. As mentioned earlier, income was under-reported by urban households. In the largest town, Blantyre, which also showed the highest average household income, all income groups reported budget deficits. In other large towns (except for Zomba) and small urban areas, almost all the high-income classes also reported budget deficits (Appendix, Tables A-I through A-5). This suggests that many sample households under-reported their income in the 1968 HIESc In the published survey data, many of the higher income urban households were thus incorrectly classified as lower income households. We assumed that in the. case of those households which reported cash receipts lower than cash dis- bursements, cash disbursements were in fact equal to actual cash receipts./- /1 In the case of very low income groups (with average cash receipts less than K600 per year), we assumed that reported cash receipts could be less than cash diobursements.  -7-- Based on thies, we then reclassified these households into higher income classes (Appendix, Tables A-12 through A-16). 18. Using the data as derived in para 17 above, we estimated distri- bution of income for urban and agricultural estates households for 1967/68 (Appendix, Table A-17). b. Rural Smallholder Households 19. Non-cash income for rural households consists mainly of food consumed in kind within the household. The NSSA data, however, show only cash expenditures for food for rural households. On the other hand, the HIES data sho; consumption of food -- both in cash and in kind - for urban and agricultural estate households. 20. We derived rural smallholder household consumption of food in kind as follows. First, assuming that per capita food consumption is the same for rural smallhoider and agricultural estate households, we estimated total food consumption (both in cash and in kind) for rural smallholder households from the HIES data on food consumption of agricultural estate households. We then subtracted food consumption in cash from the estimated total food consumption, and estimated food consumption in kind for the different income groups of rural smallholders (Appendix, Tables A-8). 21. Regarding the under-reported agricultural cash income, we assumed that the under-reporting was the result of respondents not being able to recall intermittent sales of small quantities throughout the year. This led us to assume, further, that the value of agricultural production per unit of land, which does not vary with farm size, was equal to that estimated for the smallholder group with the highest farm cash receipts per unit of land.  -8- Given this assumption, we adjusted upward the reported average farm income of smallholder households whose agricultural production per unit of land was under-estimated (Appendix, Table A-9). 22. We then totaled the food consumption in kind as derived in para. 20 the adjusted farm income as explained in para. 21 and the other cash income as reported in the NSSA (para. 13) ,and derived the adjusted average household income for rural smallholder households (Appendix, Table A-10). The total household income derived for rural smallholders from the adjusted data (K109 million) was very close to the aggregate smailholder income estimated from the national accounts data (K112 million). Using the adjusted data, the distribution of income was derived for the rural smallholder households for 1968/69 (Appendix, Table A-11). c. National Income Distribution 23. Deriving the distribution of income for the whole country required consolidation of the two sets of data for rural smallholders and urban and agricultural estate households. As noted earlier, however, the reference period for income figures in the NSSA was 1968/69, while that in the HIES, was 1967/68. A more serious problem was that the published survey data for rural smallholders ranked distribution of households according to the total cash income for rural smallholders, whereas urban and plantation households were ranked according to total cash receipts. 24. Using the survey data, we first derived current average household income by income groups of rural smallholders for 1968/69 and urban and agricultural estate households for 1967/68 respectively. Then we adjusted the 1968/69 average income of rural smallholders uniformly downward to its equivalent 1967/68 value, based on the change in the nominal value-added in  -9- the smallholder agricultural sector between 1967/68 and 1968/69. Finally, using the level of average household incomes and income ranges used to classify households, we combined the distributions of income for rural small- holders and urban and agricultural estate households. The result is a rudimen- tary estimate of Malawi's income distribution for 1967/68 (Appendix Table A-18). From this, we derived income distribution by decile household groups (Table 2). 1 Table 2: HOUSEHOLD INCOM4E DISTRIBUTION, ADJUSTED 1967/68 Cumulative % of Household Income/a Cumulative % Urban and Rural of Households Plantation Smallholders National 10 1.6 6.6 5.0 20 3.2 14.0 10.4 30 4.9 21.4 15.9 40 7.1 28.8 21.5 50 10.0 37.3 27.9 60 13.9 46.1 34.6 70 19.2 55.3 41.5 80 27.3 65.2 49.4 90 43e3 77.6 59.9 (95) (67.9) (85.6) (67.4) 100 100.0 100.0 100.0 Gini Coefficient 0.648 0.201 0.389 ha Derived from Appendix Tables A-1l, A-17 and A-18 by linear -interpolation between two adjacent classes. I -n Appendix Table A-18, the second income group includes about 30% of the households, The estimates of the income share for the second, third and fQurtt deciQle goups may not be yery accurate hecause of linear interpolation,  -10- 25. The data in Table 2 indicates a fairly equal income distribution among smallholder farm households and a highly unequal income distribution among urban and agricultural estate households. Income distribution is relatively equal for the country as a whole, in general following the income distribution pattern of the predominant, traditional agrictiltural sector. However, the data on which these income distribution estimates are based may be subject to a large margin of error. l Nevertheless, the subsistence farm- ing sector does generate about half of the total household income in Malawi, and the income from subsistence farming is normally expected to be distributed relatively equally. Therefore, income distribution among smallholder farm households is considered fairly equal,and this leads to the relatively equal distribution of income for the whole country. 26. The data in Table 2 are summarized in Table 3, which also includes the income shares of individual household members currently shown in the World Bank's Social Indicators Data Sheet for Malawi. However, these two sets of income distribution statistics (one for household groups and the other for individual hous6hold members) are not comparable. Table 3: INCOME INEQUALITY MEASURE FOR 1967/68 AND 1969 -- Income Shares % Gini Lowest Lowest Highest Highest Coeffi- 20% of 40% of 20% of 5% of Data Source cient Households Households Households Households Adjusted IET'S 0.38-9 10.4 21.5 50.6 32.6 + NSSA Data for 1967/68 Morrisson's 0.459 5.7 15.0 52.9 29.5 data for 1969 (Social Indi- cators Sheets) /1 Also, the household income does not include the imputed rents from owner- occupied dwellings, which accounted for a little more than 1% of the estimated total household income in 1967/683  -11- REFERENCES Jain, S., Size Distribution of Income: A dompilation of Data, World Bank, Washington, D.C., 1975. Mehran, F., "Dealing with Grouped Income Distribution Data," World Employment Programme Research Working Papers, WEP 2-23/WP/20, ILO, Geneva, August 1975. Morrisson., C.,, "Malawi 1969," (mimeo), Washington, D.C., 1972. National Statistical Office, Household Income Expenditure Survey for Urban Areas and Agriculttnral Estates 1968, Zomba, May 1970. _, National Sample Survey of Agriculture 1968/69, Zomba, September 1970. , National Accounts Report 1964-1970, Zomba, November 1972. World Bank, Memorandum on the Economy of Malawi, Report No.1677a-MAI, Washington, D.C., September 30, 1977.  -12- APPENDIX TABLES Table A-1: AVERAGE HOUSEHOLD INCOME AND EXPENDITURES, BLANTYRE, 1967/68 Income No. of Current /a Current C Class Households Income (Y) Expenditures (C) Y (Kwacha) (Kwacha) 0 - 100 4,818 66.6 75.4 1.13 100,1- 140 3,168 136.2 160.6 1.18 140.1- 200 3,894 181.6 202.2 1.11 200.1- 280 4,884 248.4 293.0 1.18 280.1- 400 3,300 322.4 333.8 1.04 400.1- 600 2,838 444.0 512.4 1.15 600.1-1,000 1,980 645.6 662.2 1.03 1,000.1-2,000 1,254 1,054.4 1,348.8 1.28 2,000.1+ 4,356 4,023.8 4,125.0 1.03 Total 30,492 824.0 872.9 1.06 /a After-tax income including both cash and non-cash income and net current transfer receipts. Source: fHousehold Income and Expenditure Survey for Urban Areas and Agricultural Estates, 1968, National Statistical Office, Zomba, May 1970.  -13- Table A-2: AVERAGE HOUSEHOLD INCOME AND EXPENDITURES, ZOMBA, 1967/68 Income 0 Current Current C Class Households Income CY) Expenditures (C) Y (iKwacha) (Kwacha) 0 - 100 725 80.8 89.6 1.11 I Ci` 1- 140 399 139.0 149.2 1. 07 140.1- 200 433 193.6 189.6 0.98 200.1- 280 610 263.0 258.6 0.98 280.1- 400 505 340.8 325.8 0.96 400.1- 600 878 458.8 467.6 1.02 600.1-1,000 561 645.6 618.0 0.96 1,000.1-2,000 353 1,037.8 1,107.6 1e07 2,000+ 644 3,913.2 3,545.0 0.91 Zomba 5,108 818.6 775.3 0.95 Source: Household Income and Expenditure Survey 1968, op. cit.  -14- -Table A-3: AVERAGE HOUSEHOLD INCONE AND EXPENDITURES, LILONGWE, 1967/68 Income No. of Current Current C Class Households Income (Y) Expenditures (C) Y (Kwacha) (Kwacha) 0- - 100 507 60.6 69.8 1.15 100,1- 140 425 125.6 123.8 0.99- 140.1- 200 555 159.8 175.8 1.10 200.1 280 484 216.2 257.0 1.19 280.1- 400 625 296.2 346.8 1.17 40Q.1- 60G. 1,049. 386.0 488.0 1.26 600.1-1,000 578 624.6 683.2 1.09 1,000.1-2,000 355 1,118.2 1,360.6 1.22 2,000+ 472 3,304.4 3,227.8 0.98 Lilongwe 5,050 603.8 681.2 1.08 Source: Household Income and Expenditure Survey 1968, op. cit  -15- Table A-4: AVERAGE HOUSEHOLD INCOME AND EXPENDITURE, MZUZU, 1967/68 Income No. of Current Current C Class Households Income (Y) Expenditures (C) Y (Kwacha) CKwacha) 0- 100 132 113.8 112.2 0.99 100.1- 140 188 141.6 135.0 0.95 140.1- 200 234 172.0 188.0 1.09 200.1- 280 291 259.2 264.4 1.02 280.1- 400 238 321.4 123.4 1.01 400.1- 600 301 416.6 461.4 1.11 600.1-1,000 229 577.2 660.4 1.14 1,000.1-2,000 154 952.4 1,227.4 1.29 2000.1+ 133 3,653.0 3,446.2 0.94 Mzuzu 1,900 591.5 618.7 1.05 Source: Household Income and Expenditure Survey 1968, op.. cit.  -16- Table A-5: AVERAGE HOUSEHOLD INCOME AND EXPENDITURES, SMALL URBAN AREAS, 1967/68 Income No. of Current Current C Class Households Income (Y) Expenditures (C) Y (Kwacha) (Kwacha) 0 - 100 620 97.2 113.2 1.16 100.1- 140 650 143.6 149.4 1.04 140.1- 200 1,450 185.2 191.0 1.03 200.1- 280 1,4e80 234.2 245.8 1.05 280.1- 400 1,530 328.4 331.0 1.01 400.1- 600 2,260 443.2 443.6 1.00 600.1-1,000 1,660 693.6 634.0 0.91 1,000.1-2,000 1,040 1,027.8 1,176.4 1.14 2,000.1+ 570 2,918.2 3,399.4 1.16 Small Urban Areas 10,960 555.3 589.1 1.06 Source: Household Income and Expenditure Survey 1968, op. cit.  -17- Table A-6: AVERAGE HOUSEHOLD INCOME AND EXPENDITURE, AGRICULTURAL ESTATES, 1967/68 Income No. of Current Current C Class Households Income (Y) Expenditures (C) Y (Kwacha) (Kwacha) 0 - 100 17,469 101.6 101.0 0.99 100.1- 140 5,184 132.0 135.0 1.02 140.1- 200 2,970 171.6 169.0 0.98 200.1- 280 1,188 232.4 226.8 0.98 280.1- 400 648 315.2 291.6 0.93 400.1- 600 351 391.4- 379.4 0.97 600.1-1,000 162 640.4 560.0 0.87 1,000.1-2,000 27 908.0 893.4 0.98 2,000.1+ Estates 27,999 132.8 131.2 0.99 Source: Household Income and Expenditure Survey 1968, op. cit.  -18- Table A-7: ESTIMATION OF NON-AGRICULTURAL HOUSEHOLD/a INCOME, 1968 Private consumption expenditures (in cash) 92.7 plus national savings -0.1 plus net transfer from abroad 1.8 minus corporate savings b 6.9 minus monetary agricultural income (after tax) - 20,0 minus non-agricultural income of agricultural houiseholds d 18.0 equals non-agricultural household/a income 49.5 /a Includes plantation households. /b Assumed to be equal to total domestic savings. /c Excludes plantation households; direct taxes of agricultural households were estimated from data from the NSSA. /d Derived from the NSSA and adjusted for the growth of non-agricultural income from 1968 to 1969. Source: Malawi National Statistical Office.  Table A-8: ESTIMATION OF FOOD CONSUMPTION IN KIND, AVERAGE SMALLHOLDER HOUSEHOLD, 1968/69 (Kwacha) Current Food Household Cash Food Cash Income Household Consumptio7 Food Expenditures Consumption Class (K) Size Per Capita a Consumption for Food in Kind (1) (2) (3=lx2) (4) (5=3-4) 0 - 4.0 20.0 80.0 5.5 74.5 0.01- 10.0 4.2 20.0 84.0 4.6 79.4 10.01- 20.0 4.8 20.0 96.0 9.6 86.4 20.01- 30.0 4.6 20.0 92.0 12.7 79.3 30.01- 40.0 4.9 20.0 98.0 16.5 81.5 40.01- 50.0 5.2 20.0 104.0 21.5 82.5 50.01- 60.0 5.2 22.5 117.0 17.3 9.9.7 60.01- 80.0 5.0 25.0 125.0 27.6 97.4 80.01-120.0 5.3 27.5 145.7 36.0 109.7 120.01-200.0 5.3 30.0 159.0 46,5 112.5 200.0+ 5.5 35.0 192.6 98.5 94,1 /a Estimated from the HIES data on food consumption by agricultural plantation households. per capita food consumption of K20 per year is about half of the estimated poverty line income of K41 per capita. Source: National Sample Survey of Agriculture, 1968/69, Zomba, September 1970, Table R6.13.  d K -20- Table A-9: ADJUSTMENTS FOR UNDER-REPQRTING OF INCOME, "AVEMAGE:-SALLHQLDER HOUSEHOLD, -1968/69 (Kwacha) Average Value of Farm /a Current Cas'A Holding Current Farm Food Consumed Production Farm Income- Income Class (K) Size (Acres) Cash Receipts in Kind Per Acreage Adjustments (1) (2) (3) (2+3)t(l) 0 2.6 0.l; 74.5 29.0 - 0.01- 10,Q 3.2 2.9 79.4 25.7 4.3 10.01- 20.0 3.9 8.7 86.4 24.4 5.6. 20.01- 30.0 4.5 14.7 79.3 20.8 9.2 30.01- 40.0 4.9 20.2 81.5 20.8 9.2 40.01- 50.0 4.2 20.0 82.5 24.4 5.6 50.01- 60.0 4.3 25.1 99.7 29.0 1.0 60.01- 80,0 4.3 27.7 97.4 29.1 0.9 80.01-120.0 4.5 37.0 109.7 32.6 - 120.01-200.0 5.2 53.5 112.5 31.9 200.0+ 4.8 48.9 94.1 29.8 /a Farm income adjustments were made such that the value of farm production per acre amounts to K30.  -21- Table A-10: AVERAGE HOUSEHOLD INCOME, ADJUSTED, RURAL SMALLHOLDERS, 1968/69 Current Cash Income Groups (Kwacha) 0 0.01- 10.01- 20.01- 30.01- 40.01- 50.01- 60.01- 80.01- 120.01- 200+ 10.0 20.0 30.0 40.0 50.0 60.0 80.0 120.0 200.0 1. Current farm cash income -1.7 2.3 7.0 12.1 17.1 16.0 19.0 20.8 27.2 40.8 22.0 2. Other current cash income 0.4 2.6 8.6 13.4 19.8 29.0 34.7 48.7 71.8 111.8 362.6 3. Cash transfer receipts 8.4 3.2 3.8 2.8 2,7 2.2 7.8 1.6 2.1 4.4 1.5 4. Cash transfer payments 0.1 0,1 0.2 0.4 0.9 0,8 2.2 0.9 3.0 2.3 8.8 5. Direct taxes 0.7 0.9 1.9 2.6 2.8 3.0 3.3 3.9 3.7 4.9 9.7 6. Food eonsumed in kind/a 74.5 79.4 86.4 79.3 81e5 82.5 99.7 97,4 109.7 112.5 94.1 7. Farm income/b adjustments - 4.3 5,6 9.2 9,2 5.6 1.0 0.9 - - - Net current income 80.8 90.8 109.3 113.8 126.6 131.4 156.7 164.6 204.1 262.2 461.7 (1+2+3-4-5+6+7) /a Taken from Table A-8. /b Taken from Table A-9. Source: National Sample Survey of Agriculture, 1968/69, National Statistical Office, Zomba, September 1970, Tables R6.12 and R6.13.  -22- Table A-Il: DISTRIBUTION OF HOUSEHOLD INCOME, ADJUSTED,/a RURAL SMALLHOLDERS, 1968/69 Ave. Household Households Household Income Income No. % Amount % (Kwacha) Kwacha ('000). 80.9 82,000 9.3 6,634 6.1 90.8 295,000 33.3 26,786 24.6 109.3 171,000 1923 18,690 17.1 113.8 94,000 10.6 10,697 9.8 126.6 58,000 6.6 7,343 6.7 131.4 35,000 4.0 4,599 4.2 156.7 26,000 2.9 4,074 3.7 164.6 46,000 5.2 7,572 7.0 204.1 31,000 3.5 6,327 5.8 262.2 27,000 3.0 7,079. 6,5 461.7 20,000 2.3 9,234 8.5 TOTAL 885,000 100.0 109,035 100.0 /a Household income includes both monetary and non-monetary income and net transfer receipts, after payment of direct taxes.  -23- Table A-12: DISTRIBUTION OF HOUSI1OLDS BY CASH RECEIPTS AND CASH DIS3URSEMENTS GROUPS, ADJUSTED,/a BLANTYRE, 1967/68 Cash Cash Receipts Group (K) Disbursements 0 100.1 140.1 200.1 280.1 400.1 600.1 1000.1 2000+ Group (K) -100 - 140 - 200 - 280 - 400 - 600 -1000 - 2000 2000+ 0 - 100 4,356 264 132 132 100.1- 140 330 1,980 396 132 140.1- 200 1'32 528 2,244 330 198 200.1- 280 264 924 2,508 528 66 280.1- 400 132 1,188 2,112 594 400.1- 600 462 330 1, .88 264 600.1-1,000 2,706 66 66 1,000.1-2,000 1,386 66 2,000+ 4,488 Blantyre 4,818 3,036 3,828 4,752 3,168 1,848 2,970 1,452 4,620 Average Household Income (K) /b 97.8 143.8 198.8 267.6 369.6 522.9 828.1 1532.4 4,124.9 /a Derived from Table 13 on p. 33, Household Income and Expenditure Survey for Urban Areas and Agricultural Estates, 1968. /b Estimated by multiplying median cash receipts by the ratio of reported current income to reported cash receipts (reported consumption expenditure for the highest cash receipts group).  -24- Table A-13: DISTRIBUTION QEO HOUSEHOLDS BY CASH RECEIPTS AND CASH DISBURSEMENT GROUPS, ADJUSTED,/a LILONGWE, 1967/68 Cash Receipts Group (K) Cash Disbursements 0 100.1 140.1 200.1 280.1 400.1 600.1 1000.1 Group (K) -100 - 140 - 200 - 280 - 400 600 -1000 - 2000 2000+ 0 - 100 447 35 100.1- 140 24 331 35 140.1- 200 24 47 390 24 12 12 200.1- 280 12 106 389 24 12 280.1- 400 24 12 494 35 400.1- 600 59 71 765 47 600.1-1,000 709 12 1,000.1-2,000 355 24 2,000+ 519 Lilongwe (K) 495 425 555 484 601 824 756 367 543 Average Household Income (K) /b 71.8 11,41.| 182.2 260.8 355.3 513,0 821.4 1527.3 3304.4 /a Derived from Table 13 on p. 76, Household Income and Expenditure Survey, 1968. /b Estimated by multiplying median cash receipts by the ratio of reported current income to reported cash receipts (reported current income for the highest cash receipts group),  -25- Table A-14: DISTRIBUTION OF HOUSEHOLDS BY CASH RECEIPTS AND CASH DISBURSEMENTS GROUPS, ADJUSTED, /a ZOMBA, 1967/68 Cash Cash Receipts Group (K) Disbursements 0 100,1 140,1 200.1 280.1 400.1 600.1 1000.1 Group (K) -100 - 140 - 200 - 280 - 400 - 600 -1000 - 2000 2000+ 0 - 100 631 24 100.1- 140 82 340 23 140.1- 200 12 23 351 12 200.1- 280 59 503 59 280.1- 400 12 95 363 59 400.1- 600 83 679 35 12 600.1-1,000 619 24 1,000.1-2,000 341 12 2,000+ 655 Zomba 725 387 445 610 505 738 654 377 667 Average Household Income(K) /b 80.8 155.1 217.7 299.0 387.0 547.4 840.0 1560.1 3913.2 /a Derived from Table 13, p. 119, Household Income and Expenditure Survey, 1968. /b Estimated by multiplying median cash receipts by the ratio of reported current income to reported cash receipts (reported current income for the highest cash receipts group).  -26- Table A-1S: DISTRIBUTION OF HOUSEHOLDS BY CASH RSCEIPTS AND CASH DISBURSEMENTS GROUPS, ADJUSTED;' aJXZUZU,.1967/68 Cash Cash Receipts Group (K) Disbursements 0 100.1 140.1 1 200.1 280.1 400.1 600.1 1000.1 Group (K) -100 - 140 - 200 - 280 - 400 - 600 -1000 - 2000 2000+ 0 - 100 114 13 100.1- 140 18 109 27 140.1- 200 66 145 35 200.1- 280 53 199 35 280.1- 400 9 53 163 22 400.1- 600 4 40 226 13 600.1-1,000 243 1,000.1-2,000 1158 2,000+ 155 Mzuzu 132 188 234 291 238 248 256 1L58 155 Average Household Income (K) /b 113.8 138.8 200.3 289.4 379.3 522.6 820.3 1544.8 3653.0 /a Derived from Table 13, p.161 Household Income and Expenditure Survey, 1968. /b Estimated by multiplying median cash receipts by the ratio of reported current income to reported cash receipts (reported current income for the highest cash receipts group).  -27- Table A-16: DISTRIBUTION OF HOUSEHOLDS BY CASH RECEIPTS AND CASH DISBURSEMENTS GROUPS, ADJUSTED,/a SMALL URBAN AREAS, 1967/68 Cash Receipts Group (K) Cash-r-.. C 0 100.1 140.1 200.1 280.1 400.1 600.1 1000.1 Disbursements -100 - 140 - 200 - 280 - 400 - 600 -1000 - 2000 2000+ 0 - 100 480 100.1- 140 120 520 20 140.1- 200 10 130 1,250 40 200.1- 280 10 170 950 60 280.1- 400 10 180 1,350 50 400.1- 600 10 150 2030 60 600.1-1,000 1,730 60 1,000.1-2,000 980 10 2,000+ 610 Small Urban Areas 610 660 1,450 1,180 1,560 2,080 1,790 1,040 620 Average Household Income (K) /b 97.2 158.4 206.1 274.7 374.2 529.8 829.9 1,531.8 3,399.4 /a Derived from Table 13, p. 206, Household Income and Expendituie Survey, 1968. /b Estimated by multiplying median cash receipts by the ratio of reported current income to reported cash receipts (reported consumption expenditure for the highest cash receipts group).  -28- Table A-17: DISTRIBUTION OF HOUSEHOLD INCO1E, URBAN AND PLANTATION, ADJUSTED,/a 1967/68 Ave. Househ7;d Households Household Inlcome Income (K) - No. % Amount (1000K) 99.8 24,200 29.7 2,414.5 4.8 138.6 9,900 12.1 1,371.7 2.7 191.0 9,500 11.7 1,814.4 3.6 266.4 8,500 10.4 2,264.6 4.5 366.9 6,700 8.2 2,458.1 4.9 518.4 6,100 7.5 3,162.0 6.3 823.3 6,600 8.1 5,434.0 10.8 1,539.8 3,400 4.2 5,235.4 10.4 3,959.9 6,6Q0 8.1 26,135.2 52.0 TOTAL 81,5QO 100.0 50,289.9 100.0 /a Derived from Tables A-6 and A-12 - A-16. /b Derived from Tables A-6 and A-12-,-A-16 by taking weighted averages for d.ifferent cash receipts groups.  -29- Table A-18: DFERIVATION OF HOUSEHOLD INCONZ DISTRIBUTION FOR MALAVWI, ADJUSTED, 1967/68 - Agr. Urban and Agricultural Estate Small- National No. of No. of holder No. of Average House- Income House House- Households Averag Household Income Income lholds Range ' holdsc - holds ('Q00) Incomet- Inccme (K) . Range (K)a ('000) Subdivided'- ('000) ('000) (5+6) (K) ('00) (7x8) (1) (2) (3) (4) (5) (6) (7) (8) , (9) * (80.4- 84.2 2.4 82.0 84.4 79.3 6,693 ,f ( 84.2- 98.1 8.7 295.0 303.3 89.0 27,029 99.8 80.4 -119.2 4.2 ( 98.1-109e4 7.0 171.0 178.0 107.1 19,064 (109.4-117.8 5.2 94.0 99.2 111.5 11,061 ; (117.8-119.2 0.9 58.0 60.5 124 X 1 7,508 1 119.2-126.4 1.6). 138.6 119.2-164.8 *9.9 C 126.4-141.3 3.2 35.0 38.2 128. 8 4, 920 X 141.3-157.5 3.5 26.0 29.5 153.6 4,531 C 157.5-164.8 1.6) 46.0 50.0 161.3 8,065 (164.8-180.8 2.4) 191.0 164.8-228.7 9.5 (180.8-228.7 7.1 31.0 38.1 200.1 7,624 266.4 228.7-316.7 8.5 228.7-316.7 8.5) 27.0 37.4 257.e0 9,612 366.9 316.7-442.7 6.7 (316.7-354.8 1.9) (354.8-442 7 4.8 ) 20.0 29.7 452.6. 13,442 518.4 442.7-670.9 6.1 ( 442.7-624.8/R 4.9 ) ( 624.8-670.9 1.2 - 1.2 647.0 L 776 823.3 6.6 6.6 - 6.6 823.3 5,434 1,539,8 3c4 3.4 - 3.4 1,539.8 5,235 3,959.9 6.6 6.6 - 6.6 3,959.9 26,135 TOTAL 81.5 81.5 885.0 966.5 157,129 /a The limilts of each income range were assumed to be the midpoints between two adjacent average incomes shlomn in column (1). /-b Income rl-nges it column (2) were suibdivided so as to conform to those for smallholder farm households. The latter-.income ranges were derived in a manner similar to that mentioned in footnote /a. /c It was- assumed that the households shown in column (3) were uuliformly distributed over the respective income ranges shown in colunmn (2), /d Taken fronm Table A-l1. /e Average incomes for the first 11 classes are those for smallholder farm households (the 1968Y69 average incomes shlon in Table A-li were adjusted for a change in nominal value-added in smallholder agricultural sector between 1967/68 and 19.68/69). Average incomes for the last 3 classes are for urban and agricultural, estate households. /f Estimated such that the arithmetic average of 80.4 and ]I2Z becomes 99,8 (See Mehran, 1975). /S Estimated such that the harmonic average of 354,8 and 624.8 becomes 452.6, /h The harmonic average of 624.8 ani 6.70.9.